|| Checking for direct PDF access through Ovid
The association between long working hours and alcohol drinking has been inconsistent, which may be due to an imprecise categorization of working hours and different drinking norms among occupational groups. This study aims to examine the association between working hours and drinking in different employment grades.Data from a national survey of randomly sampled Taiwanese workers in 2013 was utilised, consisting of 16 598 employees aged between 25 and 65. A total score of two or more in the CAGE questionnaire was used to identify problem drinkers. Weekly working hours were categorised into 5 groups:<40, 40, 41–48, 49–59, and ≥60. Prevalence of problem drinking was examined and compared across employment grades (i.e. managers and professionals, skilled workers, and low-skilled workers) by chi-square tests. The associations between working hours and problem drinking in different employment grades were examined by logistic regression models.The prevalence of problem drinking in managers and professionals didn’t differ between working hour categories. The prevalence of problem drinking was 3 times higher for skilled workers with <40 working hours compared to general workers. Skilled workers had the highest prevalence of problem drinking and those with 41–48 working hours had the highest odds ratio for problem drinking. The prevalence of problem drinking increased with working hours in low-skilled workers.The association between working hours and problem drinking is not linear and differs by employment grade. Interventions and studies for problem drinking in the workplace should take employment grade into consideration.