乌鸦传媒

Women's risk of new-onset systemic lupus erythematosus (SLE) was linked to habitually getting less-than-recommended sleep, according to Nurses Health Study (NHS) data, although an actual causal relationship remained uncertain.

Among women participating in the huge, long-running study, those reporting daily sleep of 5 hours or less were diagnosed with lupus at more than double the rate seen in participants usually getting 7-8 hours of sleep (HR 2.47, 95% CI 1.29-4.75), according to May Choi, MD, PhD, of the University of Calgary in Canada, and colleagues.

Other factors that also seemed to promote risk of lupus included the presence of pain and depressive symptoms, the researchers reported in .

Choi and colleagues suggested that "sleep deprivation" may be causative in lupus, saying their findings "highlight the potential role of adequate sleep in disease prevention." But the study design did not permit a firm conclusion about causality, and it appeared just as plausible that the causal pathway runs in other directions -- for example, that lupus pathology prior to formal diagnosis contributed to diminished sleep.

Certainly, given that the research drew on the (including the two cohorts established in 1976 and 1989), in which shift work would be common among participants, the notion of sleep deprivation as a cause of physical illness was compelling. Furthermore, with nearly 250,000 participants and standardized data collection and follow-up, these data should be ideal for examining such relationships (as evidenced by the that have come out of the NHS).

After excluding participants with insufficient data, Choi and colleagues examined lupus incidence and sleep duration for some 186,000 NHS participants, with a total of some 4.2 million person-years of follow-up. A total of 187 cases of lupus were diagnosed in these women.

About 9,600 reported regular sleep duration of 5 hours daily or less. Compared with participants reporting more sleep, these women tended to exercise more and drink less alcohol and were more likely to be nonwhite. More of the short-sleepers reported a history of depression (except in comparison with NHS participants who said they typically slept more than 8 hours daily), and more reported substantial bodily pain (scores of 60 or less on the Short Form-36 health questionnaire's pain scale). Those who typically got no more than 5 hours of sleep also were more likely to have worked shifts for at least 6 years.

The association between short sleep duration and lupus risk was diminished only slightly in multivariate analyses that adjusted for shift work, depression, and pain (HR 2.13, 95% CI 1.11-4.10, relative to participants getting 7-8 hours of sleep), reducing the chance that any of those co-variates were the real causative factors. As well, the association remained strong (HR 3.14) in a sensitivity analysis that excluded lupus diagnoses made less than 4 years after the sleep-duration assessment, in an attempt to rule out "reverse causation" (i.e., "whereby underlying disease might affect sleep duration," the investigators explained).

Choi and colleagues expressed confidence that insufficient sleep is a likely contributor to lupus risk, not only on the basis of these findings, but also from other research that, for example, found increased expression of pro-inflammatory cytokines and immune-cell overactivity in the context of restricted sleep. The researchers also proposed that work-related stress -- such as from shift rotations -- can promote autoimmune disease, as some previous studies have suggested.

Still, the study's observational nature meant that causality could not be proved. The authors also acknowledged that the small number of incident lupus cases was a limitation, as was the highly educated, mostly white (more than 90%) profile of NHS participants. And much of the data on sleep habits and health history were self-reported by participants.

Comment