Introduction

Atherosclerosis and associated cardiovascular risk factors are known to develop in childhood and adolescence [1,2,3]. Many studies have shown that cardiovascular disease (CVD) risk factors in childhood, including dyslipidemia, progress into adult life and are associated with increased risk of premature CVD [4,5,6,7]. A large multi-cohort study evaluating long-term trajectories of cardiovascular health (defined by the American Heart Association (AHA) in 2010, based on body mass index (BMI), blood pressure (BP), cholesterol, glucose, diet, physical activity, smoking) showed that those with poor cardiovascular health in childhood experienced a more rapid progression in adulthood, and greater subclinical atherosclerosis [8]. Recently, a large prospective cohort study involving participants in the International Childhood and Cardiovascular Cohort consortium showed childhood CVD risk factors were also associated with fatal and nonfatal cardiovascular events in midlife [9•]. However, optimum management of risk factors in childhood has been shown to reduce risk of CVD in adulthood [10•]. Thus, it is essential to target cardiovascular risk factors, including dyslipidemia, in childhood with the goal of improving outcomes.

Screening and management guidelines for youth, as well as position statements, have previously been published with the intent of identifying and optimally managing cardiovascular risk factors at an early age with the goal of preventing or reducing the burden of atherosclerosis, the underlying cause of CVD in adulthood (see Table 1) [11,12,13,14,15,16,17,18,19,20, 21•]. However, lipid screening in youth has been underutilized, and abnormal lipid values are undertreated [22,23,24,25,26,27]. A survey of US pediatricians conducted by the American Academy of Pediatrics (AAP) in 2013–2014 showed that only 30% performed universal screening at age 9–11 years. In that survey, 62% of responders considered lack of comfort in prescribing statins to be a major barrier to treatment, while 27% reported that lack of access to lipid specialists was a major barrier [22]. Suboptimal rates of screening fail to identify children who would likely benefit from early intervention.

Table 1 Timeline of pediatric cholesterol guidelines, position statements, clinical reports
Full size table

Management algorithms have been created to provide tools for primary care providers and subspecialists to aid in management of dyslipidemia, as there is a lack of access to lipid specialists. Unique clinics targeting dyslipidemia in the pediatric population are often limited to large academic centers in medium- to large-sized cities [28]. We will review the currently available algorithms, discuss pitfalls, propose enhanced algorithms, and highlight future directions.

Definitions and Normative Values

Dyslipidemia is characterized by abnormal concentrations of lipids in the blood, which include elevated levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (non-HDL-C), triglycerides (TG), and low levels of high-density lipoprotein cholesterol (HDL-C), while the term hyperlipidemia generally refers to an increase in one or more lipid values. Hypercholesterolemia more specifically refers to high TC, high LDL-C, and/or high non-HDL-C. While the main focus has been to identify youth with elevated levels of LDL-C, studies have shown a predominant combined dyslipidemic pattern in children including moderately to severely elevated TG and non-HDL-C, reduced HDL-C, and minimally elevated or normal LDL-C, with increased small, dense LDL particles. Dyslipidemia places children at risk for early atherosclerotic disease [18, 29].

Normative lipid values in children are available and derived from population-based data from the Lipid Research Clinics Prevalence Study, a multicenter collaborative in the 1970s, which evaluated fasting lipoprotein profiles in over 13,000 children aged 0–19 years, and also from National Health and Nutrition Examination Survey III which evaluated over 7000 children aged 4–19 years from 1988–1994 [30,31,32]. Lipid levels vary by age and sex, along with race/ethnicity. Lipid levels initially increase from birth until 2 years of life, and then remain relatively stable until adolescence. During puberty, TC and LDL-C concentrations decrease initially, and then increase in the later teen years. Males experience a decrease in HDL-C where females have stable HDL-C until menopause. African American children tend to have higher TC and HDL-C and lower TG compared to other racial/ethnic groups [30,31,32].

Based on normative data, definitions for dyslipidemia in youth were initially published in 1992 from the National Cholesterol Expert Panel [33]. The updated recommendations from NHLBI in 2011 also incorporate data from the Bogalusa Heart study evaluating non-HDL levels, and these cutoffs matched the NCEP pediatric panel cutoffs for LDL-C (see Table 2) [18]. Of note, these abnormal definitions are based on population distributions, not health outcomes.

