Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Interventions for promoting physical activity in people with neuromuscular disease

Authors' declarations of interest

Version published: 24 May 2021 Version history

https://doi.org/10.1002/14651858.CD013544.pub2
Collapse all Expand all

Background

The World Health Organization (WHO) recommends that people of all ages take regular and adequate physical activity. If unable to meet the recommendations due to health conditions, international guidance advises being as physically active as possible. Evidence from community interventions of physical activity indicate that people living with medical conditions are sometimes excluded from participation in studies. In this review, we considered the effects of activity‐promoting interventions on physical activity and well‐being in studies, as well as any adverse events experienced by participants living with inherited or acquired neuromuscular diseases (NMDs). 

Objectives

To assess the effects of interventions designed to promote physical activity in people with NMD compared with no intervention or alternative interventions.

Search methods

On 30 April 2020, we searched Cochrane Neuromuscular Specialised Register, CENTRAL, Embase, MEDLINE, and ClinicalTrials.Gov. WHO ICTRP was not accessible at the time.

Selection criteria

We considered randomised or quasi‐randomised trials, including cross‐over trials, of interventions designed to promote physical activity in people with NMD compared to no intervention or alternative interventions. We specifically included studies that reported physical activity as an outcome measure. Our main focus was studies in which promoting physical activity was a stated aim but we also included studies in which physical activity was assessed as a secondary or exploratory outcome.

Data collection and analysis

We used standard Cochrane procedures.

Main results

The review included 13 studies (795 randomised participants from 12 studies; number of participants unclear in one study) of different interventions to promote physical activity. Most studies randomised a minority of invited participants. No study involved children or adolescents and nine studies reported minimal entry criteria for walking. Participants had one of nine inherited or acquired NMDs.

Types of intervention included structured physical activity support, exercise support (as a specific form of physical activity), and behaviour change support that included physical activity or exercise. Only one included study clearly reported that the aim of intervention was to increase physical activity. Other studies reported or planned to analyse the effects of intervention on physical activity as a secondary or exploratory outcome measure. Six studies did not report results for physical activity outcomes, or the data were not usable. We judged 10 of the 13 included studies at high or unclear risk of bias from incomplete physical activity outcome reporting.

We did not perform a meta‐analysis for any comparison because of differences in interventions and in usual care. We also found considerable variation in how studies reported physical activity as an outcome measure. The studies that reported physical activity measurement did not always clearly report intention‐to‐treat (ITT) analysis or whether final assessments occurred during or after intervention. Based on prespecified measures, we included three comparisons in our summary of findings.

A physical activity programme (weight‐bearing) compared to no physical activity programme

One study involved adults with diabetic peripheral neuropathy (DPN) and reported weekly duration of walking during and at the end of a one‐year intervention using a StepWatch ankle accelerometer. Based on the point estimate and low‐certainty evidence, intervention may have led to an important increase in physical activity per week; however, the 95% confidence interval (CI) included the possibility of no difference or an effect in either direction at three months (mean difference (MD) 34 minutes per week, 95% CI –92.19 to 160.19; 69 participants), six months (MD 68 minutes per week, 95% CI –55.35 to 191.35; 74 participants), and 12 months (MD 49 minutes per week, 95% CI –75.73 to 173.73; 70 participants). Study‐reported effect estimates for foot lesions and full‐thickness ulcers also included the possibility of no difference, a higher, or lower risk with intervention.

A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme

One study involved adults with DPN and reported duration of walking over 48 hours at the end of four weeks' intervention using a t‐shirt embedded PAMSys sensor. It was not possible to draw conclusions about the effectiveness of the intervention from the very low‐certainty evidence (MD –0.64 hours per 48 hours, 95% CI –2.42 to 1.13; 25 participants). We were also unable to draw conclusions about impact on the Physical Component Score (PCS) for quality of life (MD 0.24 points, 95% CI –5.98 to 6.46; 35 participants; very low‐certainty evidence), although intervention may have made little or no difference to the Mental Component Score (MCS) for quality of life (MD 5.10 points, 95% CI –0.58 to 10.78; 35 participants; low‐certainty evidence).

A functional exercise programme compared to a stretching exercise programme

One study involved adults with spinal and bulbar muscular atrophy and reported a daily physical activity count at the end of 12 weeks' intervention using an Actical accelerometer. It was not possible to draw conclusions about the effectiveness of either intervention (requiring compliance) due to low‐certainty evidence and unconfirmed measurement units (MD –8701, 95% CI –38,293.30 to 20,891.30; 43 participants). Functional exercise may have made little or no difference to quality of life compared to stretching (PCS: MD –1.10 points, 95% CI –5.22 to 3.02; MCS: MD –1.10 points, 95% CI –6.79 to 4.59; 49 participants; low‐certainty evidence).

Although studies reported adverse events incompletely, we found no evidence of supported activity increasing the risk of serious adverse events.

Authors' conclusions

We found a lack of evidence relating to children, adolescents, and non‐ambulant people of any age. Many people living with NMD did not meet randomised controlled trial eligibility criteria. There was variation in the components of supported activity intervention and usual care, such as physical therapy provision. We identified variation among studies in how physical activity was monitored, analysed, and reported. We remain uncertain of the effectiveness of promotional intervention for physical activity and its impact on quality of life and adverse events. More information is needed on the ITT population, as well as more complete reporting of outcomes. While there may be no single objective measure of physical activity, the study of qualitative and dichotomous change in self‐reported overall physical activity might offer a pragmatic approach to capturing important change at an individual and population level.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

What interventions promote physical activity in people living with neuromuscular disease?

Review question

We looked at the evidence on ways to increase physical activity in people with neuromuscular disease. We only included studies that measured physical activity as an outcome. We collected and analysed all relevant studies to answer this question.

Background

Physical activity includes any type of everyday activity, such as work, travel, or exercise. Based on well‐known health benefits, international guidance recommends that everyone does enough regular physical activity. This guidance is relevant to people with health conditions, including neuromuscular diseases, which affect the way that muscles and nerves work. However, many people with neuromuscular disease live with disability and face barriers to being more physically active. Research studies also sometimes exclude people who have health conditions.

Study characteristics

This review included 13 studies (795 participants from 12 studies; the number of participants was unclear in one study). The studies mostly involved adults who were able to walk, although our review question included people of any age who were able to move around with or without assistance. Only one study intervention had a stated aim to increase physical activity. The main focus of most studies was to assess the effects of physical activity interventions on other aspects of health, fitness, and well‐being. Interventions involved supported exercise or other types of physical activity, advice about being more active, or talking therapies that included changing everyday activity behaviour. Physical therapists and other health and fitness professionals provided the support for most interventions.

Results and certainty of the evidence

The evidence related to adults with nine different types of neuromuscular disease. All 13 studies planned to measure physical activity but the results from six studies were not reported or usable. We found important differences in the interventions and in the usual care provided in different places. We also found differences in how studies measured and reported outcomes. Missing information on physical activity may have affected the results. In terms of time spent physically active, we are uncertain whether any intervention promoted physical activity in people with neuromuscular disease. We are also uncertain about the impact of physical activity interventions on quality of life and any harms. However, we found no evidence that physical activity intervention increased serious harms. Our uncertainty in the evidence is due to limitations in study design and because the results were imprecise or did not directly address our review question.

The evidence is current to 30 April 2020.

Authors' conclusions

Implications for practice

Physical inactivity is associated with an increased risk of many major noncommunicable diseases in the general population. However, we remain unclear about the effectiveness of physical activity‐promoting intervention specifically in people with neuromuscular disease (NMD). This review also highlights important variation in usual care, such as physical therapy provision, for people living with NMD. Most studies reported adverse events incompletely. However, we found no evidence of supported activity increasing the risk of serious adverse events or adverse events leading to study discontinuation.

Implications for research

The findings from our review and other reviews highlight particular challenges in addressing potentially conflicting intentions of inclusiveness and of minimising heterogeneity in the evaluation of interventions designed to promote physical activity; the inevitable question being 'whose' physical activity is being promoted? While larger randomised controlled trials (RCTs) might be one approach to tackling these challenges, our findings also raise questions about the way physical activity is measured, reported, and analysed, irrespective of sample size. The inclusion of people with NMD in broader public health interventions could potentially support more inclusive trials. However, there is also a further question of whether or not interventions that 'promote' physical activity are synonymous with interventions that motivate individual choice to participate in physical activity, and whether this matters to study participants and non‐participants. For example, some people living with NMD may find they do not have an option to be more active outside of a research setting if they do not have access to physical, psychological, social, or financial support. In terms of non‐health‐related incentives for participating in research, one included study paid a cash incentive for participation and several studies reported payment of travel expenses and gym membership. This financial cost might need to be weighed up against potential costs associated with the treatment of chronic diseases and complications attributed in part to physical inactivity. However, it is unclear to what extent intervention affected financial as opposed to physical, psychological, or social barriers to participants' physical activity and well‐being. Qualitative research could help to further explore these potential barriers to physical activity within participants who complete research interventions and those non‐participants lost at recruitment or during intervention.

Across different study populations, we anticipate that further technological changes in the way everyday physical activity is monitored through devices could play an increasing role in future studies of interventions for promoting physical activity. In this review, we found a lack of evidence relating to children, adolescents, and non‐ambulant people of any age. While there may be no single objective measure of physical activity, the study of qualitative and dichotomous change in self‐reported overall physical activity might offer a pragmatic approach to capturing important changes at an individual and population level. For evaluating promotional strategies, more complete reporting of outcomes is needed, including analysis of the intention‐to‐treat effect. Assessment of the potential for reducing all‐cause mortality risk, complications associated with NMD, and the primary and secondary prevention of other chronic diseases could also help us to better understand the value of interventions that promote physical activity in people living with NMD.

Summary of findings

Open in table viewer
Summary of findings 1. A physical activity programme (weight‐bearing) compared to no physical activity programme in people living with NMD

Physical activity programme compared to no physical activity programme

Patient or population: people with NMD

Setting: primary care, endocrinology, or podiatry practices in central Missouri, USA

Intervention: physical activity programme (weight‐bearing)

Comparison: no physical activity programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no physical activity programme

Risk with physical activity programme

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, during intervention

Follow‐up: 3 months

The mean time spent walking was 526 minutes per week

MD 34 minutes more
(92.19 fewer to 160.19 more)

—

69
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, during intervention

Follow‐up: 6 months

The mean time spent walking was 511 minutes per week

MD 68 minutes more
(55.35 fewer to 191.35 more)

—

74
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, unclear if during or after intervention

Follow‐up: 12 months

The mean time spent walking was 500 minutes per week

MD 49 minutes more
(75.73 fewer to 173.73 more)

—

70
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life

—

—

—

—

—

Outcome not measured.

Adverse events/serious adverse events

—

—

—

—

—

No comparative data between groups available for all types of adverse event. However, the study reported rate ratios specifically for foot lesions and ulcers in participants with diabetic peripheral neuropathy.

Over 12 months, the reported rate ratio for all types of foot lesions (ignoring multiple lesions/episode) was 1.24 (95% CI 0.70 to 2.19; 1 study, 70 participants). Based on the point estimate, intervention may have led to higher rate of foot lesions; however, the 95% CI included the possibility of no difference or an effect in either direction.

Over 12 months, the reported rate ratio for all full‐thickness foot ulcers (ignoring multiple lesions/episode) was 0.96 (95% CI 0.38 to 2.42; 1 study, 70 participants). Based on the point estimate, intervention may have led to a lower rate of full‐thickness foot ulcers; however, the 95% CI included the possibility of no difference or an effect in either direction.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MD: mean difference; NMD: neuromuscular disease; RCT: randomised controlled trial.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422109324271426071.

aDowngraded once for study limitations associated with an unclear risk of bias in random sequence generation.
bDowngraded once for imprecision associated with a wide CI.

Open in table viewer
Summary of findings 2. A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD

Sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme

Patient or population: people with NMD

Setting: USA and Qatar

Intervention: sensor‐based exercise programme

Comparison: no sensor‐based exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no exercise programme

Risk with exercise programme

Time spent walking (hours per 48 hours, activity monitor) assessed with: final scores, after intervention

Follow‐up: 4 weeks

The mean time spent walking was 4.12 hours

MD 0.64 hours fewer (2.42 fewer to 1.13 more)

—

25
(1 RCT)

⊕⊝⊝⊝
Verylowa,b

—

Quality of life (SF‐12 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 weeks

The mean quality of life (SF‐12 PCS) was 40.12 points

MD 0.24 points higher (5.98 lower to 6.46 higher)

—

35
(1 RCT)

⊕⊝⊝⊝
Verylowa,b

—

Quality of life (SF‐12 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 weeks

The mean quality of life (SF‐12 MCS) was 47.3 points

MD 5.1 points higher (0.58 lower to 10.78 higher)

—

35
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐12: 12‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422110856458804259.

aDowngraded once for study limitations associated with a high risk of selective reporting and attrition bias.
bDowngraded twice for imprecision associated with a very wide CI.
cDowngraded once for imprecision associated with a wide CI.

Open in table viewer
Summary of findings 3. A functional programme compared to a stretching programme in people living with NMD

Functional programme compared to stretching programme

Patient or population: people with NMD

Setting: Bethesda, Maryland, USA

Intervention: functional exercise programme

Comparison: stretching exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with stretching programme

Risk with functional programme

Physical activity (unspecified count per day, activity monitor) assessed with: final scores, during intervention

Follow‐up: 12 weeks

The mean physical activity (unspecified counts per day, activity monitor) was 70,498 counts

MD 8701 counts lower (38,293.3 lower to 20,891.3 higher)

—

43
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, unclear if during or after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 12 weeks

The mean quality of life (SF‐36 PCS) was 34.1 points

MD 1.1 points lower
(5.22 lower to 3.02 higher)

—

49
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, unclear if during or after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 12 weeks

The mean quality of life (SF‐36 MCS) was 54.4 points

MD 1.1 points lower
(6.79 lower to 4.59 higher)

—

49
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No usable adverse event data available.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422114406971747153.

aDowngraded once for study limitations associated with a high risk of attrition bias.
bDowngraded once for imprecision associated with a wide CI.

Background

Description of the condition

People with neuromuscular disease (NMD) are part of a clinically heterogeneous population with inherited or acquired disorders of muscle, peripheral nerve, neuromuscular junction, or anterior horn cell of the spinal cord (Fowler 2002; Öksüz 2011). Diagnosis is based on genetic testing where possible, biopsy, and established clinical criteria. In most types of primary NMD (e.g. Duchenne muscular dystrophy), prevalence rates are estimated to vary between 1 and 10 per 100,000 population. The estimated prevalence is higher for Charcot‐Marie‐Tooth (CMT) disease and postpolio syndrome (PPS), at over 10 per 100,000 (Deenen 2015). More than one in four people are estimated to have diabetic peripheral neuropathy (DPN) as a secondary complication of type 2 diabetes (Lu 2020). The number of people affected by DPN may rise further with an increasing global prevalence of type 2 diabetes that is, in part, attributed to less physically active lifestyles (Saeedi 2019).

NMD manifests with different patterns of disease activity and progression, sometimes requiring therapeutic intervention, assistive technology, and aids to support movement. Muscle weakness and limitations in activities of daily life are common features but everyday activity, and quality of life, may also be impacted by other factors, such as foot lesions, difficulties with balance, pain, and fatigue. Secondary disuse weakness and cardiovascular deconditioning may develop over time, which increases the risk of further chronic health problems and complications (Aitkens 2005; Apabhai 2011; Dal Bello‐Haas 2013; Fowler 2002; Jimenez‐Moreno 2017; Kilmer 2005; McDonald 2002; Öksüz 2011; Phillips 2009; Ramdharry 2017; Voet 2013; White 2004; WHO 2020a). According to the World Health Organization (WHO), physical inactivity is one of the main risk factors for non‐communicable diseases mortality (WHO 2020b).

Description of the intervention

WHO recommends regular and adequate physical activity, which is based on a minimum duration, intensity, frequency, and type of physical activity in different age groups. For people unable to meet the recommendations due to health conditions, WHO advises being as physically active as possible (WHO 2020a). In muscle‐wasting conditions, recommendations for exercise include more specific information on precautions and progression, as well as guidance on duration, intensity, frequency, and type of exercise (MDUK 2014).

This review includes any intervention that aims to promote physical activity in adults or children with NMD. We used the WHO definition of physical activity as "any bodily movement produced by skeletal muscles that requires energy expenditure" – including all movement during leisure time, while working, and travelling (WHO 2020a; WHO 2020b). As previously highlighted by the American College of Sports Medicine (ACSM), 'physical activity' and 'exercise' are sometimes used interchangeably, but the latter is a specific form of physical activity that consists of "planned, structured, and repetitive bodily movement done to improve or maintain one or more components of physical fitness" (ACSM 2010).

Exercise is often the form of physical activity studied in NMD. However, there are other potential types of lifestyle intervention (as per the WHO definition, such as monitoring, advice, and support) that may also help to promote physical activity (Foster 2005; Foster 2013; Richards 2013a; Richards 2013b). At a population level, a multi‐component approach is often taken, involving policy and environmental changes, as well as behavioural and informational interventions (Baker 2015). In this review, we considered the promotion of physical activity to include any strategy or approach that contributes to people with NMD becoming more physically active.

