O3E.2\u2005Risk of mental health disorders in human service occupations: a register based study of 445,651 norwegians

Abstract

Objectives Mental disorders contribute strongly to disability. Work in human service occupations has been considered a risk factor in several studies. We aimed at quantifying this relationship in a prospective follow-up of all employed persons born in Norway, 1967–1976. Methods We conducted follow-up in several national registries. Based upon the ISCO98 four-digit code, we classified 2007 occupations into customer contact, client/patient contact, and reference (no contact). Client/patient contact was subdivided into health care, education and social work. We collected mental disorder diagnoses (ICD-10 F00-F99), in particular affective (F30-F39) and stress-related (F40-F48), in the Norwegian Patient Registry, 2008–2011. Four-year prevalence differences (PD) per 100 across occupational categories were estimated in binomial regression models adjusted for year of birth, marital history, current family pattern, and education level. Results Among 4\u200945\u2009651 employed persons, 18% held customer-related occupations while 25% held occupations with client/patient contact. The total four-year prevalence of mental disorder was 8.6/100 (38,207 patients). Affective and stress-related prevalences were 5.3 and 6.5 for women, and 3.0 and 3.2 for men, respectively. Adjusted PD estimates showed positive associations between client/patient contact occupations and mental disorders: for women, affective and stress-related disorder PD estimates were 1.1 (95% confidence interval (CI); 0.9–1.3) and 1.3 (CI; 1.0–1.5), respectively. The corresponding PD estimates for men were 1.7 (CI; 1.5–2.0) and 1.5 (CI; 1.2–1.7). We found strongest associations for women in social work (PD 2.3 for stress-related disorders) and men in health care (PD 2.6 for affective disorders). Associations with other mental disorder categories were weak, as were associations with customer contact occupations. Conclusions Affective and stress-related morbidity was clearly associated with occupations involving client/patient contact for both sexes. We cannot dismiss health selection as an alternative to a causal effect of work-related factors. This register-based study is not well suited to provide mechanistic explanations.