Sedentary Behaviour in Hospitalised Older People: A Scoping Review

Background: Sedentary behaviour (SB) can delay hospitalised older adults’ recovery from acute illness and injuries. Currently, there is no synthesis of evidence on SB among hospitalised older people. This scoping review aimed to identify and map existing literature on key aspects of SB among hospitalised older adults, including the prevalence, measurement and intervention strategies for SB and sedentary behaviour bouts (SBBs) as well as healthcare professionals, patients and carers’ perspectives on interventions. Methods and analysis: Several electronic databases were searched between January 2001 and September 2020. The Joanna Briggs Institute (JBI) framework was used to conduct this scoping review. Results: Out of 1824 articles, 21 were included comprising 16 observational studies, 3 randomised controlled trials, 1 comparative study, and 1 phase-1 dose-response study. The sample size ranged from 13 to 393, with all 1435 participants community-dwelling before hospitalisation. Only two studies focused on measuring SB and SBBs as a primary outcome, with others (n = 19) reporting SB and SBB as a sub-set of physical activity (PA). Older adults spent an average of 86.5%/day (20.8 h) sedentary. Most studies (n = 15 out of 21) measured SB and SBB using objective tools. Conclusion: Hospitalised older people spent most of their waking hours sedentary. Studies explicitly focused on SB and SBB are lacking, and the perspectives of patients, carers and healthcare professionals are not clarified. Future hospital-based studies should focus on interventions to reduce SB and SBB, and the perspectives of healthcare professionals, patients and carers’ taken into account.


Introduction
Hospitalised older people are at risk of developing iatrogenic conditions independent of the primary reason for admission [1]. Some of these include functional decline and increased frailty [2], resulting in hospital-acquired disability (HAD) [3]. One main reason for HAD is sedentary behaviour (SB) [4], which has been shown to result in increased loss of muscle mass and function [5,6]. The effects of SB during hospital admission persist post-discharge [7], with previously active older people more sedentary at home post-discharge as a result of HAD [8]. In hospitals, when older people conduct activities such as sitting, lying and reclining whilst awake, they are said to be exhibiting SB [9]. Sedentary behaviour bout (SBB) [9] on the other hand relates to the extent people are sedentary and is said to be occurring when less than 100 steps/min are made for a period of one minute or more. Being aware of the extent of SB and SBB is an essential first step when highlighting the need for effective interventions aimed at reducing SB and SBB in hospitalised older people.
Health professionals are embracing campaigns such as #EndPJparalysis with the hope of tackling the epidemic of sedentary behaviour (SB) in hospitals, encouraging patients to exchange hospital gowns or pyjamas ("PJ") for street clothes and to get moving [10]. Despite the roll-out of such campaigns, no reviews have explored current evidence related to SB in hospitalised older adults. Describing the state of current evidence on SB could enable researchers and practitioners to develop strategies to prevent, decrease and alleviate the undesired consequences of SB for hospitalised older adults and their caregivers.
This scoping review aimed to identify and map (classify, organise and synthesise) relevant peer-reviewed articles on SB and SBB in hospitalised older adults. A scoping review methodology was used due to the broad nature of the topic and to enable the identification of currently available evidence to guide the development of future research studies.

Methodology
According to our previously published protocol paper [11], we followed the Joanna Briggs Institute (JBI) framework for scoping reviews [12]. The findings of this review are reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis extension for Scoping Reviews (PRISMA-ScR) checklist [13] (Supplementary Materials S1).

Population
This review included studies involving participants aged ≥65 years (or mean age of ≥65 years).

Concept
Studies considering various characteristics of SB and SBB including prevalence, assessment tools (objective or subjective), interventions as well as perspectives of healthcare professionals, patients and carers regarding intervention, were included.

Context
Settings considered were in-patient hospital and rehabilitation. Community settings, nursing homes or residential aged care were excluded. Studies from a broad range of research methodologies covering cross-sectional, observational, interventional, qualitative and mixed methods studies were included.
Articles from January 2001 to September 2020 were searched using Medical Subject Heading (MeSH) terms and keywords, such as sedentary behaviour, hospital and aged (Supplementary Materials S2). The following electronic databases were searched: PEDro, PubMed, Web of Science, MEDLINE Ovid, Cochrane, Scopus, Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsychInfo, Embase, Ageline, JBI and clinical trials registries for interventions.