How the intervention might work

In studies of apparently healthy populations, short‐ to medium‐term improvements in self‐reported physical activity outcomes and cardiorespiratory fitness follow physical activity interventions compared with no intervention, attention control (e.g. general health check of an equivalent duration), minimal intervention, or a combination of these (Foster 2005). There is also some evidence in favour of particular modes of intervention delivery, such as use of technologies with support from a trained professional (Foster 2013). However, this evidence excludes people with known medical conditions, and findings after community‐level interventions have been inconsistent (Baker 2015; Foster 2005; Foster 2013). For apparently healthy populations (within which as many as one in two people with diabetes mellitus globally are thought to be undiagnosed (Saeedi 2019)), increasing and maintaining regular physical activity is likely to be beneficial in terms of reducing all‐cause mortality risk, as well as for the primary and secondary prevention of noncommunicable diseases, such as diabetes mellitus, cardiovascular disease, colon and breast cancer, osteoporosis, and depression, as well as risk factors such as hypertension and obesity. At a mechanistic level, routine physical activity has been associated with enhanced mental well‐being, reduced blood pressure, and improvement in glucose control and other biomarkers for inflammation and cardiovascular disease risk (Warburton 2006). These effects might reduce the need for pharmacological or other treatment, the associated costs, and possible adverse effects. While the risk of chronic conditions will increase with age, the benefits of physical activity have been shown across the lifespan, with recommended 'doses' adjusted for children, adults, and older adults, as well as for those already living with chronic conditions and disability (Warburton 2006; Warburton 2017; WHO 2020a).     

The effect of interventions to promote physical activity may be different in people with certain medical conditions, such as NMD, compared with those living without such conditions. Ambulatory status may also vary but the potential to be more physically active applies to non‐ambulant as well as ambulant people with NMD. Several studies have highlighted that people with particular types of NMD are less physically active than apparently healthy controls without a diagnosed NMD, and have higher perceived barriers to becoming physically active (Aitkens 2005; Apabhai 2011; Heutinck 2017; McCrory 1998; Phillips 2009; Ramdharry 2017). This could suggest differences in the effect of physical activity in terms of biological mechanism or facilitation at an individual or community level. People with different types of NMD may also respond differently to physical activity interventions because of the clinical heterogeneity of their conditions (Voet 2013) (with variable disease pattern, severity, and progression), as well as differences in the timing of disease onset in relation to developmental and ageing processes (e.g. childhood versus adult onset of NMD). Furthermore, people with NMD who are non‐ambulant may be at a greater risk from waking behaviours with low energy expenditure in sitting, reclining, or lying (referred to as sedentary behaviour) than those who are ambulant; this could have an impact on health outcomes that is independent of recommended doses of everyday physical activity. As such, sedentary behaviour could confound the effect of increased physical activity in terms of the risk for all‐cause mortality and chronic disease. However, there is evidence to suggest that physical activity can attenuate, if not negate, risks associated with prolonged sitting (Ekelund 2016).

The type and dose of physical activity can affect health benefits and complications, which in turn may shape everyday physical activity behaviour. Peak performance measures are often a primary outcome in studies involving physical activity although it may be unclear whether the intervention has actually changed everyday physical activity as an outcome. In terms of potential adverse effects of physical activity, there is currently limited evidence to assess the risk in NMD. Increasing physical activity may not always be appropriate for all people with NMD. The overworking of muscles affected by NMD could increase the risk of muscle damage and impairment. For example, overexertion can lead to myalgia (muscle pain), myoglobinuria (muscle protein in the urine, associated with muscle breakdown), weakness, and fatigue in people with muscle disease (MDUK 2014). For some, there may also be particular concerns about weight‐bearing activity, for example, in relation to managing falls risk or foot lesions in DPN. Focusing on exercise intervention as a specific form of physical activity, one systematic review of studies in NMD found no evidence of serious adverse events (Stefanetti 2020). Another systematic review in people with muscle disease highlighted that adverse event data from five included randomised controlled trials (RCT) was incomplete (Voet 2013). Six years later, an update of that review found low‐ to very low‐certainty evidence relating to adverse events (Voet 2019). There was no RCT evidence for exercise intervention in one systematic review involving people with McArdle disease (Quinlivan 2011). In peripheral neuropathy, one systematic review (including three RCTs) found one incidence of lower limb pain with exercise intervention, which was attributed to the aggravation of arthritis (White 2004). Although one systematic review of RCTs in amyotrophic lateral sclerosis (ALS) found no reported adverse effects due to exercise, fatigue and rapid deterioration resulting in death were reasons given for participants dropping out from one of the two included studies (Dal Bello‐Haas 2013).

Why it is important to do this review

The purpose of this review was to better understand the effects of different approaches for people living with NMD to become more physically active as part of a management strategy for health and well‐being.

Objectives

To assess the effects of interventions designed to promote physical activity in people with NMD compared with no intervention or alternative interventions.

Methods

Criteria for considering studies for this review

Types of studies

We included parallel RCTs involving people with any type of NMD. We included randomised cross‐over studies that matched our inclusion criteria. In cross‐over studies, participants each undergo more than one intervention. This study design is considered suitable for assessing "a temporary effect in the treatment of stable, chronic conditions" (Higgins 2020), and so may be suitable in some but not all types of NMD (i.e. not those where progression is expected to lead to a clinically important decline within the timescale of the study).

We planned to include quasi‐RCTs, defined as trials that allocated participants to groups using methods such as alternation, use of a case record number, or date of attendance. We referred to other types of evidence in the 'Discussion' only.

We included studies reported as full text and those published as abstract only. We also sought unpublished data for inclusion. There were no language restrictions.

Types of participants

We accepted studies that included adults, children, or both, with NMD. We considered studies in which NMDs had been diagnosed by any established criteria, and studies that did not describe diagnostic criteria or predated genetic diagnosis. As part of the spectrum of NMD, we included genetic or acquired peripheral nerve disorders, muscle diseases, neuromuscular junction, and motor neuron disorders. We excluded mechanical nerve compression conditions, such as carpal tunnel syndrome. We reported comorbidities where this information was available.

If studies included a subset of participants with NMD, we planned to contact the investigators or study sponsors to gather any relevant subgroup data not reported. If they were unable or unwilling to provide subgroup data, we would not have included these studies in the meta‐analysis. As a protocol deviation, we limited eligible study populations with a subset of participants with NMD to those study populations with neurological disorders, including a subset of participants with NMD. See Differences between protocol and review.

Types of interventions

We included studies of any practical (e.g. exercise or environmental adaptation), informational, or motivational intervention that was designed to promote physical activity, compared with no intervention, or another intervention designed to promote physical activity (Foster 2005). This included studies of any mode of delivery, dose, duration, or intensity, in a community setting. We included co‐interventions if they were provided to each group equally. For the purposes of this review, very brief interventions that might promote physical activity, such as general health checks, were included as interventions although these have also been defined as an attention control comparison elsewhere (Foster 2005). We reported details of supervisory support provided as part of an intervention, and we reported any concurrent treatment and care where this information was provided. We would have performed subgroup analyses to explore differences in the delivery of interventions if sufficient data had been available.

Potential interventions included one or a combination of the following (Foster 2005; Foster 2013; Richards 2013a; Richards 2013b):

  • one‐to‐one advice or support;

  • group advice or support;

  • telephone advice or support;

  • Internet‐based, mobile apps, or telehealth (remote) advice or support;

  • written advice or support;

  • self‐directed or unsupervised participation in a prescribed physical activity programme;

  • supervised physical activity in the home;

  • supervised physical activity in a facility;

  • monitoring device (e.g. accelerometer or pedometer);

  • other intervention designed to promote physical activity.

Types of outcome measures

Participation in physical activity was the main focus of this review and measurement of physical activity was an inclusion criterion for the review. However, the authors acknowledge that increases in physical activity may also lead to changes in quality of life and adverse events, which we explored as a secondary focus. We reported details of outcome measure assessment where this information was provided.

Primary outcomes

  • Physical activity: measured by self‐report or objectively, using monitoring devices (e.g. accelerometer or pedometer).

In accordance with the WHO definition of physical activity, primary outcomes related to everyday activity and included: overall physical activity, measured by self‐report using standardised questionnaires (e.g. the WHO Global Physical Activity Questionnaire, which collects information on activity at work or doing household chores, recreational activities, and travel to and from places); total time spent in physical activity, measured by self‐report or objectively (minutes/week); estimated total energy expenditure, measured by self‐report or objectively (calories or joules/week); step count, measured objectively (steps/week) (Foster 2005; Foster 2013; Richards 2013a; Richards 2013b).

We planned to prioritise the inclusion of physical activity outcomes in the following order: overall physical activity > total time spent in physical activity > estimated total energy expenditure > step count. We would have performed subgroup analysis to explore differences in physical activity measured subjectively and objectively if sufficient data had been available.

Secondary outcomes

  • Quality of life: measured by self‐report using standardised questionnaire scales (e.g. 36‐item Short Form Health Survey (SF‐36)).

  • Adverse events:

    • increase in pain: measured by self‐report;

    • any other, emergent and intervention‐related;

    • leading to discontinuation from study.

  • Serious adverse events:

    • hospitalisation;

    • all‐cause death.

As a protocol deviation, we included available data on serious adverse events a defined by one included study as 'serious adverse events'; see Differences between protocol and review.

To avoid potential selection bias, we planned to only analyse final values where studies did not report change scores. If either final or change scores were reported incompletely (e.g. without a measure of variability), we prioritised the reporting of those results reported most completely. As a protocol deviation we also prioritised the reporting of unadjusted final scores over adjusted change scores from one study; see Differences between protocol and review and Characteristics of included studies table. If we had sufficient data for meta‐analysis, we planned to make comparisons at the following time points:

  • less than six weeks from baseline;

  • six weeks to less than six months from baseline;

  • six to 12 months from baseline;

  • over 12 months from baseline.

We did not exclude study data available at multiple time points. If an included study reported multiple measures for the same outcome domain, we included the data for each of these measures. We planned to base the minimal important difference (MID) for outcomes on established values in the literature where possible. Where MIDs were unavailable, we reported this in the interpretation of the outcomes in the 'Results' and 'Discussion'.

Search methods for identification of studies

Electronic searches

On 30 April 2020, the Cochrane Neuromuscular Information Specialist searched the following databases:

  • Cochrane Neuromuscular Specialised Register via the Cochrane Register of Studies (CRS‐Web; Appendix 1);

  • Cochrane Central Register of Controlled Trials (CENTRAL) via CRS‐Web (Appendix 2);

  • MEDLINE (1946 to 30 April 2020; Appendix 3);

  • Embase (1974 to 2020 Week 17; Appendix 4);

  • US National Institutes for Health Clinical Trials Registry, ClinicalTrials.Gov (Appendix 5);

The WHO International Clinical Trials Registry Portal (ICTRP; apps.who.int/trialsearch/) was not accessible at the search date; however, most of its content is indexed in CENTRAL. We searched all databases from their inception, and we imposed no restriction on language of publication.

Searching other resources

We searched review articles for additional references but not the reference lists of included studies; see Differences between protocol and review. We also searched for errata or retractions of included studies.

Data collection and analysis

Selection of studies

One review author (KJ) imported all results of the search into Covidence software for dual screening (Covidence). Two review authors (KJ and FH) independently screened titles and abstracts of all potential studies identified by the search for inclusion and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. During selection of studies, the review author team clarified eligibility of particular conditions not specified in the search strategies and in accordance with the scope of Cochrane Neuromuscular. Review authors also clarified the exclusion of studies, including exercise‐based studies, that did not explicitly refer to physical activity measurement or promotion. As an extension to the protocol, one review author (KJ) noted eligibility rationale for all judgements on potentially eligible records initially identified in Covidence, and both review authors (FH and KJ) performed a second screen of these records (see Differences between protocol and review). This additional round of screening narrowed down the potentially eligible records. We retrieved the full‐text study reports/publications, and two review authors (KJ and JN) independently screened the full text and identified studies for inclusion. We identified and recorded reasons for exclusion of the ineligible studies. We resolved any disagreement through discussion and consulted a third review author (GR) to confirm eligibility of a study population. The Information Specialist for Cochrane Neuromuscular (FS) identified and excluded duplicates and we collated multiple reports of the same study so that each study rather than each report was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Moher 2009), Characteristics of included studies table, and Characteristics of excluded studies table.

Data extraction and management

We used a data extraction form for study characteristics and outcome data that had been piloted on at least one study in the review. We planned to apply the TIDieR Checklist (Template for intervention Description and Replication; The EQUATOR Network), but focused on completion of the Checklist by included studies. We considered other intervention‐reporting guidance (including the CERT framework (Consensus on Exercise Reporting Template; Slade 2016) and MARS (Mobile App Rating Scale; Stoyanov 2015)) as part of our discussion of the evidence. At least one review author (KJ and FH, JN, JM, or GR) extracted the following study characteristics from included studies: study design and setting, characteristics of participants, eligibility criteria, intervention details, outcomes assessed, source(s) of study funding, and any conflicts of interest among investigators.

Two review authors (KJ and FH, JN, JM, or GR) extracted outcome data from included studies and one review author (KJ) transferred data into Review Manager 5 and RevMan Web (Review Manager 2020; RevMan Web 2020). If a review author had been involved in a potential included study, another uninvolved review author completed data extraction instead. As a protocol deviation, the first author (KJ) contributed to data extraction for one study despite involvement due to constraints on co‐author availability (see Differences between protocol and review). We noted in the Characteristics of included studies table if outcome data were not reported in a usable way. We resolved any disagreements by discussion. We planned for a third review author to check the outcome data entries and spot‐check study characteristics for accuracy against the trial report. As a protocol deviation, the first author (KJ) completed these checks (see Differences between protocol and review).

If full text reports had required translation, the translator would have extracted data directly using a data extraction form, or authors would have extracted data from the translation provided. Where possible, a review author would have checked numerical data in the translation against the study report.

Assessment of risk of bias in included studies

Two review authors (KJ and JN, FH, JM, or GR) independently assessed risk of bias for each outcome using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We resolved any disagreements by discussion. If a review author had involvement in any potential included studies, we planned for a third review author to complete the assessment instead. As a protocol deviation, one review author involved in an included study (KJ) contributed to its risk of bias assessment (see Differences between protocol and review). We assessed the risk of bias according to the following domains:

  • random sequence generation;

  • allocation concealment;

  • blinding of participants and personnel;

  • blinding of outcome assessment;

  • incomplete outcome data;

  • selective outcome reporting;

  • other bias.

We graded each potential source of bias as high, low, or unclear risk and provided a quote from the study report together with a justification for our judgement in the risk of bias table. We summarised the risk of bias judgements across different studies for each of the domains listed; we planned to consider all outcomes separately as some domains may have different risks of bias for different outcomes but we presented assessments at the study level with reference to outcomes of specific relevance to the review. If information on risk of bias had related to unpublished data or correspondence with a trialist, we would have noted this in the risk of bias table.

When considering treatment effects, we considered the risk of bias for the evidence that contributed to that outcome.

Assessment of bias in conducting the systematic review

We conducted the review according to the published protocol (Jones 2020), and reported any deviations from it in the Differences between protocol and review section.

Measures of treatment effect

The main effect of interest in this review was the effect of assignment to the intervention rather than adherence, which is a different review question. Therefore, we planned to limit our meta‐analysis to the intention‐to‐treat (ITT) population if we had sufficient data.

Dichotomous data

We analysed dichotomous data as risk ratios (RR) with 95% confidence intervals (CIs). If we had found rare events (zero in either arm or less than 1%), we would have used the Peto odds ratio (Peto OR) with 95% CIs. To assess absolute risk where there were zero events in the control arm, we would have calculated the risk difference (RD) with 95% CIs in Review Manager 5 (Review Manager 2020).

Continuous data

We presented continuous data as mean difference (MD) with 95% CIs. Where studies reported a mean with standard error or 95% CI, we calculated the standard deviation (SD) using Review Manager 5 (Review Manager 2020). If we had undertaken meta‐analysis, we would have considered using the standardised mean difference (SMD) with 95% CIs for results across studies with outcomes that were conceptually the same but measured in different ways (including physical activity questionnaires and health‐related quality of life questionnaires). Where necessary, we would have combined final values and change scores in the same analysis if reporting the MD but not when reporting the SMD. We would have entered data presented as a scale with a consistent direction of effect.

We planned to undertake meta‐analyses only where this was meaningful (i.e. if the interventions, participants, and underlying clinical question were similar enough for pooling to make sense).

Unit of analysis issues

Multiple arm studies

Where a study reported results for multiple arms, we planned to only include arms eligible for this review (although we would list additional arms in the Characteristics of included studies table). If two comparisons (e.g. intervention A versus intervention C and intervention B versus intervention C) had been included in the same meta‐analysis, we intended to avoid double‐counting by combining groups to create a single pair‐wise comparison (Higgins 2020). However, this approach was not found to be helpful on application (e.g. comparison of exercise versus advice versus usual care). If meta‐analysis had been appropriate, an alternative approach could have been to split the control group between multiple arms. Where the review included more than one comparison that could not be included in the same analysis, we reported the results for each comparison separately.

Cross‐over studies

As there may be carry‐over in the effect of physical activity promotion and a period effect in some neuromuscular conditions, we planned to only include first‐period data from cross‐over studies.

Cluster‐randomised controlled trials

We did not expect or find any eligible studies that were cluster‐RCTs; if we had found any cluster‐RCTs, we would have discussed these narratively in the review.

Within‐patient trial designs

We did not expect or find any eligible studies that used within‐patient trial designs (e.g. an uncontrolled before‐and‐after design as distinct from a cross‐over design); if we had found any, we would have considered these narratively in the review.

Dealing with missing data

We emailed investigators from two included studies to try to verify key study characteristics and obtain relevant missing numerical outcome data where possible (e.g. if a study was available as an abstract only).

Assessment of heterogeneity

If we had undertaken meta‐analysis, we would have used the I² statistic to measure heterogeneity among the studies in each analysis. We planned to use the rough guide to interpretation as outlined in Chapter 11 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020), as follows:

  • 0% to 40%: might not be important;

  • 30% to 60%: may represent moderate heterogeneity;

  • 50% to 90%: may represent substantial heterogeneity;

  • 75% to 100%: considerable heterogeneity.

