http://orcid.org/0000-0002-0926-8869
ABSTRACT
Background: High levels of sedentary time increases the risk of cardiovascular disease, including recurrent stroke.
Objective: This study aimed to identify factors associated with high sedentary time in community-dwelling people with stroke.
Methods: For this data pooling study, authors of published and ongoing trials that collected sedentary time data, using the activPAL monitor, in community-dwelling people with stroke were invited to contribute their raw data. The data was reprocessed, algorithms were created to identify sleep-wake time and determine the percentage of waking hours spent sedentary. We explored demographic and stroke-related factors associated with total sedentary time and time in uninterrupted sedentary bouts using unique, both univariable and multivariable, regression analyses.
Results: The 274 included participants were from Australia, Canada, and the United Kingdom, and spent, on average, 69% (SD 12.4) of their waking hours sedentary. Of the demographic and stroke-related factors, slower walking speeds were significantly and independently associated with a higher percentage of waking hours spent sedentary (p = 0.001) and uninterrupted sedentary bouts of >30 and >60 min (p = 0.001 and p = 0.004, respectively). Regression models explained 11–19% of the variance in total sedentary time and time in prolonged sedentary bouts.
Conclusion: We found that variability in sedentary time of people with stroke was largely unaccounted for by demographic and stroke-related variables. Behavioral and environmental factors are likely to play an important role in sedentary behavior after stroke. Further work is required to develop and test effective interventions to address sedentary behavior after stroke.
Introduction
Stroke is the second most common cause of death and the third leading cause of disability worldwide,Citation1,Citation2 with the burden expected to increase during the next 20 years.Citation1 Almost 40% of the people with stroke have a recurrent stroke within 10 years,Citation3 making secondary prevention vital.Citation3,Citation4 High amounts of sedentary time have been found to increase the risk of cardiovascular disease,Citation5–Citation11 particularly when the sedentary time is accumulated in prolonged bouts.Citation12–Citation15 Sedentary behavior, is defined as “any waking behavior characterized by an energy expenditure ≤1.5 Metabolic Equivalent of Task (METs) while in a sitting, reclining or lying posture”.Citation16,Citation17 Studies in healthy people, as well as people with diabetes and obesity, have shown that reducing the total amount of sedentary time and/or breaking up long periods of uninterrupted sedentary time, reduces metabolic risk factors associated with cardiovascular disease.Citation6,Citation9,Citation10,Citation12–Citation15 Recent studies have shown that people living in the community after stroke spend more time each day sedentary, and more time in uninterrupted bouts of sedentary time compared to age-matched healthy peers.Citation18–Citation20 Reducing sedentary time and breaking up long sedentary bouts with short bursts of activity may be a promising intervention to reduce the risk of recurrent stroke and other cardiovascular diseases in people with stroke.
To develop effective interventions, it is important to understand the factors associated with sedentary time in people with stroke. Previous studies have found associations between self-reported physical function after stroke and total sedentary time, but inconsistent results with regards to the relationship of age, stroke severity, and walking speed with sedentary time.Citation20,Citation21 These results are from secondary analyses of single-site observational studies, not powered to address associations, and inconsistent in the methods used to determine waking hours; thus making direct comparisons between studies difficult.Citation20,Citation21 Individual participant data pooling, with consistent processing of wake time data, allows novel exploratory analyses of larger datasets with greater power.
By pooling all available individual participant data internationally, this study aimed to comprehensively explore the factors associated with sedentary time in community-dwelling people with stroke. Specifically, our research questions were: (1) What factors are associated with total sedentary time during waking hours after stroke? (2) What factors are associated with time spent in prolonged sedentary bouts during waking hours?
Methods
Study design
This was an exploratory data pooling study, in which existing individual participant data were used for secondary analyses. By searches of databases, trial registries, and word of mouth, potentially eligible datasets were identified, and authors were invited to contribute their individual participant data and raw activity monitor data. The study was approved by the Human Research Ethics Committee of The University of Newcastle (H-2016–0427).
