9. Lipids and Lipoproteins

INTRODUCTION

This section of the Guidelines provides recommendations to pediatric care providers on lipid management in their patients. The section begins with background information about the association between dyslipidemia and atherosclerosis and the changing clinical picture of dyslipidemia in childhood. This is followed by the Expert Panel's written synopses of the evidence review relative to lipids in five subsections:

1. Relationship between dyslipidemia and atherosclerosis
2. Lipid and lipoprotein assessment in childhood and adolescence
3. Overview of the dyslipidemias
4. Dietary treatment of dyslipidemias
5. Pharmacologic treatment of dyslipidemias

This evidence review and the development process for the Guidelines are outlined in Section I. Introduction and are described in detail in Appendix A. Methodology. As described, the evidence review here augments a standard systematic review where the findings from the studies reviewed constitute the only basis for recommendations with each study described in detail. This evidence review combines a systematic review with an Expert Panel consensus process that incorporates and grades the quality of all relevant data based on preidentified criteria. Because of the large volume constituted by the included studies and the diverse nature of the evidence, the Expert Panel also provides a critical overview of the studies reviewed for each of the five subsections, highlighting those that in its judgment provide the most important information. Detailed information from each study has been extracted into the evidence tables, which will be available at http://www.nhlbi.nih.gov/guidelines/cvd_ped/index.htm. The conclusions of the Expert Panel's review of the evidence are summarized and graded at the end of each subsection, followed by age-specific recommendations. Where evidence is inadequate, recommendations are a consensus of the Expert Panel. References are listed sequentially at the end of this section, with references from the evidence review identified by unique PubMed identifier (PMID) numbers in bold text. Additional references do not include the PMID number.

BACKGROUND

Since the previous guidelines for lipid management in children and adolescents from the National Cholesterol Education Program (NCEP) were published in 1992,^[1] both the knowledge base surrounding dyslipidemia in childhood and the clinical picture have changed. A series of critical observational studies, which are summarized below, have demonstrated a clear correlation between lipoprotein disorders and the onset and severity of atherosclerosis in children, adolescents, and young adults.^[2],[3],[4]  Over that time period, a major increase in the prevalence of obesity has led to a much larger population of children with dyslipidemia. At the time of the original guidelines, the focus was almost exclusively on identification of children with elevated low-density lipoprotein cholesterol (LDL–C). Since then, the predominant dyslipidemic pattern in childhood is a combined pattern associated with obesity, with moderate to severe elevation in triglycerides (TG), normal to mild elevation in LDL–C, and reduced high-density lipoprotein cholesterol (HDL–C). Both dyslipidemic patterns have been shown to be associated with initiation and progression of atherosclerotic lesions in children and adolescents, as demonstrated by pathology and imaging studies.^{[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12],[13],[14],[15]}  Identification of children with dyslipidemias, which place them at increased risk for accelerated early atherosclerosis, must include a comprehensive assessment of serum lipids and lipoproteins.

OVERVIEW OF THE EVIDENCE FOR A RELATIONSHIP BETWEEN DYSLIPIDEMIA AND ATHEROSCLEROSIS

Postmortem pathology studies of atherosclerosis in children, adolescents, and young adults demonstrate that early atherosclerotic lesions of fatty streaks and fibrous plaques are significantly related to elevations in total cholesterol (TC), LDL–C, and non-HDL–C; lower levels of HDL–C; and the presence and intensity of other risk factors.^{[2],[3],[4],[5],[6],[7]}  The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) study, described in detail in Section II. State of the Science, evaluated the presence of atherosclerosis at postmortem in adolescents and young adults ages 15–34 years who died accidentally.^[3],[4],[5]  The extent of atherosclerosis in the aorta and coronary arteries was correlated with the presence of abnormal lipid levels, obesity, a measure of hypertension, and evidence of cigarette smoking. Using a risk score derived from these results, non-HDL–C was shown to be the major correlate of coronary atherosclerosis in this age group, with a 30 milligram per deciliter (mg/dL) increase in non-HDL–C equivalent to 2 years of vascular aging.^[6],[7]  Based on imaging studies assessing subclinical atherosclerosis, abnormal levels of lipids and lipoproteins are associated with endothelial dysfunction assessed by flow-mediated dilation (FMD) in the brachial artery, coronary artery calcium (CAC), and increased carotid intima-media thickness (cIMT)—all of which are considered precursors of advanced atherosclerosis.^{[8],[9],[10],[11],[12],[13],[14],[15]}  Childhood levels of TC/HDL–C, LDL–C, HDL–C, and TG are each predictors of CAC and cIMT.^{[9],[11],[12]}  Overall risk factor scores, lipids, and increased body mass index (BMI) have been shown to be significant longitudinal predictors of CAC and increased cIMT.^[16]  In subjects from the Cardiovascular Risk in Young Finns (Young Finns) Study, the baseline cardiovascular (CV) risk profile predicted both CAC and cIMT from adolescence through young adulthood. The Young Finns study also found that dyslipidemia in childhood with elevated LDL–C and TG levels predicted increased cIMT independently and synergistically with other CV disease (CVD) risk factors and the metabolic syndrome. Young adults with a TC level >200 mg/dL had five times the risk of developing CVD events 40 years later compared with those who had a TC level <172 mg/dL.^[16]  As part of the metabolic syndrome, childhood dyslipidemia has been shown to predict development of the metabolic syndrome, type 2 diabetes and adult CV disease at 25 year follow-up.^[17],[18]   The effects of risk factors, including dyslipidemia, on coronary lesion severity are multiplicative rather than simply additive.^[2]

In adults with elevated LDL–C but without CVD, convincing evidence suggests that lipid-lowering therapy with statins significantly decreases the incidence of major coronary and cerebrovascular events.^[19]  In children, no randomized clinical trials (RCTs) address whether treating dyslipidemias in children and adolescents will reduce CVD events in later life. However, increasing evidence indicates that lipid-lowering interventions in childhood delay the atherosclerotic process. In one trial, healthy male children treated with a low saturated fat, low cholesterol diet from infancy had enhanced vascular endothelial function at age 11 years compared with controls; this effect was not seen in females.^[20]  In a separate 2-year study, lowering TC and LDL–C levels with a low-fat diet and statin therapy in children and adolescents with heterozygous familial hypercholesterolemia (FH) was associated with a significantly smaller increase in cIMT than that seen in children treated with diet and placebo.^[21]  Followup of participants in this trial who continued on statin therapy for a mean of 4.5 years revealed that younger age at initiation was associated with subsequently smaller cIMT, suggesting that earlier initiation of statin therapy delays progression of the atherosclerotic process in children with FH assessed noninvasively.^[22]  In another RCT, impaired endothelial function in FH children, as judged by FMD, improved significantly in those with LDL–C lowered by simvastatin therapy versus those on placebo to a level similar to that in normal non-FH controls.^[23]

OVERVIEW OF THE EVIDENCE FOR LIPID AND LIPOPROTEIN ASSESSMENT IN CHILDHOOD AND ADOLESCENCE

In the past, NCEP guidelines were based on standard serum measures of TC, very low-density lipoprotein cholesterol (VLDL–C), HDL–C, LDL–C, and TG, with recommendations focused on TC and LDL–C. Since that time, knowledge about lipoprotein heterogeneity and apolipoproteins as predictors of CVD has increased significantly. This evidence review assessed whether measures of any of these in youths are better predictors of subclinical atherosclerosis in adults.

Apolipoproteins B and A–1

In adults, apolipoprotein B (apoB), the major apolipoprotein of LDL–C, and apolipoprotein A–I (apoA–1), the major apolipoprotein of HDL–C, are predictors of the development of CVD and response to treatment to prevent CVD.^[24]  The level of total apoB includes all the apoB-containing lipoproteins, chylomicrons, and VLDL–C and their remnants:  intermediate-density lipoprotein cholesterol (IDL–C), LDL–C, and lipoprotein(a) (Lp(a)). When present in increased amounts, all the apoB-containing lipoproteins are considered atherogenic. Since there is one molecule of apoB on each apoB-containing lipoprotein particle, apoB provides the most accurate assessment of the total number of LDL–C particles. ApoB and apoA–1 are determined using well-standardized immunochemical methods.^[24],[25] These apolipoproteins have been studied in children and adolescents; cut points for apoB and apoA–1—empirically derived from the Third National Health and Nutrition Examination Survey (1988–1994) (NHANES III)—are shown in Table 9–1.^[25]

In the Bogalusa Heart Study, tracking of apoB and apoA–1 over 4 years was compared with tracking for LDL–C and HDL–C. The correlations for apoB and apoA–1 were significant but of somewhat lower magnitude than those for LDL–C and HDL–C.^[26]  Thus, on a population basis, there was no clear advantage of using apoB and apoA–1 over LDL–C and HDL–C to assess tracking. Measurement of apoB and apoA–1 and the ratio of apoB to apoA–1 might provide additional useful information for selective screening, particularly in youths with a family history of premature CVD in parents.^[27],[28]  This may be related to the fact that elevated apoB is often the first expression of familial combined hyperlipidemia (FCHL) in adolescents and young adults, before the onset of overt combined dyslipidemia.^[29]  In the Bogalusa study, no improved prediction of cIMT over that obtained with LDL–C and TC/HDL–C was observed when apoB, apoA–1, or the apoB/apoA–1 ratios were used.^[8]  However, the latest report from the Young Finns study indicates that apoB and apoA–1 levels in childhood were both better predictors of cIMT and brachial endothelial function in adult life than were LDL–C or HDL–C levels.^[14]

Non-HDL–C

Non-HDL–C has emerged as a useful combined measure of the cholesterol content of all the atherogenic apoB-containing lipoproteins. TC and HDL–C can be measured accurately in plasma from nonfasting patients with non-HDL–C calculated by subtracting HDL–C from TC. The coefficient of variability for non-HDL–C thus reflects the variability of measuring both TC and HDL–C. This variability is theoretically less than that for estimated LDL–C, which includes the variability from the measurement of TC, HDL–C, and TG. Percentiles and NCEP-equivalent cut points for non-HDL–C have been determined in children from the Bogalusa study and are shown in Table 9–1.^[30]

In adults, non-HDL–C has been shown to be a better independent predictor of CVD than LDL–C.^[31]  In a longitudinal cohort of subjects (N = 1,163) from the Bogalusa study, studied as both children ages 4–5 years old and adults 27 years later, non-HDL–C (p = 0.52) and LDL–C (p = 0.58) were the best predictors of adult levels.^[32] The odds ratios (ORs) of developing dyslipidemia in adulthood, on the basis of childhood levels of non-HDL–C and LDL–C, were 4.49 and 3.46, respectively, independent of baseline BMI and BMI change over 27 years. At equivalent cut points, childhood high-risk non-HDL–C and LDL–C levels were significantly associated with increased obesity, high LDL–C, and high TG in adulthood. However, only childhood high-risk non-HDL–C status was associated with low HDL–C, hyperinsulinemia, and, marginally, hyperglycemia. Thus, childhood non-HDL–C appears to predict adult dyslipidemia, as well as nonlipid CVD risk factors, better than LDL–C.^[32]

In the pathology studies reported in the PDAY study, non HDL–C and HDL–C levels were the best lipid predictors of pathologic atherosclerotic lesions, both significantly associated with fatty streaks in the thoracic aorta and abdominal aorta and in the right coronary artery and with raised lesions in all three sites;^[33] non-HDL–C and HDL–C levels were more strongly associated with pathologic lesions than either apoB or apoA–1.

In the Bogalusa study, levels of non-HDL–C, LDL–C, TC/HDL–C, apoB, and apoB/apoA–1 in childhood emerged as significant predictors of subclinical atherosclerosis assessed by higher cIMT measurements in adulthood, but ORs were highest for LDL–C and non-HDL–C.^[15]  Overall, childhood non-HDL–C was as good as, or better than, other lipoprotein measures in predicting cIMT in adulthood.

Apolipoprotein E Polymorphism

Apolipoprotein E (apoE) binds to receptors on the surface of liver cells, promoting the hepatic uptake of remnant lipoproteins of both dietary and hepatic origins. Human apoE exists as three major isoforms—E₂, E₃, and E₄—each of which is specified by an independent allele at the locus for the apoE gene. Children with the rarest allele, apoE–2, generally have lower levels of TC and LDL–C, lower BMI and percentage of body fat, and lower insulin but higher HDL–C levels than those with apoE–3 or apoE–4.^{[34],[35],[36]}  Tracking of plasma lipid and lipoprotein is influenced to some degree by the apoE polymorphism. Children with apoE–4 have the highest LDL–C levels, but apoE–4 does not appear to influence the response to a low-cholesterol, low-fat diet^[36] or to the addition of plant stanols.^[37]

Lp(a) Lipoprotein

Lp(a) consists of a molecule of LDL–C in which its apoB moiety is connected through a disulfide bond to apo(a), a glycoprotein homologous to plasminogen. When present in elevated amounts, Lp(a) appears to be atherogenic because of its high cholesterol content and thrombogenic by virtue of the inhibition of the conversion of plasminogen to plasmin at the surface of endothelial cells.^[38]  Lp(a) is most accurately measured by an enzyme-linked immunosorbent assay (or ELISA) that is independent of apo(a) size differences, with the upper limit of normal by this method being 75 nanomoles per liter.

In adults, higher Lp(a) levels may be an independent risk factor for coronary artery disease (CAD), pulmonary vascular disease, ischemic stroke, and aortic aneurysm.^[39]  Elevated Lp(a) levels appear to particularly contribute to risk when combined with high LDL–C levels. In some families, isolated elevated Lp(a) levels have been seen with premature CAD and normal lipid and lipoprotein levels. In the Bogalusa study, Lp(a) was measured in 2,438 children.^[40] Mean Lp(a) levels were 1.7 times higher in Blacks than in Whites. White children with a history of parental myocardial infarction had significantly higher Lp(a) levels than did those with a negative family history, but there was no such association in Black children. Nowak-Gottl studied 1,002 household members of 282 White pediatric patients with a first acute ischemic stroke. Significant heritability estimates (but not environmental estimates) were found for Lp(a).^[41] In children with stroke, Lp(a) levels are significantly elevated in about half of cases with either ischemic or hemorrhagic stroke.^[42]

Advanced Lipoprotein Testing

The plasma levels of VLDL–C, LDL–C, and HDL–C subclasses and their sizes have been determined in children and adolescents by nuclear magnetic resonance spectroscopy^{[43],[44],[45]} and by vertical-spin density-gradient ultracentrifugation^[46] in research studies, but cut points derived from these methods for the diagnosis and treatment of dyslipidemia in youths are not currently available.

