We examined if childhood non–HDL-C status predicts high common carotid artery IMT in adulthood.
Abstract
BACKGROUND:
Elevated non–high-density lipoprotein cholesterol (HDL-C) levels are used to identify children at increased cardiovascular risk, but the use of non–HDL-C in childhood to predict atherosclerosis is unclear. We examined whether the National Heart, Lung, and Blood Institute classification of youth non–HDL-C status predicts high common carotid artery intima-media thickness in adulthood.
METHODS:
We analyzed data from 4 prospective cohorts among 4582 children aged 3 to 19 years who were remeasured as adults (mean follow-up of 26 years). Non–HDL-C status in youth and adulthood was classified according to cut points of the National Heart, Lung, and Blood Institute and the National Cholesterol Education Program Adult Treatment Panel III. High carotid intima-media thickness (cIMT) in adulthood was defined as at or above the study visit-, age-, sex-, race-, and cohort-specific 90th percentile of intima-media thickness.
RESULTS:
In a log-binomial regression analysis adjusted with age at baseline, sex, cohort, length of follow-up, baseline BMI, and systolic blood pressure, children with dyslipidemic non–HDL-C were at increased risk of high cIMT in adulthood (relative risk [RR], 1.29; 95% confidence interval [CI], 1.07–1.55). Compared with the persistent normal group, the persistent dyslipidemia group (RR, 1.80; 95% CI, 1.37–2.37) and incident dyslipidemia (normal to dyslipidemia) groups (RR, 1.45; 95% CI, 1.07–1.96) had increased risk of high cIMT in adulthood, but the risk was attenuated for the resolution (dyslipidemia to normal) group (RR, 1.17; 95% CI, 0.97–1.41).
CONCLUSIONS:
Dyslipidemic non–HDL-C levels predict youth at risk for developing high cIMT in adulthood. Those who resolve their non–HDL-C dyslipidemia by adulthood have normalized risk of developing high cIMT in adulthood.
What’s Known on This Subject:
Elevated non–high-density lipoprotein cholesterol (HDL-C) levels are used to identify children at increased cardiovascular risk, but the use of non–HDL-C in childhood to predict atherosclerosis is unclear.
What This Study Adds:
Non–HDL-C levels associate with future risk of preclinical atherosclerosis from the age of 15 years, suggesting a later age for the initial universal lipid screening among the pediatric population than is currently recommended in the National Heart, Lung, and Blood Institute’s expert panel guidelines.
Non–high-density lipoprotein cholesterol (HDL-C) is considered a simpler and more effective screening tool of atherosclerotic cardiovascular disease risk than low-density lipoprotein cholesterol (LDL-C).1,2 Estimated LDL-C does not include all classes of atherogenic lipoproteins,3 may underestimate those with low levels,4 and requires overnight fasting.2 Consequently, non–HDL-C is increasingly used5 and has been specified as a secondary therapy target among patients with the metabolic syndrome or diabetes.6
We have previously reported that elevated LDL-C levels in children and adolescents (youth) were associated with high carotid intima-media thickness (cIMT), a marker of preclinical atherosclerosis, in adulthood.7 For total cholesterol, in our previous analyses it was suggested that measures from those aged ≥12 years, but not in younger children, were associated with adult cIMT.8 The case for non–HDL-C measurement in youth is further strengthened by data in which it is shown that youth non–HDL-C is a better predictor of adult dyslipidemia,9 nonlipid risk factors,9 and cIMT10 than LDL-C. Moreover, non–HDL-C is independently associated with obesity indices11,12 and might provide a more sensitive measure of dyslipidemia than LDL-C among those with overweight or obesity.13 Recognizing the potential value of non–HDL-C measurement, the 2011 National Heart, Lung, and Blood Institute (NHLBI) expert panel recommended universal (population-wide) screening of nonfasting non–HDL-C levels first at age 9 to 11 years and again at age 18 to 21 years to identify youth with dyslipidemia at risk for accelerated atherosclerotic disease.14 In contrast to previous guidelines on lipid screening,15,16 the NHLBI guidelines were the first to recommend universal versus selected screening for lipid disorders and incorporate cutoffs for non–HDL-C levels, derived from population-based data in the Bogalusa Heart Study.12 However, these data were cross-sectional, and it is unknown whether the NHLBI cutoffs for non–HDL-C predict future preclinical atherosclerosis.
Using data from 4 population-based prospective cohorts beginning in youth with follow-up into adulthood, we examined if the NHLBI classification of youth non–HDL-C status is associated with adult cIMT. We also compared associations for LDL-C and examined whether resolution of elevated youth non–HDL-C status by adulthood reduces the risk of developing high cIMT.
Methods
Study Sample
The study sample was drawn from 4 prospective cohorts in the i3C Consortium.17 These were the Bogalusa Heart Study (LA), the Insulin Study (Minneapolis), the Cardiovascular Risk in Young Finns Study (YFS) (Finland), and the Childhood Determinants of Adult Health (CDAH) Study (Australia). Study characteristics and methods have been previously described.18–23 Although loss to follow-up varied by cohort, previous analyses have suggested the representativeness of the cohorts has largely been maintained.17 In total, 4582 participants with non–HDL-C data from their first study visit in youth when aged 3 to 19 years and longitudinal ultrasound data from adulthood when aged 19 to 51 years were included. Local ethics committees reviewed and approved the individual cohort studies that we analyzed, and participants in those studies (or their legal guardians) provided written informed consent. The present analysis conformed to the Declaration of Helsinki.
