Received: 21 April 2020 | Revised: 19 June 2020 | Accepted: 29 June 2020 DOI: 10.1111/sms.13762 ORIGINAL ARTICLE Sedentary behavior, physical activity, and mental health in older adults: An isotemporal substitution model Mark A. Tully1 | Ilona McMullan2 | Nicole E. Blackburn3 | Jason J. Wilson2 | Brendan Bunting4 | Lee Smith5 | Frank Kee6 | Manuela Deidda7 | Maria Giné-Garriga8,9 | Laura Coll-Planas10 | Dhayana Dallmeier11,12 | Michael Denkinger 11 | Dietrich Rothenbacher 13 | Paolo Caserotti14 1 School of Health Sciences, Institute of Mental Health Sciences, Ulster University, Newtownabbey, UK 2 School of Health Sciences, Ulster University, Newtownabbey, UK 3 School of Health Sciences, Institute of Nursing and Health Research, Centre for Health and Rehabilitation Technologies, Ulster University, Newtownabbey, UK 4 School of Psychology, Ulster University, Newtownabbey, UK 5 The Cambridge Centre for Sport and Exercise Sciences, Anglia Ruskin University, Cambridge, UK 6 Centre for Public Health, School of Medicine, Dentistry and Biomedical Science, Queen's University Belfast, Belfast, UK 7 Health Economics & Health Technology Assessment, Institute of Health & Wellbeing, University of Glasgow, Glasgow, UK 8 Department of Sport Sciences, Faculty of Psychology, Education and Sport Sciences Blanquerna, Universitat Ramon Llull, Barcelona, Spain 9 Department of Physical Therapy, Faculty of Health Sciences Blanquerna, Universitat Ramon Llull, Barcelona, Spain 10 Fundació Salut i Envelliment, Universitat Autònoma de Barcelona, Barcelona, Spain 11 Geriatric Centre Ulm/Alb-Donau, AGAPLESION Bethesda Clinic Ulm, Geriatric Research Unit Ulm University, Ulm, Germany 12 Dept. of Epidemiology, Boston University School of Public Health, Boston, MA, USA 13 Institute of Epidemiology and Medical Biometry, Ulm University, Ulm, Germany 14 Department of Sports Science and Clinical Biomechanics, Center for Active and Healthy Ageing (CAHA), University of Southern Denmark, Odense, Denmark Correspondence Mark A. Tully, School of Health Sciences, Introduction: Regular moderate-to-vigorous physical activity (MVPA) is associ- Institute of Mental Health Sciences, Ulster ated with improved mental health, but the evidence for the effect of reducing sed- University, Shore Road, Newtownabbey, Co entary behavior (SB) or increasing light PA (LPA) in older adults is lacking. Using Antrim BT37 0QB, UK. Email: m.tully@ulster.ac.uk isotemporal substitution (IS) models, the aim of this paper was to investigate the effect of substituting SB with LPA or MVPA on associations with mental health in Funding information older adults. European Commission, Grant/Award Number: 634270 Methods: Data from 1360 older adults (mean age 75.18 years) in four countries were utilized. PA and SB was measured using ActiGraph wGT3X-BT + acceler- ometers worn for 7 days. Self-rated mental health was measured using the Hospital and Anxiety Depression Scale (HADS). IS models estimated cross-sectional associa- tions when 30 minutes of one behavior was substituted with another. Models were adjusted for age, sex, marital status, and educational attainment. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2020 The Authors. Scandinavian Journal of Medicine & Science In Sports published by John Wiley & Sons Ltd Scand J Med Sci Sports. 2020;30:1957–1965.  wileyonlinelibrary.com/journal/sms | 1957 1958 |    TULLY et al. Results: Substituting 30 minutes of SB with LPA (β −.37; 95% CI −0.42, −0.32) or MVPA (β −.14; 95% CI −0.21, −0.07) and substituting LPA with MVPA (β −.11; 95% CI −0.18, −0.04) were associated with improvements in anxiety. However, sub- stituting 30 minutes of SB with LPA (β .55; 95% CI 0.49, 0.62) was associated with increased depression. Conclusion: Replacing 30 minutes of SB with LPA or MVPA was associated with improved anxiety symptoms in older adults. Greater benefits were observed when shifting SB and LPA to MVPA. KEYWORDS isotemporal, light physical activity, mental health, moderate-vigorous physical activity, older adults 1 | IN TRO D U C T ION MVPA and therefore are useful for modeling this.17 A recent review of IS studies has demonstrated that the majority of Regular physical activity is associated with a wide range evidence relates to substituting SB with MVPA, resulting in of health benefits for older adults.1,2 There is considerable potential benefits to mortality, general health, mental health, previous evidence demonstrating an association between adiposity, fitness, and cardiometabolic biomarkers.18 In a moderate-to-vigorous intensity physical activity (MVPA) study of 32 900 US women from the Nurses' Health Study, and mental health, including reduced levels of anxiety and Mekary et al16 concluded that reallocating time spent watch- depression.3,4 A systematic review of 12 prospective studies ing TV to brisk walking was associated with a significantly including 128,553 participants reported a significant positive lower risk of depression. However, this was not shown for association between SB and the risk of depression.5 Evidence the reallocation of time spent watching TV to slow walking. is accumulating that demonstrates an association between Rethorst et al19 demonstrated that replacing SB with vigorous both sedentary behavior (SB) and light-intensity physical intensity activity, but not LPA or moderate intensity activity, activity (LPA) with mental health,6,7 though the evidence is was significantly associated with a decrease