OVERVIEW OF THE EVIDENCE FOR NORMAL DISTRIBUTION PATTERNS OF LIPIDS AND LIPOPROTEINS

The Lipid Research Clinics (LRC) Prevalence Study collected lipid and lipoprotein values in children and adolescents from ages 0 to 19 years at multiple centers in the United States and Canada from 1970 to 1976. In that study, the mean TC level was approximately 160 mg/dL, and the mean LDL–C level was 100 mg/dL. The 95th percentiles for these two measures were 200 mg/dL for TC and 130 mg/dL for LDL–C.^[47]  These values were used in developing  recommendations in National Cholesterol Education Program:  Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents, which was published in 1992.^[1]

The NHANES III collected cholesterol levels in more than 7,000 U.S. children ages 0–19 years from 1988 to 1994.^[48]  Over the intervening time period following the LRC study—just over a decade—lipid levels in the pediatric population had increased significantly. The mean TC level was 171 mg/dL, and the 95th percentile was 216 mg/dL; the 95th percentile for LDL–C was 152 mg/dL. This evaluation included significant numbers of African American, Hispanic American, and Mexican American subjects. African American children and adolescents were shown to have significantly higher TC and HDL–C levels and lower TG levels compared with the other racial/ethnic groups of children in the survey.^[48] Although the percentiles of lipid levels varied by race, the risk of atherosclerosis (as measured by cIMT) was equally related to lipid levels and risk factors in African Americans and Whites, so results were not reported separately.^[10]

Lipid levels change with normal growth and maturation. Lipoproteins are very low in cord blood at birth and rise slowly in the first 2 years of life.^[49],[50]  After age 2 years, lipid and lipoprotein levels are relatively stable until adolescence. During puberty, TC and LDL–C levels decrease with increasing age before rising in the late-teen years and again in the third decade of life.^[51]  HDL–C levels decrease during puberty in males but not in females. From the Bogalusa study, there are differences in lipoprotein levels between Blacks and Whites during childhood, with higher levels of TC and HDL–C and lower levels of VLDL–C and TG in Black children and adolescents.^[52]  Recent evaluations have developed age- and gender-specific distribution curves for lipoproteins from the NHANES III data linked to CVD risk.^[53],[54]  The distribution curves reflect the changes noted with normal growth and maturation. It has been suggested that these lipid curves, similar to growth curves, be used to account for normal maturational changes and to allow accurate selection of high-risk thresholds. Alternatively, designating the 50th percentile of the pooled NHANES results as "borderline high" and the 75th percentile as "high," results in thresholds similar to these derived values. The cut points for plasma lipid, lipoprotein, and apolipoprotein levels in children and adolescents are shown in Table 9–1 and for young adults in Table 9–2.

Table 9–1. Acceptable, Borderline-High, and High Plasma Lipid, Lipoprotein and Apolipoprotein Concentrations (mg/dL) For Children and Adolescents^*

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

^* Values for plasma lipid and lipoprotein levels are from the National Cholesterol Education Program (NCEP) Expert Panel on Cholesterol Levels in Children.1 Non-HDL-C values from the Bogalusa Heart Study are equivalent to the NCEP Pediatric Panel cut points for LDL-C.^[30] Values for plasma apoB and apoA-1 are from the National Health and Nutrition Examination Survey III.
⁺ The cut points for high and borderline high represent approximately the 95th and 75th percentiles, respectively.^{[1],[25],[30]} Low cut points for HDL-C and apoA-1 represent approximately the 10th percentile.^[25]

Table 9-2. Recommended Cut Points for Lipid and Lipoprotein Levels (mg/dL) in Young Adults^*

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

^* Values provided are from the Lipid Research Clinics Prevalence Study.^[47] The cut points for TC, LDL-C, and non-HDL-C represent the 95th percentile for subjects ages 20-24 years and are not identical with the cut points used in the most recent National Cholesterol Education Program's Adult Treatment Panel III,^[55] which are derived from combined data on adults of all ages. The age-specific cut points given here are provided for pediatric care providers to use in managing this young adult age group. For TC, LDL-C, and non-HDL-C, borderline high values are between the 75th and 94th percentiles, whereas acceptable values are <75th percentile. The high TG cut point represents approximately the 90th percentile, with borderline high between the 75th and 89th percentiles; acceptable is <75th percentile. The low HDL-C cut point represents roughly the 25th percentile, with borderline low between the 26th and 50th percentiles; acceptable is >50th percentile.

OVERVIEW OF THE EVIDENCE FOR TRACKING OF LIPID AND LIPOPROTEIN LEVELS FROM CHILDHOOD INTO ADULT LIFE

An important factor in considering lipid assessment in childhood is the accuracy of childhood lipid levels in predicting adult results. This evidence review identified 13 prospective screening cohort studies that assessed tracking of elevated lipid and lipoprotien levels from childhood into adulthood, with significant tracking identified in 12 of the 13 studies. From the Bogalusa study, more than 3,000 children ages 5–14 years at baseline were followed for 12 years. Lipid and lipoprotein levels tracked well statistically, with the best correlations for TC and LDL–C levels after age 12 years. Children with TC levels above the 75th percentile had approximately a 50 percent rate of falling into a similar percentile as adults, which was more than twice that predicted by chance alone.^[51],[56]  The best predictors of elevated TC and LDL–C levels in adults were childhood elevations in TC and LDL–C levels.^[56]  In a stepwise multiple logistic regression, incremental increases in TC and BMI independently predicted incremental increases in adult TC.^[53]Similarly, in a later report from the Bogalusa study, obesity, insulin level, and TC level were highly correlated, and increasing levels of obesity predicted elevated lipid levels.^[57]

The 16-year experience of the Beaver County Lipid Study also demonstrated that the overall correlation (r = 0.44) between baseline and followup TC levels was significant; females had a higher correlation than males (0.51).^[58] In an RCT of dietary intervention in children ages 7  months through 5 years, tracking of TC levels was significant for both the diet group and the control group, and the only gender effect was stronger HDL–C tracking for boys.^[34]  In an epidemiologic study of Finnish children followed for more than 12 years, all lipid and lipoprotein levels had significant tracking, with correlations ranging from 0.48 to 0.59 for TC, LDL–C, and HDL–C.^[59]

In the LRC study,^[47] more than 1,700 subjects had their initial TC levels drawn in grades 1–12 and then again 30 years later. Sensitivities for elevated TC and LDL–C levels were 44 percent and 43 percent, respectively, and specificities were 85 percent and 86 percent, respectively. Sensitivity and specificity were not improved by selecting children with a positive family history of early CVD or high cholesterol level.^[60]  Pubertal changes caused sensitivities and specificities to be lowest at ages 14–16 years regardless of lipid status. CVD events were too infrequent to allow testing of the ability of childhood cholesterol levels to predict future CVD in a still relatively young adult cohort.

The Muscatine Study followed more than 14,000 children with two measures of TC levels and other risk factors, the first between ages 8 and 18 years and the followup between 20 and 30 years later. Although TC tracked well from adolescence to adulthood, many adolescents identified as high risk would not be considered high risk in adulthood.^[61]  For children with TC levels above the 75th percentile on two occasions, 75 percent of females and 56 percent of males would not qualify for treatment as adults. For children with TC levels above the 90th percentile on two occasions, 43 percent of females and 70 percent of males had lipid levels above the 75th percentile as adults—that is, the level designated as requiring intervention in adults.^[61]

In summary, the vast majority of epidemiologic studies indicate that there is strong statistical tracking of TC and LDL–C levels from childhood to adulthood. Clinically, this means that approximately half of children with lipid levels above the 75th percentile in childhood will have elevated lipid levels as adults.^[62]  In general, the higher the childhood result and the older the postpubertal age at which the value is obtained, the better the correlation with results in adult life:  A TC level above 200 mg/dL will identify children at risk for more marked hypercholesterolemia with 90 percent confidence.^[63]

OVERVIEW OF THE EVIDENCE FOR DYSLIPIDEMIAS IN CHILDHOOD AND ADOLESCENCE

Dyslipidemias are abnormalities in lipoprotein metabolism associated with any abnormal level of lipoproteins. There are many different types of dyslipidemias, which are influenced by genetics and environmental factors, including nutrition, physical inactivity, smoking, social factors, etc. Dyslipidemia also can be secondary to other specific causes that affect lipoprotein metabolism; these are listed in Table 9–3. The presence of dyslipidemia is an established risk factor for the development of atherosclerosis in both children and adults, but the incidence of CV clinical events due to atherosclerosis is extremely rare in children.

Table 9–3. Causes of Secondary Dyslipidemia

• EXOGENOUS
  • Alcohol
  • Drug therapy:
  • Corticosteroids
    • Isoretinoin
    • Beta-blockers
    • Some oral contraceptives
    • Select chemotherapeutic agents
    • Select antiretroviral agents
• ENDOCRINE/METABOLIC
  • Hypothyroidism/hypopituitarism
  • Diabetes mellitus types 1 and 2
  • Pregnancy
  • Polycystic ovary syndrome
  • Lipodystrophy
  • Acute intermittent porphyria
• RENAL
  • Chronic renal disease
  • Hemolytic uremic syndrome
  • Nephrotic syndrome
• INFECTIOUS
  • Acute viral/bacterial infection^*
  • Human immunodeficiency virus infection (HIV)
  • Hepatitis
• HEPATIC
  • Obstructive liver disease/cholestatic conditions
  • Biliary cirrhosis
  • Alagille syndrome
• INFLAMMATORY DISEASE
  • Systemic lupus erythematosis
  • Juvenile rheumatoid arthritis
• STORAGE DISEASE
  • Glycogen storage disease
  • Gaucher's disease
  • Cystine storage disease
  • Juvenile Tay-Sachs disease
  • Niemann-Pick disease
• OTHER
  • Kawasaki disease
  • Anorexia nervosa
  • Post solid organ transplantation
  • Childhood cancer survivor
  • Progeria
  • Idiopathic hypercalcemia
  • Klinefelter syndrome
  • Werner's syndrome

^*Delay measurement until ≥ 3 weeks postinfection.

The known dyslipidemias are defined by age, gender, and racial cutoffs based on population distributions and known genetic disorders and are outlined in Table 9–4. Genetic lipid disorders include FH, FCHL, familial defective apoB (FDB), familial hypertriglyceridemias, and hypoalphalipoproteinemia. The genetic disorders may be the result of a single-gene defect but more commonly are due to oligogenic defects involving several more genes, which lead to abnormal lipoprotein metabolism.^[16]

Table 9–4. Summary of Major Lipid Disorders in Children and Adolescents 

^*These are the two lipid and lipoprotein disorders seen most frequently in childhood and adolescence; the latter most often manifests with obesity.
HDL-C = high-density lipoprotein cholesterol; IDL-C = intermediate-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol; TC = total cholesterol; TG = triglyceride; VLDL-C = very-low-density lipoprotein cholesterol.

Disorders Affecting LDL Receptors

There are five known genetic disorders causing elevated LDL–C that are expressed in children and that cause early atherosclerosis and premature CVD; they include FH, FDB, autosomal recessive hypercholesterolemia, sitosterolemia, and mutations in proprotein convertase subtilisin-like kexin type 9. These disorders arise from either gene mutations that affect LDL receptor activity or abnormalities in the LDL receptor itself. The presence of these disorders indicates a significantly elevated risk for premature atherosclerosis and CVD events in adulthood.^[16] Of these genetic disorders affecting LDL receptor activity, only FH occurs commonly enough to be a concern for pediatric care providers.

FH is an autosomal dominant disorder that causes isolated LDL–C elevation due to gene mutations in the LDL receptor. Homozygous FH (hoFH) is very rare, with a prevalence of approximately 1:1 million children and is associated with extremely high LDL–C levels (four to eight times higher than normal). Children with hoFH usually develop CVD by the second decade of life.^[16] Heterozygous FH has a prevalence of approximately 1:500 children in the United States. In families with known FH, children with LDL–C levels above 160 mg/dL are likely to have FH.^[16] Untreated FH is associated with premature atherosclerosis and CVD events, with 25 percent of females and 50 percent of males experiencing clinical CVD by age 50 years.

Combined Dyslipidemia

Multiple phenotypes of VLDL–C overproduction and associated TG and LDL–C elevations have been described. These include FCHL, familial dyslipidemic hypertension, hyperapoB, and LDL subclass pattern B. VLDL–C overproduction presents with the lipid pattern of normal to modest elevation of TC and LDL–C, moderate to moderately severe elevation of TG, and reduced HDL–C, with increased numbers of small, dense LDL–C particles.^[16] Roughly 20–30 percent of obese children have evidence of this dyslipidemic pattern.^{[64],[65],[66],[67],[68],[69]}  Since publication of the 1992 NCEP Pediatric Guidelines, the presence of elevated TG-rich remnants, often reflected as elevated total TG or non-HDL–C, has become a recognized risk factor for CVD.^[70]

From the standpoint of lipoprotein metabolism, elevated TG in the fasting state most often reflects increased levels of VLDL–C production from the liver as a consequence of metabolic alterations associated with obesity. As the TG in VLDL–C are hydrolyzed by lipoprotein lipase (LPL) and its cofactor apolipoprotein C–II (apoC–II), a series of VLDL–C remnants of different sizes is produced, ending with IDL–C. IDL–C can be removed directly from plasma by the interaction of apoE with the LDL receptor, or the TG on IDL–C can be hydrolyzed by LPL and hepatic lipase (HL), producing LDL–C. Elevated IDL–C may promote atherosclerosis by its conversion to LDL–C. As well, IDL–C can be small enough to cross the endothelial barrier and enter the vascular wall, where its cholesterol component is atherogenic. In the nonfasting state, patients with elevated VLDL–C often have delayed removal of TG-enriched chylomicrons and chylomicron remnants because both VLDL–C and chylomicrons are competing for LPL. This further accentuates postprandial hypertriglyceridemia.^[71]  Finally, elevated TG levels are usually accompanied by low HDL–C levels, further providing an atherosclerotic milieu, and are commonly associated with nonlipid risk factors, such as obesity, hypertension, insulin resistance, and enhanced thrombogenesis.^[16],[67]

Elevated TG may be due to enhanced production of VLDL–C, decreased hydrolysis, or a combination of both. The most common cause of elevated TG is increased VLDL–C synthesis. This leads to an enhanced transfer of TG from VLDL to both LDL and HDL in exchange for cholesteryl esters (CEs) via the CE transfer protein. As the TG on LDL–C and HDL–C are hydrolyzed, smaller cholesterol-depleted particles are produced. Overproduction of VLDL–C leads to an increased number of atherogenic small, dense LDL–C particles and low HDL–C. Small HDL–C particles are more avidly removed by the kidney, reducing the number of HDL–C particles available for reverse cholesterol transport. Increased VLDL–C production most often results from enhanced hepatic uptake of free fatty acid (FFA) from plasma, leading to overproduction of TG and apoB. The elevated FFA is derived from adipose tissue due to insulin resistance and the decreased inhibition of hormone-sensitive lipase by insulin.