We would have also considered the following factors: the overlap of CIs in forest plots, whereby poor overlap is expected to indicate heterogeneity; the Chi² test included in forest plots, for which a large result relative to the degrees of freedom is expected to indicate heterogeneity; a low P value for heterogeneity (less than 0.10) in forest plots.

If we identified substantial unexplained heterogeneity, we would have reported it and explored possible causes narratively and by prespecified subgroup analysis.

Assessment of reporting biases

We note that small‐study effects can bias results even in the absence of heterogeneity. If we had been able to pool more than 10 studies, we would have created and examined a funnel plot to explore possible small‐study biases, as detailed in the Cochrane Handbook for Systematic Reviews of Interventions (Page 2021).

Data synthesis

If we had undertaken meta‐analysis, we would have used the Mantel‐Haenszel (M‐H) method to meta‐analyse dichotomous data, and the inverse variance method to meta‐analyse continuous data. We planned to use a random‐effects model in Review Manager 5 and RevMan Web (Review Manager 2020; RevMan Web 2020), on the assumption that different studies were estimating different, yet related, intervention effects (Deeks 2020).

Subgroup analysis and investigation of heterogeneity

If we had sufficient data, we planned to carry out the following subgroup analyses to investigate clinically plausible differences in the intervention effect (Deeks 2020):

  • in different types of NMD, including muscle disease, peripheral nerve disorders, neuromuscular junction disorders, and motor neuron disorders (analysis 1);

  • adults (aged 18 years or older) versus children (aged less than 18 years) versus mixed adults and children (analysis 2);

  • ambulatory (independent walking and occasional use of an assistive device) versus non‐ambulatory (habitual use of an assistive device or wheelchair) versus mixed ambulatory and non‐ambulatory (analysis 3);

  • regular supervisory support (operational definition: at least two scheduled sessions with identified personnel to plan and progress activity) versus no regular supervisory support (analysis 4);

  • subjectively measured physical activity versus objectively measured physical activity (analysis 5).

On reviewing the evidence, we removed one prespecified subgroup analysis for comparing general health visits with other interventions designed to promote physical activity because we would have considered this as a main comparison (see Differences between protocol and review). If meta‐analysis had been undertaken, we would have had sufficient subgroup information available for subgroup analysis 1 only. We did not attempt to meta‐analyse different physical activity outcome measures as a single domain.

We planned to use the following primary outcomes in subgroup analyses.

  • Overall physical activity, measured by self‐report using standardised questionnaires (e.g. Global Physical Activity Questionnaire).

  • Total time spent in physical activity, measured by self‐report or objectively (minutes/week).

  • Estimated total energy expenditure, measured by self‐report or objectively (calories or joules/week).

  • Step count, measured objectively (steps/week).

For subgroup analyses 1 to 4, we prioritised the inclusion of physical activity outcomes using the order above. For subgroup analysis 5, we planned to only include comparable measures of physical activity (total time; total energy expenditure).

We would have used the formal test for subgroup differences in Review Manager 5 or RevMan Web (Review Manager 2020; RevMan Web 2020). Overlap of CIs and a high I² statistic would indicate a difference between subgroups, and suggest there could be differential effects of interventions to promote physical activity in different types of NMD.

Sensitivity analysis

We initially planned to carry out the following sensitivity analyses to investigate the robustness of findings to the decisions made in obtaining them (Deeks 2020). If sufficient data had been available for meta‐analysis of the primary outcome, we would have repeated the analysis:

  • excluding unpublished studies (if there were any);

  • excluding studies that did not describe diagnostic criteria for NMDs;

  • excluding studies at high risk of bias for missing data;

  • excluding the data from cross‐over studies;

  • using a fixed‐effect model.

If we had sufficient data for meta‐analysis, a sensitivity analysis for adjusted and unadjusted results may also have been helpful.

Reaching conclusions

We based our conclusions only on findings from the quantitative or narrative synthesis of included studies for this review. We avoided making recommendations for practice and, in our implications for research, suggested priorities for future research and outlined remaining uncertainties in the area.

Summary of findings and assessment of the certainty of the evidence

We created summary of findings tables using GRADEpro GDT software (GRADEpro GDT), and intended to present the following outcomes.

  • Physical activity: overall physical activity measured by self‐report, using standardised questionnaires (e.g. Global Physical Activity Questionnaire) at less than six weeks from baseline.

  • Physical activity: overall physical activity measured by self‐report using standardised questionnaires (e.g. Global Physical Activity Questionnaire) at six weeks to less than six months from baseline.

  • Physical activity: total time spent in physical activity measured by self‐reported minutes/week at less than six weeks from baseline.

  • Physical activity: total time spent in physical activity measured by self‐reported minutes/week at six weeks to less than six months from baseline.

  • Quality of life: measured by self‐report using standardised questionnaire scales (e.g. SF‐36) at less than six weeks from baseline.

  • Quality of life: measured by self‐report using standardised questionnaire scales (e.g. SF‐36) at six weeks to less than six months from baseline.

  • Adverse events leading to discontinuation from study.

We presented results for three of the eight main comparisons in this review, using one summary of findings table for each comparison. In the absence of any usable evidence on overall physical activity, we did not include this in the summary of findings. The three main comparisons in people living with NMD included: a physical activity programme compared to no physical activity programme; a sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme; and a functional exercise programme compared to a stretching exercise programme. We created additional tables for four other comparisons that did not report the prespecified physical activity outcomes for summary of findings tables. These comparisons included: an aerobic exercise programme compared to no aerobic exercise programme; an aerobic exercise programme compared to cognitive behavioural therapy (CBT); CBT compared to no CBT; CBT with or without an exercise programme compared no CBT and no exercise programme. We did not create a summary of findings table for another comparison of a weight‐bearing exercise programme with a non‐weight‐bearing exercise programme because this comparison included no summary of findings table outcomes. Physical activity parameters were the primary outcome of interest for assessing the effect of interventions to promote physical activity. As a protocol deviation, we included any measure of total time spent in physical activity (i.e. using an activity monitor or self‐reported measure; see Differences between protocol and review). In addition, we included well‐being and safety aspects, which could influence intervention uptake and adherence across a spectrum of NMDs. We planned to prioritise standardised, self‐reported outcome measurement and time points that would include both very brief interventions and short‐ to medium‐length programmes of physical activity promotion. Although prioritising outcomes can assist decision‐makers, as a protocol deviation, we reported outcomes at multiple time points to also include longer term follow‐up. We included multiple time points for outcomes (as per the studies) in the absence of a single, appropriate standard for outcome reporting in people with different types of NMD (see Differences between protocol and review). We also reported mental and physical component summary scores for quality of life (as per the studies) to help capture the impact of physical and communication‐based approaches to physical activity promotion. Two review authors (KJ and JN, FH, or GR) used the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to independently assess the certainty of the body of evidence (studies that contributed data for the prespecified outcomes). We used methods and recommendations described in Chapters 11 and 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2011a; Schünemann 2011b).

For assessing imprecision consistently in continuous outcomes, we considered the CI width for the intervention effect in relation to a cut‐off of ± 0.5 SD of the control group risk. We resolved any disagreements in GRADE judgements by discussion. If a review author had involvement in any included studies, we planned for a third person to complete the assessment instead. As a protocol deviation, the first review author (KJ) was involved in an included study and contributed to its assessment (see Differences between protocol and review). We considered outcomes to have high‐certainty evidence if the five GRADE factors were not present to any serious degree, but downgraded the certainty to moderate, low, or very low according to review author interpretation. We downgraded evidence once if a GRADE consideration was serious and twice if very serious. We justified all decisions to downgrade the certainty of the evidence using footnotes and made comments to aid readers' understanding of the review where necessary. If we had undertaken meta‐analysis, we planned to use a median control group risk across studies but also report the second highest and second lowest control group risks as representative rates for assumed risk per row of the table (i.e. low‐, moderate‐, and high‐risk populations) where there was potentially important variation. We would have provided a source or rationale and corresponding time duration for the control group risk, indicating the types of participants in which this might apply. In the absence of meta‐analyses, we included narratively synthesised evidence within the summary of findings tables.

Results

Description of studies

We identified RCTs (including cross‐over trials) that explicitly aimed to promote physical activity in people living with NMD. We also identified studies in which physical activity was measured as an outcome, irrespective of the aim of the study. This approach facilitated the consideration of evidence in which 'promotion' could be evaluated as both an intentional and consequential action, although our main focus was the ITT population. We broadly identified three strata of interventions compared with each other or with usual care: structured physical activity support; structured exercise support (as a specific form of physical activity); and structured behaviour change support, including physical activity or exercise. These interventions focused on assessing benefits and harms within the included study population.

Results of the search

The Information Specialist for Cochrane Neuromuscular (FS) ran the search strategies for this review as published in the protocol and reported in Appendix 1; Appendix 2; Appendix 3; Appendix 4; and Appendix 5. The results of this search, performed on 30 April 2020, were as follows: 23,362 records identified through searching databases and 17,123 records screened following deduplication. See flow diagram in Figure 1.

Figure 1
Flow diagram. RCT: randomised controlled trial.

Flow diagram. RCT: randomised controlled trial.

Among 28 narrative and systematic reviews of potential relevance, we checked for physical activity outcome reporting in six other Cochrane Reviews (Bartels 2019; Dal Bello‐Haas 2013; Koopman 2015; Mehrholz  2015; Quinlivan 2011; Voet 2019). We identified no additional RCTs from the Cochrane Reviews for consideration in this review. However, we discussed the findings of this review within the context of evidence from other published reviews (see Discussion). We also evaluated the available trial information for 77 potentially relevant ongoing studies. We found five records for ongoing studies and two records for completed studies awaiting classification (see Characteristics of ongoing studies and Characteristics of studies awaiting classification tables). We did not seek unpublished data from ongoing studies or those awaiting classification but we plan to do so as part of the updating process of this review. We found additional published information from one of the studies awaiting classification (see Discussion). In total, we found 94 conference abstracts or full‐text reports, of which 13 studies (32 papers) met the inclusion criteria. We included seven studies in the quantitative analysis (three of which we included in summary of findings tables and four of which we summarised as additional tables). See Characteristics of included studies and Figure 1. None of the included studies (written in English) required translation.

Included studies

We included 13 studies (795 randomised participants from 12 studies; the number of participants eligible for the review in Elsworth 2011 was unclear). Elsworth 2011 reported inclusion of 26 participants with NMDs and 10 with other conditions including cerebral palsy, traumatic brain injury, and transverse myelitis. On contacting study investigators for more information, we understood 'other' diagnoses included two people with mitochondrial cytopathy, one with CMT disease, and one with polymyostasis. We were unable to ascertain participants' allocated intervention and decided not to seek additional outcome data for quantitative analysis.

The remaining 12 included studies compared interventions in people with a particular NMD (Andersen 2015; Andersen 2017; Grewal 2015; Koopman 2016; Lemaster 2008; Mueller 2013; Okkersen 2018; Shrader 2015; Van Groenestijn  2019; Voet 2014; Wallace 2019; White 2016). Participants had inherited or acquired NMDs including ALS, CMT disease type 1A, DPN, facioscapulohumeral muscular dystrophy (FSHD), inclusion body myositis (IBM), PPS, myotonic dystrophy type 1 (DM1), spinal and bulbar muscular atrophy (SBMA), and stable inflammatory immune‐mediated neuropathy (IN). Three studies involved participants whose NMD (DPN) was secondary to diabetes (Grewal 2015; Lemaster 2008; Mueller 2013). Four studies reported major comorbidities in some participants that may have prevented their participation in other studies (Andersen 2017; Lemaster 2008; Mueller 2013; Okkersen 2018) (see Characteristics of included studies table). Twelve studies excluded children and adolescents and one study did not specify eligibility by age, although participants were aged over 60 years on average. Nine studies included only ambulant participants and the other four studies did not specify baseline ambulatory status of participants but included ambulatory outcome measures.

One of the 13 included studies reported the aim of intervention being to increase physical activity (Lemaster 2008). Two included studies reported the effects of intervention on physical activity as a primary outcome measure (Lemaster 2008; Mueller 2013). All other included studies reported the effects of intervention on physical activity as either a secondary or exploratory outcome measure. Eight studies reported registration with clinical trial registers, of which four were partly or entirely conducted in the Netherlands (Koopman 2016; Okkersen 2018; Van Groenestijn  2019; Voet 2014), one in the USA (Mueller 2013), and four were partly or entirely conducted in the UK (Elsworth 2011; Okkersen 2018; Wallace 2019; White 2016). Five studies published a protocol; of these, four were set partly or entirely in the Netherlands (Koopman 2016; Okkersen 2018; Van Groenestijn  2019; Voet 2014), and two were set partly or entirely in the UK (Okkersen 2018; White 2016). One study published a description of interventions using the template for intervention and replication (TIDieR) checklist and guide by the EQUATOR Network (The EQUATOR Network) (Okkersen 2018).

Across 11 full reports of studies of people with NMD (excluding Elsworth 2011 and White 2016), 737 randomised participants were included of 2777 people invited or assessed for eligibility (less than 27%). As noted previously, relatively few studies reported everyday physical activity as a primary outcome. Other primary outcomes reported included measures of self‐reported fatigue and quality of life, and performance measures for postural stability, fitness, and functional ability. Aside from reimbursement for travel expenses and gym membership in several studies, one included study paid an additional cash incentive for study participation (Mueller 2013). This financial incentive could contribute to the effectiveness of physical activity‐promoting intervention but its analysis as a possible effect modifier is beyond the scope of this review.

How randomised controlled trial interventions relate to aspects of physical activity promotion

Elsworth 2011 compared the effects on physical activity of a 12‐week exercise programme with a Physical Activity Support System versus no exercise programme and no Physical Activity Support System in adults with neurological conditions including NMD. A stated aim of this parallel RCT was to assess the feasibility and safety of the activity‐supported intervention. Physical activity was one of the primary outcome measures, although not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the UK with potential participants recruited through local neurological services and the Dementias and Neurodegenerative Diseases Research Network (DeNDRoN). Of 103 people assessed for eligibility, 99 were randomised into the study (96%). Two people became unwell, one could not be contacted, and another declined to participate after further discussion.

  • Baseline characteristics and comparability: these data were not available for the subpopulation of participants with NMD. Overall baseline characteristics appeared similar between groups, in accordance with reporting by study authors. 

  • Physical activity outcome measurement: 'experienced physical activity' was reported using the Physical Activity Scale for the Elderly (PASE; ranging from 0 to 400+ with higher scores indicating a better outcome). Physical activity was also measured using a Step Activity Monitor (SAM) for eight days. Change scores in the composite PASE rating and daily steps (count) were reported over 12 weeks between assessments at baseline and at the end of intervention, with measurements reported to be completed after assessment visits. Final scores were also reported after 12 weeks and 24 weeks (three months' follow‐up). No results were reported for the subgroup of participants with NMDs.

Three studies involved adults with DPN (Grewal 2015; Lemaster 2008; Mueller 2013). Grewal 2015 compared the effects on physical activity of a four‐week sensor‐based, interactive exercise programme with no sensor‐based, interactive exercise programme in adults living with DPN. A stated aim of this parallel RCT was to assess the effects of intervention on physical activity. However, physical activity was neither a primary outcome nor used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the USA (Arizona) and Qatar, with potential participants recruited through outpatient clinics. Of 54 people assessed for eligibility, 39 were randomised into the study (72%). Eight people (15%) met exclusion criteria and seven people (13%) declined to participate.

  • Baseline characteristics and comparability: 19 people were randomised to the intervention group (42% male, ethnicity not reported) and 20 to the control group, of whom 16 supplied baseline characteristics (50% male, ethnicity not reported). The mean age of the intervention group was 62.6 years (SD 7.98) and 64.9 years (SD 8.50) in the control group. Baseline characteristics appeared similar between groups, in accordance with reporting by study authors.

  • Physical activity outcome measurement: a PAMSys activity sensor was positioned in the chest pocket of a custom‐made t‐shirt that was worn for 48 hours. Time spent walking (hours per 48 hours) and daily steps (count, activity monitor) were reported as final scores after four weeks, with the assessment assumed to be after the intervention.

Lemaster 2008 compared the effects on physical activity of a 12‐month physical activity programme (weight‐bearing) with no physical activity programme in adults living with DPN. A stated aim of this parallel RCT was to 'encourage participants to gradually increase total daily weight‐bearing steps'. Physical activity was the primary outcome measure and also used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the USA (Missouri) with recruitment through primary care, endocrinology, or podiatry practices of participants aged 50 years and over who received diabetes or foot care. Of 260 people invited to participate, only 79 were randomised into the study (30%). One hundred and five people (40%) declined to participate or could not be contacted; it is understood that 44 people (17%) did not meet inclusion criteria and 50 people (19%) met exclusion criteria.

  • Baseline characteristics and comparability:  41 people were randomised to the intervention group (53% male and 92% white ethnicity) and 38 to the control group (47% male and 93% white ethnicity). The mean age of the intervention group was 66.6 years (SD 10.4) and 64.8 years (SD 9.4) in the control group. Baseline characteristics appeared to be similar between groups, in accordance with reporting by study authors.

  • Physical activity outcome measurement: participants wore a StepWatch accelerometer on the ankle for 14 days. Time spent walking (minutes per week), daily steps (count) and steps taken in 30‐minute bouts (count) were reported as final scores after three months, six months, and 12 months. Assessment was during intervention at three months and six months, but it was unclear if the final assessment was during or after the intervention.

