Exploring the causal effect of placental physiology in susceptibility to mental and addictive disorders: a Mendelian randomization study

Abstract

Background: Epidemiological studies have linked low birth weight to psychiatric disorders, including substance use disorders. Genomic analyses suggest a role of placental physiology on psychiatric risk. We investigated whether this association is causally related to impaired trophoblast function.

Methods: We conducted a two-sample summary-data Mendelian randomization study using as instrumental variables those genetic variants strongly associated with birth weight, whose effect is exerted through the fetal genome, and are located near genes with differential expression in trophoblasts. Eight psychiatric and substance use disorders with >10,000 samples were included as outcomes. The inverse variance weighted method was used as the main analysis and several sensitivity analyses were performed for those significant results.

Results: The inverse variance weighted estimate, based on 14 instrumental variables, revealed an association, after correction for multiple tests, between birth weight and broadly defined depression (β = -0.165, 95% CI = -0.282 to -0.047, P = 0.0059). Sensitivity analyses revealed the absence of heterogeneity in the effect of instrumental variables, confirmed by leave-one-out analysis, MR_Egger intercept, and MR_PRESSO. The effect was consistent using robust methods. Reverse causality was not detected. The effect was specifically linked to genetic variants near genes involved in trophoblast physiology instead of genes with fetal effect on birth weight or involved in placenta development.

Conclusion: Impaired trophoblast functioning, probably leading to reduced fetal brain oxygen and nutrient supply, is causally related to broadly defined depression. Considering the therapeutic potential of some agents to treat fetal growth restriction, further research on the effect of trophoblast physiology on mental disorders may have future implications in prevention.

Keywords: Mendelian randomization; birth weight; depression; fetal neurodevelopment; mental disorders; placental physiology; psychiatric genetics; trophoblast invasion.

Figure 1

Directed acyclic graph illustrating our Mendelian randomization approach. Gray arrows correspond to the IV assumptions. Green arrow indicates the causal effect of interest. By selecting as IVs SNPs with specific characteristics, we avoid dynastic effects (dashed black arrows), as it is improbable that SNPs accomplishing these characteristics act on the outcome via parents’ phenotype.

Figure 2

Flow chart of the study.

Figure 3

Forest plot of the main results of the Mendelian randomization (MR) analysis. (A) Results of the MR analysis using the different psychiatric disorders as outcome. The size of each square is proportional to the effective sample size of each GWAS ( Supplementary Table S2 ). (B) Results of the MR analysis in which independent SNPs significantly associated with each of the psychiatric disorders were taken as IVs to assess their effect on birth weight, evaluating the inverse causality hypothesis. In case of just one IV, the Wald ratio test is shown. The other results are based on the IVW method.

Figure 4

Leave-one-out sensitivity analysis for depression as outcome. The y-axis indicates the SNP that is removed in each analysis. The x-axis shows the beta (dot) and 95% CI (line) for each analysis.

Figure 5

Scatter plot showing the beta effects of SNPs on exposure (birthweight) and outcome (depression). The estimates are represented as dots, with 95% CI represented by horizontal and vertical lines, respectively. The slope of each colored line corresponds to the estimated causal effect by each MR method.

Figure 6

Results of the MR analysis based on subsets of IVs according to trophoblast cell type.