Purpose: Equivocal results regarding the role of leptin in colorectal cancer (CRC) and adenoma (CRA) have been reported. A case-control study investigating the association of leptin with CRC risk and clinicopathological characteristics along with meta-analysis of published data on both CRC and CRA were conducted.
Methods: Pubmed and Embase were searched for the meta-analysis, comprising 28 case-control studies amounting 3,614 CRC and 1,215 CRA cases, along with 5,220 controls. Meticulous contact with the authors of individual studies was undertaken for the provision of additional data. Pooling of standardized mean differences (SMD), relative risks (RR) and 95 % CI (random effects models), subgroup, sensitivity, and meta-regression analyses were conducted.
Results: The meta-analysis suggested positive association of serum leptin with CRA (RR, 95 % CI 1.35, 1.03 to +1.76), but not CRC either at the pooled analysis on SMDs or RRs (SMD, 95 % CI 0.18, -0.04 to +0.40; RR, 95 % CI 1.04, 0.65 to +1.65). Significant heterogeneity between studies on CRC as well as between studies on CRA providing SMD was noted. Subgroup, meta-regression and sensitivity analyses highlighted potential methodology-, design-, size- and quality-related effect modifiers.
Conclusions: Meta-analysis of current evidence suggests positive association of serum leptin with CRA but not with CRC risk. Given the case-control nature of available studies, the limited number of studies on serum leptin and CRA, and the heterogeneity of CRC studies, carefully designed, prospective studies preferably reporting RRs adjusted for a variety of confounders may be warranted.