Risk Factor Assessment
In the YFS at baseline, serum cholesterol and triglycerides were measured using fully enzymatic Boehringer CHOD-PAP kits with an OLLI 3000 analyzer. Subsequently, an Olympus System reagent analyzer in a clinical chemistry analyzer (AU400; Olympus) was used. Serum HDL-C was measured by the dextran sulfate 500 000 method. In CDAH, serum total cholesterol and triglycerides were determined according to the Lipid Research Clinics Program,24 and HDL-C was analyzed after precipitation of apolipoprotein B–containing lipoproteins with heparin-manganese.25 In the Bogalusa Heart Study, HDL-C and triglycerides were measured by using chemical procedures with a Technicon Auto Analyzer II (Technicon Instrument Corp), according to the Lipid Research Clinics Program.24,26 Commencing after baseline, these variables were determined by enzymatic procedures by using the Abbott VP instrument (Abbott Laboratories).27 Serum concentrations of LDL-C and HDL-C were analyzed by a combination of heparin-calcium precipitation and agar-agarose gel electrophoresis procedures. In the Insulin Study, serum lipids were analyzed in the University of Minnesota laboratory with a Cobas FARA.28 HDL-C was determined after precipitation of non–high-density lipoproteins with a dextran-sulfate magnesium precipitating reagent. Triglycerides were determined with a standard glycerol blanked enzymatic triglyceride method. For all cohorts, non–HDL-C was calculated as total cholesterol – HDL-C, and LDL-C was calculated using the Friedewald formula.29 The coefficient of variation for within-assay precision in the YFS was 2.2% for total cholesterol, 2.3% for HDL-C, and 3.8% for serum triglycerides. Both of the US cohorts and CDAH used chemical and enzymatic procedures meeting the performance requirements of the Lipid Clinics Program and Lipid Standardization Program of the Centers for Disease Control and Prevention, which routinely monitors the accuracy of measurements of total cholesterol, triglyceride, and HDL-C concentrations. Height and weight were measured and used to calculate BMI as weight (kilograms)/(height [meters]2). Cole’s international BMI cutoffs30 were used to denote weight status. Systolic blood pressure at baseline was measured by using a standard mercury sphygmomanometer.31 Youth smoking habits were assessed by questionnaire. Those who had smoked ≥1 cigarette per day in youth (<20 years old) were considered smokers.
Ultrasound Measurements
B-mode ultrasound studies of the left common carotid artery were performed at follow-up examinations by using standardized protocols in each study. Details of the ultrasound data, protocols, and reproducibility have been described elsewhere.7,32
Exposure and Outcome Definitions
In youth, non–HDL-C status was defined as normal if <3.10 mmol/L (<120 mg/dL), elevated if 3.10 to <3.75 mmol/L (120– <145 mg/dL), and dyslipidemia if ≥3.75 mmol/L (≥145 mg/dL), and LDL-C status was defined as normal if <2.85 mmol/L (<110 mg/dL), elevated if 2.85 to <3.36 mmol/L (110– <130 mg/dL), and dyslipidemia if ≥3.36 mmol/L (≥130 mg/dL) according to cut points from the NHLBI expert panel.14 Non–HDL-C status in adulthood was defined as normal if <4.91 mmol/L (<190 mg/dL) and dyslipidemia if ≥4.91 mmol/L (≥190 mg/dL), and LDL-C was defined as normal if <4.14 mmol/L (<160 mg/dL) and dyslipidemia if ≥4.14 mmol/L (≥160 mg/dL). Change in non–HDL-C and LDL-C status between youth and adulthood was defined as persistent dyslipidemia (dyslipidemia at both time points), incident dyslipidemia (normal to dyslipidemia), resolution (dyslipidemia to normal), and persistent normal (normal at both time points). The latest available measurement of cIMT was used, and high cIMT in adulthood was defined as at or above the follow-up year-, age-, sex-, race-, and cohort-specific 90th percentile. Because there was a small number of participants in some categories after stratification by age, sex, race, cohort, and follow-up years (in which the proportion of high-risk cIMT ranged between 10% and 20%), the combined average rate for the pooled data was higher than the expected 10%.
Statistical Analysis
Univariable and/or multivariable modified Poisson regression models (using a robust error variance) were used to estimate the relative risk (RR) and 95% confidence intervals (CIs) for the association of youth lipids and changes in their status between youth and adulthood with adult risk of having high cIMT. Because weight status is thought to influence the predictive use of non–HDL-C versus LDL-C, we performed a sensitivity analysis stratified by Cole BMI weight categories.30 There were significant interactions between youth non–HDL-C and LDL-C status with cohort and youth age but not sex. Therefore, we also conducted multivariable modified Poisson regression models for associations between youth non–HDL-C and LDL-C status and adult risk of high intima-media thickness (IMT) stratified by cohort and youth age groups (3–8, 9–11, 12–14, 15–17, and 18+ years). Youth age groups were based on current risk screening age windows used by the NHLBI in which universal screening occurs from age 9 to 11 years and again at 18+ years.14 All multivariable analyses were adjusted for age, BMI, and systolic blood pressure at baseline, sex, cohort, and length of follow-up. Logistic regression models (all covariates adjusted above were included in the model) were used to obtain area under receiver-operating curve (AUC) values to estimate and compare the predictive use of youth non–HDL-C and LDL-C on adult risk of having high IMT. All analyses were re-run after additional adjustment for youth smoking or exclusion of those individuals having lipid-lowering medication (n = 155, all in adulthood: 112 from Bogalusa Heart Study, 2 from CDAH, and 41 from YFS), and risk ratios remained essentially similar in these analyses. As a sensitivity analysis, funnel plots were generated for the association of non–HDL-C or LDL-C status in youth with carotid artery IMT ≥90th percentile in adulthood. Analyses were performed in Stata version 15.1 (Stata Corp, College Station, TX). A 2-tailed P value <.05 was considered statistically significant.
