A latent class analysis of health risk behaviours in the UK Police Service and their associations with mental health and job strain

Irizar, P. and Gage, S.H. and Fallon, V. and Goodwin, L. (2022) A latent class analysis of health risk behaviours in the UK Police Service and their associations with mental health and job strain. BMC Psychiatry, 22 (1): 426. ISSN 1471-244X

Abstract

BACKGROUND: Health risk behaviours (e.g., harmful drinking and smoking) often cluster together and can be associated with poor mental health and stress. This study examined how health risk behaviours cluster together in individuals in a high stress occupation (UK Police Service), and the associations with mental health and job strain. METHODS: Data was obtained from the Airwave Health Monitoring Study (25,234 male and 14,989 female police employees), which included measures of health risk behaviours (alcohol use, diet, smoking status, physical activity), poor mental health (depression, anxiety, post-traumatic stress disorder [PTSD]), and job strain (low, high, active, passive). Classes of health risk behaviours were identified using Latent Class Analysis (LCA) and the associations with mental health and job strain were analysed through multinomial logistic regressions. RESULTS: For men and women, a 5-class solution was the best fit. Men and women with depression, anxiety, and/or PTSD (analysed as separate variables) had at least double the odds of being assigned to the "high health risk behaviours" class, compared to those with no mental health problem. Compared to those reporting low strain, men and women reporting high strain had increased odds of being assigned to the "low risk drinkers with other health risk behaviours" classes. CONCLUSIONS: These finding highlight the importance of holistic interventions which target co-occurring health risk behaviours, to prevent more adverse physical health consequences. Police employees with poor mental health are more likely to engage in multiple health risk behaviours, which suggests they may need additional support. However, as the data was cross-sectional, the temporal associations between the classes and mental health or job strain could not be determined.