Although the molecular pathways regulating energy balance are beginning to be illuminated, the causes of obesity remain elusive. In part, this reflects the fact that obesity is a heterogeneous group of disorders. At one level, the pathophysiology of obesity seems simple: a chronic excess of nutrient intake relative to the level of energy expenditure. However, due to the complexity of the neuroendocrine and metabolic systems that regulate energy intake, storage, and expenditure, it has been difficult to quantitate all the relevant parameters (e.g., food intake and energy expenditure) over time in human subjects.
Role of genes versus environment
Obesity is commonly seen in families, and the heritability of body weight is similar to that for height. Inheritance is usually not Mendelian, however, and it is difficult to distinguish the role of genes and environmental factors. Adoptees more closely resemble their biologic than adoptive parents with respect to obesity, providing strong support for genetic influences. Likewise, identical twins have very similar BMIs whether reared together or apart, and their BMIs are much more strongly correlated than those of dizygotic twins. These genetic effects appear to relate to both energy intake and expenditure. Currently, identified genetic variants, both common and rare, account for less than 5% of the variance of body weight.
Whatever the role of genes, it is clear that the environment plays a key role in obesity, as evidenced by the fact that famine prevents obesity in even the most obesity-prone individual. In addition, the recent increase in the prevalence of obesity in the United States is far too rapid to be due to changes in the gene pool. Undoubtedly, genes influence the susceptibility to obesity in response to specific diets and availability of nutrition. Cultural factors are also important—these relate to both availability and composition of the diet and to changes in the level of physical activity. In industrial societies, obesity is more common among poor women, whereas in underdeveloped countries, wealthier women are more often obese. In children, obesity correlates to some degree with time spent watching television. Although the role of diet composition in obesity continues to generate controversy, it appears that high-fat diets may, when combined with simple, rapidly absorbed carbohydrates, promote obesity. Specific genes are likely to influence the response to specific diets, but these genes are largely unidentified.
Additional environmental factors may contribute to the increasing obesity prevalence. Both epidemiologic correlations and experimental data suggest that sleep deprivation leads to increased obesity. Changes in gut microbiome with capacity to alter energy balance are receiving experimental support from animal studies, and a possible role for obesigenic viral infections continues to receive sporadic attention.
Specific genetic syndromes
For many years, obesity in rodents has been known to be caused by a number of distinct mutations distributed through the genome. Most of these single-gene mutations cause both hyperphagia and diminished energy expenditure, suggesting a physiologic link between these two parameters of energy homeostasis. Identification of the ob gene mutation in genetically obese (ob/ob) mice represented a major breakthrough in the field. The ob/ob mouse develops severe obesity, insulin resistance, and hyperphagia, as well as efficient metabolism (e.g., it gets fat even when ingesting the same number of calories as lean litter mates). The product of the ob gene is the peptide leptin, a name derived from the Greek root leptos, meaning thin. Leptin is secreted by adipose cells and acts primarily through the hypothalamus. Its level of production provides an index of adipose energy stores (Fig. 20-4). High leptin levels decrease food intake and increase energy expenditure. Another mouse mutant, db/db, which is resistant to leptin, has a mutation in the leptin receptor and develops a similar syndrome. The ob gene is present in humans where it is also expressed in fat. Several families with morbid, early-onset obesity caused by inactivating mutations in either leptin or the leptin receptor have been described, thus demonstrating the biologic relevance of the leptin pathway in humans. Obesity in these individuals begins shortly after birth, is severe, and is accompanied by neuroendocrine abnormalities. The most prominent of these is hypogonadotropic hypogonadism, which is reversed by leptin replacement in the leptin-deficient subset. Central hypothyroidism and growth retardation are seen in the mouse model, but their occurrence in leptin-deficient humans is less clear. Mutations in the leptin or leptin receptor genes do not play a prominent role in common forms of obesity.
The physiologic system regulated by leptin. Rising or falling leptin levels act through the hypothalamus to influence appetite, energy expenditure, and neuroendocrine function and through peripheral sites to influence systems such as the immune system.
