Diurnal patterns of objectively measured physical activity and sedentary behaviour in older men

Participants

The British Regional Heart Study (BRHS) is a prospective cohort of 7735 men recruited from a single local primary care centre in 24 British towns in 1978-80 (age 40-59 years). In 2010-2012, all surviving cohort members resident in the UK (n = 3137) were invited to attend a further physical examination including measurements of weight and height and to participate in a study of objectively measured physical activity. The National Research Ethics Service (NRES) Committee for London provided ethical approval. Participants provided informed written consent to the investigation, which was performed in accordance with the Declaration of Helsinki [13].

Objective physical activity assessment

Diurnal patterns of PA and SB

The intensity of PA indexed by both accelerometer CPM and the number of steps peaked around 10 am and then declined until a small further increase at 2 pm followed by a long decline until 9 pm and then a small increase after 10 pm (Fig. 2). Figure 3 shows the average percentage of time spent in PA of different intensities (SB, LIPA and MVPA) for each hour of the day. LIPA and MVPA were positively correlated within each hour: r = 0.21 (p < 0.001), and conversely, hourly SB was negatively correlated with LIPA (r = -0.88, p < 0.001) and with MVPA (r = -0.64, p < 0.001).

The pattern in daily variation of MVPA and LIPA closely followed the pattern observed for CPM. At around 10 am, the proportions of each hour spent in LIPA and MVPA peaked (at approximately 30 % and 8 % of all activity respectively) and then declined until 1 pm, followed by a slight increase in the afternoon around 2-3 pm and then a long decline until around 9 pm, when light activity accounted for only approximately 10 % and MVPA 1 % of each hour. Conversely, SB levels increased throughout the morning, with a steeper increase before 1 pm and then a small dip around 3 pm, followed by a slow increase to a peak of over 80 % spent in SB between 8-9 pm, followed by a slight decline after 9 pm (Fig. 3). Among men who completed log diaries, commonly reported activities around 11 am were gardening, shopping, moderate housework and do-it-yourself (DIY). Similar activities were reported around 3 pm, although more men reported gardening and fewer men reported housework in the afternoon.

Univariable descriptive plots show hourly patterns of different intensities of physical activity stratified by age, mobility limitations, chronic diseases, BMI, geriatric depression score and smoking status (Figs. 4 and 5) and by social isolation, social class, active transport and vision problems, season (winter vs summer) and day of the week (Figs. 6 and 7). In most cases, the patterns of mean percentage of SB, LIPA and MVPA by hour of the day followed a consistent pattern of peaks and dips at the same time of day. Descriptive statistics on sedentary and MVPA bouts (plots not presented) showed consistent results with patterns in Figs. 4, 6, 7. 49.0 % of the sedentary bouts lasting ≥60 min over a valid week occurred in the evenings; most started between 8-9 pm (13.6 %) or 9-10 pm (14.0 %). Conversely, most (59.5 %) of MVPA bouts lasting ≥10 min over a valid week occurred in the morning and in particular when the peaks of MVPA were reported, at 10 am (15.9 %) and 11 am (16.4 %).

Associations of social and demographic factors with PA and SB

The MVPA results were reported as RRs rather than beta coefficients, due to non-normality of the outcome distribution. The decline in MVPA was particularly marked over the course of the day; compared to the morning, levels of MVPA (>1040 CPM, Table 3 Model 1) declined substantially in the afternoon (RR = 0.75, 95 % CI 0.54-0.59) and in the evening (RR = 0.17, 0.15-0.18). Additionally, men with moderate or more severe mobility limitations compared to the reference category (no mobility limitations) had a RR of 0.50 (95 % CI 0.43, 0.57) for MVPA indicating that men with mobility limitations were half as likely to spend time doing MVPA compared to people with no limitations. Moreover, men who were older, did not use active transport, were obese, depressed, had more chronic health conditions, and were smokers had lower levels of MVPA.

Similar associations were seen when these analyses were repeated with a higher cut point (1952 CPM) to define MVPA (Table 3, Model 2). The largest RRs (risks of having low MVPA levels) were for being over 80 compared to less than 75 years, for the category “moderate/severe limitations or unable to do” if compared with no mobility limitations and use of active transport versus car/public transport. However, when using the 1952 CPM cut-point, MVPA level no longer differed by depression or smoking status.

Factors associated with modified diurnal patterns of PA and SB

Four factors were significantly associated with each of SB, LIPA and MVPA levels: age, mobility limitations, chronic diseases, and BMI. The effects of older age, obesity, mobility limitations and chronic diseases on LIPA, MVPA and SB appeared to be more marked in the morning than in the afternoon and evening, independent of the lower overall levels of PA observed in these subgroups. Interaction tests were performed to establish whether these associations differed by period of the day. The tests for interaction (Wald tests) were all statistically significant (p < 0.05 for age, chronic conditions and BMI and p <0.001 for mobility limitations).