in depressive inconsistent.8 In addition, there is a lack of research on the symptoms in a cohort of 16 415 Hispanic/Latino men and relationship between physical activity and mental health in women, aged 18-74 years. In contrast, Curtis et al did not older adults.1 find an association between substituting SB with either LPA Levels of physical activity decline with age,9,10 meaning or MVPA and symptoms of depression, anxiety, or stress.20 that older adults spend 79% of their waking day in SB,9,10 There has been a lot less research investigating the associ- defined as any waking behavior characterized by an energy ation between reallocating SB time with physical activity on expenditure ≤1.5 metabolic equivalents (METs), while in a depression and anxiety in older adults. Del Pozo Cruz et al21 sitting, reclining, or lying posture.11 The rest of their day is have demonstrated that reallocating 60 minutes from SB to spent either in LPA, MVPA, or sleeping. With current ev- MVPA was associated with small reductions in depressive idence supporting a dose response relationship between symptoms in a sample of 3233 adults and older adults from physical activity and health, current public health recommen- the USA. Similarly, Dillon et al observed and an association dations encourage people to “sit less and move more.”12,13 between a reallocation of 30 minutes of SB to LPA and de- However, less evidence exists as to the specific benefits of creased anxiety symptoms but not depression in 396 adults substituting SB with either LPA or MVPA on mental health, aged 50-69.22 No statistically significant associations with or indeed the volume of substituted activity that may be con- depressive symptoms were observed when sedentary behav- fer these benefits.14 This may be in part due to the financial ior was substituted with either LPA or MVPA. In a study of and time costs of conducting randomized controlled trials. 276 older adults, Yasunaga et al23 concluded that replacing Previous studies of the association between health outcomes 30 minutes of SB with the same amount of LPA is statisti- and SB have generally adjusted for minutes of MVPA of par- cally negatively associated with depression measured using ticipants, without consideration LPA or of the how changes in the geriatric depression scale in a study of 276 Japanese older MVPA impact on the time spent in other behaviors.15 adults aged 65-85 years. However, data from European coun- The isotemporal substitution (IS) approach as suggested tries are lacking. by Mekary et al's16 simultaneously model the replacement of On average, older adults only spend 3% of their waking one activity for another and so has the potential to show the day in MVPA.9,10 Health-related barriers, such as frailty, are effects of substituting SB for physical activity such as LPA or the main reasons given by older adults for not undertaking TULLY et al.    | 1959 MVPA, but they also report other barriers such as habits, car- University (Spain), The Regional Committees on Health ing duties, low motivation, and misperceptions of physical Research Ethics for Southern Denmark (Denmark), Office for activity and ageing.24 Therefore, for inactive older adults, Research Ethics Committees in Northern Ireland (Northern targeting substituting SB with LPA may be a more feasible, Ireland), and the Ethical Review Board of Ulm University attractive, or realistic behavior change to target in the first in- (Germany). Participation was voluntary, and all participants stance. However, a lack of studies using IS models to explore provided informed consent before the start of the study. the benefits of substituting SB with LPA on mental health in older adults has been identified.18 Given both the lack of research investigating the relation- 2.2 | Outcome measures ship between substituting SB for LPA and mental health in older adults, and the potential attractiveness of LPA for older 2.2.1 | Physical activity and adults, there is a need for research to examine these areas. sedentary behavior Using a larger sample of older adults than in the previous study of older adults, drawn from four different countries, At baseline, all participants were asked to wear an accelerom- the aim of this paper was to investigate the associations be- eter (ActiGraph wGT3X-BT+; ActiGraph, LLC, Pensacola, tween Actigraph measured SB, LPA, and MVPA with mental FL) on their dominant hip during waking hours for seven health, anxiety, and depression. This was done by exploring consecutive days and were told to remove them during any the effect of substituting SB with LPA or MVPA on associa- water-based activities such as bathing or swimming, and dur- tions with mental health in older adults using an isotemporal ing sleep time. The devices were initialized to collect data at approach. 