High TG in combination with elevated LDL–C and reduced HDL–C is the dyslipidemia seen as one of the components of the metabolic syndrome.^[72]  Elevated non-HDL–C also will be present in this dyslipidemic phenotype. The presence of this cluster of findings in childhood predicts the development of type 2 diabetes mellitus (T2DM), the metabolic syndrome, and premature clinical CVD in adulthood.^[17],[18]  The pediatric aspects of the metabolic syndrome cluster are addressed separately later in Section XII. Risk Factor Clustering and the Metabolic Syndrome.

No single gene defect has been identified with the combined dyslipidemia disorders, which appear to be oligogenic in origin, with expression exacerbated due to lifestyle factors, especially obesity. In pediatric lipid clinics to which children are referred because of dyslipidemia, combined hyperlipidemia is seen about three times as often as FH and is usually associated with obesity.^[73] In families identified because of an adult proband with clinical CVD and a lipid abnormality (types IIa, IIb, IV), the expression of combined hyperlipidemia often is delayed until the third decade of life. However, combined hyperlipidemia appears to be expressed in adolescents as an elevated apoB level.^[29]  A recent report from the longitudinal Young Finns study revealed that, at 21-year followup, subjects with the combined dyslipidemia pattern beginning in childhood had significantly increased cIMT compared with normolipidemic controls, even after adjustment for other risk factors; cIMT was further increased when the dyslipidemia occurred in the context of the metabolic syndrome.^[9]

Given the association with obesity, combined dyslipidemia is an increasingly common problem.^[65] In a recent study of overweight children, TG levels were significantly elevated in 18 percent of boys and 29 percent of girls, with the degree of elevation directly correlated with the severity of insulin resistance.^[74],[75]  The combined dyslipidemia pattern is now the most common form of dyslipidemia seen in childhood, and in longitudinal studies, it has been shown to persist into adulthood.^[67],[76]  Normal values for TG are <100 mg/dL in children younger than age 10 years and <130 mg/dL at ages 10–18 years (see Table 9–1). Obesity and insulin resistance are usually associated with TG levels between 100 and 400 mg/dL.^[77]  TG values >500 mg/dL usually identify an underlying rare genetic abnormality and are addressed below. Acute conditions associated with severe inflammation and/or endothelial injury and chronic conditions, such as human immunodeficiency virus (HIV) infection and cancer chemotherapy, can be associated with marked TG elevation. Profound hypertriglyceridemia also may occur transiently with ketoacidosis in type 1 diabetes mellitus (T1DM).

Severe Hypertriglyceridemia

Severe elevation in TG to ≥500 mg/dL is rare in childhood and is usually associated with genetically based recessive metabolic defects, including defects in LPL and apoC–II. Severe elevations in TG to >1,000 mg/dL are associated with increased risk for pancreatitis. With LPL and apoC–II deficiency, massive increases in chylomicrons and VLDL–C can occur, producing TG ≥1,000 mg/dL and as high as 5,000–10,000 mg/dL. Such profound increases in TG can produce pancreatitis and eruptive xanthomas but are not associated with premature atherosclerosis because the TG-enriched particles are too large to enter the vascular wall. Finally, TG ≥500 mg/dL can be seen in HL deficiency.^[16] In this condition, HDL–C levels are actually elevated. However, so are the TG-enriched remnants, and premature atherosclerosis can occur in adulthood.

A fasting TG level of ≥ 500 mg/dL often indicates postprandial elevations to >1,000 mg/dL; children with this degree of hypertriglyceridemia present a special clinical problem that requires treatment by a lipid specialist to prevent pancreatitis. These children require a very low-fat diet (<10 percent fat) undertaken with a nutritionist to ensure adequate intake of essential fatty acids. Medium-chain TG, which are absorbed directly into the portal system and do not require chylomicrons for transport to the liver, can have a significant effect on TG, especially in the LPL defect. Neither LPL nor apoC–II deficiency responds to lipid-altering medications. Patients with HL deficiency will respond to lipid-lowering medication; this is addressed below in the subsection about pharmacologic therapy.

Low HDL–C Disorders

HDL–C varies inversely with the risk for CVD, and low HDL–C is an independent predictor of increased risk. In childhood, low HDL–C is usually expressed as part of combined dyslipidemia accompanied by obesity, as described previously. It can also be reduced significantly due to the presence of sedentary lifestyle, cigarette smoke exposure, inherited defects of low HDL–C production, or increased catabolism. Rare genetic forms of low HDL–C include familial hypoalphalipoproteinemia, apoA–1 mutations, Tangier disease, and lecithin cholesterol acyltransferase deficiency. Some, but not all, forms of low HDL–C disorders are associated with premature CVD.^[16]

OVERVIEW OF THE EVIDENCE FOR SCREENING FOR LIPID DISORDERS IN CHILDHOOD AND ADOLESCENCE

Screening for dyslipidemia in childhood is based on the concept that early identification and control of dyslipidemia throughout youth and into adulthood will substantially reduce clinical CVD risk beginning in young adult life. The primary objectives of screening for dyslipidemia are the identification of children and adolescents who are (1) at the highest risk for premature CVD because of extreme lipid abnormalities secondary to inherited or acquired cholesterol disorders and (2) at increased risk because of dyslipidemia that is often associated with other risk factors, such as a family history of CVD or obesity. As described previously, the evidence that children with dyslipidemia are at significant risk for becoming adults with dyslipidemia with an increased risk for early CVD is strong. Accurate identification allows early treatment efforts to focus on children and adolescents at defined risk for accelerated atherosclerosis.

In 2007, the U.S. Preventive Services Task Force (USPSTF) published a major systematic review on screening and treatment for lipid disorders in children.^[62]  The review noted that in the included studies, family history questions were not standardized and had limited diagnostic accuracy. In addition, with the dispersion of families, knowledge of family history for medical problems was often incomplete. The reviewers concluded that the evidence demonstrated that using family history as a primary factor to identify children for screening would miss the majority of children with inherited dyslipidemias, including approximately 50 percent of those with FH. Although overweight has been shown to be the best of the known risk factors for predicting combined dyslipidemia, the review concluded that the use of other risk factors, alone or in combination, had not been evaluated adequately to assess their ability to identify children with dyslipidemia. The review noted that currently recommended screening strategies had low adherence rates by pediatric health care providers and parents of children at risk. In addition, studies did not adequately identify the optimal age and frequency of testing.^[62]

In Section IV. Family History of Early Atherosclerotic Cardiovascular Disease of these Guidelines, a positive family history of early CVD was identified as important information implying increased risk for future CVD in offspring. In the previous NCEP Pediatric Panel guidelines, family history of early CVD was used as the screening tool to define the need for lipid assessment.^[1]  Since that time, a number of studies have evaluated the limitations of this approach. There is no standardized methodology to assess the family history of CVD, and family histories are often inaccurate and/or incomplete. Using the recommendations from the NCEP Pediatric Panel,^[1] the proportion of children who have a family history of premature CVD that will support lipid screening is between 25 and 55 percent. Those studies that used a family history measure to screen for elevated TC levels found that this method for screening misses between 30 and 60 percent of children with high TC levels. From the Bogalusa study, when a positive family history of premature CVD was present, there was a higher risk that the progeny would have abnormal LDL–C levels, but the additional sensitivity gained was minimal.^[78],[79]  Late in adolescence, children with a family history of CVD have been shown to have higher TC, LDL–C, TG, and blood glucose levels and higher body weight.^[80]  However, a negative family history does not rule out dyslipidemia in children. As noted in the USPSTF systematic review, although a positive family history of early coronary heart disease has been shown to predict increased risk for future CVD, inaccurate and incomplete family history reporting make it neither sensitive nor specific enough to use as a predictive screening tool for childhood dyslipidemia.^[62] Overweight in children is associated with significant adverse effects on risk factors, primarily combined dyslipidemia with elevated TG and low HDL–C levels; these abnormalities track into adulthood.^[67],[76]  Although overweight is the risk factor most predictive of dyslipidemia, the magnitude of the effect is variable.^{[64],[65],[66],[67]}

In the past, fasting TC levels have been chosen as the initial screening test by most health care organizations and guidelines. Using the 95th percentile as abnormal, TC levels in the LRC study have 69 percent sensitivity and 98 percent specificity in accurately assessing LDL–C elevations.^[47] Using NHANES data, TC levels had 50 percent sensitivity and 90 percent specificity in detecting elevated LDL–C levels. As described previously, correlations for TC, LDL–C, and HDL–C levels with future measures range from approximately 0.4 to 0.6.^[62]  Approximately half of those with TC levels above the 75th percentile in childhood will have elevated TC levels in adulthood.

The issue of appropriate cutoffs for children screened for lipid disorders was addressed in analyses by both the NCEP Pediatric Panel^[1] and the NHANES III.^[48]  In a more recent analysis,^[53] TC, LDL–C, HDL–C, and TG levels from more than 1,700 participants in three population-based prospective cohort studies were used to compare the ability of single NCEP cut points with multiple NHANES cut points in adolescence to predict abnormal levels in adulthood. NCEP cut points were found to be more predictive of adult high TC, LDL–C, and TG levels than NHANES results but were less predictive of low HDL–C levels. The likelihood of an adult having abnormal lipids was significantly higher in those adolescents with borderline or high lipoprotein levels compared with those with normal levels, and the increase in risk for adult levels was directly correlated and graded according to adolescent levels. Acceptable, borderline, and elevated lipid levels in childhood and young adulthood are shown in Tables 9–1 and 9–2.

Race and gender have both been shown to affect lipid results. Analysis of more than 4,000 children and adolescents from the Bogalusa study revealed that after controlling for overweight, White males had significant adverse changes in TC, LDL–C, VLDL–C, and HDL–C levels on entering adulthood, with less significant changes for White and Black females and Black males.^[52]  By age 26 years, 9 percent of White males, 8 percent of White females, 2 percent of Black males, and 6 percent of Black females had abnormal lipid profiles, with White males having a dramatic worsening of the TC/HDL–C ratio.^[52] Also from the Bogalusa study, there were racial differences in TG and VLDL–C levels between Blacks and Whites, with higher VLDL–C and TG levels in Whites and a modest difference related to higher HDL–C levels in Blacks.^[45],[81]  White female children and Black males had higher HDL–C levels than White males, although the absolute differences are modest.^[45],[81]  Differing distributions of individual risk factors in different groups is not in itself a reason for different standards for evaluation and/or management. Race and/or ethnic group-specific recommendations would be indicated only if there were evidence of a different relationship between risk factor level and future risk of CVD. At this time, there is insufficient evidence linking lipid levels to atherosclerosis by race or ethnic group, so similar cut points are recommended for determining risk status.

The number of children with dyslipidemia continues to increase along with population increases in overweight and decreases in insulin sensitivity. Cardiovascular risk factors cluster in children and are strongly correlated with body fatness.^[82],[83]  Childhood overweight is clearly correlated with abnormal lipid levels.^{[69],[77],[83],[84],[85]} Other conditions—such as diabetes, nephrotic syndrome, chronic renal disease, inflammatory disease, hypothyroidism, and other secondary causes of dyslipidemia, known to be associated with accelerated atherosclerosis—should indicate a higher frequency of testing (see Table 9–3 and Section XI. Diabetes Mellitus and Other Conditions Predisposing to the Development of Accelerated Atherosclerosis).^[86]  Children with these conditions need to be evaluated for dyslipidemia when the diagnosis of the primary condition is made.

As described previously, non-HDL–C is now increasingly used in evaluating adults for dyslipidemia. In analyses of two large pediatric cohorts from the Bogalusa study, non-HDL–C was shown to be both sensitive and specific for identifying those who will have elevated LDL–C levels and other dyslipidemias as adults. Children in the top quartile of non-HDL–C were approximately four times more likely to have dyslipidemia as adults.^[15],[32] A non-HDL–C level above the 95th percentile was 86–96 percent sensitive and 96–98 percent specific for detecting an elevated LDL–C level in both African American and Hispanic children.^[62]  In a separate study, the top quartile of non-HDL–C levels correlated with the top decile of cIMT, as well as did any other lipoprotein measure.^[15] Non-HDL–C levels appear to be a sensitive test for screening, with the additional advantage of being readily available in the nonfasting state.^[32]  As with TC and LDL–C, levels at which risk are identified could be defined by the 75th and 95th percentiles, as shown in Tables 9–1 and 9–2. A recent observational study found that non-HDL–C was as powerful as any other lipoprotein measure for predicting the presence of atherosclerosis in children and adolescents.^[15]  For both children and adults, non-HDL–C levels appear to be more predictive of persistent dyslipidemia, and therefore atherosclerosis and future events, than TC, LDL–C, or HDL–C levels alone.^[32]

Risks/Harms Associated With Lipid Screening

No studies have identified any consistent harm from screening for cholesterol in children and adolescents. A concern is whether screening abnormalities may cause labeling of children, although the evidence is not sufficient to demonstrate any adverse effects. Although one small nonrandomized study showed some possible behavior changes in children identified with dyslipidemia, this has not been substantiated in any of the many other screening studies, observational trials, or clinical trials.^[62]

There is a significant rate of lack of compliance with screening and followup recommendations by both clinicians and parents of children with abnormal levels. A number of factors have been suggested, including inconvenience, discomfort with the screening tests, refusal by the child or parent, concerns about upsetting the child, resistance regarding dietary and lifestyle changes, and other unidentified factors.

