The bi-directional association between bipolar disorder and obesity: Evidence from Meta and bioinformatics analysis

Abstract

Background: The globally high prevalence of both obesity and bipolar disorder makes the bidirectional relationship between the two disorders a pivotal phenomenon; hence, a meta-analysis to synopsize their co-occurrence is indispensable. Psychotropic-induced obesity has been reported to be an important factor linking bipolar disorder and obesity. Nonetheless, the molecular signature of this connection is perplexing.

Methods: Here, we leverage both meta-analysis and bioinformatics analysis to provide a conspectus and deduce the molecular signature of obesity in bipolar disease patients following psychotropic treatment. Searches were performed on a diverse collection of databases through June 25, 2020. The Newcastle-Ottawa Scale was used to rate the quality of the studies. Analysis of OR, 95% CI, and tests of homogeneity were carried out with STATA software. For the bioinformatics analysis, the LIMMA package which is incorporated into the Gene Expression Omnibus database was used.

Results: Our search yielded 138 studies, of which 18 fitted our inclusion criteria. Individuals who are obese have an increased risk of developing bipolar disorder (pooled adjusted OR = 1.32, 95% CI = 1.01-1.62). In a manner analogous to this, the pooled adjusted odds ratio reveals that patients with bipolar disorder have an increased chance of obesity (OR = 1.68, 95% CI = 1.35-2). To deduce the molecular signature of obesity in bipolar disorder patients following psychotropic treatment, three data sets from the Gene Expression Omnibus database (GSE5392, GSE87610, and GSE35977) were integrated and the genes obtained were validated by a cohort of known single nucleotide polymorphism of obesity via direct overlap. Results indicate genes that are activated after psychotropic treatment. Some of these genes are CYBB, C3, OLR1, CX3CR1, C3AR1, CD53, AIF1, LY86, BDNF, ALOX5AP, CXCL10, and the preponderance falls under mesodermal and PI3K-Akt signaling pathway. The ROC analysis reveals a strong discriminating value between the two groups (UBAP2L AUC = 0.806, p = 1.1e-04, NOVA2 AUC = 0.73, p = 6.7e-03).

Conclusion: Our study shows unequivocal evidence of a bi-directional association between bipolar disorder and obesity, but more crucially, it provides a snapshot of the molecular signature of obesity in bipolar patients as a result of psychotropic medication.