30 hz, analyzed using ActiLife v6.13.3 software with nor- mal filter, and summarized using 10-second epochs. Non- wear time was defined using the Choi algorithm26 which 2 | M ATE R IA L S A N D ME T HODS uses a 2-window system, a 90-minute window for checking for consecutive zero counts, and another 30-minute up- and 2.1 | Participants down-stream window for checking for more than 2 minutes of non-zero counts.26 The study included the results from This cross-sectional analysis is based on baseline data avail- participants with at least four valid days including at least able from the SITLESS study, which is a prospective trial one weekend day.27 Furthermore, to be included, each day of community-dwelling men and women aged ≥65 years had to include at least 600 minutes (10 h/d) of wear time as from Denmark, Spain, Germany, and Northern Ireland.25 in previous studies.27 Activity counts, accelerations due to Participants were eligible if they were able to walk for ≥2 min- body movement, were classified using set of cutpoints to utes (with or without a walking aid), scored four or above calculate the amount and intensity of SB and physical activ- on the Short Physical Performance Battery (SPPB), reported ity.27 SB was classified as <100 counts per minute (CPM), undertaking regular MVPA ≤30 minutes on 5 days per week, daily LPA was 100-2019 CPM, and daily MVPA was >2020 and/or spend 6-8 hours per day sitting (eg, watching TV or CPM. Values are the average daily time spent in SB, LPA, working at the computer). Participants were excluded if they and MVPA.28 had three or more errors on a six-item cognitive impairment questionnaire to identify moderate or severe dementia, had a medical condition that contraindicated to undertaking an ex- 2.2.2 | Mental health ercise intervention such as unstable medical conditions (eg, uncontrolled hypertension), could not commit to attend 75% Self-rated mental health was measured using the Hospital and of the intervention exercise sessions, and/or had participated Anxiety Depression Scale (HADS)29 is a self-reported ques- in an exercise program in the six months prior to their entry tionnaire consisting of 14 items split across anxiety and de- into the study. A total of 2660 older adults were recruited pression. A high score indicates high psychological distress. across the four countries via media (letters and social media), general practitioners, and other health professionals from pri- mary care, and senior center community groups. Of those, 2.3 | Statistical analysis 45.15% did not participate in the study (27.3% of those were excluded based on eligibility assessment). A total of 1360 We investigated the hypothesis that replacing SB with older adults provided baseline data. Further details can be physical activity (LPA or MVPA) results in improvements found in the study protocol.25 SITLESS has been approved in HADS anxiety and HADS depression scores. All anal- by the Ethics and Research Committee of each intervention yses were carried out in Mplus (version 7.4; Muthen & site: The Ethics and Research Committee of Ramon Llull Muthen, Los Angeles, CA). A robust form of weighted least 1960 |    TULLY et al. squares (WLSMV) was used to correct for non-normaility. Since the total hours in a day are fixed, SB, LPA, and WLSMV was used to optimize the relationship between the MVPA are inter-related where increasing one results in the dichotomous variable and the latent construct as this makes reduction in another. These substitutional relationships need the relationship between the factor(s) and the observed vari- to be considered when considering how a particular activ- able non-linear, with a Probit link function. A model-based ity is associated with health outcomes. The IS approach estimation strategy involves using the fit indices to under- enables researchers to simultaneously model a specific ac- stand how well the model describes the data and missing tivity being performed and an activity being displaced in an data is based on the entire data set including observations equal time-exchange manner. The single-activity model an- with missing data. Missing data were assumed to be missing alyzes SB, LPA, and MVPA separately without considering at30 random and a robust WLSMV approach uses a model- the other activity types, adjusting for total wear time and co- based estimation strategy to address missing data.31 variates. The partition model analyzes SB, LPA, and MVPA To address the inconsistency between studies regarding simultaneously, without adjusting for total wear time. The the dimensionality of the HADS depression and HADS anx- coefficient for one type of activity represents the effect of iety subscales,32 a bifactor model including all items loaded increasing this type of activity while holding the other ac- onto a general distress factor (GD), an anxiety (HA), and a tivities constant in this model. The model represents the depression (HD) factor as outlined by Norton et al33 was used effects of adding rather than substituting an activity type, (Figure 1). Model fit was evaluated using a root mean square because the total wear time is not included in the model error of approximation (RMSEA) ≤0.05 with an upper limit (thus is not held constant). The IS model assesses the ef- (90% CI) ≤0.08; a comparative fit index (CFI) ≥0.95; a fect of substituting one activity with another for the equal Tucker Lewis index (TLI) ≥0.95; and a standardized root amount of time (eg, replacing LPA with SB, by removing mean square residual (SRMR) ≤0.05.34,35 Where the levels of SB from the model). It should be noted that as this data are fit indices were not achieved, the modification indices were cross-sectional, the observed effects of substitution are not examined, and where appropriate, the necessary adjustments temporal. were made. The single-activity model assessed the association be- Following this, three multiple linear regression models tween SB, LPA, or MVPA and mental health separately, comprising single-activity, partition, and IS models were adjusting for accelerometer wear time and other covariates