Warning: Working Long Hours May Rot Your Brain

Long work hours and sleep interruptions are expected occupational hazards in medicine. And, it's clear that sleep deprivation impairs performance. However, data on persistent impairment from long working hours have been scarce, as most studies are cross-sectional. Also, most existing data pertain to long shifts, rather than total weekly work hours.

The Whitehall II study [Note: Free full access to this manuscript is available at the publisher's website] fills in some of the gaps. The authors studied 2,214 British civil servants (average age at enrollment: 52.1 years old). Five different cognitive measures were obtained at baseline and 5 year follow up evaluations. This study was rather unusual in the breadth of additional factors evaluated as covariates. In addition to the typical age, gender, marital status, highest level of education, etc, the authors also determined job rank, a wide variety of physical and psychological health factors, and potentially adverse lifestyle choices. The statistical analysis was detailed and rigorous.

The authors broke work hours into three groups: 35-40 hours per week: 853 (39% of total participants); 41-55 hours: 1,180 (53%); Over 55 hours: 181 (8%). The primary outcomes of interest were found in comparing the "over 55 hours per week" group to the "40 hours or less" group. Participants in the longest work hours group had lower vocabulary scores at both baseline and follow-up, and lower reasoning scores at the follow up assessments. These findings persisted when work hours were analyzed as a continuous variable, and when all demographic factors were taken into account.

Of most concern to readers is the reasoning task deterioration. The authors note that this test is associated with executive functions, a cognitive domain especially vital to effective medical practice. The authors also note that the established adverse effects of long work hours on sleep, cardiac status and other physical health indicators, as well as possible increases in adverse behaviors, are probably not the mechanism mediating the cognitive declines. The magnitude of the declines is of much concern: "...the 0.6- to 1.4-unit difference in aspects of cognitive functioning [...] is similar in magnitude to that of smoking, a risk factor for dementia." Thus, these impairments are of clinical, in addition to statistical, significance. Adding to the concern: The authors wonder if larger declines, and worsening in other areas, might be seen with a longer follow up interval.

This study has clear, and not pleasant, implications for people working in healthcare. Those of us in medicine work long hours. Those long hours may take a toll on the cognitive function we need to adequately perform our jobs.