Reciprocal Associations between Job Strain and Disturbed Sleep—Opportunities for Sleep Health

Job strain may be used as a model to address fundamental questions about sleep and sleep health.¹ A recent meta-analysis of more than a dozen European cohort studies reported that job strain is extremely common, with prevalence rates of 13% to 22%.² Clearly articulated theoretical models and measurement tools, combined with existing longitudinal cohort studies, have contributed to a wealth of data on the causes and consequences of job strain. These data have identified discrete stressors that contribute to job strain and might be used, in turn, to better understand physiological, psychological, social, behavioral, and environmental factors that disturb sleep and its downstream consequences to health and functioning. Associations among job strain and sleep have been less systematically evaluated. While an early survey of US adults identified job strain as the leading cause of self-reported sleep difficulties,³ more recent work suggests a bidirectional association between job strain and sleep.⁴

In this issue of SLEEP, Åkerstedt and colleagues report a reciprocal association between job strain and sleep over a two-year period, such that work demands were prospectively associated with disturbed sleep and disturbed sleep was prospectively associated with work demands.⁵ These associations were observed despite relatively stable reports of work demands and disturbed sleep over the two-year study period and despite statistical adjustment for factors that might explain these associations including perceived stress and other job characteristics such as work schedule, work hours, social support at work, and physical and environmental demands in the work environment. An additional strength of the study is the use of a nationally representative sample of working adults such that psychosocial, work, and sleep characteristics were more generalizable to the population than job strain studies based on non-representative groups (e.g., shift workers or workers from lower socioeconomic strata).^6,7

The reciprocal link between work demands and disturbed sleep described by Åkerstedt and colleagues highlight the inextricable nature of the stress-sleep relationship.⁵ While observational and experimental studies suggest that psychological stress can disrupt sleep,^8–11 an emerging literature suggests that disturbed sleep may also contribute to psychological stress. For example, short sleep duration and experimental sleep deprivation have both been associated with emotion dysregulation and increased physiological reactivity to stress.^12–16 As noted by Åkerstedt and colleagues⁵ in their Discussion, these findings suggest that workplace policies to redress job strain may reap benefits for sleep, as well as its downstream consequences to health and functioning. In addition, their data suggest that workplace interventions that target disturbed sleep may attenuate multiple aspects of job strain including perceived demands, low control, and poor social support.

Assessment of job strain and disturbed sleep during two waves of data collection by Åkerstedt and colleagues allowed careful control of possible confounders but did not allow formal tests of full mediation,¹⁷ as acknowledged by the study authors. It is hoped that these investigators will probe the pathways through which work demands contribute to later sleep disturbances, and vice versa, as they collect additional waves of data in the Longitudinal Occupational Survey of Health (SLOSH) study, which is still ongoing. One missed opportunity in the current report is the examination of effect modifiers. While examination of mediators informs the development and optimization of interventions, effect modifiers, or moderators, inform prevention efforts and precision medicine. The questions they can help answer are, “In whom, and under what circumstances, does job strain lead to disturbed sleep, and vice versa?” and “In whom should we intervene to improve disturbed sleep and/ or job strain?”

In early theoretical work that has received strong empirical support in the ensuing decades, Arthur Spielman identified reactivity to stress as a key individual difference in vulnerability to insomnia.¹⁸ We and others have reported that increased reactivity to stress, including stress-related intrusive thoughts and avoidance behaviors, predicts disrupted sleep and increased physiological hyperarousal in patients with insomnia.^11,19,20 More recently, Drake and colleagues and investigators at Laval University have shown that individual differences in vulnerability to stress-related sleep disturbances, or sleep reactivity, are associated with the pathophysiology and clinical course of insomnia.^8,21,22 It is thus plausible that participants in the SLOSH study at greatest risk for disturbed sleep during Wave 2 of data collection might be those who endorsed both elevated levels of perceived stress and frequent work demands during Wave 1 of data collection. Adverse work characteristics assessed in SLOSH such as long working hours and poor social support, each of which has been associated with disturbed sleep in other studies,^7,23 may also plausibly interact with work demands to increase the risk of disturbed sleep. Conversely, workplace characteristics may interact with sleep to influence job strain. Although there was no main effect of shift work on disturbed sleep in the present study, SLOSH participants at risk for increased work demands at Wave 2 of data collection might be those who endorsed both disturbed sleep and shift work during Wave 1 of data collection.

Future studies should seek to identify factors that reliably modify prospective associations among job strain and sleep, and vice versa. These data may then be used to develop targeted interventions to reduce job strain and improve sleep in at-risk individuals, with possible benefits to long-term health, functioning, and productivity in the workplace. There is also a need for more longitudinal data gathered over multiple waves of data collection, which is challenging to US based researchers given the length of funding cycles for NIH-sponsored research. Another approach would be to conduct pragmatic trials based on existing, scalable interventions such as brief behavioral treatment of insomnia (BBTI)²⁴ or internetbased cognitive behavioral treatment of insomnia (CBTI)²⁵ to determine whether improvements in sleep benefit job strain. New analytic approaches for identifying mediators and moderators using a precision medicine framework may also be applied to the data on job strain and sleep to more accurately characterize their reciprocal relationship, identify mechanisms through which one influences the other, in whom, and under what circumstances.²⁶ While much work has yet to be done, the sleep science questions addressed by Åkerstedt and colleagues⁵ in this issue of SLEEP hold promise for advancing our basic understanding of sleep, enhancing sleep health, and influencing social policy.²⁷

CITATION

Hall MH. Reciprocal associations between job strain and disturbed sleep—opportunities for sleep health. SLEEP 2015;38(7):1007–1008.

DISCLOSURE STATEMENT

Dr. Hall has indicated no financial conflicts of interest.