This cross-sectional exploratory study involved health care workers of various skill types and levels. We tested the hypothesis that the prevalence of diseases, sleep complaints, and insufficient time for nonprofessional activities (family, leisure, and rest) are higher among night than day workers. Data collection was carried out in two public hospitals using questionnaires and other forms. Night work was explored as a risk factor, considering a night worker as one who had at least one night job on the occasion of the research. Data were assessed by a univariate analysis. The association between work schedule and the dependent variables--health conditions, sleep complaints, and insufficient time for nonprofessional activities--was evaluated through the estimation of the prevalence ratio, with a confidence interval of 95%. Two hundred and fifty-eight female nursing personnel participated; 41.5% were moonlighters, and only 20 worked a shift of less than 12h in length. Reports of migraine and need of medical care the 2 weeks before the survey were more prevalent among day than night workers (PR=0.71; CI=0.55-0.92 and PR=0.71; CI=0.52-0.95, respectively). Migraine headaches occurred less frequently among night than day workers as confirmed by comparing the reports of the night workers and day workers whose work history was always day shifts (PR = 0.74; CI = 0.57-0.96). Reports of mild emotional disorders (mild depression, tension, anxiety, or insomnia) were less frequent among night (PR=0.76; CI=0.59-0.98) and ex-night workers (PR=0.68; CI=0.50-0.91) than day workers who never had worked a night job. The healthy worker effect does not seem to explain the results of the comparisons between day and night workers. The possible role of exposure by day workers to some risk factors, such as stress, was suggested as an explanation for these results. No significant difference was observed between night and day workers as to sleep complaints, a result that may have been influenced by the nature of the shift-work schedule (no successive night shifts) and possibly nap taking during the night shift. Moreover, the long work hours and moonlighting of the healthcare workers, which is common in Brazil, may have masked other possible differences between the day and night workers. Among night workers, a significant relation was found between years working nights (more than 10 yrs) and high cholesterol values (PR = 2.58; CI = 1.07-6.27), a result that deserves additional study. Working nights more than four times per 2-week span was related to complaints about insufficient time for children (PR= 1.96; CI = 1.38-2.78) and rest/leisure (PR= 1.54; CI = 1.20-1.99). These results can be related to the "social value of time," as evenings and nights are when families usually spend time together. The complexity of the professional life and the consequent heterogeneity of the group of workers under shift-work schemes confound the results. More in-depth study of the questions raised here demands a more sophisticated epidemiological treatment and larger sample size.
