Chapter 9: Obesity, Metabolic Syndrome, and Dyslipidemia

Obesity is an increasingly common condition in childhood and adolescence, and it has emerged as one of the most serious public health concerns in the 21st century. The growing prevalence of childhood obesity has brought with it an increase in the incidence of obesity-related comorbid disease prior to adulthood. The profound personal and economic impacts of childhood obesity require a thorough examination of this complex topic leading to serious efforts toward prevention and early intervention.

Definitions of Overweight and Obesity in Children

Discussion of this topic must begin with a consensus as to how overweight and obesity are to be defined in the pediatric population. Multiple different indices and techniques can be used to estimate the degree of adiposity. The most commonly used measure is body mass index (BMI), defined as kilograms (kg) of body weight per height in square meters (m²). Adults are defined as overweight if their BMI is 25.0 to 29.9 kg/m² and obese if it is greater than 30.0 kg/m². This index of adiposity, however, is less accurate in children that have not yet achieved full adult height. The number of children defined as obese using the adult criteria would be greatly underestimated. Therefore, the BMI percentile is a more accurate index of body mass in the pediatric age group. BMI percentile curves have been generated using data from the 2000 National Health and Nutrition Examination Survey (NHANES) and take into account both age and gender (Figures 9-1A and 9-1B). The Centers for Disease Control and Prevention (CDC) define children as overweight if the BMI is greater than 95th percentile and, at risk for overweight, if it is between the 85th and 95th percentile. The BMI percentile graphs only include children older than 2 years. For children younger than 2 years, standard weight-for-length curves can be used to assess body mass.

Prevalence of Overweight and Obesity in Children

Multiple national surveys have indicated a significant increase in the frequency of overweight children over the past 30 years. With current prevalence of overweight and obese children ranging from 12% to 30% in developed nations and from 2% to 12% in the developing world, overnutrition is now one of the most common health problems in childhood.¹ The most recent NHANES data indicate that 31.8% of American children and adolescents aged 2 to 19 years have a BMI more than 85th percentile, with 16.9% having a BMI greater than 95th percentile.² Although there was a significant increase in the prevalence of obesity in the 1980s and 1990s, significant increases were not seen in the latest NHANES data set (2007-2008) except in the highest BMI cohort (97th percentile) of 6- through 19-year-old males. No change has been seen in the prevalence of obesity in females since the 1999-2000 NHANES study. The risk of overweight and obesity is not evenly distributed, with low-income groups and minorities most affected in industrialized societies. In the United States, for example, childhood obesity rates are 24% among African American and 21% among Hispanic populations versus 14% among Caucasians.²

Prevalence of Obesity-Related Disease

Increased adiposity can lead to both immediate complications in children and adolescents and long-term health consequences as adults. Insulin resistance frequently accompanies obesity, specifically abdominal obesity, reflecting increased visceral and hepatic fat deposition. Insulin resistance is directly related to the degree of adiposity and is further influenced by underlying genetic predisposition and ethnicity. Many of the consequences of obesity, including impaired glucose tolerance (IGT) and type 2 diabetes mellitus (T2DM), dyslipidemia, hypertension, polycystic ovarian syndrome (PCOS), and acanthosis nigricans, are related to insulin resistance. Other consequences of obesity not related to insulin resistance include pulmonary, gastrointestinal, and orthopedic complications. In addition, obese children often suffer from psychosocial complications including depression and low self-esteem.

Impaired glucose tolerance and type 2 diabetes mellitus

The prevalence of T2DM is increasing in parallel with the epidemic of childhood obesity. Previously considered an adult disease, the SEARCH for Diabetes in Youth Study Group reported an incidence rate of 11.8 per 100,000 persons per year in the 15- to 19-year age group with the majority of these new cases occurring in African American, Native American, and Hispanic ethnic groups.³ The prevalence of prediabetes in obese pediatric populations is also on the rise. Prediabetes is defined by the American Diabetes Association as either impaired glucose tolerance (IGT), defined as a fasting glucose greater than 100 mg/dL, a 2-hour postoral glucose load level between 140 and 200 mg/dL, or a hemoglobin A1c between 5.7% and 6.5%. Based on NHANES data collected in their 2005-2006 report, approximately 30% of obese adolescents have elevated fasting glucose with 8% exhibiting IGT.⁴

Hypertension

The relationship between hypertension and obesity is well established in adults. Population, meta-analyses, and screening studies are demonstrating similar relationships in children.⁵ The rate of hypertension in school-aged children is as high as 10% among children with BMI greater than 97th percentile.⁶ Although the mechanisms of hypertension in obese children are not completely understood, increased circulating volume and cardiac output have been shown to be important factors. Additional contributions from insulin resistance, sympathetic nervous system activity, inflammation, the renin-angiotensin-aldosterone system, and structural changes in the kidney and cardiovascular system have also been proposed.⁷

Dyslipidemia

Dyslipidemia is an emerging comorbidity of childhood obesity. Unlike familial hypercholesterolemia (FH), which is characterized by an elevated low-density lipoprotein cholesterol (LDL-C) and a family history of hypercholesterolemia, the pattern of dyslipidemia associated with pediatric obesity consists of a combination of hypertriglyceridemia (TG), decreased high-density lipoprotein cholesterol (HDL-C), and normal to mildly elevated LDL-C.⁸ NHANES data indicate that this pattern is highly prevalent (42.9%) in children with BMI greater than 95th percentile.⁹ The nomenclature, secondary atherogenic dyslipidemia, is applied when hypertriglyceridemia, low HDL-C, and normal to elevated LDL-C are found in conjunction with components of the metabolic syndrome.¹⁰ In normal-weight children, the same lipid profile is called familial combined dyslipidemia.

The association of dyslipidemia with endothelial dysfunction and accelerated formation of atherosclerotic lesions is well established. Hypertriglyceridemia in adults has been identified as an independent risk factor for coronary artery disease, and is also associated with increased risk of acute pancreatitis.¹¹

Polycystic ovary syndrome

Adolescent females with obesity often present with signs and symptoms consistent with hyperandrogenism and ovarian dysfunction, including amenorrhea or oligomenorrhea, acne, and hirsutism. Hyperandrogenism appears to primarily result from hyperinsulinemia. PCOS affects approximately 5% of adult women, primarily those who are overweight or obese, with most women first becoming symptomatic during adolescence (see Chapter 3 for more information on PCOS).

Other obesity-related complications

Gastrointestinal disease, including fatty infiltration of the liver and cholelithiasis, may occur in overweight children and correlates with the severity of adiposity. In studies of obese children, 10% to 25% have been found to have elevation of liver enzymes and approximately 50% to 75% have evidence of fatty infiltration of the liver on ultrasound,¹² which can progress to liver fibrosis and cirrhosis. Nonalcoholic fatty liver disease (NAFLD) is now the third most common reason for liver transplant accounting for one in every eight liver transplants.¹³ The duration of obesity, but not its severity, is correlated with the extent of fibrosis. In addition, NAFLD is much more common in obese Hispanic adolescents than in Caucasians. Obese African American adolescents have the lowest prevalence of NAFLD. Furthermore, familial aggregation studies, heritability studies, and genome-wide scans suggest that there is a significant genetic underpinning for the risk of developing NAFLD in association with obesity.¹⁴

Obesity is also a risk factor for cholelithiasis and gallbladder disease. Although these are collectively rare in children, almost 50% of cases of cholecystitis in adolescents are associated with obesity.

Pulmonary disease, including both asthma and obstructive sleep apnea, is common in overweight children. The incidence of sleep apnea in obese children is four to five times that seen in nonobese children.¹⁵ Severe sleep apnea may ultimately lead to cor pulmonale.

