NIOSHTIC-2 Publications Search
Extended work schedules and health outcomes in the U.S.
Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, R01-OH-007564, 2005 Feb; :1-80
Occupational Safety and Health Issue - This project evaluated the associations between exposure to extended work hours and the risk of occupational injury; nonoccupational injury; back disorder; ischemic heart disease; and Psychiatric Disorders. Importance of the Problem - Bureau of Labor Statistics reports have shown that the proportion of persons having very long workweeks increased through the 1990s. The prevalence of long workweeks varies considerably by occupation, for both men and women. For example, extended work schedules are more highly concentrated in the managerial, professional, sales, and transportation occupations. Among adults, lack of sleep, which is a direct consequence of extended work schedules, has been implicated in injuries and car crashes. However, little is known about the relation, if any, between the increase in hours spent working on events such as occupational and nonoccupational injury. Approach - The project was a retrospective follow-up study that utilized the 1996 through 2000 Medical Expenditure Panel Survey (MEPS) household component, Panels 1 through 5. The MEPS is a nationally representative sample of the civilian noninstitutionalized population of the US. A series of nested case-control studies were carried out. Initial steps involved the creation of a workers' cohort. These were subjects drawn from the MEPS Household Component's Consolidated File and Jobs File. To be eligible for inclusion, study subjects had to have been approximately 17 years of age and had at least one (1) job for any duration at any time during the follow-up period. For each investigation cases and controls were selected. For occupational and nonoccupational injury, cases were identified through the MEPS Medical Conditions File as having reported an injury (ICD9 800-999) during the follow-up period. Additional information about whether or not the injury occurred at work enabled differentiation between work and nonwork related injuries. Using the MEPS Medical Conditions File and Medical Events Files, incident cases of back disorder (ICD9 720-724), ischemic heart disease (ICD9 410-414, 416), and psychiatric disorder (ICD9 290, 292, 294-314) were identified and their diagnosis date recorded. Prevalent (preexisting) cases were excluded from case identification. For all investigations controls were matched (5:1) on age (+/- 1 year) at injury date (or diagnosis date), race and gender. Exposure was assessed for the 30-day period before the case injury date (or diagnosis date) for the case and their matched controls. Subjects' average number of jobs (1, >1), peak number of jobs (1, >1), worked rotating shifts (yes, no), average hours worked per week and peak hours worked per week were calculated and categorized as being approximately 40,40,41-50, or >50 hours per week. Conditional logistic regression was used to calculate risk ratios (RRs) and 95 % confidence intervals (CIs) for the associations between the series of outcomes and measures of extended work hours derived over the 30-day period before the date of event. Stratified analysis was carried out to determine the effect in different occupational groupings (white- and blue-collar) and gender groups. Final adjusted measures of association were calculated for the overall group and by occupational and gender subgroups. Variables in the final multivariate models included: average hours worked, white-collar versus blue-collar, working rotating shifts, and average number of jobs > 1; and peak hours worked, white-collar versus blue-collar, working rotating shifts, and peak number of jobs > 1. Key Findings - There were 74,900 subjects included in the 1996-2000 MEPS Household Component. Of that number, 54,029 were 17 years of age and older during the follow-up period and of these 38,847 held at least one (1) job during the follow-up period. There were 2,438 cases of occupational injury. We found that average number of jobs greater than one (1) resulted in significant increased injury risks but that this effect was restricted to white-collar workers (RR 2.69,95% CI 2.05-3.52). For average number of jobs >1, adjusted RR 1.72 (95% CI 1.43-2.08). Our examination of extended hours worked indicated that increased risk was consistent both overall and across occupational and gender groups. The risk of work related injury related to average number of hours worked was strongest among blue-collar workers (relative to those who worked < 40 hours per week). After adjusting for number of days worked, the following RRs were found: 40 hours/week RR 1.24, 95% CI 1.02-1.50; 41-50 hours/week RR 1.43, 95% CI 1.15-1. 79; > 50 hours/week RR 1.60, 95% CI 1.27-2.02. In the other subgroups results were generally consistent, that is increasing injury risks with increasing hours worked, in other subgroups (white-collar, men and women). Final adjusted RRs for average hours worked per week were 40 hours/week RR 1.10, 95% CI 0.97-1.26; 41-50 hours/week RR 1.35, 95% CI 1.151.57; > 50 hours/week RR 1.40,95% CI 1.18-1.66). Working rotating shifts was also significantly associated with slight increased risks (RR 1.24, 95% CI 1.07-1.44). There were 4,210 cases of nonoccupational injury. Results for non-work related injuries were unremarkable when considering average or peak hours worked per week. Adjusting for total days worked, men who worked> 50 hours/week were at significant decreased risk (RR 0.74,95% CI 0.63-0.87). Also, those subjects that worked rotating shifts were not at increased risk. However, elevated and significant associations for non-work related injury were seen for women who had >1 job (RR 1.33, 95% CI 1.05-1.68). There were 271 incident cases of back disorder (ICD9 720-724). For extended hours worked elevated risks were seen primarily among blue-collar workers but confidence intervals were wide and included the null (1.0). Relative to those who worked < 40 hours per week and adjusting for number of days worked: 40 hours/week RR 1.50, 95% CI 0.62-3.64; 41-50 hours/week RR 2.46,95% CI 0.87-6.98; > 50 hours/week RR 1.60, 95% CI 0.50-5.12). Further subgroup analysis by gender indicated that increased risks were principally among women with men having nonsignificant protective effects. Final adjusted RRs for average hours worked per week were 40 hours/week RR 1.15, 95% CI 0.71-1.87; 41-50 hours/week RR 2.14,95% CI 1.173.93; > 50 hours/week RR 1.62, 95% CI 0.73-3.59). There were 117 incident cases of ischemic heart disease (ICD9 410-414, 416) and measures of associations for extended hours worked, average number of jobs and working rotating shifts were not remarkable. There were 429 incident cases of psychiatric disorder (ICD9 290, 292, 294-314). Study results suggest that overall there was no association between incident psychiatric' disorder and extended work hours. However subgroup analysis showed significant two (2) fold increased risk among men who had more than one (1) average/peak jobs (RR 2.30,95% CI 1.21-4.40). Final adjustment for other potential confounders removed the statistical significance of this association (RR 2.12,95% CI 0.93-4.81).
Injuries; Injury-prevention; Accident-prevention; Accident-rates; Accident-statistics; Traumatic-injuries; Back-injuries; Musculoskeletal-system-disorders; Heart; Cardiovascular-disease; Cardiopulmonary-system-disorders; Psychological-disorders; Psychological-effects; Epidemiology; Statistical-analysis
Gerald McGwin, Jr., Department of Epidemiology and International Health, School of Public Health, University of Alabama, Birmingham, AL 35294
Final Grant Report
NTIS Accession No.
Work Environment and Workforce: Organization of Work
National Institute for Occupational Safety and Health
University of Alabama, Birmingham, Alabama
Page last reviewed: September 2, 2020
Content source: National Institute for Occupational Safety and Health Education and Information Division