Results
Characteristics are shown in Supplemental Table 6. Youth characteristics among those who had elevated non–HDL-C or LDL-C are presented in Table 1. Of those with elevated youth non–HDL-C levels, 2335 (94%) also had elevated youth LDL-C levels. Moreover, 150 participants had elevated youth non–HDL-C but not elevated LDL-C levels, and 70 had elevated youth LDL-C but not elevated non–HDL-C levels. A total of 47% of those with elevated youth non–HDL-C levels had elevated levels as adults, compared with 57% of those with elevated youth LDL-C levels having elevated levels as adults (Table 1).
TABLE 1.
Characteristics Among Those With Elevated non–HDL-C and LDL-C in Youth
| Elevated non–HDL-C in Youth | Elevated LDL-C in Youth | |
|---|---|---|
| n (%)a | 2485 (54) | 2405 (52) |
| Boys, n (%)b | 1079 (43) | 1045 (43) |
| Mean age in y, (SD) | 10.7 (4.6) | 10.6 (4.6) |
| Mean non–HDL-C, mmol/L (SD) | 3.98 (0.68) | 4.00 (0.69) |
| Mean LDL-C, mmol/L (SD) | 3.64 (0.68) | 3.68 (0.66) |
| Mean total cholesterol, mmol/L (SD) | 5.50 (0.77) | 5.53 (0.76) |
| Mean HDL-C, mmol/L (SD) | 1.52 (0.36) | 1.54 (0.36) |
| Mean triglycerides, mmol/L (SD) | 0.77 (0.41) | 0.73 (0.37) |
| Mean BMI, kg/m2 (SD) | 18.1 (3.6) | 17.9 (3.5) |
| Mean systolic blood pressure, mm Hg (SD) | 110.9 (12.4) | 110.7 (12.4) |
| Elevated non–HDL-C in youth, n (%) | 2485 (100) | 2335 (97) |
| Elevated LDL-C in youth, n (%) | 2335 (94) | 2405 (100) |
| Elevated non–HDL-C in adulthood, n (%) | 1164 (47) | 1135 (47) |
| Elevated LDL-C in adulthood, n (%) | 1408 (57) | 1378 (57) |
To convert non–HDL-C, LDL-C, and total cholesterol from mmol/L to mg/dL, multiply values by 38.67. To convert triglycerides from mmol/L to mg/dL, multiply values by 88.57.
Percent of the total number of study participants (N = 4582).
Percent of the number of participants with elevated youth non–HDL-C (n = 2485) or LDL-C (n = 2405).
Pooled RR and their 95% CIs for high adult cIMT according to non–HDL-C and LDL-C status in youth are shown in Table 2. Compared with those classified as not having dyslipidemia, youth with non–HDL-C dyslipidemia or LDL-C dyslipidemia were at increased risk of high cIMT in adulthood. Additional adjustment for youth BMI and systolic blood pressure (model 2, Table 2) did not appreciably change the effect estimates. The AUC was comparable between non–HDL-C and LDL-C models, with AUC of 0.622 and 0.620, respectively (P > .5). The AUC was similar between non–HDL-C and LDL-C when stratified by BMI status (normal weight: non–HDL-C AUC of 0.630 versus LDL-C AUC of 0.628, P > .5; overweight or obese: non–HDL-C AUC of 0.552 versus LDL-C AUC of 0.549, P > .5). A sensitivity analysis using funnel plots showed an asymmetry in the scatter of small studies, with more studies showing a lower-magnitude association between youth non–HDL-C or LDL-C with adult risk of cIMT (Supplemental Fig 1). In an additional pooled analysis based on 3 smaller cohorts (ie, excluding the Young Finns data), youth non–HDL-C dyslipidemia or LDL-C dyslipidemia was not associated with the risk of high cIMT in adulthood (Supplemental Table 7).
TABLE 2.
RR and 95% CIs From Pooled Data for the Association of non–HDL-C or LDL-C Status in Youth With Carotid Artery IMT ≥90th Percentile in Adulthood (n = 4582)
| n Out of Na | % | Model 1 | Model 2 | |
|---|---|---|---|---|
| RR (95% CI) | RR (95% CI) | |||
| Non–HDL-C | ||||
| Normal | 302 out of 2097 | 14.4 | Reference | Reference |
| Elevated | 148 out of 1159 | 12.8 | 1.08 (0.89–1.31) | 1.05 (0.87–1.28) |
| Dyslipidemia | 197 out of 1326 | 14.9 | 1.35 (1.12–1.63)b | 1.29 (1.07–1.55)b |
| LDL-C | ||||
| Normal | 317 out of 2177 | 14.6 | Reference | Reference |
| Elevated | 119 out of 945 | 12.6 | 1.05 (0.85–1.29) | 1.03 (0.84–1.27) |
| Dyslipidemia | 211 out of 1460 | 14.5 | 1.30 (1.08–1.57)b | 1.26 (1.04–1.51)b |
Model 1 is adjusted for age at baseline, sex, cohort, and length of follow-up. Model 2 includes model 1 covariates plus baseline BMI and systolic blood pressure.
