Birth weight, liver function, cardiovascular diseases, and type 2 diabetes mellitus

Abstract

Implications for population health of the observed relation of lower birth weight with poorer liver function and with higher risk of several non-communicable diseases (NCDs) including cardiovascular disease (CVD) and type 2 diabetes mellitus (T2DM) remain controversial. Similarly, the implications of observational associations of poorer liver function with higher risk of CVD and T2DM are also unclear, although they raise the possibility of liver function mediating the association between low birth weight and NCD risk, but the mechanism and potential targets of intervention are poorly understood given observational studies are open to confounding and selection bias. To clarify the role of liver function and birth weight in CVD and T2DM. I used two-sample Mendelian randomization (MR) to assess the associations of genetic variants predicting liver enzymes (alanine transaminase (ALT), alkaline phosphatase (ALP), and gamma glutamyltransferase (GGT)) with coronary artery disease (CAD) /myocardial infarction (MI) and T2DM. To elucidate the pathway from liver function to T2DM risk, I assessed the associations of ALT and ALP at ~17.5 years with adiposity indices (body mass index (BMI), waist circumference (WC) and waist-hip ratio (WHR)) and body composition (muscle mass and body fat percentage) in the population-representative “Children of 1997” birth cohort. As validation, two-sample MR was also used to investigate the associations of genetic variants predicting liver enzymes (ALT, ALP, and GGT) with adiposity indices (BMI, WC, and WHR) and body composition (fat-free mass and fat mass). To further clarify, from a lifecourse perspective, the association of birth weight with liver function was assessed observationally using multivariable linear regression in the “Children of 1997” birth cohort which has little socio-economic patterning of birth weight, and additionally using twin status as an instrumental variable predicting birth weight in an instrumental variable analysis design for validation. Lastly, I assessed the association of birth weight with body composition observationally in the “Children of 1997” and used MR for validation. This thesis suggests a causal role of ALT in increasing T2DM risk. ALT was inversely associated with BMI, but not with WC or WHR. ALP and GGT were not clearly associated with BMI, WC or WHR. ALT was inversely associated with fat-free mass and fat mass. ALP was not clearly associated with fat-free mass or fat mass. GGT was positively associated with fat-free mass and fat mass. Birth weight was negatively associated with liver function (i.e., ALT) and positively associated with fat-free mass, but the association with fat mass was unclear. Liver function, mainly indicating by ALT, might play a key role in T2DM, possibly mediated by muscle mass. Meanwhile, from a lifecouse perspective, liver function might mediate the effect of birth weight on T2DM via muscle mass. From an etiological perspective, these findings inform understanding of NCDs and highlight potential interventions.