conducted to examine the association between SB, LPA, and that are known to be associated with both activity and mental MVPA with both HADS anxiety and HADS depression. SB health (age, sex, marital status, educational attainment, phys- and physical activity (LPA and MVPA) were standardized ical health component of the 12-item Short Form Health (SF- using 30 minutes as a unit for activity in analyses. Research 12) questionnaire38). suggests that 30 minutes rather than 60 minutes of LPA or Following this, partition models were prepared, whereby MVPA is more feasible for older adults,36,37 so IS models the coefficient for one activity type is the effect of increas- assessed the effect of replacing 30 minutes of one form of ing that activity type while holding all others constant. The activity with the equal time of another. model represents the effects of adding, not substituting a F I G U R E 1 HADS bifactor model.33 GD, general distress; HA, anxiety; HD, depression; HADS, Hospital and Anxiety Depression Scale; 1-14 are individual HADS question items TULLY et al.    | 1961 particular activity type because total wear time is excluded in For HADS anxiety, in both single and partition models, this model and so is not held constant. higher SB, LPA, and MVPA levels were associated with Finally, in IS models with total wear time included, one lower scores on HADS anxiety scale which in this model of the activities is eliminated from the model to ascertain the indicates lower symptoms of anxiety (Table 2). Replacing effect of replacing 30 minutes of one activity with 30 minutes 30 minutes of SB with LPA or MVPA in the IS model was of another activity. associated with significant improvements in anxiety, with apparently greater dividends from replacing SB with LPA (−0.37) than MVPA (−0.14). In addition, replacing 30 min- 3 | R E S U LTS utes of LPA with MVPA results in significant improvements in anxiety symptoms (Table 2). As displayed in Table 1, of the included participants The single model analyses showed that higher levels of (n = 1360), 62% were female and the mean age was 75 (6.3) both SB and MVPA were associated with higher scores on years. Participants showed on average low levels of anxiety HADS depression scale which in this model indicates higher and depression and good mental and physical health com- symptoms of depression (Table 3). In contrast, a higher level pared to normative values.38,39 of LPA was associated with a lower depression score. In the partition model, higher levels of SB and LPA were associated with higher levels of symptoms of depression (Table 3). In TABLE 1 Characteristics of study participants the IS model, replacing 30 minutes of SB with LPA (0.55), n or Mean (SD), % replacing 30 minutes of LPA with SB (0.58), or replacing 30 minutes of MVPA with either SB (0.38) or LPA (0.28) Sex was associated with increased depression (Table 3). Male 520 38.2% Female 840 61.8% Age (y) 75.18 (6.17) 4 | DISCUSSION Marital status Married 690 50.6% This study including 1360 older adults from four European Single (including widowed, divorced, 618 45.4% countries aimed to investigate the effect of substituting SB separated) with LPA or MVPA on associations with mental health in Education attainment older adults using an IS approach. Our results demonstrate Primary education 320 24.0% that in this cross-European cohort of older adults, substituting Secondary education 712 53.3% 30 minutes of either SB or LPA with MVPA was associated with improved symptoms of anxiety. In addition, replacing University 303 22.7% 30 minutes of SB with LPA was associated with improved Country of residence anxiety scores. The findings for symptoms of depression Denmark 338 24.9% were inconsistent across the single, partition and IS modeled. Germany 345 25.4% For example, the single model demonstrated that a more time Spain 356 26.2% spent in SB was associated with higher depression scores, but Northern Ireland 321 23.6% replacing SB with LPA was also associated with a signifi- Total accelerometer wear time per day 28.74 (2.33) cantly higher depression score. (30 min/d) For anxiety, the largest effect size was seen when substi- SB per day (30 min/d) 22.62 (2.52) tuting 30 minutes of SB with LPA (β −.37), which was over LPA per day (30 min/d) 5.36 (1.76) two and a half times that of substituting 30 minutes of SB MVPA per day (30 min/d) 0.76 (0.66) with MVPA (β −.14) and over three times that of substitut- SF12-physical score 44.96 (9.10) ing 30 minutes of LPA with MVPA (β −.11). To achieve the same effect size as substituting 30 minutes of SB with LPA HADS anxiety score 4.96 (3.65) would require substituting 79 minutes of SB or 101 minutes HADS depression score 4.05 (3.29) of LPA, respectively. The β value represents the effect of Note: In this table, a high score on HADS indicates good mental health, and a substituting 30 minutes of one type of PA for another on the high score on SF-12 indicates good physical health. Columns may not sum up to latent construct of HADS, which in our model is scored from total of 1,360 participants because of missing data in some variables. Abbreviations: HADS Anxiety, Depression, 19 items; LPA, light-intensity 1 to 4. So for example, by reducing SB by 30 minutes and re- physical activity; MVPA, moderate-vigorous intensity behavior; SB, sedentary placing it with LPA, the effect on HADS anxiety will be a de- behavior; SF12-Physical health component, 12 items. crease of 