Number of participants with high IMT out of all participants.
Statistical significance, P < .05.
Age-stratified results for the association between youth non–HDL-C and LDL-C status with adult high cIMT are shown in Table 3. Those with dyslipidemia aged 15 to 17 years had significantly increased risk of adult high cIMT, as did those aged 18 years or over with elevated non–HDL-C or LDL-C status.
TABLE 3.
RR and 95% CIs for the Association of non–HDL-C and LDL-C Status in Youth With Carotid Artery IMT ≥90th Percentile in Adulthood Stratified by Youth Age Group (n = 4582)
| Youth Age Group, ya | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 3–8 | 9–11 | 12–14 | 15–17 | 18+ | |||||||||||
| n Out of Nb | % | RR (95% CI) | n Out of Nb | % | RR (95% CI) | n Out of Nb | % | RR (95% CI) | n Out of Nb | % | RR (95% CI) | n Out of Nb | % | RR (95% CI) | |
| Non–HDL-C | |||||||||||||||
| Normal | 117 out of 633 | 18.5 | Reference | 54 out of 392 | 13.8 | Reference | 77 out of 627 | 12.3 | Reference | 47 out of 342 | 13.7 | Reference | 7 out of 104 | 6.7 | Reference |
| Elevated | 45 out of 371 | 12.1 | 0.89 (0.63–1.24) | 28 out of 238 | 11.8 | 0.95 (0.61–1.49) | 31 out of 240 | 12.9 | 1.27 (0.83–1.95) | 25 out of 191 | 13.1 | 1.11 (0.70–1.75) | 19 out of 119 | 16.0 | 2.43 (1.10–5.38)c |
| Dyslipidemia | 68 out of 477 | 14.3 | 1.10 (0.80–1.52) | 45 out of 280 | 16.1 | 1.42 (0.92–2.20) | 34 out of 263 | 12.9 | 1.36 (0.91–2.03) | 32 out of 177 | 18.1 | 1.72 (1.08–2.72)c | 18 out of 130 | 13.9 | 2.20 (0.98–4.94) |
| LDL-C | |||||||||||||||
| Normal | 116 out of 629 | 18.4 | Reference | 56 out of 398 | 14.1 | Reference | 86 out of 673 | 12.8 | Reference | 49 out of 359 | 13.7 | Reference | 10 out of 119 | 8.4 | Reference |
| Elevated | 34 out of 296 | 11.5 | 0.84 (0.59–1.21) | 25 out of 188 | 13.3 | 1.05 (0.66–1.67) | 21 out of 190 | 11.1 | 1.04 (0.65–1.66) | 20 out of 169 | 11.8 | 1.08 (0.66–1.77) | 19 out of 102 | 18.6 | 2.25 (1.12–4.55)c |
| Dyslipidemia | 80 out of 556 | 14.4 | 1.15 (0.85–1.57) | 46 out of 324 | 14.2 | 1.24 (0.80–1.92) | 35 out of 267 | 13.1 | 1.35 (0.90–2.03) | 35 out of 182 | 19.2 | 1.97 (1.24–3.15)c | 15 out of 132 | 11.4 | 1.40 (0.66–2.99) |
All analyses were adjusted for age, BMI, and systolic blood pressure at baseline, sex, cohort, and length of follow-up.
The age when non–HDL-C or LDL-C levels were first measured in youth.
Number of participants with high IMT out of all participants.
Statistical significance, P < .05.
Table 4 shows results for adult high cIMT by youth non–HDL-C and LDL-C status, stratified by cohort. Within each cohort, associations were similar for both elevated and dyslipidemia non–HDL-C and LDL-C classifications. Between each cohort, there was large heterogeneity in effect estimates for elevated and dyslipidemia status in both non–HDL-C and LDL-C, with only participants in the YFS showing a consistent and graded increase in risk for adult high cIMT on the basis of youth non–HDL-C or LDL-C status.
TABLE 4.
RR and 95% CIs for the Association of non–HDL-C and LDL-C Status in Youth With Carotid Artery IMT ≥90th Percentile in Adulthood by Cohorts (n = 4582)
| CDAH | Bogalusa | Insulin Study | YFS | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n Out of Na | % | RR (95% CI) | n Out of Na | % | RR (95% CI) | n Out of Na | % | RR (95% CI) | n Out of Na | % | RR (95% CI) | |
| Non–HDL-C | ||||||||||||
| Normal | 48 out of 234 | 20.5 | Reference | 177 out of 1068 | 16.6 | Reference | 34 out of 209 | 16.3 | Reference | 43 out of 586 | 7.3 | Reference |
| Elevated | 20 out of 107 | 18.7 | 0.88 (0.55–1.41) | 25 out of 150 | 16.7 | 0.96 (0.66–1.39) | 8 out of 43 | 18.6 | 0.85 (0.38–1.87) | 95 out of 859 | 11.1 | 1.51 (1.07–2.14)b |
| Dyslipidemia | 16 out of 60 | 26.7 | 1.21 (0.74–1.99) | 11 out of 74 | 14.9 | 0.71 (0.41–1.22) | 2 out of 33 | 6.1 | 0.27 (0.07–1.11) | 168 out of 1159 | 14.5 | 1.92 (1.39–2.65)b |
| LDL-C | ||||||||||||
| Normal | 51 out of 248 | 20.6 | Reference | 178 out of 1075 | 16.6 | Reference | 38 out of 234 | 16.2 | Reference | 50 out of 620 | 8.1 | Reference |
| Elevated | 15 out of 85 | 17.7 | 0.85 (0.51–1.42) | 23 out of 140 | 16.4 | 0.93 (0.63–1.38) | 5 out of 32 | 15.6 | 0.65 (0.28–1.49) | 76 out of 688 | 11.1 | 1.41 (1.00–1.97)b |
| Dyslipidemia | 18 out of 68 | 26.5 | 1.20 (0.74–1.93) | 12 out of 77 | 15.6 | 0.75 (0.44–1.25) | 1 out of 19 | 5.3 | 0.26 (0.04–1.92) | 180 out of 1296 | 13.9 | 1.70 (1.26–2.29)b |
All analyses were adjusted for age, BMI, and systolic blood pressure at baseline, sex, cohort, and length of follow-up.
