Harnessing the Power of Population Cohorts to Study the Relationship Between Endocrine-Metabolic Disorders and Depression

Being diagnosed with an endocrine-metabolic disorder such as type 1 diabetes, type 2 diabetes, polycystic ovary syndrome (PCOS), or obesity is known to increase the risk of a subsequent diagnosis of major depressive disorder (hereafter referred to as depression), and vice versa (1, 2). However, the causal and functional relationship between endocrine-metabolic disorders and depression is still poorly understood, and the studies to date have at times reached seemingly contradictory conclusions. For instance, recent causal inference studies have suggested that type 2 diabetes and obesity can lead to a subsequent depression diagnosis (3, 4), whereas another study found higher (relative) carbohydrate intake to reduce depression risk in a causal manner (5). Furthermore, the gut microbiome, shown to be influenced by the host’s genetics (6), may be associated with both diabetes and depression (7, 8). Taken together, the literature depicts a remarkably complex relationship between endocrine-metabolic disorders and depression, with various potential pathways of bidirectional correlations (see Figure 1).

To help solve this important puzzle, Leone et al., reporting in this issue of the Journal (9), utilize a population cohort of 2.2 million individuals born in Sweden and their first-degree relatives to characterize the genetic and environmental relationship between endocrine-metabolic disorders and depression. The authors performed two types of analyses. First, they examined the association between endocrine-metabolic disorders and depression in case subjects and the extent to which these conditions coaggregated in family members, who, to some degree, share both genetics and environment. Despite the low observed frequency of some of the endocrine-metabolic conditions and their co-occurrences, the relatively large sample size in the Swedish registers allowed for the detection of significant positive associations between all endocrine-metabolic disorders and depression in case subjects as well as in their family members. Second, the authors performed a quantitative analysis using the siblings to estimate (additive) heritabilities, as well as genetic and environmental correlations between endocrine-metabolic disorders and depression. The authors found genetic correlations for the sibling pairs (see the article’s supplementary Table S12) to be 0.31 between type 2 diabetes and depression, 0.21 between obesity and depression, and 0.34 between PCOS and depression. However, the estimated genetic correlation between autoimmune disorders and depression was smaller, with 0.13 for hypothyroidism and approximately 0 for type 1 diabetes. Interestingly, shared environmental effects for siblings were small, suggesting that the parental environment does not play a large role in the observed associations. The large sample size of the quantitative genetic analysis (more than 778,000 sibling pairs) allowed Leone et al. to estimate the genetic overlap with more accuracy than previous studies, clearly demonstrating a positive association between endocrine-metabolic disorders and depression that is due to both genetics and the environment. These results are consistent with recent findings (10, 11) and bring clarity to a previously unclear picture of how these serious conditions with notable population prevalence relate to each other.

Recent efforts toward erasing the boundaries between psychiatric and somatic disorders have highlighted the inaccuracy of the dichotomous classification of the former as solely falling under the domain of the brain and the nervous system. One such example is the negative genetic correlations reported for anorexia nervosa with metabolic (e.g., type 2 diabetes, insulin resistance, circulating leptin levels) and anthropometric measures (e.g., body mass index, overweight, obesity), reestablishing this deadly eating disorder as a metabo-psychiatric disorder as opposed to one having a purely psychiatric origin (12). Considering our improved understanding of the complex interplay between monoamines (13), cytokines, and the microbiome (14) influencing the risk of depression, Leone and colleagues’ findings also point toward the need for a more comprehensive evaluation of the pathophysiology of depression. Importantly, this study is a crucial first step toward setting the path for future work establishing a causal relationship between endocrine-metabolic disorders and depression. Although making a causal inference, especially in a retrospective cohort study, is very difficult, individual medical trajectories could provide important clues, such as the temporal relationship between two diagnoses (i.e., the order of diagnosis). With the goal of identifying modifiable causal risk factors that could be targeted in public health efforts and preventive treatments, understanding the causal relationship between comorbid conditions such as endocrine-metabolic disorders and depression is critical.