Background: For purposes of this report, multifactorial dyslipidemia refers to dyslipidemias involving elevated total cholesterol (TC) or low-density lipoprotein cholesterol (LDL-C) that are not familial hypercholesterolemia (FH). There is evidence that elevated TC and LDL-C concentrations in childhood, and especially adolescence, are associated with markers of atherosclerosis in young adults.
Purpose: We conducted this systematic evidence review on the benefits and harms of screening adolescents and children for multifactorial dyslipidemia to support the U.S. Preventive Services Task Force (USPSTF) in updating its previous recommendation.
Data sources:
We searched MEDLINE, the Cochrane Central Register of Controlled Trials, and PubMed for trials published between January 1, 2006 and December 31, 2014 relevant to all key questions (KQs) posed by the USPSTF. We supplemented our searches with reference lists from existing systematic reviews, cohort studies, suggestions from experts, and ongoing trials registered in
Study Selection: Investigators independently evaluated 7,137 abstracts and 537 articles against a priori inclusion criteria. Investigators also independently critically appraised each study using design-specific quality criteria based on USPSTF methods. Only fair- or good-quality studies that met the a priori criteria were chosen for each KQ. Differences between investigators were resolved by consensus.
Data Extraction and Analysis: One investigator abstracted data from 16 included studies into evidence tables and a second reviewer verified the accuracy of these data. We qualitatively summarized the evidence regarding screening for multifactorial dyslipidemia and its treatment.
Results: We found no direct evidence for an effect of screening children on adult health outcomes (KQ1), intermediate health outcomes (KQ2), harms of screening (KQ4), or effect of treatment on adult health outcomes (KQ5). Studies met the inclusion criteria for four KQs.
KQ3. What is the diagnostic yield of screening for multifactorial dyslipidemia in children and adolescents?
Neither selective nor universal screening of children for multifactorial dyslipidemia has been evaluated in randomized, controlled trials (RCTs). Fair evidence indicates that a screening TC concentration of 200 mg/dL (the National Cholesterol Education Program's [NCEP's] recommended cut point for high TC) has a positive predictive value of 77 percent for a diagnosis of multifactorial dyslipidemia. Recent nationally representative prevalence estimates suggest a simulated diagnostic yield of 5.4 percent for 8- to 12-year-olds and 6.5 percent for 13- to 17-year-olds. Simulated diagnostic yields ranged between 4 and 12 percent for different age and body mass index subgroups. Based on large, recent U.S. studies, the highest diagnostic yield appears to be in obese children (12.3%), children ages 9 to 11 years (7.2%), and adolescents ages 16 to 19 years (7.2%). The Coronary Artery Risk Detection in Appalachian Communities study was conducted in the same age group recommended by the National Heart, Lung, and Blood Institute expert panel for universal testing, so the diagnostic yields of 5.8 percent, 8.9 percent, and 12.3 percent for healthy weight, overweight, and obese 10- and 11-year-olds are reasonable estimates of what might be seen for those subgroups if screening among 9- to 11-year-olds was more widely adopted.
KQ6. Does treatment of multifactorial dyslipidemia with lifestyle modifications and/or lipid-lowering medications in children and adolescents improve intermediate outcomes (i.e., reduce lipid concentrations or reverse or slow the progression of atherosclerosis) in childhood and adolescence?
We identified only one RCT of a dietary intervention, so the body of evidence on this question is fair at best. This trial provided good-quality evidence of a modest effect of dietary counseling for a low-fat, low-cholesterol diet on lipid levels in children ages 8 to 10 years with mild-to-moderate dyslipidemia. After 1 year of relatively intensive counseling, mean differences between groups compared to baseline were a decrease of 6 mg/dL in mean TC and a decrease of 5 mg/dL in mean LDL-C; both differences were statistically significant. Both between-group differences were reduced to about 3 mg/dL at year 3, and by year 5 were no longer statistically significant. Adherence to the diet during the intervention was good, and some evidence indicated that diet quality improved in children in the intervention group.
A small, 4-week RCT of flaxseed found no cholesterol-lowering effect. We found no evidence to support any other type of treatment for multifactorial dyslipidemia, including nutritional supplements or medications.
KQ7. What are the harms of treatment of multifactorial dyslipidemia with lifestyle modifications and/or lipid-lowering medications in children and adolescents?
Dietary changes made by children in the intervention group of the Dietary Intervention Study in Children (DISC) did not adversely affect children's nutritional status, growth, or development.
KQ8. What is the association between intermediate outcomes in childhood and adolescence and future incidence of myocardial infarction (MI) and stroke events in adults?
In a single, high-quality longitudinal study that included data on adolescents and young adults from the National Health and Nutrition Examination Survey (NHANES), TC concentration at ages 12 to 39 years was not associated with death before age 55 years. The study did find that very high TC concentrations (≥240 mg/dL) were significantly associated with premature death in women only, although this estimate was based on a small number of deaths. The subgroup of females with concentrations this high is likely dominated by individuals who have FH, for whom premature coronary heart disease deaths are expected. The meaning of this finding is unclear because of the small number of deaths in this subgroup.
Limitations: Screening (KQ3)
The availability of confirmatory testing for only one of the included studies required that the diagnostic yield be estimated from simulations. The screening evidence draws heavily from epidemiologic studies rather than screening trials. The prevalence of elevated cholesterol concentrations across broad age ranges, as is often reported, is less meaningful than age group–specific rates given the known variation by age. Only one school-based study was relevant to the primary care setting. The use of the fixed NCEP thresholds for defining elevated TC and LDL-C concentrations makes the prevalence of dyslipidemia (and diagnostic yield) difficult to interpret because they do not account for variability by age and sex. Finally, incomplete tracking of lipid concentrations between childhood, adolescence, and adulthood is another limitation of this body of evidence. Multifactorial dyslipidemia in children and adolescents younger than age 20 years is likely to resolve in the young adult years; thus, the importance of dyslipidemia identified in childhood or adolescence for health in adulthood is unclear.
Screening trials for detecting multifactorial dyslipidemia that include confirmatory testing are needed. Many studies were excluded because they used cut points for cholesterol concentrations that were lower than accepted cut points. Even the accepted fixed NCEP 1992 cut points for LDL-C and TC concentrations may be inadequate because they may over- or underidentify children and adolescents, depending on age and sex. We found no trials that compared either universal or selective screening to no screening. Data on prevalence and diagnostic yield in racially/ethnically diverse populations of children are lacking. No studies addressed harms of screening.
Treatment (KQ6 and KQ7)
The DISC trial selected children with lower levels of hyperlipidemia, making it difficult to generalize results to children with TC concentrations greater than 200 mg/dL or LDL-C concentrations greater than 130 mg/dL. This trial also targeted 8- to 10-year-olds, so the impact of dietary intervention on adolescents is unknown. The intensity of the counseling intervention limits the generalizability of this treatment to primary care settings, where trained nutritional counselors may not be part of the health care team. Finally, the clinical importance of the small impact on cholesterol levels in the 1 to 3 years of the study time frame is unclear.
The evidence regarding treatment has several gaps. We found only one high-quality RCT of dietary intervention for children with multifactorial dyslipidemia. Larger RCTs with long-term followup are needed. Rigorous trials of dietary supplements and medications to reduce levels of atherogenic lipids in children and adolescents are also needed. Our search revealed no trials of statins in this population. The treatment studies reviewed for inclusion also relied on various TC and LDL-C cut points, often below the standard NCEP thresholds. Few studies followed participants for a year or longer.
Outcomes (KQ8)
Outcomes for adolescents were not reported separately from those for young adults. The single included study for KQ8 reported all-cause and endogenous-cause mortality but not cardiovascular mortality, an outcome which would be more closely linked to the causal pathway. Long-term followup cohort studies of children are needed to better understand the association between pediatric dyslipidemia and adult MI and stroke.
Conclusions: We found no direct evidence for an effect of cholesterol screening on intermediate or health outcomes. Only one study provided a diagnostic yield for TC screening (5.8%). Simulated diagnostic yields in data from large U.S. population-based studies show variation in TC concentrations by age and body mass index. There were no studies of diagnostic yield in selective screening and no studies on the harms of screening.
No evidence was found for an effect of treatment on health outcomes in adulthood (MI and stroke). Dietary counseling may lower TC and LDL-C concentrations by 5 to 7 mg/dL over 3 years, but this intervention was relatively intensive and the effect on lipids dissipates by 5 years. No studies of lipid-lowering medications met the inclusion criteria. There is fair evidence of the safety of dietary intervention in 8- to 10-year-olds.
In one longitudinal study using NHANES data combined for both sexes, neither very high nor moderately increased TC concentrations in 12- to 39-year-olds was independently associated with death before age 55 years.
Research needs include randomized trials of screening strategies with confirmatory testing and long-term followup, as well as rigorous RCTs of promising medications, supplements, and dietary interventions with long-term followup. Long-term followup of pediatric cohorts is needed to better establish the long-term health risks conferred by elevated concentrations of TC and LDL-C. Although not the focus of this systematic review, our findings support a re-examination of the commonly accepted fixed NCEP thresholds indicating elevated TC and LDL-C concentrations and a reconsideration of age- and sex-specific thresholds.