0.37, indicating better mental health. Consequently, 1962 |    TULLY et al. TABLE 2 The associations of SB, LPA, and MVPA with HADS anxiety score SB LPA MVPA β 95% CI β 95% CI β 95% CI * * * Single-activity model −.44 (−0.48, −0.39) −.39 (−0.44, −0.34) −.27 (−0.34, −0.19) * * * Partition model −.33 (−0.39, −0.26) −.16 (−0.22, −0.09) −.23 (−0.29, −0.17) Isotemporal model * * Replace SB with Dropped −.37 (−0.42, −0.32) −.14 (−0.21, −0.07) * * Replace LPA with −.42 (−0.47, −0.37) Dropped −.11 (−0.18, −0.04) * * Replace MVPA with −.33 (−0.39, −0.26) −.16 (−0.22, −0.09) Dropped Note: All models adjusted for covariates of age, sex, marital status, educational attainment, SF-12 physical health component. Single and substitution models adjusted for all covariates including total accelerometer wear time. β = Regression coefficients correspond to a 30 min increment of each activity. A high score on HADS Anxiety indicates high anxiety in this model. Abbreviations: LPA, light-intensity physical activity; MVPA, moderate-to-vigorous physical activity; SB, sedentary behavior. *P < .05. TABLE 3 The associations of SB, LPA, and MVPA with HADS depression score SB LPA MVPA β 95% CI β 95% CI β 95% CI * * * Single-activity model .59 (0.52, 0.65) −.30 (−0.51, .21 (0.10, 0.32) −0.43) * * Partition model .38 (0.29, 0.47) .28 (0.19, 0.37) −.02 (−0.09, 0.06) Isotemporal model * Replace SB with Dropped .55 (0.49, 0.62) .06 (−0.04, 0.16) * Replace LPA with .58 (0.51, 0.65) Dropped .04 (−0.06, 0.14) * * Replace MVPA with .38 (0.30, 0.47) .28 (0.19, 0.37) Dropped Note: All models adjusted for covariates of age, sex, marital status, educational attainment, SF-12 physical health component. Single and substitution models adjusted for all covariates including total accelerometer wear time. β = Regression coefficients correspond to a 30 min increment of each activity. A high score on HADS Depression indicates high anxiety in this model. Abbreviations: LPA, light-intensity physical activity; MVPA, moderate-to-vigorous physical activity; SB, sedentary behavior. *P < .05. a decrease of 0.37 within a range of 1 to 4 would be relatively recommendations have been based largely on epidemiolog- meaningful in terms of better mental health. ical evidence.5 However, the evidence from IS analyses has On the other hand, the results for depression did not follow shown inconsistent findings. Yasunaga et al23 have shown this pattern. Substituting 30 minutes of SB with LPA or LPA that substituting SB with either LPA or MVPA is associated with MVPA led to similar increases in the effect size for depres- with improvements in depression, but Curtis et al did not find sive symptoms. However, no effect was observed for substitut- this relationship for depression, anxiety, or stress.20 Dillon ing 30 minutes of SB with MVPA. This adds to our knowledge et al22 observed that reallocating 30 minutes of SB to LPA of the range of benefits for older adults if they replace SB with was associated with a decrease in anxiety symptoms but not LPA in terms of anxiety. This effect was considerably larger depression in 396 adults aged 50-69. than effects sizes for replacing SB with MVPA. However, for We have demonstrated that the message to sit less and depression, MVPA appears to still be important. move more for older adults would result in significant bene- Taken together, these findings support current health fit to mental health and symptoms of anxiety, but not symp- recommendations that inactive adults should be to initially toms of depression. Hallgren et al40 have posited that there encouraged to increase their LPA, which may be a more re- may be different associations with depression between dif- alistic and sustainable change. There has been a shift in the ferent types of SB, based on the mental stimulation they emphasis of national physical activity guidelines toward offer. They have proposed that mentally passive SB (eg, encouraging people to “sit less and move more” for a num- TV viewing) may be associated with depression whereas ber of health benefits, including mental health.12,13 These as mentally active SB (eg, reading) may be beneficial. This TULLY et al.    | 1963 may explain why we have shown mixed findings for symp- reflection of the actual physical activity carried out by older toms of depression. adults in a community setting. Additionally, in a study of 3233 US adults and older Multicollinearity is a disadvantage of this approach. adults, Del Pozo Cruz et al21 used compositional analysis to Although some researchers have suggested that a correlation show that reallocating 60 minutes of SB to MVPA was as- coefficient of >.6 should be carefully considered when in- sociated with small reductions in depression symptoms, but cluded in a model,47 others view the removal of confounding reallocating SP to LPA was not associated with changes in as more important and think that collinear variables of up depressive symptoms. The small effect sizes Del Pozo Cruz to .90 can still be reasonably controlled in various models.48 identified suggest that one reason we did not see an effect was Regardless, the goodness-of-fit tests for our models were that the time period of 30 minutes in the current study may be adequate, and results were still able to distinguish important too small to detect a change. differences in change in different substitution models. Although the recognized definition of SB is that it in- The use of IS in SB research is becoming