Obesity is a risk factor for the development of orthopedic disorders, including both Blount disease and slipped capital femoral epiphysis (SCFE). Blount disease, also termed tibia vara, occurs when the inner part of the tibia just below the knee fails to develop normally, causing angulation of the tibia. Unlike benign tibia vara that tends to correct with age, Blount disease is progressive and can cause severe bowing of one or both legs. Adolescent-onset Blount disease affects African American males almost exclusively, with a prevalence of 2.5% among the obese adolescent black male population.¹⁶

SCFE occurs when the epiphysis of the proximal femur begins to slip off the corresponding metaphysis. SCFE presents as a limp combined with hip pain and/or referred pain to the knee. The affected limb appears to be externally rotated on examination, and there is pain with passive manipulation. Data from a number of studies indicate that 51% to 72% of all SCFE cases occur in obese individuals.¹⁶

Psychosocial effects of obesity

The most prevalent consequence of childhood obesity may be the stigmatization experienced by the obese child. Teasing is common by both peers and family members, and even health care professionals may display negative biases in their interactions with obese children and adolescents. Multiple studies have shown poor body image, low self-esteem, depression, and even suicidal ideation in children and adolescents teased about their weight.¹⁷ Severely overweight children have demonstrated health-related quality-of-life scores comparable to children with cancer.¹⁸ The social and economic impacts of childhood obesity may be far-reaching, with at least one study showing that women with a history of overweight in adolescence ultimately completed less schooling, were less likely to be married, had lower household incomes, and had higher rates of poverty than matched controls who had not been overweight.¹⁹

Economic impact

The growing economic impact of childhood obesity is staggering. In youth aged 6 to 17 years, financial costs related to treatment of obesity-associated diseases more than tripled from a cost-of-living adjustment of $35 million in 1979-1981 to $127 million in 1997-1999. A recent study estimated that approximately 21% of annual US healthcare spending in 2010 (or $190.2 billion) was related to costs of treatment for overweight, obesity, and their sequelae in both children and adults, with 14 billion attributed to childhood obesity alone.²⁰

Progression of Childhood Obesity Into Adulthood

Obese children tend to become obese adults. Those at highest risk are adolescents in the highest weight categories who come from families with at least one obese parent.^{1, 21} The probability of adult obesity (BMI > 30 kg/m²) is more than 50% among children older than 13 years whose BMI percentiles meet or exceed the 95th percentile for age and gender.²¹

Pathogenesis

Genetic and environmental components

Obesity is clearly a “genetic” disorder that is heavily influenced by environmental factors. Twin studies indicate that 50% to 80% of the tendency to gain weight is genetically determined.²² The adiposity of a child’s parents also has a profound effect on that of the child.²³ Furthermore, data from the Bogalusa heart study indicate that the increase in BMI observed in children over the past few decades is not evenly distributed among the population.²⁴ When age-, sex-, and race-adjusted data collected from children in the 1970s are compared to a cohort of children in the 1990s, the adjusted BMI that defines the lower percentiles (ie, the thinnest individuals) changes very little. However, there is a significant further increase in adiposity in individuals already at higher body mass in BMI percentiles (Figure 9-2). Data from the National Health and Examination Survey (NHES) and NHANES III studies demonstrate a similar phenomenon. In summary, there has been little change over the years in adiposity in those individuals genetically predisposed to thinness. Adverse environmental changes in our society appear to have the greatest impact on individuals with a genetic predisposition to obesity.

There is no question that changes in the environment have had a profound effect on the prevalence of overweight children. These changes have adversely impacted energy balance, particularly in the genetically susceptible individual. The typical modern Western diet is characterized by high-fat, calorie-dense foods with larger portions. Physical activity has declined in parallel with the increase in the incidence of overweight children.²⁵ Television, computers, and video games have changed how children entertain themselves, with more time spent in sedentary activities. The resulting decrease in caloric expenditure combined with an increase in caloric intake is having profound consequences.

The “thrifty” gene hypothesis

Genetic traits predisposing to obesity likely evolved from genetic modifications that provided a selective advantage for survival. Individuals with a genetic predisposition toward obesity preferentially shunt consumed calories toward storage in adipose tissue and further protect energy stores during starvation by efficiently decreasing energy expenditure.²⁶ Food-seeking behaviors are also more pronounced. In other words, individuals susceptible to obesity have a “thrifty” phenotype that optimizes utilization of energy resources, a selective advantage during famine.

In contrast to those with a “thrifty” phenotype, thin individuals tend to be less efficient with caloric resources, that is, have a “wasteful” phenotype that tends to store less energy as adipose tissue and have less intense food-seeking behaviors. Because genes that comprise the “thrifty” phenotype confer selective advantage during times of famine, these genes would become more prevalent in populations subjected to starvation. There are many historical examples of enrichment of “thrifty” genes that exist in populations.²⁷ Unfortunately, the emergence of a Western diet and sedentary lifestyle has resulted in profound increases of obesity-related complications in these metabolically efficient, thrifty ethnic populations.

Weight set point and regulation of energy homeostasis

Physiologic mechanisms exist that tightly regulate food intake and energy expenditure, even in most overweight and obese individuals. The existence of a tightly regulated feedback system can be demonstrated in long-term clinical studies of weight loss. Weight loss induced by very-low-calorie diets and/or behavior modification is rarely maintained over time. Most individuals tend to gain weight back to a point at or near the prestudy weight. A similar phenomenon can be demonstrated in individuals who are overfed and forced to gain weight.²⁸ These subjects gain adipose mass, but they are generally unable to maintain this degree of adiposity when the study is stopped. They rapidly lose weight and, as a group, end up near their prestudy weight. These observations imply that most individuals have a genetically determined “weight set point.”

The weight set point is maintained in most individuals by adjustments to metabolic rate in response to changes in body mass. Total energy requirements to maintain weight stability in both non–weight-reduced obese and nonobese subjects are similar. Obese individuals who have undergone weight loss, however, require 10% to 15% fewer calories to maintain weight stability than a “non–weight-reduced” person of similar body composition and weight.²⁶ Further weight loss results in further decline in total energy expenditure in an attempt by the individual to protect adipose stores. A sustained decrease in caloric intake in overweight subjects eventually results in a new steady-state body weight, rather than continued weight loss, as the decline in daily energy expenditure equilibrates with the decline in energy consumption (Figure 9-3A). Furthermore, the diet-reduced adipose mass leads ultimately to an increased appetite and difficulty in maintaining the lower caloric intake. A resumption of the “prediet” caloric intake will almost certainly result in a return to the previous obese body weight set point.

A tightly regulated feedback system that controls body weight is absolutely necessary to keep an individual’s weight relatively stable over time. Thin and normal-weight individuals have feedback systems that regulate appetite and metabolic rate tightly in response to changes in adiposity (Figure 9-3B and C). Feedback systems in individuals genetically predisposed to obesity are not as tightly regulated allowing for greater increases in body mass. If these homeostatic mechanisms did not exist, small, but persistent, increases in caloric intake would result in significant obesity. It has been estimated that approximately 1 lb of adipose tissue is added for every 3500 kcal consumed in excess of caloric requirement for energy expenditure and growth. Assuming no compensatory increase in metabolic rate with increased adipose mass, as little as 50 extra kcal/day for 10 years would result in a 50-lb gain in adipose mass. A caloric intake of 200 kcal/day in excess of caloric expenditure would result in a 100-lb weight gain in as little as 5 years (Figure 9-4). Fortunately, most individuals defend “ideal weight” by alteration of both energy expenditure and food intake. Therefore, the weight of most children tends to follow standard published growth curves. However, if the homeostatic mechanisms are not intact, dramatic weight gain can occur.

The neuroendocrinology of weight regulation

Energy balance and weight set point regulation are accomplished by a complex feedback system. Adequacy of energy stored in the adipocyte is communicated centrally by the adipocyte-derived hormone, leptin. Secreted from adipocytes in proportion to the degree of adiposity, leptin acts at the hypothalamus to inhibit feeding behavior, decrease insulin secretion, and increase metabolic rate.²⁹ As an animal gains weight, leptin levels rise resulting in decreased appetite and increased metabolic rate. Likewise, weight loss leads to a decline in leptin levels and an increase in food-seeking behavior with a decrease in basal metabolic rate.

Leptin signaling in the hypothalamus is primarily via receptors located in the arcuate nucleus. The arcuate nucleus lacks a blood-brain barrier allowing direct communication of peripheral “metabolic” signals including both nutrients and hormones. Leptin signals centrally by inhibiting neurons that contain neuropeptides (neuropeptide Y [NPY] and agouti-related peptide [AgRP]) that stimulate appetite/decrease metabolic rate, while stimulating neurons that contain neuropeptides (eg, MSH derived from proopiomelanocortin [POMC]) that inhibit appetite/increase metabolic rate through binding of the MC4-R receptor (Figure 9-5A).

Peripheral signals other than leptin also appear to play a role in appetite regulation and energy homeostasis. Ghrelin, an orexigenic hormone secreted by the stomach, stimulates NPY/AgRP neurons centrally.³⁰ Ghrelin levels rise in the fasted state and decrease with feeding. Another gastrointestinal peptide, peptide YY (PYY), is secreted from the intestine.³¹ In contrast to ghrelin, PYY appears to be upregulated with feeding. PYY inhibits NPY/AgRP neurons while stimulating POMC neurons. It is theorized that, while leptin serves as a more general indicator of energy stores, ghrelin and PYY modulate leptin signaling to provide short-term feedback on caloric intake (Figure 9-5B). Insulin also appears to signal satiety via central nervous system (CNS) mechanisms similar to leptin. Therefore, postprandial increases in insulin in response to a carbohydrate meal may also act to decrease further food intake.

This complex interplay between peripheral signals of “energy status” and central control of feeding and metabolism must be intact to maintain normal weight homeostasis. If it is loosely regulated or interrupted by genetic defects or hypothalamic injury, energy imbalance and obesity can result.

