Physical activity, sedentary time and sleep and associations with mood states, shift work disorder and absenteeism among nurses: an analysis of the cross-sectional Champlain Nurses’ Study

Introduction

Nurses are the largest occupational group in Canada’s health system including 282,300 publicly employed nurse managers and registered nurses (Infometrica Limited, 2017). Although nursing is perceived to be a physically active occupation, nurses have been found to spend more than half their time being sedentary on work days and very few achieve current physical activity guidelines (at least 150 min per week of moderate-to-vigorous intensity physical activity (MVPA) in bouts of 10 min or more) (Shields & Wilkins, 2006; Ratner & Sawatzky, 2009; Tremblay et al., 2011; Reed et al., 2018; Prince et al., 2019). Nurses working rotating shifts rarely obtain optimal amounts of sleep (i.e., at least 7 h) and have greater odds of reporting fair or poor health when compared to nurses who work days only (Shields & Wilkins, 2006; Rogers, 2008; Edward, 2009). Shift work, though necessary, may present a barrier to obtaining adequate sleep, being physically active and reducing sedentary time with the potential to affect nurses’ work performance and mental health (Atkinson et al., 2008).

Shift work has been linked to stress, depression, absenteeism and dissatisfaction in the workplace among nurses (Shields & Wilkins, 2006; Ferri et al., 2016). Nurses’ absenteeism rates are considerably higher than any other occupational group in Canada (Yassi, Gilbert & Cvitkovich, 2005; Shields & Wilkins, 2006; Infometrica Limited, 2017). In 2016, 24,600 nurses were absent due to their own illness or disability on a weekly basis and 61% reported taking time off for health reasons (Shields & Wilkins, 2006; Infometrica Limited, 2017). Shift workers have also been found to be at an increased risk for insomnia, chronic fatigue, anxiety, depression as well as adverse cardiovascular and gastrointestinal effects (Atkinson et al., 2008). This includes shift work disorder, a sleep disorder characterized by excessive daytime sleepiness and/or insomnia associated with shift work (Barger et al., 2012). The mechanism underlying shift work’s association with poor health may be related to changes in desynchronized circadian rhythms, sleep disturbances and psychological stress (Loprinzi, 2015). Tolerance to shift work may be affected by personality, neuroticism, rigidity of sleeping habits, commitment to shift work (how people schedule their lives) and physical fitness (Knauth & Härmä, 1992).

Sedentary time describes time spent in activities awake where the energy expenditure is ≤ 1.5 metabolic equivalents such as sitting (Tremblay et al., 2017). Greater amounts of sedentary time have been associated with lower quality and efficiency of sleep, and various health outcomes (e.g., poor metabolic profile) which are independent of physical activity levels (Loprinzi, 2015; Prince et al., 2016; Biswas et al., 2015). Currently, no Canadian guidelines for sedentary time among adults exist, but evidence has shown greater risk of cardiometabolic disease and mortality when sedentary time exceeds 8 h a day (Biswas et al., 2015; Ekelund et al., 2016; Prince et al., 2016). One study revealed nurses are approaching this value; they spend 7.5 waking hours/day being sedentary (Prince et al., 2016).

Workplace interventions, albeit few, typically focus on increasing employee physical activity levels, but fail to acknowledge the significance of addressing other health-enhancing physical behaviours (Admi et al., 2008). It is important to consider our cumulative movement behaviours and their relationship with health outcomes rather than each behaviour separately (Giacobbi, Hausenblas & Frye, 2005; Admi et al., 2008; Chaput et al., 2014; Chastin et al., 2015). Research has suggested ideal combinations of sleep, MVPA and sedentary time (i.e., optimal sleep/high MVPA, and high MVPA/low sedentary time) are associated with better overall health (Reed et al., 2018). For example, it is important that adults engage in sufficient health-enhancing physical activity, limit sedentary time (especially leisure screen time) and obtain adequate amounts of sleep (Chastin et al., 2015). Individuals who exhibit the least ideal combinations (i.e., non-optimal sleep/low MVPA/high sedentary time) are known to have less desirable measures of adiposity and cardiometabolic health; however, the effects of intermediate combinations on MVPA and mood states (e.g., tension-anxiety, depression-dejection, anger-hostility, fatigue, vigor and confusion-bewilderment) are unknown (McNair, Lorr & Droppelman, 1971; Rao, Orpana & Krewski, 2016). Emerging research has shown that when individuals engage in all three healthy movement behaviours, compared with engaging in none or only one, there are significant effects on work-related factors such as presenteeism; (Giacobbi, Hausenblas & Frye, 2005). There has been no known work to examine how combinations of sleep, physical activity and sedentary time relate to the work performance and mental health of nurses. The objective of this study was to help fill this gap in the literature by determining whether combinations of sleep, MVPA and sedentary time are associated with absenteeism, mood states and shift work disorder in a sample of Canadian nurses.

