Browse/search for people

Publication - Dr Pauline Emmett

    Dietary patterns and depressive symptoms in a UK cohort of men and women

    a longitudinal study


    Northstone, K, Joinson, C & Emmett, P, 2018, ‘Dietary patterns and depressive symptoms in a UK cohort of men and women: a longitudinal study’. Public Health Nutrition, vol 21., pp. 831-837


    Objective There is evidence to suggest that individual components of dietary intake are associated with depressive symptoms. Studying the whole diet, through dietary patterns, has become popular as a way of overcoming intercorrelations between individual dietary components; however, there are conflicting results regarding associations between dietary patterns and depressive symptoms. We examined the associations between dietary patterns extracted using principal component analysis and depressive symptoms, taking account of potential temporal relationships. Design Depressive symptoms in parents were assessed using the Edinburgh Postnatal Depression Scale (EPDS) when the study child was 3 and 5 years of age. Scores >12 were considered indicative of the presence of clinical depressive symptoms. Diet was assessed via FFQ when the study child was 4 years of age. Setting Longitudinal population-based birth cohort. Subjects Mothers and fathers taking part in the Avon Longitudinal Study of Parents and Children when their study child was 3-5 years old. Results Unadjusted results suggested that increased scores on the 'processed' and 'vegetarian' patterns in women and the 'semi-vegetarian' pattern in men were associated with having EPDS scores ≥13. However, after adjustment for confounders all results were attenuated. This was the case for all those with available data and when considering a sub-sample who were 'disease free' at baseline. Conclusions We found no association between dietary patterns and depressive symptoms after taking account of potential confounding factors and the potential temporal relationship between them. This suggests that previous studies reporting positive associations may have suffered from reverse causality and/or residual confounding.

    Full details in the University publications repository