Number of participants with high IMT out of all participants.
Statistical significance, P < .05.
Table 5 shows the pooled RR and their 95% CIs for high adult cIMT according to youth and adult non–HDL-C and LDL-C status. Compared with the persistent normal non–HDL-C group, those with persistent or incident non–HDL-C dyslipidemia had increased risk of high cIMT in adulthood; increased but weaker associations were observed for resolution and incident non–HDL-C dyslipidemia groups. On the basis of LDL-C classification in youth and adulthood, only those with persistent dyslipidemia had significantly increased risk of high adult cIMT compared with the normal LDL group (Table 5). We observed similar risk estimates after further adjustment for youth BMI and systolic blood pressure (model 2, Table 5).
TABLE 5.
RR and 95% CIs From Pooled Data for the Association of Change in non–HDL-C or LDL-C Status Between Youth and Adulthood With Carotid Artery IMT ≥90th Percentile in Adulthood (n = 4582)
| n Out of Na | % | Model 1 | Model 2 | |
|---|---|---|---|---|
| RR (95% CI) | RR (95% CI) | |||
| Non–HDL-C | ||||
| Persistent normal | 411 out of 3056 | 13.5 | Reference | Reference |
| Incident dyslipidemia | 39 out of 200 | 19.5 | 1.45 (1.07–1.96)b | 1.47 (1.09–1.98)b |
| Resolution | 147 out of 1082 | 13.6 | 1.21 (1.00–1.46)b | 1.17 (0.97–1.41) |
| Persistent dyslipidemia | 50 out of 244 | 20.5 | 1.80 (1.37–2.37)b | 1.74 (1.33–2.28)b |
| LDL-C | ||||
| Persistent normal | 400 out of 2900 | 13.8 | Reference | Reference |
| Incident dyslipidemia | 36 out of 222 | 16.2 | 1.14 (0.83–1.56) | 1.15 (0.84–1.58) |
| Resolution | 161 out of 1189 | 13.5 | 1.20 (0.99–1.45) | 1.16 (0.96–1.40) |
| Persistent dyslipidemia | 50 out of 271 | 18.5 | 1.60 (1.21–2.10)b | 1.55 (1.18–2.04)b |
Model 1 is adjusted for age at baseline, sex, cohort, and length of follow-up. Model 2 includes model 1 covariates plus baseline BMI and systolic blood pressure.
Number of participants with high IMT out of all participants.
Statistical significance, P < .05.
Discussion
These longitudinal data suggest that youth non–HDL-C levels are associated with high cIMT, in adulthood. In age-stratified analyses, this relationship was observed for non–HDL-C measurements performed from the age of 15 years. The predictive use for high adult cIMT was similar using either LDL-C or non–HDL-C.
We have previously reported that elevated youth LDL-C levels7 and total cholesterol8 were associated with high cIMT in adulthood. For LDL-C we observed that only persistent dyslipidemia was related to higher cIMT.7 For total cholesterol, our previous analyses suggested that the measures in those aged ≥12 years, but not in younger children, were associated with high adult cIMT.32 The present data for non–HDL-C are in line with these observations but provide important additional information that non–HDL-C levels in youth, especially among those aged ≥15 years, are predictive of cIMT. In analyses in which both youth and adult non–HDL-C levels are taken into account, those with elevated adult non–HDL-C levels had high cIMT independent of their youth non–HDL-C status. Among those individuals with elevated non–HDL-C in youth but not in adulthood, the increased risk for developing high cIMT was partly attenuated (RR, 1.17; 95% CI, 0.97–1.41). In cohort-specific analyses, our findings were somewhat different, particularly for the US cohorts. Only in the YFS did we observe consistent relations between youth lipids with adult carotid IMT, whereas pooled analyses restricted on 3 smaller cohorts showed no associations. Possible explanations for this are complex and may include the larger sample and case number size (as suggested by our funnel plots), longer follow-up time, higher childhood lipid levels in the YFS (highest in the world in 1970s33), providing larger lifetime lipid exposure, and differences in lipid measurement and IMT methodology across the cohorts.
The 2011 NHLBI Expert Panel Guidelines14 were the first to suggest universal lipid screening in youth, initially at age 9 to 11 years and again at 18 to 21 years, using non–HDL-C as the preferred lipid measure. We observed that youth non–HDL-C measurements are comparable to LDL-C in predicting preclinical atherosclerosis. However, there is still limited evidence on population-based interventions in youth to reduce cIMT over the long-term. Screening using non–HDL-C comes with the benefit of not requiring the measurement of triglycerides to calculate LDL-C and the subsequent advantage of not requiring patients to provide a fasting sample. Concerning the optimal age for lipid screening, our observational data suggest that neither non–HDL-C nor LDL-C levels at the age of 9 to 11 years are associated with subsequent cIMT because associations only became evident from age 15 years onwards. However, our findings need to be interpreted with caution because effect estimates were inconsistent across ages. In pooled estimates, elevated lipids significantly related with later cIMT only among those aged 18 to 19 years and dyslipidemic levels only among 15- to 17-year-olds.
