Ammarin ThakkinstianLaor ChailurkitDaruneewan WarodomwichitWipa RatanachaiwongSukit YamwongSuwannee ChanprasertyothinJohn AttiaPiyamitr SritaraBoonsong OngphiphadhanakulMahidol UniversityMedical and Health OfficeUniversity of Newcastle Faculty of Medicine and Health Sciences2018-11-092018-11-092014-01-01Clinical Endocrinology. Vol.81, No.2 (2014), 197-20313652265030006642-s2.0-84899659664https://repository.li.mahidol.ac.th/handle/123456789/34767Objective Fetuin-A is associated with body mass index (BMI) as well as components of the metabolic syndrome. However, it is unclear if fetuin-A affects BMI or the other way around. We therefore assessed the causal association between fetuin-A and BMI or vice versa, utilizing a bidirectional Mendelian randomization approach. Design and Methods This was a study of 2558 subjects from the Electricity Generating Authority of Thailand (EGAT) cohort. Two polymorphisms, that is, rs2248690 in the alpha2-Hereman-Schmid glycoprotein (AHSG) gene and rs9939609 in the fat mass and obesity-Associated (FTO) gene were genotyped. Bidirectional causal models were constructed using a two-stage least-square instrumental variable (IV) regression. First, rs2248690 locus was used as the instrumental variable for the effect of circulating fetuin-A on BMI, and then, the FTO rs9939609 locus was used as the instrumental variable for the effect of BMI on circulating fetuin-A. Results Among the 2558 subjects, the prevalence of the minor AHSG (T) and FTO (A) alleles was 17·9% and 22·1%, respectively. The AHSG rs2248690 locus was highly related to serum fetuin-A levels (P < 0·001). Likewise, the FTO rs9939609 locus and BMI were highly associated (P < 0·001). Mendelian randomization analyses showed that circulating fetuin-A, instrumented by the AHSG rs2248690 locus, was associated with BMI (coefficient = 2·26; 95% CI: 0·39, 4·12). In contrast, BMI, instrumented by the FTO rs9939609 locus, was not associated with circulating fetuin-A (coefficient = 0·0007; 95% CI: -0·0242, 0·0256). Conclusion Our findings suggest a causal association leading from circulating fetuin-A to BMI. There was no evidence of reverse causality from BMI to fetuin-A. © 2013 John Wiley & Sons Ltd.Mahidol UniversityMedicineArticleSCOPUS10.1111/cen.12303