The association of adult height with the risk of cardiovascular disease and cancer in the population of Sardinia

The relationship between body height and the risk of non‒communicable diseases such as cardiovascular disease and cancer has been the subject of much debate in the epidemiological literature. Concerns have recently arisen over spurious associations due to confounding factors like birth cohort, especially in the context of epidemiological transition. The population of Sardinia represents an interesting case study, as the average physical stature of inhabitants was the lowest recorded in Europe until a few decades ago. In this population we tested whether height is an independent risk factor for cardiovascular disease and cancer. We analysed the stature of 10,427 patients undergoing endoscopy for any reason, for whom a detailed clinical history of cardiovascular disease and/or malignancies had been documented. Poisson regression modelling was used to test the association between stature and disease risk. When patients were subdivided according to sex and height tertiles, the risk of cardiovascular disease proved significantly greater for subjects in the lowest tertile irrespective of sex (men: 1.87; 95%CI 1.41‒2.47; women: 1.23; 95%CI 0.92‒1.66) and smaller for those in the highest tertile (men: 0.51; 95%CI 0.35‒0.75; women: 0.41; 95%CI 0.27‒0.61). However, after adjusting the risk for birth cohort and established risk factors, it mostly resulted in non-significant values, although the overall trend persisted. Similar results were obtained for all-cancer risk (relative risk for men and women in the lowest tertile: 1.44; 95%CI 1.09–1.90 and 1.17; 95%CI 0.93–1.48, in the highest tertile: 0.51; 95%CI 0.36–0.72 and 0.62; 95%CI 0.47–0.81, respectively) as well as for some of the most common types of cancer. We concluded that the risk of developing cardiovascular disease and malignancies does not vary significantly with stature in the Sardinian population, after adjusting for birth cohort and more obvious risk factors.