[Comparison of sickness absences across industries: A cross-sectional study using the JILPT Data Archive]

Assessing the risk of employee health problems according to firm characteristics (e.g., industry) can be used by companies to identify groups of workers with health problems and develop health-related policies. Previous studies have examined differences in the prevalence of diseases across industries; however, studies using sickness absences, which reduce productivity, are scarce. The purpose of this study was to identify differences in sickness absence rates across industries.

With permission for secondary use of archived data from the Japan Institute for Labor Policy and Training (JILPT), we obtained data from private companies with 50 or more regular employees nationwide. Negative binomial regression was conducted using the number of sickness absences attributed to mental health, cancer, lifestyle-related diseases (e.g., diabetes, hypertension, heart disease, cerebrovascular disease), and intractable diseases as the objective variables. The results were confirmed by an ordinal logistic regression. Firm characteristics other than industry were adjusted for firm size, age structure, medical examinations, labor unions, and flexible work systems.

The incidence of sickness absences due to mental health was high in the information and communications, medical care/welfare, and education/learning support industries because of lifestyle-related diseases, and heart disease was high in the transportation/postal industry. Cancer was high in the medical care/welfare industry. While older worker age groups had a lower incidence of mental health issues, a higher incidence of physical illnesses, excluding intractable diseases, was observed. The presence of a labor union was associated with sickness absences due to mental health, cancer, and cerebrovascular disease, and the availability of a flexible work system was associated with sickness absences due to mental health and heart disease.

The three industries with a high incidence of mental health leave had a high percentage of professional/technical workers and a common background of heavy workloads. In addition, role ambiguity, particularly in the education and medical industries, could lead to stress. The association between the transportation/postal industry and sickness absences due to lifestyle-related diseases and heart disease was probably influenced by work style and lifestyle, whereas the association between the medical care/welfare industry and sickness absences due to cancer was probably influenced by the high percentage of women in this industry. This study could not be adjusted for several variables examined in other studies, such as sex, and caution should be exercised when interpreting the results, especially regarding absences due to cancer.