CONCLUSIONS AND GRADING OF THE EVIDENCE REVIEW FOR LIPID ASSESSMENT IN CHILDHOOD AND ADOLESCENCE

• Combined evidence from autopsy studies, vascular studies, and cohort studies strongly indicates that abnormal lipid levels in childhood are associated with increased evidence of atherosclerosis (Grade B).
• The evidence review supports the concept that early identification and control of dyslipidemia throughout youth and into adulthood will substantially reduce clinical CVD risk beginning in young adult life. Preliminary evidence in children with heterozygous FH with markedly elevated LDL–C indicates that earlier treatment is associated with reduced subclinical evidence of atherosclerosis (Grade B).
• Multiple prospective screening cohort studies have demonstrated the normal lipid and lipoprotein distributions in childhood, adolescence, and young adult life (Tables 9–1 and 9–2) (Grade B). Cohort studies have also demonstrated significant tracking of elevated lipid levels from childhood to adulthood, with lipid and lipoprotein results in childhood predictive of future adult lipoprotein profiles; the strongest statistical correlation occurs between results in late childhood and the third and fourth decades of life (Grade B).
• TC and LDL–C levels fall as much as10–20 percent or more during puberty (Grade B). Based on this normal pattern of change in lipid and lipoprotein levels with growth and maturation, age 10 years (age range 9–11 years) is a stable time for lipid assessment in children (Grade D). For most children, this age range will precede onset of puberty.
• Significant evidence exists that using family history of premature CVD or of cholesterol disorders as the primary factor in determining lipid screening for children misses approximately 30–60 percent of children with dyslipidemias, and accurate and reliable measures of family history are not available (Grade B). In the absence of a clinical or historic marker, identification of children with lipid disorders that predispose them to accelerated atherosclerosis requires universal lipid assessment (Grade D).
• Non-HDL–C has been identified as a significant predictor of the presence of atherosclerosis, as powerful as any other lipoprotein cholesterol measure in children and adolescents. For both children and adults, non-HDL–C appears to be more predictive of persistent dyslipidemia, and therefore atherosclerosis and future events, than TC, LDL–C, or HDL–C alone. A major advantage of non-HDL–C is that it can be accurately calculated in a nonfasting state and is therefore very practical to obtain in clinical practice. The evidence supports use of non-HDL–C as a screening tool for identification of a dyslipidemic state in childhood (Grade B).
• In terms of other lipid measurements:  (1) at this time, most but not all studies indicate that measurement of apoB and apoA–1 for universal screening provides no additional advantage over measuring non-HDL–C, LDL–C, and HDL–C; (2) measurement of Lp(a) is useful in the assessment of children with both hemorrhagic and ischemic stroke; (3) in offspring of a parent with premature CVD and no other identifiable risk factors, elevations of apoB, apoA–1, and Lp(a) have been noted; and (4) measurement of lipoprotein subclasses and their sizes by advanced lipoprotein testing has not been shown to have sufficient clinical utility in children at this time (Grade B).                               
• Obesity is commonly associated with a combined dyslipidemia pattern, with mild elevations in TC and LDL–C, moderate to severe elevation in TG, and low HDL–C. This is the most common dyslipidemic pattern seen in childhood, and lipid assessment of overweight and obese children identifies an important proportion with significant lipid abnormalities (Grade B).
• Dyslipidemias can be acquired genetically but also secondary to specific conditions, such as diabetes, nephrotic syndrome, chronic renal disease, postorthotopic heart transplant, history of Kawasaki disease with persistent coronary involvement, chronic inflammatory disease, hypothyroidism, and other causes, as outlined in Table 9–3. There is impressive evidence for accelerated atherosclerosis both clinically and as assessed with noninvasive methods in some of these conditions, which accordingly have been designated as special risk diagnoses for accelerated atherosclerosis (Table 9–7); management of these is described in Section XI. Diabetes Mellitus and Other Conditions Predisposing to the Development of Accelerated Atherosclerosis. Lipid evaluation of these patients contributes to risk assessment and identifies an important proportion with dyslipidemia (Grade B).
• The complete phenotypic expression of some inherited disorders, such as FCHL, may be delayed until adulthood. Evaluation in children and adolescents from high-risk families with FCHL that begins in childhood and continues through adulthood (per NCEP adult treatment guidelines) will lead to early detection of those with abnormalities (Grade B).

Age-specific recommendations for lipid assessment are outlined in Table 9–5. Specific management for children with identified dyslipidemia is outlined in the algorithms in Figures 9–1 and 9–2. Definitions of the risk factors and special risk conditions for use with the recommendations and in the algorithms appear in Tables 9–6 and 9–7. The advantages of identifying dyslipidemia and initiating treatment in childhood are the potential for increased reversibility or slowing of the disease process, the knowledge that lifestyle change and attention to risk are more readily accomplished than with individuals in their twenties and thirties, and the fact that regular contact with the health care system is routine in this age group. Late adolescence is often the last time for many years that young adults will routinely undergo health assessment, at the precollege or preemployment physical. It therefore represents an opportunity to diagnose lipid disorders and to advise the young adult about his or her CV risk profile and a healthy lifestyle pattern. When medication is recommended, the decision occurs in the context of the complete CV risk profile of the patient and the sociocultural milieu of the family.

The first step proposed for management of children with identified lipid abnormalities is a focused intervention to improve diet and physical activity. Conclusions of the evidence review and recommendations for dietary management of dyslipidemias are provided in the next subsection.

Table 9–5. Evidence-Based Recommendations for Lipid Assessment

Grades reflect the findings of the evidence review.
Recommendation levels reflect the consensus opinion of the Expert Panel.

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL–C), high-density lipoprotein cholesterol (HDL–C), and non-HDL–C by 38.6; for triglycerides (TG), divide by 88.6.

^a Interval between FLP measurements: after 2 weeks but within 3 months.
^b Use Table 9-1 for interpretation of results; use lipid algorithms in Figures 9-1 and 9-2 for management of results.
^c Disregard TG and LDL-C in nonfasting sample.
^d Use Table 9-1 for interpretation of results; use lipid algorithms in Figures 9-1 and 9-2 for management of results.
^e Lipid screening is not recommended for those ages 12–16 years because of significantly decreased sensitivity and specificity for predicting adult LDL–C levels and significantly increased false-negative results in this age group. Selective screening is recommended for those with the clinical indications outlined.
^f Interval between FLP measurements: after 2 weeks but within 3 months.
^g Use Table 9-1 for interpretation of results of 7- to 19-year-olds and lipid algorithms in Figures 9-1 and 9-2 for management. Use Table 9-2 for interpretation of results of 20- to 21-year-olds and ATP III algorithms for management
^h Disregard TG and LDL-C in nonfasting sample.
ⁱ Interval between FLP measurements: after 2 weeks but within 3 months

Table 9–6. Risk Factor (RF) Definitions for Dyslipidemia Algorithms

(+) Family history: myocardial infarction, angina, coronary artery bypass graft/stent/angioplasty, sudden cardiac death in parent, grandparent, aunt, or uncle, male <55 years, female <65 years

High-Level RFs:

• Hypertension requiring drug therapy (BP ≥ 99th percentile (%ile) + 5 mmHg)
• Current cigarette smoker
• BMI ≥ 97th %ile
• Presence of high-risk conditions (Table 9-7)

(Diabetes mellitus [DM] is also a high-level risk factor but it is classified here as a high-risk condition to correspond with Adult Treatment Panel III recommendations for adults that DM is considered a CVD equivalent.)

Moderate-Level RFs:

• Hypertension not requiring drug therapy
• BMI ≥ 95th %ile, < 97th %ile "
• HDL-C < 40 mg/dL
• Presence of moderate-risk conditions (Table 9-7)

Table 9–7. Special Risk Conditions

High Risk:

• Diabetes mellitus, type 1 and type 2
• Chronic kidney disease/end-stage renal disease/post renal transplant
• Postorthotopic heart transplant
• Kawasaki disease with current aneurysms

Moderate Risk:

• Kawasaki disease with regressed coronary aneurysms
• Chronic inflammatory disease (systemic lupus erythematosus, juvenile rheumatoid arthritis)
• Human immunodeficiency virus infection
• Nephrotic syndrome

OVERVIEW OF THE EVIDENCE FOR DIETARY TREATMENT OF DYSLIPIDEMIA

BACKGROUND

In the first NCEP guidelines addressing lipids in children published in 1992, the NCEP Pediatric Panel recommended a prudent diet (the NCEP Step I diet), with no more than 30 percent of calories from fat, less than 10 percent of calories from saturated fat, and cholesterol intake less than 300 milligrams per day (mg/d) for all healthy U.S. children older than age 2 years.^[1]  Children with dyslipidemias, primarily those with elevated LDL–C levels, were to be treated first with the Step I diet; then, if after 3 months they failed to achieve therapeutic goals, with a more stringent diet (NCEP Step II diet). The NCEP Step II diet recommended no more than 30 percent of calories from fat, less than 7 percent of calories from saturated fat, and less than 200 mg/d of dietary cholesterol. Calories were to be sufficient to maintain normal growth and development. These recommendations were based primarily on epidemiologic and clinical studies. At that time, few RCTs addressed the effects of diet modification in children, particularly during infancy and adolescence, the periods of most rapid growth and development. The increasing prevalence of obesity in childhood has led to a large population of children with combined dyslipidemia who also need dietary management. The evidence review for these Guidelines identified a large number of observational studies and RCTs that, when combined, provide a substantial body of information on which to base new recommendations.

EVIDENCE FOR DIETARY TREATMENT OF HYPERCHOLESTEROLEMIA BY AGE GROUP

Infant Feeding

A meta-analysis of 37 observational cohort and cross-sectional studies compared the effect of breast-feeding versus formula-feeding on TC levels in adolescents and adults.^[87] Although the mean TC level was higher in breast-fed versus formula-fed infants, this difference did not persist into childhood or adolescence. In adults, the TC levels of those who were breast fed as infants were lower than in those who were formula fed. Short-term feeding studies, all RCTs with small sample sizes, varied the fat and cholesterol contents of infant formula, with subsequent changes in levels of TC, LDL–C, TG, and HDL–C in infancy; there were no differences in lipoprotein profiles postweaning.^{[87],[88],[89],[90]}

Infancy Feeding Beyond Weaning

Many of the data on the safety and efficacy of a diet low in saturated fat and cholesterol starting in infancy come from the Special Turku Coronary Risk Factor Intervention Project (STRIP), in which 7-month-old Finnish infants (N = 1,062) were randomized into either a group receiving intensive counseling from a nutritionist for a diet with total fat at 30–35 percent of calories, a 1:1:1 intake ratio of saturated fatty acid/monounsaturated fatty acid/polyunsaturated fatty acid (PUFA), cholesterol <200 mg/d, protein (10–15 percent), and carbohydrate 50–60 percent or into a group receiving basic health education and no instructions on the use of fats.^{[91],[92],[93],[94]} Breastfeeding or formula feeding was advised until age 12 months; after age 12 months, the recommended beverage was fat-free milk supplemented with vegetable fat to maintain total fat intake at the recommended level until age 2 years. The children were followed with serial evaluations, including dietary assessment using 4-day food records, to early adolescence. At baseline, there was no difference in total fat or saturated fat consumption between the groups. At the first postrandomization lipid evaluation at age 13 months, the diets of intervention subjects contained a mean of 26 percent of calories from fat, with 9 percent from saturated fat compared with 28 percent and 13 percent, respectively, in the diets of control subjects, a significant difference between groups. This change was associated with significantly lower TC and LDL–C levels in the intervention group, with no differences in measures of growth and development.^[91],[92]

A short-term study varied the fat content of cow's milk in toddlers between ages 12 months and 2 years. Increasing the vegetable fat content increased plasma linoleic acid and alpha linoleic acid concentrations with no change in long-chain PUFA, arachidonic acid, or docosahexaenoic acid (DHA).^[95]

Infancy to Ages 5, 7, and 11 Years

When assessed at ages 3 and 5 years, the STRIP intervention group consistently had lower intakes of total fat and cholesterol, higher ratios of polyunsaturated to saturated fat and unsaturated to saturated fat, and higher intakes of protein and carbohydrate than the control group.^[92] These dietary differences were associated with significantly lower levels of TC, LDL–C, HDL–C, apoB, and apoA–1 in the intervention group. There were no differences between the groups in mean energy intake, relative weight, relative height, or neurologic development.^[93]  The dietary differences between the intervention and control groups were maintained at age 7 years.^[94]  However, only the boys had significantly lower levels of TC, LDL–C, apoB, and TG. At age 11 years, with a 55 percent followup rate, intervention boys and girls again had significantly lower intake of saturated fat and higher polyunsaturated fat to saturated fat ratios than controls.²⁰  There were no differences in weight, BMI, or physical activity. In intervention males, TC levels were 4.6 percent lower, and LDL–C levels were 9.6 percent lower than control males, but again, there were no significant lipid differences between groups for females. Of note, intervention boys had significantly greater endothelial function, as judged by FMD, than control boys, even after adjusting for differences in LDL–C levels.^[20] 

A clinically initiated, home-based, parent-child autotutorial (PCAT) dietary education program directed at increasing dietary knowledge and reducing fat consumption and LDL–C levels was assessed in an RCT of 4- to 10-year-old boys and girls with borderline high or high LDL–C levels.^[96] Intervention families received either individualized diet counseling or use of tape-recorded nutrition messages aimed at achieving a total dietary fat of less than 30 percent of calories, saturated fat less than 10 percent of calories, and cholesterol less than 300 mg/d; control subjects received usual care. At baseline, cholesterol intake averaged 156 mg/d in the PCAT tutorial group, 163 mg/d in the dietary counseling group, and 176 mg/d in the control group. After 3 months, those in the PCAT and the dietary counseling groups, compared with those in the high cholesterol control group, had significantly lower intakes of total fat as percentage of calories (-1.5 percent in PCAT; -1.6 percent in diet counseling; +0.2 percent in high cholesterol controls) and saturated fat (-0.8 percent in PCAT; -1.0 percent in dietary counseling; no change in high-cholesterol control group). Cholesterol intake averaged 133 mg/d in the intervention group and138 mg/d in the dietary counseling group and was essentially unchanged at 183 mg/d in the usual care control group. Mean LDL–C levels decreased significantly more in the PCAT intervention group by 10 mg/dL, compared with 4.1 mg/dL in the dietary counseling group and 3.4 mg/dL in the control group. These results were maintained at 1-year followup.^[97]  Another pediatric office-based nutritional education program also effectively decreased total fat, saturated fat, and cholesterol intakes, with significant decreases in TC and LDL–C levels after 16 weeks.^[98]

In prepubertal children with FH, a restricted diet with 23 percent ± 5 percent of energy from total fat, 8 percent ± 2 percent from saturated fat, 5 percent ± 1 percent from polyunsaturated fat, 8 percent ± 2 percent from monounsaturated fat, 15 percent ± 2 percent from protein, 62 percent ± 5 percent from carbohydrate, and cholesterol 67 ± 28 mg/1,000 kcal for 1 year lowered TC and LDL–C levels by 4 percent and 5.5 percent, respectively. HDL–C, TG, apoB, ferritin, weight for height, and height velocity were unchanged.^[99]

The Child and Adolescent Trial for Cardiovascular Health was a group randomized school trial designed to examine the outcomes of multilevel and multicomponent health behavior intervention in 56 intervention and 40 control public schools in California, Louisiana, Minnesota, and Texas. The trial followed 5,106 initially third-grade students from ethnically diverse backgrounds.^[100]  In half of the intervention schools, there were school food service modifications to lower fat and sodium contents plus enhanced physical education and classroom health curricula; the other half received the same intervention plus family education. Compared with control schools, intervention schools had a significant decrease in total fat, from 38.9 percent to 31.9 percent of energy in cafeteria lunches, and an increase in the amount of vigorous physical activity. However, after this 2.5 year intervention, there was no difference between the intervention and control groups in TC levels, the primary outcome. There was no evidence of any deleterious effect on growth or development.

