Diabetes, plasma glucose and incidence of colorectal cancer in Chinese adults: a prospective study of 0.5 million people

Abstract

Background: Diabetes is associated with higher risk of colorectal cancer (CRC). Uncertainty remains about the relevance of duration of diabetes and about the association of blood glucose with CRC risk among individuals without diabetes.

Methods: The prospective China Kadoorie Biobank recruited 512 713 participants in 2004-2008 from 10 diverse areas in China. After 10 years of follow-up, 3024 incident cases of CRC (1745 colon, 1716 rectal) were recorded among 510 136 participants without prior cancer at baseline. Cox regression was used to estimate adjusted HRs for CRC associated with diabetes (previously diagnosed or screen-detected) and, among those without previously diagnosed diabetes, with levels of random plasma glucose (RPG).

Results: Overall 5.8% of participants had diabetes at baseline. Individuals with diabetes had an adjusted HR of 1.18 (95% CI 1.04 to 1.33) for CRC, with similar risk for colon and rectal cancer (1.19 [1.01 to 1.39] vs 1.14 [0.96 to 1.35]). The HRs decreased with longer duration of diabetes (p for trend 0.03). Among those without previously diagnosed diabetes, RPG was positively associated with CRC, with adjusted HRs per 1 mmol/L higher baseline RPG of 1.04 (1.02 to 1.05) for CRC, again similar for colon and rectal cancer (1.03 [1.01to 1.05] and 1.04 [1.02 to 1.06], respectively). The associations of diabetes and RPG appeared stronger in men than in women, but the differences were non-significant (p for heterogeneity 0.3 and 0.2).

Discussion: Among Chinese adults, diabetes and higher blood glucose levels among those without known diabetes are associated with higher risk of CRC.

Keywords: Chinese; blood glucose; colorectal cancer; diabetes.

Figure 1

Adjusted HRs for colorectal cancer by levels of RPG among individuals without previously diagnosed diabetes at baseline. RPG levels for participants without previously diagnosed diabetes at baseline were classified as ≤5.5 (reference), 5.6–6.0, 6.1–6.9, 7.0–7.7 and ≥7.8 mmol/L. HRs were plotted against the mean usual RPG level in each group. Mean usual RPG levels were calculated as the mean RPG levels at the resurvey according to baseline RPG categories. The size of the boxes is proportional to the inverse of the variance of the log HRs. The models were stratified by age-at-risk, sex and region and adjusted for age at baseline, education, smoking, alcohol, total physical activity and fasting time. 1 mmol/L=18 mg/dL. RPG, random plasma glucose.

Figure 2

Adjusted HRs for colorectal cancer associated with diabetes in population subgroups. The models were stratified by age-at-risk, sex and region and adjusted for age at baseline, education, smoking, alcohol, total physical activity and BMI where appropriate. Boxes represent subgroup-specific estimates and diamonds represent the overall HR. The sizes of the boxes are proportional to the inverse of the variance of the log HRs. For WHR, obese participants included men with a WHR<0.85 and women with a WHR<0.90. BMI, body mass index; MET, metabolic equivalent of task; WHR, waist-to-hip ratio.

Figure 3

Adjusted HRs for colorectal cancer associated with 1 mmol/L higher RPG among individuals without previously diagnosed diabetes in population subgroups. Conventions as in figure 2. 1 mmol/L=18mg/dL. BMI, body mass index; MET, metabolic equivalent of task; RPG, random plasma glucose; WHR, waist-to-hip ratio.