increasingly cludes all time spent sitting during waking hours,11 it is popular as it offers novel insights into the potential benefits comprised of a range of behaviors, undertaken for vary- from changing behaviors with the cost or duration of an inter- ing different reasons. A systematic review of qualitative vention. There is some debate in the literature as to the most studies indicated that although older adults understand the appropriate way to approach developing IS models.18 We have health consequences of prolonged SB, they also recognized not accounted for sleep time in our analysis. Compositional that some SBs may confer cognitive or social benefits.41 data analysis seeks to address this by modeling substitutions A systematic review of 18 studies (n = 40 228) concluded to and from sleep as well as the other behaviors, which has that there is still some uncertainty about the relationship been proposed to be a more robust method.49 In response, between sedentary behavior and cognition in older adults Mekary et al50 have highlighted that the statistical properties but posited that the underlying mechanisms are likely to of IS models encompass those of models that have been de- be similar to physical activity.42 There is a need for fur- rived from the approach, such as compositional data analysis ther research to examine the relationship between different and that IS models convey their findings in absolute terms domains of SB and cognitive and mental health in older which make them easier to communicate and inform physical adults. The increasing use of accelerometers to measure activity guidelines. Irrespective of which statistical model is SB is improving the quality of the data,43 but questionnaire used, there is a need to test the assumptions in the model data have a role in complementing Actigraph measured on prospective longitudinal data. The data included in this data to allow an understanding of the contribution of differ- study were cross-sectional; therefore, causal inferences about ent types of SB. Furthermore, Actigraph measured physical the associations that were observed cannot be made. It is not activity, while providing a measure of overall activity, do possible to rule out reverse causality, or the possibility that an not allow for the comparison of different types of physical unmeasured confounding factor may explain the associations activity (group exercise, active travel, domestic physical observed. activity, outdoor recreation) which conceivably could also differentially impact on depression through the associated social benefits.44 5 | PERSPECTIVE A key strength of this study was the use of confirmatory factor analysis (CFA) (configural, metric, and scalar invari- Regular physical activity in older adults is associated with a wide ance) to assess the reliability and the variance of each ob- range of health benefits.1,2 Current public health recommenda- served measure within the composite measures of HADS. tions encourage older adults to “sit less and move more.”12,13 This approach addresses unreliability and consequently Using IS, we have demonstrated that replacing 30 minutes of measurement error and so increases the robustness of the SB with LPA or MVPA was associated with improved men- findings.45,46 tal health in older adults. Greater benefits were observed when Also, this study measured physical activity using an accel- shifting SB and LPA to MVPA. These results should be used to erometer-based measure of activity rather than self-report in encourage older adults for a more physical active lifestyle. order to provide objective evidence of links with depression, anxiety, and mental health. Self-report measures are more ACKNOWLEDGEMENTS commonly used in observational epidemiology studies ex- The authors wish to express their appreciation for all those ploring physical activity and health outcomes. Consequently, who volunteered for this study and Mrs Nicky Laird for her bias caused by using self-reported physical activity measures administrative support. We also thank the Northern Ireland such as social desirability or problems in recall which limit Clinical Research Facility and the Municipality of Odense the accuracy of self-reports in older adults are removed. for facilitating the assessments in the United Kingdom and Therefore, the results from this study provide a more accurate Danish site, respectively. 1964 |    TULLY et al. The work described in this publication was part of 2. Cunningham C, O' Sullivan R, Caserotti P, Tully MA. the SITLESS project, and supported and funded by the Consequences of physical inactivity in older adults: a systematic review of reviews and meta-analyses. Scand J Med Sci Sports. European Union program Horizon 2020 (H2020-Grant 2020;30(5):816-827. 634270). Consortium members of the participating organi- 3. Schuch FB, Stubbs B, Meyer J, et al. Physical activity protects from zations of the SITLESS project: Antoni Salvà Casanovas, incident anxiety: a meta-analysis of prospective cohort studies. Àlex Domingo, Marta Roqué, and Laura Coll-Planas: Depress Anxiety. 2019;36(9):846-858. Health and Ageing Foundation of the Autonomous 4. Schuch FB, Vancampfort D, Firth J, et al. Physical activity and University of Barcelona, Barcelona, Spain; Maria Giné- incident depression: a meta-analysis of prospective cohort studies. Garriga, Miriam Guerra-Balic, Carme Martin-Borràs, Am J Psychiatry. 2018;175(7):631-648. 