RESUMO
PURPOSE: The prevalence of Diabetes Type 2 is on the rise internationally. Currently, Fasting Plasma Glucose (FPG) and HbA1c are both used to determine if an individual is diabetic or prediabetic. We aimed to describe the prevalence of diabetes, prediabetes, and glycemic control in a population-based sample of elderly Hispanic and non-Hispanic White participants in New Mexico. METHODS: To do this, we compared HbA1c with FPG using Chi-Square analysis across gender and ethnicity to provide information for future health care policy. We also performed non-parametric regression using a locally weighted smoothing technique to investigate the relationship between FPG and HbA1c levels. RESULTS: Our analysis identifies a large variation between the sensitivity of HbA1c and FPG in the identification of both prediabetes and diabetes. Interestingly, 95% of diabetics defined by FPG are also defined by HbA1c, representing overlap between the two measures. When comparing the prevalence of prediabetes between the two measures, the overlap of FPG with HbA1c was only 30% and HbA1c identifies more individuals as prediabetic than FPG. Prevalence of diabetes was also higher when defined by HbA1c compared to FPG and the overall agreement between HbA1c and FPG appears to be poor particularly by sex and ethnicity (K=0.22-0.34). Glycemic control was poor overall with Hispanics displaying a larger amount of uncontrolled diabetes. CONCLUSION: We compared HbA1c and FPG by gender and ethnicity and showed both measures of diabetes differ in their sensitivity across ethnic groups. Our results suggest that using HbA1c, rather than FPG, results in higher rates of prediabetes and diabetes, a finding with numerous implications for health care practice.