Brown fat and nonshivering thermogenesis

The type of fat present may also play a role in determining body mass and obesity. There is increasing evidence to support the existence of brown fat in adult humans.³² This mitochondria-rich fat cell was once thought to exist only in newborns but has now been found in diffuse locations within white adipose tissue in adults. White adipose tissue functions as an energy storage depot while brown adipose tissue has a role in the generation of heat from existing energy stores. Activation and uncoupling of the mitochondria in brown adipose tissue result in the generation of heat in response to mild-to-moderately decreased ambient temperatures, a phenomenon termed nonshivering thermogenesis. A well-known trait in other mammals and in newborns, it has recently been established that nonshivering thermogenesis also occurs in many adult humans. Although similar in function, it is likely that the brown adipose tissue discovered in adult humans differs somewhat from that found in other animals including rodents. Classic brown fat found in rodents derives from a common lineage with skeletal muscle. However, the brown fat found in adult humans may derive from cells in white adipose depots that take on a brown adipose cell-like appearance after cold stimulation. In addition, these cells, sometimes referred to as “beige adipose tissue,” are regulated differently than are classic brown adipose cells. Furthermore, this beige adipose tissue seems to increase in white adipose tissue when exposed to the hormone irisin, a peptide secreted from muscle in response to exercise. These discoveries have led to significant research into methods to safely recruit brown and beige adipose tissue in obese individuals as a potential therapeutic approach to obesity treatment.

White adipose tissue as an endocrine organ

White adipose tissue was long thought simply to be an energy storage depot. The discovery of leptin, as detailed previously, led to a new understanding of white adipose tissue as an important mediator of a number of metabolic functions. Since this discovery, the list of hormones and cytokines that are released from white adipose tissue, often referred to as “adipokines,” grows longer every year (Table 9-1).³³ Collectively, these hormones and inflammatory cytokines mediate many of the metabolic and cardiovascular effects associated with obesity. For example, adiponectin, omentin, apelin, and chemerin may be cardioprotective and inhibit inflammation and endothelial dysfunction while leptin, resistin, and visfatin may contribute to obesity-mediated atheroschlerosis and inflammation. It is very likely that the role of these hormones differs in normal-weight and thin individuals compared to obese individuals. For instance, adiponectin levels are higher in normal-weight subjects and considerably lower in obese subjects, making the cardioprotective role of adiponectin somewhat less relevant in obesity than it is in normal-weight individuals. The relative imbalance of these adipocyte-derived hormones and cytokines in obese states likely contributes to the higher incidence of cardiovascular disease and abnormalities of glucose tolerance seen in obese subjects.

Definition of Pediatric Metabolic Syndrome

The metabolic syndrome is defined as a group of factors related to insulin resistance, which predict increased risk for T2DM and cardiovascular disease. The metabolic syndrome definition was initially developed to describe a combination of abdominal obesity, glucose intolerance, hypertension, and dyslipidemia in adults. However, the definition has recently been adapted to include adolescents^{34, 35} (Table 9-2). The adult-specific criteria were adapted for adolescents to include age- and gender-specific criteria. Although presence of the metabolic syndrome in adults predicts risk for T2DM and cardiovascular disease, the same future risk of cardiovascular disease cannot necessarily be extended to adolescents. Long-term prospective studies to demonstrate a direct connection between the presence of metabolic syndrome in adolescents and cardiovascular disease as adults have yet to be designed. However, adolescents with metabolic syndrome tend to become adults with metabolic syndrome. Therefore, until conclusive studies are available, metabolic syndrome-related risk of future cardiovascular disease is inferred and may warrant early aggressive interventions to increase insulin sensitivity in adolescents with metabolic syndrome.

Prevalence of Pediatric Metabolic Syndrome

Estimates of the prevalence of the metabolic syndrome in adolescents depend on the definition used (see Table 9-2). The International Diabetes Federation (IDF) criteria require an increase in waist circumference, emphasizing the role of abdominal adiposity as a predisposing and presumed necessary factor in the pathogenesis of the metabolic syndrome. The Modified Adult Treatment Panel (ATP III) criteria treat waist circumference as an independent and not absolutely necessary factor, implying other factors may predispose an individual to developing the metabolic syndrome independently of significantly increased abdominal fat. It is difficult to directly compare data generated using the IDF criteria with the results from ATP III given their significantly different definitions of metabolic syndrome. Presently, there is no consensus on which definition best reflects the pathophysiology of the metabolic syndrome.

An analysis of data from 2001-2006 NHANES found that 8.6% of US adolescents from 12 to 19 years of age met the ATP III criteria for metabolic syndrome.³⁶ This included 10.8% males and 6.1% females. More Hispanic adolescents (11.2%) had the metabolic syndrome than did Caucasians (8.9%) or African Americans (4.0%). The low prevalence among African American adolescents would be expected to translate to a lower risk of cardiovascular disease in black adults compared to white adults. However, multiple studies indicate that the prevalence of metabolic syndrome-associated morbidities is higher in black adults than in whites further emphasizing the need to use caution in making a direct association with pediatric metabolic syndrome and adult disease.

Pathogenesis and theRole of Adipokines

In adolescents and some children, obesity is often associated with features of the metabolic syndrome, including dyslipidemia, abnormal glucose homeostasis, and hypertension (see Table 9-2). However, body fat distribution is likely to be the most important risk factor for these obesity-related disorders. Increased visceral and hepatic fat, the classic “apple” body type, rather than subcutaneous fat, the classic “pear” body type, is associated with the metabolic syndrome, suggesting distinct biological differences in these two adipose depots.

The specific mechanisms responsible for central adiposity leading to the metabolic syndrome and insulin resistance are not known. Current evidence suggests that metabolic products of omental and mesenteric adipose depots, including inflammatory cytokines and adipocyte hormones, are secreted and released directly into the portal venous system (Figure 9-6). Free fatty acids released from these sites can directly induce hepatic insulin resistance and provide substrate for lipoprotein synthesis. Mediators of inflammation, predominantly tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6), are elevated in central obesity and decline with weight loss.³⁷ TNF-α signaling cascades appear to directly inhibit insulin-mediated tyrosine phosphorylation of insulin receptor substrate (IRS)-1 leading to insulin resistance in liver and other tissues.³⁸ TNF-α has also been associated with the development of early atherosclerosis. Adiponectin, an adipocyte hormone or “adipokine,” has both insulin-sensitizing and anti-inflammatory actions.³⁹ Unfortunately, levels of adiponectin decline with increasing adiposity. Leptin, an adipokine that increases with increasing adiposity, also promotes insulin sensitivity, likely via central mechanisms within the hypothalamus. However, obesity appears to induce a state of central leptin resistance leading to an inability of high levels of leptin to augment insulin sensitivity with increasing fat mass.

Increasing visceral adiposity leads to decreased hepatic insulin sensitivity, increased gluconeogenesis, increased triglyceride/VLDL synthesis, and decreased HDL cholesterol via the mechanisms discussed previously. In muscle, decreased insulin sensitivity leads to decreased glucose utilization. Increased glucose production and decreased utilization lead to compensatory increased insulin levels to maintain euglycemia. Eventually, insulin production is inadequate leading to glucose intolerance. Ongoing glucose and lipid exposure to the β-cell can lead to further decline in β-cell function and the development of T2DM. Hyperinsulinemia, in addition to obesity-induced inflammatory factors, leads to vascular dysfunction and hypertension. When coupled with an atherogenic lipid profile, risk for future cardiovascular disease is markedly increased.

Definition and Classifications of Dyslipidemia

Dyslipidemias are disorders of lipoprotein metabolism, resulting in abnormal serum levels of LDL-C, HDL-C, and/or triglycerides (TG). NHANES has providedpopulation-based gender- and age-specific standards for lipid and lipoprotein concentrations in children and adolescents.⁴⁰ Values exceeding the 95th percentile for TG and LDL-C or below the 5th percentile for HDL-C (Table 9-3) are considered abnormal.⁷ Dyslipidemias in children may be hereditary, secondary, or associated with environmental factors such as diet and physical inactivity.

Primary inherited dyslipidemias

Major disorders of monogenic primary hypercholesterolemia are summarized in Table 9-4. Other monogenic dyslipidemias, leading to insufficient LDL-C levels (abetalipoproteinemia, hypobetalipoproteinemia, proprotein convertase subtilisin/kexin type 9 loss-of-function mutations, and primary bile acid malabsorption), are very rare and not addressed in detail in this chapter.

Within the general population, the most common primary dyslipidemia is familial hypercholesterolemia (FH), an autosomal dominant disorder related to mutations of the gene encoding the LDL receptor (LDLR). While the homozygous form is rare (1:10⁶), the heterozygous form is one of the most common inherited disorders, with an estimated worldwide prevalence of 1:500 (0.2%). The diagnosis is established on clinical findings, although genetic analysis will often identify a causative mutation when applied to those meeting clinical criteria. In children, the diagnosis of heterozygous FH is based on an elevated LDL-C level in the setting of a positive family history of hypercholesterolemia and/or premature cardiovascular disease (a parent, grandparent, aunt, or uncle; age < 55 years for male or < 65 years for female relatives). Most children with heterozygous FH are asymptomatic. However, those with homozygous mutations often manifest xanthomata, yellow lipid-filled nodules, typically located on the eyelids or over a joint, and may experience myocardial ischemia as early as the first decade of life.

