Abstract P128: Serum Lipid Levels at Adolescence can Predict Adult Dyslipidemia in South Korea - The Kangwha Study
Introduction: Several studies have examined tracking pattern of lipid profile level during long follow-up periods in Western countries. However, there have been few such studies in East Asia. Hypothesis: We assessed the hypothesis that there exists tracking pattern of lipid profile level from adolescence to adulthood, and lipid measurements in adolescence can predict adult dyslipidemia in South Korea. Methods: The Kangwha Study was a community-based prospective cohort study that started in 1986 in Kangwha County, South Korea . A total of 400 participants (54% women) whose serum total cholesterol, triglyceride, and high density lipoprotein (HDL) cholesterol level were measured at least once during adolescence (1992-1996), and repeatedly measured at least once during adulthood (2005-2015) were enrolled in our study. Body mass index, waist circumstance, and blood pressure were measured at all measurements. Family history of cardiovascular disease, smoking history, and presence of adult dyslipidemia were checked at adulthood. The tracking pattern of lipid profile level was determined by tracking coefficients (low: <0.30; moderate: 0.30-0.59; moderately high: 0.60-0.89; high: ≥0.90). The tracking coefficients were calculated by Generalized Estimating Equation. The predictability of adult dyslipidemia was assessed by multiple logistic regression and area under curve (AUC) value. Additional analyses were performed to find out whether repeated lipid measurements during adolescence can enhance the predictability of adult dyslipidemia or not. Results: The presence of adult dyslipidemia was 26.3% (105 of 400). Mean age of study participants at enrollment is 13.8 years (SD, 1.6 years), and that at adulthood is 30.1 years (SD, 3.7 years). When adjusted for age, body mass index, waist circumstance, and blood pressure, the tracking coefficient of total cholesterol was 0.59 (95% confidence interval (CI), 0.54-0.63), that of triglyceride was 0.39 (95% CI, 0.28-0.49), and that of HDL cholesterol was 0.51 (95% CI, 0.46-0.55). The AUC value of our multiple logistic regression model on adult dyslipidemia without lipid profile levels at adolescence was 0.77 (95% CI, 0.72-0.83), and that with lipid profile levels at adolescence was 0.80 (95% CI, 0.75-0.85). P value for AUC comparison was significant (p=0.02). In additional analyses, using the average lipid profile levels in multiple lipid measurements at adolescence did not significantly improve the AUC value (p>0.09). Conclusion: In conclusion, moderate tracking patterns of serum lipid profile level were shown in this study. Serum lipid profile measurements at adolescence could help the prediction of adult dyslipidemia. The results of this study supported the need of lipid profile screening at adolescence.