Given that most working adults spend around a third of their day at work, the occupational environment is likely to have a bearing on time spent sitting, sleeping and being active, and on subsequent health risks. Long periods spent sitting, sleep duration outside the recommended range (e.g., 7–8 h) and a lack of moderate to vigorous physical activity (MVPA) are associated with increased cardio-vascular disease risk, incident type 2 diabetes and premature mortality [1
]. The duration and timing (night/day) of paid work, as well as the demands of work in terms of seated or active tasks, have been associated with overall daily sitting, activity and sleep, and with transport and leisure-time physical activity. Hence relationships between these three behaviours (sitting, sleep, physical activity) are complex and likely to be intertwined with patterns and duration of work.
Several small studies have examined associations between occupation, work hours and sitting time. One Australian study has shown that call centre workers spend significantly more time in sedentary behaviour while at work, than office and customer services employees [6
]. Other studies have shown that managers and professionals report more occupational sitting time than technicians and trade workers, but that leisure sitting time is similar in these employee groups [7
]. Hours of paid work may also be influential, as one study has shown that women who work part-time sit less during the whole day than those who work full-time [9
]. There is also some evidence that more time spent sitting at work is associated with more time sitting at home, as shown by a study that found female University employees who sat most at work, also sat most for television viewing, on work and non-workdays [10
]. As most studies to date have involved small samples in selected workplaces, larger studies are required to clarify the relationships between work hours and sitting time in different occupational groups.
A recent review has examined associations between occupation and physical activity (PA). It found strong associations of leisure time physical activity with occupation type, with higher leisure-time MVPA in professional than in blue collar occupations in most studies, although findings were mixed. In the same review there was a negative association between hours worked and leisure time MVPA [11
]. The authors stressed that confounding factors such as hours of work, work demands, and work-related physical activity were often not accounted for [11
]. The issue of whether leisure-time PA is influenced or compensated by occupational activity is unclear. In a study of office, call centre and customer service employees in Australia, all employees engaged in more sedentary behaviour and less light intensity activity on work days than non-work days [6
] but MVPA was higher on work days than non-work days [6
]. Conversely, another study showed that MVPA in office workers was very similar on work and non-work days [12
]. The amount of leisure-time MVPA may also be influenced by the physical demands of work; however, findings have been equivocal, with high workplace PA associated with low [7
] or high levels of non-occupational PA [11
], while other studies showed no difference [13
]. These mixed results suggest that the relationship between occupation and non-work activity is currently unclear and should be investigated.
There is also evidence to suggest that occupational factors may impact sleep quality and duration. Data from the 2009 U.S. Behavioural Risk Factor Surveillance System showed that adults who worked more than 40 h per week were 65% more likely to report insufficient sleep (<7 h per day) than those who worked less than 40 h per week [14
] and that those working in the manufacturing sector were more likely than people in the general working population to sleep less than 6 h per night [14
]. Working night shifts has also been associated with short sleep duration, as one study found that 44% of night shift workers report sleeping less than 6 h per day compared with 30% of day shift workers [14
]. Thus, sleep duration may, not surprisingly, differ according to level and type of employment.
While it is apparent that patterns of sitting, sleeping and being active vary markedly across groups in different work settings, these patterns may be even more complex in women. This is because women are more likely to juggle paid and unpaid work, Australian Bureau of Statistics data show more women work part-time than men, and that women spend more time in household duties than men of the same work status (part-time or full-time) [15
]. There is also some evidence that women’s health may be particularly at risk from sedentary behaviour, as shown by associations of television viewing with metabolic syndrome [16
] and all-cause mortality [17
] in women, but not in men. Finally, the occurrence and nature of sleep disorders appears to be different in women and men [18
] and gender differences in achieving 150 min/week of MVPA have also been noted [11
Time spent sleeping, active and sedentary are mutually exclusive but as time in a day is finite these are likely to compete with or influence each other, particularly on work days where work tasks involve either active or sedentary behaviours. To date, no research has assessed whether time spent sitting, sleeping, and being physical active, differs among women in different occupational groups. Given that there are now more women in the paid workforce [21
], it is important to understand patterns of sitting, sleep and PA in women, so that appropriate workplace health interventions can be developed. Data from the Australian Longitudinal Study on Women’s Health provide a unique opportunity to examine how occupation and work hours are associated with time spent in these three key health behaviours in a large population cohort of young and mid-aged working women. It is hypothesised that women in different occupations and with varied paid work hours may have different patterns of sleep, PA and sitting on work days, but that on non-work days their patterns will be similar.