In addition to LDLR mutations, autosomal dominant hypercholesterolemia (ADH) may result from rare mutations in two other genes involved in LDL metabolism, the apolipoprotein B (ApoB) and the proprotein convertase subtilisin/kexin type 9 (PCSK9) genes. The clinical presentation of these conditions resembles that of heterozygous FH.⁴¹

Autosomal recessive hypercholesterolemia (ARH) and sitosterolemia are inherited hypercholesterolemias resulting in increased LDL-C levels. Sitosterolemia may present with xanthelasmas in childhood. These are small, yellow, lipid-filled collections typically present on or around the eyelids. When xanthelasmas progress to larger, nodular collections, they are classified as xanthomata. These can be distinguished from ADH disorders by identifying normal or very slightly elevated lipid levels in the parent.⁴¹

Inherited hypertriglyceridemias are represented mainly by Fredrickson type I hyperchylomicronemia, but they also include Fredrickson type V (VLDL and chylomicron remnants) and, to a lesser extent, type IV (excess VLDL) dyslipidemias. All are associated with a reduction in lipoprotein lipase (LPL) activity. LPL is a key enzyme in lipoprotein metabolism, responsible for hydrolysis of the triglyceride core of chylomicrons and VLDLs. The common clinical presentation is characterized by hypertriglyceridemia and may include milky serum, eruptive cutaneous xanthomata, which are small, yellowish-orange papules that may appear all over the body, hepatomegaly, lipemia retinalis, and recurrent pancreatitis which, in the most severe forms, may present in childhood.⁴²

Familial HDL deficiency is characterized by a reduction in serum HDL levels and an increased risk of cardiovascular disease prior to age 50. This condition results from a single copy of a mutation in the ABCA1 gene. The more severe form of inherited HDL deficiency, Tangier disease, results from two copies of the altered gene. The mutations alter the ABCA1 protein, preventing the release ofcholesterol and phospholipids from cells. The accumulation of intracellular lipids leads to impairment of cell function or cell death and may be recognized as ayellowish-orange discoloration of the tonsils. The inability to transport cholesterol and phospholipids out of cells results in a reduction in serum HDL levels, which has been linked to an increased risk of cardiovascular disease.⁴¹

Secondary dyslipidemias

Obesity and overweight are the most common secondary causes of hypertriglyceridemia (adiposity-related hypertriglyceridemia). The metabolic effects of a positive energy-balance diet with a high-fat or high glycemic index content, leading to adipocyte hypertrophy and/or visceral adipose tissue accumulation, result in adverse metabolic consequences contributing to processes such as dyslipidemia, T2DM, and hypertension.

Other causes of secondary dyslipidemias include hypothyroidism, nephrotic syndrome, hepatic disorders, and immunologic conditions.⁴³ Special populations of children at high risk for dyslipidemias include those with cancer, solid-organ transplant, human immunodeficiency virus (HIV), seizures, and immunologic disorders. This may be due in part to the association between dyslipidemia and chronic treatment with corticosteroids, immunosuppressants, antiretroviral regimens, anticonvulsants, retinoids, estrogens, and psychotropic medications (Table 9-5). Poorly controlled diabetes may also lead to dyslipidemia, particularly hypertriglyceridemia. This can be reversed with improved blood sugar control. It is not uncommon to see significantly elevated triglyceride levels acutely normalized with insulin administration.

Relevance of Pediatric Dyslipidemia to Adult Disease

The role of dyslipidemia in the pathogenesis of endothelial dysfunction and accelerated formation of atherosclerotic lesions has been well described. High serum levels of atherogenic lipoproteins (LDL, VLDL, IDL, lipoprotein [a], and lipoprotein remnants) enhance cholesterol delivery to endothelial plaques, promote atherosclerotic progression, and increase the risk of plaque rupture resulting in an increased risk of myocardial infarction or stroke. A series of critical observational studies have demonstrated a clear correlation between lipoprotein disorders and the onset and severity of atherosclerosis in children, adolescents, and young adults.⁴⁴ Prevalence and cohort studies have established the epidemiologic concept of tracking, specifically that concentrations of serum lipids and lipoproteins after two years of age may be used to predict levels in adulthood.

The concept of hypertriglyceridemia as an independent risk factor of coronary artery disease remains controversial. What has been demonstrated is the amplification of risk when hypertriglyceridemia is combined with elevated LDL-C levels. This effect is likely to be secondary to several mechanisms and comorbidities of hypertriglyceridemia, many of which are interdependent. In addition to cardiovascular disease, severe hypertriglyceridemia is associated with pancreatitis (most often in patients with an underlying genetic defect). Marked increases in chylomicrons, the TG-rich lipoprotein particles that predominantly carry postprandial TG, are hypothesized to impair circulatory flow in pancreatic capillary beds, leading to ischemia-induced disruption in acinar structure and exposing the TG-rich particles to pancreatic lipase, leading to necrosis, edema, and inflammation.

The established risk factors for the development of atherosclerosis include dyslipidemia, hypertension, obesity, diabetes, and cigarette smoking. Epidemiologic studies have demonstrated that maintenance of a low-risk lipid profile through childhood, adolescence, and young adulthood until age 50 is associated with exceedingly low lifetime risk of adverse cardiovascular outcomes.⁴⁵ A large and growing body of evidence suggests that early detection of lipid abnormalities and intervention could play a role in reducing the prevalence of adult cardiovascular diseases.

Phenotypic Patterns of Pediatric Dyslipidemia

With the emergence of the epidemic of pediatric obesity, the predominant dyslipidemic pattern in childhood is a combined phenotype of the overweight child with hypertriglyceridemia and lowHDL-C, referred to as adiposity-related hypertriglyceridemia (or atherogenic dyslipidemia). While the LDL-C may be mildly to moderately elevated in this setting, it is typically not in a range requiring pharmacologic intervention. The phenotypic pattern of an asymptomatic, active, normal-weight child with a lipid profile of high LDL-C, normal TG and HDL-C, and a family history of early heart disease characterizes the patient with familial hyperlipidemia (FH). Children with normal weight, a family history of dyslipidemia and a lipid profile of high LDL-C, hypertriglyceridemia, and low HDL-C are classified as type II combined familial hyperlipidemia (Table 9-6).

Obesity/Metabolic Syndrome-Related Dyslipidemia

Dyslipidemia is an important comorbidity of childhood obesity. The pattern of dyslipidemia associated with pediatric obesity consists of a combination of hypertriglyceridemia, decreased HDL-C, and normal to mildly elevated LDL-C.⁸ NHANES data indicate that this pattern is highly prevalent (42.9%) in children with BMI of greater than 95th percentile.

The presence of hypertriglyceridemia and low HDL-C in addition to visceral adiposity (central obesity), hyperglycemia, and hypertension represents a clustering of atherogenic risk factors often described as a “metabolic syndrome.” The pathogenesis and specific definitions of the pediatric metabolic syndrome are detailed earlier in this chapter.

High TG levels are often associated with and may contribute to low HDL-C levels. Elevated TG is initially carried in the form of triglyceride-rich VLDL. In hypertriglyceridemia, VLDL particles acquire cholesterol esters from HDL in exchange for TG. These TG-rich HDL particles became a target for hepatic lipase that removes much of the TG from the HDL, creating a much smaller HDL particle lacking significant cholesterol esters. This results in significantly lower circulating HDL-C levels in the presence of hypertriglyceridemia.

High HDL-C levels are generally associated with decreased CHD risk. Low HDL-C is associated with decreased flux of cholesterol from atherosclerotic plaques, or possibly due to effects related to anti-inflammatory properties associated with HDL particles.

Recommendations for Screening

The most current recommendation endorsed by the American Academy of Pediatrics is to screen overweight (BMI ≥ 85th percentile) children with a fasting lipid profile beginning at 2 years of age. If values are within the reference range on initial screening, the patient should be retested in 3 to 5 years. Universal screening of all children, regardless of weight or other risk factors, is recommended once between 9 and 11 years of age and again between 17 and 18 years of age.⁴³ The new universal screening guidelines have generated significant controversy based on concerns of lack of conclusive evidence that treatment of pediatric dyslipidemia would prevent cardiovascular mortality in adults, as well as the potential adverse psychological impact of screening, insufficient cost analysis, and the unknown risks of long-term lipid-lowering therapy.

The lipid and lipoprotein cut points to be used in screening are presented in Table 9-3. An abnormal nonfasting non–HDL-C measurement should be followed by a fasting lipid profile.