Mueller 2013 compared the effects on physical activity of a 12‐week weight‐bearing exercise programme with a non‐weight‐bearing exercise programme in people living with DPN. A stated aim of this parallel RCT was to assess the effects of intervention on physical activity, which was one of the primary outcome measures and also used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the USA with potential participants recruited through a database of previous participants, the Washington University School of Medicine Research Participant Registry, cable television commercials, a newspaper story, and recruitment posters displayed in a Diabetes Treatment Center and on area commuter trains. Of 265 people invited to participate, only 29 were randomised into the study (11%). Ninety people (34%) did not have diabetes or neuropathy, 84 people (32%) declined to participate, could not be contacted, had a time conflict or lack of interest; 43 people (16%) were excluded due to other illnesses, orthopaedic issues, or inability to exercise; and 19 people (7%) did not meet inclusion criteria.

  • Baseline characteristics and comparability: 15 people were randomised to the weight‐bearing group (67% male, ethnicity not reported) and 14 to the non‐weight‐bearing group (50% male, ethnicity not reported). The mean age of the weight‐bearing group was 65.2 years (SD 12.8) and 63.9 years (SD 12.5) in the non‐weight‐bearing group. We noted the difference in proportion of males, but baseline characteristics appeared to be broadly similar between groups, in accordance with reporting by study authors.

  • Physical activity outcome measurement: participants wore a StepWatch accelerometer on the ankle for 14 days with data used from a seven‐day period in which at least eight hours of activity were recorded and at least one weekend day. Daily steps (count) were reported as final scores after 16 weeks with the assessment completed after intervention.

Three studies involved adults with FSHD (Andersen 2015; Andersen 2017; Voet 2014). Andersen 2015 compared the effects on physical activity of a 12‐week exercise programme with a protein supplement versus an exercise programme with a placebo supplement versus neither intervention in adults with FSHD type 1. This parallel RCT did not state that it aimed specifically to promote or assess the effects of intervention on physical activity. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in Denmark with potential participants recruited through the Copenhagen Neuromuscular Center and from the Rehabilitation Centre for Neuromuscular Diseases in Denmark. Of 140 people assessed for eligibility, only 41 were randomised into the study (29%). Fifty people (36%) did not respond on contact, 40 people (29%) met exclusion criteria, and nine people declined to participate (6%).

  • Baseline characteristics and comparability: 18 participants were randomised to a protein supplement and training group of whom 13 were included in baseline characteristics (62% male, ethnicity not reported), 13 to a placebo supplement and training group (54% male, ethnicity not reported) and 10 to no intervention, of whom nine had included baseline data (56% male, ethnicity not reported). The mean age of the protein‐supplemented training group was 42.6 years (range 24 to 55), 45.7 years (range 22 to 63) in the placebo‐supplemented training group and 51.3 years (range 24 to 65) in the no intervention group. Baseline fitness and walking speed were reported to be lower in the non‐intervention group than the training groups but demographics appeared similar between groups.

  • Physical activity outcome measurement: participants wore a SenseWear Pro3 accelerometer for three days before and after 16 weeks of intervention. Physical activity was measured as daily steps (1000 counts) and daily energy expenditure (1000 kilocalories (kcal)). A Bouchard diary was also used to estimate daily energy expenditure (1000 kcal). Final scores were reported after the intervention but the results were not usable because the mean data were only reported with ranges.

Andersen 2017 compared the effects on physical activity of eight weeks of high‐intensity interval training (HIT) with no HIT in adults with FSHD type 1. This parallel RCT stated that it aimed specifically to promote or assess the effects of intervention on physical activity. In fact, the study investigators specified that participants could not change their activity during the study, although this was not an exclusion criterion. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in Denmark with potential participants recruited through the Copenhagen Neuromuscular Center. Of 97 people assessed for eligibility, only 13 were randomised into the study (13%). Forty‐three people (44%) met exclusion criteria, 25 people (26%) declined to participate, and 16 people (16%) did not respond on contact. 

  • Baseline characteristics and comparability: six participants were randomised to a supervised HIT group  (67% male, ethnicity not reported) and seven to a usual care group of whom six were included in baseline characteristics (83% male, ethnicity not reported). The mean age of the HIT group was 53 years (SD 15) and 46 years (SD 9) in the usual care group. Baseline self‐reported physical activity (Metabolic Equivalent of Task (MET) hours/week) was lower in the training group than usual care, but characteristics appeared broadly similar between groups, in accordance with reporting by study authors. 

  • Physical activity outcome measurement: an Omron Walking Style Pro pedometer was worn for four to seven days at baseline (before exercise intervention) and prior to the follow‐up assessment at eight weeks. Physical activity was also reported using the International Physical Activity Questionnaire (IPAQ). However, narrative results were reported only for physical activity measured as steps/day.

Voet 2014 compared the effects on physical activity of a 16‐week aerobic exercise programme versus CBT versus neither intervention in adults with FSHD. This parallel RCT did not state that it aimed to promote or assess the effects of intervention on physical activity. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place at nine healthcare institutions in the Netherlands. People with FSHD were invited to participate if they had participated in any previous study at the centre, were registered in a Dutch neuromuscular database, or participated in a patient support organisation. Of 337 people invited to participate, only 57 were randomised into the study (17%). One hundred and ninety‐nine people declined to participate (59%), 84 people did not respond on contact (25%), and 37 people either met exclusion criteria or did not meet inclusion criteria (11%). 

  • Baseline characteristics and comparability: 20 people were randomised to aerobic exercise training (60% male, ethnicity not reported), 13 to CBT (62% male, ethnicity not reported), and 24 to usual care (71% male, ethnicity not reported). The median age of the exercise group was 59 years (range 21 to 68), 49 years (range 24 to 69) in the CBT group and 52 years (range 20 to 79) in the usual care group. Baseline characteristics appeared to be similar between groups, in accordance with reporting by study authors. 

  • Physical activity outcome measurement: 'experienced physical activity' was reported using the physical activity subscale of the Checklist Individual Strength (CIS‐Activity), which includes three questions about activity over the previous two weeks and scores each question on a seven‐point Likert scale (higher scores indicate a poorer outcome). An actometer (model unreported) was also worn on the ankle for 12 days and nights and registered data analysed across this period. The CIS‐Activity subscale ratings and body accelerations per five‐minute period (count) were reported as change scores over 16 weeks with assessments completed before and after intervention but these were not usable because participants originally in the usual care group were later amalgamated with other intervention groups.

Van Groenestijn  2019 planned to compare the effects on physical activity of a 16‐week aerobic exercise programme with no aerobic exercise programme in adults with ALS. A secondary aim of this parallel RCT was to assess the effects of intervention on activity limitations. Study investigators planned to measure physical activity but the outcome was removed because participants did not complete the questionnaire.

  • Recruitment: the study took place in the Netherlands with consecutive participants screened at five rehabilitation centres or rehabilitation departments of academic hospitals. Of 325 people assessed for eligibility, only 57 were randomised into the study (18%). One hundred and seventy‐six people (54%) did not meet eligibility criteria, 77 people (24%) declined to participate, and 15 people (5%) enrolled in a CBT trial.

  • Baseline characteristics and comparability: 27 people (67% male, ethnicity not reported) were randomised to aerobic exercise therapy and 30 to usual care (73% male, ethnicity not reported). The mean age of the exercise group was 60.9 years (SD 10.0) and 59.9 years (SD 10.7) in the usual care group. Baseline characteristics appeared to be similar between groups, which is consistent with reporting by study authors although they undertook a propensity‐matched analysis for baseline inequalities.

  • Physical activity outcome measurement: study investigators applied the LASA Physical Activity Questionnaire (LAPAQ) to estimate METs per day. As a protocol deviation, no results were reported.

Koopman 2016 compared the effects on physical activity of a four‐month aerobic exercise programme versus CBT versus neither intervention in adults living with PPS. A stated aim of this parallel RCT was to assess the effects of intervention on improving activities. However, physical activity was neither a primary outcome nor used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the Netherlands with potential participants recruited through seven hospitals and rehabilitation centres. Of 490 people invited to participate, only 68 were randomised into the study (14%). Two hundred people (41%) declined to participate, 129 people (26%) did not respond on contact, and 93 people (19%) did not meet inclusion criteria.

  • Baseline characteristics and comparability: 23 people were randomised to the exercise therapy group, of whom 22 were included in baseline characteristics (41% male and 87% white ethnicity), 23 to CBT (43% male and 96% white ethnicity), and 22 to usual care (50% male and 82% white ethnicity). The mean age of the exercise group was 56.9 years (SD 8.9), 60.1 years (SD 8.2) in the CBT group, and 60.1 years (SD 8.2) in the usual care group. Baseline characteristics appeared to be similar between groups, in accordance with reporting by study authors. 

  • Physical activity outcome measurement: participants wore a StepWatch activity monitor for seven days. Daily steps (count) were reported as final scores after four, seven, and 10 months, with all assessments completed after intervention.

Okkersen 2018 compared the effects on physical activity of 10 months of CBT with or without an exercise programme versus no CBT and no exercise programme in adults with DM1. A stated aim of this parallel RCT was to assess the effects of intervention on health status. However, physical activity was neither a primary outcome nor used in a power calculation to determine the study sample size.

  • Recruitment: the multi‐centre study took place in France, Germany, the Netherlands, and the UK. Potential participants were recruited through DM1 registries, from clinics via their treating neurologists, or independent volunteering through patient organisations. Of 344 people assessed for eligibility, 255 were randomised into the study (74%). Eighty‐nine people (26%) were ineligible because they either met exclusion criteria or did not meet inclusion criteria.

  • Baseline characteristics and comparability: 128 people were randomised to CBT (55% male, ethnicity not reported) and 127 to standard care (53% male, ethnicity not reported). The mean age of the CBT group was 44.8 years (SD 11.7) and 46.4 years (SD 1.3) in the standard care group. Baseline characteristics appeared to be similar between groups, in accordance with reporting by study authors.

  • Physical activity outcome measurement: a GENEActiv tri‐axial accelerometer was worn on the ankle for seven to 14 consecutive days. Data were only analysed for days with at least 23 hours of registered activity over at least seven days. The first and last days of recorded activity were excluded 'to avoid confounding factors related to distribution or delivery procedures'. Physical activity, interpreted as mean magnitude of ankle acceleration over 24 hours and over five hours of highest and lowest activity, was reported as final scores after five, 10, and 16 months. The first assessment was completed during intervention and the final assessment postintervention, but it was unclear whether the assessment at 10 months was completed during or after the intervention.

Shrader 2015 compared the effects on physical activity of a 12‐week functional exercise programme with a stretching exercise programme in adults with SBMA. This parallel RCT did not state that it aimed specifically to promote or assess the effects of intervention on physical activity. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in Maryland, USA. The method of recruiting participants was not reported. Of 61 people assessed for eligibility, 54 people were randomised into the study (89%) and seven people met exclusion criteria.

  • Baseline characteristics and comparability: 27 people were randomised to a functional exercise group of whom 24 were included in baseline characteristics (sex and ethnicity not reported), and 27 to a stretching exercise group of whom 26 had included baseline data (sex and ethnicity not reported). The mean age of the functional exercise group was 53.8 years (SD 10.0) and 56.5 years (SD 8.1) in the stretching exercise group. Baseline characteristics appeared similar between groups, in accordance with reporting by study authors. 

  • Physical activity outcome measurement: an Actical accelerometer was worn for the first and last 10 days of the trial. Data were weighted by the number of days recorded and only included if there were at least six days of registered activity. Total physical activity was reported as an unspecified count per day. Final scores were reported after 12 weeks, with the assessment completed during intervention.

Wallace 2019 compared the effects on physical activity of a 16‐week aerobic exercise training programme versus regular telephone contact that included review of activity in adults with CMT type 1A  and IBM. A stated aim of this cross‐over RCT was to explore the secondary physical and non‐physical effects of exercise intervention. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study took place in the UK with potential participants recruited through clinics and research databases of the National Hospital for Neurology and Neurosurgery, plus national clinics of colleagues from the British Myology Society for people with IBM. Of 404 people assessed for eligibility, only 45 people (28 with CMT and 17 with IBM) were randomised into the study (11%). One hundred and fifty‐six people (39%) did not meet inclusion criteria, 71 people (18%) declined to participate, and 132 people (33%) were excluded for other reasons.

  • Baseline characteristics and comparability: 23 participants (with CMT and IBM) were randomised to an aerobic exercise training group during the first period of the cross‐over study, of whom 21 participants were included in baseline characteristics (67% male, ethnicity not reported). Twenty‐two participants were randomised to a control group during the first period, of whom 20 were included in the baseline data (65% male, ethnicity not reported). The mean age of the exercise group was 46.3 years (95% CI 37.2 to 55.4) in participants with CMT and 65.4 years (95% CI 59.1 to 71.8) in participants with IBM; in the control group, the mean age was 45.3 years (95% CI 35.9 to 54.6) in participants with CMT and 57.1 years (95% CI 50.4 to 63.9) in those with IBM. Baseline characteristics appeared similar between groups of participants with the same condition although exercising participants with IBM were slightly older than those who did not exercise. The study authors noted that, as a measure of fitness, peak oxygen uptake (VO2peak) did not fully return to baseline after an eight‐week washout period among participants with CMT.

  • Physical activity outcome measurement: the study used a SenseWear activity monitor to measure physical activity over seven days. Physical activity duration over 3 METs (minutes) was reported as a final score following 12 weeks of training, after the intervention. The timing of physical activity review differed between groups because of the cross‐over design of the study. The IPAQ was also used before and after intervention but with sitting time (minutes) reported as an outcome rather than physical activity. Participants' first period data over 12 weeks (before cross‐over) were not available separately.

White 2016 compared the effects on physical activity of a home exercise programme with written advice about physical activity in adults with IN. Based on the study protocol, a stated aim of this parallel RCT was to assess the effects of the intervention on activity limitation. Included as a secondary outcome measure, physical activity was not used in a power calculation to determine the study sample size.

  • Recruitment: the study protocol set out to recruit people attending selected specialist peripheral nerve clinics in the South East and West Midlands of England and people with IN who accessed the Guillain‐Barré syndrome and Associated Inflammatory Neuropathy (GAIN) charity website or newsletter. Fifty‐eight people were randomised into the study. No further recruitment information was available at the time of preparing this review.

  • Baseline characteristics and comparability: these data were not available.

  • Physical activity outcome measurement: the seven‐question IPAQ‐short was a prespecified outcome for measuring physical activity after 12 weeks' intervention and 12 months after intervention but no results were available.

See Characteristics of included studies table for a list of all outcome measures and further information.

Excluded studies

We excluded 62 conference abstracts or full‐text reports because the studies were not RCTs, they involved a different study population, or they did not report physical activity outcome measurement (see Characteristics of excluded studies table). We translated one excluded conference abstract from Spanish (Cejudo‐Ramos 2000). We also excluded a full report in Chinese without translation as it was identifiable as a review (Zhang 2005). 

Risk of bias in included studies

We completed risk of bias assessments based on full reports, with the exception of one study published only as an abstract (White 2016). We judged nine of the 13 studies to have a high risk of bias in at least one domain. Incomplete physical activity outcome reporting was one of the main reasons for this high risk of bias judgement. Two studies had at least two domains with an unclear risk of bias assessment, and two studies had a low risk of bias in all domains. See Characteristics of included studies table and, for a risk of bias summary, Figure 2.

Figure 2
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Most studies were at low risk of bias associated with random sequence generation. We judged two studies to have an unclear risk of bias because randomisation was reported by postcode or type of clinical site (Andersen 2017; Lemaster 2008).

In terms of allocation concealment, four studies had an unclear risk of bias (Andersen 2015; Andersen 2017; Shrader 2015; White 2016); the other studies were at low risk of bias. 

Blinding

We judged that it was not applicable to blind participants and personnel involved in supporting any study intervention for promoting physical activity and so we considered this at low risk of bias throughout. However, the blinding of outcome assessors was feasible. We judged most studies to have a low risk of bias associated with the blinding of outcome assessors. We found that two studies had an unclear risk of detection bias (Grewal 2015; Okkersen 2018), and we found one study to have a high risk of detection bias because the study was reported to be unblinded (Andersen 2017).

Incomplete outcome data

We judged nine of 13 studies to have a high risk of bias associated with incomplete physical activity outcome data because of large proportions of missing data or unexplained missing data (Andersen 2015; Andersen 2017; Grewal 2015; Koopman 2016; Okkersen 2018; Shrader 2015; Van Groenestijn  2019; Voet 2014; Wallace 2019). One study had an unclear risk of attrition bias (White 2016), and the other studies were at low risk of attrition bias.

Selective reporting

We judged seven studies at low risk of selective reporting. We found three studies had an unclear risk of selective reporting (Andersen 2017; Wallace 2019; White 2016). Three studies were at high risk of selective reporting because available results were reported incompletely (Grewal 2015; Koopman 2016; Van Groenestijn  2019). In terms of selective non‐reporting, two studies had prespecified outcome measures (including the EuroQol (EQ‐5D) and LAPAQ) that were not reported in the results (Koopman 2016; Van Groenestijn  2019). Another study prespecified activities of daily living assessment but, as a protocol deviation, the study investigators reported they had forgotten to delete this outcome domain from the list; we did not judge this to be reporting bias (Okkersen 2018).

Other potential sources of bias

We judged one study at unclear risk of other potential sources of bias because the study was published as an abstract only (White 2016). We did not identify any of the other 12 studies to have other potential sources of bias.

Effects of interventions

See: Summary of findings 1 A physical activity programme (weight‐bearing) compared to no physical activity programme in people living with NMD; Summary of findings 2 A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD; Summary of findings 3 A functional programme compared to a stretching programme in people living with NMD

Interventions with physical activity as a primary outcome

A physical activity programme (weight‐bearing) compared to no physical activity programme in people living with NMD

One RCT involving people with DPN contributed data for this comparison (Lemaster 2008).