Comparisons with other studies

To date there has been little work using hourly accelerometer data to examine diurnal physical activity patterns among older adults. Our results showing that physical activity peaks in the morning are consistent with recent studies [2, 4, 6–8, 11], although our study extends the literature by investigating more intensities of activity (SB, light PA and MVPA) and bouts of activities. In our study the overall PA levels (measured as CPM) over the course of the day were similar to other smaller studies using the same measurement device [4, 6–8, 11]. We examined PA patterns between 7.00 am and 10.59 pm and a similar period (6 am-10 pm or 7 am-9 pm) was analysed in other studies due to sparse data in early morning and late evening [2]. In line with our findings, a study of 38 healthy active adults (mean age 70 years) reported significantly fewer minutes of MVPA in the evening than in the morning or afternoon [11] and that longer bouts of activity occurred in the morning (6 am-12 pm) more often than afternoon or evening. The AGES-II study of 579 adults aged 73–98 from Iceland reported that the majority of PA occurred between 8 am and 4 pm on an average day [4], which fits with our findings. In line with our study, they also reported that sedentary time was similar across all age groups, except for the oldest age group (>85 years old) who were the most sedentary and PA levels declined with increasing age and BMI, but other social and health factors were not taken into account. Our findings about age modifying the daily patterns in physical activity fit with data from the AGES-II study [4] and from the Baltimore Longitudinal study of Ageing [6], which also found that older age groups had a steeper decline in in PA levels over the course of the day. Our finding that activity levels were lower in the mornings in obese than normal weight men mirrors data from a study of Canadian adults aged 20-79 years [2]. To our knowledge other studies have not investigated how presence of chronic conditions and mobility limitations affect the diurnal patterns of physical activity and sedentary behaviour in older adults.

Strengths and limitations

This study investigates how hourly levels of objectively measured SB, LIPA and MVPA vary over the course of the day and how daily activity patterns are modified by a wide range of demographic and health characteristics. It is particularly important to investigate LIPA in population based samples of older adults, because of the high proportion of time spent in light activity. Our findings about the correlates of LIPA offer a new contribution to the ongoing debate about whether and how the PA guidelines should include recommendations on LIPA as well as MVPA [27, 28].

This study benefits from using a large scale population-based cohort of free-living older men rather than a special at risk population, which should increase generalizability. The response rate achieved in this study is comparable with other studies on objective measurements of daily physical activity patterns [4]. Men who did not accept our invitation were about two years older and had higher BMI measured 10 years earlier; implying that overall PA (e.g. total counts or number of steps) might be lower in the general population. Our study is however limited by studying only white European men, who, based on existing literature, would be expected to have higher levels of PA, particularly MVPA, compared with women [9]. Therefore our results may not be generalizable to older women or ethnic minority populations [29]. Our study did not report detailed information on mode of activity, which was self-reported only during the first three days of accelerometer wearing. However, the importance of this information is recognized and future studies could investigate further the particular types of activities carried out during the highest and lowest peaks of activity [30]. A further area for future study would be the assessment of seasonal patterns of PA and SB (although these were not an objective of this study). Future analysis is needed to determine whether or not seasonal variations in PA and SB are observed even after adjustment for confounding of weather variables (e.g. temperature or sunshine duration). We investigated whether diurnal patterns in PA and SB were modified by season (winter vs. summer) and day of the week (Sunday vs Saturday or Monday-Friday). Whilst PA levels are generally lower (and SB levels higher) in winter and on Sundays, but we did not find evidence of effect modification on diurnal patterns.

Implications

The marked variations in PA occurring on a within-day basis provide information which could be helpful in planning interventions to increase PA levels. Older adults do most of their MVPA and light activity during the morning. Thus, one possible strategy for interventions aiming to increase these intensities of activity would be either to focus on the morning when people are already active and when variability in activity levels are greatest, aiming to increase the intensity or duration of existing physical activity bouts. Alternatively, interventions could focus on the afternoon period, aiming to stimulate physical activity of comparable intensity to that occurring in the morning. It is unlikely that low levels of activity in the evening can be changed, particularly in the winter months if it is dark in the late afternoons and evenings. Indeed, the combination of darkness and visual problems have been previously investigated as potential causes of falls [31]. Likewise with sedentary behaviours, our findings suggest that the period in the late afternoon and early evenings are periods with high levels of SB and when bouts of SB are likely to be longest, so it may be particularly valuable to focus on efforts to break up long sedentary bouts at these times of day. Our investigation showed that age and health status affected these diurnal patterns suggesting that PA policies might be targeted by sub-groups. Among older and disabled men, lower levels of MVPA were observed in morning and afternoon than in younger healthy men, the morning peak was more reduced than the afternoon peak, suggesting that with increasing age, the higher morning peak in moderate to vigorous activity may be particularly difficult to maintain. Longitudinal analyses could offer additional insights and determine if there are independent effects on health of MVPA or SB at different times of the day.