Signs and Symptoms

In order to properly monitor for the development of overweight and obesity, all children should have height and weight measured and weight for length (prior to age 2 years) or BMI (after age 2 years) plotted on the growth curve at every health maintenance visit. Children and adolescents at risk for overweight, with BMI greater than 85th percentile, should receive a medical evaluation. Even before the BMI reaches the 85th percentile, primary care physicians should consider anticipatory guidance, which appears particularly important in the 2- to 5-year age group. Rapid weight gain in early life is a strong predictor of later obesity and represents an excellent opportunity for intervention. Initiation of a diagnostic workup for secondary causes of obesity is usually unnecessary when rapid weight gain is accompanied by upward crossing of height percentiles, but it is particularly appropriate for those children who show a concomitant decrease in linear growth velocity. Nutritional obesity typically stimulates both weight gain and linear growth, so the absence of appropriate height gain in the face of excessive weight gain should trigger an investigation of possible pathologies. For instance, children with hypothyroidism may report increased fatigue, excessively dry skin, constipation, or cold intolerance. Striae, easy bruising, and central fat distribution are potentially consistent with glucocorticoid excess. Early-onset hyperphagic obesity with declining height velocity may signal rapid-onset obesity with hypothalamic dysregulation, hypoventilation, and autonomic dysregulation (ROHHAD) syndrome. A similar clinical picture with the addition of neural crest tumors is termed ROHHADNET syndrome. Although several candidate genes have been investigated, including those associated with other central hypoventilation syndromes, the genetic basis of ROHHAD/ROHHADNET syndrome is currently unknown.

Those individuals with a CNS lesion affecting the hypothalamus or pituitary might report severe nocturnal headaches and emesis, vision changes, polyuria and polydipsia, slower than normal recovery from minor illness, and precocious or delayed puberty. Finally, children with genetic syndromes linked to obesity (eg, Prader-Willi syndrome [PWS]) may have linear growth less than expected for family height genetics, dysmorphic features, hypotonia, and developmental delay.

In the assessment of overweight children, with BMI greater than the 95th percentile, attention to signs and symptoms of comorbid conditions is warranted. Acanthosis nigricans is often present in folds of the skin such as the nape of the neck, axillae, and inner aspect of the elbow in children with insulin resistance. Patients may report polydipsia, polyuria, nocturia, and nocturnal enuresis as clues to the development of DM. Patients who report snoring, sleep disruption, and daytime somnolence may require workup for obstructive sleep apnea. Female adolescent patients may complain of menstrual irregularity, acne, and hirsutism as indicators of possible polycystic ovary syndrome. Overweight children may also report orthopedic complaints such as hip, knee, and/or back pain, and lower extremity deformity.

History and Physical

A careful history and physical examination can largely distinguish individuals with familial obesity from those with obesity secondary to hormone disorders, CNS lesions, or genetic syndromes (Figure 9-7). Significant findings in the early life history such as decreased fetal movement, postnatal hypotonia, and failure to thrive in infancy and early childhood suggest a diagnosis of PWS. Important elements of the neonatal history include hypoglycemia and micropenis and/or cryptorchidism that may be associated with growth hormone (GH) deficiency. Delayed achievement of milestones, learning disabilities, and mental retardation may be associated with a number of genetic syndromes linked to obesity. A marked increase in appetite and food-seeking behaviors beginning at about 2 to 4 years of age, sometimes accompanied by polyuria, polydipsia, and temperature dysregulation, can be associated with ROHHAD syndrome.

In terms of the more immediate history, children who have experienced a sudden increase in weight velocity should be questioned about severe headaches, visual changes, and recurrent vomiting to look for evidence of a CNS lesion. Although quite rare, tumors of the pituitary and/or hypothalamus have been reported to cause obesity. These tumors are usually associated with other findings such as easy bruising, bleeding, and growth failure (ie, Cushing disease) or growth failure and lack of genital development (ie, Fröhlich syndrome).

For all overweight children, the physician should inquire as well about signs and symptoms of hypothyroidism, DM, and obstructive sleep apnea as noted earlier. Menstrual history should be reviewed in adolescent females, keeping in mind that both primary and secondary amenorrhea may be manifestations of hyperandrogenism.

The history should include a review of any available previous growth records. A history of poor height velocity despite increased weight gain is more consistent with hormonal disorders and genetic syndromes, while normal or increased height velocity suggests familial obesity (Figure 9-8).

The final component of the obesity evaluation is an assessment of caloric intake, feeding behaviors, and energy expenditure. A nutritionist can obtain a diet history that details both nutrient composition and total caloric intake. Often, the total caloric intake of an overweight child falls within the broad range of normal for age but still represents more calories than that individual needs for weight maintenance. Identifying maladaptive behaviors is the key to designing an effective treatment plan. Questions regarding food-seeking behaviors, appetite, eating habits, exercise, television watching, and the like are all important in identifying areas for behavioral intervention. It is important to define these behaviors in the context of the family, rather than just the child. Frequently, problems with obesity exist within the entire family. Interventions aimed only at the pediatric patient are unlikely to be successful unless family behavioral patterns are changed.

The physical examination of the overweight child or adolescent should include blood pressure measurement at every visit, using an appropriately sized cuff. General body habitus should be noted, including a description of body fat distribution (apple vs pear body shape), as well as the presence of moon facies, buffalo hump, or any dysmorphic features. On head and neck examination, an obstructed oropharynx (eg, enlarged tonsils) suggests an increased risk for obstructive sleep apnea. The abdominal examination may be difficult depending on the degree of adiposity, but a concerted effort should be made to palpate the liver, as hepatomegaly may be a sign of fatty infiltration of the liver. Tanner staging should be performed with attention to evidence of either precocious or delayed puberty. While precocious pubertal development might suggest the presence of a CNS lesion, early onset of puberty is rather common in obese girls and delayed puberty is often seen in obese adolescent boys. Skin examination should include a search for acanthosis nigricans and skin tags (acrochordons) seen primarily on the posterior neck (nape) and in the axillae, as well as documentation of striae or excessively dry skin. Finally, a neurologic examination, including assessment of visual fields by confrontation, is important if the history suggests the possibility of a CNS lesion.

Diagnostic Tests

The basic evaluation of all overweight children and adolescents should include lipid screening as detailed earlier. Patients who have had a rapid increase in weight velocity or who meet the criteria for overweight or obesity reviewed previously should have a further workup, including a multichemistry panel containing random glucose, transaminases, and bicarbonate (CO₂) (Table 9-7). Prolactin should be included as a marker of possible hypothalamic dysfunction in hyperphagic patients. Thyroid function testing should be done in the setting of poor linear growth to rule out hypothyroidism but should not be routinely ordered in all overweight patients. TSH is often mildly elevated in overweight children and adolescents, and may be secondary to obesity rather than causative. TSH elevation may be more closely tied to the volume of visceral adipose tissue, possibly due to a central effect of elevated leptin levels on TRH release. This is theorized to be a compensatory mechanism to try to maintain normal weight by subtly increasing TSH and thus thyroxine levels, rather than reflecting mild hypothyroidism leading to obesity. TSH concentrations often return to normal with weight loss.

Screening for T2DM and IGT can be accomplished by measuring a fasting plasma glucose (FPG) and/or hemoglobin A1c, or by performing an oral glucose tolerance test (OGTT). Historically, the OGTT was considered the “gold standard” for assessment of glucose metabolism in obese children and adolescents, but, in fact, this test is known to have poor reproducibility and is inconvenient to perform. Because the OGTT requires fasting, it usually cannot be performed in the context of the initial clinical assessment visit. In 2010, the ADA added a hemoglobin A1c cutoff to its list of diagnostic criteria for T2DM and IGT.⁴⁶ There has been some controversy in the pediatric community about how well hemoglobin A1c correlates with FPG and OGTT results,⁴⁷ but a recent position statement from the Pediatric Endocrine Society (PES) Drugs and Therapeutics Committee supports use of any of these methodologies to identify children and adolescents with T2DM or IGT. These tests should be considered especially if there is a history of polydipsia and polyuria, but asymptomatic T2DM or IGT can also occur. Repeat screening should be considered annually, particularly in children older than 10 years. Interpretation of these tests is detailed in Table 9-8.

Additional laboratory workup might include a fasting insulin level and free testosterone level if there is suspicion of polycystic ovary syndrome, gonadotropins, and sex steroids if there is concern about precocious or delayed puberty; a 24-hour urine for free cortisol and creatinine if there is a suggestion of glucocorticoid excess; and a karyotype or other genetic testing if there is suspicion that a syndrome may exist.

Excessive daytime sleepiness and severe snoring may indicate sleep apnea. If the history is consistent with sleep apnea or if an elevated serum CO₂ is detected on a venous blood gas or multichemistry panel, polysomnography is strongly recommended. Other studies such as videotaping and nocturnal pulse oximetry are helpful if positive but are of poor predictive value if negative. If sleep apnea is present, an electrocardiogram (ECG) should be obtained to evaluate for the presence of right ventricular hypertrophy followed up by an echocardiogram and cardiology referral if indicated.