Study Selection
Articles were screened and included using a two-step approach. Two independent reviewers (U.J. and L.Y.) screened the abstracts. Then, the full text of relevant articles was reviewed for inclusion (Supplementary Materials S3). Disagreements were clarified by discussing the differences, and a third reviewer (J.D.) was invited if a consensus could not be reached.

Data Extraction
Two independent reviewers (U.J. and L.Y.) extracted data using a data extraction tool developed according to the aims of the study (Supplementary Materials S4). Data extracted included: number of participants, duration of SB and SBBs, assessment of SB and SBBs, study design, study population and intervention details.

Presentation of the Results/Data Mapping
Findings are presented as a narrative summary and in a tabular format.

SB and SBB Measurements
Only two studies measured SB and SBB as a primary outcome [24,34] (Table 1). Both studies focused on patients with stroke with varying sample sizes (i.e., n = 19 and n = 104) and used different methods to investigate SB and SBB. Barrett et al. [34] met the current definition of SB by excluding sleep time, measuring SB and SBB over seven days but with a smaller sample size of 19. The larger study [n = 104] by Sjoholm et al. [24], explored SB and SBB using behavioural mapping for 9 h (8 am to 5 pm) on one day.
The only wearable devices validated for the measurement of SB and SBB in hospital are ActivPAL and PAL2 [35,36]. ActivPAL was the most commonly (6 out of 8 studies) used of the two wearable devices. Six studies utilised ActivPAL and PAL2 as the sole measurement strategy [15][16][17][18]23,27], or in combination with another wearable device called Tractivity [30], or in combination with behavioural mapping [14].
Three studies reported on the acceptability of wearable devices (Table 2), two relating to the PAL 2 sensor (attached to thigh) [14,18] and one to the Usense sensor (attached to lower back) [26]. The two studies with PAL2 worn on the lower thigh reported acceptability of 81% [18] and 90%, respectively [14]. Primary reasons for declining or removing a sensor in all three studies were discomfort [14,18,26] with other reasons such as not wanting to be monitored, lack of interest, inconvenience, anxiety and interference with toileting/indwelling catheter reported in one study [18].

Proportion of SB and SBB
Overall, the mean duration of SB from all the studies was 86.5%/day (20.8 h)

Studies with SB and/or SBB as Primary Outcome
Sjoholm et al. revealed that participants were sedentary for 74% (6.2 h) over nine hours between 8 am and 5 pm [24] (Table 1). In this study, SBB of >1 h was noted among 54% of participants; the average SBB was 38 min (10-490 min), and the mean number of SBBs per patient was 9.9.
The smaller study by Barrett et al. [34] that measured SB over seven days reported participants were sedentary for 85.6% of the observed duration (12. 8 h/day). There was no significant difference in participants SB between admission and discharge, but an increase in participants SB on the weekend to 89.8% (13.5 h/day). Including sleep periods, this study reported average SBBs of 5.31 h (318.8 ± 219.9 min) [34].
A study among 58 patients with stroke using ActivPAL [23] demonstrated that SBB increased by 4.32 min daily over seven days among those with severe stroke and older participants. The study included sleep time and excluded patients with dementia

Perspectives of Stakeholders about Interventions
Despite our comprehensive search strategy and broad inclusion criteria, we did not identify studies reporting stakeholders (healthcare practitioners), end-users (patients and their formal or informal carers) perspectives on interventions to reduce SB or SBB within hospital settings.

Intervention Strategies for SB and SBB
No intervention studies specifically focused on reducing SB and SBB as the primary aim.