At six months, 18 participants (45%) in the physical activity programme and 13 participants (35%) in the control group adhered to more than half of the study protocol elements. At 12 months, only seven participants (18%) in the physical activity programme and nine participants (24%) in the control group adhered to more than half of the study protocol elements. The study reported ITT analysis of randomised participants, irrespective of adherence, but did not specify the assumptions made for ITT analysis.

Time spent walking (minutes per week, activity monitor)

After three months (during intervention): the MD was 34 min per week (95% CI –92.19 to 160.19; 1 study, 69 participants; low‐certainty evidence; Analysis 1.1) in favour of the physical activity programme but the CIs included the possibility of an effect favouring either the physical activity programme or no physical activity programme (summary of findings Table 1).

After six months (during intervention): the MD was 68 min per week (95% CI –55.35 to 191.35; 1 study, 74 participants; low‐certainty evidence; Analysis 1.1) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme (summary of findings Table 1).

After 12 months (unclear if during or after intervention): the MD was 49 min per week (95% CI –75.73 to 173.73; 1 study, 70 participants; low‐certainty evidence; Analysis 1.1) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme (summary of findings Table 1).

We downgraded the certainty of the evidence once for study limitations associated with an unclear risk of bias in random sequence generation, and once for imprecision associated with a wide CI. We did not identify an MID from the literature although WHO guidance emphasises at least meeting recommended levels of physical activity (WHO 2020a; WHO 2020b). On the premise that any increase in physical activity is considered important, the intervention may have led to an important increase in physical activity.

Daily steps (count, activity monitor)

After three months (during intervention): the MD was 178 steps per day (95% CI –459.81 to 815.81; 1 study, 69 participants; Analysis 1.2) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

After six months (during intervention): the MD was 408 steps per day (95% CI –243.40 to 1059.40; 1 study, 74 participants; Analysis 1.2) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

After 12 months (unclear if during or after intervention): the MD was 262 steps per day (95% CI –407.40 to 931.40; 1 study, 70 participants; Analysis 1.2) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

Steps taken in 30‐minute bouts (count, activity monitor)

After three months (during intervention): the MD was 50 steps in 30‐minute bouts (95% CI –27.66 to 127.66; 1 study, 69 participants; Analysis 1.3) in favour of the physical activity programme but the CIs included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

After six months (during intervention): the MD was 83 steps in 30‐minute bouts (95% CI –20.95 to 186.95; 1 study, 74 participants; Analysis 1.3) in favour of the physical activity programme but the CIs included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

After 12 months (unclear if during or after intervention): the MD was 33 steps in 30‐minute bouts (95% CI –69.56 to 135.56; 1 study, 70 participants; Analysis 1.3) in favour of the physical activity programme but the CIs included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

Quality of life

The study did not report quality of life.

Adverse events

There were no comparative data between groups for any type of adverse event. However, we narratively reported rate ratios calculated in the study specifically for foot lesions and ulcers (see Comments in summary of findings Table 1).

Over 12 months, the reported rate ratio for all types of foot lesions (ignoring multiple lesions/episode) was 1.24 (95% CI 0.70 to 2.19; 1 study, 70 participants) in favour of no physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

Over 12 months, the reported rate ratio for all full‐thickness foot ulcers (ignoring multiple lesions/episode) was 0.96 (95% CI 0.38 to 2.42; 1 study, 70 participants) in favour of the physical activity programme but the CI included the possibility of an effect favouring either the physical activity programme or no physical activity programme.

These study‐reported data on foot lesions and full‐thickness ulcers were limited by imprecision.

Serious adverse events

There were no data on the number of participants with serious adverse events.

A weight‐bearing exercise programme compared to a non‐weight‐bearing exercise programme in people living with NMD

One RCT involving people with DPN contributed data for this comparison (Mueller 2013). We did not create a summary of findings table for this comparison because none of the study outcomes matched prespecified outcome measurement for inclusion.

The mean proportion of participants attending all exercise sessions in the weight‐bearing programme was 83.4% (SD 11), compared with 83.3% (SD 10.8) in the non‐weight‐bearing programme. The study reported ITT analysis of randomised participants but did not specify assumptions made for ITT analysis.

Daily steps (count, activity monitor)

After 16 weeks (after intervention): the MD was –485 steps per day (95% CI –1773.66 to 803.66; 1 study, 29 participants; Analysis 2.1) in favour of the non‐weight‐bearing exercise programme but the CI included the possibility of an effect favouring either the non‐weight‐bearing exercise programme or the weight‐bearing exercise programme.

Quality of life

The study did not report quality of life.

Adverse events

There were no comparative data available between groups for any type of adverse event. We calculated RRs specifically for foot lesions and ulcers in the absence of study‐reported rate ratios, and reported these narratively.

Participants with foot lesions after 16 weeks: the RR was 1.31 (95% CI 0.54 to 3.17; 1 study, 29 participants) in favour of the non‐weight‐bearing exercise programme but the CI included the possibility of an effect favouring either the non‐weight‐bearing exercise programme or the weight‐bearing exercise programme.

Participants with foot ulcers after 16 weeks: the RR was 0.50 (95% CI 0.05 to 4.90; 1 study, 29 participants) in favour of the weight‐bearing exercise programme but the CI included the possibility of an effect favouring either the weight‐bearing exercise programme or the non‐weight‐bearing exercise programme.

Serious adverse events

There were no data on the number of participants with serious adverse events.

Interventions with physical activity as a secondary or exploratory outcome

A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD

One RCT involving people with DPN contributed data for this comparison (Grewal 2015).

The study did not report on participants' adherence. Physical activity was reported for a subgroup of participants without further details, and those with active foot ulcers, among other contraindications, were excluded from the study analysis (see Characteristics of included studies table). Other outcomes were reported using per‐protocol analysis.

Time spent walking (hours per 48 hours, activity monitor)

After four weeks (interpreted to be after intervention): the MD was –0.64 hours per 48 hours (95% CI –2.42 to 1.13; 1 study, 25 participants; very low‐certainty evidence; Analysis 3.1) in favour of no exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (summary of findings Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of selective reporting and attrition bias, and twice for imprecision associated with a very wide CI.

Daily steps (count, activity monitor)

After four weeks (interpreted to be after intervention): the MD was 1788 steps per day (95% CI –3440.55 to 7016.55; 1 study, 25 participants; Analysis 3.2) in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme.

Quality of life (12‐item Short Form Health Survey, Physical Component Score, questionnaire)

After four weeks (interpreted to be after intervention): the MD was 0.24 points (95% CI –5.98 to 6.46; 1 study, 35 participants; very low‐certainty evidence; Analysis 3.3) in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (summary of findings Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of selective reporting and attrition bias, and twice for imprecision associated with a very wide CI.

Quality of life (12‐item Short Form Health Survey, Mental Component Score, questionnaire)

After four weeks (interpreted to be after intervention): the MD was 5.10 points (95% CI –0.58 to 10.78; 1 study, 35 participants; low‐certainty evidence; Analysis 3.4) in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (summary of findings Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of selective reporting and attrition bias, and once for imprecision associated with a wide CI.

We did not identify an anchor‐ or distribution‐based MID for the SF‐12 quality of life questionnaire in people with DPN or NMD. Based on the effect estimates and low‐certainty evidence, we found that the sensor‐based interactive exercise programme may have made little or no difference to the Mental Component Score (MCS) and Physical Component Score (PCS) for quality of life.

Adverse events

There were no comparative data available between groups for any type of adverse event.

Serious adverse events

The study did not report serious adverse events.

An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD

Two RCTs involving people with PPS or ALS contributed data for this comparison but we did not meta‐analyse them because they measured outcomes in different ways (Koopman 2016; Van Groenestijn  2019). We did not include physical activity outcomes in a summary of findings table because they did not match prespecified outcome measurements (Table 1).

Open in table viewer
Table 1. An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD

Aerobic exercise programme compared to no aerobic exercise programme

Patient or population: people with NMD

Setting: the Netherlands

Intervention: aerobic exercise programme

Comparison: no aerobic exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no aerobic exercise programme

Risk with aerobic exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS)
assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 33.6 points

MD 1.8 points higher (2.9 lower to 6.5 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS)
assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.5 points

MD 0.1 points lower
(6.86 lower to 6.66 higher)

—

37
(1 RCT)

⊕⊝⊝⊝
Verylowa,c

—

Disease‐specific quality of life (ALSAQ‐40, questionnaire; lower = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean disease‐specific quality of life (ALSAQ‐40; lower = better quality of life) was 2.48 points monthly

MD 1.06 points monthly lower
(2.55 lower to 0.43 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS; higher = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean quality of life (SF‐36 PCS, questionnaire; higher = better quality of life) was –0.5 points monthly

MD 0.51 points monthly lower
(1.36 lower to 0.34 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS; higher = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean quality of life (SF‐36 MCS; higher = better QoL) was –0.09 points monthly

MD 0.23 points monthly higher
(0.64 lower to 1.1 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS, questionnaire) was 33.2 points

MD 1.1 points higher
(3.74 lower to 5.94 higher)

—

36
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 51.7 points

MD 1.9 points lower
(8.74 lower to 4.94 higher)

—

36
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 34.5 points

MD 1.3 points higher
(3.71 lower to 6.31 higher)

—

34
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 4.4 points lower
(12.18 lower to 3.38 higher)

—

34
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

ALSAQ‐40: Amyotrophic Lateral Sclerosis Assessment Questionnaire;CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; QoL: quality of life; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422111384441723949.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded once for imprecision associated with a wide CI.
cDowngraded twice for imprecision associated with a very wide CI.

Koopman 2016 reported using ITT analysis with no imputation of missing data, under the assumption that data were missing at random. In the absence of measurement of outcome data from all randomised participants, the analysis could be described as an available‐case analysis. The amount of missing data varied across different time points and outcome measures. The study reported that additional per‐protocol analyses of a subset of participants who completed more than 47 of 63 (75%) exercise sessions showed similar effect estimates as ITT analysis. In the ITT population, the median number of exercise sessions completed was 57 (range 8 to 63).

Van Groenestijn  2019 reported ITT analysis of randomised participants, irrespective of whether they received the allocated intervention and  completed follow‐up assessment. The assumptions made for ITT analysis were not specified. For outcomes at six months, including the ALS Assessment Questionnaire (ALSAQ‐40), the study reported additional per‐protocol analyses for a subset of participants who completed more than 75% of exercise sessions and attended the follow‐up assessment. The study reported that quality of life results favoured the exercise programme over usual care. Of the 18 participants who started the exercise programme, 11 (61%) attended at least 75% of the sessions, although one person died before follow‐up, for reasons considered unrelated to the exercise programme. The reasons reported for non‐attendance at exercise sessions included psychosocial problems, clavicular fracture due to a fall, time constraints, perceived lack of benefit, holiday, and an unrelated medical procedure.

Daily steps (count, activity monitor)

In Koopman 2016 after four months (after intervention): the MD was –197 steps per day (95% CI –2332.21 to 1938.21; 1 study, 36 participants; Analysis 4.1) in favour of no exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme.

After seven months (three months after intervention): the MD was –118 steps per day (95% CI –2010.18 to 1774.18; 1 study, 34 participants; Analysis 4.1) in favour of no exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme.

After 10 months (six months after intervention): the MD was 205 steps per day (95% CI –1618.68 to 2028.68; 1 study, 32 participants; Analysis 4.1) in favour of the exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme.

Disease‐specific quality of life (ALS Assessment Questionnaire)

In Van Groenestijn  2019 (slope over time, up to six months after intervention): the MD was –1.06 points monthly (95% CI –2.55 to 0.43; 1 study, 57 participants; low‐certainty evidence; Analysis 4.2) in favour of the exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Physical Component Score, questionnaire)

In Van Groenestijn  2019 (slope over time, up to six months after intervention): the MD was –0.51 points monthly (95% CI –1.36 to 0.34; 1 study, 57 participants; low‐certainty evidence; Analysis 4.3) in favour of no exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Mental Component Score. questionnaire)

In Van Groenestijn  2019 (slope over time, up to six months after intervention): the MD was 0.23 points monthly (95% CI –0.64 to 1.10; 1 study, 57 participants; low‐certainty evidence; Analysis 4.4) in favour of the exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Physical Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was 1.80 points (95% CI –2.90 to 6.50; 1 study, 37 participants; low‐certainty evidence; Analysis 4.5)  in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1).

After seven months (three months after intervention): the MD was 1.10 points (95% CI –3.74 to 5.94; 1 study, 36 participants;  low‐certainty evidence; Analysis 4.5) in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1).

After 10 months (six months after intervention): the MD was 1.30 points (95% CI –3.71 to 6.31; 1 study, 34 participants; low‐certainty evidence; Analysis 4.5) in favour of the exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1).

We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Mental Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was –0.10 points (95% CI –6.86 to 6.66; 1 study, 37 participants; very low‐certainty evidence; Analysis 4.6) in favour of no exercise programme but the CIs included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and twice for imprecision associated with a very wide CI.

After seven months (three months after intervention): the MD was –1.90 points (95% CI –8.74 to 4.94; 1 study, 36 participants; low‐certainty evidence; Analysis 4.6) in favour of no exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

After 10 months (six months after intervention): the MD was –4.40 points (95% CI –12.18 to 3.38; 1 study, 34 participants; low‐certainty evidence; Analysis 4.6) in favour of no exercise programme but the CI included the possibility of an effect favouring either the exercise programme or no exercise programme (Table 1). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Adverse events

Koopman 2016 and Van Groenestijn  2019 reported results for the training group but there were no comparative data available between groups for any type of adverse event.

Serious adverse events

There were no comparative data available between groups.

An aerobic exercise programme compared to CBT in people living with NMD

One RCT involving people with PPS contributed data for this comparison (Koopman 2016). We did not include physical activity outcomes in a summary of findings table because they did not match prespecified outcome measurements (Table 2).

Open in table viewer
Table 2. An aerobic exercise programme compared to CBT in people living with NMD

Aerobic exercise programme compared to CBT

Patient or population: people with NMD

Setting: the Netherlands

Intervention: aerobic exercise programme

Comparison: CBT

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with CBT

Risk with aerobic exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS) assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 34.9 points

MD 0.5 points higher (4.19 lower to 5.19 higher)

—

40
(1 RCT)

⊕⊝⊝⊝
Very lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 1.2 points higher (3.9 lower to 6.3 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS) was 36 points

MD 1.7 points lower (6.58 lower to 3.18 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 49.8 points

MD 3.2 points higher (3.03 lower to 9.43 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 36.1 points

MD 0.3 points lower (4.88 lower to 4.28 higher)

—

38
(1 RCT)

⊕⊝⊝⊝
Very lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 48 points

MD 1.3 points higher (6.34 lower to 8.94 higher)

—

38
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CBT: cognitive behavioural therapy; CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422112777684393547.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded twice for imprecision associated with a very wide CI.
cDowngraded once for imprecision associated with a wide CI.

Koopman 2016 reported using ITT analysis with no imputation of missing data, under the assumption that data were missing at random. In the ITT population, the median number of CBT sessions was seven (range zero to 12). According to the study protocol, the total number of sessions could vary from 12 to 16 depending on the modules taken. The median proportion of exercise sessions completed was 90% (range 13% to 100%).

Daily steps (count, activity monitor)

In Koopman 2016 after four months (after intervention): the MD was 449 steps per day (95% CI –1317.15 to 2215.15; 1 study, 36 participants; Analysis 5.1) in favour of the exercise programme but the CI included the possibility of an effect favouring either CBT or the exercise programme.

After seven months (three months after intervention): the MD was –11 steps per day (95% CI –1852.50 to 1830.50; 1 study, 35 participants; Analysis 5.1) in favour of CBT but the CI included the possibility of an effect favouring either CBT or the exercise programme.

After 10 months (six months after intervention): the MD was 137 steps per day (95% CI –1604.76 to 1878.76; 1 study, 35 participants; Analysis 5.1) in favour of the exercise programme but the CIs included the possibility of an effect favouring either CBT or the exercise programme.

Quality of life (36‐item Short Form Health Survey, Physical Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was 0.50 points (95% CI –4.19 to 5.19; 1 study, 40 participants; very low‐certainty evidence; Analysis 5.2) in favour of the exercise programme but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and twice for imprecision associated with a very wide CI.

After seven months (three months after intervention): the MD was –1.70 points (95% CI –6.58 to 3.18; 1 study, 37 participants; low‐certainty evidence; Analysis 5.2) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

After 10 months (six months after intervention): the MD was –0.30 points (95% CI –4.88 to 4.28; 1 study, 38 participants; very low‐certainty evidence; Analysis 5.2) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and twice for imprecision associated with a very wide CI.

Quality of life (36‐item Short Form Health Survey, Mental Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was 1.20 points (95% CI –3.90 to 6.30; 1 study, 40 participants;  low‐certainty evidence; Analysis 5.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2).

After seven months (three months after intervention): the MD was 3.20 points (95% CI –3.03 to 9.43; 1 study, 37 participants; low‐certainty evidence; Analysis 5.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2).

After 10 months (six months after intervention): the MD was 1.30 points (95% CI –6.34 to 8.94; 1 study, 38 participants; low‐certainty evidence; Analysis 5.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 2).

We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Adverse events

Koopman 2016 reported results for the training group but there were no comparative data available between groups for any type of adverse event.

Serious adverse events

There were no comparative data on serious adverse events available between groups.

CBT compared to no CBT in people living with NMD

One RCT involving people with PPS contributed data for this comparison (Koopman 2016). We did not include physical activity outcomes in a summary of findings table because they did not match prespecified outcome measurements (Table 3).

