Diet quality and depressive symptoms. Assessing the direction of the association in a population-based cohort study

Emerging evidence links a poor diet with mental ill-health although the direction of this association is unclear. The aim was to examine the bidirectional prospective relationships between core (and non-core food consumption, and symptoms of depression.

Depressive symptoms (Mental Health Index-5, MHI-5), current/prior depression and consumption of core (recommended food groups) and non-core (discretionary) foods were assessed in the population-based 2013 and 2017 Household Income and Labour Dynamics in Australia cohort study. Three cross-lagged linear models assessed associations between all three baseline variables in 2013, alternating 2017 variables as outcomes.

In the population (n = 10,003; 48.3% women; 48.5[15.7] years), core food score in 2013 was associated with MHI-5 (β:0.102, 95%CI: 0.010,0.193) in 2017, while the non-core food score was not (β:-0.030, 95%CI:-0.099,0.160). Depressive symptom score in 2013 was not associated with either food score in 2017. Current/prior diagnosis of depression in 2013 was associated with core (β:-0.198, 95%CI:-0.329,-0.067) but not non-core (β:-0.036, 95%CI: -0.151,0.080) food score in 2017.

Results may not be generalizable to the whole population due to some selection bias, self-report depression diagnosis may have led to misclassification of previous mental illness, and core and non-core food scores are not validated measures of diet quality.

There is a prospective association between core food consumption and depressive symptoms. This association is of small magnitude and we cannot discount insufficient core food consumption reflecting an effect of prior mental illness. Our results suggest that, for depression, public health focus should be on improving core food intake.