Study selection
Datasets from studies were included if they met the following criteria;
Included adults with stroke who were living in the community,
Measured sedentary behavior using the activPAL monitor (PAL Technologies Ltd, Glasgow, United Kingdom),
The ethical approval and informed consent for the data collection permitted the use of the data for secondary analyses,
The available data was not influenced by any form of intervention.
Authors of original studies provided de-identified datasets. Factors included in the datasets were mapped by one author (WH) in consultation with the co-authors. A list of factors of interest was created a priori (see ), based on previous research in determinants of sedentary time and consideration of other relevant stroke-related factors.Citation20–Citation28 For each dataset, we determined which factors were measured and what measurement instrument was used. Where different measurement instruments were used for the same factor, we sought valid methods to categorize or dichotomize data to facilitate data pooling (see supplementary Box 1 for the conversion methods). Where the original studies included repeated measures, we included data from one time-point only and used the time-point with the least missing data or at baseline in the case of intervention trials.
Box. 1 Factors of interest determined a priori.
Activity monitor data
We chose to only include data on sedentary time that was measured using the activPAL monitor (PAL Technologies Ltd, Glasgow, United Kingdom) because it is highly reliable (Intraclass correlation coefficient 0.79–0.99) and valid (98–100% accuracy) for measuring sedentary time and posture transitions during daily life in people with stroke.Citation29–Citation31 The ActivPAL uses an inclinometer worn on the anterior side of the thigh to determine if someone is either sedentary (sitting, lying or reclining), standing or walking making it a highly valid and accurate monitor to determine sedentary time.Citation29–Citation31 A conversion to METs is also possible.Citation29–Citation31 Event files from all participants were combined into one dataset. To identify waking hours, a custom algorithm was developed based on previously published codes.Citation32 The algorithm aggregated sleep time based on the largest bout of sitting/lying time within a 24-h period and then aggregated adjacent bouts of sitting/lying time where these bouts were interrupted by short bursts of activity, i.e. to account for getting up to the toilet overnight (see Appendix for more details). Our previous work has found that any three days of monitoring, regardless of weekend or weekday, is sufficient to accurately represent habitual physical activity over seven days.Citation33 We therefore included participants with at least three days of valid (>8 h day) waking wear time.Citation33 We excluded days in which more than 18 h of wake time were identified.
Data processing and analyses
From the activPAL data during waking hours, the percentage of total sedentary time and the percentage of waking hours spent in prolonged bouts of sedentary time was determined. Two variables were created for prolonged bouts: percentage of sedentary time in bouts >30 min and percentage of sedentary time in bouts >60 min.Citation9,Citation10,Citation12,Citation18 Linear regressions (adjusting for age, gender, and study) were conducted to determine the association of individual factors with percentage of total sedentary time, percentage of sedentary time in bouts >30 min, and percentage of sedentary time in bouts >60 min. All factors and residuals (from regression analyses) were checked for normality and where needed the appropriate transformations were computed. Factors that were found significantly associated with univariable regressions (p < 0.05) were included in the multivariable regressions. We first determined the coverage of factors across studies and then conducted the multivariable regressions with the best coverage of factors across studies and the highest sample sizes. To avoid collinearity, if correlations between independent factors were higher than r = 0.850 one factor was removed from the analyses.Citation34,Citation35 Both forward and backward stepwise linear regressions were run. Based on the 1:10 rule by Peduzzi et al.,Citation36 a sample of at least n = 250 was needed to be able to include all the factors we identified a priori (Box 1). All analyses were conducted with R statistical software, version 3.3.3 and IBM SPSS statistics version 22.
Results
Participant characteristics
Ten datasets were identified that met the inclusion criteria and we were able to obtain individual participant data from 9 (90%), including n = 350 individual participants (). In all, n = 274 (78%) individual participants contributed at least three days of valid activPAL data. There were no differences in demographics between the original (n = 350) and final (n = 274) sample (). On average, participants spent 69 (Standard Deviation 12)% of waking hours sedentary, 40 (SD 16)% of waking hours in sedentary bouts >30 min and 23 (SD 15)% of waking hours in sedentary bouts >60 min. Only age and gender were reported in all studies; other variables were reported in between 3 (33%) and 8 (89%) of included studies (Supplementary Table 1).
