Amerindians, particularly Pima Indians, are at high risk for type 2 diabetes (T2D). Our previous studies have shown that lipid levels (high triglycerides or TG, and low HDL-cholesterol, or HDL-C) are important predictors of T2D incidence. There are few data on the extent to which apolipoprotein levels, particularly apolipoprotein CIII (apoC-III), which is a potential therapeutic target to treat high TG, predict T2D. Thus, we examined whether circulating apolipoprotein levels predict T2D risk in addition to known risk factors. We measured serum TG, HDL-C and 5 apolipoprotein levels (apoA-I, apoA-V, apoB100, apoC-III, apoE), in 1,916 nondiabetic Pima Indians at baseline (mean age=31.2±10 years, 43% male). After mean follow-up of 6.26 years, ∼24% of subjects (n=454) developed T2D.

BMI, HDL-C and TG predicted T2D, when adjusted for sex and age (hazard ratio [HR] per s.d. unit was 1.05, 0.74 and 1.18, respectively; all p<0.0005). TG had moderate correlation with apoC-III (r=0.68), apoE (r=0.56) and HDL-C (r=-0.31). With adjustment for age and sex, only apoA-V predicted T2D incidence (HR=1.12 P=0.02). When effects of BMI, HDL-C, TG and each one of 5 apoliproteins were evaluated together, the apoA-V effect was attenuated (HR=1.09, p=0.08). However, higher apoC-III was protective for T2D incidence (HR for apoC-III=0.83, p=0.007; HR=1.43, 0.87, and 1.29 for BMI, HDL-C, and TG, respectively, all p<0.05). Formal significance testing for apoC-III effect showed that it has distinct predictive value for T2D incidence in addition to TG (0.008). As TG and apoC-III are positively correlated, apoC-III may be a risk factor for T2D. However, our findings suggest that, controlled for BMI, TG and HDL-C levels, apoA-I, apoA-V, apoB100 and apoE add little information about T2D risk prediction, whereas higher apoC-III is associated with reduced T2D risk. The mechanisms underlying this association are not presently known.

Disclosure

W. Hsueh: None. M. Walter: None. R. Nelson: None. W.C. Knowler: None. R.L. Hanson: None.

Funding

National Institutes of Health

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.