Examining the Causal Role of Leptin in Alzheimer Disease: A Mendelian Randomization Study

<b><i>Background:</i></b> Observational evidence regarding the role of leptin in Alzheimer disease (AD) is conflicting. We sought to determine the causal role of circulating leptin and soluble plasma leptin receptor (sOB-R) levels in AD using a separate-sample Mendelian randomization study. <b><i>Methods:</i></b> Single nucleotide polymorphisms (SNPs) independently and solely predictive of log-transformed leptin (rs10487505 [<i>LEP</i>], rs780093 [<i>GCKR</i>], rs900400 [<i>CCNL1</i>], rs6071166 [<i>SLC32A1</i>], and rs6738627 [<i>COBLL1</i>]) and of sOB-R (rs1137101 [<i>LEPR</i>], rs2767485 [<i>LEPR</i>], and rs1751492 [<i>LEPR</i>]) levels (ng/mL) were obtained from 2 previously reported genome-wide association studies. We obtained associations of leptin and sOB-R levels with AD using inverse variance weighting with fixed effects by combining Wald estimates for each SNP. Sensitivity analyses included using weighted median and MR-Egger methods and repeating the analyses using only SNPs of genome-wide significance. <b><i>Results:</i></b> Using inverse variance weighting, genetically predicted circulating leptin levels were not associated with AD, albeit with wide confidence intervals (CIs): odds ratio (OR) 0.99 per log-transformed ng/mL; 95% CI 0.55-1.78. Similarly, the association of sOB-R with AD was null using inverse variance weighting (OR 1.08 per log-transformed ng/mL; 95% CI 0.83-1.41). Results from our sensitivity analyses confirmed our findings. <b><i>Conclusions:</i></b> In this first Mendelian randomization study estimating the causal effect of leptin on AD, we did not find an effect of genetically predicted circulating leptin and sOB-R levels on AD. As such, this study suggests that leptin is unlikely to be a major contributor to AD, although the wide CIs preclude a definitive assessment.