Risk factors for sick leave in Brazilian oil industry workers.

Objective: Studies point to a complex and multifactorial aetiology for sickness absence, but there is still insufficient evidence on how occupational and non-occupational risk factors contribute to this phenomenon. This study aimed to identify occupational and non-occupational factors related to sickness absence in oil industry workers, as well as their population attributable fractions (PAF). Methods: A retrospective cohort study of 2028 oil industry workers was conducted in the state of Bahia, Brazil, between January 1, 2012 and December 31, 2016. Study data were obtained from workers´ medical register. The dependent variable was the first event of sick leave for five or more days, and the independent variables were gender, age, years of work, type of work activity, level of education, work schedule, awkward working postures, body mass index, smoking, physical activity and alcohol consumption. Cox regression analysis identified risk factors to the outcome and PAF were calculated. Results: Being female (HR 1.79), being older (HR 1.44), non-graduated ones (HR 1.44), being of operational staff (HR 1.18), overweight (HR 1.13), obesity (HR 1.39), smoking (HR 1.61) and former smoking (HR 1.30) were risk factors for sickness absence in this cohort. Of all sickness absence in the study population, 6% could be attributed to overweight, 10% to obesity, and 8% to smoking. Conclusion: To prevent sick leave, we recommend programs focused on combating smoking and encouraging healthy habits, aiming at weight control and the prevention of chronic diseases. Population attributable fractions point that a universal approach, aimed at the total population will be more cost-effective than health promotion programmes restricted for high-risk populations. [ABSTRACT FROM AUTHOR]