ALBERT J. STUNKARD AND THOMAS A. WADDEN
Incidence and Epidemiology
Physical Examination and Laboratory Findings
Obesity is a condition characterized by excessive accumulation of fat in the body. In clinical practice, overweight (weight corrected for height) is used as a surrogate for body fat. This practice is reasonable because body weight and fat are highly correlated, especially at greater degrees of overweight. Traditionally, overweight has been defined in terms of tables of ideal or desirable body weight. It is now defined in terms of the body mass index (BMI = weight in kilograms divided by height in meters squared), with overweight defined as a BMI of greater than 25 and obesity as a BMI greater than 30. Table 36.1 shows the BMI values for different heights and weights in inches and pounds, and Table 36.2 depicts the World Health Organization’s classification of obesity. Unlike many “real” diseases and like hypertension, obesity represents one arm of a distribution curve of body weight, with no physiologically defined cutoff point. In the future, diagnosis may be based on newer methods of measuring body fat. Until then, for most practical purposes, the eyeball test is sufficient: if a person looks fat, the person is fat.
TABLE 36.1. BODY MASS INDEX
TABLE 36.2. MODIFIED WHO CLASSIFICATION OF OVERWEIGHT AND OBESITY
INCIDENCE AND EPIDEMIOLOGY
Obesity is one of the most pervasive public health problems in the United States today. One-third of all adults are overweight, and this high prevalence occurs in all ethnic and racial groups at all ages and in both sexes. It is even higher among persons with lower levels of education and certain ethnic or racial groups. Figure 36.1 shows the prevalence of overweight among three racial or ethnic groups by age, separately for men and women. A striking finding is the very high prevalence (almost 60%) among middle-aged African-American and Mexican-American women; the value for white women is 33.5%.
FIGURE 36.1. Prevalence of obesity by gender, age, and ethnic status according to the former criteria of a body mass index greater than 27.8 for men and 27.3 for women. (From Kuczmarski RJ, Flegal KM, Campbell SM, et al. Increasing prevalence of overweight among US adults: the national health and nutrition examination surveys, 1960–1991. JAMA 1994;272:205, with permission.)
For men and women, the prevalence of overweight increases with each 10-year increment of age until 50 to 59 years of age, when it begins to fall progressively at older ages. Overweight is also highly correlated with socioeconomic status; increasing body weight is associated with decreasing socioeconomic status, particularly among women. The high prevalence of overweight in the United States is compounded by a striking 30% increase in the past decade. The problem extends across the entire spectrum of body weight. During the past decade, the average American gained 3.6 kg, and the proportion of persons with healthier, lighter weights decreased significantly. Even children have not been spared the rapid increase in body weight; 25% of American children are overweight.
What causes obesity? In one sense, the answer is simple—consuming more calories than are expended as energy. In another sense, the answer is most elusive. The causes of obesity are to be found in the regulation of body weight (which is primarily the regulation of body fat), and we still have only an imperfect understanding of how this regulation is achieved. We do know that weight is regulated with great precision. During a lifetime, the average person consumes at least 60 million kcal. A gain or loss of 20 lb, or 72,000 kcal, would represent an error of no more than 0.001%. The determinants of obesity or, as we have suggested, an elevated body weight set point can be divided into genetic, environmental, and regulatory factors.
The existence of numerous forms of genetic obesity in experimental animals and the ease with which adiposity can be produced by the selective breeding of farm animals suggest that genetic factors can play an important role in human obesity. The stunning advances in our knowledge during the past decade have made it clear that genetic factors do play an important role in human obesity. The first studies, using the classic twin method, estimated very high levels of heritability; the percentage of the variance in body weight accounted for by genetic influences was about 80%. Even studies of identical twins separated at birth, a method that avoided some of the bias in classic twin studies, estimated heritability at 66%. These studies are still widely cited, but there is a growing consensus that they overestimate the influence of heredity.
The results of adoption studies and complex segregation analysis agree on a heritability of BMI of about 33%, a value now viewed as a more reasonable estimate than that of the twin studies. Genetic influences may play a more important part in the determination of regional fat distribution than of total body fat, with a particularly strong influence on the critical visceral fat depot.