Table 1. Characteristics of studies that provided data.
Table 2. Participant demographics.
Factors associated with total sedentary time
The results of the univariable regression (adjusting for age, gender, and study) for percentage of total sedentary time are shown in . Body mass index (p = 0.048), stroke severity (p = 0.035), walking speed (p < 0.001), walking capacity (p < 0.001), walking aid use (p < 0.001), degree of independence in activities of daily living (p = 0.014), and anxiety (p = 0.028) were all significantly associated with percentage of total sedentary time. As walking speed and walking capacity were highly correlated (r = 0.897), and more data were available across the datasets for walking speed, only walking speed was included in the multivariable regression analyses. Only walking speed remained significant in the multivariable regression model (p = 0.001, see ), which explained 14% of the variance in percentage of total sedentary time.
Table 3. Univariate regressions.
Table 4. Multivariable regression.
Factors associated with time spent in prolonged sedentary bouts
The results of the univariable regression (adjusting for age, gender, and study) for percentage of sedentary time in bouts >30 min and percentage of sedentary time in bouts >60 min are shown in . Body mass index (p = 0.024 and p = 0.038), stroke severity (p = 0.019 and p = 0.016), walking speed (both p < 0.001), walking capacity (both p < 0.001), walking aid use (p < 0.001 and p = 0.009), and independence in activities of daily living (p = 0.003 and p = 0.005) were significantly associated with percentage of sedentary time in bouts >30 min and percentage of sedentary time in bouts >60 min. Fatigue was significantly associated only with percentage of sedentary time in bouts >60 min (p = 0.044).
Walking capacity was removed from the multivariable regression because of the high correlation with walking speed. In the multivariable regressions (), only walking speed was significantly associated with percentage of sedentary time in bouts >30 min (p = 0.001) and percentage of sedentary time in bouts >60 min (p = 0.004). For percentage of sedentary time in bouts >30 min, body mass index (p = 0.049) was also found to be significantly associated. The models explained 19% of the variance in percentage of sedentary time in bouts >30 min and 11% of the variance in percentage of sedentary time in bouts >60 min.
There was a wide range in time since the stroke in our dataset (1 to 237 months) and these data were highly skewed. To check whether this confounded results, we categorized the time since stroke into three epochs (1 to 3 months, 3 to 6 months and >6 months) and re-ran the regression models for the percentage of total sedentary time using this ordinal variable. This did not change the results.
Discussion
We pooled data from 274 individuals from three countries and found that people with stroke spent on average 69% of waking hours sedentary. Slower walking speed was the only factor independently associated with more total sedentary time, and more time spent in prolonged bouts of sedentary behavior. However, our models accounted for only a small proportion of the variance in sedentary behavior, suggesting that other factors not measured in the participants included in this study are also important.
Our findings in relation to walking speed are consistent with a previous study which found both slower walking speed, and other measures of poorer physical function (in this case the Stroke Impact Scale) were associated with greater sedentary time.Citation21 However, walking speed may also be a proxy measure for general health and co-morbidities.Citation39–Citation41 In older people, walking speed is an important predictor of a number of adverse outcomes such as falls, activities of daily living difficulties, disability, institutionalization, comorbidities, and mortality.Citation39–Citation43 Further research is needed to determine whether there is a direct causal pathway between slow walking speed and high sedentary time, or if it is a proxy measure of general health. It is possible that interventions aimed at improving the walking abilities of people with stroke might help reduce the total sedentary time and the time spent sedentary in prolonged bouts. However, this premise requires testing in clinical trials.
We found few other factors were independently associated with high sedentary behavior. This is in contrast to previous studies. In older adults without stroke, age, gender, education level, living arrangements, body mass index, smoking status, and independence in activities of daily living, were all found to be associated with sedentary behavior.Citation22,Citation25–Citation27 In previous studies of people with stroke both age and stroke severity were associated with sedentary behavior.Citation20,Citation21 In people with multiple sclerosis, both disease severity and physical ability are reported to be associated with high sedentary time.Citation44 Taken together, this suggests that the factors associated with high sedentary time may differ between population groups. This is important to consider when developing interventions to reduce sedentary behavior.
