Associations of objectively measured physical activity and sedentary time with all-cause mortality in Japanese older adults: a 10-year prospective study

    To examine the associations of accelerometer-measured physical activity and sedentary time with all-cause mortality in older Japanese adults.

    A total of 1723 independent Japanese adults aged ≥65 years were followed from 2011 to 2021. Moderate-to-vigorous physical activity (MVPA), light physical activity (LPA) and sedentary time were measured using a triaxial accelerometer secured to participants’ waists.

    Over a median follow-up of 9.9 years, 336 deaths were recorded. When examined as tertiles, higher MVPA (both ≥10 and <10 min bouts) and LPA were associated with a lower mortality risk. Additional adjustment for MVPA attenuated the associations of LPA, but the HRs in the highest tertile remained significant. Longer sedentary time was significantly associated with an increased mortality risk, but not after adjusting for MVPA. In spline analyses, a linear dose–response association with all-cause mortality was observed for LPA, while the HRs declined progressively with higher levels of MVPA until approximately 80 min/day, beyond which they levelled out. Replacing 10 min/day of sedentary time with MVPA but not LPA was associated with a 12% lower risk of morality (HR 0.88; 95% CI 0.83 to 0.93); the HR for replacing 10 min/day LPA with MVPA was 0.89 (95% CI 0.84 to 0.95). These results were materially unchanged when excluding deaths within the first 5 years of follow-up.

    Physical activity, regardless of intensity, was associated with a lower all-cause mortality risk among older Japanese adults. Replacing sedentary time or LPA with MVPA was associated with a lower mortality risk. The mortality benefit started from a low MVPA dose and additional benefits were associated with higher doses.

    Data are available on reasonable request. The datasets used and/or analysed during this study are available from the corresponding author on reasonable request.

    http://creativecommons.org/licenses/by-nc/4.0/

    This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    WHAT THIS STUDY ADDS

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    The WHO 2020 guidelines on Physical Activity and Sedentary Behaviour provide updated evidence-based recommendations for older adults.1 The guidelines acknowledge the benefits of light physical activity (LPA) and the detrimental effects of sedentary behaviour.1 These updates reflect an evolving evidence base, particularly the growing number of studies using device-based assessments of physical activity and sedentary time, which reduce the well-known biases (eg, recall bias) associated with self-reported measurement.2 3 However, evidence supporting physical activity guidelines for older adults is primarily based on studies including middle-aged adults,1 despite differences in the dose-response health effects between younger and older adults.4

    Furthermore, epidemiological evidence on the associations of device-based measures of LPA and sedentary time remains limited and controversial.5–7 Recently, one meta-analysis of studies among older adults observed that accelerometer-measured moderate-to-vigorous physical activity (MVPA), LPA and sedentary time were significantly associated with mortality.8 However, this meta-analysis was limited by a short study period (all three studies regarding LPA had observation periods ≤5 years), which could result in reverse causality.8 In fact, a recent methodological study showed that the exclusion of the first 5 years of follow-up is needed to minimise the risk of reverse causation on estimates of the physical activity–mortality association, particularly for studies on LPA and sedentary time.9 However, prior studies typically excluded only the first 2 years of observations because of short study periods overall. Indeed, the WHO Guidelines Development Group recently revealed the dose–response association of LPA and/or sedentary behaviour with health outcomes as an important evidence gap.10

    Additionally, all studies using device-based assessments of different intensities of physical activity and sedentary time have been conducted in the USA and Western Europe.8 11 However, there exist differences in physical activity levels and their association with mortality risk across regions and countries.12 13 The few prior studies on the association between physical activity or sedentary time and mortality risk in East Asian older adults that do exist are primarily based on self-reported measurements,14–17 which are subject to measurement error and recall bias,2 3 or pedometers,18 which are unable to account for sedentary time and intensities of physical activity. Therefore, the lack of data on the associations of device-based measures of different intensities of physical activity and sedentary time and mortality risk in East Asia limits generalisability of the current evidence.