Table 2 Population-based acceptable, borderline-abnormal, and abnormal lipid concentrations in pediatrics
Full size table

Screening Guidelines

The goal of screening is early detection and intervention of children at risk of developing CVD. Earlier guidelines targeted youth with an informative family history, risk factors, and medical conditions associated with high risk of premature cardiovascular disease. However, 30–60% of children with dyslipidemia are missed if family history is the only criteria used [18]. In addition, population screening at younger ages is highly sensitive for detection of high-risk conditions such as familial hypercholesterolemia (FH). A meta-analysis showed that population screening in children 1–9 years of age resulted in a detection rate of 96% for infants and children with FH, with a false positive rate of 1% [34]. Thus, the National Heart, Lung, and Blood Institute (NHLBI) guidelines added universal screening at 9–11 years, and in those with normal initial screening, repeat testing at 17–21 years [18]. While the U.S. Preventative Services Task Force concluded that evidence was insufficient to either recommend or not recommend universal cholesterol screening in children, the NHLBI guidelines have been endorsed by many organizations including the AAP, the National Lipid Association (NLA), the AHA, and the American College of Cardiology (ACC) [18, 20, 35]. Thus, universal screening is widely recommended to detect and create the opportunity of early treatment in children with dyslipidemia.

Management Algorithms Based on Guidelines and Position Statements

The 1992 NHLBI guidelines for cardiovascular risk reduction were the first to address lipid screening, definitions, and management in the pediatric population. These recommendations were updated in 2011, by an expert panel assembled to critically review subsequent evidence targeting both prevention of risk factors and future CVD by optimally managing risk factors. The intended audience were primary care providers and dietitians [18]. The AHA released a scientific statement in 2006, stratifying management of CVD risk reduction based on high-risk conditions/diseases that are associated with accelerated atherosclerosis and early CVD, directed to pediatric care providers and subspecialists who manage complex primary disease processes. This position statement was last updated in 2019 [21•]. We will summarize the dyslipidemia management algorithms that were derived from these guidelines and position statements.

The 2011 NHLBI dyslipidemia algorithm approaches management based upon the presence of an abnormal fasting or non-fasting lipid panel [18]. In general, if lipids levels are confirmed to be abnormal, the panel recommended an initial trial of lifestyle changes; if lifestyle alone was unsuccessful in reaching the desired goal, then pharmacology should be considered. At very elevated levels (LDL-C greater than or equal to 250 mg/dL and/or TG greater than or equal to 500 mg/dL), a lipid specialist should be consulted. The guidelines include statins, as well as instructions as to how to initiate and monitor statin therapy. It also lists additional agents to consider under direction of a lipid specialist, such as bile acid sequestrants and fibric acid derivatives. However, the list of lipid-lowering agents is limited to the medications which were available in the year that it was published. Also of note, currently none of the pharmacotherapies for hypertriglyceridemia is U.S. Food and Drug Administration (FDA) approved for use in childhood.

The 2019 AHA algorithm begins with, and focuses on, patient risk stratification [21•]. Patients are stratified by disease condition as high-risk, moderate-risk, or at-risk. These updated risk categories are based upon the magnitude of the risk for atherosclerotic-related pathology compared to the general population. This is followed by global risk factor assessment. If 2 or more risk factors are present, then patients are reassigned to the next higher risk category. Once the risk category is established, then specific treatment goals are recommended. Depending on the risk category, there are different treatment thresholds, recommendations, and treatment goals. Treatment recommendations generally vary in terms of when to start pharmacotherapy in addition to lifestyle changes. Treatment recommendations for elevated LDL-C include statins, additional agents such as cholesterol absorption inhibitors, proprotein convertase subtilisin/kexin type 9 serine protease (PCSK9) agents, and LDL apheresis for homozygous FH (HoFH). Pharmacotherapy recommendations for elevated triglycerides include off label use of fenofibrate, omega-3 fatty acids, and statins if non-HDL (or apolipoprotein B) is elevated.

Pitfalls and Future Directions

Despite having management algorithms, dyslipidemia in children is still poorly recognized and inadequately treated. From 1999 to 2016, the proportion of youths aged 6–19 years with ideal concentrations of lipids and apolipoprotein B, for example, has been reported to be only 47–51%. Those with adverse levels of lipids or apolipoprotein B comprised 19–25% [36]. There are many possible reasons for this that have previously been extensively discussed in the literature; however, we will focus on potential pitfalls of the algorithms.

The Pediatric Endocrine Society (PES) sponsors several special interest groups to address issues concerning endocrine-related disease in children, including lipid disorders. A PES subcommittee (Lipid Special Interest Group) created several algorithms in an attempt to fill gaps which are not addressed by existing guidelines and algorithms. Examples of these are included here as condition-specific algorithms.

