LAUSR

AIMS To examine potential associations between the glycaemic index (GI), glycaemic load (GL), and carbohydrates and the incidence risk of type-2 diabetes (T2D) and the effect modification of obesity among Korean adults aged ≥40 years. METHOD Incidence rate ratios (IRRs) and 95% confidence intervals (CIs) for T2D were estimated in 8,310 participants using a modified Poisson regression model. Dietary indices were averaged using repeated dietary assessments during follow-up. RESULT After adjusting for potential confounders, a positive association between GI and T2D was found among women (IRR=1.63, 95% CI=1.06-2.51 in the highest tertile (T3) vs. the lowest tertile (T1) for GI, p trend=0.0310), but not for GL and carbohydrate intake. This positive association with GI was stronger in obese women (IRR=1.91, 95% CI: 1.15-3.19 in T3 vs. T1, p trend=0.0137 for body mass index ≥ 23 kg/m2; IRR=2.35, 95% CI: 1.01-5.48, p trend=0.0350 for waist circumference (WC) ≥ 85 cm). In men, there was no association before stratification by obesity, but IRRs of GI (T3 vs. T1) were significant and stronger with increased WCs (IRR=2.26, 95% CI: 1.02-4.98, p trend=0.0439 for WC ≥90). CONCLUSION GI may be positively associated with the incidence of T2D in women, particularly in obese women. The association of GI with T2D incidence risk may also be positive even in men with high WC.