Methods

Statistical analysis

All analyses were conducted using SPSS v24 (IBM Corp, NY, USA). Univariate analyses were used to produce descriptive results including means (standard deviations) and proportions. Between groups differences by sex and for missing/non-missing data were examined using independent t-tests and chi-square tests. Analyses were conducted using data from nurses with complete data (nurses with complete data were no different than those with incomplete data). Data were tested for normality visually using plots and Shapiro–Wilks test of normality. Behavioural patterns were categorized into four groups (see Fig. 1) to compare combinations of ‘more healthy’ and ‘less healthy’ behaviours. Kruskal–Wallis one-way analyses of variance were used to compare mean outcomes for absenteeism and POMS across all four behavioural comparison groups with multiple comparisons conducted using a Bonferroni correction. Chi-square tests were used to compare shift work disorder across the four groups using adjusted residuals for multiple comparisons with p < 0.006 required for significance. As none of the variables of interest were normally distributed, non-parametric tests (independent samples Mann–Whitney U tests) were used for two-way comparisons between absenteeism/POMS and MVPA, sedentary time and sleep. Chi-square tests were used for two-way comparisons between shift work disorder and all three movement behaviours.

Results

A total of 342 nurses (83% of overall sample) from the Champlain Nurses’ Study had complete MVPA, sedentary time and sleep data; 159 had valid data and worked non-day shifts. Characteristics of this sample of nurses are shown in Table 1. Nurses’ mean age was 43 ± 12 years and 94% were women. The majority were married (68%), of White ethnicity (87%), considered overweight (56%), worked 8-hour shifts (56%), and in urban hospital locations (82%). Half worked fixed shifts (days, evenings or nights only) with most working fixed day shifts. No statistically significant differences in participant characteristics, exposures or outcomes were observed between men and women (p > 0.05 each). Complete versus missing data for all outcomes (shift work disorder, absenteeism, POMS) did not significantly differ by age, sex, shift length, shift type, sleep, sedentary time or MVPA. Missing data for the POMS questionnaire were more likely to come from nurses working in urban hospitals. Additionally, missing data for shift work disorder were more likely to be from nurses working 8-hour and fixed shifts.

Nurses spent an average of 444 ± 11 minutes/day sedentary (70% sedentary for < 8 hours/day), 14 ± 15 minutes/day in bouts ≥ 10 min of MVPA (23% meeting MVPA guidelines) and reported an average of 8 h 39 min ± 1 h 5 min of sleep (58% were ‘optimal’ sleepers) and 1.2 ± 0.3 sleep-wake cycles per 24-hour period (Table 2). On average, nurses had 8 valid sleep log entries (range of 4-9 entries). None of the covariates were significantly associated with the exposures and outcomes. Of the four comparison groups, 10.2% of nurses were characterized as belonging to Group 1, 39.5% belonging to Group 2, 41.2% belonging to Group 3 and 9.1% belonging to Group 4. Approximately half the sample achieved at least two of three healthy behaviours (i.e., high MVPA/high sedentary time/optimal sleep, high MVPA/low sedentary time/non-optimal sleep or low MVPA/low sedentary time/optimal sleep) (Fig. 1).

No significant difference between the four behavioural comparison groups (combinations of MVPA, sedentary time and sleep) were found for absenteeism (Table 3). Significant associations between behaviour groups and POMS total mood disturbance and vigor subscale and shift work disorder were observed. Group 1 had a significantly lower mood disturbance compared to Group 3 and greater vigor compared to Group 3 and Group 4. Pairwise comparisons for shift work disorder did not identify specific between-group differences. Meeting the MVPA guidelines was associated with significantly lower POMS scores (i.e., lower mood disturbance) (median [IQR] = 0.4 [4.5] vs. 1.3 [4.4], Z =  − 2.294, df = 1, p = 0.022), lower anger, higher vigor and lower fatigue compared to not meeting the guidelines (Table 4). Low sedentary time was also associated with lower POMS scores (median [IQR] = 0.6 [4.4] vs. 1.4 [4.3], Z = 2.028, df = 1, p = 0.043), higher vigor and lower fatigue versus higher sedentary time (Table 4). Significant and negative correlations were observed between sedentary time and sleep (ρ = − 0.244, p < 0.001), and MVPA in bouts and sedentary time (ρ = − 0.112, p = 0.038).

Discussion

This was the first study, to our knowledge, to assess the synergistic associations of sleep, sedentary time and MVPA with absenteeism, shift work disorder and mood states among a large sample of hospital nurses. On average, sleep time was within the recommended 7 to 9 h of sleep, sedentary time was slightly below the 8-hour threshold, but MVPA was substantially lower than the recommended 150 min per week when using ≥ 10-minute bouts. We observed significant differences in total mood disturbance by health behaviour groups, however, no significant differences between groups were found for absenteeism or shift work disorder.