In our analyses, the regression models accounted for only a small proportion of the variance in sedentary behavior. It is likely that environmental and behavioral factors may also influence sedentary time in people with stroke, and this should be taken into consideration when designing interventions to reduce sedentary behavior in this population. Such interventions will need to be carefully developed and include strategies to address both the factors influencing sedentary behavior, and the barriers and motivations to increase light, moderate, and vigorous physical activity. Systematic reviews of clinical trials in other populations (healthy and older adults, those with diabetes or obesity) have highlighted the importance of developing interventions specifically targeted to reduce sedentary time, as such programs are more effective for reducing sedentary time compared with interventions that aim to increase physical activity alone.Citation45,Citation46 An international consensus framework for sedentary behavior research across all population groups,Citation23 as well as qualitative research involving people with stroke,Citation47 highlights the importance of the environment, psychology (including motivation), education, and behavior as determinants of sedentary time. Development of effective interventions to address high levels of sedentary time in people with stroke will need to take all these factors into consideration.
Strengths and limitations
We pooled all available individual participant activity monitor data, and completed a novel exploratory analysis on a large dataset, with sufficient statistical power. We choose this novel data pooling methodology (instead of for instance meta-analyses) to be able to conduct independent secondary analyses using raw data. This also allowed the inclusion of data from ongoing and unpublished studies. We did not complete systematic literature searches, meaning that it is possible that some potentially relevant datasets were missed. The extensive international collaboration that was the foundation of this study allows confidence that we captured the vast majority of trials that have included activPAL data. The large dataset provides confidence in the results. We re-processed all raw activity monitor files using a custom-built algorithm to consistently and systematically identify sleep-wake time without manual error.Citation32 We decided to use only data in which the activPAL was used to measure sedentary time. This decision was based on the fact that different activity monitors use different methods to determine the sedentary time and movement, and therefore combining raw data from different monitors would introduce bias.Citation48,Citation49 Two studies have shown the incompatibility of data from different monitors.Citation48,Citation49 Only including activPAL data provides confidence in the validity of data between datasets. We acknowledge that this reduced the number of datasets we were able to include. Since the activPAL is highly reliable in the determination of sedentary behavior it is a commonly used monitor and therefore enabled the inclusion of most of the data that is available.
While we pooled all the available individual participant data, not all factors of interest we identified a priori were available. Furthermore, even where the same construct (for example, depression, anxiety, physical ability) was measured, the variability in the outcome measures used necessitated categorizing or dichotomizing data. To facilitate comparability of research findings and future data pooling studies, greater consistency in outcome measurement tools used is required.Citation50,Citation51 The international Stroke Recovery and Rehabilitation Round Table group recently conducted a consensus project and have published recommendations for a core dataset for all stroke recovery and rehabilitation trials.Citation50
Though the cut-offs of 30 and 60 min, used as an outcome variable for prolonged sedentary time, in their origin are arbitrary they have been used in previous studies on the risk of sedentary behavior.Citation12–Citation15 These studies have shown that the risk of cardiovascular disease increases even more when the sedentary time is accumulated in these prolonged bouts.Citation12–Citation15 Therefore these cut-offs provide a standard metric for prolonged sedentary time.
This study included only people with stroke living in the community, and for the most part only those able to walk independently, therefore results have limited generalizability beyond this group.
Conclusion
We found that variability in sedentary time of people with stroke was largely unaccounted for by demographic and stroke-related variables. Behavioral and environmental factors are likely to play an important role in sedentary behavior after stroke. Further work is required to develop and test effective interventions to address sedentary behavior after stroke.
Disclosure of interest
Authors have no conflicts of interest to declare.