For preventive interventions, our findings that the adverse effects of youth dyslipidemia are attenuated if lipid status is improved or normalized by adulthood are encouraging. These data indicate there is a window for change in late adolescence and young adulthood in which individual- and public health–focused programs might have long-term preventive effects. In addition, because the effect of elevated childhood non–HDL-C was not completely attenuated or reversed in the resolution group, primordial prevention will continue to be an important goal. We have shown in these cohorts that although lipid levels track, or persist, well from youth to adulthood,34 those able to change from high risk in youth to normal risk in adulthood coincide with healthful lifestyle changes, such as lower gains in fatness and improvements in cardiorespiratory fitness relative to their peers, not smoking, and upward mobility in education.35–38 Higher gains in BMI from youth to adulthood have been associated with more adverse adult lipid levels irrespective of genetic susceptibility.39 Indeed, the genetic effect on life-course lipid levels tends to persist, or slightly weaken, with age, suggesting greater importance of lifestyle factors at different life stages.39 We have also shown evidence of an infant-onset dietary counseling intervention aimed primarily at improving fat quality in the diet but also promoting intake of fruit, vegetables, and whole grains to associate with a higher likelihood of achieving dietary guidelines and reduced lipid levels even into adulthood.40,41 These data provide insight on lifestyle and environmental factors that could be targeted at individual or population-wide prevention.
The main strength of this study is the use of pooled data on youth risk factors and adult cIMT from 4 international longitudinal cohorts. The study also has some limitations. First, because the study cohorts are composed of relatively young adults at follow-up, we were not able to study associations with cardiovascular events. Instead, we have used cIMT as a surrogate end point with the risk stratification groupings, not based on absolute risk of cardiovascular events (as in adult risk score algorithms) but on high cIMT (defined as ≥90th percentile). However, in older adults, cIMT has been shown to predict subsequent cardiovascular disease events.42 Second, study participants were predominantly white, and the results may not be generalizable to other ethnicities. Third, observational studies are prone to bias when trying to establish causality. Finally, concerning the possible confounding factors, most cohorts did not have data available on some possible childhood confounders, such as socioeconomic factors.
Conclusions
In our analysis, we show that elevated non–HDL-C levels in youth are related to high cIMT in adulthood. In age-stratified analyses, a significant association was observed if non–HDL-C levels were measured at the age of 15 to 19 years. The predictive use of youth non–HDL-C and LDL-C were similar. It is also demonstrated in the data that individuals with normal non–HDL-C in youth but elevated non–HDL-C in adulthood had high cIMT, whereas those with dyslipidemia in youth but normal non–HDL-C as adults had the proportion of high cIMT comparable to those who never had dyslipidemia.
Glossary
- AUC
area under receiver-operating curve
- CDAH
Childhood Determinants of Adult Health
- CI
confidence interval
- cIMT
carotid intima-media thickness
- HDL-C
high-density lipoprotein cholesterol
- IMT
intima-media thickness
- LDL-C
low-density lipoprotein cholesterol
- NHLBI
National Heart, Lung, and Blood Institute
- RR
relative risk
- YFS
Young Finns Study
Footnotes
Drs Juonala, Wu, and Magnussen conceptualized and designed the study, analyzed the data, drafted the initial manuscript, and reviewed and revised the manuscript; Drs Sinaiko, Woo, Urbina, Jacobs, Steinberger, Prineas, Burns, Bazzano, Venn, Viikari, Hutri-Kähönen, Daniels, Dwyer, and Raitakari designed and executed data collection, collected data, and critically reviewed and revised the manuscript for important intellectual content; Drs Koskinen, Sabin, and Burgner conceptualized and designed the study, interpreted the data, and critically reviewed and revised the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by a grant from the National Institutes of Health and National Heart, Lung, and Blood Institute (R01 HL121230) and the Australian National Health and Medical Research Council Project (grant APP1098369). Dr Magnussen was supported by a National Heart Foundation of Australia Future Leader Fellowship (100849). The Cardiovascular Risk in Young Finns Study has been financially supported by the Academy of Finland (grants 126925, 121584, 124282, 129378, 117787, and 41071); the Social Insurance Institution of Finland; Kuopio, Tampere, and Turku University Hospital medical funds; the Juho Vainio Foundation; the Paavo Nurmi Foundation; the Finnish Foundation of Cardiovascular Research; the Finnish Cultural Foundation; the Sigrid Juselius Foundation; and the Yrjö Jahnsson Foundation. The Childhood Determinants of Adult Health baseline study received grant support from the Commonwealth Departments of Sport, Recreation, and Tourism and Health; The National Heart Foundation; and the Commonwealth Schools Commission. The Childhood Determinants of Adult Health follow-up study received grant support from the National Health and Medical Research Council (APP211316), the National Heart Foundation (GOOH 0578), the Tasmanian Community Fund (D0013808), and Veolia Environmental Services. Sponsors included Sanitarium, ASICS, and Target. Funding bodies and sponsors did not play a role in the study design; collection, analysis, and interpretation of data; writing of the manuscript; or the decision to submit the manuscript for publication. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2020-0159.
