Genome-wide pleiotropy between depression and body mass index

Introduction Obesity and depression are public health problems with increasing prevalence worldwide. Previous studies indicate a phenotypic association between increased Body-Mass-Index (BMI) and higher rates of depression. However, it remains unclear how much of the association is due to environment or shared biological and genetic factors. In this study we apply a multilevel bioinformatics pipeline to summary statistics from recent GWASs of both traits to identify pleiotropic SNPs, genes and pathways. Moreover, we examined possible sex-specific pleiotropic effects. Finally, we could support some of our results using phenotype and genetic data from our general population sample from the SHIP study.

Methods Summary statistics of GWASs on MDD (Wray et al. 2018, cases: 135 458, controls: 344 901) and BMI (Locke et al. 2015, cases: 59 851, controls: 113 154) were used to identify pleiotropic SNPs employing a sum ranking method. Pleiotropic p-values for genes and pathways were calculated using the extended Simes test (GATES) and hybrid set-based test (HYST) as implemented in the software KGG (Knowledge-based mining system for Genome-wide Genetic studies). Regression analyses on the general population sample of the SHIP study (n = 5749) were performed to assess the association between the identified pleiotropic SNPs and BMI and MDD, respectively. Further regression analyses were performed to test the joint effect of SNPs and BMI on MDD status as well as the joint effect of SNPs and MDD status on BMI. All analyses were performed for the complete samples as well as for men and women separately.

Results We identified 243 SNPs with significant pleiotropic effect after correction for multiple testing. They were located in seven independent loci with lead SNPs rs7531118 (Chr 1), rs4714293 (Chr 6), rs1627536 (Chr 9), rs12411886 (Chr 10), rs11625397 (Chr 14), rs7243785 (Chr 18) and rs427943 (Chr 21). On gene-level we identified 16 pleiotropic genes with NEGR1 (Neuronal Growth Regulator 1) on chromosome 1 being the most significant (p_corrected = 1.19 × 10⁻⁶). Our analyses yielded 12 pathways with significant pleiotropic p-value, including the regulation of endogenous sterols and the purine metabolism. Of the seven pleiotropic loci, those around SNPs rs7531118, rs1627536 and rs11625397 also showed nominal significance in the association analyses on the SHIP study sample.

Conclusion The results of this study support the hypothesis that shared genetic factors play a relevant role in the frequent joint occurrence of obesity and depression. The identified pleiotropic genes and pathways indicate an important role of the HPA-axis, energy homeostasis and the endocrine system in the depression-obesity relationship and should be studied further to understand the biological mechanisms underlying this comorbidity.