Supplemental Material
Download MS Word (19.3 KB)Additional information
Funding
References
- Donnan GA, Fisher M, Macleod M, Davis SM. Stroke. Lancet. 2008;371:1612–1623. doi:10.1016/S0140-6736(08)60694-7
- Feigin VL, Forouzanfar MH, Krishnamurthi R, et al. Global and regional burden of stroke during 1990–2010: findings from the global burden of disease study 2010. Lancet. 2014;383:245–254. doi:10.1016/S0140-6736(13)61612-8
- Johnston SC, Rothwell PM, Nguyen-Huynh MN, et al. Validation and refinement of scores to predict very early stroke risk after transient ischaemic attack. Lancet. 2007;369(9558):283–292. doi:10.1016/S0140-6736(07)60150-0
- Mohan KM, Wolfe CD, Rudd AG, Heuschmann PU, Kolominsky-Rabas PL, Grieve AP. Risk and cumulative risk of stroke recurrence: a systematic review and meta-analysis. Stroke. 2011;42(5):1489–1494. doi:10.1161/STROKEAHA.110.602615.
- van der Ploeg HP, Chey T, Korda RJ, et al. Sitting time and all-cause mortality risk in 222,497 Australian adults. Arch Intern Med. 2012;172:494–500. doi:10.1001/archinternmed.2011.2174
- Neville O, Healy GN, Matthews CE, Dunstan DW. Too much sitting: the population-health science of sedentary behavior. Exerc Sport Sci Rev. 2010;38(3):105–113. doi:10.1097/JES.0b013e3181e373a2.
- Ford ES, Caspersen CJ. Sedentary behaviour and cardiovascular disease: a review of prospective studies. Int J Epidemiol. 2012;41(5):1338–1353. doi:10.1093/ije/dys078.
- Biswas A, Oh PI, Faulkner GE, et al. Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: a systematic review and meta-analysis. Ann Intern Med. 2015;162(2):123–132.doi:10.7326/M14-1651
- Bauman AE, Chau JY, Ding D, Bennie J. Too much sitting and cardio-metabolic risk: an update of epidemiological evidence. Curr Cardiovasc Risk Rep. 2013;7:293–298. doi:10.1007/s12170-013-0316-y
- Dunstan DW, Howard B, Healy GN, Owen N. Too much sitting – a health hazard. Diabetes Res Clin Pract. 2012;97(3):368–376. doi:10.1016/j.diabres.2012.05.020.
- Tremblay MS, Colley RC, Saunders TJ, Healy GN, Owen N. Physiological and health implications of a sedentary lifestyle. Appl Physiol Nutr Metab. 2010;35(6):725–740. doi:10.1139/H10-079.
- Healy GN, Dunstan DW, Salmon J, et al. Breaks in sedentary time: beneficial associations with metabolic risk. Diabetes Care. 2008;31(4):661–666.doi:10.2337/dc07-2046
- Benatti FB, Ried-Larsen M. The effects of breaking up prolonged sitting time: a review of experimental studies. Med Sci Sports Exerc. 2015;47(10):2053–2061. doi:10.1249/MSS.0000000000000654.
- Chastin SF, Egerton T, Leask C, Stamatakis E. Meta-analysis of the relationship between breaks in sedentary behavior and cardiometabolic health. Obesity (Silver Spring). 2015;23(9):1800–1810. doi:10.1002/oby.21180.
- Healy GN, Matthews CE, Dunstan DW, et al. Sedentary time and cardio-metabolic biomarkers in US adults: NHANES 2003-06. Eur Heart J. 2011;32:590–597. doi:10.1093/eurheartj/ehq451
- Pate RR, O‘Neill JR, Lobelo F. The evolving definition of “sedentary”. Exerc Sport Sci Rev. 2008;36(4):173–178. doi:10.1097/JES.0b013e3181877d1a.
- Tremblay MS, Aubert S, Barnes JD, et al. SBRN terminology consensus project participants. Sedentary Behavior Research Network (SBRN) - terminology consensus project process and outcome. Int J Behav Nutr Phys Act. 2017;14(1):75.doi:10.1186/s12966-017-0525-8
- English C, Healy GN, Coates A, Lewis L, Olds T, Bernhardt J. Sitting and activity time in people with stroke. Phys Ther. 2016;96(2):193–201. doi:10.2522/ptj.20140522.