References
- 1.Sniderman AD, Williams K, Contois JH, et al. . A meta-analysis of low-density lipoprotein cholesterol, non-high-density lipoprotein cholesterol, and apolipoprotein B as markers of cardiovascular risk. Circ Cardiovasc Qual Outcomes. 2011;4(3):337–345 [DOI] [PubMed] [Google Scholar]
- 2.Frost PH, Havel RJ. Rationale for use of non-high-density lipoprotein cholesterol rather than low-density lipoprotein cholesterol as a tool for lipoprotein cholesterol screening and assessment of risk and therapy. Am J Cardiol. 1998;81(4A):26B–31B [DOI] [PubMed] [Google Scholar]
- 3.Miller M, Ginsberg HN, Schaefer EJ. Relative atherogenicity and predictive value of non-high-density lipoprotein cholesterol for coronary heart disease. Am J Cardiol. 2008;101(7):1003–1008 [DOI] [PubMed] [Google Scholar]
- 4.Quispe R, Hendrani A, Elshazly MB, et al. . Accuracy of low-density lipoprotein cholesterol estimation at very low levels. BMC Med. 2017;15(1):83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ramjee V, Sperling LS, Jacobson TA. Non-high-density lipoprotein cholesterol versus apolipoprotein B in cardiovascular risk stratification: do the math. J Am Coll Cardiol. 2011;58(5):457–463 [DOI] [PubMed] [Google Scholar]
- 6.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA. 2001;285(19):2486–2497 [DOI] [PubMed] [Google Scholar]
- 7.Magnussen CG, Venn A, Thomson R, et al. . The association of pediatric low- and high-density lipoprotein cholesterol dyslipidemia classifications and change in dyslipidemia status with carotid intima-media thickness in adulthood evidence from the cardiovascular risk in Young Finns study, the Bogalusa Heart study, and the CDAH (Childhood Determinants of Adult Health) study. J Am Coll Cardiol. 2009;53(10):860–869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Juonala M, Magnussen CG, Venn A, et al. . Influence of age on associations between childhood risk factors and carotid intima-media thickness in adulthood: the Cardiovascular Risk in Young Finns Study, the Childhood Determinants of Adult Health Study, the Bogalusa Heart Study, and the Muscatine Study for the International Childhood Cardiovascular Cohort (i3C) Consortium. Circulation. 2010;122(24):2514–2520 [DOI] [PubMed] [Google Scholar]
- 9.Srinivasan SR, Frontini MG, Xu J, Berenson GS. Utility of childhood non-high-density lipoprotein cholesterol levels in predicting adult dyslipidemia and other cardiovascular risks: the Bogalusa Heart Study. Pediatrics. 2006;118(1):201–206 [DOI] [PubMed] [Google Scholar]
- 10.Frontini MG, Srinivasan SR, Xu J, Tang R, Bond MG, Berenson GS. Usefulness of childhood non-high density lipoprotein cholesterol levels versus other lipoprotein measures in predicting adult subclinical atherosclerosis: the Bogalusa Heart Study. Pediatrics. 2008;121(5):924–929 [DOI] [PubMed] [Google Scholar]
- 11.Dai S, Eissa MA, Steffen LM, Fulton JE, Harrist RB, Labarthe DR. Associations of BMI and its fat-free and fat components with blood lipids in children: Project HeartBeat!. Clin Lipidol. 2011;6(2):235–244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Srinivasan SR, Myers L, Berenson GS. Distribution and correlates of non-high-density lipoprotein cholesterol in children: the Bogalusa Heart Study. Pediatrics. 2002;110(3):e29. [DOI] [PubMed] [Google Scholar]
- 13.Dai S, Yang Q, Yuan K, et al. . Non-high-density lipoprotein cholesterol: distribution and prevalence of high serum levels in children and adolescents: United States National Health and Nutrition Examination Surveys, 2005-2010. J Pediatr. 2014;164(2):247–253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Expert Panel on Integrated Guidelines for Cardiovascular Health and Risk Reduction in Children and Adolescents; National Heart, Lung, and Blood Institute . Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics. 2011;128(suppl 5):S213–S256 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.American Academy of Pediatrics. American Academy of Pediatrics. National cholesterol education program: report of the expert panel on blood cholesterol levels in children and adolescents. Pediatrics. 1992;89(3, pt 2):525–584 [PubMed] [Google Scholar]
- 16.Daniels SR, Greer FR; Committee on Nutrition . Lipid screening and cardiovascular health in childhood. Pediatrics. 2008;122(1):198–208 [DOI] [PubMed] [Google Scholar]
- 17.Dwyer T, Sun C, Magnussen CG, et al. . Cohort Profile: the international childhood cardiovascular cohort (i3C) consortium. Int J Epidemiol. 2013;42(1):86–96 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Moran A, Jacobs DR Jr., Steinberger J, et al. . Changes in insulin resistance and cardiovascular risk during adolescence: establishment of differential risk in males and females. Circulation. 2008;117(18):2361–2368 [DOI] [PubMed] [Google Scholar]
- 19.Rasmussen-Torvik LJ, Pankow JS, Jacobs DR Jr., Steinberger J, Moran A, Sinaiko AR. Development of associations among central adiposity, adiponectin and insulin sensitivity from adolescence to young adulthood. Diabet Med. 2012;29(9):1153–1158 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Berenson GS, Srinivasan SR, Bao W, Newman WP III, Tracy RE, Wattigney WA. Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. The Bogalusa Heart Study. N Engl J Med. 1998;338(23):1650–1656 [DOI] [PubMed] [Google Scholar]
