OBJECTIVE: The National Cholesterol Education Program (NCEP) definition of the metabolic syndrome was modified to be described as a continuous variable and adapted to the characteristics of a Hispanic population. RESEARCH DESIGN AND METHODS: Age/sex population percentiles for every component of the NCEP criteria were included in this approach using population-based data from a Mexican nationwide survey (2,158 subjects). One point was given per decile for every component. The total number of points accumulated was used to classify subjects. The predictive power for incident diabetes was evaluated using the 7-year follow-up results of the Mexico City Diabetes Study. RESULTS: Our population-based method had a significantly better prognostic power compared with the original and the updated NCEP definitions (area under the receiver operating characteristic curve 0.746 vs. 0.697 and 0.723, respectively, P < 0.05). Using individuals with </=1 component of the NCEP definition as reference, the odds ratio was greater in the upper quartile of the points scale (>/=39 points) (12.71 [95% CI 5.67-28.49]) compared with that calculated for the original (9.52 [4.69-19.31]) and the updated (11.14 [5.33-23.30]) NCEP criteria. The major advantage of our approach is the detection of subjects at the extremes of the range of diabetes risk and the ability to estimate this risk as a continuum. CONCLUSIONS: Our method adapts the NCEP criteria to the characteristics of a Hispanic population. It improves the predictive power of the NCEP criteria for future diabetes.
OBJECTIVE: The National Cholesterol Education Program (NCEP) definition of the metabolic syndrome was modified to be described as a continuous variable and adapted to the characteristics of a Hispanic population. RESEARCH DESIGN AND METHODS: Age/sex population percentiles for every component of the NCEP criteria were included in this approach using population-based data from a Mexican nationwide survey (2,158 subjects). One point was given per decile for every component. The total number of points accumulated was used to classify subjects. The predictive power for incident diabetes was evaluated using the 7-year follow-up results of the Mexico City Diabetes Study. RESULTS: Our population-based method had a significantly better prognostic power compared with the original and the updated NCEP definitions (area under the receiver operating characteristic curve 0.746 vs. 0.697 and 0.723, respectively, P < 0.05). Using individuals with </=1 component of the NCEP definition as reference, the odds ratio was greater in the upper quartile of the points scale (>/=39 points) (12.71 [95% CI 5.67-28.49]) compared with that calculated for the original (9.52 [4.69-19.31]) and the updated (11.14 [5.33-23.30]) NCEP criteria. The major advantage of our approach is the detection of subjects at the extremes of the range of diabetes risk and the ability to estimate this risk as a continuum. CONCLUSIONS: Our method adapts the NCEP criteria to the characteristics of a Hispanic population. It improves the predictive power of the NCEP criteria for future diabetes.