Daily sleep, weekly working hours, and risk of work-related injury: US National Health Interview Survey (2004-2008).

The impact on health and safety of the combination of chronic sleep deficits and extended working hours has received worldwide attention. Using the National Health Interview Survey (NHIS), an in-person household survey using a multistage, stratified, clustered sample design representing the US civilian, non-institutionalized population, the authors estimated the effect of total daily self-reported sleep time and weekly working hours on the risk of a work-related injury. During the survey period 2004-2008, 177,576 persons (ages 18-74) sampled within households reported that they worked at a paid job the previous week and reported their total weekly work hours. A randomly selected adult in each household (n = 75,718) was asked to report his/her usual (average) total daily sleep hours the prior week; complete responses were obtained for 74,415 (98.3%) workers. Weighted annualized work-related injury rates were then estimated across a priori defined categories of both average total daily sleep hours and weekly working hours. To account for the complex sampling design, weighted multiple logistic regression was used to independently estimate the risk of a work-related injury for categories of usual daily sleep duration and weekly working hours, controlling for important covariates and potential confounders of age, sex, race/ethnicity, education, type of pay, industry, occupation (proxy for job risk), body mass index, and the interaction between sleep duration and work hours. Based on the inclusion criteria, there were an estimated 129,950,376 workers annually at risk and 3,634,446 work-related medically treated injury episodes (overall injury rate 2.80/100 workers). Unadjusted annualized injury rates/100 workers across weekly work hours were 2.03 (< or =20 h), 3.01 (20-30 h), 2.45 (31-40 h), 3.45 (40-50 h), 3.71 (50-60 h), and 4.34 (>60 h). With regards to self-reported daily sleep time, the estimated annualized injury rates/100 workers were 7.89 (<5 h sleep), 5.21 (5-5.9 h), 3.62 (6-6.9 h), 2.27 (7-7.9 h), 2.50 (8-8.9 h), 2.22 (9-9.9 h), and 4.72 (>10 h). After controlling for weekly work hours, and aforementioned covariates, significant increases in risk/1 h decrease were observed for several sleep categories. Using 7-7.9 h sleep as reference, the adjusted injury risk (odds ratio [OR] for a worker sleeping a total of <5 h/day was 2.65 (95% confidence interval [CI]: 1.57-4.47), for 5-5.9 h 1.79 (95% CI: 1.22-2.62), and for 6-6.9 h 1.40 (95% CI: 1.10-1.79). No other usual sleep duration categories were significantly different than the reference; however, for >10 h of usual daily sleep, the OR was marginally significantly elevated, 1.82 (95% CI: 0.96-3.47). These results suggest significant increases in work-related injury risk with decreasing usual daily self-reported sleep hours and increasing weekly work hours, independent of industry, occupation, type of pay, sex, age, education, and body mass.