    Therefore, the main purpose of this 10-year prospective study was to examine the dose–response relationships of MVPA, LPA and sedentary time with all-cause mortality among community-dwelling older Japanese adults. The updated WHO guidelines indicate that MVPA need not occur in 10 min bouts,1 although few prospective studies have assessed how the duration of MVPA bouts relates to clinical endpoints, such as all-cause mortality.19 Hence, we also examined whether MVPA in bouts <10 min protects against mortality risks, or whether only ≥10 min bouts of MVPA is critical. The updated guidelines also indicate that replacing sedentary time with any intensity of physical activity (including LPA) has health benefits, a recommendation that is based largely on cross-sectional ‘replacement’ studies (ie, isotemporal substitution).1 We further examined whether replacing sedentary time with LPA is associated with a lower risk of all-cause mortality or whether replacing sedentary time with MVPA is needed to protect against mortality risk using an isotemporal substitution model.

    This prospective study used data from the Sasaguri Genkimon Study.20 21 Details of this study have been described elsewhere.20 21 Briefly, all residents aged ≥65 years without disability in Sasaguri town, a suburb in Fukuoka Prefecture in southern Japan, in January 2011 met the inclusion criteria (n=4979). After excluding subjects who had died or moved out of the district (n=66) by the onset of the study in 2011, 4913 subjects were invited and 2629 consented. Of these, 1723 participants were included in the study after excluding 17 with a medical history of dementia or Parkinson’s disease, 861 without valid accelerometer data and 28 without complete data on covariates (online supplemental figure 1). Online supplemental table 1 shows the characteristics of included participants and excluded subjects due to without valid accelerometer data in the present study.

    As of January 2011, the distributions of age, sex, education and occupation in Sasaguri town were similar to that of the general population in Japan.22 Our research team comprises four men and two women from China and Japan.

    Methods for baseline accelerometer data collection and processing have been previously described.21 Briefly, participants wore a waist-mounted triaxial accelerometer (Active Style Pro HJA-350IT, Omron Healthcare, Kyoto, Japan) for 7 days, removing it for sleeping or any water activities. The non-wear period was calculated using a modified SAS macro program provided by the National Cancer Institute.23 24 Accelerometer data were considered valid when there were ≥4 valid wear days (≥10 hours of wear time per day).25

    The intensity of activities was directly estimated as a metabolic equivalent (MET) every 60 s using a built-in algorithm of the accelerometer, which was developed and validated in adults aged 20–59 years.26 27 A validation study showed that the degree of correlation between predicted and actual METs in older adults (r=0.85, p<0.001) was comparable to younger adults aged 20–59 years (r=0.88, p<0.001), although the prediction errors were greater in older adults (−0.6±0.6 vs −0.1±0.5 METs in younger adults), particular at higher intensity activities.28 Sedentary time, LPA and MVPA were defined as time spent in activities of ≤1.5 METs, 1.6–2.9 METs and ≥3 METs, respectively.29 MVPA was categorised as bouts of ≥10 min (accumulated in sustained bouts ≥10 min with an allowance for up to 2 min below threshold) and <10 min. Time spent in moderate physical activity (MPA, 3–5.9 METs) or vigorous physical activity (VPA, ≥6 METs) was also calculated.29 We also calculated sedentary bout length by dividing total sedentary time by the number of sedentary bouts.30 Time of physical activities and sedentary time were averaged across valid days and expressed in min/day.

    The occurrence of all deaths from baseline to 31 March 2021 was ascertained from the Sasaguri municipal government office by checking local registries that had linked records with the Japanese National Vital Statistics System.

    Information on age and sex was obtained from the municipality office. Years of formal education, living alone (yes/no), current smoking status (yes/no) and current drinking status (yes/no) were obtained using a questionnaire. Multimorbidity was defined as having two or more of the following chronic diseases, as self-reported on the questionnaire: hypertension, stroke, heart disease, diabetes mellitus, hyperlipidaemia, respiratory disease, digestive disease, kidney disease, osteoarthritis or rheumatism, trauma fracture, cancer, ear disease and eye disease. We used the question ‘Do you normally walk continuously for 15 min?’ with an answer of ‘yes/no’ to determine low walking ability.31 Body mass and height were measured in light clothes without shoes, and body mass index (BMI) was calculated (kg/m2).

    Participants were censored at the date of death, loss to follow-up because of moving out of town or the end of the follow-up period (31 March 2021). Cox proportional hazard models were used to estimate HRs and 95% CIs for the associations of MVPA, LPA and sedentary time with all-cause mortality in the full sample and after excluding deaths within the first 5 years of follow-up to minimise the risk of reverse causality.

