VALIDATION OF A CONTINUOUS MEASURE OF CARDIOMETABOLIC RISK AMONG ADOLESCENTS

Kaigang Li,Denise L. Haynie,Xiang Gao,Leah M. Lipsky,Tonja Nansel,Ronald J. Iannotti,Federico E. Vaca,Bruce G. Simons-Morton
DOI: https://doi.org/10.1515/jpem-2020-0600
2021-01-01
Abstract:Abstract Objectives We validated a continuous cardiometabolic risk (CMR) measure among adolescents. Methods Five metabolic syndrome (MetS) components including waist circumference, triglycerides, high-density lipoprotein cholesterol, fasting blood glucose, and mean arterial pressure were assessed in a national cohort of U.S. adolescents (n=560; 16.5 ± 0.5 y/o at baseline) in 10th grade (2010, Wave 1 (W1)), and follow-up assessments four (W4) and seven (W7) years later. Separately by wave, linear regressions were fitted to each MetS component controlling for age, sex, and race/ethnicity, and yielded standardized residuals (Z-scores). Wave-specific component Z-scores were summed to obtain composite CMR Z-scores. Four- and seven-year CMR change (CMR-diff W1–W4 and W1–W7). and average CMR risk (CMR-avg; (W1 + W4)/2 and (W1 + W7)/2) were calculated using the CMR Z-scores. W7 MetS was determined using adult criteria. Student’s t-test and receiver operating characteristic (ROC) curve were conducted. Results Participants meeting the adult criteria for MetS at W7 (74 of 416, 17.8%) had statistically significant (p<0.01) higher values for W1 CMR Z-scores (0.92 vs. −0.21), W4 CMR Z-scores (1.69 vs. −0.28), W7 CMR Z-scores (2.21 vs. −0.55), W1–W4 CMR-avg (1.53 vs. −0.27), W1–W7 CMR-diff (1.29 vs. −0.21), and W1–W7 CMR-avg (1.46 vs. −0.48) than those not meeting MetS criteria. Most results were similar for males and females in the sex-stratified analyses. The areas under the ROC curve were 0.61, 0.71, and 0.75 for W1, W4 and W7 Z-scores. Conclusions Findings support the validity of the continuous CMR Z-scores calculated using linear regression in evaluating and monitoring CMR profiles from adolescence to early adulthood.
What problem does this paper attempt to address?