Inverse Association Between Betel Quid Use and Diabetes in Rural Bangladesh

Objectives: Betel quid is used as a mild stimulant in many parts of South and East Asia and the Pacific. In observational studies, its use has been associated with elevated risk for diabetes, but studies in animal models suggest some component(s) of betel quid could reduce risk. Methods: We assessed associations between betel quid use and diabetes (glycated hemoglobin, HbA_1c ≥ 6.5%) among a cross-sectional sample of 410 men and 717 non-pregnant women in Matlab, Bangladesh. Results: In multivariable logistic regression, betel quid use was inversely associated with diabetes among men (aOR: 0.45; 95% CI: 0.26, 0.79) but not women (aOR: 0.88; 95% CI: 0.51, 1.52). There was a dose–response relationship between frequency of betel quid use and diabetes among men (aOR: 0.73; 95% CI: 0.59, 0.89), but not women (aOR: 0.96; 95% CI: 0.77, 1.18). Betel quid use was inversely associated with diabetes as an ordinal variable (no diabetes/prediabetes/diabetes) among men (aOR: 0.55; 95% CI: 0.36, 0.82) and women (aOR: 0.66; 95% CI: 0.47, 0.94). Structural equation modeling suggested that some of the inverse association was mediated by food source (market vs. household production). Conclusions: These models support the hypothesis that betel quid use could decrease, rather than increase, risk for diabetes in the Bangladeshi context, particularly among men (who have more frequent betel quid use than women). Heterogeneity in betel quid preparation across settings, multifactorial effects of betel quid use, and publication bias may contribute to differences between these findings and other observational studies.