One pitfall is not only the complexity of the comprehensive management algorithms, but there are also many different organizations with varying recommendations, creating even more confusion for the treating clinician on optimal management for certain patient populations. We propose simplifying clinical management by creating condition-specific algorithms. For example, dyslipidemia is frequently seen in patients with type 1 or type 2 diabetes and data show dyslipidemia in these populations is undertreated and undermanaged [37,38,39,40]. Therefore, we have created a diabetes-specific lipid algorithm incorporating recommendations from the AHA position statement, NHLBI guidelines, and American Diabetes Association (ADA) standards of care [18, 21•, 41, 42]. This includes starting statin pharmacotherapy simultaneously with lifestyle changes for LDL ≥ 130 mg/dL [21•]. We have not included an age-specific limit on when to start statin therapy, a departure from the ADA standards of care, due to availability of FDA-approved statins starting at age 7 years for HoFH and age 8 years for HeFH [43]. We also recommend initiation of pharmacotherapy earlier for TG > 400 mg/dL or TG ≥ 150 and non-HDL ≥ 145, with omega-3 fatty acids and considering fibrates, consistent with AHA recommendations, which differs slightly from the NHLBI guidelines and ADA recommendations. We have also incorporated non-HDL-C goals into this algorithm as it is a significant predictor of presence of atherosclerosis and persistent dyslipidemia and is accurate in a non-fasting state [18]. However, the primary endpoints are still LDL-C or TG levels, as most studies of medications and reducing cardiovascular risk have targeted primarily LDL-C and/or TG levels. In general, our algorithm provides more clinically relevant guidance than the AHA recommendations, for the clinician who manages this defined population. We see potential for creation of additional algorithms in the setting of other high-risk conditions (such as chronic kidney disease, Kawasaki disease, childhood cancer survivor, solid organ transplant) that may be useful to other pediatric subspecialists. In general, there is a need for simplification of algorithms to help with feasibility of guideline adherence (Fig. 1).

Fig. 1
figure 1figure 1

Diabetes and lipids algorithm. This algorithm was developed by PES members to address dyslipidemia in type 1 and type 2 diabetes mellitus patients, based on the AHA position statement, NHLBI guidelines, and ADA standards of care. This algorithm includes a second page with instructions for starting statins, including recommendations for baseline labs, and counseling on side effects and contraindications, based on NHLBI guidelines. There is also a dose range table with anticipated lipid-lowering effects, incorporated from Jones et al. 1998. There are also general precautions about starting fibrate therapy, including counseling on side effects, drug interactions, and contraindications [18, 21•, 41, 42, 53]. *Side effects: muscle pain/cramps, weakness, myopathy. †Potential medication interactions: cyclosporine, niacin, fibric acid derivative, erythromycin, azole antifungal, HIV protease inhibitor. ‡Contraindicated in pregnancy, known teratogen. For female patients consider contraception or gynecology referral. Also contraindicated in nursing mothers, acute liver disease, hypersensitivity to statin. §Currently statins are approved for children starting age 7 years for HoFH and age 8 years for HeFH. Toxicity = AST, ALT > 3 × ULN, or CK > 10 × ULN (check CK if concern for myopathy). #2 × dose = ↓ 6% LDL-C. αOf note, TG lowering medications in adults with diabetes have shown inconsistent results with improved CVD benefit [54]. βSide effects: muscle toxicity, hepatotoxicity. If on fibrate and statin, higher risk of muscle toxicity. γDrug interactions: coumarin anticoagulants, immunosuppressants, bile acid resins. δContraindications: severe renal dysfunction, acute liver disease, gall bladder disease, hypersensitivity to fibrate, nursing mothers. T1D type 1 diabetes mellitus, T2D type 2 diabetes mellitus, LDL-C low-density lipoprotein cholesterol, non-HDL non-high-density lipoprotein, TG triglyceride, y years of age, CHILD-1 Cardiovascular Health Integrated Lifestyle-1, CHILD-2 Cardiovascular Health Integrated Lifestyle-2, ALT alanine transaminase, AST aspartate aminotransferase, CK creatinine kinase, ULN upper limit of normal, HoFH homozygous familial hypercholesterolemia, HeFH heterozygous familial hypercholesterolemia, EPA eicosapentaenoic acid, DHA docosahexaenoic acid

Full size image

Another potential pitfall is the lack of consistent implementation of the algorithms in clinical practice, again perhaps from the complexity of algorithms. It has been reported that establishing clinical effectiveness of a clinical innovation is not enough to guarantee uptake into routine use and it may take at least 17–20 years to implement into clinical practice, with fewer than 50% of clinical innovations making it into general usage [44, 45]. Perhaps we need more research focusing on implementation science in this field.

