Feedback from physical activity monitors to enhance amount of physical activity in adults—a protocol for a systematic review and meta-analysis

Physical activity is defined as any bodily movement that requires energy expenditure [1]. According to the World Health Organization (WHO), one in four adults does not meet the recommendations for physical activity and physical inactivity is estimated to be responsible for 9% of the world’s premature deaths [2]. Globally, inactivity costs 54 billion USD in direct health care cost every year [1]. Furthermore, strong evidence exists that physically active people will have reduced rates of all-cause mortality, coronary heart disease, high blood pressure, stroke, metabolic syndrome, type 2 diabetes, breast cancer, colon cancer, depression, and falls [2]. A global plan from the WHO focused on reducing the relative prevalence of physical inactivity by 15% in 2030. In order to progress towards this goal WHO states that effective community-based interventions are crucial [1].

Physical activity monitors (PAMs) have been found to be an effective facilitator to increase physical activity [3] and reduce sedentary time [4] and have a positive influence on weight loss among individuals in weight loss programs [5]. Furthermore, the feedback from PAMs are suggested to motivate behavioral change [6], and therefore, PAM-based interventions hold great potential for increasing physical activity. PAMs were originally designed to quantify the level of physical activity [4] and have been used in research since the 1960s [5]. Due to the technological development, the accessibility ,and relatively low cost of PAMs, they are incorporated in many devices and could potentially be one of the keys to effective large-scale community-based interventions and thereby increasing physical activity on a global level [1]. Earlier systematic reviews and meta-analyses have shown that pedometers can increase physical activity among outpatient adults [7] and reduce sedentary time in adults [4]. However, there are two reasons for performing a new systematic review and meta-analysis on the effect of PAM-based interventions on physical activity. Firstly, since recent systematic reviews were conducted, technology has improved rapidly and many of today’s PAMs can monitor number of steps taken and provide real-time feedback on several measures. As a result of this, the amount of scientific papers investigating PAMs impact on physical activity has increased equivalent. Secondly, with the increasing body of evidence, it might be possible to do further subgroup analyses investigating which study and participant characteristics that can explain effect size heterogeneity in the literature.

An update of the body of evidence regarding PAMs effect on daily physical activity, moderate to vigorous physical activity, and time spent sedentary with an investigation of the intervention, study, and participants characteristics is deemed necessary in order to progress towards WHO’s declared goal to reduce the prevalence of physical inactivity [1], to inform future studies, and to inform clinical guidelines.

The objective of this systematic review and meta-analysis is to estimate the effect on daily physical activity, moderate to vigorous physical activity, and time spent sedentary when using feedback from PAMs in interventions, compared to control interventions where the participants do not receive feedback from PAMs, in participants aged 18–64. Subsequently, the impact of intervention, study, and participants characteristics on the effect of PAM will be investigated.

Reported adverse events and drop-outs will be extracted.

Data will be extracted at post-intervention and follow-up.

An electronic search for eligible studies in the electronic databases MEDLINE, EMBASE, SPORTDiscus, CINAHL, and Cochrane Central Register of Controlled Trials (CENTRAL) was conducted in July 2018, according to the PROSPERO protocol.

The search matrix illustrated in Table 1 will consist of a combination of relevant keywords and MeSH/Thesaurus terms for (1) PAMs and (2) study design.

No restrictions on language or publication-time will be applied. If relevant studies are identified in other language than Danish, English, Swedish, Norwegian, and German, a relevant translator will be contacted. The authors of unobtainable studies or studies with missing data will be contacted.

The detailed search strategy for each database can be found in the Additional file 1.

Searching references of eligible studies and relevant journals (pearl growing) will be conducted independently by two reviewers (CK and VW) in order to include relevant articles not captured by the search strings. The database ClinicalTrials.​gov will be searched to locate ongoing studies.

The technology platform, Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at www.​covidence.​org), will be used to import citations from the literature searches, screening of title and abstracts, screening of full text, assessing risk of bias in included studies, and extracting the data. The analyses will be conducted in Stata Statistical Software, version 15.

The selection of studies will be done by merging search results from the databases and removing duplicates. Two review authors (CK and VW) will then independently screen titles and abstracts. Articles judged as eligible by at least one of the reviewers will be screened in independently full-text by the same review authors (CK and VW). Any inconsistencies between authors will be discussed and solved with consultation of a third author (RTL).

Data on the following items will be extracted from all included studies.

Source: Study ID, protocol ID, review author, citation details and contact details.

Two review authors (VW and CK) will independently assess the risk of bias in included studies, using the RoB 2.0 tool [11]. Disagreement between reviews authors will be solved by including a third reviewer (RTL). The risk of bias assessment for each study will be presented using a table with judgment and support for judgment.

If a study has more than one intervention group relevant for inclusion in the systematic review, the intervention groups will be included as two separate studies and the control group will be separated.

Treatment effect, measured as continuous data, will be expressed as mean difference with 95% confidence intervals for outcomes measured with the same outcome measurement instrument or as standardized mean difference with 95% confidence intervals, when different measurement instruments are used in included the studies. Dichotomous outcomes, such as adverse events and drop-outs, will be analyzed and expressed as risk ratios with 95% confidence intervals.

The heterogeneity of the results will be examined using Cochrane Q test and quantified as I² values and the between study variance τ². Assessment of small study bias will be done by calculating an Egger’s test score and illustrated with a funnel plot. If small study bias is found, by a positive Egger’s test, a metatrim analysis will be conducted and an adjusted effect size will be calculated. For all statistical analysis, an alpha level of 0.05 will be considered statistically significant.

We will explore heterogeneity by conducting sub-group analyses and stratified analyses on the following nominal variables:

We will explore heterogeneity on continuous data by performing univariate meta-regressions on the following variables:

The impact of risk of bias will be analyzed on all outcomes by stratifying on overall risk of bias, defined by the RoB 2.0 tool (low/some concerns/high) [11].

We will create a summary of findings table with effect sizes on all outcomes. Two reviewers (RTL and JC) will independently rate the quality of the body of evidence using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach including downgrading and upgrading rating, which will be described in the footnotes in the table [12, 13].

This systematic review and meta-analysis will be conducted according to the current recommendations from the Cochrane Handbook and reported according to the PRISMA statement [16, 17]. The methodological quality of this review and the expected large body of evidence should be used to generate trustworthy recommendations regarding the use of PAM-based physical activity interventions. The use of PAMs has already been documented [4, 7], but as innovative health care solutions and health care technology are rapidly growing fields, this systematic review will provide an updated meta-analysis of the use of PAM in interventions to increase physical activity.