Skip to main content
. 2018 Jan 3;187(7):1380–1391. doi: 10.1093/aje/kwx376

Table 4.

Association Between Type 2 Diabetes Mellitus and Risk of Incident Cancer According to Time From Diabetes Diagnosis to Baseline, China Kadoorie Biobank Study, 2004–2013a

Type of Cancer No. of Events Among Persons Without T2DM Time From T2DM Diagnosis to Baseline P for Trendb
≤5 Years >5 Years
No. of Events HRc 95% CI No. of Events HRc 95% CI
All cancers 16,006 391 1.09 0.99, 1.21 413 1.10 0.99, 1.21 0.02
Esophagus 1,572 19 0.71 0.45, 1.11 19 0.73 0.46, 1.17 0.06
Stomach 2,061 28 0.63 0.44, 0.92 44 0.96 0.71, 1.30 0.20
Colon and rectum 1,721 41 0.87 0.64, 1.19 49 0.94 0.70, 1.25 0.44
Liver 1,746 57 1.56 1.20, 2.04 43 1.18 0.87, 1.61 0.02
Lung 3,217 83 1.08 0.86, 1.34 98 1.12 0.91, 1.37 0.23
Female breastd 1,344 28 0.94 0.64, 1.38 45 1.56 1.15, 2.12 0.02

Abbreviations: CI, confidence interval; HR, hazard ratio; T2DM, type 2 diabetes mellitus.

a Participants with screen-detected T2DM (n = 13,954) and an implausible age of diagnosis (n = 20) were excluded from the analysis, and the total sample size was 494,918.

bP values for trend were from a likelihood ratio test comparing the model with time since diabetes diagnosis as an ordered categorical variable to the model without it.

c The model stratified by the sex, age (5-year intervals), and study area of the study population. Results were adjusted for education, parental history of cancer, body mass index, cigarette smoking, alcohol drinking, and physical activity.

d For women only (n = 291,587). The model stratified by the age (5-year intervals) and study area of the study population. Results were adjusted for education, parental history of cancer, menopausal status, body mass index, cigarette smoking, alcohol drinking, and physical activity.