Mutations in several other genes cause severe obesity in humans (Table 20-1); each of these syndromes is rare. Mutations in the gene encoding proopiomelanocortin (POMC) cause severe obesity through failure to synthesize α-MSH, a key neuropeptide that inhibits appetite in the hypothalamus. The absence of POMC also causes secondary adrenal insufficiency due to absence of adrenocorticotropic hormone (ACTH), as well as pale skin and red hair due to absence of α-MSH. Proenzyme convertase 1 (PC-1) mutations are thought to cause obesity by preventing synthesis of α-MSH from its precursor peptide, POMC. α-MSH binds to the type 4 melanocortin receptor (MC4R), a key hypothalamic receptor that inhibits eating. Heterozygous loss-of-function mutations of this receptor account for as much as 5% of severe obesity. Loss of function of MRAP2, a protein required for normal MC4R signaling, has been found in rare cases of severe obesity. These six genetic defects define a pathway through which leptin (by stimulating POMC and increasing α-MSH) restricts food intake and limits weight (Fig. 20-5). The results of genomewide association studies to identify genetic loci responsible for obesity in the general population have so far been disappointing. More than 40 replicated loci linked to obesity have been identified, but together they account for less than 3% of interindividual variation in BMI. The most replicated of these is a gene named FTO, which is of unknown function, but like many of the other recently described candidates, is expressed in the brain. Because the heritability of obesity is estimated to be 40–70%, it is likely that many more loci remain to be identified. It is possible that epistatic interactions between causative loci or unknown gene-environment interactions explain the poor success at identifying causal loci.
A central pathway through which leptin acts to regulate appetite and body weight. Leptin signals through proopiomelanocortin (POMC) neurons in the hypothalamus to induce increased production of α-melanocyte-stimulating hormone (α-MSH), requiring the processing enzyme PC-1 (proenzyme convertase 1). α-MSH acts as an agonist on melanocortin-4 receptors to inhibit appetite, and the neuropeptide AgRp (Agouti-related peptide) acts as an antagonist of this receptor. Mutations that cause obesity in humans are indicated by the solid green arrows.
In addition to these human obesity genes, studies in rodents reveal several other molecular candidates for hypothalamic mediators of human obesity or leanness. The tub gene encodes a hypothalamic peptide of unknown function; mutation of this gene causes late-onset obesity. The fat gene encodes carboxypeptidase E, a peptide-processing enzyme; mutation of this gene is thought to cause obesity by disrupting production of one or more neuropeptides. AgRP is coexpressed with NPY in arcuate nucleus neurons. AgRP antagonizes α-MSH action at MC4 receptors, and its overexpression induces obesity. In contrast, a mouse deficient in the peptide MCH, whose administration causes feeding, is lean.
TABLE 20-1SELECTED OBESITY GENES IN HUMANS AND MICE ||Download (.pdf) TABLE 20-1 SELECTED OBESITY GENES IN HUMANS AND MICE
|GENE ||GENE PRODUCT ||MECHANISM OF OBESITY ||IN HUMAN ||IN RODENT |
|Lep (ob) ||Leptin, a fat-derived hormone ||Mutation prevents leptin from delivering satiety signal; brain perceives starvation ||Yes ||Yes |
|LepR (db) ||Leptin receptor ||Same as above ||Yes ||Yes |
|POMC ||Proopiomelanocortin, a precursor of several hormones and neuropeptides ||Mutation prevents synthesis of melanocyte-stimulating hormone (MSH), a satiety signal ||Yes ||Yes |
|MC4R ||Type 4 receptor for MSH ||Mutation prevents reception of satiety signal from MSH ||Yes ||Yes |
|AgRP ||Agouti-related peptide, a neuropeptide expressed in the hypothalamus ||Overexpression inhibits signal through MC4R ||No ||Yes |
|PC-1 ||Prohormone convertase 1, a processing enzyme ||Mutation prevents synthesis of neuropeptide, probably MSH ||Yes ||No |
|Fat ||Carboxypeptidase E, a processing enzyme ||Same as above ||No ||Yes |
|Tub ||Tub, a hypothalamic protein of unknown function ||Hypothalamic dysfunction ||No ||Yes |
|TrkB ||TrkB, a neurotrophin receptor ||Hyperphagia due to uncharacterized hypothalamic defect ||Yes ||Yes |
A number of complex human syndromes with defined inheritance are associated with obesity (Table 20-2). Although specific genes have limited definition at present, their identification will likely enhance our understanding of more common forms of human obesity. In the Prader-Willi syndrome, a multigenic neurodevelopmental disorder, obesity coexists with short stature, mental retardation, hypogonadotropic hypogonadism, hypotonia, small hands and feet, fish-shaped mouth, and hyperphagia. Most patients have reduced expression of imprinted paternally inherited genes encoded in the 15q11-13 chromosomal region. Reduced expression of Snord116, a small nucleolar RNA highly expressed in hypothalamus, may be an important cause of defective hypothalamic function in this disorder. Bardet-Biedl syndrome (BBS) is a genetically heterogeneous disorder characterized by obesity, mental retardation, retinitis pigmentosa, diabetes, renal and cardiac malformations, polydactyly, and hypogonadotropic hypogonadism. At least 12 genetic loci have been identified, and most of the encoded proteins form two multiprotein complexes that are involved in ciliary function and microtubule-based intracellular transport. Some evidence suggests that mutations might disrupt leptin receptor trafficking in key hypothalamic neurons, causing leptin resistance.