Assessment of actual body composition (lean mass vs fat mass as well as fat distribution) provides much richer data than does height, weight, and BMI alone. Indices such as BMI percentile and weight-for-height percentile are only indirect measures of adiposity. Individuals with increased muscle mass may have a BMI greater than 95th percentile despite a normal amount of adiposity. “Direct” measures of adiposity are available, although many of these methods are primarily used in a research setting. These methods vary from less accurate techniques easily performed in a clinic (ie, skinfold thickness and bioelectrical impedance) to very accurate techniques (ie, DEXA [dual-energy x-ray absorptiometry] scanning, isotope dilution, magnetic resonance spectroscopy, and underwater weighing), most of which are only available in clinical research centers. Although this information is interesting and potentially useful, it is generally not practical for the general pediatric practice setting.

Increased adiposity in childhood and adolescence is not always the result of a genetic predisposition toward weight gain superimposed on a calorie-rich environment. Other conditions can also lead to obesity in children (Tables 9-9 and 9-10). Although endocrine disorders and syndromes associated with obesity are found in fewer than 5% of children referred for evaluation of obesity, it is important to consider these etiologies in the evaluation of the overweight child. Hypothyroidism, GH deficiency, and glucocorticoid excess all may result in an increase in adiposity. In addition, some boys with hypogonadism and low testosterone levels have a body composition that favors fat deposition over development of lean body mass. In contrast to most children with obesity who tend to be tall with advanced bone ages, children with the above hormone deficiencies or with glucocorticoid excess tend to have significantly decreased height velocities and delayed bone ages.

A number of genetic syndromes are also associated with increased weight gain. PWS, most commonly resulting from either a deletion of the paternal allele at the chromosomal region 15q11-13 or uniparental disomy of the maternal allele, is characterized by obesity, hypogonadism, developmental delay, short stature, and hypotonia. Affected children frequently fail to thrive in the first years of life because of early difficulties with swallowing but rapidly gain weight thereafter. Bardet-Biedl syndrome is characterized by obesity, postaxial polydactyly, retinitis pigmentosa, hypogonadism, developmental delay, and short stature. This is an autosomal recessive disorder that is caused by at least three separate mutations that have been mapped to genes on chromosomes 3, 15, and 16. Other genetic disorders, including Carpenter, Cohen, Alström, and Down syndromes, along with Albright hereditary osteodystrophy, are associated with excessive weight gain. As with endocrine disorders leading to obesity, these syndromes are generally associated with short stature. Many are also characterized by developmental delay.

Hypothalamic injury may also result in rapid weight gain and obesity. Lesions involving the arcuate nucleus and/or the paraventricular nucleus of the hypothalamus can disrupt the ability of the brain to detect peripheral adiposity signals. Individuals with the hypothalamic obesity syndrome develop increased food-seeking behavior with lack of satiety. Furthermore, they still tend to gain weight even with “normal” caloric intake, suggesting a decrease in energy expenditure. Craniopharyngiomas are the most common lesions associated with the hypothalamic obesity syndrome. Although occasionally the tumor itself directly damages the circuits that control hunger, satiety, and energy expenditure, surgical removal of these tumors, with resultant iatrogenic hypothalamic damage, more commonly results in obesity. Other forms of traumatic brain damage, including anoxic brain injury, meningitis, and severe head trauma, can, on occasion, result in varying degrees of obesity. The hypothalamic obesity syndrome may be seen without trauma to the CNS as in the ROHHAD/ROHADDNET syndrome described earlier.

The majority of cases of obesity, however, have an “idiopathic” basis, although it is clear that there are major genetic influences that determine body size. The tendency toward developing “idiopathic” or “familial” obesity is polygenic. Despite the fact that many of the genes that regulate weight and energy homeostasis are known, at present it is unclear how often defects in these genes contribute to common obesity. Only a minority of single gene mutations has been found in the coding regions of the known weight regulatory genes that could account for obesity. Genome scans looking for obesity genes, however, have uncovered a number of genetic loci that may contribute to the obesity tendency. There is a strong association between body fat and specific DNA markers on chromosomes 1p, 2p, 3p, 6p, 7q, and 11q.⁴⁸ If human syntenic regions corresponding to mouse quantitative train locus (QTL) maps are included, all human chromosomes, except 9, 18, 21, and Y, have at least one candidate locus for genes regulating weight homeostasis. The large number of obesity loci in the human genome emphasizes the fact that the cause of most human obesity is likely to be polygenic.

At present, single gene mutations account for fewer than 10% of all cases of obesity. By far the most common of these single gene mutations are defects in the melanocortin-4 receptor (MC4R) gene, accounting for approximately 6% of cases of severe obesity with early onset during childhood.⁴⁹ While all individuals with homozygous defects have severe obesity, detailed family studies of heterozygous individuals indicate a variable penetrance of the MC4R gene mutation. In one study, only 68% of individuals with heterozygous mutations were classified as obese,⁵⁰ indicating a large effect of other genetic and environmental modifiers on the actual degree of adiposity in heterozygous individuals.

The onset of obesity in subjects with MC4R mutations occurs in early childhood. Affected individuals tend to have increased lean body mass and increased bone density. Increased linear growth is also observed although no changes in GH secretion have been noted. Thyroid function is normal. Unlike what is seen in leptin deficiency, puberty is not affected and fertility is preserved.

Severe hyperinsulinemia occurs in all obese individuals with MC4R defects.⁵⁰ In addition, subjects with MC4R deficiency appear to be hyperphagic when compared to unaffected siblings.^{49, 50} Many of these individuals exhibit food-seeking behaviors. Unlike the MC4R knockout mouse, humans with MC4R gene mutations exhibit no deficits in energy expenditure.⁵⁰ However, it remains possible that differences in energy expenditure are either too small to detect or would only be apparent after caloric restriction and a decrease in percent body fat to normal.

The obesity and hyperinsulinemia phenotype in MC4R deficiency appears to improve with age.⁵⁰ The decrease in insulin levels parallels an improvement in the hyperphagia. The effect of the MC4R gene defect on long-term obesity-associated morbidity and mortality is presently unknown.

Single gene mutations other than MC4R that are known to result in obesity are collectively quite rare. Mutations have been described in leptin and in the leptin receptor, resulting in early-onset obesity, hyperphagia, hyperinsulinemia, delayed puberty, and short stature.⁴⁹ POMC gene mutations result in childhood-onset obesity, adrenal insufficiency, and red hair.⁴⁹ Finally, prohormone convertase gene mutations have been shown to result in extreme childhood obesity, abnormal glucose homeostasis with very low insulin levels, hypogonadotropic hypogonadism, and adrenal insufficiency.⁵¹

Treatment

Treatment of overweight and obese children and adolescents remains a challenging proposition. Inherited weight set points resist efforts to change body composition. Loss of adipose tissue brings about compensatory decreases in metabolic rate and increases in appetite. Nevertheless, the primary focus of treatment in obese children is decreased caloric intake in combination with increased energy expenditure through increased activity (Figure 9-9). In children who are still growing, it is important that any limit in caloric intake be carefully determined to allow enough calories for adequate growth. Treatment should begin early before the development of severe obesity. The treatment plan should involve the entire family and progress in a stepwise manner as habits are changed and skills are learned.

Lifestyle change

There remains considerable debate regarding whether fats or carbohydrates should be the primary nutritional component limited in the diet to achieve decreased caloric intake. Energy expenditure and body composition do not appear to change with isoenergetic constant protein meals even with extreme changes in the fat-to-carbohydrate ratio. When limiting caloric intake to induce weight loss, however, some studies suggest a benefit of limiting carbohydrate instead of limiting fat to achieve target calorie goals. Although reported weight loss between low-carbohydrate and low-fat diets is similar, paradoxically the low-fat diet results in a less favorable result on lipid profiles.⁵² Diets high in carbohydrate result in increased hepatic lipogenesis and increased triglycerides.⁵³ There may also be a shift to smaller more atherogenic LDL particles with low-fat diets.

Alternatively, low-fat diets may have some advantages over low-carbohydrate diets. Fat consumption may be less satiating than carbohydrate, and a high fat-to-carbohydrate ratio in the diet may promote overconsumption. This would result in a positive energy balance and weight gain in susceptible individuals. Fat is more easily absorbed from the intestine than is carbohydrate and fecal energy loss is lower in diets with a higher percentage of dietary fat. Fatty foods are calorically denser with a greater amount of calories per gram than are carbohydrates. Thus, more calories can be consumed over a short period with fatty foods than with high-carbohydrate foods. These findings would suggest a benefit of low-fat diets.

More recently, there has been growing debate over whether high-protein diets might have some advantage for weight reduction. A 2012 meta-analysis of 74 randomized controlled trials suggests slightly greater weight loss, BMI reduction, and decrease in waist circumference in adults consuming higher-protein diets. However, the magnitude of these differences was small, at least in the short term, with an effect size of only 1.2 kg for weight loss over a 3-month interval. This is not likely of clinical significance unless the effect is cumulative over longer periods, and such data are not available at present.⁵⁴ In a prospective study of energy-restricted high protein versus high carbohydrate, low-fat diets combined with1 year of cognitive behavioral therapy in adults, there was no difference in weight loss between the groups.⁵⁵ This highlights the fact that, regardless of the actual composition of the diet, any decrease in caloric intake should be of some benefit to the overweight patient provided adequate protein is supplied to allow for growth. Conflicting results of various studies make it difficult to recommend one method of decreased caloric intake over another.

