Background
The treatment of behavioural and psychological symptoms of dementia (BPSD) is a key challenge in inpatient dementia care [
1,
2]. During the course of the disease nearly every patient suffers from an exacerbation of behavioural and psychological symptoms. Due to extreme caregiver and environmental burden, such phases often lead to admission into specialised Dementia Care Units (DCU) in old age psychiatric departments.
The application of antipsychotic medication is a standard treatment method for behavioural disturbances in inpatient dementia care. Owing to potential side effects of neuroleptics [
3], clinicians seek non-pharmacological approaches for the treatment of BPSD [
4]. A conceptual review [
5] and an observational, cross-sectional trial [
6] describe a direct link between physical inactivity and increased behavioural disturbances. Considering these findings, more physical activation in DCU is recommended, possibly leading to a reduced use of sedative medication and physical constraints. This non-pharmacological approach shows the positive effects of physical activation and may help to reduce critical side-effects of psychopharmacotherapy.
Unlike the good evidence for exercise as a key-factor in the prevention of dementia [
7‐
9], research focussing on the effects of exercise in dementia care is in its ‘infancy’ [
10]. While exercise trials focussing on functional performance and activities of daily living show promising results [
11,
12], there is a paucity of clinical trials investigating the effects of exercise on BPSD. While trials in long-term nursing home care reveal promising effects on some BPSD - especially on mood, agitation and circadian disturbances - there are only a few clinical studies in inpatient hospital dementia care with consistent results [
13]. The Cochrane Review by Forbes
et al. [
11] includes only one trial investigating the effects of exercise on BPSD [
14]. This 1-year nursing-home exercise trial, conducted with an overall participation-rate of 30 %, showed no effects on BPSD [
14]. As the low level of participation and consequently low level of physical activation could be a determining factor explaining only minimal effects on BPSD, future research should focus on approaches with better exercise adherence and higher levels of physical activation [
6].
In addition to physical activation, the organisation of non-pharmacological treatment in the course of a day is a key component of inpatient dementia care [
15]. Disease-related changes in circadian motor behaviour, that is, wandering, agitation or sundowning, affect almost every patient in advanced stages of dementia [
16]. In this context, sundowning is denoted as an increase of BPSD in the late afternoon and early evening hours. As a low level of daytime activity is linked to a higher fragmentation of circadian rest-activity cycles [
17], structured exercise programmes may help to synchronise and reduce circadian rhythm disturbances [
16,
18].
Taking into account the assumed positive effects of physical exercise on BPSD and the necessity of providing day structure through activities in dementia care, our research group has conducted a pilot study. The aim of this pilot study was to investigate the feasibility of the planned day-structuring physical activation programme in a hospital setting. Eighteen patients suffering from acute exacerbation of behavioural symptoms in dementia were included. These patients received a 2-week complex day-structuring exercise programme in addition to their usual treatment - the same intervention as planned for the randomised controlled trial (RCT). Two thirds of these patients accomplished more than 180 minutes of exercise per week. In this pilot-study we found reductions in some BPSD; furthermore we found stabilisation of day-structure and a decrease in caregiver burden. Based on these preliminary findings, our planned RCT trial aims to investigate the effects of a day-structuring exercise programme in inpatient hospital dementia care.
Objectives
The primary objective of this trial is to investigate the effects of a day-structuring exercise programme on BPSD in inpatient hospital dementia care. The promising results of our pilot-study regarding the feasibility and effects of the
exercise-carrousel on BPSD in hospital old age psychiatry care, led to the following study-hypothesis:
The intervention group, carrying out a 2-day-structuring exercise intervention in addition to treatment as usual (TAU), is hypothesized to show reduced BPSD at follow-up. In contrast, the control group, receiving social stimulation in addition to TAU, is not expected to show such changes.
In addition to investigating the putative effects of the day-structuring exercise on BPSD, the following secondary-endpoint parameters are investigated: caregiver burden, total dose and frequency of application of antipsychotic and sedative medication, circadian rhythms of motor behaviour, saliva cortisol-levels and brain-derived neurotrophic factor.
Discussion
The treatment of BPSD is challenging as patients with dementia are prone to suffer from severe side effects of pharmacotherapy [
47]. Furthermore, psychopharmacotherapy often has no significant effects on BPSD or may even worsen behavioural symptoms in the patients [
48]. Therefore, non-pharmacological approaches for the treatment of BPSD are essential. Establishing non-pharmacological treatment and especially physical activation therefore represent a fruitful and promising approach [
5]. Hence, we plan to implement a hospital-based day-structuring exercise programme for demented patients suffering from exacerbated behavioural symptoms.
To our knowledge, this is the first randomised controlled trial in inpatient hospital dementia care investigating the effects of multiple exercise sessions in the course of a day. Whereas former intervention trials and routine inpatient dementia care provide exercise sessions in a rigid time schedule for approximately one hour per session [
14], this trial offers four flexible short-time exercise sessions throughout a day. In a hospital setting, ward routines and intra-daily variations of the patients’ mood or behaviour are most reasons for non-participation in exercise sessions. If a patient has no motivation to take part in an exercise session, or is hindered due to nursing or medical routines, there is in most cases no further chance to take part in guided physical activation throughout the same day. In order to circumvent these problems, the
exercise-carrousel schedule offers multiple low-threshold exercise sessions during a day. From our clinical experience, this exercise programme should increase the patients’ willingness for participation. Therefore, we expect a better exercise-adherence compared to routine care and former trials focussing on physical activation in dementia [
14]. In accordance with Christofoletti
et al. [
6], this programme represents an innovative strategy to increase exercise-adherence in dementia patients, thereby possibly leading to decreased BPSD.
