OBJECTIVETo compare the differences of two recommended diagnostic criteria for metabolic syndrome (MS) in a health check-up population aged 12-19 years in Taiwan province.

METHODThe study data were supplied by the MJ Health Screening Center, which is a private membership chain clinic with 4 health screening centers around the Taiwan Island and provides periodic health examination to its members. The database included a self-administered questionnaire for health history, asking about demographic, socioeconomic, medical, and lifestyle information, and clinical and laboratory measures for every member. A total of 1629 members (873 boys and 756 girls, respectively) received a health check-up first time at MJ centers were recruited from 2005 to 2006. MS detection rate and agreement rate was calculated according to two definitions, respectively. The distributions of MS components and the aggregation of risk factors were further analyzed.

RESULT(1) The range of age-adjusted detection rate of MS for two definitions were 4.05% (5.84% for boys, 1.98% for girls) and 8.35% (10.42% for boys, 5.95% for girls), respectively. It was 0.94% , 14.20% and 36.59% for criterion I among adolescents who were overweight (BMI over 95th percentile), at risk of overweight (BMI between 85th and 95th percentile) and normal weight (BMI below the 85th percentile), respectively; while 3.61%, 25.93% and 53.66% for criterion II. (2) The range of five MS components were 9.09% (low-HDL-C)-16.39% (high blood pressure) for definition I, while 0.98% (high FBG)-27.13% (high WC) for definition II. (3) Of the total subjects, 2.76%, 1.04% and 0.25% were presented with three, four and five MS risk factors for definition I; while 6.69%, 1.60% and 0.34% for definition II, separately. (4) The most common clinical symptom complex of MS was "obesity, hypertension and low-HDL-C" for criterion I, "high TG, obesity and low-HDL-C" for criterion II. (5) The MS diagnostic criterions of I and II were in moderate accordance with agreement rate of 94.35%, Kappa index was 0.518.

CONCLUSIONOur findings reveal that there were relatively large differences in detection and aggregation of risk components on MS when using two recommended definitions, the detection rate of MS in adolescents depends strongly on the parameters chosen and their respective cut-off points. In order to avoid possible relevant under- or over-estimation of the prevalence, it seems advisable that the use of unversally specific cut-off values seems to be more appropriate to give more reliable results.