Adult dyslipidemia prediction is improved by repeated measurements in childhood and young adulthood. The Cardiovascular Risk in Young Finns Study.

BACKGROUND

Prediction of adult dyslipidemia has been suggested to improve with multiple measurements in childhood or young adulthood, but there is paucity of specific data from longitudinal studies.

METHODS AND RESULTS

The sample comprised 1912 subjects (54% women) from the Cardiovascular Risk in Young Finns Study who had fasting lipid and lipoprotein measurements collected at three time-points in childhood/young adulthood and had at least one follow-up in later adulthood. Childhood/young adult dyslipidemia was defined as total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (non-HDL-C) or triglycerides (TG) in the highest quintile, or high-density lipoprotein cholesterol (HDL-C) in the lowest quintile. Adult dyslipidemia was defined according to European cut-points (TC > 5.0 mmol/L, LDL-C >3 mmol/L, Non-HDL-C >3.8 mmol/L, HDL-C <1.0 mmol/L (in men)/<1.2 mmol/L (in women) and TG > 1.7 mmol/L). With the exception of triglycerides, Pearson correlation coefficients for predicting adult levels significantly improved when two lipid or lipoprotein measurements in childhood/young adulthood were compared with one measurement (all P < 0.01). For triglycerides, there was significant improvement only when three measurements were considered (P = 0.004). Two measurements significantly improved prediction of dyslipidemia levels in adulthood for non-HDL-C, LDL-C, HDL-C and TG compared with one measurement (P < 0.05 for improved area-under the receiver-operating characteristic curve). Risk of dyslipidemia in adulthood grew according to the number of times a person had been at risk in childhood.

CONCLUSIONS

Based on these results, it seems that compared to a single measurement two lipid measures in childhood/early adulthood significantly improve prediction of adult dyslipidemia. A lack of dyslipidemia in childhood does not strongly exclude later development of dyslipidemia. Multiple measurements increase the prediction accuracy, but the incremental prognostic/diagnostic accuracy of especially third measurement is modest.