A third potential pitfall of the algorithms is that they may not address the complexity of treating rare lipid disorders. There are no current algorithms for pediatric patients that include additional recommendations if first-line therapy is insufficient to achieve treatment goals, or if a patient exhibits statin intolerance. Newer pharmacotherapy options have been approved for use in children and adolescents in recent years. Thus, we propose a “Beyond Statin” supplemental algorithm to close this information gap, which includes new medications including proprotein convertase subtilisin/kexin type 9 serine protease (PCSK9) inhibitors, and angiopoietin-like 3 (ANGPTL3) inhibitors (see Fig. 2) [46, 47]. This algorithm addresses the gap of treating rare lipid disorders, such as HoFH, but also encompasses other high-risk conditions when statins are insufficient (Fig. 2).

Fig. 2
figure 2

Beyond statins algorithm. This algorithm was developed by PES members to address management of hypercholesterolemia in the event that statins are insufficient, or a patient is intolerant to statins. This algorithm includes medications for HoFH that have been very recently approved by the FDA. The decision branch points for therapeutic options are based on age and available information on genetic mutations, since PCSK9 inhibitor response depe-ds on LDL receptor function. The PES algorithm for members includes a table summarizing anticipated LDL-C lowering effects and drug safety information [46, 47, 55]. *Genetic testing is currently not an exclusive criteria for defining familial hypercholesterolemia, and not all patients with clinical FH have identifiable mutations. Thus, the decision branch encompasses patients whose genetics are unknown. †In severe cases, initial LDL-C goal may be ≥ 50% reduction from baseline, and clinical judgment is required to evaluate responsiveness. ‡At this time Inclisiran is only FDA-approved for age 18 years and older. §HoFH is clinically diagnosed when a person has an untreated LDL-C level > 500 mg/dL, with either the presence of cutaneous or tendinous xanthomas before the age of 10 years or documentation of untreated LDL-C levels of > 250 mg/dL in both parents [56]. LDL-C low-density lipoprotein cholesterol, LDLR LDL receptor, PCSK 9 proprotein convertase subtilisin/kexin type 9 serine protease, HoFH homozygous familial hypercholesterolemia, ANGPTL3 angiopoietin-like 3, MTP microsomal triglyceride transfer protein

Full size image

A fourth potential pitfall to the algorithms, specifically with the 2011 NHLBI guidelines, is that data is no longer up to date. However, the bigger issue may be that there is no single organization that speaks to the issues of management of pediatric dyslipidemia and/or timely updates. To achieve the goal of improving outcomes in youth with dyslipidemia, there needs to be more collaboration among organizations including the National Lipid Association, American Heart Association, Pediatric Endocrine Society, and American Academy of Pediatrics. The latest NHLBI algorithm was published in 2011 and lacks updated evidence showing even more support for early intervention for hypercholesterolemia in children. There is now more data showing that statin therapy is effective and safe among pediatric patients. Meta-analyses have shown effectiveness of statin therapy in children with FH without adverse side effects [48, 49]. A recent large longitudinal study of pediatric patients with FH who started statin therapy and were followed for 20 years showed slowed progression of carotid media intima thickness and reduced risk of CV disease in adulthood [10•]. Another study in children with type 1 diabetes showed that statin therapy for 2–4 years was effective in improving lipid profiles without adverse side effects [50]. The AHA position statement takes the recent data into account; however, it does not provide detailed framework of drug initiation and monitoring. We have adapted the algorithm in the AHA position statement to create easy-to-follow management algorithms (see Figs. 3 and 4) that include pharmacotherapy initiation and monitoring, adapted from the NHLBI guidelines (see Fig. 1) [21•]. The AHA position statement recommends simultaneously initiation of pharmacotherapy with lifestyle in the high-risk group, as there is data supporting the fact that higher risk patients such as those with FH do benefit from early treatment [10•].

Fig. 3
figure 3

LDL-C management in high-risk pediatric patients. This algorithm was developed by PES members and includes the basic management strategy for elevated LDL-C, adapted from the AHA position statement. It also includes a second page with instructions for starting statins, including recommendations for baseline labs, and counseling on side effects and contraindications, and a dose range table with anticipated lipid-lowering effects, same as previous figure (see Fig. 1) [18, 21•, 53]. *See reference for more information on CHILD-2 LDL lowering diet [57]. LDL-C low-density lipoprotein cholesterol, HoFH homozygous familial hypercholesterolemia, T2D type 2 diabetes mellitus, T1D type 1 diabetes mellitus, ESRD end-stage renal disease, BMI body mass index, HeFH heterozygous familial hypercholesterolemia, Lp(a) lipoprotein (a), NAFLD nonalcoholic fatty liver disease, PCOS polycystic ovary syndrome, JIA juvenile idiopathic arthritis, SLE systemic lupus erythematosus, IBD inflammatory bowel disease, HIV human immunodeficiency virus, CAD coronary artery disease, M male, F female, BP blood pressure, h hours, non-HDL non-high-density lipoprotein, TC total cholesterol, HDL-C high-density lipoprotein, CHILD-2 LDL Cardiovascular Health Integrated Lifestyle-2 low-density lipoprotein diet

