Meta-Analyses of Genome-Wide Association Studies for Postpartum Depression

Abstract

Objective: Postpartum depression (PPD) is a common subtype of major depressive disorder (MDD) that is more heritable, yet is understudied in psychiatric genetics. The authors conducted meta-analyses of genome-wide association studies (GWASs) to investigate the genetic architecture of PPD.

Method: Meta-analyses were conducted on 18 cohorts of European ancestry (17,339 PPD cases and 53,426 controls), one cohort of East Asian ancestry (975 cases and 3,780 controls), and one cohort of African ancestry (456 cases and 1,255 controls), totaling 18,770 PPD cases and 58,461 controls. Post-GWAS analyses included 1) single-nucleotide polymorphism (SNP)-based heritability ([Formula: see text]), 2) genetic correlations between PPD and other phenotypes, and 3) enrichment of the PPD GWAS findings in 27 human tissues and 265 cell types from the mouse central and peripheral nervous system.

Results: No SNP achieved genome-wide significance in the European or the trans-ancestry meta-analyses. The [Formula: see text] of PPD was 0.14 (SE=0.02). Significant genetic correlations were estimated for PPD with MDD, bipolar disorder, anxiety disorders, posttraumatic stress disorder, insomnia, age at menarche, and polycystic ovary syndrome. Cell-type enrichment analyses implicate inhibitory neurons in the thalamus and cholinergic neurons within septal nuclei of the hypothalamus, a pattern that differs from MDD.

Conclusions: While more samples are needed to reach genome-wide levels of significance, the results presented confirm PPD as a polygenic and heritable phenotype. There is also evidence that despite a high correlation with MDD, PPD may have unique genetic components. Cell enrichment results suggest GABAergic neurons, which converge on a common mechanism with the only medication approved by the U.S. Food and Drug Administration for PPD (brexanolone).

Keywords: Genetics/Genomics; Mood Disorders-Postpartum.

FIGURE 1.. Results of genome-wide association meta-analyses for PPD.

(A) Manhattan plot for association tests from fixed effects meta-analysis of EUR-ancestry (17,339 PPD cases and 53,426 screened controls). Genomic position (chromosomes 1–22 and X-chromosome) is shown on the x-axis and statistical significance as -log(P) is shown on the y-axis. The solid red horizontal line indicates the genome-wide significance threshold of 5 × 10⁻⁸, and the dashed red horizontal line indicates the suggestive threshold of 1 × 10⁻⁶. (B) Association test quantile-quantile plot of observed versus expected -log10(P) values from the EUR meta-analysis. The 95% confidence interval of expected values is shown in grey. Test-statistic inflation value, λ, is 1.04. (C) Manhattan plot for association tests from trans-ancestry random effects meta-analysis (18,770 PPD cases and 58,461 screened controls). (D) Association test quantile-quantile plot of observed versus expected -log10(P) values from the trans-ancestry meta-analysis. Test-statistic inflation value, λ, is 0.94.

FIGURE 2.. Genetic correlations (rg) between PPD and psychiatric disorders, medical diseases, and biomedical traits.

Significant $r_g$ values with false discovery rate < 0.05 are shown. Error bars indicate standard error. Dashed vertical line indicates $r_g=1$.

FIGURE 3.. Cell type enrichment analyses performed using Partitioned LD Score Regression.

Nominally significant values with $\text{p}<0.05$ are shown. Labels indicate the enriched tissue or cell-type. Solid red line indicates findings with $\text{p}<0.01$.