UCSF

Aims

We assessed the relative associations of β-cell dysfunction and insulin sensitivity with baseline glycemic status and incident glycemic progression among Asian Indians in the United States.

Methods

A 5-sample oral glucose tolerance test was obtained at baseline. Normoglycemia, impaired fasting glucose (IFG), impaired glucose tolerance (IGT), and type 2 diabetes (T2DM) were defined by ADA criteria. The Matsuda Index (ISIM) estimated insulin sensitivity, and the Disposition Index (DIo) estimated β-cell function. Visceral fat was measured by abdominal CT. After 2.5years, participants underwent a 2-sample oral glucose tolerance test. Standardized polytomous logistic regression was used to examine associations with prevalent and incident glycemia.

Results

Mean age was 57±8years and BMI 26.1±4.6kg/m(2). Log ISIM and log DIo were associated with prediabetes and T2DM after adjusting for age, sex, BMI, family history of diabetes, hypertension, and smoking. After adjusting for visceral fat, only DIo remained associated with prediabetes (OR per SD 0.17, 95% CI: 0.70, 0.41) and T2DM (OR 0.003, 95% CI: 0.0001, 0.03). Incidence rates (per 1,000 person-years) were: normoglycemia to IGT: 82.0, 95% CI (40, 150); to IFG: 8.4, 95% CI (0, 41); to T2DM: 8.6, 95% CI (0, 42); IGT to T2DM: 55.0, 95% CI (17, 132); IFG to T2DM: 64.0, 95% CI (3, 316). The interaction between sex and the change in waist circumference (OR 1.8, per SD 95% CI: 1.22, 2.70) and the change in log HOMA-β (OR 0.37, per SD 95% CI: 0.17, 0.81) were associated with glycemic progression.

Conclusions

The association of DIo with baseline glycemia after accounting for visceral fat as well as the association of the change in log HOMA-β with incident glycemic progression implies innate β-cell susceptibility in Asian Indians for glucose intolerance or dysglycemia.