Background and aim: The role of diabetes mellitus as a risk factor for gastric cancer has been controversial. We studied the association between diabetic biomarkers and the risk of gastric cancer and whether these associations depend on cancer location.
Methods: In this retrospective cohort study with subjects with negative initial esophagogastroduodenoscopy findings (n = 23 218) during a routine health checkup, we measured fasting glucose and insulin levels, calculated the homeostatic model assessment insulin resistance (HOMA-IR) values, and analyzed the risk of gastric cancer in relation to diabetic biomarker tertiles and the presence of diabetes mellitus.
Results: The incidence rate of gastric cancer was 9.7 per 10 000 person-years during the mean 6.8-year follow up. Patients with diabetes, higher fasting glucose levels, or higher HOMA-IR levels were older; men, current smokers, and heavy alcohol consumers represented larger proportions of these groups. They also had high body mass index and hemoglobin A1c more often. In the multivariate-adjusted Cox regression analyses, the incidence of gastric cancer was not significantly associated with diabetes mellitus or higher diabetic biomarker levels. Compared with normal glucose levels, lower glucose levels were significantly associated with an increased risk of distal gastric cancer. The hazard ratio for fasting glucose level tertile 1 was 2.39 (95% confidence interval, 1.48-3.85) (reference, tertile 2). Lower glucose levels were not associated with a risk of proximal gastric cancer, compared with a normal glucose level.
Conclusions: Our findings suggest that fasting glucose levels have a different effect on distal and proximal gastric cancers.
Keywords: cohort studies; diabetes mellitus; gastric cancer.
© 2016 Journal of Gastroenterology and Hepatology Foundation and John Wiley & Sons Australia, Ltd.