    To examine dose–response associations, all exposure factors were first analysed as tertiles. Model 1 was adjusted for sex, age and accelerometer wear time. Model 2 was additionally adjusted for education, living alone, BMI, multimorbidity, low walking ability, smoking and drinking. Model 3 was further adjusted for other intensity of physical activities to investigate whether (1) LPA and sedentary time were associated with mortality independent of MVPA and (2) MVPA was associated with mortality independent of sedentary time. Trends in mortality risk across the tertiles of physical activities and total sedentary time were tested by assigning ordinal numbers (0, 1, 2) to each tertile and treating the tertiles as a continuous variable. The above models were then repeated to examine the associations of MVPA bout length, MPA, VPA and sedentary bout length with the risk of all-cause mortality. Because 81.8% of the participants recorded zero min of VPA, we categorised participants into two VPA groups: one group included those who recorded zero min (referent group) and the other group included those who recorded any VPA.

    Interactions of sex and age (<72 and ≥72 years (median split)) with exposure factors were tested to examine potential modifying effects. Because recent evidence suggests that the effects of sedentary time depended on MVPA levels,32 we examined the interactions between MVPA levels (<150 and ≥150 min/week as per the current physical activity guideline,1 calculated as the average daily min of valid wear days multiplied by 7) and total sedentary time with adjustment for the above-mentioned covariates.

    To visualise the shape of the dose-response curves of each physical activity measure and sedentary time (as continuous variables) with all-cause mortality, and to test possible nonlinear associations, restricted cubic spline functions were used, with three knots placed at the 5th, 50th and 95th percentile.33

    The isotemporal substitution model was used to estimate the theoretical effects of replacing a defined duration of sedentary time for the same duration with equal time of LPA and MVPA while holding accelerometer wearing time constant.34 35 Specifically, LPA, MVPA and wearing time were included in a single Cox model (each expressed continuously in 10 (30 and 60) min units) with adjustment for the covariates in model 2. In this model, sedentary time was not included and the resulting HRs for LPA and MVPA represent the consequences of replacing a defined duration of sedentary time with an equal amount of time in a given type of activity (LPA or MVPA). For a complete isotemporal substitution analysis, we also examined replacing LPA or MVPA with other activities. Isotemporal substitution model has been widely adopted as a means to account for time displacement36 and is considered by some as the standard time-use statistical method for physical activity epidemiologic research.37

    We performed sensitivity analyses by censoring up to 31 December 2019 (the start of the COVID-19 pandemic) to completely exclude any possible impacts of the COVID-19 pandemic on the association during the follow-up period.

    All statistical analyses used SAS V.9.4 (SAS Institute). The significance level was set at two-sided α=0.05.

    The mean age of the full sample (n=1723) at baseline was 73.3 (rang: 65–93) years and 38.1% of the participants were men. Table 1 shows the participants’ baseline characteristics. Over a median follow-up of 9.9 years, 336 deaths were recorded, with 126 deaths occurring within the first 5 years of follow-up. As shown in online supplemental table 2, participants who died were older, more likely to be men, less educated, and living alone, had lower BMI and higher rate of multimorbility, and were more likely to smoke. Furthermore, those who died had lower levels of physical activity but higher levels of sedentary time and longer sedentary bout length.

    Table 1

    Baseline characteristics of the participants (n=1723)

    As shown in table 2, there was a significant inverse association between MVPA and all-cause mortality, with lower mortality risk in higher tertiles of MVPA in the full sample (all models, p for trend <0.0001). In the multivariable-adjusted model (model 2), HRs for all-cause mortality in the higher tertiles of MVPA were 0.53 (95% CI 0.40 to 0.69) and 0.40 (95% CI 0.28 to 0.57), respectively, compared with the lowest tertile. Adjusting additionally for sedentary time (model 3) did not change the results. After excluding deaths (n=126) within the first 5 years of follow-up, the association of MVPA remained significant (HRs per tertile in model 3: 1.00, 0.61, 0.38; p for trend <0.0001) (table 2).

    Table 2

    Associations of physical activity and sedentary time with all-cause mortality

    For both ≥10 and <10 min bouts, participants in higher MVPA tertiles had a significantly lower risk of all-cause mortality than those in the lowest tertile in all models (online supplemental table 3). Higher levels of MPA (all models) and any VPA (models 1 and 2) were also associated with a lower risk of all-cause mortality (online supplemental table 3). However, the association of VPA was attenuated to non-significant after adjusting additionally for sedentary time (model 3).