The workplace is an opportune setting for health promotion programs, which aim to improve employee health status and productivity [28
]. Given that many women are now in the paid workforce [29
], and are juggling their work with unpaid family and caring responsibilities [15
], it is likely that time pressure might impact on the time available for sleep and physical activity. As many women are also in occupations such as office work, which require them to sit for long hours [6
], they may be exposed to a trifecta of behavioural risk factors, too much sitting, too little sleep and insufficient physical activity. This study examined how these behaviours vary in women with different occupations and work patterns, as improved understanding of the patterns of these behaviours in working women could help to identify ‘at risk’ groups for health promotion intervention. Differences in self-reported sitting time, sleep and PA were found between groups of women with different work characteristics in both the young and mid-aged cohort. As hypothesised, these differences occurred on work days, particularly. Notably on work days, full-time workers were most likely to report high sitting time, as were those in the managerial/professional or clerical/sales occupations. Work days were, however, different from non-work days in many ways. Not surprisingly, sleep durations were lower on work days than non-work days, most notably amongst those who worked shifts or nights, but also in full-time workers across occupational groups. Physical activity was not reported separately for work and non-work days, but over the week. Nonetheless, work related differences were still noted, with mid-age full-time workers and younger clerical/sales workers least likely to meet the PA guidelines.
4.1. Sitting Time
Average sitting times were comparable with other Australian data [31
] yet higher than those reported internationally [32
]. The higher levels of sitting reported in the current study are likely to reflect the multiple domain measure, which provides estimates that are about 2 h higher (per day) than when a single item sitting estimate, such as the International Physical Activity Questionnaire is used [33
]. On workdays, differences in sitting time followed expected patterns in both the young and mid aged cohorts; managers/professionals and clerical/sales workers, full time workers and regular pattern workers reported higher durations of sitting time, as has been previously shown [31
] and this appears to be driven by high work sitting on work days (Figure 1
). The amount of sitting time reported by younger women who work full-time with regular hours appeared to be particularly influenced by working longer hours, which often equate with more work-based sitting [36
]. Non-work day sitting did not differ among any of the work-related groups, even for those who sat less on work days (trades/production/labourers). Differences in non-work day sitting have not been consistently observed in previous studies [31
]. It appears that differences in sitting time in employed women are largely due to sitting for work, although this should be confirmed by studies using objective measures of sitting time.
In both cohorts, average sleep duration was consistent with previous Australian reports [37
], which have shown that shift-work is typically associated with shorter sleep durations and poorer sleep quality [39
]. This was borne out by our findings that shift and night workers in both cohorts were mostly likely to report shorter sleep on work days. In both cohorts, average sleep times were longer on non-work days than workdays, which may be a strategy for trying to catch up on sleep.