Open in table viewer
Table 3. CBT compared to no CBT in people living with NMD

CBT compared to no CBT

Patient or population: people with NMD

Setting: the Netherlands

Intervention: CBT

Comparison: no CBT

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no CBT

Risk with CBT

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 33.6 points

MD 1.3 points higher
(2.96 lower to 5.56 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.5 points

MD 1.1 points higher
(5.18 lower to 7.38 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS) was 33.2 points

MD 2.8 points higher
(2.07 lower to 7.67 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 51.7 points

MD 1.3 points higher
(4.42 lower to 7.02 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 34.5 points

MD 1.6 points higher
(3.22 lower to 6.42 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention(higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 3.1 points lower
(9.53 lower to 3.33 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference: NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422113426106228546.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded once for imprecision associated with a wide CI.

Koopman 2016 reported using ITT analysis with no imputation of missing data, under the assumption that data were missing at random. In the ITT population, the median number of CBT sessions was seven (range zero to 12) although between 12 and 16 sessions were prespecified in the study protocol.

Daily steps (count, activity monitor)

In Koopman 2016 after four months (after intervention): the MD was –646 steps per day (95% CI –2683.25 to 1391.25; 1 study, 36 participants; Analysis 6.1) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme.

After seven months (three months after intervention): the MD was –107 steps per day (95% CI –1773.16 to 1559.16; 1 study, 39 participants, Analysis 6.1) in favour of CBT but the CI included the possibility of an effect favouring either CBT or the exercise programme.

After 10 months (six months after intervention): the MD was 68 steps per day (95% CI –1672.87 to 1808.87; 1 study, 33 participants; Analysis 6.1) in favour of the exercise programme but the CI included the possibility of an effect favouring either CBT or the exercise programme.

Quality of life (36‐item Short Form Health Survey, Physical Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was 1.30 points (95% CI –2.96 to 5.56; 1 study, 41 participants; low‐certainty evidence; Analysis 6.2) in favour of CBT but the CI included the possibility of an effect favouring either CBT or no CBT (Table 3).

After seven months (three months after intervention): the MD was 2.80 points (95% CI –2.07 to 7.67; 1 study, 41 participants; low‐certainty evidence; Analysis 6.2) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or no CBT (Table 3).

After 10 months (six months after intervention): the MD was 1.60 points (95% CI –3.22 to 6.42; 1 study, 40 participants; low‐certainty evidence; Analysis 6.2) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or no CBT (Table 3).

We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Mental Component Score, questionnaire)

In Koopman 2016 after four months (recorded after intervention): the MD was 1.10 points (95% CI –5.18 to 7.38; 1 study, 41 participants; low‐certainty evidence; Analysis 6.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 3).

After seven months (three months after intervention): the MD was 1.30 points (95% CI –4.42 to 7.02; 1 study, 41 participants; low‐certainty evidence; Analysis 6.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 3).

After 10 months (six months after intervention): the MD was –3.10 points (95% CI –9.53 to 3.33; 1 study, 40 participants; low‐certainty evidence; Analysis 6.3) in favour of CBT but the CIs included the possibility of an effect favouring either CBT or the exercise programme (Table 3).

We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Adverse events

Koopman 2016  reported results for the CBT group but there were no comparative data available between groups for any type of adverse event.

Serious adverse events

The study did not report serious adverse events.

CBT with or without an exercise programme compared to no CBT and no exercise programme in people living with NMD

One study involving people with DM1 contributed data for this comparison (Okkersen 2018). We did not include physical activity outcomes in a summary of findings table because they did not match prespecified outcome measurements (Table 4).

Open in table viewer
Table 4. CBT with or without an exercise programme compared to no CBT and no exercise programme in people living with NMD

CBT with or without an exercise programme compared to no CBT and no exercise programme

Patient or population: people with NMD

Setting: Paris (France), Munich (Germany), Nijmegen (the Netherlands), and Newcastle (UK)

Intervention: CBT with or without an exercise programme

Comparison: no CBT and no exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no CBT and no exercise programme

Risk with CBT with or without an exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

NMD‐specific quality of life (INQoL) assessed with: final scores, during intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 5 months

The mean NMD‐specific quality of life (INQOL – quality of life domain, questionnaire) was 70.26 points

MD 1.05 points lower (10.44 lower to 8.34 higher)

—

218
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

NMD‐specific quality of life (INQoL) assessed with: final scores, unclear if during or after intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean NMD‐specific quality of life (INQoL) was 68.5 points

MD 1.67 points higher (7.64 lower to 10.98 higher)

—

222
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

NMD‐specific quality of life (INQoL) assessed with: final scores, after intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 16 months

The mean NMD‐specific quality of life (INQoL) was 69.32 points

MD 2.71 points higher (7.07 lower to 12.49 higher)

—

208
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events

Follow‐up: up to 14 days after the final study visit (16 months after baseline)

496 per 1000

506 per 1000
(397 to 650)

RR 1.02
(0.80 to 1.31)

255
(1 RCT)

⊕⊝⊝⊝
Very lowb,c

—

Serious adverse events

Follow‐up: up to 14 days after the final study visit (16 months after baseline)

118 per 1000

71 per 1000
(32 to 155)

RR 0.60
(0.27 to 1.31)

255
(1 RCT)

⊕⊝⊝⊝
Very lowb,c

—

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CBT: cognitive behavioural therapy; CI: confidence interval; INQoL: Individualized Neuromuscular Quality of Life; MD: mean difference; NMD: neuromuscular disease; RCT: randomised controlled trial; RR: risk ratio.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422113914345496441.

aDowngraded once for study limitations associated with a high risk of attrition bias.
bDowngraded once for indirectness because graded exercise therapy was not offered as part of intervention at all sites (variation in the intervention across different settings).
cDowngraded twice for imprecision associated with a very wide CI.

The study reported ITT analysis for the primary outcome analysis only (DM1‐Active‐c score). Other outcomes were reported using available‐case analysis. The mean number of CBT sessions completed was 10.7 (SD 3.3) of a maximum of 14 sessions. The mean number of face‐to‐face CBT sessions completed was 6.3 (SD 4.0). An activity module was indicated in all participants receiving CBT and provided to 112 people (94%). Forty‐two participants were referred to the optional exercise programme, of whom nine were non‐compliant due to lack of motivation or because they did not satisfy an aerobic exercise criterion. In total, 33/128 participants (26%) received the optional exercise programme, with two further withdrawals during the programme (one person was lost to follow‐up and the other stopped due to a malignancy).

Physical activity (interpreted as mean magnitude of ankle acceleration over 24 hours, activity monitor with Euclidian Norm Minus One metric)

After five months (during intervention): the MD was 2.08 unknown units (95% CI –1.00 to 5.16; 1 study, 154 participants; Analysis 7.1) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 10 months (unclear if during or after intervention): the MD was 1.90 unknown units (95% CI –0.97 to 4.77; 1 study, 164 participants; Analysis 7.1) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 16 months (six months after intervention): the MD was 1.26 unknown units (95% CI –2.09 to 4.61; 1 study, 139 participants; Analysis 7.1) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

Physical activity (interpreted as mean magnitude of ankle acceleration over five hours of highest activity, activity monitor with Euclidian Norm Minus One metric)

After five months (during intervention): the MD was 7.15 unknown units (95% CI –1.72 to 16.02; 1 study, 154 participants; Analysis 7.2) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 10 months (unclear if during or after intervention): the MD was 6.39 unknown units (95% CI –2.01 to 14.79; 1 study, 164 participants; Analysis 7.2) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 16 months (six months after intervention): the MD was 3.21 unknown units (95% CI –6.34 to 12.76; 1 study, 139 participants; Analysis 7.2) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

Physical activity (interpreted as average magnitude of ankle acceleration over five hours of lowest activity, activity monitor with Euclidian Norm Minus One metric)

After five months (during intervention): the MD was –0.02 unknown units (95% CI –0.36 to 0.32; 1 study, 154 participants; Analysis 7.3) in favour of no CBT and no exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 10 months (unclear if during or after intervention): the MD was 0.08 unknown units (95% CI –0.14 to 0.30; 1 study, 164 participants; Analysis 7.3) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

After 16 months (six months after intervention): the MD was 0.07 unknown units (95% CI –0.15 to 0.29; 1 study, 139 participants; Analysis 7.3) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme.

NMD‐specific quality of life (Individualized Neuromuscular Quality of life, INQoL)

After five months (during intervention): the MD was –1.05 points (95% CI –10.44 to 8.34; 1 study, 218 participants; low‐certainty evidence; Analysis 7.4) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme (Table 4).

After 10 months (unclear if during or after intervention): the MD was 1.67 points (95% CI –7.64 to 10.98; 1 study, 222 participants; low‐certainty evidence; Analysis 7.4) in favour of no CBT and no exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme (Table 4).

After 16 months (six months after intervention): the MD was 2.71 points (95% CI –7.07 to 12.49; 1 study, 208 participants; low‐certainty evidence; Analysis 7.4) in favour of no CBT and no exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme (Table 4).

We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition bias, and once for indirectness because graded exercise therapy was not offered as part of intervention at all sites (variation in the intervention across different sites).

Adverse events

Up to 14 days after the final assessment at 16 months: the RR was 1.02 (95% CI 0.80 to 1.31; 1 study, 255 participants; very low‐certainty evidence; Analysis 7.5) in favour of no CBT and no exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme (Table 4). We downgraded the certainty of the evidence once for indirectness because graded exercise therapy was not offered as part of intervention at all sites (variation in the intervention across different sites), and twice for imprecision associated with a very wide CI.

Serious adverse events

Up to 14 days after the final assessment at 16 months: the RR was 0.60 (95% CI 0.27 to 1.31; 1 study, 255 participants; very low‐certainty evidence; Analysis 7.6) in favour of CBT with/without an exercise programme but the CIs included the possibility of an effect favouring either CBT with/without an exercise programme or no CBT and no exercise programme (Table 4). We downgraded the certainty of the evidence once for indirectness because graded exercise therapy was not offered as part of intervention at all sites (variation in the intervention across different sites), and twice for imprecision associated with a very wide CI.

A functional exercise programme compared to a stretching exercise programme in people living with NMD

One RCT involving people with SBMA contributed data for this comparison (Shrader 2015). We included physical activity outcomes in a summary of findings table although the unit of measurement was unconfirmed. It is anticipated that the Actical accelerometer measured the physical activity count as a time‐based outcome.

If participants did not meet a minimal level of compliance, they would have been excluded from the study analysis (see Characteristics of included studies table). However, overall intervention compliance was reported as 88.8% with only one dropout due to non‐compliance (from the functional exercise programme). ITT analysis was not reported. The amount of missing data varied across different time outcome measures, suggesting available‐case analysis.

Physical activity (unspecified count per day, activity monitor)

After 12 weeks (during intervention): the MD was –8701 unconfirmed units (95% CI –38,293.30 to 20,891.30; 1 study, 43 participants; low‐certainty evidence; Analysis 8.1) in favour of the stretching exercise programme but the CIs included the possibility of an effect favouring either the stretching or functional exercise programme (summary of findings Table 3). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Physical Component Score, questionnaire)

After 12 weeks (unclear if during or after intervention): the MD was –1.10 points (95% CI –5.22 to 3.02; 1 study, 49 participants; low‐certainty evidence; Analysis 8.2) in favour of the stretching exercise programme but the CIs included the possibility of an effect favouring either the stretching or functional exercise programme (summary of findings Table 3). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

Quality of life (36‐item Short Form Health Survey, Mental Component Score, questionnaire)

After 12 weeks (unclear if during or after intervention): the MD was –1.10 points (95% CI –6.79 to 4.59; 1 study, 49 participants; low‐certainty evidence; Analysis 8.3) in favour of the stretching exercise programme but the CIs included the possibility of an effect favouring either the stretching or functional exercise programme (summary of findings Table 3). We downgraded the certainty of the evidence once for study limitations associated with a high risk of attrition and selection bias, and once for imprecision associated with a wide CI.

In the literature, we identified small (0.2 SD; 2.21), moderate (0.5 SD; 5.53) and large (0.8 SD; 8.85) distribution‐based MIDs that corresponded to the overall SF‐36 quality of life in people with NMDs before and after exercise intervention (Stefanetti 2020). However, we did not find an MID for the MD between interventions. Based on the effect estimates and low‐certainty evidence, we found that the functional exercise programme may have made little or no difference to PCS and MCS for quality of life when compared to the stretching exercise programme.

Quality of life (36‐item Short Form Health Survey, Vitality Component Score, questionnaire)

After 12 weeks (unclear if during or after intervention): the MD was –1.90 points (95% CI –13.14 to 9.34; 1 study, 49 participants; Analysis 8.4) in favour of the stretching exercise programme but the CIs included the possibility of an effect favouring either the stretching or functional exercise programme. We did not include this outcome in the summary of findings table as Vitality is one of eight Component Scores in the SF‐36 questionnaire.

Adverse events

The study did not report usable adverse event data.

Serious adverse events

The study did not report serious adverse event data.

We sought additional unpublished data relating to published outcome measures in three included RCTs, in accordance with the review protocol. In one study, we sought unpublished data on a subpopulation of 26 participants with motor neuron disease and other NMDs (Elsworth 2011). However, it was difficult to ascertain intervention allocation and so we did not analyse these data quantitatively. For Wallace 2019, we sought the first‐period data from the cross‐over trial but these were not available for the relevant published outcomes at the time of review development because the data had been analysed across both periods only. We did not analyse the combined first‐ and second‐period data quantitatively in this review, as per the protocol. We sought additional information in relation to one other study that had not been published as a full report, but it was not available at the time of preparing this review (White 2016).

Discussion

Summary of main results

Our review included 13 studies (795 randomised adults from 12 studies; number of participants unclear in one study). We did not perform meta‐analysis for any comparison because of differences in interventions and usual care. We also found considerable variation in how studies reported physical activity as an outcome measure. Six of the 13 included studies did not report results for physical activity outcomes, or the data were not usable in this review. The studies that reported physical activity measurement did not always clearly report ITT analysis or whether final assessments occurred during or after the completion of intervention. Based on prespecified measures, we included three comparisons in our summary of findings tables.

Two distinct studies of adults with DPN reported time spent physically active as time spent walking. One comparison of a physical activity programme (weight‐bearing) with no physical activity programme reported weekly duration of walking during and at the end of a one‐year intervention using a StepWatch ankle accelerometer. Based on the point estimate, intervention may have led to an important increase in physical activity per week; however, the 95% CI included the possibility of no difference or an effect in either direction at three months, six months, and 12 months. We assessed the evidence at low certainty. Another comparison of a brief, sensor‐based interactive exercise programme with no sensor‐based, interactive exercise programme reported duration of walking over 48 hours at the end of four weeks intervention using a t‐shirt embedded sensor. It was not possible to draw conclusions about the effectiveness of intervention based on the very low‐certainty evidence. It is unclear what minimal level of physical movement constituted objectively measured physical activity using an ankle and t‐shirt‐embedded accelerometer in these comparisons. Though an SMD could be calculated, we did not undertake such meta‐analysis for time spent physically active due to the variation between studies and low likelihood that statistical combination of these results would improve precision in a clinically meaningful way.

One study of adults with SBMA reported time spent physically active as a daily physical activity count. The comparison of a functional exercise programme with a stretching exercise programme involved measuring physical activity counts using an Actical accelerometer at the end of 12 weeks' intervention. It was not possible to draw conclusions about the effectiveness of the intervention (requiring compliance) from the low‐certainty evidence and unconfirmed measurement units.

The two exercise studies with prespecified measures of physical activity also reported on participants' quality of life. In one study, we found low‐certainty evidence that the sensor‐based interactive exercise programme may have made little or no difference to the MCS for quality of life when compared to no sensor‐based interactive exercise programme. However, we were unable to draw conclusions about the impact of intervention on the PCS for quality of life based on very low‐certainty evidence. In another study, we found that the functional exercise programme may have made little or no difference to the MCS or PCS for quality of life when compared to the stretching exercise programme, based on low‐certainty evidence. It is unclear whether the recording or recall period, or both, was during or after the intervention for final assessment of quality of life.

Although studies reported adverse events incompletely, we found no evidence of supported activity increasing the risk of serious adverse events or adverse events leading to study discontinuation. For the comparison of a physical activity programme (weight‐bearing) with no physical activity programme, one study specifically reported effect estimates for foot lesions and full‐thickness ulcers; this evidence included the possibility of no difference or a higher or lower risk with intervention.

Overall completeness and applicability of evidence

According to the WHO, physical inactivity is one of the main risk factors for noncommunicable disease mortality (WHO 2020b). For apparently healthy populations, increasing and maintaining regular physical activity is understood to be beneficial in terms of reducing all‐cause mortality risk, as well as for the primary and secondary prevention of chronic diseases, such as cardiovascular disease, diabetes mellitus, colon and breast cancer, osteoporosis, and depression, and risk factors such as hypertension and obesity. At a mechanistic level, routine physical activity has been associated with enhanced mental well‐being, reduced blood pressure, and improvement in glucose control and other biomarkers for inflammation and cardiovascular disease risk (Warburton 2006). By reducing the risk of type 2 diabetes, fewer people would be expected to develop peripheral neuropathy as a secondary complication, emphasising the need for effective prevention strategies.

Observational studies have shown that people with particular types of NMD are less physically active than apparently healthy controls, and have higher perceived barriers to becoming physically active (Aitkens 2005; Apabhai 2011; Heutinck 2017; McCrory 1998; Phillips 2009; Ramdharry 2017). This evidence further emphasises the need for effective strategies for physical activity participation in people living with NMD, in addition to prevention of secondary chronic diseases and complications.