The type of carbohydrate consumed may also be an important mediator of increased weight gain. Added sugars, generally mono- and disaccharides added to food by the manufacturer or the consumer, have been implicated in many studies as a factor in increasing adiposity both in adults and children. A large proportion of the added sugars consumed by children and adolescents are consumed in the form of sugar-sweetened beverages. Because liquid intake does not reliably trigger satiety signals, caloric intake via sugar-sweetened beverages is often not offset by a reduction in caloric intake at later meals. This results in an overall positive energy balance that quite often leads to weight gain.⁵⁶ Reducing or eliminating the intake of sugar-sweetened beverages is an essential component of lifestyle change for the overweight child and his/her family.

The metabolism of mono- and disaccharides may also account for their role in promoting obesity. High intake of fructose has been implicated in mediating obesity and numerous obesity-related complications in children and adults.⁵⁷ Most of the soft drinks on the market today contain high-fructose corn syrup. Both fructose and glucose are metabolized via glycolysis and the Krebs cycle to form energy. Alternatively, these compounds can form substrate for lipogenesis and, therefore, energy storage as adipose tissue. Glucokinase, the rate-limiting enzyme for glycolysis, is inhibited by increased adenosine triphosphate (ATP) generation. Inhibition of glucokinase tends to limit the entry of glucose into the cell when energy stores are sufficient and, therefore, less substrate is available for lipogenesis. Fructokinase, however, is not limited by increased energy stores. Therefore, fructose is available to be used and easily converted into triglycerides regardless of the energy state of the cell.

The actual recommended daily caloric intake will vary from patient to patient depending on age and degree of adiposity. Furthermore, total daily energy expenditure required for normal growth and development in each child varies. Depending on the severity of the obese condition, the daily recommended caloric intake in an overweight child should allow for either weight maintenance or a rate of weight gain less than normal for age. This allows the child to gradually approach a healthier weight without the adverse effects on height growth that can be seen with severe caloric restriction. In general, actual weight loss is not recommended in growing children unless life-threatening comorbid conditions exist.

One approach to calculating daily recommended caloric intake is to estimate the total average daily caloric intake in excess of that needed for normal growth and development. This can be estimated using the equation that, for every 3500 kcal of food consumed and not used for growth and development, 1 lb of fat is stored. If a child has gained more weight during a period than would have been expected using standard growth charts, that child has likely consumed excessive calories and converted the calories into adipose tissue. The excessive calories consumed can be calculated by determining the total excess weight gained over a defined number of days (in pounds), multiply by 3500, and divide by the number of days. This daily caloric excess can be subtracted from the patient’s current daily caloric intake as estimated by a nutritionist using dietary recall. In theory, a diet based on these calculations should provide caloric intake equal to that required for normal growth and development. In practice, these calorie calculations serve only as a starting point and often overestimate caloric needs and energy expenditure. Assuming adherence to treatment recommendations, only weekly at-home and monthly clinic weight measurements will provide information regarding whether or not a particular caloric intake is leading to weight maintenance or weight loss in any obese child.

Physical activity increases energy expenditure and is critical to successful weight loss or weight maintenance.⁵⁸ In fact, physical activity accounts for 15% to 30% of total daily energy expenditure, more than any other single component that is amenable to modification. In addition, increased activity may result in increased muscle mass, which increases resting energy expenditure. Exercise has numerous other metabolic benefits in children, including reductions in blood pressure, triglycerides, and insulin levels, all of which decrease cardiovascular risk. Overweight children need to be encouraged to become more active. Current recommendations (eg, CDC and the American Heart Association) suggest that children and adolescents should engage in ≥60 minutes of physical activity daily. Targeting sedentary behaviors (eg, decreasing television viewing and video game playing), rather than enforcing physical activity, results in an improvement in body composition.⁵⁹ Recommendations for exercise should emphasize activities that are fun. If the activities are uncomfortable or uninteresting to the child, they are unlikely to be sustained. Behavioral modification is necessary to achieve and maintain long-term weight loss (Table 9-11). Habits that encourage a sedentary lifestyle need to be changed. In addition, behaviors that result in increased consumption of calories should be altered. Behavioral interventions aimed at the child alone are unlikely to achieve long-term success. Indeed, the only intervention studies that demonstrate long-term success in maintaining weight loss are those targeted to changing behaviors of the entire family.⁵⁹

Weight-loss medications and bariatric surgery

Intensive intervention, either medical or surgical, should be discouraged except in very limited situations. In patients with documented serious comorbid conditions who have failed prolonged attempts at nutritional management and behavioral counseling, weight loss medications might be considered. Orlistat, a lipase inhibitor, is the only FDA-approved weight loss drug for children aged 12 years and older. It has a modest effect on weight, and compliance is limited by gastrointestinal side effects. In children and adolescents, the glucagon-like peptide-1 receptor agonist, exenatide, may have some benefit in the treatment of profound pediatric obesity (BMI ≥ 1.2 times the 95th percentile for age or BMI ≥ 35 kg/m²) but is not FDA-approved for this indication. Exenatide appears to work by increasing satiety and decreasing appetite and, over the short term, reducing BMI, body weight, and fasting insulin levels, with only modest short-term side effects, primarily mild nausea.⁶⁰ There is increasing concern, however, that exenatide may increase the risk of pancreatitis, pancreatic cancer, and thyroid cancers. Overall, there are minimal data regarding the safety and efficacy of other anorectic medications in pediatric patients. The short-term effects on growth and the long-term effects on general health of these medications in children are unknown. Weight loss induced by medications may require lifelong administration of these drugs for weight maintenance, because adults tend to regain previously lost weight when drugs such as orlistat and exenatide are stopped or replaced with placebo.⁶¹ Therefore, long-term potential side effects need to be considered and balanced against current health problems resulting from the overweight condition of the patient.

Bariatric surgery has been performed on a limited number of obese adolescent patients and has yielded some success. Studies have shown postoperative weight loss of 20 to 60 lb and BMI loss of up to 35% following laparoscopic gastric banding and gastric bypass surgeries.⁶² Gastric bypass appears to be more effective, with weight loss almost double that seen with banding procedures.⁶² Guidelines were published in 2009 to assist in selecting appropriate adolescent patients for bariatric surgical intervention. These guidelines state that adolescents with a BMI greater than 35 kg/m² and severe comorbidities or a BMI greater than 40 kg/m² and mild comorbidities could be considered for bariatric surgery if a number of other criteria are met, including pubertal development to Tanner stage 4 or beyond, achievement of 95% of predicted adult height, commitment to psychological evaluation before surgery, anticipated good compliance with postoperative recommendations, and failure of organized preoperative attempts at weight reduction.⁶² Surgery should be performed only at centers experienced in adolescent bariatric surgery. Close follow-up with a weight loss team is essential for long-term success. As with anorectic medications, the long-term consequences of these procedures in adolescents are unknown, and therefore bariatric surgery should be cautiously considered only in cases in which significant comorbid conditions exist and diet and exercise have failed.

Treatment of Metabolic Syndrome

True therapeutic lifestyle intervention, delivered in a multidisciplinary setting and based on cognitive behavioral therapy, has been shown to reduce the frequency of clinical findings consistent with the metabolic syndrome in obese adolescents.⁶³ Although there is evidence for improvement in features of the metabolic syndrome in adults with the use of metformin, a biguanide drug that improves insulin sensitivity, there is debate in the pediatric literature regarding to what extent metformin adds value to therapeutic lifestyle change in the management of metabolic syndrome during adolescence. Many, but not all, of the retrospective studies that have been conducted support greater BMI reductions when metformin is added to therapeutic lifestyle interventions.⁶⁴ One double-blind, placebo-controlled prospective study, however, demonstrated a 1.3% reduction in BMI, 11% reduction in fasting glucose, and 38% reduction in fasting insulin levels following 6 months of modest-dose metformin therapy in adolescents aged 12 to 19 years with hyperinsulinemia and a family history of T2DM. None of the adolescents in this study had glucose intolerance or PCOS.⁶⁵ Another study demonstrated a significant decrease in triglycerides and total cholesterol with metformin and a low-calorie diet with only 8 weeks of intervention.⁶⁶ Metformin may be prescribed at a total daily dose of up to 2000 mg, divided twice daily. The most common side effects include gastrointestinal, including nausea, bloating, diarrhea, and sometimes emesis. These may be mitigated by starting at a low dose (500 mg daily) and working up by 500 mg every 1 to2 weeks. Taking the medication with food is also helpful.

If the manifestations of the metabolic syndrome include IGT, this may be an indication in itself for the use of metformin. In light of evidence in the adult population that metformin therapy may decrease the progression of IGT to frank DM by as much as one-third,⁶⁷ many pediatric providers use metformin in children and adolescents with glucose intolerance or simply insulin resistance with normal glucose tolerance (see Figure 9-9).