    For LPA, there was also a significant inverse association with all-cause mortality in the full sample (p for trend <0.01 in model 1) (table 2). In the multivariable-adjusted model (model 2), HRs for all-cause mortality in the higher tertiles of LPA were 0.96 (95% CI 0.75 to 1.23) and 0.53 (95% CI 0.38 to 0.75), respectively, compared with the lowest tertile (p for trend <0.001). Additional adjustment for MVPA (model 3) attenuated these associations (p for trend=0.06), but participants in the highest LPA tertile still had a significantly lower risk of all-cause mortality (HR 0.67; 95% CI 0.47 to 0.94; p=0.02). After excluding deaths within the first 5 years of follow-up, a consistent pattern of associations of LPA was observed (tertile 3 vs tertile 1 in model 3: HR 0.58, 95% CI 0.37 to 0.92; p=0.02) (table 2).

    Similar but directionally opposite results were observed for sedentary time (p for trend <0.01 in models 1 and 2) (table 2). However, when additionally adjusting for MVPA (model 3), this association was no longer apparent (tertile 2 vs tertile 1: HR 1.02; 95% CI 0.75 to 1.39; tertile 3 vs tertile 1: HR 1.14; 95% CI 0.80 to 1.62; p for trend=0.44). The pattern of association for sedentary time was similar after excluding deaths within the first 5 years of follow-up (table 2). Mean sedentary bout length was not associated with all-cause mortality (online supplemental table 3).

    The associations of physical activity and sedentary time with all-cause mortality did not vary by sex (p values for interactions: 0.42–0.67) (online supplemental table 4) or age groups (p values for interactions: 0.06–0.57). The associations of sedentary time or LPA with mortality also did not vary by MVPA time (p values for interactions: 0.13–0.34).

    Spline analyses yielded a more detailed dose-response curve (figures 1–3 and online supplemental figure 2–4). After adjusting for all covariates and sedentary time, the HRs declined progressively with higher levels of MVPA until approximately 80 min/day, beyond which no further risk reduction occurred in the full sample (figure 1A). After adjusting for all covariates and MVPA, LPA was significantly associated with all-cause mortality in a linear dose-dependent fashion in the full sample (figure 2A). After adjusting for all covariates and MVPA, sedentary time was not significantly associated with all-cause mortality in the full sample (figure 3A). Similar dose-response curves of MVPA and sedentary time were observed when data were analysed after excluding deaths within the first 5 years of follow-up (figures 1B and 3B and online supplemental figures 2B and 4B). In contrast, a non-linear significant association of LPA was observed after excluding deaths within the first 5 years of follow-up (figure 2B and online supplemental figure 3B).

    Figure 1

    Figure 1

    Dose–response associations of MVPA with all-cause mortality. Models were fit using cubic restricted splines, with three knots at the 5th, 50th and 95th percentiles. Lowest value of MVPA was set as the reference. Results are trimmed at the 1st and 99th percentiles and reported as HRs (black line) and 95% CIs (shaded area). Models were adjusted for sex, age, education, living alone, body mass index, multimorbidity, low walking ability, smoking, drinking, sedentary time, and wear time. MVPA, moderate-to-vigorous physical activity.

    Figure 2

    Figure 2

    Dose–response associations of LPA with all-cause mortality. Models were fit using cubic restricted splines, with three knots at the 5th, 50th and 95th percentiles. Lowest value of LPA was set as the reference. Results are trimmed at the 1st and 99th percentiles and reported as HRs (black line) and 95% CIs (shaded area). Models were adjusted for sex, age, education, living alone, body mass index, multimorbidity, low walking ability, smoking, drinking, MVPA and wear time. LPA, light physical activity; MVPA, moderate-to-vigorous physical activity.

    Figure 3

    Dose–response associations of sedentary time with all-cause mortality. Models were fit using cubic restricted splines, with three knots at the 5th, 50th and 95th percentiles. Lowest value of sedentary time was set as the reference. Results are trimmed at the 1st and 99th percentiles and reported as HRs (black line) and 95% CIs (shaded area). Models were adjusted for sex, age, education, living alone, body mass index, multimorbidity, low walking ability, smoking, drinking, MVPA and wear time. MVPA, moderate-to-vigorous physical activity.

