Birth weight and four global-leading cancers: a linear and nonlinear Mendelian randomization study

Abstract Background: Birth weight (BW) reflects summary exposure measure for the intrauterine environment that affects fetal growth. The causal role of BW in four global-leading cancers is unclear.Objective: To apply Mendelian randomization (MR) to investigate the linear and nonlinear association between BW and four cancers (lung, colorectal, breast and prostate cancer).Methods: Two-sample summary data MR—from published genome-wide association studies for the associations of single-nucleotide polymorphisms (SNPs) with BW (sample 1), and from the UK Biobank for the associations of SNPs with cancer outcomes (236,201 participants) (sample 2)—was used. Non-linear MR—the fractional polynomial method for investigation on the nonlinear relationship between genetically proxied birth weight and risk of four global leading cancers.Results: After Bonferroni correction for multiple testing, genetically predicted BW was significantly inversely associated with prostate cancer. The odds ratio per 1 standard deviation increase in birth weight was 0.586 [95% confidence interval (CI) 0.388, 0.885; P =0.011] using the two-stage least squares (2SLS) method. Two sample IVW method confirmed the result. Non-linear MR suggested that there was suggestive evidence of L-shaped associations between genetically predicted birth weight and prostate cancer (Cochran Q P = 0.027; Quadratic test P = 0.049). We did not find significant evidence of the causal effect of birth weight on lung, colorectal and prostate cancer with linear and nonlinear MR analyses. Conclusions: Lower birth weight can be causally associated with an increased susceptibility to prostate cancer. Population-level interventions to maintain an optimal birth weight may lower prostate cancer risk in life. Further underlying mechanism exploration is also warranted.