Because the use of metformin is sometimes limited by its gastrointestinal side effects, there has been some discussion of whether other insulin-sensitizing agents might also be useful in the obese adolescent population. Thiazolidinediones (TZDs) are known in the adult population to be effective in the management of T2DM and in the prevention of progression from IGT to T2DM. There are very limited data in adolescents on the use of drugs within this class, but one prospective study demonstrated that the use of a TZD resulted in conversion from IGT to normal glucose tolerance in 58% of subjects, as compared to a conversion rate of 44% for placebo-treated subjects.⁶⁸ While this may present an interesting avenue for future research, TZDs are not currently approved for use in children or adolescents, and they have been associated with rare, but very serious hepatotoxicity and cardiotoxicity. The use of these drugs should be limited at this time to the setting of research protocols.

Treatment of Dyslipidemia

The most recent policy statement of the American Academy of Pediatrics recommends 6 months of dietary and lifestyle modification as initial therapy for children found to have dyslipidemia on screening studies (see Figure 9-9). When dietary and lifestyle modifications are insufficient, pharmacologic intervention is recommended for patients 10 years and older with an LDL-C 190 mg/dL or greater, or an LDL-C 160 mg/dL or greater with a family history of early cardiovascular disease or at least two additional risk factors of early cardiovascular disease, or an LDL 130 mg/dL or greater with DM. The most significant change in the most recent expert panel recommendations was the identification of statins as the first line of pharmacologic therapy for patients meeting criteria. Prior to initiating statin therapy, liver function studies should be performed to establish a baseline. Regardless of the absolute value of the LDL-C, if the patient meets criteria for treatment, statin therapy is initiated at a low dose. Laboratory studies are performed at approximately 3-month intervals to assess therapeutic response as well as the effects of therapy on liver function. The initial goal is to lower the LDL-C concentration to less than 160 mg/dL; however, in the presence of high-risk situations such as a family history of cardiovascular disease or a patient history of obesity, diabetes, metabolic syndrome or other factors, the target goal may be as low as 110 mg/dL.⁴³

For patients with elevated triglycerides and/or decreased HDL-C, lifestyle modifications remain the primary therapeutic approach. While the precise mechanism of action is not established, studies suggest that omega-3 fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) reduce serumTG levels through decreased hepatic secretion of TG-containing lipoproteins and enhanced clearance of TG-containing lipoproteins.⁶⁹ A daily dose of 1000 to 2000 mg DHA has been shown to lower TG, raise HDL-C, and it may paradoxically increase LDL-C. Patients with mixed dyslipidemias, including elevated TG and LDL-C, may benefit from the combination of omega-3 fatty acids and a statin.

An alternative therapy, nicotinic acid (Niacin), is not as effective as statins at decreasing serum LDL-C; however, it offers the advantage of raising HDL-C and lowering triglycerides more effectively than statins. Intense flushing is often experienced by patients taking high-dose niacin and compliance in children, primarily for this reason, is often limited.

Patients with secondary adiposity-related hypertriglyceridemia (atherogenic dyslipidemia) consisting of hypertriglyceridemia, low HDL-C, and features of the metabolic syndrome benefit most from intensive weight management and appropriate dietary changes. The presence of obesity should prompt specific evaluation of all other cardiovascular risk factors (systemic hypertension, DM, tobacco exposure, and dyslipidemia) and comorbidities (hepatic steatosis, left ventricular hypertrophy, sleep apnea, and right ventricular hypertension).

If criteria for pharmacological therapy for hypertriglyceridemia are met, the use of LDL-C-lowering medications is recommended as first-line therapy followed by consideration of fibrates.¹¹ The combination of a statin with omega-3 fatty acids has been shown to be effective in improving lipid parameters beyond that of the statin alone.

Treatment of Comorbidities

Polycystic ovary syndrome

As is true for metabolic syndrome, there are multiple proposed sets of diagnostic criteria for PCOS, including those published by the NIH in 1990, the Rotterdam criteria published in 2003, a set of criteria designed by the Androgen Excess and PCOS Society in 2009, and a revised set of Rotterdam criteria published in 2012 with special attention to the diagnostic conundrums associated with PCOS in adolescence (see Chapter 3 for more detail). All these criteria include oligomenorrhea, clinical and/or biochemical hyperandrogenism, + radiographic evidence of polycystic ovaries. Although there is considerable overlap between the populations displaying elements of the metabolic syndrome and those who meet criteria for PCOS, neither obesity nor any of the metabolic derangements of the metabolic syndrome are required to make a diagnosis of PCOS. When PCOS is seen in conjunction with the metabolic syndrome, there is evidence to suggest that metformin is beneficial. A study of obese adolescent girls with evidence of PCOS showed that the addition of metformin to a treatment regimen consisting of lifestyle intervention + oral contraceptives further reduced serum testosterone levels, decreased waist circumference, and increased HDL cholesterol.⁷⁰ However, metformin does not appear to be adequate as monotherapy to control the hyperandrogenic symptoms in most adolescents with PCOS. Combination oral contraceptives are still felt to be the medical intervention of greatest utility for this purpose. In some cases, antiandrogens such as spironolactone may also have a role.

Nonalcoholic fatty liver disease

An open-label pilot study of metformin in children with biopsy-proven nonalcoholic fatty liver disease (NAFLD) showed significant improvement in both serum aminotransferase levels and hepatic fat content assessed by magnetic resonance (MR) spectroscopy, prompting a randomized controlled trial.¹² The TONIC (Treatment of Nonalcoholic Fatty Liver Disease in Children) trial is a randomized, double-blind, placebo-controlled study comparing the effects of 800 IU vitamin E or 1000 mg metformin against placebo over 96 weeks of treatment in children aged 8 to 17 years with biopsy-confirmed NAFLD. Although neither vitamin E nor metformin was superior to placebo in achieving sustained reductions in aminotransferase levels, 48% of metformin-treated subjects had resolution of NAFLD by histopathology at the end of the 96-week treatment period, versus only 22% of those receiving placebo.⁷¹

Pulmonary comorbidities

Both reactive airway disease and sleep apnea respond positively to weight loss. Asthma should be aggressively treated with pharmacologic agents, including inhaled glucocorticoids if needed. The lowest possible dose should be used as multiple studies have demonstrated a systemic effect at moderate to high doses. Systemic absorption can increase weight gain and worsen insulin resistance. Sleep apnea should be suspected in individuals with severe snoring, morning headaches, or daytime somnolence. A sleep study should be obtained and treatment with bilevel positive airway pressure (BiPAP) considered in confirmed cases (see Figure 9-9). If sleep apnea is severe, consider performing an ECG and referring to a cardiologist if right-sided cardiac hypertrophy is detected.

When to Refer

The only therapy that has shown long-term success in the treatment of obese children is family-based behavioral modification therapy. Unfortunately, such programs are not widely available or financially accessible. When attempts to manage pediatric obesity fail in the primary care setting, many children are referred to pediatric endocrinologists. However, in a retrospective study of 587 children referred to pediatric endocrinologists for obesity evaluation and treatment, the mean percentage overweight increased over the follow-up interval of 1.9 years. While 38% of children showed improvement in percentage overweight, only five subjects (< 1%) reduced their BMI to less than 95th percentile.⁷² As such, referral should be reserved for those likely to have an etiology other than idiopathic familial obesity, and, of course, for those demonstrating obesity-induced comorbidity.

• Referral to endocrinologists. Primary care physicians should consider referral to a pediatric endocrinologist for patients who meet any of the criteria in Box 9-1.

• Referral to other subspecialists. Referral to subspecialists other than pediatric endocrinologists is warranted if the initial workup or follow-up of obese patients reveals any of the findings listed in Box 9-2.

• Reporting to child protective services. A report of medical neglect is warranted in the setting of life-threatening complications of obesity that are not being adequately addressed in the home environment.⁷³

Although positive results can sometimes be obtained in the treatment of overweight children, successful resolution of the childhood obesity epidemic will not be accomplished at the individual level. Profound changes within society have resulted in the dramatic increase in the number of overweight children. It will take equally profound changes at the policy level to reverse this trend. Needs include government support of research programs into the causes and treatment of obesity and insurance industry recognition of obesity as a medical disorder and coverage for evaluation and long-term treatment of overweight children. Schools should have adequate resources to provide physical education to all children from kindergarten through high school. Adequate funding for school lunch programs would allow nutritious low-fat and lower-calorie meals to be provided to all students. Families should be encouraged to alter their lifestyles by eating healthier, consuming less fast food, and getting more exercise. Provision of safe environments with adequate parks, bike paths, and the like for these endeavors will certainly help promote a more active lifestyle. The genetic makeup of our population will not change over short periods of time and, in fact, demographic trends suggest greatest growth in groups at higher risk for obesity and its complications. Therefore, the problem of childhood obesity now appears to be a permanent part of the pediatric healthcare landscape. Positive changes within our society, however, can have a profound effect on the prevalence of childhood obesity and result in a much healthier population now and in the future.