Additionally, we are also interested in the influences on circadian rhythm disturbances. Disrupted circadian motor behaviour is often seen in demented patients; that is, fragmentation of rest-activity cycles or wandering-phenomenon. Actually, there are recommendations to perform exercise sessions in the morning hours [
49]. As behavioural disturbances more often occur in the afternoon and evening hours, Rimmer and Smith [
49] were more concerned with behavioural disturbances that may intervene with exercise sessions at that time. In most hospital and nursing home settings, physical and occupational therapy is therefore centred in the morning hours. Consequently, in daily routines for dementia patients there is often a lack of guided activities in the afternoon and evening hours. National guidelines and international expert panels demand the provision of day structure as non-pharmacological treatment of circadian rhythm disruptions [
15,
50,
51]. To our knowledge, no trial focussing on the provision of day structure through physical activation in dementia care has been conducted so far [
15,
51].
We plan a trial that provides structured rest- and activity-sequences in the course of a day and investigates the synchronising effects on circadian rhythm disturbances. Intervention days are planned on three non-consecutive days a week in addition to usual care with daily co-therapies, nursing and medical applications. To our knowledge there is no evidence, and neither are there recommendations, for daily structured activities as treatment of circadian rhythm disturbances. Within the thrice weekly exercise schedule, we expect a day of rest to be necessary for the patients and for the evaluation of exercise effects on circadian disruptions on intervention and non-intervention-days. Based on the experiences from our pilot project, a day of rest between the intervention days seems to be useful for these older population patients.
In order to appropriately assess behavioural symptoms, circadian rhythm disruptions and the effects of the day-structuring exercise schedule, we suggest an innovative combination of body-fixed motion sensors, well-established instruments for the rating of BPSD (NPI, CMAI) and physiological parameters (saliva-cortisol, BDNF). The upcoming complementary use of hybrid motion sensors is an especially important methodological advancement evaluating the association between disrupted motor activity and behavioural symptoms in dementia [
46]. Wrist-worn accelerometers, logging and analysing the periodicity of limb movements on the basis of activity counts, are usually used to investigate movement behaviour and circadian rhythm disruptions [
52]. Sensors logging the periodicity of body postures, may be superior to activity counts in analysing these symptoms. The results of our planned trial may help to implement and improve the clinical use of body-fixed sensors in assessing circadian and behavioural symptoms in patients suffering from dementia [
46,
53].
Considering the methodological issues, a major challenge in conducting clinical trials is to prevent selection bias and achieve allocation concealment for all investigators throughout the whole trial. Therefore, all investigators involved in the assessment of patients (BPSD, motor behaviour, cortisol, BDNF) and the staff of the ward are blinded to group allocation. Trial design as well as the definition and recording of intervention adherence are important features when conducting exercise trials in patients with dementia. In accordance to Edwards
et al. [
44], we will rate exercise adherence regarding time, intensity and degree of exercise performance. We do not plan to use standardised self-reported rating scales to assess exercise exertion (for example, via Borg score [
54]), as these scores are not reliable in patients with advanced dementia [
55]. Apart from the application of body-fixed motion sensors, which provide reliable and objective information about motor behaviour during participation in exercise sessions, there is no further monitoring of exercise exertion planned for this trial. On the one hand, the monitoring of heart rate and blood pressure during exercise participation has not been feasible during the pilot project. These patients were often distracted and sometimes refused to wear the monitoring devices. On the other hand, we experienced small motor sensors fixed on the lower back to be out of the patients’ focus and therefore to be well accepted. The lack of instrumental monitoring of exercise exertion represents a limitation of this trial. Because this pragmatic method favours the provision of best possible low-threshold exercise participation, with the lowest possible obstruction before and during an exercise session, this approach has possible methodological limitations, but it is also a simple, transferable approach for routine care. Furthermore, the 2-week intervention phase only allows investigation into the short-term effects of the exercise programme. Taking into account the regular stay length of 4 to 5 weeks in inpatient old age psychiatric hospital care, the design of this trial schedules a 2-week intervention phase, due to 1 week after admission for familiarisation with the setting and delirium exclusion, and 3 days for pre- and post-assessment, respectively. Apart from the usually recommended long-term exercise schedules, the investigation of an intensive short-term
exercise-carrousel programme represents, from our point of view, an appropriate and easily transferable approach at this stage of care. Obviously, this intensive exercise intervention programme requires a greater use of therapy resources and may negatively interact with other hospital routines.
Physical activation is a key part of non-pharmacological treatment, but despite the clinicians’ advice to apply non-pharmacological methods for the treatment of BPSD [
4], there are no recommendations and guidelines for inpatient hospital dementia care. The results of our study may help to increase the evidence in the field of exercise science and non-pharmacological treatment. The
exercise-carrousel approach can further advance non-pharmacological treatment, leading to potential worthwhile developments for patients, medical staff and caregivers in dementia care.
Trial status
Recruitment began in April 2015
Competing interests
Tim Fleiner, Wiebren Zijlstra, Hannah Dauth and Peter Haussermann declare that they have no competing interests.
Authors’ contributions
All authors were involved in planning and writing the exercise carrousel study protocol, all have read and approved the final manuscript. Furthermore, TF is responsible for the implementation and performance of the exercise intervention; WZ is responsible for the implementation and analysis of motion sensors; HD is responsible for saliva and blood probes; PH is responsible for adequate in- and exclusion of patients. All authors read and approved the final manuscript.