Adolescents

The STRIP trial has results to age 14 years, at which time intervention group children still consumed less total and saturated fats and more carbohydrates and polyunsaturated fat and had lower TC and LDL–C levels than children in the control group; the difference between groups was only significant in males. These results were present at the first evaluation at age 13 months and sustained throughout the study period.^[101]  There were no harmful effects identified on growth, micronutrient intake, development, or neurologic function.^{[93],[94],[101],[102]}

The Dietary Intervention Study in Children^[103] assessed the efficacy and safety of an intervention to lower dietary intakes of total fat, saturated fat, and cholesterol in order to lower elevated LDL–C levels (between the 80th and 98th percentiles), starting in prepubertal boys (N = 362) and girls (N = 301) ages 8–10 years and continuing for 3 years. The children were randomized to an intervention group or a usual care control group in a six-center clinical trial. A behavioral-based, nutritionist-tailored intervention was used to promote a diet similar to the NCEP Step II diet—with 28 percent of calories from fat, less than 8 percent from saturated fat, less than 9 percent from polyunsaturated fat, and cholesterol less than 75 mg/1,000 kcal/d, not to exceed 150 mg/d; the control group received dietary literature. At 3-year followup, dietary total fat, saturated fat, and cholesterol were decreased significantly in the intervention group compared with the usual care control group; this was accompanied by small but significant differences in LDL–C levels (reduction of 15.4 mg/dL in the intervention group versus reduction of 11.9 mg/dL in the control group). Greater sexual maturation and BMI were found to increase the normal decrease in LDL–C level during adolescence.^[104] There were no differences between the two groups for multiple safety measures. After 3 years, the intervention was modified to include a more appropriate approach for adolescents; the significant differences between the two groups in total fat, saturated fat, and cholesterol intakes were maintained, with no differences in any of the safety measures, but LDL–C levels did not differ between the intervention and control groups at 7-year followup.^{[104],[105],[106]}

Young Adulthood, Ages 18–21 Years

Little information is available for this age group. A small study in university students with moderate LDL–C elevations showed that a multisession educational intervention significantly improved knowledge and attitudes about dietary changes compared with controls, but this was not associated with any significant decreases in TC and LDL–C levels.^[107]

OVERVIEW OF THE EVIDENCE FOR LOWERING TC AND LDL–C LEVELS WITH DIETARY SUPPLEMENTS

Plant Stanol Esters and Plant Sterol Esters

Background

Under normal conditions, few if any plant stanol esters and only very small amounts of plant sterol esters are absorbed by the human intestine. Both of these compounds inhibit the absorption of cholesterol, either by displacing cholesterol from its mixed micelle or by competing with cholesterol for high-affinity binding sites on the surfaces of intestinal cells, leading to decreased cholesterol in chylomicrons and less cholesterol delivered to the liver, a decrease in the hepatic pool of cholesterol, an induction of LDL–C receptors, and an ultimate decrease of the LDL–C level.^[16] This is associated with increased biosynthesis of hepatic cholesterol, which limits the efficacy of these compounds.

Use of Plant Stanols in the General Population

The effect of replacing dietary fat with plant stanol ester was investigated in a subset of 81, 6-year-old children from the STRIP trial. TC and LDL–C levels decreased by 5.4 percent and 7.5 percent, respectively, in those who consumed a plant stanol-enriched margarine as replacement for 20 grams per day (g/d) of dietary fat intake compared with control margarine. There was no effect on HDL–C or TG levels. These changes were accompanied by decreased cholesterol absorption. Safety was judged to be excellent.^[108]  Increasing dietary plant sterols did not alter cholesterol precursor sterol concentrations in these children.^[109]  Presence of the apoE–4 variant did not affect the response to plant stanols in the same group of 6-year-olds.^[37]  There was no significant difference in decreased cholesterol absorption between boys (36.3 percent) and girls (42.0 percent), but the compensatory increase in hepatic cholesterol synthesis was significantly higher in girls (19.5 percent) than in boys (7.6 percent). This might explain the greater decrease in LDL–C levels in boys (-9.1 percent) than in girls (-5.8 percent). These plant stanol results were confirmed in a small group of healthy 2- to 5-year-old U.S. preschool children.^[110]

Plant Sterol Esters and Plant Stanol Esters in Children With Familial Hypercholesterolemia

Five RCTs of plant sterols and plant stanols have been performed, primarily in FH prepubertal children ages 2–12 years. In each of these studies, both stanol and sterol esters lowered TC and LDL–C levels significantly, with decreased absorption of cholesterol accompanied by increased cholesterol biosynthesis.^{[111],[112],[113],[114]}

Two separate studies in children with FH assessed the effect of plant sterols on endothelial function and found that despite significant decreases in TC and LDL–C levels, there was no improvement in endothelial function, as judged by FMD.^[113],[115]  This negative result was ascribed to a relatively small decrease in LDL–C levels compared with that achieved with statins, which have been shown to improve FMD in FH children.

Other Dietary Supplements

Dietary supplementation with garlic,^[116] soy protein,^[117] and DHA^[118] did not lower LDL–C levels in hypercholesterolemic children. However, DHA supplementation restored impaired FMD. In some but not all studies, psyllium significantly lowered LDL–C levels from 5 to 10 percent, but there was no identification of dietary sources of fiber, so results are difficult to interpret.^{[119],[120],[121]}  In a single study, adding vitamins C and E to the low-fat diets of FH or FCHL children was associated with an improvement in FMD, independent of changes in blood lipid and lipoprotein levels.^[122]

 OVERVIEW OF THE EVIDENCE FOR DIETARY MANAGEMENT OF HYPERTRIGLYCERIDEMIA

Elevated TG levels are very responsive to weight loss, diet composition, and exercise. Most importantly, in overweight and obese children and adolescents with elevated TG levels, even small amounts of weight loss are associated with significant decreases in TG levels and increases in HDL–C levels.^{[123],[124],[125],[126]}  Exercise training alone, when associated with a decrease in body fat, has also been shown to be associated with a significant decrease in TG levels, with reversion to baseline when children became less active.^[127]

Regarding dietary composition, substitution of soy milk for low-fat cow's milk induced significant reductions in TG and VLDL–C levels and increased HDL–C levels in a small series of children.^[128]  In adults with hypertriglyceridemia, a low-carbohydrate, high-fat diet (40 percent carbohydrate, 39 percent total fat, 8 percent saturated fat, 15 percent monounsaturated fat) significantly decreased TG by a mean of 63 percent, with associated mean increases in LDL–C of 22 percent and HDL–C of 8 percent. A subsequent high-carbohydrate, low-fat diet (54 percent carbohydrate, 28 percent total fat, 7 percent saturated fat, 10 percent monounsaturated fat) significantly increased TG back to baseline levels.^[129]  In children, a 12-month followup study of 21-month-old children with elevated TG levels treated with a carbohydrate-restricted diet showed a decrease in sugar and carbohydrate intakes associated with a decrease in TG from a mean of 274.1 +/- 13.1 mg/dL before treatment to 88.8 +/- 13.3 mg/dL after 12 months.^[130]  In an analysis of adolescents from NHANES, the U.S. Department of Agriculture's Center for Nutrition Policy and Promotion's Healthy Eating Index (HEI) was used to provide an overall picture of dietary quality relative to the metabolic syndrome constellation of central obesity, elevated TG, elevated BP, reduced HDL–C level, and impaired fasting glucose level.^[131]  There was a significant inverse association between the overall HEI score plus the fruit intake score and the prevalence of the metabolic syndrome components. There was also a trend toward lower prevalence of the metabolic syndrome components, including elevated TG in adolescents with high activity levels, although this was not significant. The concept of glycemic load has also been evaluated in the setting of obesity and dyslipidemia in adolescents and adults. The glycemic index is a measure of the blood glucose response to a 50 g portion of a selected carbohydrate; the glycemic load is the mathematic product of the glycemic index and the carbohydrate amount.^[132]  In adolescents and young adults, there is evidence that low glycemic-load diets are at least as effective as low-fat diets in achieving weight loss, with decreased TG and increased HDL in subjects on the low glycemic-load diet.^{[133],[134],[135]}  In adolescents, a low-carbohydrate diet associated with weight loss has been shown to significantly reduce TG levels.^[136]

CONCLUSIONS AND GRADING OF THE EVIDENCE REVIEW FOR DIETARY MANAGEMENT OF DYSLIPIDEMIA

• A diet with total fat at 25–30 percent of calories, saturated fat less than 10 percent of calories, and cholesterol intake less than 300 mg/d, as recommended by the original NCEP Pediatric Panel, has been shown to safely and effectively reduce the levels of TC and LDL–C in healthy children. (Grade A). There is some evidence this is also the case when the diet begins in infancy and is sustained throughout childhood into adolescence (Grade B). The Cardiovascular Health Integrated Lifestyle Diet (CHILD 1), described in Section V. Nutrition and Diet of these Guidelines, has this composition.
• In children with identified hypercholesterolemia and elevated LDL–C levels, a more stringent diet, with saturated fat <7 percent of calories and dietary cholesterol limited to 200 mg/d, has been shown to be safe and modestly effective in lowering LDL–C levels (CHILD 2-LDL, Table 9-8) (Grade A).
• Use of dietary adjuncts, such as plant sterol or stanol esters, up to 2 g/d can safely enhance the LDL–C-lowering effects short term in children with FH (Grade A). However, long-term studies on the safety and effectiveness of plant sterol and stanol esters have not been completed. Their clinical use is therefore usually reserved for children with primary elevations of LDL–C who do not achieve LDL–C goals with dietary treatment alone. Such an approach may lower LDL–C sufficiently to avoid the necessity of drug treatment. Food products containing plant stanol esters, such as some margarines, are marketed directly to the general public. In two short-term trials, they have been shown to be safe with minimal LDL-lowering effects in healthy children (Grade B).
• Evidence for use of other dietary supplements is insufficient for any recommendation (no grade).
• In children with elevated TG, weight loss and reduction of simple carbohydrate intake are associated with decreased TG levels. (CHILD 2-TG, Table 9-8) (Grade B). When elevated TG are associated with obesity, decreased calorie intake and increased activity levels are of paramount importance.
• A behavioral approach that engages the child and family delivered by a registered dietitian has been shown to be the most consistently effective approach for achieving dietary change (Grade B).

The approach to management of dyslipidemias is staged, as in the original NCEP Pediatric Panel recommendations. For all children with identified dyslipidemia in whom the response to a low-fat/low saturated fat/low cholesterol diet has not been evaluated, the CHILD 1 diet described in Section V. Nutrition and Diet is recommended as the first step, with implementation guided by a registered dietitian. For obese children with identified dyslipidemia, age- and BMI-specific additional recommendations addressing calorie restriction and increased activity appear in Section X. Overweight and Obesity. If, after a 3-month trial of CHILD 1/lifestyle management, fasting lipid profile findings exceed the therapeutic goals in Tables 9-1 and 9-2, lipid parameter-specific diet changes outlined in Table 9-8 are recommended. Dyslipidemia management is also outlined in the algorithms in Figures 9-1 and 9-2.

Table 9–8. Evidence-Based Recommendations for Dietary Management of Elevated LDL–C, non-HDL-C and TG

Grades reflect the findings of the evidence review.
Recommendation levels reflect the consensus opinion of the Expert Panel.
Supportive actions represent expert consensus suggestions from the Expert Panel provided to support implementation of the recommendations; they are not graded.

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

ELEVATED LDL–C: CHILD 2 - LDL

^* Can be found added to some foods, such as some Margarins

ELEVATED TG OR NON-HDL–C: CHILD 2 - TG

^* If child is obese, nutrition therapy should include calorie restriction, and increased activity (beyond that recommended for all children) should be prescribed. See Section X. Overweight and Obesity for additional age-specific recommendations.
^** The Food and Drug Administration (FDA) and the Environmental Protection Agency are advising women of childbearing age who may become pregnant, pregnant women, nursing mothers, and young children to avoid some types of fish and shellfish and eat fish and shellfish that are low in mercury. For more information, call the FDA's food information line toll free at 1–888–SAFEFOOD or visit http://www.cfsan.fda.gov/~dms/admehg3.html.

MEDICATION THERAPY FOR HYPERLIPIDEMIAS

Background

1992 NCEP Recommendations/American Heart Association Scientific Statement/American Academy of Pediatrics' Committee on Nutrition Recommendations

In addition to making recommendations for screening and lifestyle management of LDL abnormalities, the 1992 NCEP Pediatric Panel report made recommendations regarding medication therapy.^[1]  It was recommended that treatment with medications be considered only in children ages 10 years and older after an adequate 6- to 12-month trial of lifestyle/dietary modification. The criteria for initiation included LDL–C levels ≥190 mg/dL or LDL ≥160 mg/dL, together with either a positive family history of premature CVD or the presence of two or more other CVD risk factors. The only medication recommended was a bile acid sequestrant. At that time, no data were available regarding the safety and efficacy of statin use in children.

Since the NCEP Pediatric Panel guidelines were published in 1992, new evidence, as described in the opening section, provides the impetus for their reevaluation and modification. In 2007, the American Heart Association (AHA) published a scientific statement with updated treatment recommendations for children and adolescents.^[137]  The AHA guidelines recommended that the presence of overweight and obesity was an additional indication for screening with a fasting lipid profile (FLP) and that the presence of overweight and lipid abnormalities indicated screening for other abnormalities associated with insulin resistance/metabolic syndrome. For patients meeting the criteria for pharmacologic therapy, the AHA recommended a statin as a first-line agent. The presence of other risk factors and high-risk conditions should be considered in making decisions regarding medication therapy, including LDL–C levels and age for beginning treatment and target LDL–C levels to be achieved. However, the AHA made no specific recommendation about how these other risk factors and conditions were to be considered; this was left to the discretion of the treating health care provider.

More recently, the American Academy of Pediatrics' Committee on Nutrition published guidelines for the screening and management of lipid abnormalities in children.^[138]  In these recommendations, pharmacologic therapy is primarily considered at or after age 10 years. Rarely, initiation of treatment with medications could be considered in FH patients ages 8 years and older. Children younger than age 8 years with extreme elevation of LDL–C levels, above 500 mg/dL (as with hoFH), also might be considered for pharmacologic therapy. The Committee also recommended that children with diabetes mellitus and LDL–C ≥130 mg/dL be considered for drug therapy.

Consideration of Associated Risk Factors/High-Risk Conditions

Both pathology studies and studies using noninvasive assessment of vascular markers have shown an exponential increase in arterial abnormalities with increasing numbers of risk factors. For such patients, intensification of therapy and revised thresholds for initiation and treatment targets should be considered, particularly if those other risk factors are present at a higher magnitude or there is a higher level of individual risk. In addition, some patients may have high-risk conditions that are associated with established premature CVD or that serve as additional accelerators to the atherosclerotic process. These high-risk conditions have been highlighted and their management discussed in an AHA scientific statement on CV risk reduction in high-risk pediatric patients.^[86]  A high-risk condition was defined as one associated with manifest CAD at younger than age 30 years; high-risk conditions included hoFH, T1DM, chronic or end-stage renal disease or postrenal transplant, Kawasaki disease complicated by persistent coronary artery aneurysms/stenoses/occlusions, and patients with orthotopic heart transplantation. It was recommended that patients with these conditions be routinely screened for CV risk factors and that those identified be aggressively managed. Moderate-risk conditions were defined as those associated with pathophysiologic evidence that the atherosclerotic process was accelerated and included FH, Kawasaki disease with regressed coronary artery aneurysms, T2DM, and chronic inflammatory disease (juvenile onset rheumatoid arthritis and systemic lupus erythematosus). It was recommended that these patients also be routinely screened for CV risk factors, although levels for initiation of intervention and therapeutic targets were less aggressive. For these Guidelines, either T1DM or T2DM is considered a high-risk condition, and HIV and nephrotic syndrome are added as moderate-risk conditions (Table 9–7). Specific management of these high-risk conditions is outlined in Section XI. Diabetes Mellitus and Other Conditions Predisposing to the Development of Accelerated Atherosclerosis.

OVERVIEW OF THE EVIDENCE FOR SAFETY AND EFFICACY OF MEDICATION THERAPY

Since the 1992 NCEP Pediatric Panel guidelines, a series of RCTs of medications to treat lipid abnormalities in children and adolescents have been completed; all of these are outlined in Table 9–10 and included in the evidence tables that will be available at http://www.nhlbi.nih.gov/guidelines/cvd_ped/index.htm. Several clinical trials of statins^{[21],[22],[23],[139],[140],[141], [142],[143],[144],[145]} and bile acid sequestrants^{[99],[146],[147],[148]} for the treatment of severe elevations of LDL–C in children with FH have been completed and are shown in Table 9–11. These studies have been conducted in children and adolescents with severe dyslipidemia or FH who met the recommendations of the 1992 NCEP guidelines for initiation of medication therapy.^[1]  FH in children was defined as having a family history of elevated LDL–C, atherosclerosis, or CAD in conjunction with having elevated LDL–C. The LDL levels for trial eligibility ranged from a lower limit of 154–189 mg/dL or 95th percentile for age and gender. Per the 1992 NCEP recommendations, almost all the studies tested drug therapy after a trial of diet. The studies have been of relatively short duration, ranging from 6 weeks to 2 years, with several longer than 4 months; one trial was extended as an open-label study to 7.4 years, with both randomized groups receiving drug therapy. Patients in early puberty have been included. Trial subjects were monitored carefully throughout the treatment period. No impact on growth, development, or sexual maturation has been identified; adverse event profiles and efficacy were similar to those noted in studies of adults. Because of problems with palatability, compliance with the bile acid sequestrants has been generally problematic. The details of the safety and efficacy of the statin medications in children are described below with the management of FH. The specific LDL–C-lowering effects of each medication are shown in Table 9–10.

There is limited published experience in children with use of niacin and fibrates, which may be useful in treating patients with combined dyslipidemias.^[149],[150] Efficacy and safety data are limited, and no data are available regarding newer formulations. In adults, cholesterol absorption inhibitors have been advocated as an adjunct to statin therapy for patients who do not reach LDL–C therapeutic targets. Since their action is independent of and complementary to that of statins, the LDL–C-lowering effect is additive. No pediatric studies of monotherapy with cholesterol absorption inhibitors had been published during the time period for this evidence review. Use of niacin, fibrates, and cholesterol-absorption inhibitors should be instituted only in consultation with a lipid specialist.

OVERVIEW OF THE EVIDENCE OF THE IMPACT OF MEDICATION THERAPY ON VASCULAR MARKERS

Although it is unlikely that studies will ever document that treatment of lipid abnormalities in youths will reduce manifest atherosclerotic disease and CV events when they become adults, there is emerging evidence using noninvasive vascular markers that lipid-lowering therapy improves arterial function and structure. From this evidence review, brachial artery reactivity by ultrasound FMD has been assessed as a measure of endothelial function in a clinical trial of simvastatin for children and adolescents with FH.^[23]  At baseline, both placebo and statin intervention FH groups had impaired FMD in response to reactive hyperemia, but this improved significantly in the group treated with simvastatin for 28 weeks. Carotid intima-media thickness by ultrasound has been evaluated as a marker of early atherosclerosis in a clinical trial of pravastatin for children and adolescents with FH.^[21]  After 2 years, cIMT had increased in the placebo group, but there was significant regression in the group treated with pravastatin. An open-label followup study of these patients for an average of 4.5 years reported that earlier initiation of statin therapy was associated with smaller cIMT at followup, after adjusting for baseline cIMT, gender, and duration of treatment.^[22]  This study included patients who had started treatment at age 8 years. These findings contrast with a nonplacebo-controlled, single-arm study with fluvastatin for 2 years, in which no significant changes in cIMT or wall stiffness were noted in response to LDL–C reduction.^[140]  It is important to note that the changes observed in these studies occurred despite the fact that patients did not necessarily achieve recommended LDL–C target cut points and had important residual elevations in LDL–C. These findings represent early evidence that therapy with statins in youth may have a significant positive impact on the atherosclerotic process. These results and findings of better endothelial function assessed by FMD in adolescent boys with lower LDL from infancy^[20] suggest the potential benefit of initiation of LDL-lowering treatment in childhood. The atherosclerotic process is not uniform over a lifetime, and it is probable that early lesions are more effectively treated and reversed than more advanced lesions; this provides some additional potential rationale for initiating drug therapy in youth with severely elevated LDL–C levels.

OVERVIEW OF THE EVIDENCE FOR MANAGEMENT OF SPECIFIC LIPID ABNORMALITIES

Heterozygous Familial Hypercholesterolemia

Heterozygous FH is associated with markedly elevated LDL–C levels and with normal or low HDL–C and usually but not always normal TG levels. FH is inherited as an autosomal dominant trait, with prevalence in the general population of 1:500 but higher in certain ethnic groups (e.g., French Canadians, South Africans, Lebanese). More than 500 mutations resulting in abnormalities of the LDL–C receptor have been identified, ranging from null alleles blocking LDL–C receptor formation and processing to defects resulting in defective receptors with diminished functionality. Alternatively, a similar phenotype is noted for patients with defects of the LDL–C receptor ligand apoB. All these abnormalities result in impaired LDL–C clearance. FH is associated with acceleration of atherosclerosis and premature CVD or events, beginning at ages thirties to forties in men and forties to fifties in women. The risk for clinical events is influenced by the presence of other risk factors or conditions. It is unclear whether FH confers an increased risk beyond that associated with the attendant lipid abnormalities; the fact that the lipid abnormalities are unrelenting from birth may impart an increased cumulative risk beyond that of other conditions and acquired abnormalities. In children with FH, LDL–C elevations are such that the vast majority will meet the criteria for treatment with medication with statins as the mainstay, as reviewed below and in Table 9–11.^{[21],[22],[23],[139],[140],[141], [142],[143],[144],[145],[151]}

A recent systematic review and meta-analysis of statin therapy in children with FH analyzed studies that included almost 800 children.^[152]  No statistically significant differences were found between statin-treated and placebo-treated children for the occurrence of adverse events, sexual development, muscle toxicity, or liver toxicity; there was a minimal difference in growth in favor of the statin group. The LDL–C level and the timing for introduction of medication therapy are outlined in the algorithm (see Figure 9–1), and recommended medications appear in Tables 9–10 and 9–11. Response to the statins can be variable and may relate to the underlying specific genetic abnormality; some patients may require additional therapy, such as bile acid sequestrants, to achieve target LDL–C levels.^{[99],[146],[147],[148],[153]} Cholesterol absorption inhibitors have also been recommended in this situation. A recent RCT of pediatric patients ages 10–17 years with FH demonstrated that coadministration of the cholesterol absorption inhibitor ezetimibe with simvastatin resulted in significantly greater reductions in LDL–C than did simvastatin alone; the combination was safe and well-tolerated up to 53 weeks.^[154]  For those with associated HDL–C and TG abnormalities, intensification of statin therapy or additional therapy with fibrates or niacin would be recommended. Any combination therapy should be undertaken in consultation with a lipid specialist. At this time, the need for therapy and monitoring is lifelong.

Homozygous Familial Hypercholesterolemia

Homozygous FH results in extreme elevations of LDL–C levels (often 5–10 times the upper limit of normal) and decreased HDL–C levels. The magnitude of LDL–C abnormalities may be influenced by which of the two mutations is inherited; these are often not concordant. CVD is usually manifest by the second decade, consisting primarily of coronary ostial stenoses and occlusions, aortic valve thickening with stenosis and/or regurgitation, and extensive atherosclerosis of the aortic root. The most common mode of presentation of hoFH is physical manifestations in infancy and early childhood, consisting of primarily fleshy cutaneous xanthomata between the fingers and toes and over the buttocks, elbows, and knees and tendonous xanthomata, most marked in the Achilles tendon, with nodularity and thickening. Because of the cutaneous manifestations, the diagnosis is often made by dermatologists; additional investigation and management should be made by a lipid specialist. A complete cardiologic investigation is indicated at the time of presentation, since important CVD already may be present, and ongoing careful monitoring is important. There are no RCTs of treatment for hoFH. Despite severely reduced LDL–C receptor capacity, patients may respond somewhat to high doses of statins and to cholesterol absorption inhibitors.^[16],[155]  However, the majority of patients will require artificial clearance of circulating LDL. LDL apheresis specifically removes LDL and is preferred to plasmapheresis, which depletes HDL as well as LDL. LDL apheresis usually is performed biweekly in medical centers with this expertise. At this time, the need for therapy and monitoring is lifelong. Liver transplantation is no longer recommended for children with hoFH because of the marked side effect profile of the procedure. The current goal of therapy is palliation of the disease until a time when effective and safe gene therapy becomes available.

Severe Primary Hypertriglyceridemias

In children with severe hypertriglyceridemia for whom diet and exercise interventions are insufficient, there are nutriceutical and medication options that can be considered. The TG level and the timing for introduction of more advanced therapy are outlined in the algorithm (see Figure 9–2), and recommended medications are shown in Tables 9–10 and 9–11. A recent systematic review demonstrated that omega-3 fish oil capsules are both safe and effective in adults, reducing TG by 30–45 percent, with significant associated increases in HDL–C.^[156]  For children, the safety of omega-3 fish oil was observed in their use in children with immunoglobulin A nephropathy and in a small series of children with dyslipidemia.^[157]  Because fish oil preparations are marketed directly to the public, pediatric care providers can expect to encounter children who are taking these supplements. Information about how to evaluate the various preparations available is provided at the bottom of Table 9–11. In adults, fibrates have been used to lower TG levels, and a small series in children demonstrated effective reductions in TG levels and an associated increase in HDL–C levels.^[150]  Finally, niacin has been used extensively in adults, but there is limited experience in children, with a single series demonstrating a high rate of side effects.^[149] The use of either fibrates or niacin in youths should be undertaken only with the assistance of a lipid specialist.

Children with sustained TG levels ≥500 mg/dL present a rare and serious clinical problem that is usually associated with an underlying genetic defect (LPL deficiency, HL deficiency, or apoC–II deficiency). They are at high risk for pancreatitis beginning in infancy. Management of these patients should always be in consultation with a lipid specialist. These children require a very low-fat diet (<10 percent fat) undertaken with a nutritionist to ensure adequate calories and intake of essential fatty acids. Medium-chain TG, which are absorbed directly into the portal system and do not require chylomicrons for transport to the liver, can have a significant lowering effect on TG levels, especially in those with defective or deficient LPL. Patients with either LPL or apoC–II deficiency do not respond to lipid-altering medications, but patients with HL deficiency will respond to fibrates, niacin, or statins. As indicated in the algorithm, management of these patients should always be in consultation with a lipid specialist.

Isolated Low HDL–C Levels

Isolated low HDL–C levels can occur as a primary abnormality in HDL–C metabolism; this is associated with an increased risk of premature CVD in affected family members. Currently, it is unclear whether this condition contributes to accelerated atherosclerosis in youth, there is no evidence concerning management of this condition in youth, and there is no evidence that treatments aimed at increasing HDL–C levels are effective and safe. There has been some beneficial effect on HDL–C levels associated with niacin therapy.^[16],[149]  There are no RCTs of medication therapy for isolated low HDL–C levels in childhood. Current recommendations for management of this abnormality include attention to lifestyle modification and to abnormalities in TG, LDL–C, and non-HDL–C levels, including lower levels at which to initiate therapy and lower therapeutic target levels. For example, some patients who present with isolated HDL–C levels can have an elevation in the number of small, dense LDL particles that can be detected by determining the apoB level under the direction of a lipid specialist. More aggressive management of other risk factors/conditions is also indicated in the presence of isolated low HDL–C levels.

Elevated Lipoprotein(a)

There is currently no medication therapy specific for elevated Lp(a), and similar to isolated low HDL–C levels, management may focus on addressing other risk factors and on more aggressively managing concomitant elevations of LDL–C, TG, and non-HDL–C.

In adults, niacin will lower Lp(a) approximately 15 percent, but this has not been studied in children.

Combined Lipid Abnormalities

The most common combination of lipid abnormalities is that associated with obesity, the triad of low HDL–C, high TG, and a mild increase in LDL–C levels, with a qualitative change in LDL molecules such that the particle is smaller and denser and particle numbers are higher, contributing to a more atherogenic milieu. In general, these patients do not meet the criteria for pharmacologic therapy, and management focuses on reducing adiposity and changing diet composition as described in the previous lipid subsection on the dyslipidemias. Diet and lifestyle management for elevated TG levels are outlined in the subsection on the dietary treatment of the dyslipidemias, the CHILD 2-TG diet (Table 9–8). However, given that some patients will have clustering of risk factors, there may be a selective role for pharmacologic therapy, most commonly directed at elevations in non-HDL–C levels. Treatment with statins, omega-3 fish oil, or fibrates might be considered, but evidence for this in children is very limited, and there have been no RCTs.

Several primary dyslipidemias are associated with elevations in VLDL–C level, which is often manifest as an elevated LDL–C level together with a high TG level. Specific therapies are not available, and pharmacologic therapy is usually guided by non-HDL–C levels. Statins, omega-3 fish oil, or fibrates should be considered as first-line agents for patients who meet the criteria for medication therapy, under the direction of a lipid specialist.

Guidelines for Initiating and Monitoring Medication Therapy

Statins (Hydroxymethylglutaryl Coenzyme A Reductase Inhibitors)

Given the widespread experience with statins in adult patients, their greater efficacy and tolerance, and the presence of a number of well-designed (albeit relatively short-term) studies in children and adolescents (see Table 9–11),^{[21],[22],[23],,[140],[141], [142],[143],[144],[145], [151]} statin therapy is recommended as the initial medication of choice for treating patients with sufficiently elevated LDL–C or non-HDL–C levels (see algorithm in Figure 9–1). The statins inhibit hydroxymethylglutaryl coenzyme A reductase, which is a rate-limiting enzyme in the endogenous cholesterol synthesis pathway. This results in a decrease in the intracellular pool of cholesterol, which signals upregulation of LDL–C receptors and increased clearance of circulating LDL–C.