- 21.Raitakari OT, Juonala M, Kähönen M, et al. . Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood: the Cardiovascular Risk in Young Finns Study. JAMA. 2003;290(17):2277–2283 [DOI] [PubMed] [Google Scholar]
- 22.Raitakari OT, Juonala M, Rönnemaa T, et al. . Cohort profile: the cardiovascular risk in Young Finns Study. Int J Epidemiol. 2008;37(6):1220–1226 [DOI] [PubMed] [Google Scholar]
- 23.Dwyer T, Magnussen CG, Schmidt MD, et al. . Decline in physical fitness from childhood to adulthood associated with increased obesity and insulin resistance in adults. Diabetes Care. 2009;32(4):683–687 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lipid Research Clinics Program Manual of Laboratory Operations: Lipid and Lipoprotein Analysis, vol. Vol. 1 Bethesda, MD: National Institutes of Health: US Dept of Health, Education and Welfare; 1974 [Google Scholar]
- 25.Dwyer T, Gibbons LE. The Australian Schools Health and Fitness Survey. Physical fitness related to blood pressure but not lipoproteins. Circulation. 1994;89(4):1539–1544 [DOI] [PubMed] [Google Scholar]
- 26.Srinivasan SR, Berenson GS. Serum Lipoproteins in Children and Methods for Study In: Lewis L, ed.. Handbook of Electrophoresis. Boca Raton, FL: CRC Press; 1983:185–204 [Google Scholar]
- 27.Allain CC, Poon LS, Chan CS, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem. 1974;20(4):470–475 [PubMed] [Google Scholar]
- 28.Sinaiko AR, Jacobs DR Jr., Steinberger J, et al. . Insulin resistance syndrome in childhood: associations of the euglycemic insulin clamp and fasting insulin with fatness and other risk factors. J Pediatr. 2001;139(5):700–707 [DOI] [PubMed] [Google Scholar]
- 29.Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502 [PubMed] [Google Scholar]
- 30.Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(7244):1240–1243 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Koskinen J, Juonala M, Dwyer T, et al. . Utility of different blood pressure measurement components in childhood to predict adult carotid intima-media thickness. Hypertension. 2019;73(2):335–341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Dengel DR, Jacobs DR, Steinberger J, Moran AM, Sinaiko AR. Gender differences in vascular function and insulin sensitivity in young adults. Clin Sci (Lond). 2011;120(4):153–160 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Knuiman JT, Hermus RJ, Hautvast JG. Serum total and high density lipoprotein (HDL) cholesterol concentrations in rural and urban boys from 16 countries. Atherosclerosis. 1980;36(4):529–537 [DOI] [PubMed] [Google Scholar]
- 34.Juhola J, Magnussen CG, Viikari JS, et al. . Tracking of serum lipid levels, blood pressure, and body mass index from childhood to adulthood: the Cardiovascular Risk in Young Finns Study. J Pediatr. 2011;159(4):584–590 [DOI] [PubMed] [Google Scholar]
- 35.Magnussen CG, Thomson R, Cleland VJ, Ukoumunne OC, Dwyer T, Venn A. Factors affecting the stability of blood lipid and lipoprotein levels from youth to adulthood: evidence from the Childhood Determinants of Adult Health Study. Arch Pediatr Adolesc Med. 2011;165(1):68–76 [DOI] [PubMed] [Google Scholar]
- 36.Porkka KV, Viikari JS, Åkerblom HK. Tracking of serum HDL-cholesterol and other lipids in children and adolescents: the Cardiovascular Risk in Young Finns Study. Prev Med. 1991;20(6):713–724 [DOI] [PubMed] [Google Scholar]
- 37.Porkka KV, Viikari JS, Taimela S, Dahl M, Åkerblom HK. Tracking and predictiveness of serum lipid and lipoprotein measurements in childhood: a 12-year follow-up. The Cardiovascular Risk in Young Finns study. Am J Epidemiol. 1994;140(12):1096–1110 [DOI] [PubMed] [Google Scholar]
- 38.Webber LS, Srinivasan SR, Wattigney WA, Berenson GS. Tracking of serum lipids and lipoproteins from childhood to adulthood. The Bogalusa Heart Study. Am J Epidemiol. 1991;133(9):884–899 [DOI] [PubMed] [Google Scholar]
- 39.Buscot MJ, Magnussen CG, Juonala M, et al. . The combined effect of common genetic risk variants on circulating lipoproteins is evident in childhood: a longitudinal analysis of the cardiovascular risk in young Finns study. PLoS One. 2016;11(1):e0146081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Laitinen TT, Nuotio J, Juonala M, et al. . Success in achieving the targets of the 20-year infancy-onset dietary intervention: association with insulin sensitivity and serum lipids. Diabetes Care. 2018;41(10):2236–2244 [DOI] [PubMed] [Google Scholar]
- 41.Niinikoski H, Lagström H, Jokinen E, et al. . Impact of repeated dietary counseling between infancy and 14 years of age on dietary intakes and serum lipids and lipoproteins: the STRIP study. Circulation. 2007;116(9):1032–1040 [DOI] [PubMed] [Google Scholar]
- 42.Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M. Prediction of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis. Circulation. 2007;115(4):459–467 [DOI] [PubMed] [Google Scholar]