TABLE 20-2A COMPARISON OF SYNDROMES OF OBESITY—HYPOGONADISM AND MENTAL RETARDATION ||Download (.pdf) TABLE 20-2 A COMPARISON OF SYNDROMES OF OBESITY—HYPOGONADISM AND MENTAL RETARDATION
|FEATURE ||SYNDROME |
|PRADER-WILLI ||LAURENCE-MOON-BIEDL ||AHLSTROM’S ||COHEN’S ||CARPENTER’S |
|Inheritance ||Sporadic; two-thirds have defect ||Autosomal recessive ||Autosomal recessive ||Probably autosomal recessive ||Autosomal recessive |
|Stature ||Short ||Normal; infrequently short ||Normal; infrequently short ||Short or tall ||Normal |
|Obesity ||Generalized ||Generalized ||Truncal ||Truncal ||Truncal, gluteal |
| ||Moderate to severe ||Early onset, 1–2 years ||Early onset, 2–5 years ||Mid-childhood, age 5 || |
| ||Onset 1–3 years || || || || |
|Craniofacies ||Narrow bifrontal diameter ||Not distinctive ||Not distinctive ||High nasal bridge ||Acrocephaly |
| ||Almond-shaped eyes || || ||Arched palate ||Flat nasal bridge |
| ||Strabismus || || ||Open mouth ||High-arched palate |
| ||V-shaped mouth || || ||Short philtrum || |
| ||High-arched palate || || || || |
|Limbs ||Small hands and feet ||Polydactyly ||No abnormalities ||Hypotonia ||Polydactyly |
| ||Hypotonia || || ||Narrow hands and feet ||Syndactyly |
| || || || || ||Genu valgum |
|Reproductive status ||1° Hypogonadism ||1° Hypogonadism ||Hypogonadism in males but not in females ||Normal gonadal function or hypogonadotropic hypogonadism ||2° Hypogonadism |
|Other features ||Enamel hypoplasia || || ||Dysplastic ears || |
| ||Hyperphagia || || ||Delayed puberty || |
| ||Temper tantrums || || || || |
| ||Nasal speech || || || || |
|Mental retardation ||Mild to moderate || ||Normal intelligence ||Mild ||Slight |
Other specific syndromes associated with obesity
Although obese patients commonly have central obesity, hypertension, and glucose intolerance, they lack other specific stigmata of Cushing’s syndrome (Chap. 8). Nonetheless, a potential diagnosis of Cushing’s syndrome is often entertained. Cortisol production and urinary metabolites (17OH steroids) may be increased in simple obesity. Unlike in Cushing’s syndrome, however, cortisol levels in blood and urine in the basal state and in response to corticotropin-releasing hormone (CRH) or ACTH are normal; the overnight 1-mg dexamethasone suppression test is normal in 90%, with the remainder being normal on a standard 2-day low-dose dexamethasone suppression test. Obesity may be associated with excessive local reactivation of cortisol in fat by 11β-hydroxysteroid dehydrogenase 1, an enzyme that converts inactive cortisone to cortisol.
The possibility of hypothyroidism should be considered, but it is an uncommon cause of obesity; hypothyroidism is easily ruled out by measuring thyroid-stimulating hormone (TSH). Much of the weight gain that occurs in hypothyroidism is due to myxedema (Chap. 7).
Patients with insulinoma often gain weight as a result of overeating to avoid hypoglycemic symptoms (Chap. 26). The increased substrate plus high insulin levels promote energy storage in fat. This can be marked in some individuals but is modest in most.