- Paul L, Brewster S, Wyke S, et al. Physical activity profiles and sedentary behaviour in people following stroke: a cross-sectional study. Disabil Rehabil. 2016;38(4):362–367.doi:10.3109/09638288.2015.1041615
- Tieges Z, Mead G, Allerhand M, et al. Sedentary behavior in the first year after stroke: a longitudinal cohort study with objective measures. Arch Phys Med Rehabil. 2015;96(1):15–23.doi:10.1016/j.apmr.2014.08.015
- English C, Healy GN, Coates A, Lewis LK, Olds T, Bernhardt J. Sitting time and physical activity after stroke: physical ability is only part of the story. Top Stroke Rehabil. 2016;23(1):36–42. doi:10.1179/1945511915Y.0000000009.
- Chastin SF, Buck C, Freiberger E, et al. DEDIPAC consortium and on behalf of the DEDIPAC consortium. Systematic literature review of determinants of sedentary behaviour in older adults: a DEDIPAC study. Int J Behav Nutr Phys Act. 2015;12:127. doi:10.1186/s12966-015-0292-3
- Chastin SF, De Craemer M, Lien N, et al. DEDIPAC consortium, expert working group and consensus panel. The SOS-framework (Systems of Sedentary behaviours): an international transdisciplinary consensus framework for the study of determinants, research priorities and policy on sedentary behaviour across the life course: a DEDIPAC-study. Int J Behav Nutr Phys Act. 2016;13:83. doi:10.1186/s12966-016-0409-3
- Bampton EA, Johnson ST, Vallance JK. Profiles of resistance training behavior and sedentary time among older adults: associations with health-related quality of life and psychosocial health. Prev Med Rep. 2015;2:773–776. doi:10.1016/j.pmedr.2015.08.017
- Chen T, Narazaki K, Haeuchi Y, Chen S, Honda T, Kumagai S. Associations of sedentary time and breaks in sedentary time with disability in instrumental activities of daily living in community-dwelling older adults. J Phys Act Health. 2016;13(3):303–309. doi:10.1123/jpah.2015-0090.
- Diaz KM, Howard VJ, Hutto B, et al. Patterns of sedentary behavior in US middle-age and older adults: the REGARDS study. Med Sci Sports Exerc. 2016 Mar;48(3):430–438. doi:10.1249/MSS.0000000000000792
- Heseltine R, Skelton DA, Kendrick D, et al. “Keeping moving”: factors associated with sedentary behaviour among older people recruited to an exercise promotion trial in general practice. BMC Fam Pract. 2015;16:67. doi:10.1186/s12875-015-0284-z
- Sardinha LB, Santos DA, Silva AM, Baptista F, Owen N. Breaking-up sedentary time is associated with physical function in older adults. J Gerontol A Biol Sci Med Sci. 2015;70(1):119–124. Epub 2014 Oct 16. doi:10.1093/gerona/glu193.
- Taraldsen K, Askim T, Sletvold O, et al. Evaluation of a body-worn sensor system to measure physical activity in older people with impaired function. Phys Ther. 2011;91:277–285. doi:10.2522/ptj.20100159
- Lyden K, Kozey Keadle SL, Staudenmayer JW, Freedson PS. Validity of two wearable monitors to estimate breaks from sedentary time. Med Sci Sports Exerc. 2012;44:2243–2252. doi:10.1249/MSS.0b013e318260c477
- Godfrey A, Culhane KM, Lyons GM. Comparison of the performance of the activPAL Professional physical activity logger to a discrete accelerometer-based activity monitor. Med Eng Phys. 2007;29:930–934. doi:10.1016/j.medengphy.2006.10.001
- Winkler EAH, Bodicoat DH, Healy GN, et al. Identifying adults’ valid waking wear time by automated estimation in ActivPAL data collected with a 24 H wear protocol. Physiol Meas. 2016;37(10):1653.doi:10.1088/0967-3334/37/10/1653
- Tinlin L, Fini N, Bernhardt J, et al. Best practice guidelines for the measurement of physical activity levels in stroke survivors. Int J Rehabil Res. 2018 Mar;41(1):14–19. doi:10.1097/MRR.0000000000000253
- de Vocht A. Basishandboek SPSS 22 IBM SPSS statistics 22. Bijleveld. 2014;192–193:9789055482412.
