Is domestic violence a risk factor for women's emotional distress in the long run? Analysis of MINIMat longitudinal data from Matlab, Bangladesh.

Background: Violence against women (VAW) contributes significantly to women's distress affecting physical and mental health of women and their families' well-being. Most analyses on distress and VAW use cross-sectional data, inhibiting any causal inference. Furthermore, different trajectories of women's distress are usually ignored while assessing the impact of VAW. This study addresses these gaps in the literature by examining the impact of different types of VAW on different stress trajectories of women.

Methods: The data are from a cohort study in Bangladesh known as the Maternal and Infant Nutrition Interventions in Matlab (MINIMat). This analysis includes a subsample of the MINIMat participants (n=1066) with data on domestic violence (DV) and distress at enrolment (2001-2020) and 10-year and 18-year follow-ups. Logistic regression analysis, latent class growth modelling and multinomial logistic regression modelling were used in the analyses.

Results: Physical violence, emotional violence and controlling behaviour were identified as risk factors for women's distress. Three distress trajectories identified within the sample are: (1) persistently low distress, (2) persistently increasing distress and (3) high midlife distress. Women who reported ever experiencing physical violence (OR 1.69; 95% CI 1.17 to 2.44) and controlling behaviour (OR 1.41; 95% CI 1.43 to 1.91) at enrolment were more likely to have persistently increasing distress compared with women in the persistently low distress trajectory. Women who reported ever experiencing emotional violence (OR 2.23; 95% CI 1.20 to 4.13) and controlling behaviour at enrolment (OR 2.31; 95% CI 1.30 to 4.10) were more likely to be in the high midlife distress trajectory.

Conclusions: Although all the studied types of DV caused distress in the long run, exposure to physical DV and controlling behaviour had the worst effect by ever-increasing distress over an 18-year period. Exposure to emotional DV and controlling behaviour also increased distress over a 10-year period. These findings are valuable for targeting the most vulnerable groups in resource-poor settings.