Full size image
Fig. 4
figure 4

TG management in high-risk pediatric patients. This algorithm was developed by PES members and applies the risk stratification approach utilized in the AHA position statement [21•] to the management of hypertriglyceridemia. The narrative and tabular information regarding TG evaluation is integrated into the algorithm along with the composition of a recommended diet (CHILD-2 TG), and the pharmacotherapy options. *See reference for more information on CHILD-2 TG lowering diet [57]. TG triglyceride, CHILD-2 TG Cardiovascular Health Integrated Lifestyle-2 triglyceride, EPA eicosapentaenoic acid, DHA docosahexaenoic acid, non-HDL non-high-density lipoprotein, ApoB apolipoprotein B-100, HDL-C high-density lipoprotein cholesterol, HoFH homozygous familial hypercholesterolemia, T2DM type 2 diabetes mellitus, T1DM type 1 diabetes mellitus, ESRD end-stage renal disease, BMI body mass index, HeFH heterozygous familial hypercholesterolemia, Lp(a) lipoprotein (a), CKD chronic kidney disease, NAFLD nonalcoholic fatty liver disease, PCOS polycystic ovary syndrome, JIA juvenile idiopathic arthritis, SLE systemic lupus erythematosus, IBD inflammatory bowel disease, HIV human immunodeficiency virus, CAD coronary artery disease, M male, F female, BP blood pressure

Full size image

Lastly, there is a lack of outcome studies in the pediatric literature, in comparison to adults, which hampers our ability to predict whether pediatric algorithms are safe and effective. Therefore, precise risk assessment tools are not available for pediatric care. Even our definitions of dyslipidemia and treatment thresholds in youth are generally “borrowed” from adult data, and based on normative data, not outcomes-based data. Given the nature of preventive care for children, and the long interval prior to a CVD-related event, other approaches are needed such as prospective longitudinal and indirect studies of risk. Unlike adult algorithms that are based on risk scores generated from risk of atherosclerotic CVD-related morbidity and mortality, pediatric algorithms are not based on expected risk of atherosclerotic CVD [18, 51]. Perhaps this nebulous concept of atherosclerotic CVD risk has prevented wider acceptance and implementation of these guidelines and algorithms. We need to create an improved system for developing a more specific risk score for patients to individualize treatment.

There are a few studies evaluating proposed risk scores in youth. The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) risk score was developed from initial observations of well-defined risk factors, measured postmortem in young adults, which correlated to the amount of advanced atherosclerosis in coronary arteries and abdominal aorta in women and men dying accidentally. This risk score was applied to a longitudinal cohort, Cardiovascular Risk Development in Young Adults (CARDIA), who were young adults, age 18–30 years, and followed for 30 years. At 15 years of follow-up, they found that the PDAY score was highly predictive of atherosclerotic cardiovascular disease events later in life [52]. It remains to be seen if this can be applied to the younger population. Recently, a large international prospective study over 35 years analyzed five risk factors including BMI, systolic BP, total cholesterol level, TG level, and youth smoking, with the use of age- and sex-specific z-scores, and with a combined-risk z-score that was calculated as the unweighted mean of the five risk z-scores [9•]. A comparable adult combined-risk z-score was analyzed jointly with the childhood risk factors. It was found that the childhood combined-risk z-score and the change in the combined-risk z-score between childhood and adulthood were associated with CV events in midlife. This was the first study showing childhood risk factors were associated with clinical events in adulthood. It remains unclear if this type of risk score can be generalized for wider use. Nevertheless, these studies are promising.

Conclusions

Despite guideline and statement-based algorithms for screening and management of dyslipidemia in youth, there are still low rates of screening and treatment. Algorithms have been created to facilitate initiation of management among youth by general pediatricians, pediatric endocrinologists, and preventative cardiologists. We propose a few enhanced algorithms based upon expert opinion to address potential gaps and pitfalls to the existing algorithms, including a diabetes-specific and beyond statin algorithm, to add more resources for clinicians. We also acknowledge the need for further risk assessment tools in youth for more tailored dyslipidemia management.