    As shown in table 3, replacing 10 min/day of sedentary time with 10 min/day of LPA was not significantly associated with all-cause mortality risk in the full sample (HR 0.99; 95% CI 0.97 to 1.001, p=0.08). Replacement of 10 min/day of sedentary time with MVPA was significantly associated with a 12% lower risk of all-cause mortality (HR 0.88; 95% CI 0.83 to 0.93; p<0.0001). Each 10 min substitution of LPA with MVPA was associated with an 11% lower risk for all-cause mortality (HR 0.89, 95% CI 0.84 to 0.95; p<0.001). Similar results were observed in the isotemporal substitution analysis after excluding deaths within the first 5 years of follow-up (table 3). To account for non-linear associations between MVPA and mortality, we restricted the isotemporal substitution analysis among participants with MVPA<80 min in the full sample (n=1473; n of deaths=318). The cut-off was determined by visual inspection of the restricted cubic spline for MVPA to identify the group in which the association was approximately linear. Similar results were observed (online supplemental table 5). The HRs and 95% CIs for 30 and 60 min replacement in the full sample are detailed in online supplemental table 6.

    Table 3

    Multivariable adjusted HRs of all-cause mortality associated with 10 min changes in time spent in sedentary, LPA and MVPA

    When censoring up to 31 December 2019 to completely exclude any possible impact of the COVID-19 pandemic, the results remained robust (online supplemental tables 7 and 8).

    To our knowledge, we showed for the first time that higher levels of physical activity at any intensity were associated with a lower risk of all-cause mortality among older Japanese adults. By contrast, higher sedentary time was associated with a higher mortality risk, although this association did not persist after adjustment for MVPA. Moreover, higher MVPA was associated with a lower mortality risk, irrespective of bout duration (ie, <10 min or ≥10 min). Finally, we found that replacing sedentary time with MVPA was associated with a lower risk of all-cause mortality and that replacing LPA with MVPA was also associated with a lower risk of all-cause mortality. Consistent findings were observed after excluding deaths occurring within the first 5 years of follow-up.

    Our study found a curvilinear relationship between MVPA and all-cause mortality among older Japanese adults. The dose-response curve showed that the mortality benefit started from a dose well below the currently recommended minimum level (150 min/week), and that the maximum risk reduction was observed at approximately 80 min/day (equal to 560 min/week), beyond which benefits levelled out. This dose-response curve is consistent with a recent meta-analysis of accelerometer-measured MVPA and mortality based on a mixed sample of middle-aged and older populations from the USA and Western Europe.11 Compared with our study, this meta-analysis found the maximal risk reduction was seen at about 24 min/day (168 min/week).11 The authors also noted that wide CIs were observed at the end of the dose-response curve owing to the small number of people with high levels of MVPA. Therefore, this lower maximal threshold may reflect the fact that individuals in Western countries tend to be less physically active than those in East Asia.13 Indeed, the maximal risk reductions of all-cause mortality were seen at approximately 3000–4500 MET min/week of self-reported MVPA (ie, 1000–1500 min/week) in a recent 4-year study among Japanese older adults.16 Our study provides new evidence by demonstrating the detailed dose–response association between accelerometer-measured MVPA and mortality in older adults, suggesting that even a low level of MVPA can provide health benefits, and additional benefits are associated with higher levels.

    We found that MVPA, accumulated in either <10 min or ≥10 min bouts, was associated with a lower mortality risk. This finding adds to the existing limited evidence6 38 39 and supports recent changes to WHO guidelines that MVPA need not be accumulated in bouts ≥10 min. In addition, we found participants who engaged in any VPA had a lower risk of all-cause mortality after adjusting for multiple confounding factors, including sex, age and multimorbidity. However, most participants did not record any VPA, limiting the interpretation of these observations.

    One of the key findings of the present study was that LPA was associated with all-cause mortality among older adults, independent of MVPA. There have been controversial findings regarding associations between accelerometer-measured LPA and all-cause mortality with some studies showing significant associations6 40 41 and others showing no association.5 42 One recent meta-analysis showed that accelerometer-measured LPA was associated with mortality among older adults based on three studies with follow-ups of 2.3–5 years,8 raising concerns about reverse causality. Our study extends the existing evidence by demonstrating that the independent association of LPA is consistent in an older population with a 10-year follow-up and analyses excluding the first 5 years of follow-up as recommended to minimise reverse causation bias.9 However, the wide CIs surrounding the point estimates in the dose-response curve in the full sample make us uncertain about the magnitude of the effect. Thus, interpreting the present results also warrants caution.