The starting dose used for each of the statin medications in the RCTs is shown in Table 9–11. These are preparation specific. Use of these medications requires that practitioners become familiar with the dose recommendations for one of the statins. Statin use should begin with the lowest available dose given once daily. If LDL–C target levels are not achieved with at least 3 months of compliant use, then the dose may be increased by one increment. The risk and effectiveness of dose escalation have been explored in several of the statin clinical trials in children, with no additional safety issues identified.

Adverse effects from statins are rare at standard doses but include myopathy and hepatic enzyme elevation. Asymptomatic hepatic enzyme elevation is fairly common in adults on statin therapy but is reversible with medication change and is not clearly associated with increased risk of liver disease. In the meta-analysis of statin use in children, evidence of hepatic enzyme elevation did not differ between the statin and placebo groups.^[152]  Myopathy—muscle pain and weakness with creatine kinase elevations more than 10 times the upper limits of normal range—typically occurs in fewer than 1 in 10,000 adult patients. Evidence of muscle toxicity did not differ between the statin and placebo groups in the meta-analysis of statin use in children.^[152] Rhabdomyolisis, a very rare occurrence in adults on statin therapy reported at 3 per 100,000 person-years, did not occur in any of the pediatric trials but the total number of subjects is too small to evaluate that risk. The risk of rhabdomyolysis increases with use of higher doses and interacting drugs. Drug interactions with statins occur primarily with drugs that are metabolized by the cytochrome P–450 system, the primary mode of metabolism for the majority of statins. Drugs that potentially interact with statins include fibrates, azol antifungals, macrolide antibiotics, antiarrhthymics, and protease inhibitors. When statin use is initiated, prescribing information must be routinely consulted for potential drug interactions. Patients need to be cautioned about potential future medication interactions, and pediatric care providers need to assess this whenever any new medication is introduced.

In addition to significantly lowering LDL–C levels, statins may increase HDL–C levels modestly and lower TG modestly—effects that are considered beneficial. In adults, treatment with statins also decreases inflammation as judged by the lowering of high-sensitivity C-reactive protein, considered to be a pleiotropic effect of statins. This test is not established for use in the management of lipid disorders in children and adolescents. The statins do not influence levels of the essential fatty acids necessary for early central nervous system maturation and have not been shown to affect neurodevelopment, although studies in infants and very young children have not yet been conducted. Clinical trials have included both male and female children studied over the time of puberty and have shown no impact on sexual maturation or height velocity. Specific guidelines for the use of statins are given in Table 9–12.

Bile Acid Sequestrants

Bile acid sequestrants were the initial medication of choice recommended in the original NCEP Pediatric Guidelines. The rationale was that these agents were not systemically absorbed and thus were believed to be safer for children and adolescents. The sequestrants bind bile salts within the intestinal lumen and prevent their enterohepatic reuptake in the terminal ileum, resulting in a depletion of bile salts in the liver and a signal for increased production. Since bile salts are synthesized from intracellular cholesterol in the liver, the intracellular pool of cholesterol becomes depleted, signaling increased production of LDL–C receptors and increased clearance of circulating LDL–C to replenish the intracellular cholesterol pool for increased production of bile salts. The sequestrants are available in tablet and powder formulations (to be swallowed with or mixed with a liquid of preference). Studies of bile acid sequestrants (cholestyramine, colestipol, colesevelam) in children and adolescents with FH and hence more extreme elevations of LDL–C levels, show reductions of LDL–C levels of 10–20 percent and sometimes a modest elevation in TG levels (see Table 9–11).^{[99],[147],[148],[153]}  The primary adverse effects of the bile acid sequestrants tend to be gastrointestinal in nature, including bloating, nausea, diarrhea, and constipation; these significantly affect compliance. Since the bile acid sequestrants reduce bile salts, which are important for intestinal lipid absorption, there has been some concern regarding malabsorption of fat-soluble vitamins (A, D, E),^[99][158] and routine supplementation with a daily multivitamin and folate may be indicated. Also, bile acid sequestrants may interfere with the absorption of some medications; this potential interaction should be specifically evaluated whenever any additional medication is needed.

The efficacy and adverse effects of bile acid sequestrants are somewhat dose dependent. The degree of LDL–C reduction is often insufficient to achieve LDL–C target levels for the majority of those patients who meet the criteria for pharmacologic therapy, and tolerance and compliance have been reported to be variable in children and may be influenced by the formulation used.^{[99][147],[148]} The initial tablet dose (colestipol) is five tablets (5 g)/d and for the powder (cholestyramine) one packet (4 g)/d. The dose can be titrated upward according to tolerance to a maximum of 20 g/d. The dosage can be divided throughout the day according to the patient's preferences and can be taken with meals. Tablets should not be chewed or divided. Colesevelam is in tablet form with starting dose of 1.875 g/day; dose can be up-titrated to 3.75 g/day; in the clinical trial, compliance with colesevelam was satisfactory. Since the formulation can be a matter of personal preference, patients intolerant or noncompliant with one formulation may do better with the other. Fasting lipid profiles and growth and maturation should be monitored every 6–12 months. No other laboratory testing for safety is required.

The bile acid-binding sequestrants may be used in combination with a statin for patients who fail to meet LDL–C target levels with either medication alone. One pediatric study assessed this combination and showed no increase in adverse effects.^[148]  As expected, the efficacy of the two agents together appears to be additive. Compliance should be closely monitored, since noncompliance with bile acid sequestrants may cause some patients to become noncompliant with the more effective statin as well.

Cholesterol Absorption Inhibitors

Ezetimibe, a cholesterol absorption inhibitor, lowers LDL–C levels by upregulating LDL–C receptors in a manner similar to the bile acid sequestrants, although the mechanism of action differs. Intestinal cholesterol absorption is inhibited at the level of the intestinal villus, resulting in inhibition of absorption of dietary cholesterol. Absorption of bile salts occurs in the ileum by a different mechanism and is not affected by the use of cholesterol absorption inhibitors. Studies in adults have shown that the effect is additive when used with statins, giving an additional 20 percent lowering of LDL–C levels. Adverse effects are minimal, and a dose of 10 mg/d is recommended. At present, there is no evidence that the addition of ezetimibe to a statin provides benefit in the prevention of atherosclerosis or events, and there are no published studies of its use as monotherapy in children. A study of ezetimibe for patients with hoFH, including some children, showed good efficacy and safety regarding lipid lowering.^[155]  A recent RCT of pediatric subjects ages 10–17 years with FH demonstrated that coadministered ezetimibe and simvastatin resulted in significantly greater reductions in LDL–C levels than did simvastatin alone; the combination was safe and well-tolerated for up to 53 weeks.^[154] Currently, ezetimibe might be considered for children and adolescents who, under the care of a lipid specialist, do not meet LDL–C therapeutic targets on a statin alone or as potential monotherapy for patients with lower LDL–C levels who meet the criteria for pharmacologic therapy. In 2009, the U.S. Food and Drug Administration released a safety alert about potential adverse effects of ezetimibe in combination with statins in adults. Until more data are available on its safety, ezetimibe should be used only in consultation with a lipid specialist. Results of future trials should clarify safety issues surrounding use of ezetimibe in children.

CONCLUSIONS AND GRADING OF THE EVIDENCE REVIEW FOR USE OF MEDICATION TO TREAT DYSLIPIDEMIA

When medication is recommended, this should always be in the context of the complete CV risk profile of the patient and in consultation with the patient and the family.

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

• Decisions regarding the need for medication therapy should be based on the average of results from at least two fasting lipid profiles obtained at least 2 weeks but no more than 3 months apart (Grade C) (Figure 9–1).
• The cut points used to define the level at which drug therapy should be considered from the 1992 NCEP Pediatric Guidelines have been used as the basis for multiple drug safety and efficacy trials in dyslipidemic children (Grade B):
  • LDL–C ≥ 190 mg/dL after a 6-month trial of lifestyle management (CHILD 1 à CHILD 2-LDL) for children ages ≥ 10 years.
  • LDL–C 160–189 mg/dL after a 6-month trial of lifestyle/diet management (CHILD 1 à CHILD 2-LDL) in a child ≥ age 10 years with a positive family history of premature CVD/events in first-degree relatives (Table 9–6) or at least one high-level risk factor or risk condition or at least 2 moderate-level risk factors or risk conditions (Tables 9–6, 9–7, and 9–12) (Figure 9–1).
  • LDL–C 130–190 mg/dL in a child ≥ age 10 years with a negative family history of premature CVD in first-degree relatives and no high-level or moderate-level risk factor or risk condition, management should continue to focus on lifestyle changes (CHILD 1à CHILD 2-LDL) based on lipid profile findings (Figure 9–1) plus weight management if BMI is at least the 85th percentile.
  • The goal of LDL-lowering therapy in childhood and adolescence is LDL–C below the 95th percentile (≤130 mg/dL).
• Children with homozygous FH and extremely elevated LDL–C levels (>500 mg/dL) have undergone effective LDL-lowering therapy with biweekly LDL apheresis under the care of lipid specialists in academic medical centers (Grade C).
• Multiple cohort studies have shown benefits of LDL-lowering therapy in children at high risk for accelerated atherosclerosis (Table 9-7). Children and adolescents with chronic kidney disease, T1DM or T2DM, Kawasaki disease with coronary aneurysms, or postcardiac transplantation) should be considered for initiation of medication therapy. (Grade C) (see Section XI. Diabetes Mellitus and Other Conditions Predisposing to the Development of Accelerated Atherosclerosis).
• The bile acid sequestrants are medications that bind bile salts within the intestinal lumen and prevent their enterohepatic reuptake in the terminal ileum, resulting in a depletion of bile salts in the liver and a signal for increased production. Since bile salts are synthesized from intracellular cholesterol in the liver, the intracellular pool of cholesterol becomes depleted, signaling increased production of LDL receptors and increased clearance of circulating LDL–C to replenish the intracellular cholesterol pool for increased production of bile salts. Studies of bile acid sequestrants in children and adolescents ages 6–18 years with LDL–C levels from 131 to 190 mg/dL show TC reduction of 7-17 percent and reduction of LDL–C of 10–20 percent, sometimes with a modest elevation in TG levels. The bile acid sequestrants commonly have gastrointestinal side effects, and these significantly affect compliance. However, they are safe and moderately effective (Grade A).
• Statin medications inhibit hydroxymethylglutaryl coenzyme A reductase, which is a rate-limiting enzyme in the endogenous cholesterol synthesis pathway. This results in a decrease in the intracellular pool of cholesterol, which signals upregulation of LDL receptors and increased clearance of circulating LDL–C. As a group, statins have  been shown to reduce LDL–C in children and adolescents with marked LDL–C elevation or FH  (defined as elevated LDL–C in the child in conjunction with a family history of elevated LDL–C and/or atherosclerosis or CAD) when used from 8 weeks to 2 years for children ages 8–18 years. The lower LDL–C level for eligibility into the statin trials was >190 mg/dl or > 160 mg/dl with 2 or more additional risk factors, after a trial period on diet. Trial subjects were monitored carefully throughout treatment; adverse impacts on growth, development, or sexual maturation were not seen, and adverse event profiles and efficacy were similar to those in studies of adults (Grade A).
• Adverse effects from statins are rare at standard doses but include myopathy and hepatic enzyme elevation. In the meta-analysis of statin use in children, evidence of hepatic enzyme elevation and muscle toxicity did not differ between the statin and placebo groups. Routine monitoring of hepatic enzymes and clinical assessment for muscle toxicity are strongly recommended for children and adolescents on statin therapy (Table 9–12). The risk of adverse events increases with use of higher doses and interacting drugs, the latter occurring primarily with drugs that are metabolized by the cytochrome P–450 system, which is the primary mode of metabolism for the majority of statins. Drugs that potentially interact with statins include fibrates, azol antifungals, macrolide antibiotics, antiarrhthymics, and protease inhibitors (Grade A).
• Bile acid-binding sequestrants may be used in combination with a statin for patients who fail to meet LDL–C target levels with either medication alone. One pediatric study assessed this combination and showed no increase in adverse effects. The efficacy of the two agents together appears to be additive (Grade B).
• There is limited published experience in children with use of niacin and fibrates, which have been useful in treating adult patients with combined dyslipidemias. Efficacy and safety data are limited, and no data are available regarding newer formulations. In adults, cholesterol absorption inhibitors have been advocated as an adjunct to statin therapy for patients who do not reach LDL–C therapeutic targets. Since their action is independent of and complementary to that of statins, the LDL–C-lowering effect is additive. No pediatric studies of monotherapy with cholesterol absorption inhibitors had been published during the time period for this evidence review. Use of niacin, fibrates, and cholesterol absorption inhibitors should be instituted only in consultation with a lipid specialist (Grade C).
• Medication therapy is rarely needed for children with elevated TG who respond well to weight loss and lifestyle changes (Grade B). When TG levels exceed 500 mg/dL, patients are at risk for pancreatitis and require care in consultation with a lipid specialist (Grade B). In adults, use of omega-3 fish oil has been shown to lower TG by 30–40 percent and to raise HDL by 6–17 percent. Experience with fish oil in children is limited to small case series with no safety concerns identified; there have been no RCTs of fish oil in children (Grade D).

AGE-BASED RECOMMENDATIONS FOR MEDICATION THERAPY OF CHILDREN WITH DYSLIPIDEMIA

Children Younger Than Age 10 Years

• Children younger than age 10 years should not be treated with a medication unless they have a severe primary hyperlipidemia or a high-risk condition that is associated with serious medical morbidity (homozygous hypercholesterolemia/ LDL–C ≥400 mg/dL; primary hypertriglyceridemia with TG ≥500 mg/dL; evident CVD in the first two decades of life; post cardiac transplantation) (Grade C).

Children Ages 10–21 Years (see algorithms, Figures 9–1 and 9–2)

• Decisions regarding the need for medication therapy should be based on the average of results from ≥ two FLPs obtained at least 2 weeks but no more than 3 months apart. (Grade C) (Figure 9–1)
• Children with average LDL–C ≥ 250 mg/dL or average TG > 500 mg/dL should be referred directly to a lipid specialist (Grade B).
• Children with lipid abnormalities should have a detailed family history taken and be assessed for causes of hyperlipidemia, additional risk factors, and risk conditions (Grade C) (Tables 9–3, 9–6, and 9–7).
• Children with lipid abnormalities (other than LDL–C ≥250 mg/dL or TG ≥500 mg/dL) should be initially managed for 3–6 months with diet changes (CHILD 1àCHILD 2-LDL or CHILD 2-TG, Table 9–8) based on specific lipid profile findings (Figures 9–1 and 9–2); if BMI is ≥ 85th percentile, add increased physical activity, reduce screen time, and restrict calories. Assessment for associated secondary causes (Table 9–3), additional risk factors, or high-risk conditions (Tables 9–6 and 9–7) is recommended. Children at high risk who are unlikely to achieve lipid targets with this strategy alone (severe primary dyslipidemia, cardiac transplantation) should concomitantly be considered for initiation of medication therapy (Grade C) (Section XI. Diabetes Mellitus and Other Conditions Predisposing to the Development of Accelerated Atherosclerosis).