Craniopharyngioma and other disorders involving the hypothalamus
Whether through tumors, trauma, or inflammation, hypothalamic dysfunction of systems controlling satiety, hunger, and energy expenditure can cause varying degrees of obesity (Chap. 4). It is uncommon to identify a discrete anatomic basis for these disorders. Subtle hypothalamic dysfunction is probably a more common cause of obesity than can be documented using currently available imaging techniques. Growth hormone (GH), which exerts lipolytic activity, is diminished in obesity and is increased with weight loss. Despite low GH levels, insulin-like growth factor (IGF) I (somatomedin) production is normal, suggesting that GH suppression may be a compensatory response to increased nutritional supply.
Pathogenesis of common obesity
Obesity can result from increased energy intake, decreased energy expenditure, or a combination of the two. Thus, identifying the etiology of obesity should involve measurements of both parameters. However, it is difficult to perform direct and accurate measurements of energy intake in free-living individuals; and the obese, in particular, often underreport intake. Measurements of chronic energy expenditure are possible using doubly labeled water or metabolic chamber/rooms. In subjects at stable weight and body composition, energy intake equals expenditure. Consequently, these techniques allow assessment of energy intake in free-living individuals. The level of energy expenditure differs in established obesity, during periods of weight gain or loss, and in the pre- or postobese state. Studies that fail to take note of this phenomenon are not easily interpreted.
There is continued interest in the concept of a body weight “set point.” This idea is supported by physiologic mechanisms centered around a sensing system in adipose tissue that reflects fat stores and a receptor, or “adipostat,” that is in the hypothalamic centers. When fat stores are depleted, the adipostat signal is low, and the hypothalamus responds by stimulating hunger and decreasing energy expenditure to conserve energy. Conversely, when fat stores are abundant, the signal is increased, and the hypothalamus responds by decreasing hunger and increasing energy expenditure. The recent discovery of the ob gene, and its product leptin, and the db gene, whose product is the leptin receptor, provides important elements of a molecular basis for this physiologic concept (see above).
What is the status of food intake in obesity? (do the obese eat more than the lean?)
This question has stimulated much debate, due in part to the methodologic difficulties inherent in determining food intake. Many obese individuals believe that they eat small quantities of food, and this claim has often been supported by the results of food intake questionnaires. However, it is now established that average energy expenditure increases as individuals get more obese, due primarily to the fact that metabolically active lean tissue mass increases with obesity. Given the laws of thermodynamics, the obese person must therefore eat more than the average lean person to maintain their increased weight. It may be the case, however, that a subset of individuals who are predisposed to obesity have the capacity to become obese initially without an absolute increase in caloric consumption.
What is the state of energy expenditure in obesity?
The average total daily energy expenditure is higher in obese than lean individuals when measured at stable weight. However, energy expenditure falls as weight is lost, due in part to loss of lean body mass and to decreased sympathetic nerve activity. When reduced to near-normal weight and maintained there for a while, (some) obese individuals have lower energy expenditure than (some) lean individuals. There is also a tendency for those who will develop obesity as infants or children to have lower resting energy expenditure rates than those who remain lean. The physiologic basis for variable rates of energy expenditure (at a given body weight and level of energy intake) is essentially unknown.
Another component of thermogenesis, called nonexercise activity thermogenesis (NEAT), has been linked to obesity. It is the thermogenesis that accompanies physical activities other than volitional exercise such as the activities of daily living, fidgeting, spontaneous muscle contraction, and maintaining posture. NEAT accounts for about two-thirds of the increased daily energy expenditure induced by overfeeding. The wide variation in fat storage seen in overfed individuals is predicted by the degree to which NEAT is induced. The molecular basis for NEAT and its regulation is unknown.
Leptin in typical obesity
The vast majority of obese persons have increased leptin levels but do not have mutations of either leptin or its receptor. They appear, therefore, to have a form of functional “leptin resistance.” Data suggesting that some individuals produce less leptin per unit fat mass than others or have a form of relative leptin deficiency that predisposes to obesity are at present contradictory and unsettled. The mechanism for leptin resistance, and whether it can be overcome by raising leptin levels or combining leptin with other treatments in a subset of obese individuals, is not yet established. Some data suggest that leptin may not effectively cross the blood-brain barrier as levels rise. It is also apparent from animal studies that leptin-signaling inhibitors, such as SOCS3 and PTP1b, are involved in the leptin-resistant state.