- Fields, Andy Discovering Statistics Using SPSS. SAGE Publications Ltd; London, 2009. Vols. 223–224, p. 223-224.
- Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379.
- Jones TM, Dear BF, Hush JM, Titov N, Dean CM. myMoves program: feasibility and acceptability study of a remotely delivered self-management program for increasing physical activity among adults with acquired brain injury living in the community. Phys Ther. 2016;96(12):1982–1993. doi:10.2522/ptj.20160028
- Mahendran N, Kuys SS, Brauer SG. Recovery of ambulation activity across the first six months post-stroke. Gait Posture. 2016;49:271–276. doi:10.1016/j.gaitpost.2016.06.038
- Viccaro LJ, Perera S, Studenski SA. Is timed up and go better than gait speed in predicting health, function, and falls in older adults? J Am Geriatr Soc. 2011;59(5):887–892. doi:10.1111/j.1532-5415.2011.03336.x.
- Callisaya ML, Blizzard L, Schmidt MD, et al. Gait, gait variability and the risk of multiple incident falls in older people: a population-based study. Age Ageing. 2011 Jul;40(4):481–487. doi:10.1093/ageing/afr055
- Tabue-Teguo M, Le Goff M, Avila-Funes JA, et al. Walking and psychomotor speed in the elderly: concordance, correlates and prediction of death. J Nutr Health Aging. 2015;19(4):468–473.doi:10.1007/s12603-014-0560-y
- Donoghue OA, Savva GM, Cronin H, Kenny RA, Horgan NF. Using timed up and go and usual gait speed to predict incident disability in daily activities among community-dwelling adults aged 65 and older. Arch Phys Med Rehabil. 2014;95(10):1954–1961. doi:10.1016/j.apmr.2014.06.008.
- Abellan van Kan G, Rolland Y, Andrieu S, et al. Gait speed at usual pace as a predictor of adverse outcomes in community-dwelling older people an International Academy on Nutrition and Aging (IANA) task force. J Nutr Health Aging. 2009;13(10):881–889.
- Veldhuijzen van Zanten JJ, Pilutti LA, Duda JL, Motl RW. Sedentary behaviour in people with multiple sclerosis: is it time to stand up against MS? Mult Scler. 2016;22(10):1250–1256. doi:10.1177/1352458516644340.
- Prince SA, Saunders TJ, Gresty K, Reid RD. A comparison of the effectiveness of physical activity and sedentary behaviour interventions in reducing sedentary time in adults: a systematic review and meta-analysis of controlled trials. Obes Rev. 2014;15(11):905–919. doi:10.1111/obr.12215.
- Martin A, Fitzsimons C, Jepson R, et al. Interventions with potential to reduce sedentary time in adults: systematic review and meta-analysis. EuroFIT consortium. Br J Sports Med. 2015;49(16):1056–1063.doi:10.1136/bjsports-2014-094524
- Ezeugwu VE, Garga N, Manns PJ. Reducing sedentary behaviour after stroke: perspectives of ambulatory individuals with stroke. Disabil Rehabil. 2017;39(25):2551–2558. doi:10.1080/09638288.2016.1239764.
- Fanchamps MHJ, van Den Berg-Emons HJG, Stam HJ, Bussmann JBJ. Sedentary behavior: different types of operationalization influence outcome measures. Gait Posture. 2017;54:188–193. doi:10.1016/j.gaitpost.2017.02.025
- Kozey-Keadle S, Libertine A, Lyden K, Staudenmayer J, Freedson PS. Validation of wearable monitors for assessing sedentary behavior. Med Sci Sports Exerc. 2011;43(8):1561–1567. doi:10.1249/MSS.0b013e31820ce174.
