New-Onset Type 2 Diabetes Mellitus and Cancer Risk: A Matched Cohort Study in China Kadoorie Biobank

Abstract

The prevalence of type 2 diabetes mellitus (T2DM) is rising rapidly in China and is linked to increased cancer risk, but causality remains unclear due to biases. We examined the causal effect of T2DM on cancer risk using bias-minimizing methods. We conducted a matched cohort study within China Kadoorie Biobank (median follow-up 49 months in men and 53 months in women). To minimize bias, we included only new-onset T2DM, applied sequential longitudinal matching, used pre-diagnosis BMI, and accounted for detection time bias. Stratified Cox models estimated sex-specific time-split hazard ratios (tsHRs). Results were triangulated with two-sample Mendelian Randomization (MR) in East Asians. After 1:3 matching, 8657 men and 13,680 women with T2DM were matched to unexposed individuals. T2DM was associated with increased risk of total cancers in men (tsHR 1.57, 95% CI 1.38–1.78) and women (tsHR 1.29, 95% CI 1.14–1.46). Site-specifically, T2DM was associated with liver cancer (men: 68 cases, tsHR 2.12, 95% CI 1.42–3.15; women: 43 cases, tsHR 2.39, 95% CI 1.42–4.04) and pancreatic cancer (men: 36 cases, tsHR 2.57, 95% CI 1.35–4.92; women: 33 cases, tsHR 3.95, 95% CI 1.92–8.13). No significant associations were observed for colorectal, lung, stomach, or breast cancers. In multivariable MR, genetic liability to T2DM was associated with pancreatic cancer after adjusting for BMI (OR 1.07, 95% CI 1.01–1.14, p = 0.03). In this large Chinese population, we found evidence for causal associations between T2DM and pancreatic cancer only, whereas evidence for other cancers was weak or discordant.