Objectives:
This study aims to evaluate the effect of selenium deficiency on depressive disorders with adjustment for possible confounders. Its importance among non-dietary and dietary risk factors for depression is discussed using empirical evidence.

Methods:
A structural equation model was fitted using diagonally weighted least squares estimation with adjusted chi-square test statistics (WLSMV). The average daily intake of selenium and other nutrients was calculated to verify their possible association with self-reported depressive disorders. The effect of dietary patterns was adjusted for possible confounders, including the presence of chronic diseases, life problems, pain levels, physical activity, and income. The study was performed on a sample of 9,354 men and women aged 45–65 of the Polish-Norwegian Study (PONS) cohort.

Results:
The model shows a significant effect of low selenium intake (standardized total effect of 0.133), high lipids intake (0.102) and low iron intake (0.065) on depressive disorders. Other dietary factors fail to make a significant contribution to depressive disorders, according to the model (p > 0.05). Among the considered non-dietary risk factors, home stress (0.181), pain (0.179) and low income (0.178) show a strong correlation with depression. Pain mediates a small part of the effect of morbidity (0.140). Depressive disorders are also associated with work problems (0.123) and low physical activity (0.024).

Conclusions:
Selenium intake is most strongly related to depression among all the dietary factors considered. In the model, the effect of dietary risk factors on depressive disorders is moderate when compared to non-dietary variables. Chronic pain, low income and morbidity are the main correlatives of depressive disorders.