Endocrine Abstracts

Background: Many studies indicated an increased cancer incidence and cancer mortality in subjects with diabetes mellitus. But most of studies have been surveyed in western countries and have problems in applying the results to Korean. We aimed to evaluate the site specific cancer risk of diabetic patients in Korea over 30 years, and investigate causal relationship and temporal relationship by analyzing the organ-specific cancer risk in diabetics according to duration.

Methods: Using a database provided by NHIS, we conducted a retrospective, population-based cohort study for adults over 30 years from January 1, 2005 to December 31, 2013. We calculated the hazard ratio (HR) for each cancer by using Cox regression. And we compared HR for each cancer according the duration of diabetes divided in to 6month, 6months to 3years and more than 3years to verify the possibility of detection bias or reverse causation.

Results: The incidence of total cancer per 1,000 Subjects with diabetes was higher than with non-diabetes group. Hazard ratio(HR) for cancer of pancreas, liver, kidney, bladder, colorectum, thyroid, lung, stomach and leukemia were significantly higher in diabetes group after adjusted. HR for cervix, endometrium and breast cancer in female diabetes group were higher than those of non-diabetes group. HR for laryngeal cancer in male was higher in diabetes group. The total cancer risk was higher in < 6 months of diabetes duration (HR 2.03; 95% confidence interval [95% CI]. 1.99–2.07), whereas the hazard ratio decreased with increasing duration of diabetes, raging from 1.19 (95% CI, 1.18–1.21) between 6 months and 3 years to 1.12 (95% CI, 1.11 to 1.13) at over 3 years, but remained significantly higher than the non-diabetic group. HR for pancreas, liver, colorectum, prostate and endometrium were higher in diabetes group for entire duration of diabetes. When excluding diabetes duration less than 6months, HR for stomach, colorectum, liver, pancreas, kidney, bladder, thyroid, prostate, endometrium and cervix showed similar pattern including diabetes duration less than 6 months.

Conclusion: Cancer incidence risk increases in T2DM patients, and the phenomenon is more prominent in short durations, i.e. shortly after diagnosis of diabetes. As the duration increases, incidence risk of cancer varies depending on the site of cancer and patient’s gender. Thus, in diabetes patients, cancer screening should be individualized regarding duration of diabetes, sex, and the site of cancer.