Finding macronutrients that matter

Abstract

Dietary guidelines on macronutrients recommend restricting sugar and sugar sweetened beverage (SSB) intake to prevent obesity, and restricting saturated fatty acids intake and replacement with polyunsaturated fatty acids (PUFA) to prevent cardiovascular disease (CVD). However, most evidence concerning the effects of sugar on obesity and CVD is from observational studies in Western settings where diet and health are socially patterned, thus is open to confounding by social-economic position. Different PUFAs may also have distinct effects on CVD. However, evidence about the effects of specific PUFAs remains controversial, while the effect of arachidonic acid (AA), a major dietary long-chain n-6 PUFA from animal foods, is less studied. Moreover, dietary recommendations are not sex-specific even though disease risk and longevity differ by sex. This thesis aimed to examine associations of SSB frequency with adiposity in the non-Western setting of Hong Kong with less clear confounding, validated by assessing associations of genetically instrumented blood sucrose and fructose with adiposity, ischemic heart disease (IHD) and its risk factors, and to examine associations of genetically instrumented plasma phospholipid AA with CVD and its risk factors and cancer. Where relevant and possible I also examined differences by sex. I examined the associations of SSB frequency at 11 and 13 years, assessed using a food frequency questionnaire, with subsequent body mass index z-score and overweight/obesity to 18 years, and with waist circumference, waist-to-hip ratio, and body fat percentage at 16–19 years in a population-representative birth cohort “Children of 1997” (n=3628). As validation, Mendelian Randomization (MR), which minimizes confounding, was used to examine associations of genetically instrumented blood sucrose and fructose with IHD risk and its risk factors (i.e., type 2 diabetes, adiposity, blood pressure, lipids, and glycaemic traits). Univariable and multivariable MR were used to examine associations of genetically instrumented plasma phospholipid AA with CVD and its risk factors (i.e., blood pressure, lipids, adiposity, and markers of inflammation and coagulation) and cancer, and mediators of associations of AA with IHD and stroke. SSB intake frequency at 11 or 13 years was not associated with subsequent adiposity, but a small association of SSB frequency with adiposity cannot be excluded. Genetically instrumented blood sucrose or fructose was not associated with IHD or its risk factors, although a very small effect cannot be excluded. Genetically instrumented AA was associated with atherosclerotic CVD (ASCVD), including IHD and ischemic stroke, in men but not women, and with venous thromboembolism. AA was associated with blood lipids but no other CVD risk factors. Associations of AA with IHD and ischemic stroke were attenuated after adjusting for apolipoprotein B. AA was also nominally associated with cancer of the respiratory/intrathoracic organs, skin, and genital organs, in men but not women. In conclusion, I found only a minor role of SSB in adiposity in Hong Kong Chinese children with low SSB consumption. AA has adverse effects on ASCVD, likely mediated by apolipoprotein B, and on venous thromboembolism, cancer of the respiratory/intrathoracic organs, skin, and genital organs , with potentially stronger effects in men than women.