Skip to main content
. 2025 Sep 19;104(38):e44619. doi: 10.1097/MD.0000000000044619

Table 4.

Mendelian randomization estimates of the causal effect of epilepsy on obesity using pan-UKB data.

Method Number of SNPs MR analysis Heterogeneity test MR-Egger intercept P
OR LCI UCI P Cochran’s Q I2 P
Epilepsy (N = 64,629, Ncase = 27,559, Ncontrol = 42,436) on obesity (N = 4,20,531, Ncase = 15,917, Ncontrol = 4,04,614)
 IVW 40 1.06 0.91 1.25 .441 52.4 27.4 .061
 MR-Egger 40 0.64 0.38 1.08 .106 57.8 32.5 .027 .055
 Weighted Median 40 1.09 0.89 1.33 .394
GGE (N = 49,388, Ncase = 6952, Ncontrol = 42,436) on obesity (N = 4,20,531, Ncase = 15,917, Ncontrol = 4,04,614)
 IVW 19 1.05 0.93 1.18 .445 30.8 44.9 .021
 MR-Egger 19 1.05 0.56 1.97 .872 30.8 41.6 .030 .985
 Weighted Median 19 1.04 0.91 1.20 .570
FE (N = 57,375, Ncase = 14,939, Ncontrol = 42,436) on obesity (N = 4,20,531, Ncase = 15,917, Ncontrol = 4,04,614)
 IVW 28 1.24 1.06 1.46 .008 37.4 30.6 .068
 MR-Egger 28 1.39 0.72 2.66 .337 37.6 28.2 .084 .734
 Weighted Median 28 1.25 1.02 1.53 .030

FE = focal epilepsy, GGE = generalized genetic epilepsy, IVW = inverse-variance weighted, LCI = lower confidence interval, MR = Mendelian randomization, OR = odds ratio, SNP = single nucleotide polymorphism, UCI = upper confidence interval.