LDL–C:  Treatment for children with severe elevation of LDL–C is based on assessment of lipid levels and associated risk factors or risk conditions (Tables 9–6 and 9–7; Figures 9–1 and 9–2):

• Children with average LDL–C ≥ 250 mg/dL should be referred directly to a lipid specialist (Grade B).
• If LDL–C remains ≥ 190 mg/dL after a 6-month trial of lifestyle/diet management (CHILD 1→CHILD 2-LDL) for children ages 10 years and older, statin therapy should be considered (Grade A) (Figure 9–1) (Table 9–12).
• If LDL–C remains 130–190 mg/dL in a child age 10 years or older with a negative family history of premature CVD in first-degree relatives and no high-level or moderate-level risk factor or risk condition (Tables 9–6 and 9–7), management should continue to focus on diet changes (CHILD 2-LDL) based on lipid profile findings (Figure 9–1) plus weight management if BMI is ≥ 85th percentile. Pharmacologic therapy is not generally indicated, but treatment with bile acid sequestrants might be considered, the latter in consultation with a lipid specialist (Grade B).
• If LDL–C remains 160–189 mg/dL after a trial of lifestyle/diet management (CHILD 1→CHILD 2-LDL) in a child age 10 years or older with a positive family history of premature CVD/events in first-degree relatives (Table 9–6) or at least one high-level risk factor or risk condition or at least two moderate-level risk factors or risk conditions (Tables 9–6 and 9–7), then statin therapy should be considered (Grade B) (Figure 9–1) (Table 9–12).
• If LDL–C remains ≥ 130 to 159 mg/dL after a trial of lifestyle/diet management (CHILD 1→CHILD 2-LDL) in a child age 10 years or older with at least two high-level risk factors or risk conditions or at least one high-level risk factor or risk condition together with at least two moderate-level risk factors or risk conditions (Tables 9–6 and 9–7), then statin therapy should be considered (Grade C) (Figure 9–1) (Table 9–12).
• For children ages 8 and 9 years with LDL–C persistently ≥ 190 mg/dL after a trial of lifestyle/diet management (CHILD 1→CHILD 2-LDL), together with multiple first-degree family members with premature CVD/events, or the presence of at least one high-level risk factor or risk condition or the presence of at least two moderate-level risk factors or risk conditions (Figure 9–1) (Tables 9–6 and 9–7), statin therapy might be considered (Grade B) (Table 9–12).
• Statin use should begin with the lowest available dose given once daily. If LDL–C target levels are not achieved with at least 3 months of compliant use, then the dose may be increased by one increment (usually 10 mg). If LDL–C target levels are still not achieved with at least 3 months of compliant use, then the dose may be further increased by one increment. The risk and effectiveness of dose escalation has been explored in several of the statin clinical trials in children with no additional safety issues identified (Grade B). Alternatively, a second agent such as a bile acid sequestrant or cholesterol absorption inhibitor may be added under the direction of a lipid specialist (Grade B) (Table 9–12).
• Children taking a statin should have routine clinical monitoring for symptoms of muscle toxicity and assessment of hepatic transaminases and creatine kinase (Grade A) (Table 9–12).
• Pediatric care providers should be on the alert for, and children and their families should be counseled about, potential medication interactions (Grade D) (Table 9–12).
• Females taking a statin should be counseled about risks associated with pregnancy and appropriate contraception strategies if indicated. Use of oral contraceptives in combination with statins is not contraindicated (Grade D) (Table 9–12).

TG, non-HDL–C:  Children with elevated TG or elevated non-HDL–C after control of LDL–C are managed based on lipid levels (Figure 9–2):

• Children with average fasting levels of TG ≥ 500 mg/dL or any single measurement ≥ 1,000 mg/dL related to a primary hypertriglyceridemia should be treated in conjunction with a lipid specialist; the CHILD 2-TG diet (Table 9–8) should be started and use of fish oil, fibrate, or niacin to prevent pancreatitis should be considered (Grade D) ( Figure 9–2) (Tables 9–10 and 9–11).
• Children with fasting levels of TG 200–499 mg/dL after a trial of lifestyle/diet management with CHILD 1→CHILD 2-TG (Table 9–8) should have non-HDL recalculated and be managed to a goal of less than 145 mg/dL (Grade D).
• Children with fasting levels of TG 200–499 mg/dL, non-HDL ≥ 145 mg/dL, after a trial of lifestyle/diet management with CHILD 1→CHILD 2-TG (Table 9–8), including increased fish intake, may be considered for fish oil supplementation (Grade D) (Table 9–10).
• Children ages 10 years or older with non-HDL–C levels ≥145 mg/dL after the LDL–C goal is achieved may be considered for further intensification of statin therapy or additional therapy with a fibrate or niacin, in conjunction with referral to a lipid specialist (Grade D) (Figure 9–1) (Tables 9–10 and 9–11).
• Children with severe or complex mixed dyslipidemias, particularly where multiple medications are being considered, should be referred for consultation with a lipid specialist (Grade D) (Figures 9–1 and 9–2).

The age-specific recommendations for pharmacologic management of dyslipidemia are summarized in Table 9–9.

Table 9–9. Evidence-Based Recommendations for Pharmacologic Treatment of Dyslipidemia

Grades reflect the findings of the evidence review.
Recommendation levels reflect the consensus opinion of the Expert Panel.
When medication is recommended, this should always be in the context of the complete cardiovascular risk profile of the patient and in consultation with the patient and the family.

NOTE: Values given are in mg/dL. To convert to SI units, divide the results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

^* Consideration of drug therapy based on the average of ≥2 FLPs, obtained at least 2 weeks but no more than 3 months apart.
^** If average LDL–C ≥190 mg/dL after CHILD 2-LDL and child is age 8–9 years with + FHx OR ≥1 high-level RF/RC OR ≥ 2 moderate-level RFs/RCs, statin therapy may be considered.
^*** Consideration of drug therapy based on the average of ≥2 fasting lipid profiles obtained at least 2 weeks but no more than 3 months apart.
^† If child is obese, nutrition therapy should include calorie restriction and increased activity beyond that recommended for all children. See Section X. Overweight and Obesity for additional age-specific recommendations.

Table 9–10. Medications for Managing Hyperlipidemia

Table 9–11. Clinical Trials of Lipid-Lowering Medication Therapy in Children and Adolescents

Bile acid binding resins

HMG CoA reductase inhibitors (statins)

 Other agents

ABBREVIATIONS: AHA = American Heart Association; CAD = coronary artery disease; d = day; FHx = family history; g = grams; mg = milligrams; NA = not available; NC = not calculated; TC = total cholesterol; FH = heterozygous familial hypercholesterolemia; FCHL = familial combined hyperlipidemia; RCT = randomized controlled trial. tx = treatment

** There is only one FDA-approved fish oil preparation, but there are many generic forms of fish oil capsules that are commercially available. The University of Wisconsin maintains a preventive cardiology patient education Web site http://www.heartdecision.org. The "fish oil" section includes information about the content of various preparations. The Web site is updated every 6 months (https://www.heartdecision.org/chdrisk/v_hd/patient_edu_docs/Fish_Oil_11-2007.pdf

Table 9–12. Recommendations for Use of HMG- CoA reductase Inhibitors (Statins) in Children and Adolescents

Figure 9-1. Dyslipidemia Algorithm: TARGET LDL-C (Low-Density Lipoprotein Cholesterol)

NOTE: Values given are in mg/dL. To convert to SI units, divide results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

Figure 9-1 Description

The figure is a flow chart with 13 labeled boxes linked by arrows. The chart flows in one direction with arrows pointing downward and lateral arrows to one or more boxes.

Below the flow chart is described as lists in which the possible next steps are listed beneath each box label.

1. Fasting lipid profile (FLP) x 2^*, average
   1. Forward to LDL-C > 130, < 250 mg/dL ^** → Target LDL-C TG > 100, < 500 mg/dL, < 10 y → Target TG > 130, < 500 mg/dL, 10-19 y
2. TG ≥ 500 mg/dL,
   → Consult lipid specialist
   1. Lateral to LDL-C ≥ 250 mg/dL → Consult lipid specialist
3. LDL-C ≥ 250 mg/dL
   → Consult lipid specialist
   1. Lateral to TG ≥ 500 mg/dL, → Consult lipid specialist
4. LDL-C ≥ 130, < 250 mg/dL ^** → Target LDL-C
   TG≥ 100, < 500 mg/dL, < 10 y → Target TG
   ≥ 130, < 500 mg/dL, 10-19 y (see TG algorithm, Figure 9-2)
   1. Forward to FLP
   2. Boxed text with no arrow between the two boxes: Exclude secondary causes. Evaluate for other risk factors (RFs). Start Cardiovascular Health Integrated Lifestyle Diet (CHILD 1) → CHILD 2-LDL (Table 9-8) + lifestyle change x 6 months^***
5. FLP
   1. Forward to LDL-C < 130 mg/dL
   2. Forward to LDL-C ≥130 to -189 mg/dL Family history (FHx) (-) No other RFs
   3. Forward to LDL-C ≥ 190 mg/dL
   4. Forward to LDL-C ≥ 160 to -189 mg/dL FHx (+) or 1 high-level RF or ≥ 2 moderate-level RFs
   5. Forward to LDL-C ≥ 130 to -159 mg/dL + 2 high-level RFs or 1 high-level + ≥ 2 moderate-level RFs OR clinical CVD
6. LDL-C < 130 mg/dL
   → Continue CHILD 2-LDL
   → Repeat FLP q. 12 months
7. LDL-C ≥130 to -189 mg/dL Family history (FHx) (-) No other RFs
   → Continue CHILD 2 LDL, Follow q. 6 m with FLP, FHx/ RF update
8. LDL-C ≥ 190 mg/dL
   → Initiate statin therapy (Tables 9-11, 9-12)
   1. Forward to Follow with FLPs, related chemistries per Table 9-12
9. LDL-C ≥ 160 to -189 mg/dL FHx (+) or 1 high-level RF or ≥ 2 moderate-level RFs
   →Initiate statin therapy (Tables 9-11, 9-12)
   1. Forward to Follow with FLPs, related chemistries per Table 9-12
10. LDL-C ≥ 130 to -159 mg/dL + 2 high-level RFs or 1 high-level + ≥ 2 moderate-level RFs OR clinical CVD
    → Initiate statin therapy (Tables 9-11, 9-12)
    1. Forward to Follow with FLPs, related chemistries per Table 9-12
11. Follow with FLPs, related chemistries per Table 9-12
    1. Forward to → LDL-C still≥130 mg/dL, TG <200 mg/dL, refer to lipid specialist for addition of second lipid-lowering agent; monitor per Table 9-12 → In high LDL-C patients, if non-HDL-C ≥145 mg/dL after effective LDL-C treatment, → Target TG (Figure 9-2)
12. → LDL-C still ≥130 mg/dL, TG <200 mg/dL, refer to lipid specialist for addition of second lipid-lowering agent; monitor per Table 9-12
    → In high LDL-C patients, if non-HDL-C ≥145 mg/dL after effective LDL-C treatment,
     → Target TG (Figure 9-2)

Figure 9-1 Footnotes:

^* Obtain FLPs at least 2 weeks but no more than 3 months apart.
^** Per Table 5, use of drug therapy is limited to children ≥10 y with defined risk profiles.
^*** In a child with LDL-C > 190 mg/dL and other RFs, trial of CHILD 2 LDL diet may be abbreviated.

Figure 9-2. Dyslipidemia Algorithm: TARGET TG (Triglycerides)

NOTE: Values given are in mg/dL. To convert to SI units, divide results for total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C) high-density lipoprotein cholesterol (HDL-C), and non-HDL-C by 38.6; for triglycerides (TG), divide by 88.6.

Figure 9-2 Description

The figure is a flow chart with 9 labeled boxes linked by arrows. The chart flows in one direction with arrows pointing downward and lateral arrows to one or more boxes.

Below the flow chart is described as lists in which the possible next steps are listed beneath each box label.

1. Fasting lipid profile (FLP) x 2^*, average results
   1. Forward to LDL-C ≥ 130, < 250 mg/dL ^** → Target LDL-C TG≥ 100, < 500 mg/dL, < 10 y → Target TG ≥ 130, < 500 mg/dL, 10-19 y
2. TG ≥ 500 mg/dL,
   → Consult lipid specialist
   1. Lateral to LDL-C ≥ 250 mg/dL → Consult lipid specialist
3. LDL-C ≥ 250 mg/dL
   → Consult lipid specialist
   1. Lateral to TG ≥ 500 mg/dL, → Consult lipid specialist
4. LDL-C ≥ 130, < 250 mg/dL ^** → Target LDL-C (see LDL algorithm, Figure 9-1)
   TG≥ 100, < 500 mg/dL, < 10 y → Target TG
   ≥ 130, < 500 mg/dL, 10-19 y
   1. Forward to TARGET TGs → Cardiovascular Health Integrated Lifestyle Diet (CHILD 1) → CHILD 2 TG diet (Table 9-8) + lifestyle modification with weight loss goal as needed × 6 months
5. TARGET TGs → Cardiovascular Health Integrated Lifestyle Diet (CHILD 1) → CHILD 2 TG diet (Table 9-8) + lifestyle modification with weight loss goal as needed × 6 months
   1. Forward to FLP
6. FLP
   1. Forward to TG <100 mg/dL <10 y, <130 mg/dL, 10-19 y
   2. Forward toTG ≥100, <200 mg/dL, <10 y ≥130, <200 mg/dL, 10-19y
   3. Forward to TG ≥200-499 mg/dL
7. TG <100 mg/dL <10 y,
   <130 mg/dL, 10-19 y
   → Continue CHILD 2-TG + lifestyle change
   →Reassess q .12 m
8. TG ≥100, <200 mg/dL, <10 y
   ≥130, <200 mg/dL, 10-19y
   → Intensify CHILD 2-TG + wt loss ' Increase dietary fish content**
   → Repeat FLP in 6 m
9. TG ≥200-499 mg/dL
   → If LDL-C target achieved and non-HDL ≥145 mg/dL → lipid specialist for drug therapy (statin+/-fibrate+/-nicotinic acid)
   → Consider omega-3 fish oil therapy

Figure 9-2 Footnotes:

^* Obtain FLPs at least 2 weeks but no more than 3 months apart.
^** The Food and Drug Administration (FDA) and the Environmental Protection Agency are advising women of childbearing age who may become pregnant, pregnant women, nursing mothers, and young children to avoid some types of fish and shellfish and eat fish and shellfish that are lower in mercury. For more information, call the FDA's food information line toll free at 1-888-SAFEFOOD or visit http://www.cfsan.fda.gov/~dms/admehg3.html.

Back to Top

Back to Table of Contents