5. Huang Y, Li L, Gan Y, et al. Sedentary behaviors and risk of de- Javier Jerez-Roig, Guillermo R. Oviedo, Marta Santiago- pression: a meta-analysis of prospective studies. Transl Psychiatry. Carrés, Oriol Sansano, and Guillermo Varela: Faculty of 2020;10(1):26. Psychology, Education, and Sport Sciences Blanquerna, 6. Hoare E, Milton K, Foster C, Allender S. The associations between Ramon Llull University, Barcelona, Spain; Emma sedentary behaviour and mental health among adolescents: a sys- McIntosh and Manuela Deidda: Health Economics and tematic review. Int J Behav Nutr Phys Act. 2016;13(1):108. Health Technology Assessment, University of Glasgow, 7. Hallgren M, Nguyen TT, Owen N, et al. Associations of sedentary Glasgow, United Kingdom; Dietrich Rothenbacher, behavior in leisure and occupational contexts with symptoms of Michael Denkinger, Katharina Wirth, Dhayana Dallmeier, depression and anxiety. Prev Med. 2020;133:e106021. 8. de Rezende LF, Rey-Lopez JP, Matsudo VK, Luiz OC. Sedentary and Jochen Klenk: Institute of Epidemiology and Medical behavior and health outcomes among older adults: a systematic re- Biometry, Ulm University, Ulm, Germany; Frank Kee: view. BMC Public Health. 2014;14:333. Centre for Public Health, School of Medicine, Dentistry and 9. Giné-Garriga M, Sansano-Nadal O, Tully MA, et al. Accelerometer- Biomedical Sciences, Queen's University Belfast, Belfast, measured sedentary and physical activity time and their correlates United Kingdom; Mark A. Tully, Jason J. Wilson, Nicole E. in european older adults: the SITLESS study. J Gerontol Series A. Blackburn, and Ilona McMullan: School of Health Sciences, 2020. Ulster University, Northern Ireland, United Kingdom; 10. Smith L, Gardner B, Fisher A, Hamer M. Patterns and correlates of Paolo Caserotti and Mathias Skjødt: Department of Sport physical activity behaviour over 10 years in older adults: prospec- tive analyses from the English Longitudinal Study of Ageing. BMJ Science and Clinical Biomechanics, University of Southern Open. 2015;5(4):e007423. Denmark, Odense, Denmark; Guillaume Lefebvre: SIEL, 11. Tremblay MS, Aubert S, Barnes JD, et al. Sedentary Behavior Sport Initiative et Loisir Bleu association, Strasbourg, Research Network (SBRN) – terminology consensus project pro- France; Denise González: SIEL, Sport Initiative et Loisir cess and outcome. Int J Behav Nutr Phys Act. 2017;14(1):75. Bleu Association, Barcelona, Spain. 12. Piercy KL, Troiano RP, Ballard RM, et al. The physical activity guidelines for Americans. JAMA. 2018;320(19):2020-2028. ORCID 13. Department of Health and Social Care. UK Chief Medical Officers' Physical Activity Guidelines. London, UK: Department of Health Mark A. Tully https://orcid.org/0000-0001-9710-4014 and Social Care; 2019. Nicole E. Blackburn https://orcid. 14. Smith L, Ekelund U, Hamer M. The potential yield of non-exercise org/0000-0003-4379-6312 physical activity energy expenditure in public health. Sports Med. Jason J. Wilson https://orcid.org/0000-0001-5558-8399 2015;45(4):449-452. Lee Smith https://orcid.org/0000-0002-5340-9833 15. Gupta N, Mathiassen SE, Mateu-Figueras G, et al. A comparison Frank Kee https://orcid.org/0000-0002-0606-8167 of standard and compositional data analysis in studies addressing Maria Giné-Garriga https://orcid. group differences in sedentary behavior and physical activity. Int J org/0000-0003-4449-3524 Behav Nutr Phys Act. 2018;15(1):53. 16. Mekary RA, Lucas M, Pan A, et al. Isotemporal substitution analy- Laura Coll-Planas https://orcid. sis for physical activity, television watching, and risk of depression. org/0000-0001-5204-8717 Am J Epidemiol. 2013;178(3):474-483. Dhayana Dallmeier https://orcid. 17. Mekary RA, Willett WC, Hu FB, Ding EL. Isotemporal substi- org/0000-0003-3665-7023 tution paradigm for physical activity epidemiology and weight Michael Denkinger https://orcid. change. Am J Epidemiol. 2009;170(4):519-527. org/0000-0002-8097-060X 18. Grgic J, Dumuid D, Bengoechea EG, et al. Health outcomes associ- Dietrich Rothenbacher https://orcid. ated with reallocations of time between sleep, sedentary behaviour, org/0000-0002-3563-2791 and physical activity: a systematic scoping review of isotemporal substitution studies. Int J Behav Nutr Phys Act. 2018;15(1):69. 19. Rethorst CD, Moncrieft AE, Gellman MD, et al. Isotemporal anal- R E F E R E NC E S ysis of the association of objectively measured physical activity 1. Bangsbo J, Blackwell J, Boraxbekk CJ, et al. Copenhagen with depressive symptoms: results from Hispanic Community Consensus statement 2019: physical activity and ageing. Br J Health Study/Study of Latinos (HCHS/SOL). J Phys Act Health. Sports Med. 2019;53(14):856-858. 2017;14(9):733-739. TULLY et al.    | 1965 20. Curtis RG, Dumuid D, Olds T, et al. The association between reduce sedentary time in older Scottish adults: results of a feasibil- time-use behaviors and physical and mental well-being in adults: a ity and pilot study. Prev Med. 2013;57(5):718-720. compositional isotemporal substitution analysis. J Phys Act Health. 37. Gardiner PA, Eakin EG, Healy GN, Owen N. Feasibility of reducing 2020;17(2):197-203. older adults' sedentary time. Am J Prev Med. 2011;41(2):174-177. 21. Del Pozo CB, Alfonso-Rosa RM, McGregor D, Chastin SF, 38. Ware JE, Kosinski M, Keller SD. A 12-item short-form health sur- Palarea-Albaladejo J, Del Pozo CJ. Sedentary behaviour is asso- vey: construction of scales and preliminary tests of reliability and ciated with depression symptoms: compositional data analysis validity. Med Care. 1996;34(3):220-233. from a representative sample of 3233 US adults and older adults 39. Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of assessed with accelerometers. J Affect Disord. 2020;265:59-62. the Hospital Anxiety and Depression Scale. An updated literature 22. Dillon CB, McMahon E, O'Regan G, Perry IJ. Associations be- review. J Psychosom Res. 2002;52(2):69-77. tween physical behaviour patterns and levels of depressive 40. Hallgren M, Dunstan DW, Owen N. Passive versus mentally ac- symptoms, anxiety and well-being in middle-aged adults: a tive sedentary behaviors and depression. Exerc Sport Sci Rev. cross-sectional study using isotemporal substitution models. BMJ 2020;48(1):20-27. Open. 2018;8(1):e018978. 41. Rawlings GH, Williams RK, Clarke DJ, et al. Exploring adults' 23. Yasunaga A, Shibata A, Ishii K, Koohsari MJ, Oka K. Cross- experiences of sedentary behaviour and participation in non-work- sectional associations of sedentary behaviour and physical activity place interventions designed to reduce sedentary behaviour: a on depression in Japanese older adults: an isotemporal substitution thematic synthesis of qualitative studies. BMC Public Health. approach. BMJ Open. 2018;8(9):e022282. 2019;19(1):1099. 24. Park CH, Elavsky S, Koo KM. Factors influencing physical activ- 42. Olanrewaju O, Stockwell S, Stubbs B, Smith L. Sedentary ity in older adults. J Exerc Rehabil. 2014;10(1):45-52. behaviours, cognitive function, and possible mechanisms 25. Gine-Garriga M, Coll-Planas L, Guerra M, et al. The SITLESS in older adults: a systematic review. Aging Clin Exp Res. project: exercise referral schemes enhanced by self-management 2020;32(6):969-984. strategies to battle sedentary behaviour in older adults: study pro- 43. Prince SA, Cardilli L, Reed JL, et al. A comparison of self-reported tocol for a randomised controlled trial. Trials. 2017;18(1):221. and device measured sedentary behaviour in adults: a systematic re- 26. Choi L, Liu Z, Matthews CE, Buchowski MS. Validation of accel- view and meta-analysis. Int J Behav Nutr Phys Act. 2020;17(1):31. erometer wear and nonwear time classification algorithm. Med Sci 44. Rapp K, Mikolaizak S, Rothenbacher D, Denkinger MD, Klenk J. Sports Exerc. 2011;43(2):357-364. Prospective analysis of time out-of-home and objectively measured 27. Migueles JH, Cadenas-Sanchez C, Ekelund U, et al. Accelerometer walking duration during a week in a large cohort of older adults. data collection and processing criteria to assess physical activity Eur Rev Aging Phys Act. 2018;15:8. and other outcomes: a systematic review and practical consider- 45. Bollen KA. Structural Equations with Latent Variables. New York, ations. Sports Med. 2017;47(9):1821-1845. NY, Chichester, UK: Wiley; 1989. 28. Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, McDowell 46. McCoach DB, Black AC, O'Connell AA. Errors of inference in M. Physical activity in the United States measured by accelerome- structural equation modeling. Psychol Sch. 2007;44(5):461-470. ter. Med Sci Sports Exerc. 2008;40(1):181-188. 47. McGee D, Reed D, Yano K. The results of logistic analyses when 29. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. the variables are highly correlated: an empirical example using diet Acta Psychiatr Scand. 1983;67(6):361-370. and CHD incidence. J Chronic Dis. 1984;37(9–10):713-719. 30. Finney S, DiStefano C. Non-normal and categorical data in 48. Pischon T, Girman CJ, Rifai N, Hotamisligil GS, Rimm EB. Structural equation modeling: A second courseNonnormal and Association between dietary factors and plasma adiponectin con- categorical data in structural equation modeling. In: Hancock GR, centrations in men. Am J Clin Nutr. 2005;81(4):780-786. Mueller RO (eds.), Quantitative Methods In Education and the 49. Biddle GJH, Edwardson CL, Henson J, et al. Associations of Behavioral Sciences: Issues, Research, And Teaching. Structural physical behaviours and behavioural reallocations with markers of Equation Modeling: A Second Course. Charlotte, NC: IAP metabolic health: a compositional data analysis. Int J Environ Res Information Age Publishing; 2013: 439-492. Public Health. 2018;15(10):2280. 31. Brown TA. Confirmatory Factor Analysis for Applied Research. 50. Mekary RA, Ding EL. Isotemporal substitution as the gold stan- New York; NY: Guilford Press; 2006. dard model for physical activity epidemiology: why it is the most 32. Coyne JC, van Sonderen E. No further research needed: aban- appropriate for activity time research. Int J Environ Res Public doning the Hospital and Anxiety Depression Scale (HADS). J Health. 2019;16(5):797. Psychosom Res. 2012;72(3):173-174. 33. Norton S, Cosco T, Doyle F, Done J, Sacker A. The Hospital Anxiety and Depression Scale: a meta confirmatory factor analy- How to cite this article: TullyMA, McMullan I, sis. J Psychosom Res. 2013;74(1):74-81. Blackburn NE, et al. Sedentary behavior, physical 34. Hoyle RH. Structural Equation Modeling: Concepts, Issues, and activity, and mental health in older adults: An Applications. Thousand Oaks, CA: Sage; 1995. isotemporal substitution model. Scand J Med Sci Sports. 35. Kline RB. Principles and Practice of Structural Equation Modeling. New York, NY: Guilford Press; 2015. 2020;30:1957–1965. https://doi.org/10.1111/sms.13762 36. Fitzsimons CF, Kirk A, Baker G, Michie F, Kane C, Mutrie N. Using an individualised consultation and activPAL feedback to