Discussion
Findings from this study revealed that while hospitalised older people spent most of their waking hours sedentary, the prevalence of SB and SBB cannot be established due to poor adherence to current definitions of SB and SBB and a lack of consensus on how best to measure SB and SBB within hospital and in-patient rehabilitation settings. Additionally, no intervention studies have focused explicitly on reducing SB or SBB as primary aims, and no study has sought the perspectives, attitudes and experiences of patients, carers and health professionals, especially crucial if translatable interventions are to be developed. Thus, the findings from this review suggest there remains a need for developing hospital-based interventions to reduce SB and SBBs by taking into account the perspectives of healthcare professionals, older patients and their carers.
Only two studies focused primarily on SB and SBBs, revealing a gap in knowledge on the prevalence of and interventions for SB and SBBs among hospitalised older adults. Possibly, the non-recognition of SB and SBBs as distinct risk-factors from PA in the development of geriatric-specific poor health outcomes [37] was a reason for the small number of hospital studies focused on SB. We, therefore, have limited understanding of the prevalence, timing, frequency and determinants of SB in hospitalised older adults. In recent years there has been an increasing trend in exploring the extent of SB in older people, especially in the community. For example, Stamatakis et al. [38] found that 67% of older people spend <8.5 h sedentary while awake using an objective measure, while a combined analysis of seven studies that explored self-reported sitting time found that almost 60% of older people were sedentary for >4 h/day [39]. Thus, there is a need to explore the extent of SB and SBBs in hospital, to generate evidence and guide the development of appropriate interventions.
Currently, no hospital-based intervention studies have focused on reducing SB and SBBs in older people. The four intervention studies in this review focused on PA [17,18,27,31] and reported SB as secondary outcomes, a possible reason interventions had no significant impact on SB. Interventions exclusively targeting SB and SBBs are required to decrease SB. Purely advising patients to increase PA [40] may not address SB (and SBB) because a person might be active for a few minutes but spend the rest of the day sedentary. A better approach could be to specifically intervene to reduce SB whilst also increasing activity [41] especially for hospitalised older people who may lack the fitness to achieve moderate to vigorous levels of PA. Activities that interrupt periods of SB could include sit-to-stands, walking to the lounge room to socialise with other patients or relatives and walking to the restroom rather than using a commode.
With few studies exploring acceptability, it is unclear if older people are comfortable wearing devices or being observed. Interventions must be acceptable to end-users if there is to be adherence to and translation of the intervention into healthcare [42]. Thus, understanding the perspectives of stakeholders and end-users regarding interventions on ubiquitous health issues as SB is critical to reducing SB among hospitalised older people successfully. With better understanding and integration of the perspectives of end-users and stakeholders, investigators can design SB interventions that are effective and acceptable to both groups [42]. The lack of end-users perspectives is likely contributing to the dearth of intervention studies explicitly focused on reducing SB and SBB in hospitalised older people.
There is a lack of validated wearables and subjective tools for measuring SB in hospitalised older people. A previous systematic review has reported a dearth of valid and reliable objective tools to measure SB in older adults [43]. Regarding subjective measures, the behavioural mapping technique is validated to describe activity in older patients with stroke [44] while the IPAQ-SF is unsuitable for assessing SB in older adults [45]. For objective tools, only PAL2 has been validated to measure SB in natural hospital activities [35], while ActivPAL has only been validated to measure SB during a controlled test protocol of 1 h [36]. There are concerns that for longer durations in conventional hospital settings, the measurement error may be larger, and the validity of the results may be different from that in the controlled protocol [36]. It is suggested that there is a need for validity and reliability studies focused on older people where the required wear time, as well as position of devices for objective methods, are better clarified [43].
Consensus on the best methods to measure SB and SBB in hospitals is important. A previous systematic review exploring older people's SB found discrepancies between subjective and objective measures [39]. The behavioural mapping technique, for example, classifies SB (rolling in bed, transfers and sitting in bed/chair) as PA and its method of calculating SBBs (as time spent in SB/number of transitions from lying to sitting) underestimates it. Data capture, be it with subjective methods or wearable devices, should focus on awake periods and exclude sleep time. The heterogeneity of methods with many studies failing to meet or to report if they had met currently accepted definitions makes comparisons between studies difficult. Therefore, the true extent of SB in hospitals whilst high [86.5%/day (20.8 h)] is currently not known.

Strengths and Limitations
This scoping review has some limitations. The inclusion of studies with a mean age ≥65 years may have resulted in the exclusion of findings from studies involving younger cohorts. A focus on English only studies might have resulted in missing important research in other languages.

Conclusions
This scoping review has highlighted the need for hospital-based studies focused on SB and SBB. Importantly, there is a need to better understand the perspectives of healthcare professionals, carers and patients, as this would support the design and investigation of translatable interventions to reduce SB and SBB in hospitals. While older patients spend most of their waking hours sedentary, the prevalence of SB and SBB in hospitalised older people is unclear due to poor adherence to the current definitions and a lack of consensus on how best to measure SB and SBB in hospital. Future hospital-based studies should focus on designing context-specific interventions to reduce SB and SBBs and prioritise seeking healthcare professionals, patients and carers' perspectives of SB and SBB interventions to aid subsequent intervention design and implementation. This may improve adherence to and translatability of such interventions into clinical practice. The opportunity for further research on this topic is extensive and crucial if this health issue is to be addressed.