- Kwakkel G, Lannin NA, Borschmann K, et al. Standardized measurement of sensorimotor recovery in stroke trials: consensus-based core recommendations from the stroke recovery and rehabilitation roundtable. Int J Stroke. 2017;12(5):451–461.doi:10.1177/1747493017711813
- Ali M, English C, Bernhardt J, Sunnerhagen KS, Brady M. VISTA-rehab collaboration. More outcomes than trials: a call for consistent data collection across stroke rehabilitation trials. Int J Stroke. 2013;8(1):18–24. doi:10.1111/j.1747-4949.2012.00973.x.
Appendix
Sleep/Non-wear time identification algorithm
Objective
Identify the single daily longest period of sleep/non-wear activity in order to delineate what is considered as wake period.
Methods
The simple prescription given by Elisabeth Winkler et al. (Winkler et al. (2016)) was used as a starting point.
Recorded data consists of activPAL timestamped events, typified as sitting/lying, standing and walking. Events represent the longest continuous uninterrupted activity of each class. There is one event per step.
It was observed during initial implementation of Winkler‘s prescription that sleep period patterns for this cohort exhibit a more interrupted pattern, requiring a more flexible approach to correctly identify periods. The algorithm was modified as shown below.
Pseudocode:
Definitions
SL: sleep period. A sleep period consists of a “chain” of “nearby” events, primarily of class lying, that accounts for the longest aggregated resting period in a 24hr interval. The meaning of “chain” and “nearby” is made precise through the pseudocode. A sleep period is defined by its start and end times, which must be start and end times of lying-class events, and all events encompassed in between. duration(SL) is the total accumulated time in lying events in SL.
e1, e2 represent generic lying events. A lying event carries an “aggregation opportunity window” of length of 12 min + 10% of event duration, capped at 45 min. Longer events have longer opportunity windows to be aggregated into the sleep event chain. The opportunity window of a lying event is denoted below as opp.window(e).
Ev is the list of all events in a 24hr interval for an individual, from noon to noon next day.
LEv is the list of lying events longer than 30 min in Ev, to be considered for aggregation in the sleep period (“long lying events”)
Tlev is the total time accumulated in long lying events in the day. Used in considering an alternative chain of lying events for the sleep period.
Algorithm
Note: how to read pseudocode. A simplified pseudocode of the algorithm is shown below. while and for each imply a loop, if imply testing a conditiont; the level of indentation indicates the actions included in the repeating part of the loop or the true outcome of the test. For clarity, abnormal termination conditions are excluded from the algorithm below.
Input: Ev
Output: SL
LEv = get lying events longer than 30 minutes from Ev
Tlev = sum of event duration for events in LEv
e1 = find longest event in LEv
A:
initialise sleep chain SL with e1
mark e1 as used
while there are unused events in LEv and SL modified since last pass
for each unused event e2 in LEv, in descending duration order
if opp.window(e2) overlaps SL
add e2 to SL
mark e2 as used
endif
endfor
endwhile
if duration(SL) < 0.4 Tlev and there are unused events in LEv
e1 = find longest unused event from LEv
mark all events in LEv as unused
mark e1 as used
restart from A:
endif
References
Grolemund, Garrett, and Hadley Wickham. 2011. “Dates and Times Made Easy with lubridate.” Journal of Statistical Software 40 (3): 1–25. http://www.jstatsoft.org/v40/i03/.
James, David, and Kurt Hornik. 2017. Chron: Chronological Objects Which Can Handle Dates and Times. https://CRAN.R-project.org/package = chron.
R Core Team. 2017. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/.
Winkler, Elisabeth A. H., Danielle H. Bodicoat, Genevieve N. Healy, Kishan Bakrania, Thomas Yates, Neville Owen, David W. Dunstan, and Charlotte L. Edwardson. 2016. “Identifying Adults’ Valid Waking Wear Time by Automated Estimation in ActivPAL Data Collected with a 24 H Wear Protocol.” Physiological Measurement 37 (10): 1653. doi:10.1088/0967–3334/37/10/1653.