![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: The prevalence of prediabetes among adults from the National Health and Nutrition Examination Survey (NHANES) based on blood tests is 37.5%, and is about three to fourfold higher than self-reported prevalence estimates from the Behavioral Risk Factor Surveillance System (BRFSS). Self-reported prediabetes prevalence from the BRFSS is therefore a significant underestimate of true population prevalence. Given the inability of states to assess population estimates of prediabetes using blood tests, many rely solely on self-reported data from BRFSS. Alternative methods to blood tests beyond BRFSS self-reports alone are needed to more accurately estimate overall prevalence of prediabetes and target diabetes prevention efforts. The objective of this study is to use the Diabetes Prevention Recognition Program (DPRP) prediabetes screening test in combination with self-reports of prediabetes from the BRFSS to estimate overall population prevalence of prediabetes.
METHODS: The study uses data from the 2012 North Carolina BRFSS on respondents without a history of diabetes. Those with a self-reported history of prediabetes were considered to have “diagnosed prediabetes”. The DPRP prediabetes screening test was applied to respondents without a history of prediabetes. Published positive predictive values for the DPRP prediabetes screening test were used to obtain a range of possible estimates of “undiagnosed prediabetes” prevalence among respondents at high risk for prediabetes (screening test score of 9 or greater). The resulting estimates of undiagnosed prediabetes were combined with diagnosed prediabetes estimates to obtain a range of estimates for population prediabetes prevalence. All analyses were performed using STATA version 13.0. Sampling weights and ‘svy’ commands were used to account for complex sampling design.
RESULTS: The prevalence of diagnosed prediabetes was 8.8% (95% CI: 8.1, 9.5). The estimated prevalence of undiagnosed prediabetes ranged from 11.6% (10.8, 12.4) to 21.7% (20.7, 22.7). The combined overall prevalence of prediabetes ranged from 19.1% (18.1, 20.0) to 28.3% (27.2, 29.4).
CONCLUSIONS: Using the DPRP prediabetes screening test in conjunction with self-reported prediabetes prevalence from BRFSS produces estimates for population-level prevalence of prediabetes that are closer to those obtained from blood tests. This approach may be a way for state public health programs to better maximize their Diabetes Prevention Program efforts. Further analysis and comparisons with NHANES data is needed to fully validate this approach.