In addressing the review question, we found fundamental evidence gaps for physical activity promotion among people living with NMD. Most included studies randomised only a minority of invited participants, which might suggest a broad recruitment strategy or strict inclusion criteria. However, the effectiveness of physical activity promotion could be affected by the initial study promotion and skewed by a low recruitment rate. Included studies involved people with nine of several hundred recognised types of NMD associated with varying severities of disease, disability, and impact on life expectancy. We also found some differences in the classification of NMDs. For example, one included study reported motor neuron disease separately from NMD, and other studies in the wider literature excluded DPN, unlike in this review. Four studies reported major comorbidities in some participants, which may have precluded their participation in other clinical trials. The restriction of this review to studies of people with NMD or neurological disorders including NMD, meant that we had a less heterogeneous review population than the general population but excluded evidence from other potentially relevant community interventions. Other Cochrane Reviews have aimed to address broader public health interventions for promoting physical activity but may not include people living with medical conditions such as NMD. Only one study in this review involved participants with different types of neurological disorders of whom a subgroup had NMD. We did not attempt to analyse the few data (unpublished) from this subgroup. If sufficient data had been available, the absence of stratification by condition may still have limited the usability or applicability of the evidence. All other included studies compared interventions in people with a specific NMD. The evidence related to adults in studies conducted over 2011 to 2019, largely facilitated through specialist healthcare settings in Western Europe. Twelve studies excluded children and adolescents and one study did not specify eligibility by age, although participants were aged over 60 years on average. The absence of evidence in children and adolescents, as well as in non‐ambulant people of any age, perhaps reflects a more general lack of clinical trials in these populations (Joseph 2015).

Also fundamental to the review question, we have shown how little research to date has addressed the effectiveness of promotional strategies for physical activity in people with NMD, as distinct from assessing the effects of exercise and compliance. As such, we have detailed promotional aspects of included RCTs in the study tables and main results of this review. We found that all included interventions and comparators related to the clinical care of people living with NMD. However, usual care varied across different conditions and clinical settings, for example, in national recommendations for physical therapy. We also found differences in study eligibility criteria in relation to baseline measures of physical activity. While all the studies intended to measure physical activity as an outcome, only one study clearly reported the aim of intervention being to promote physical activity. The main focus of 12 studies was to determine the effects of intervention on other aspects of health, fitness, and well‐being, such as fatigue, peak exercise capacity, or quality of life. We decided against meta‐analysis of any comparison, in part due to the different combinations of physical activity support and differences in usual care depending on the condition and setting.

We found that the studies reviewed tended to report physical activity as a secondary or exploratory outcome. We also found considerable variation in the way that physical activity was reported as an outcome measure. In the review, we applied a hierarchy of physical activity outcome measures for data extraction as per the protocol. We decided not to prioritise a specific time point because of the potential significance of different time points for different study populations, interventions, and comparators. Where specified, we also found variation in the timing of outcome measurement during or after intervention. We prioritised the reporting of measures of physical activity duration so that results could include participants irrespective of ambulatory status. However, nine studies included only ambulant participants and the other four studies did not specify baseline ambulatory status of participants but included ambulatory outcome measures. The number of daily steps taken was the most commonly reported measure of physical activity in included studies. This perhaps demonstrates the challenge of identifying a single primary outcome measure for physical activity across the spectrum of disability expected in NMD, including people who may use upper extremity physical activity to mobilise with assistive devices and wheelchairs.

In this review, we limited physical activity outcomes to self‐reported and objectively measured everyday physical activity as per the protocol. Assessment of the effects of interventions on performance measures such as timed walking distance and exercise capacity are addressed in other Cochrane Reviews. However, we still found considerable variation in the reporting of objectively measured physical activity outcomes and a lack of self‐reported physical activity outcomes. We did not attempt to meta‐analyse these measures as an overarching physical activity domain. Typically, studies each reported a single measure of physical activity, although accelerometers can collect data on multiple parameters (e.g. step count, energy expenditure, and duration of different intensities of physical activity). As a result, the chance of finding a statistically significant change may increase and there could be a risk of selective reporting if the chosen parameter is not prespecified in a protocol. To minimise the risk of selective reporting in the review process, we avoided prioritisation of shorter‐length follow‐up for outcomes reported at multiple time points. As highlighted above, we reported multiple time points for outcomes (as per the studies) in the absence of a single, appropriate standard for outcome reporting in people with different types of NMD. We also reported mental and physical component summary scores for quality of life (as per the studies) to help capture the impact of physical and communication‐based approaches to physical activity promotion. In our interpretation of the evidence, we did not identify or apply any anchor‐based MIDs for quality of life outcomes in people with NMD. We considered statistical, distribution‐based MIDs for overall quality of life before and after exercise interventions but these did not provide a MID for the MD between interventions (Stefanetti 2020). We also did not find any established MID for adverse events or physical activity outcomes within the review population. Most included studies reported adverse events incompletely and we found some differences in study exclusion criteria that may have affected the data for adverse events and other outcomes. For example, two studies monitored foot ulcers in participants with DPN (Lemaster 2008; Mueller 2013), whereas another study excluded people with active foot ulcers (Grewal 2015). In terms of time spent physically active, WHO guidance emphasises at least meeting recommended amounts of physical activity (WHO 2020a; WHO 2020b). As such, we interpreted the evidence on the premise that any increase in time spent physically active is considered important. More emphasis on qualitative and dichotomous change in self‐reported overall physical activity of people with NMD in future interventional studies might offer a pragmatic approach to capturing the multiple dimensions of physical activity participation as well as important change at an individual and population level.

Quality of the evidence

We assessed most included studies at high risk of bias due to incomplete outcome data. Missing data on physical activity outcomes was a particular concern. Studies highlighted technical problems with data retrieval from activity monitors and incomplete questionnaires. Although several studies reported an ITT analysis, the assumptions with this approach were not reported fully, which made it unclear how missing data were handled.

We are uncertain about whether the interventions promoted physical activity in people with NMD in terms of time spent physically active. We are also uncertain whether there was a benefit or harm of any intervention over another intervention or over usual care in terms of quality of life and adverse events. We assessed the certainty of the evidence as low to very low due to study limitations and because the results were imprecise or the comparison did not directly address the review question.

Potential biases in the review process

We conducted a thorough search for published RCTs and sought additional unpublished data in accordance with the review protocol. However, it is possible that further interventions for physical activity promotion may have been identified through searching other databases, such as CINAHL and AMED. Based on the completed search, we anticipate that we could have missed studies if the abstract did not explicitly state an aim to promote physical activity or omitted physical activity outcomes. Several studies reported pain or fatigue outcomes but we did not report these as adverse events unless studies reported them in this way despite our narrative reporting of foot lesions. In addition, we did not consider qualitative evidence in this review. However, one included study undertook interviews during and after intervention with a subset of participants (Wallace 2019). Another included study surveyed participants after intervention to understand better their perspective on the value of the exercise programme and their current exercise and skin‐monitoring habits (Mueller 2013).

Review authorship did not include people living with NMD although all review authors have clinical research experience of working with people who have neuromuscular conditions. Four review authors (JN, GG, KJ, and GR) had varying levels of involvement as personnel in one of two included RCTs in this review (Okkersen 2018; Wallace 2019), which might potentially bias the review process. Two review authors with study involvement (KJ and JN) contributed to the initial screening of full reports for inclusion, following the eligibility criteria previously published in the review protocol. We tried to minimise further potential biases in the review process through dual independent data extraction and dual independent risk of bias assessment by review authors not involved in the studies where possible. Due to logistical constraints, the first review author (KJ) contributed to data extraction and risk of bias assessment for one study, despite involvement. In addition, the first review author (KJ) and another review author (JN, FH, or GR) not involved in any included studies undertook GRADE assessments. 

Agreements and disagreements with other studies or reviews

Physical activity interventions have been widely investigated at a community level and in subpopulations living with certain health conditions. However, we found limited reporting on physical activity promotion in RCTs of people with NMD. Other types of non‐randomised trial design may be more pragmatic for longer‐term follow up of outcomes such as all‐cause mortality risk and the primary and secondary prevention of other chronic diseases.

Through the review process, we identified some challenges in defining and communicating 'promotion' as potentially distinct from 'increasing', 'supporting', or 'encouraging' physical activity alone. We observed that our search results appeared to focus more on 'exercise' over 'physical activity' and 'compliance' or 'adherence' over the acceptability of intervention to participants. Although we did not search for qualitative evidence in this review, we identified qualitative methods such as the evaluation of motivational interviewing that might help to better understand the influence on physical activity promotion of communication between study participants and those involved in study delivery and usual care (NCT03515356).

In 2005, one Cochrane Review focussed on a population that was sedentary but "free from pre‐existing medical condition or with no more than 10% of subjects with pre‐existing medical conditions that may limit participation in physical activity" (Foster 2005). Given these exclusions, the evidence is less directly applicable to people living with NMD and a broader group of people with chronic conditions, who may benefit from physical activity intervention. The review included no evidence for objectively measured physical activity. The review authors meta‐analysed 19 studies of self‐reported physical activity, measured as a continuous variable in a variety of ways (e.g. weekly energy expended, scoring on the PASE, and total hours of physical activity per week). These outcome measures might be considered too different to meta‐analyse, highlighting a lack of consistency in the way that physical activity is measured, reported, and analysed as an outcome. However, the authors found a short‐ to medium‐term positive effect of physical activity interventions in terms of an SMD in self‐reported physical activity. The review authors also meta‐analysed 10 studies of self‐reported physical activity measured as a dichotomous outcome; this evidence indicated moderate statistical heterogeneity. Combined with marked variation in the components of interventions and in the personnel supporting them, we found that it remains difficult to apply conclusions from a broader population that involves more participants. As noted by the review authors, translation of the evidence into practice is further complicated by a potential difference in the motivation of people who participate in research studies compared with those who do not.

Also published in 2005, another Cochrane Review examined physical activity interventions by setting (home or centre‐based) (Ashworth 2005). Focusing on an older adult population with certain medical conditions or risk factors, the review authors decided not to meta‐analyse the available evidence due to the heterogeneity of the studies. The primary outcome for this review was functional ability but secondary outcomes included measures of long‐term maintenance of physical activity. We also chose not to meta‐analyse physical activity outcomes on the basis of heterogeneity between studies. 

In 2013, several Cochrane Reviews investigated different methods of delivery for promoting physical activity, including face‐to‐face, web‐based, and remote interventions in apparently healthy study populations. When comparing face‐to‐face interventions with placebo or minimal intervention, review authors meta‐analysed self‐reported physical activity as a dichotomous and continuous outcome (Richards 2013a). The review authors cautioned that limited conclusions could be drawn about the effectiveness of components of interventions due to clinical and statistical heterogeneity despite some evidence in favour of face‐to‐face interventions. Similarly, caution was also advised when review authors compared remote and web 2.0 interventions with a placebo or minimal intervention for physical activity promotion although there was some evidence in favour of the use of technologies with support from a trained professional (Foster 2013). When face‐to‐face interventions were compared with remote and web 2.0 interventions for promoting physical activity, only one study met eligibility criteria and measured cardiorespiratory fitness, but included no measure of physical activity (Richards 2013b).

As we found in this review, most studies in other Cochrane Reviews of interventions for promoting physical activity have focused on adults. However, in 2013, one Cochrane Review assessed the effects of school‐based programmes for promoting physical activity and fitness in children and adolescents (Dobbins 2015). Review authors reported physical activity outcomes as dichotomous and continuous measures. As in this review, the duration of physical activity was a primary outcome either measured objectively or by self‐report. The proportion of participants engaging in moderate‐to‐vigorous physical activity was another primary outcome, again measured objectively or by self‐report. Also similar to this review, the authors decided not to meta‐analyse the available physical activity outcomes on the basis of variations in populations, interventions, and outcomes. The review authors emphasised caution in the interpretation of generally small effects supported by low‐certainty evidence.

A later Cochrane Review investigated community‐wide, multi‐component interventions for physical activity promotion (Baker 2015). Eligible study populations "must have been free living and not part of any institutionalised community, such as those who were mentally ill, the frail or bedridden elderly population, or those incarcerated in prison." Dichotomous and continuous measures of combined physical activity outcomes were reported narratively. The review authors reported that they did not conduct a meta‐analysis due to the heterogeneity in the populations and study designs. They assessed certainty of the evidence for combined physical activity outcomes as ranging from high to low. More recently, one Cochrane Review of workplace populations compared a pedometer intervention with minimal intervention and concluded that exercise can have a positive effect but authors did not perform meta‐analysis because of very high statistical heterogeneity (Freak‐Poli 2020). For comparisons with minimal and alternative physical activity interventions, the authors judged the certainty of the evidence for combined physical activity outcomes to be very low.

Several Cochrane Reviews have assessed the effects of exercise as a form of physical activity in people living with specific types of NMD. Among these reviews, two included the secondary outcome measurement of physical activity, although no data were available at that time (Bartels 2019; Quinlivan 2011). Quality of life appears to be more commonly reported than physical activity outcomes in reviews of RCTs that have involved physical activity. The reason for this might be that quality of life assessment has more standardised methods, such as the SF‐36. Quality of life is also used to determine quality‐adjusted life years, which can inform economic evaluations in healthcare decision‐making (along with measures of disability for determining disability‐adjusted life years). Use of published checklists and guidance has been proposed elsewhere to standardise assessment and reporting (Jimenez‐Moreno 2017; Slade 2016; Stoyanov 2015; The EQUATOR Network). In this review, quality of life was measured using disease‐specific and general population questionnaires. Uncertainty about whether the timing of the assessment period was during or after intervention complicated interpretation of changes in quality of life. For example, the recall period might be one week or four weeks, which could relate to the intervention period, the period when intervention stopped, or both. One review of psychosocial interventions designed to improve the quality of life and well‐being of people living with NMD found that most included studies reported a short‐term benefit of intervention but advised caution in the interpretation of the evidence because of widespread study limitations (Walklet 2016). For adverse event reporting, we found that studies in this review sometimes monitored the intervention arm only. However, in Okkersen 2018, we found very low‐certainty evidence that the intervention might reduce the risk of serious adverse events when compared with no intervention, which could carry implications for further research. It is unclear whether this finding, if confirmed, relates to closer monitoring alone or specific components of the intervention. Nevertheless, a reduction in all‐cause serious adverse events with intervention would be consistent with general population studies that support increased physical activity for reducing all‐cause mortality risk and the primary or secondary prevention of chronic diseases (Warburton 2006). In our review, one study undertook qualitative interviews of a subset of participants who identified that study personnel had helped to facilitate greater participation, but financial cost was a barrier to some people continuing exercise beyond the study (Wallace 2019). Another study surveyed participants after intervention and found that less than half of respondents continued to exercise three to seven days per week after approximately 15 months, despite the majority reporting that the exercise programme had been beneficial to them (Mueller 2013).

We excluded many studies involving physical activity that did not report physical activity outcomes. Of particular note, we excluded one study that reported social activity engagement but not physical activity (Veenhuizen 2019). One study awaiting classification assessed the effects of peer support on physical activity among adults with DPN (ChiCTR‐IPR‐15006127). Another study awaiting classification focused on physical activity among people who have muscular dystrophy and use wheelchairs (NCT00866112). In contrast, the studies in the current review focused on adults who were able to walk at study entry. However, Nary 2011 identified recruitment challenges due to strict eligibility criteria and a rare disease population, which may also be relevant to studies with low uptake in this review. While included studies largely focused on supported exercise, physical activity, and CBT, the ongoing studies we identified (mostly in DPN) investigate the use of insoles, aids, and devices, as well as supported exercise, and physical activity.

Flow diagram. RCT: randomised controlled trial.

Figures and Tables -
Figure 1

Flow diagram. RCT: randomised controlled trial.

