Live fast die young : health consequences of early puberty

Abstract

Cardiovascular disease (CVD) is one of the leading causes of morbidity and mortality globally whose rates have increased with economic development. CVD remains poorly understood and cannot be entirely explained by adult risk factors, giving rise to interest in a life course perspective. Age of puberty has declined substantially in developed settings and is now declining elsewhere with rapid economic development. Earlier age of puberty is consistently associated with CVD in adulthood. However, whether the associations are causal, or would propagate to the future generations remains unclear. This thesis used Hong Kong’s large (n=8327) population-representative birth cohort “Children of 1997” in a recently developed non-Western setting, in which age of puberty is among the lowest globally, to examine associations of onset and duration of puberty with CVD risk factors at age ~17 years; maternal age of menarche with birth weight, body mass index (BMI), height and blood pressure, and whether the associations were mediated by adiposity or pubertal development. In a complimentary analysis I also used large publicly available genetic studies to estimate unconfounded associations of childhood adiposity, age of menarche, ischaemic heart disease (IHD) and its risk factors. I used multivariable linear regression to assess the associations of clinically assessed timing and duration of puberty with CVD factors. I used multivariable linear regression and generalised estimating equations to examine the adjusted association of maternal age of menarche with birth weight and with BMI and height at different growth phases, as well as blood pressure in adolescence. I assessed whether these associations varied by sex or maternal birthplace. Mendelian randomisation was used to obtain unconfounded estimates of the associations of childhood BMI with Tanner stage in boys and girls and of age of menarche with IHD and diabetes. For boys, later age of onset of pubic hair development and voice break were associated with lower low-density lipoprotein (LDL) cholesterol, later age of voice break was associated with lower glycated haemoglobin A1c (HbA1c), and longer duration of puberty was associated with lower HbA1c. Maternal age of menarche was negatively associated with BMI in childhood and at puberty, height at puberty, and systolic blood pressure in adolescence, which was largely explained by adiposity. Genetically predicted childhood BMI was positively associated with pubertal stage and earlier menarche in girls but not boys and with IHD and diabetes. Genetically predicted age of menarche was inversely associated with IHD and diabetes, but these associations were not robust to adjustment for childhood BMI. In this thesis, childhood adiposity appeared to partly drive earlier puberty and blood pressure, and hence should be a public health priority. However, the role of advancing pubertal timing in later CVD could not be completely ruled out. I suggest future pharmacological agents could consider targeting the reproductive axis to offer cardiovascular benefits, based on the evolutionary biology paradigm of growth and reproduction trading-off against longevity, and hence major chronic diseases.