    Although the accumulated evidence,6 30 40 41 including a recent meta-analysis,11 suggests higher levels of accelerometer-measured sedentary time are associated with a greater risk of all-cause mortality, even after adjusting for MVPA,11 a large body of evidence reports no association5 35 42 43 or an MVPA-dependent association.32 While our study found longer sedentary time was associated with an increased risk of all-cause mortality, this result was not independent of MVPA. Our isotemporal substitution analysis showed that sedentary time was associated with higher mortality risk when replacing MVPA, suggesting that the adverse effects of sedentary time on health outcomes are likely due to the displacement of physical activity, particularly MVPA.44 45 These findings are consistent with studies focused on mortality in adults aged 50 years or older35 or older women.5 46 For example, Schmid et al found that sedentary time was not associated with all-cause mortality risk after adjusting for physical activity, and that replacing 30 min of sedentary time with LPA or MVPA led to a reduction in all-cause mortality risk ranging from 12% to 49%.35 However, this study did not formally examine whether replacing physical activity with sedentary time would increase the risk of all-cause mortality.35 Our study, in conjunction with these findings, does not indicate that excessive sedentary time is an independent risk factor for mortality; however, it does support the notion that the harmful effects of excessive sedentary time may be due to low physical activity.47 48

    In addition, our isotemporal substitution analysis showed that replacing LPA with MVPA was significantly associated with lower mortality risk, which is in line with several studies conducted in adults aged 50 years or older35 39 49 50 or older women.46 For example, one of the studies using data from the UK Biobank found that a 10 min increase in MVPA coupled with a 10 min decrease in LPA per day was associated with a lower risk of all-cause mortality (HR 0.96, 95% CI 0.94 to 0.97).39 Overall, our study consistently supports the current WHO guideline that emphasises the need for older adults to engage in MVPA.

    Our findings provide further support for the updated guidelines that promote physical activity, particularly MVPA, among older adults. However, reductions in sedentary time as a stand-alone public health message should be promoted with caution.

    The major strengths of our study include its prospective cohort design, long follow-up period, relatively large, population-based sample and objective measurement of exposure factors using triaxial accelerometers. The comparable findings when excluding deaths occurring within the first 5 years suggest the findings are robust for reverse causation bias.

    First, the accelerometer used in this study could not distinguish between standing and sitting, which may have led to an overestimation of sedentary time and an underestimation of the association of sedentary time. Waist-mounted accelerometers may also underestimate METs during activities with little or no movement at the waist/hip. In addition, the built-in algorithm of the accelerometer used in the present study has been shown to underestimate the intensity of activities among older adults, particularly for MVPA,28 which may result in an underestimation of MVPA and the associations of MVPA. Second, exposure factors and covariates were measured only at baseline and their changes during follow-up were not considered. This may have biased our results towards the null. Third, although we accounted for a wide range of confounding factors, residual confounding may still exist (eg, health conditions or occupations that influence both physical activity and risk of mortality). Fourth, the exclusion of participants from the final sample mainly resulted from inadequate accelerometer wear time or days. Participants in the final sample may, therefore, have been more physically active than the overall sample, which could have caused an underestimation of the associations. Finally, we urge caution in generalising the current findings because the study was undertaken in a single town in Japan.

    This 10-year prospective study of older adults demonstrated that higher levels of physical activity, regardless of intensity, were associated with a decreased risk of all-cause mortality. The mortality benefit started from a low MVPA dose and additional benefits were associated with higher doses. Sedentary time was associated with higher mortality risk if it replaced MVPA. Replacing sedentary time or LPA with MVPA was associated with lower all-cause mortality. Taken together, the present findings support the updated WHO guidelines that ‘doing some physical activity is better than doing none’ in addition to reinforcing the importance of MVPA.

    Data are available on reasonable request. The datasets used and/or analysed during this study are available from the corresponding author on reasonable request.

    Consent obtained directly from patient(s).

    This study involves human participants and was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Fukuoka Institute of Technology, Japan (ethical approval code: hm03-20-01). We obtained written informed consent from all participants. Participants gave informed consent to participate in the study before taking part.

    We would like to thank Dr. Shuzo Kumagai who initiated the Sasaguri Genkimon Study, Ms. Yuka Haeuchi, Ms. Eri Shiokawa and the municipal staff in the primary caregiving division in Sasaguri who helped us coordinate the survey in the community.

    Read the full text or download the PDF: