- Title
-
Screening for Lipid Disorders in Children and Adolescents: An Evidence Update for the U.S. Preventive Services Task Force
- Description
-
Background: Familial hypercholesterolemia (FH) is a genetic disorder of lipoprotein metabolism characterized by highly elevated low density lipoprotein cholesterol (LDL-C) levels early in life and is associated with substantial long-term cardiovascular risk. Multifactorial dyslipidemia includes dyslipidemias that are not FH that may be associated with environmental factors, with or without an inherited component. Lipid screening in childhood and adolescence can lead to early diagnosis of FH and non-FH multifactorial dyslipidemia. The long-term potential benefits of lipid screening and subsequent treatment are uncertain.
Purpose: To systematically review evidence for the effectiveness and harms of screening and treatment of pediatric dyslipidemia due to FH and multifactorial dyslipidemia.
Data Sources: We searched MEDLINE, and the Cochrane Central Register of Controlled Clinical Trials to identify literature that was published between January 2015 and May 16, 2022. Studies included in the 2016 review for the USPSTF were re-evaluated for potential inclusion. We supplemented our searches with reference lists from the previous review, relevant existing systematic reviews, suggestions from experts, and Clinicaltrials.gov to identify ongoing trials. We conducted ongoing surveillance for relevant literature through March 24, 2023.
Study Selection: Two investigators independently reviewed 7,058 abstracts and 272 full-text articles against prespecified inclusion criteria. We included English-language publications of studies conducted among children and adolescents 20 years of age or younger in countries categorized as “Very High” on the Human Development Index. For studies evaluating the benefits and harms of lipid screening, we included RCTs of universal or selective screening using a serum lipid panel compared to no screening or usual care that reported a health outcome (MI, ischemic stroke, CVD mortality, or all-cause mortality), intermediate outcome (serum lipid concentrations, atherosclerosis markers, BMI), intermediate behavioral outcome (physical activity, sedentary behavior, dietary intake), or harm. For studies evaluating the diagnostic yield of serum lipid screening, we included recent, large U.S. cohort studies that conducted universal or selective lipid screening and reported screen positivity for any stated threshold of abnormal lipids based on a single lipid test or the positive predictive value of a first elevated screening lipid result for a second confirmatory test. For studies evaluating the benefits and harms of treatment, we included RCTs of lipid-lowering medications, behavioral counseling interventions, and dietary supplements that had a comparator group of no treatment, placebo, or usual care and reported a health outcome, intermediate outcome, intermediate behavioral outcome, or harm. One investigator abstracted data into an evidence table and a second investigator checked these data.
Data Analysis: Random effects meta-analysis was used to evaluate the lipid-lowering efficacy of interventions with sufficient evidence to warrant pooled analyses. Other analyses for each key question were qualitative.
Results: 43 studies were eligible for inclusion (n=491,516). Twenty-six studies (n=437,000) were in children and adolescents with familial hypercholesterolemia (FH), 9 studies (n=143,265) in children and adolescents with multifactorial dyslipidemia, and 9 studies (n=10,624) were among children and adolescents with FH or multifactorial dyslipidemia.
Familial Hypercholesterolemia:
Direct Screening Benefits and Harms (KQ 1, 3): There were no randomized screening trials directly addressing the effectiveness and harms of screening for FH in children and adolescents.
Screening Yield (KQ 2): No studies performed a confirmatory lipid or genetic test; thus, evidence is limited to screen-positivity (prevalence) rather than diagnostic yield of lipid screening for identifying FH. We included three fair-quality U.S. studies (n=395,465) reporting prevalence of FH of 0.2 to 0.4 percent (1:250 to 1:500) using diagnostic criteria exclusively based on lipid levels (LDL-C ≥190 mg/dL or TC ≥270 mg/dL). One study showed that targeted screening in those with a family history would miss many cases of children with LDL-C ≥160 mg/dL (prevalence in those with family history: 1.2%, prevalence in those without family history: 1.7%).
Treatment Benefits (KQ4): We included 22 fair- to good-quality trials (n=2,257) examining the effectiveness of various lipid-lowering treatments for FH including pharmacotherapy, behavioral counseling, and dietary supplements. Ten fair- to good-quality randomized, controlled trials (RCTs) (N=1,230) of statins comprised the largest body of evidence addressing FH treatment with followup up to 2 years. Pooled analyses demonstrated that statins were associated with an 81-82 mg/dL greater mean difference in total cholesterol (TC) and LDL-C compared to placebo at up to 2 years followup. Within-trial comparisons demonstrated that higher doses were generally associated with greater reductions in TC and LDL-C compared to lower doses, but confidence intervals overlapped. Pooled analysis showed no statistically significant difference in HDL-C. Individual trials showed mixed results for triglycerides (TG). We included one good- and two fair-quality bile acid sequestrant trials (n=332) trials demonstrating a significantly greater reduction in TC ranging from -22.1 to -40.6 mg/dL and LDL-C ranging from -13.2 to -45.9 mg/dL compared to placebo at 8 weeks. Bile acid sequestrants were not associated with statistically significant reductions in TG and results were mixed for HDL-C, with some variation in effect by dose. We included one good quality ezetimibe trial (n=138) showing a statistically significant 63.0 to 65.0 mg/dL mean reduction in TC and LDL-C, and non-HDL-C. Changes in HDL-C and TG were not significant. We included one very small fair-quality fibrate trial (N=14) reporting a statistically significant 84.9 mg/dL mean reduction in TC but no significant differences in HDL-C or TG at 13 weeks; however, this drug is not available in the U.S and not FDA-approved in children. One good quality PCSK9 inhibitor trial (n=158) demonstrated that evolocumab was associated with a statistically significant 38.3 percent reduction in LDL-C and absolute mean reduction of 68.6 mg/dL with 60.2 percent greater absolute difference in achievement of goal LDL-C <100 mg/dL compared to placebo at 24 weeks. One previously included trial of a statin and ezetimibe drug combination compared to a statin alone (n=248) showed that the two-drug intervention was associated with a 37.5 to 40.1 mg/dL greater reduction in TC, LDL-C, and non-HDL-C, and a 9.5 mg/dL median difference in percent change of TG compared to the single-drug intervention control group at 33 weeks.
We included one very small fair-quality behavioral counseling trial (n=21) in an FH population that reported no statistically significant improvement in lipid levels, overlapping confidence intervals for physical activity outcomes, and mixed results for dietary outcomes at 12 weeks.
We included four fair-quality randomized crossover supplement trials (n=116) in FH populations. The two trials of plant sterol food spreads demonstrated statistically significant reductions of 20.5 to 30.5 mg/dL in TC and 22.4 to 30.1 mg/dL in LDL-C at 4 to 8 weeks. The two trials of omega-3 fatty acids did not show a statistically significant difference in lipid level changes between the intervention and control groups.
Treatment Harms (KQ5): Harms reported in statin trials were similar in the intervention and control groups; however, most studies were relatively short term and small, with few events leading to imprecise estimates. Transaminitis (elevations in alanine transaminase [ALT] or aspartate transaminase [AST]) of three times or more the upper limit of normal occurred in 0 to 4.5 percent of participants in the intervention groups and 0 to 1.9 percent in control groups. The largest trial (n=214) with 2-year followup reported no cases in the statin group and only 2 cases of AST more than 3 times the upper limit of normal in the control group. In the 10-year observational followup of this trial, transaminitis at this threshold was similarly rare (ALT: 1 case of >3 times elevation in the statin group; AST: 1 case of >3 each in the statin and control group). Abnormal creatine kinase of 10 times or greater the upper limit of normal was reported as zero in two trials, up to 4.5 percent in the statin groups, and up to 1.7 percent in the control groups. One trial’s 10-year observational followup reported no instances of elevated creatine kinase in participants on statins and in two non-FH siblings not taking statins. One fair-quality observational study evaluated the association of statins and new onset diabetes (n=9,393), showing no difference in new diabetes diagnoses over 9 years followup in those taking statins compared to controls. One fair-quality observational study (n=943) reported ALT more than 3 times the upper limit of normal with a frequency of 4.4 percent in the statin group and 1.5 percent in the control group over 3.5 years of observation. No significant differences between Tanner stages or other hormonal adverse events were reported in the RCTs or longer observational followup.
Harms in the non-statin trials were similar in the intervention and control groups; however, for bile acid and fibrate trials, the trials were generally small with few events. The diet and physical activity counseling intervention did not mention harms and three supplement trials in FH reported that there were no adverse events.
Multifactorial Dyslipidemia:
Direct Screening Benefits and Harms (KQ 1, 3): There were no randomized screening trials directly addressing the effectiveness and harms of screening for multifactorial dyslipidemia in children and adolescents.
Screening Yield (KQ 2): No studies performed a confirmatory lipid test; thus, evidence is limited to screen-positivity (prevalence) rather than diagnostic yield of lipid screening for identifying multifactorial dyslipidemia. We included five fair-quality studies (n=142,257) reporting prevalence of multifactorial dyslipidemia showing that lipid abnormalities are common, being generally more common for the parameters of HDL-C and TG. Prevalence ranged from 7.1 to 9.4 percent for elevated TC (≥200 mg/dL), 6.4 to 7.4 percent for elevated LDL-C (≥130 mg/dL), 12.1 to 22.2 percent for low HDL-C (<40 mg/dL), 8.0 to 17.3 percent for elevated TG (using various thresholds), and 6.4 to 13.0 percent for elevated non-HDL-C (≥145 mg/dL). Prevalence of any lipid abnormality in 6- to 19-year-olds was 19.2 percent based on NHANES data (2013-2016, n=4,381). Older age and higher BMI (≥95th percentile) were associated with higher prevalence of multifactorial dyslipidemia. Prevalence by sex was inconsistent across the cohorts and for different lipid measures.
Treatment Benefits (KQ4): We included four fair- to good-quality trials (n=1,008) examining the effectiveness of various lipid lowering treatments for multifactorial dyslipidemia. There were no included trials of drug interventions in child and adolescent populations with multifactorial dyslipidemia. We included two behavioral counseling trials (n=934), one fair-quality and one good-quality. These trials demonstrated statistically significant greater reductions in TC (3-6 mg/dL) and improvements in dietary intake outcomes in the intervention group compared to the control group in the short-term, but findings did not persist at longer follow-up.
We included two fair-quality supplement intervention trials (n=74) in populations with multifactorial dyslipidemia examining flaxseed and fish oil. These trials reported no statistically significant difference in TC or LDL-C, and flaxseed was associated with a statistically significant worsening of TG and HDL-C in the intervention group. There were no differences in BMI or total caloric intake.
Treatment Harms (KQ5): The two behavioral counseling trials in children with multifactorial dyslipidemia (n=934) reported no adverse effects in terms of growth and development, nutrient adequacy, and psychosocial outcomes in the dietary intervention group compared to the control group. The flaxseed trial (n=32) reported no adverse events; the fish oil trial (n=42) reported gastrointestinal symptoms, fishy taste, and frequent nose bleeds. Most trials reporting growth and development harms were limited by short duration.
Familial Hypercholesterolemia and Multifactorial Dyslipidemia:
Treatment Benefits (KQ4): We included seven fair- to good-quality supplement trials (n=288) which evaluated a wide range of supplement interventions in populations of children and adolescents with FH or multifactorial dyslipidemia. Only one trial, which evaluated the fiber glucomannan, showed a statistically significant improvement in TC, LDL-C and non-HDL-C (-10 to -11 mg/dL). Two other fiber trials, however, showed no statistically significant improvements in TC or LDL-C. One psyllium fiber trial showed a 60.2 mg/dL reduction in TG while other fiber trials showed no difference in TG. The trials of hempseed, probiotics, and hazelnuts showed no statistically significant reductions in any lipid parameter.
Treatment Harms (KQ5): Five of the seven supplement trials reported harms with two trials reporting no adverse events. The fiber trials reported various gastrointestinal side effects of 0 to 22.2 percent in intervention groups and 0 to 5.0 percent in control groups, and the probiotic trial reported three cases of abdominal pain (5.4% v 2.8%).
Limitations: No studies performed a confirmatory lipid or genetic test; thus, evidence is limited to screen-positivity (prevalence) rather than diagnostic yield of lipid screening for FH. FH diagnostic criteria were limited to lipid levels alone, which is inconsistent with treatment trial criteria which also included genetic, family, or clinical history components in addition to lipid levels. Treatment trials were generally small with relatively short followup, with most trial durations of less than 6 months. Only one statin trial had a followup as long as 2 years. With the exception of statins evaluated in the FH population, the bodies of evidence for any specific intervention in either the FH or multifactorial dyslipidemia population were extremely sparse, often consisting of just one to three studies. Behavioral counseling and supplement trials were generally small, with short-term followup leading to uncertainty regarding long-term adherence and benefit persistence. Outcomes for treatment trials were limited to intermediate outcomes with insufficient followup periods to assess long-term health effects or harms.
Conclusions: There is no direct evidence from population-based screening trials addressing the benefits and harms of pediatric lipid screening for intermediate, behavioral, or health outcomes. Dyslipidemia is common in pediatric populations with a prevalence of 19.2 percent for any lipid abnormality and heterozygous FH prevalence estimated at 0.2 to 0.4 percent (1:250 to 1:500). The body of evidence on treatment benefit is strongest for statins in FH children and adolescents based on mostly small, short-term studies with the longest trial of 2 years showing beneficial effects on TC and LDL-C. Most of the evidence for statin harms is from small, short-term studies and limited longer-term evidence showing few withdrawals due to adverse events, slightly higher rates of liver and musculoskeletal lab elevations, and no significant differences in Tanner staging or hormonal adverse events between statin and placebo groups. The trials of bile acids, fibrates, and PCSK-9 inhibitors in FH populations show reductions in one or more lipid parameters and are generally associated with low withdrawals due to adverse events. There is scant evidence on behavioral counseling interventions and supplements in populations with FH; two small plant sterol supplement trials show improvement in TC and LDL-C at 4-8 weeks. The body of evidence on treatment of multifactorial dyslipidemia is sparse, being limited to two short-term behavioral counseling interventions showing modest short-term benefits in lipid levels that did not persist with longer followup and two supplement studies of flaxseed and fish oil showing no benefit in lipid levels. Supplement trials recruiting both FH and multifactorial dyslipidemia populations show mixed results for fiber supplements. Fiber supplements were commonly associated with gastrointestinal side effects; otherwise, four of the seven supplement trials in populations with FH or multifactorial dyslipidemia reported no adverse events, no serious adverse events, or no AEs leading to withdrawals.
- Attribution
-
N/A
-
Authors of Report
-
Janelle M. Guirguis-Blake, MD
Corinne V. Evans, MPP
Erin L. Coppola, MPH
Nadia Redmond, MSPH
Leslie A. Perdue, MPH
-
Methodology description
-
Systematic Review
- PROSPERO
-
N/A
- DOI
-
N/A
- Notes
-
AHRQ Publication No. 22-05301-EF-1
- Funding Source
-
Agency for Healthcare Research and Quality