4.3. Physical Activity
Previous reviews have identified that blue collar workers report higher levels of occupational PA and lower levels of leisure time PA than white collar and professional workers [11
]. However, in the current study, after adjusting for covariates, it was not the ‘blue collar’ (used to describe the trades/production/labourer category in this study) workers, but the clerical/sales workers, who were significantly less likely to be physically active, when compared with managers/professionals in the young cohort. This finding is consistent with a previous report which found that managers/professionals were more likely to be active than other occupational groups, in women of similar age [41
]. In the mid-aged cohort, there was no association between occupational category and physical activity, although full-time workers were less likely (than part-time) to achieve physical activity recommendations. This might be expected as there is potentially more leisure-time available with part-time work. As the nature of the association between occupational category and physical activity is inconsistently observed in other studies [7
], it may be that life-stage is important. Even the mid-age women who worked full time were surprisingly active; this may reflect their lesser involvement in child care and household work than the younger women. Others have reported that MVPA declines with age [20
], but previous ALSWH reports have shown that women tend to increase their activity levels when children leave home [42
]. The presence of young children may be a factor in the differences noted in achieving physical activity guidelines between the young and mid-aged cohorts. In contrast to the mid aged cohort, full-time workers in the young cohort were more likely to achieve guidelines in unadjusted analyses, but this finding was diminished after adjustment for demographic variables. This may have been due to the fact that part-time workers in the young cohort were more likely to have children in the household (85% compared with 30% for full-time workers; Table 1
). Previous research has shown that having young children decreases participation in physical activity [43
4.4. Similarities and Differences between Cohorts
In terms of which work groups reported more sitting, short or long sleep, and not meeting physical activity guidelines, the two cohorts were remarkably similar. For both cohorts, manager/professionals and clerical/sales workers reported more sitting than trades/labourers, full-time more than part-time and those who worked regular hours more than shift/night workers. For sleep, those working shifts or at night were more likely to report short sleep on work days in both cohorts. Clerical/sales workers in the young cohort and the full-time workers in the mid-aged cohort were less likely to achieve physical activity guidelines. Interestingly, both these groups also reported higher sitting times, so it is possible that these groups are at even greater health risk as high sitting time and low physical activity have detrimental effects on health and mortality [45
]. Patterns of sleeping and sitting across work and non-work days were also similar in the two cohorts. For most groups, participants reported sitting less and sleeping longer on non-work days than work days. It is possible that away from the restrictions imposed by work conditions and weekday schedules that women engage in less sitting time and attempt to increase sleep duration on non-work days. However, attempts to catch up sleep on non-work days by sleeping for longer may also present other challenges to maintaining a healthy sleep schedule and wellbeing [46
]. Twenty-four-hour data collection would clarify the very variable movement and sleep patterns on non-work and work days.
Several limitations should be acknowledged when interpreting the findings of this study. As with all cohort studies, there has been attrition over time, so that the included women were less representative of the general population than the starting cohort in 1996 and particularly in the young cohort, there is a bias towards more educated women [47
]. However, the very large sample size and inclusion of women from across all occupational categories and work patterns is a strength. Another limitation is that all the data were self-reported, though the measures of PA and sitting have acceptable psychometric properties. In relation to sleep, we did not include measures of sleep quality, timing, chronotype or presence of sleep disorders. As the study only examines the associations cross-sectionally we cannot infer causality.
The categories used for occupation type were broad and this may lead to some dilution or masking of effects. For example, clerical, sales and retail were collapsed into one category; however, we know that office workers (likely to include clerical workers) sit for longer at work than retail workers [6
While most of the findings reported here appear to reflect ‘conventional thinking’, they do provide insight for targeting public health campaigns. Occupational characteristics were associated with higher sitting, shorter sleep and less activity, suggesting an important role for workplaces in promoting healthier behaviours. Differences in sitting time are largely driven by sitting for work and those groups that experience high sitting time at work (manager/professionals, clerical/sales and full-time workers) should be targeted. Workplace modifications including sit-stand desks have been shown to be effective at reducing workplace sitting [48
]. Those at health risk from short sleep are the shift and night workers, who may benefit from cognitive behavioural sleep programs [49
]. The groups at risk from low physical activity varied in the two cohorts. The combination of low physical activity and high sitting time in young clerical sales workers and mid-aged full-time workers was of particular concern. These workers could benefit from a two-pronged approach to increase physical activity and decrease sitting time, perhaps replacing one behaviour with the other, which has been shown to have a beneficial effect in studies that used 24 h monitor data [51