Navigate to figure in ReviewOpen in new tab

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figures and Tables -
Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 1: Time spent walking (minutes per week, activity monitor)

Figures and Tables -
Analysis 1.1

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 1: Time spent walking (minutes per week, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 2: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 1.2

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 2: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 3: Steps taken in 30‐minute bouts (count, activity monitor)

Figures and Tables -
Analysis 1.3

Comparison 1: A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD, Outcome 3: Steps taken in 30‐minute bouts (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 2: A weight‐bearing (WB) exercise programme compared to a non‐WB exercise programme in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 2.1

Comparison 2: A weight‐bearing (WB) exercise programme compared to a non‐WB exercise programme in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 1: Time spent walking (hours per 48 hours, activity monitor)

Figures and Tables -
Analysis 3.1

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 1: Time spent walking (hours per 48 hours, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 2: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 3.2

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 2: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 3: Quality of life (Physical Component Score, questionnaire)

Figures and Tables -
Analysis 3.3

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 3: Quality of life (Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 4: Quality of life (Mental Component Score, questionnaire)

Figures and Tables -
Analysis 3.4

Comparison 3: A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD, Outcome 4: Quality of life (Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 4.1

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 2: Disease‐specific quality of life (ALSAQ‐40, questionnaire)

Figures and Tables -
Analysis 4.2

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 2: Disease‐specific quality of life (ALSAQ‐40, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 3: Quality of life (SF‐36 Physical Component Score, questionnaire)

Figures and Tables -
Analysis 4.3

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 3: Quality of life (SF‐36 Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 4: Quality of life (SF‐36 Mental Component Score, questionnaire)

Figures and Tables -
Analysis 4.4

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 4: Quality of life (SF‐36 Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 5: Quality of life (SF‐36 Physical Component Score, questionnaire)

Figures and Tables -
Analysis 4.5

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 5: Quality of life (SF‐36 Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 6: Quality of life (SF‐36 Mental Component Score, questionnaire)

Figures and Tables -
Analysis 4.6

Comparison 4: An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD, Outcome 6: Quality of life (SF‐36 Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 5.1

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Figures and Tables -
Analysis 5.2

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Figures and Tables -
Analysis 5.3

Comparison 5: An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Figures and Tables -
Analysis 6.1

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 1: Daily steps (count, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Figures and Tables -
Analysis 6.2

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Figures and Tables -
Analysis 6.3

Comparison 6: Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 1: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 24 hours, activity monitor with Euclidian Norm Minus One metric)

Figures and Tables -
Analysis 7.1

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 1: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 24 hours, activity monitor with Euclidian Norm Minus One metric)

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 2: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of highest activity, activity monitor with Euclidian Norm Minus One metric)

Figures and Tables -
Analysis 7.2

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 2: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of highest activity, activity monitor with Euclidian Norm Minus One metric)

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 3: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of lowest activity, activity monitor with Euclidian Norm Minus One metric)

Figures and Tables -
Analysis 7.3

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 3: Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of lowest activity, activity monitor with Euclidian Norm Minus One metric)

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 4: NMD‐specific quality of life (INQoL)

Figures and Tables -
Analysis 7.4

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 4: NMD‐specific quality of life (INQoL)

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 5: Adverse events

Figures and Tables -
Analysis 7.5

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 5: Adverse events

Navigate to figure in ReviewOpen in new tab

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 6: Serious adverse events

Figures and Tables -
Analysis 7.6

Comparison 7: Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD, Outcome 6: Serious adverse events

Navigate to figure in ReviewOpen in new tab

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 1: Physical activity (unspecified count per day, activity monitor)

Figures and Tables -
Analysis 8.1

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 1: Physical activity (unspecified count per day, activity monitor)

Navigate to figure in ReviewOpen in new tab

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Figures and Tables -
Analysis 8.2

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 2: Quality of life (SF‐36 Physical Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Figures and Tables -
Analysis 8.3

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 3: Quality of life (SF‐36 Mental Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 4: Quality of life (SF‐36 Vitality Component Score, questionnaire)

Figures and Tables -
Analysis 8.4

Comparison 8: A functional exercise programme compared to a stretching exercise programme in people living with NMD, Outcome 4: Quality of life (SF‐36 Vitality Component Score, questionnaire)

Navigate to figure in ReviewOpen in new tab
Summary of findings 1. A physical activity programme (weight‐bearing) compared to no physical activity programme in people living with NMD

Physical activity programme compared to no physical activity programme

Patient or population: people with NMD

Setting: primary care, endocrinology, or podiatry practices in central Missouri, USA

Intervention: physical activity programme (weight‐bearing)

Comparison: no physical activity programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no physical activity programme

Risk with physical activity programme

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, during intervention

Follow‐up: 3 months

The mean time spent walking was 526 minutes per week

MD 34 minutes more
(92.19 fewer to 160.19 more)

—

69
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, during intervention

Follow‐up: 6 months

The mean time spent walking was 511 minutes per week

MD 68 minutes more
(55.35 fewer to 191.35 more)

—

74
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Time spent walking (minutes per week, activity monitor)
assessed with: final scores, unclear if during or after intervention

Follow‐up: 12 months

The mean time spent walking was 500 minutes per week

MD 49 minutes more
(75.73 fewer to 173.73 more)

—

70
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life

—

—

—

—

—

Outcome not measured.

Adverse events/serious adverse events

—

—

—

—

—

No comparative data between groups available for all types of adverse event. However, the study reported rate ratios specifically for foot lesions and ulcers in participants with diabetic peripheral neuropathy.

Over 12 months, the reported rate ratio for all types of foot lesions (ignoring multiple lesions/episode) was 1.24 (95% CI 0.70 to 2.19; 1 study, 70 participants). Based on the point estimate, intervention may have led to higher rate of foot lesions; however, the 95% CI included the possibility of no difference or an effect in either direction.

Over 12 months, the reported rate ratio for all full‐thickness foot ulcers (ignoring multiple lesions/episode) was 0.96 (95% CI 0.38 to 2.42; 1 study, 70 participants). Based on the point estimate, intervention may have led to a lower rate of full‐thickness foot ulcers; however, the 95% CI included the possibility of no difference or an effect in either direction.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MD: mean difference; NMD: neuromuscular disease; RCT: randomised controlled trial.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422109324271426071.

aDowngraded once for study limitations associated with an unclear risk of bias in random sequence generation.
bDowngraded once for imprecision associated with a wide CI.

Figures and Tables -
Summary of findings 1. A physical activity programme (weight‐bearing) compared to no physical activity programme in people living with NMD
Navigate to table in Review
Summary of findings 2. A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD

Sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme

Patient or population: people with NMD

Setting: USA and Qatar

Intervention: sensor‐based exercise programme

Comparison: no sensor‐based exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no exercise programme

Risk with exercise programme

Time spent walking (hours per 48 hours, activity monitor) assessed with: final scores, after intervention

Follow‐up: 4 weeks

The mean time spent walking was 4.12 hours

MD 0.64 hours fewer (2.42 fewer to 1.13 more)

—

25
(1 RCT)

⊕⊝⊝⊝
Verylowa,b

—

Quality of life (SF‐12 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 weeks

The mean quality of life (SF‐12 PCS) was 40.12 points

MD 0.24 points higher (5.98 lower to 6.46 higher)

—

35
(1 RCT)

⊕⊝⊝⊝
Verylowa,b

—

Quality of life (SF‐12 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 weeks

The mean quality of life (SF‐12 MCS) was 47.3 points

MD 5.1 points higher (0.58 lower to 10.78 higher)

—

35
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐12: 12‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422110856458804259.

aDowngraded once for study limitations associated with a high risk of selective reporting and attrition bias.
bDowngraded twice for imprecision associated with a very wide CI.
cDowngraded once for imprecision associated with a wide CI.

Figures and Tables -
Summary of findings 2. A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD
Navigate to table in Review
Summary of findings 3. A functional programme compared to a stretching programme in people living with NMD

Functional programme compared to stretching programme

Patient or population: people with NMD

Setting: Bethesda, Maryland, USA

Intervention: functional exercise programme

Comparison: stretching exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with stretching programme

Risk with functional programme

Physical activity (unspecified count per day, activity monitor) assessed with: final scores, during intervention

Follow‐up: 12 weeks

The mean physical activity (unspecified counts per day, activity monitor) was 70,498 counts

MD 8701 counts lower (38,293.3 lower to 20,891.3 higher)

—

43
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, unclear if during or after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 12 weeks

The mean quality of life (SF‐36 PCS) was 34.1 points

MD 1.1 points lower
(5.22 lower to 3.02 higher)

—

49
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, unclear if during or after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 12 weeks

The mean quality of life (SF‐36 MCS) was 54.4 points

MD 1.1 points lower
(6.79 lower to 4.59 higher)

—

49
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No usable adverse event data available.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422114406971747153.

aDowngraded once for study limitations associated with a high risk of attrition bias.
bDowngraded once for imprecision associated with a wide CI.

Figures and Tables -
Summary of findings 3. A functional programme compared to a stretching programme in people living with NMD
Navigate to table in Review
Table 1. An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD

Aerobic exercise programme compared to no aerobic exercise programme

Patient or population: people with NMD

Setting: the Netherlands

Intervention: aerobic exercise programme

Comparison: no aerobic exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no aerobic exercise programme

Risk with aerobic exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS)
assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 33.6 points

MD 1.8 points higher (2.9 lower to 6.5 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS)
assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.5 points

MD 0.1 points lower
(6.86 lower to 6.66 higher)

—

37
(1 RCT)

⊕⊝⊝⊝
Verylowa,c

—

Disease‐specific quality of life (ALSAQ‐40, questionnaire; lower = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean disease‐specific quality of life (ALSAQ‐40; lower = better quality of life) was 2.48 points monthly

MD 1.06 points monthly lower
(2.55 lower to 0.43 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS; higher = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean quality of life (SF‐36 PCS, questionnaire; higher = better quality of life) was –0.5 points monthly

MD 0.51 points monthly lower
(1.36 lower to 0.34 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS; higher = better quality of life) assessed with: slope over time, after intervention

Follow‐up: 6 months

The mean quality of life (SF‐36 MCS; higher = better QoL) was –0.09 points monthly

MD 0.23 points monthly higher
(0.64 lower to 1.1 higher)

—

57
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS, questionnaire) was 33.2 points

MD 1.1 points higher
(3.74 lower to 5.94 higher)

—

36
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 51.7 points

MD 1.9 points lower
(8.74 lower to 4.94 higher)

—

36
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 34.5 points

MD 1.3 points higher
(3.71 lower to 6.31 higher)

—

34
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 4.4 points lower
(12.18 lower to 3.38 higher)

—

34
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

ALSAQ‐40: Amyotrophic Lateral Sclerosis Assessment Questionnaire;CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; QoL: quality of life; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422111384441723949.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded once for imprecision associated with a wide CI.
cDowngraded twice for imprecision associated with a very wide CI.

Figures and Tables -
Table 1. An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD
Navigate to table in Review
Table 2. An aerobic exercise programme compared to CBT in people living with NMD

Aerobic exercise programme compared to CBT

Patient or population: people with NMD

Setting: the Netherlands

Intervention: aerobic exercise programme

Comparison: CBT

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with CBT

Risk with aerobic exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS) assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 34.9 points

MD 0.5 points higher (4.19 lower to 5.19 higher)

—

40
(1 RCT)

⊕⊝⊝⊝
Very lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, recorded after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 1.2 points higher (3.9 lower to 6.3 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS) was 36 points

MD 1.7 points lower (6.58 lower to 3.18 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 49.8 points

MD 3.2 points higher (3.03 lower to 9.43 higher)

—

37
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 36.1 points

MD 0.3 points lower (4.88 lower to 4.28 higher)

—

38
(1 RCT)

⊕⊝⊝⊝
Very lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 48 points

MD 1.3 points higher (6.34 lower to 8.94 higher)

—

38
(1 RCT)

⊕⊕⊝⊝
Lowa,c

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CBT: cognitive behavioural therapy; CI: confidence interval; MCS: Mental Component Score; MD: mean difference; NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422112777684393547.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded twice for imprecision associated with a very wide CI.
cDowngraded once for imprecision associated with a wide CI.

Figures and Tables -
Table 2. An aerobic exercise programme compared to CBT in people living with NMD
Navigate to table in Review
Table 3. CBT compared to no CBT in people living with NMD

CBT compared to no CBT

Patient or population: people with NMD

Setting: the Netherlands

Intervention: CBT

Comparison: no CBT

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no CBT

Risk with CBT

Time spent physically active

—

—

—

—

—

Outcome not measured.

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 PCS) was 33.6 points

MD 1.3 points higher
(2.96 lower to 5.56 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 4 months

The mean quality of life (SF‐36 MCS) was 52.5 points

MD 1.1 points higher
(5.18 lower to 7.38 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 PCS) was 33.2 points

MD 2.8 points higher
(2.07 lower to 7.67 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 7 months

The mean quality of life (SF‐36 MCS) was 51.7 points

MD 1.3 points higher
(4.42 lower to 7.02 higher)

—

41
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 PCS) assessed with: final scores, after intervention (higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 PCS) was 34.5 points

MD 1.6 points higher
(3.22 lower to 6.42 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Quality of life (SF‐36 MCS) assessed with: final scores, after intervention(higher = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean quality of life (SF‐36 MCS) was 52.4 points

MD 3.1 points lower
(9.53 lower to 3.33 higher)

—

40
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events/serious adverse events

—

—

—

—

—

No comparative data available between groups for any type of adverse event.

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: confidence interval; MCS: Mental Component Score; MD: mean difference: NMD: neuromuscular disease; PCS: Physical Component Score; RCT: randomised controlled trial; SF‐36: 36‐item Short Form Health Survey.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422113426106228546.

aDowngraded once for study limitations associated with a high risk of attrition and selection bias.
bDowngraded once for imprecision associated with a wide CI.

Figures and Tables -
Table 3. CBT compared to no CBT in people living with NMD
Navigate to table in Review
Table 4. CBT with or without an exercise programme compared to no CBT and no exercise programme in people living with NMD

CBT with or without an exercise programme compared to no CBT and no exercise programme

Patient or population: people with NMD

Setting: Paris (France), Munich (Germany), Nijmegen (the Netherlands), and Newcastle (UK)

Intervention: CBT with or without an exercise programme

Comparison: no CBT and no exercise programme

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with no CBT and no exercise programme

Risk with CBT with or without an exercise programme

Time spent physically active

—

—

—

—

—

Outcome not measured.

NMD‐specific quality of life (INQoL) assessed with: final scores, during intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 5 months

The mean NMD‐specific quality of life (INQOL – quality of life domain, questionnaire) was 70.26 points

MD 1.05 points lower (10.44 lower to 8.34 higher)

—

218
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

NMD‐specific quality of life (INQoL) assessed with: final scores, unclear if during or after intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 10 months

The mean NMD‐specific quality of life (INQoL) was 68.5 points

MD 1.67 points higher (7.64 lower to 10.98 higher)

—

222
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

NMD‐specific quality of life (INQoL) assessed with: final scores, after intervention (lower = better quality of life)

Scale: 0–100

Follow‐up: 16 months

The mean NMD‐specific quality of life (INQoL) was 69.32 points

MD 2.71 points higher (7.07 lower to 12.49 higher)

—

208
(1 RCT)

⊕⊕⊝⊝
Lowa,b

—

Adverse events

Follow‐up: up to 14 days after the final study visit (16 months after baseline)

496 per 1000

506 per 1000
(397 to 650)

RR 1.02
(0.80 to 1.31)

255
(1 RCT)

⊕⊝⊝⊝
Very lowb,c

—

Serious adverse events

Follow‐up: up to 14 days after the final study visit (16 months after baseline)

118 per 1000

71 per 1000
(32 to 155)

RR 0.60
(0.27 to 1.31)

255
(1 RCT)

⊕⊝⊝⊝
Very lowb,c

—

*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CBT: cognitive behavioural therapy; CI: confidence interval; INQoL: Individualized Neuromuscular Quality of Life; MD: mean difference; NMD: neuromuscular disease; RCT: randomised controlled trial; RR: risk ratio.

GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

See interactive version of this table: gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_422113914345496441.

aDowngraded once for study limitations associated with a high risk of attrition bias.
bDowngraded once for indirectness because graded exercise therapy was not offered as part of intervention at all sites (variation in the intervention across different settings).
cDowngraded twice for imprecision associated with a very wide CI.

Figures and Tables -
Table 4. CBT with or without an exercise programme compared to no CBT and no exercise programme in people living with NMD
Navigate to table in Review
Comparison 1. A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1.1 Time spent walking (minutes per week, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.1.1 3 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.1.2 6 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.1.3 12 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.2 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.2.1 3 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.2.2 6 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.2.3 12 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.3 Steps taken in 30‐minute bouts (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.3.1 3 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.3.2 6 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

1.3.3 12 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 1. A physical activity (PA) programme (weight‐bearing) compared no PA programme in people living with NMD
Navigate to table in Review
Comparison 2. A weight‐bearing (WB) exercise programme compared to a non‐WB exercise programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

2.1 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 2. A weight‐bearing (WB) exercise programme compared to a non‐WB exercise programme in people living with NMD
Navigate to table in Review
Comparison 3. A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

3.1 Time spent walking (hours per 48 hours, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

3.2 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

3.3 Quality of life (Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

3.4 Quality of life (Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 3. A sensor‐based, interactive exercise programme compared to no sensor‐based, interactive exercise programme in people living with NMD
Navigate to table in Review
Comparison 4. An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

4.1 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.1.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.1.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.1.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.2 Disease‐specific quality of life (ALSAQ‐40, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.3 Quality of life (SF‐36 Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.4 Quality of life (SF‐36 Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.5 Quality of life (SF‐36 Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.5.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.5.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.5.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.6 Quality of life (SF‐36 Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.6.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.6.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

4.6.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 4. An aerobic exercise programme compared to no aerobic exercise programme in people living with NMD
Navigate to table in Review
Comparison 5. An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

5.1 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.1.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.1.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.1.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.2 Quality of life (SF‐36 Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.2.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.2.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.2.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.3 Quality of life (SF‐36 Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.3.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.3.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

5.3.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 5. An aerobic exercise training programme compared to cognitive behavioural therapy (CBT) in people living with NMD
Navigate to table in Review
Comparison 6. Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

6.1 Daily steps (count, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.1.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.1.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.1.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.2 Quality of life (SF‐36 Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.2.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.2.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.2.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.3 Quality of life (SF‐36 Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.3.1 4 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.3.2 7 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

6.3.3 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 6. Cognitive behavioural therapy (CBT) compared to no CBT in people living with NMD
Navigate to table in Review
Comparison 7. Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

7.1 Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 24 hours, activity monitor with Euclidian Norm Minus One metric) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.1.1 5 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.1.2 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.1.3 16 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.2 Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of highest activity, activity monitor with Euclidian Norm Minus One metric) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.2.1 5 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.2.2 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.2.3 16 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.3 Physical activity (unclear units – interpreted as mean magnitude of ankle acceleration over 5 hours of lowest activity, activity monitor with Euclidian Norm Minus One metric) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.3.1 5 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.3.2 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.3.3 16 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.4 NMD‐specific quality of life (INQoL) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.4.1 5 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.4.2 10 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.4.3 16 months

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

7.5 Adverse events Show forest plot

1

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected

7.6 Serious adverse events Show forest plot

1

Risk Ratio (M‐H, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 7. Cognitive behavioural therapy (CBT) with/without an exercise programme compared to no CBT and no exercise programme in people living with NMD
Navigate to table in Review
Comparison 8. A functional exercise programme compared to a stretching exercise programme in people living with NMD

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

8.1 Physical activity (unspecified count per day, activity monitor) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

8.2 Quality of life (SF‐36 Physical Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

8.3 Quality of life (SF‐36 Mental Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected

8.4 Quality of life (SF‐36 Vitality Component Score, questionnaire) Show forest plot

1

Mean Difference (IV, Random, 95% CI)

Totals not selected
Figures and Tables -
Comparison 8. A functional exercise programme compared to a stretching exercise programme in people living with NMD
Navigate to table in Review