levated serum glucose and cholesterol levels have been widely shown to be linked to increased cardio- vascular mortality and all-cause mortality. Hyper- cholesterolemia is also a well-established risk factor for cardiovascular disease among patients with type 2 diabetes. The association between high cholesterol levels and increased risk of cancer has been a long-standing area of interest and in- vestigation because the pathway for cholesterol synthesis may produce various tumorigenic compounds and because choles- terol serves as a precursor for the synthesis of many sex hor- mones linked to increased risk of various cancers. However, disease prediction based on biological hypotheses often is not well reflected in actual clinical risk. Such complexity is underscored by divergent findings from epidemiologic studies on serum cholesterol levels and cancer risk. In this issue of CMAJ, Yang and colleagues 1 conducted a prospective study of the association between low-density lipoprotein (LDL) cholesterol levels and risk of cancer among Chinese patients with type 2 diabetes. Results indicated a con- sistent and striking V-shaped association. The lowest risk of cancer was observed among those with an intermediate LDL cholesterol level (≥ 2.80 to < 3.80 mmol/L). About a 50% higher risk of cancer was observed among patients with an LDL cholesterol level either above or below that range. Over- all, the findings were robust in both crude and adjusted mod- els, and results were consistent among users and nonusers of statins. A V-shaped association was also observed between LDL cholesterol and all-cause mortality. What warrants careful consideration is whether the associ- ation is biologically causal or merely confounded by other risk factors, including socioeconomic status. A V-shaped as- sociation between LDL cholesterol and cancer risk suggests that multiple mechanisms are involved. LDL cholesterol is unlikely the sole or direct causal factor. More than a decade ago, Jacobs and colleagues 2 observed that low total cholesterol levels were associated with an in- creased risk of cancer. However, this finding was considered to be attributable to confounding or reverse causation by sub- clinical cancer. Low cholesterol levels are commonly ob- served in people who are ill (e.g., patients with cancer) and those with unhealthy lifestyle characteristics, such as smoking and heavy drinking. 3 Furthermore, 3 meta-analyses of ran- domized controlled trials showed that the use of statins, which inhibits cholesterol synthesis, was not associated with increased or decreased risk of overall or major cancers.