Group 1 (most optimal combination of health behaviours) had a significantly lower mood disturbance compared to Group 3 (2/3 unhealthy behaviours) and greater vigor compared to Group 3 and Group 4 (all unhealthy behaviours). A closer look at differences in individual behaviours found lower total mood disturbance (higher POMS scores) and affected mood subscales including lower anger, higher vigor and lower fatigue in those who met MVPA guidelines compared to those who did not meet guidelines. We also observed lower mood disturbance in those with less than 8 h of sedentary time compared to those with greater than 8 h of sedentary time including higher vigor and lower fatigue. These findings are consistent with a previous study which found that within the composition of movement behaviours, physical activity has the strongest positive association with better health outcomes (Chastin et al., 2015).

Contrary to our hypothesis, we did not find a significant difference in nurses’ levels of absenteeism and shift work disorder across the four groups. Given that these nurses volunteered to participate in this study, it is possible that they represent a biased sample of younger, healthier individuals with a reduced likelihood of absenteeism and shift work disorder.

Given that the number of hours in a day is finite, this leads us to believe that time spent in one behaviour is co-dependent of time spent in others (Colley, Michaud & Garriguet, 2018). We found that high sedentary time was related to less sleep while lower sedentary time was related to greater MVPA. These associations demonstrate that the behaviours are not entirely independent. The effect of time spent in one behaviour may, therefore, depend on the composition of the rest of the day and may also displace time spent in another behaviour. Of this sample, 197 (57.6%) nurses obtained optimal levels of sleep and 239 (70%) nurses engaged in less than 8 h of sedentary time, showing that the majority of nurses achieved what is considered healthier levels of sleep and sedentary time (Edward, 2009; Ekelund et al., 2016). Only 23% of nurses had adequate/higher levels of MVPA, which is likely why we observed the strongest positive association with mood state.

A closer look at this sample revealed that mental health conditions such as depression (9.1%) and anxiety (7.0%) were prevalent (Reed et al., 2018). Findings from this study revealed that nurses’ MVPA levels were significantly related to measures of mood states, which can potentially affect their mental health. When nurses had higher levels of MVPA, they reported better mood state scores. It is well established that increased levels of physical activity can significantly increase positive mood states and reduce negative mood (Giacobbi, Hausenblas & Frye, 2005). The literature suggests different exercise intensities (e.g., moderate intensity/high frequency and high intensity/moderate frequency) may engender more positive mood states and mental health through a dose–response relationship (Katula, Blissmer & McAuley, 1999; Shephard, 2001). Previous research has shown that physical activity plays a role in regulating stress and psychological well-being in shift workers (medical interns) in health care (Kalmbach et al., 2018). These workers have reported poorer mood when they are less active, therefore, a reduction in physical activity may increase the risk for poorer mood outcomes (Kalmbach et al., 2018).

There is an increased interest in the association between time spent being sedentary and mental health, however, this field is in its infancy (Hamer & Smith, 2018). Findings from this study revealed that nurses’ sedentary time was significantly associated with higher POMS total mood disturbance scores. When nurses were less sedentary (< 8 hours/day), they had lower mood disturbance compared to those who were highly sedentary (≥ 8 hours/day). Emerging research has also shown a consistent association between greater sedentary time and poorer mental health, independent of changes in MVPA (Hamer, Coombs & Stamatakis, 2014; Endrighi, Steptoe & Hamer, 2016). The mechanisms underlying the effect of sedentary time on mental health are not completely understood but may be associated with enhanced pro-inflammatory responses due to acute stress and transient mood disturbances induced by greater time spent sedentary (Endrighi, Steptoe & Hamer, 2016). Past interventions targeting sedentary behaviour demonstrate that decreased sedentary time can improve mental well-being and may benefit mood, stress and sleep (Ellingson et al., 2018).

Conclusion

Emerging research has revealed the importance of healthy movement behaviours and the relationship between ideal combinations of sleep, MVPA and sedentary time and mental-health related outcomes. There is less work looking at the combined effects of these behaviours on aspects of work performance and mental health, especially in the workplace. In our sample of Canadian nurses, combinations of these behaviours were significantly associated with mood state, however, they were not significantly associated with absenteeism or shift work disorder. Furthermore, nurses that met physical activity guidelines and had lower levels of sedentary time had significantly lower total mood disturbance (better mood state scores). Our findings underscore the importance of nurses achieving optimal levels in all three behaviours, in particular, obtaining adequate levels of MVPA and limiting sedentary time, as this may be beneficial in decreasing total mood disturbance and minimizing the effects of shift work disorder. Future work is needed to address the lower physical activity and high sedentary behaviour levels of nurses and to better understand how these behaviours can be improved to optimize the mental health of the health workforce.