Obesity has been considered to be a polygenic disorder. This belief has been challenged by the discovery of major genes responsible for five forms of genetic obesity in mice, and an active search is under way to determine if they participate in human obesity. This dramatic entry of molecular biology into research on the genetics of human obesity will undoubtedly make a major contribution to our understanding of this disorder.
The fact that genetic influences account for only one-third of the variation in body weight means that the environment exerts an enormous influence. One measure of the extent of this influence is the dramatic increase in the prevalence of obesity during the past decade. One surprising aspect of the nature of environmental influences has become clear. Adoption studies have revealed that the early home environment, in which children learn their eating habits and which had been blamed for the origins of obesity, plays no role in determining obesity in adult life. The environment in which a person is living, however, makes a profound contribution to body weight.
Some of the most systematic studies of the environmental determinants of obesity are those of socioeconomic status. There is a strong inverse correlation between socioeconomic status and obesity, particularly among women, with lower socioeconomic status apparently favoring the development of obesity. Longitudinal studies on both sides of the Atlantic have shown that growing up in a lower-class environment is a powerful risk factor for the development of obesity, particularly for females.
The social influences on obesity involve energy intake and energy expenditure. Increased food intake plays a major role. After years of uncertainty about the contribution of food intake to the development of obesity, the introduction of doubly labeled water to measure energy expenditure, and so energy intake, has made it clear that obesity is associated with increased food intake. An increase in the fat content of the diet, from 32% in 1910 to 43% in 1985, has mirrored the increase in obesity in the United States. Dietary fat promotes obesity in at least four ways: the greater caloric density of fat (9 kcal versus 4 kcal for carbohydrate and protein); the high palatability of dietary fat; the 25% greater efficiency (compared with carbohydrate and protein) with which dietary fat is converted into body fat; and the likelihood that, unlike carbohydrate and protein, dietary fat is not regulated. Excessive intake of fat at one meal is not followed by decreased intake at the next meal, as is the case with carbohydrate and protein.
The second environmental factor promoting obesity is the sedentary lifestyle so prevalent in the United States today. A decline in physical activity has paralleled the increase in obesity, which is far more prevalent today than it was at the turn of the century. This increase in prevalence has occurred despite a substantial decrease in average food intake. The major drop in energy expenditure reflects a decline in physical activity.
This diminished physical activity has resulted from the proliferation of labor-saving devices that have transformed the nature of work and of leisure activities. It is the rare person today who puts in long hours at heavy labor, and even the home presents fewer opportunities for physical activity. The Bell Telephone Company has estimated that, in the course of a year, an extension phone saves a person 70 miles of walking, the energy equivalent of 2 lb of fat. For genetically predisposed persons, decreased physical activity means increased body weight.
Obese persons are considerably less active than are persons of normal weight. Programs that increase physical activity also decrease body weight, but because some of the decline in body weight in these programs is probably a result of diminished food intake, animal studies provide more direct evidence of the influence of physical activity. As in humans, physical activity is limited in most forms of experimental obesity, and intensifying this activity helps control the obesity. When the tendency toward obesity is strong, as in the obese hyperglycemic mouse (ob/ob) and the Zucker obese rat (fa/fa), exercise can mitigate, but not prevent, the development of obesity. When the tendency toward obesity is weaker, as in the yellow obese mouse (AY) or in rats with small lesions in the ventromedial hypothalamus, exercise can prevent the development of obesity in some cases.
GENES AND ENVIRONMENT
Genetic and environmental determinants of obesity are not in conflict. Neither acts alone to determine a clinical outcome. This outcome is determined instead by the combination of genetic vulnerability and adverse environmental events. This combination is diagrammed in Fig. 36.2, in which the small inner circle represents persons who are genetically predisposed to a disorder. The wedge represents adverse environmental conditions to which these individuals may be exposed. The model indicates that only those genetically predisposed persons who are exposed to adverse environmental conditions are clinically affected.
FIGURE 36.2. Model for the interaction of genetic vulnerability and environmental challenge. (From Stunkard AJ. Genetic contributions to human obesity. In: McHugh PR, McKusick VA, eds. Genes, brain, and behavior. New York: Raven Press, 1991:205, with permission.)
Four determinants of obesity can be classified by their effect on the regulation of body weight: adipose tissue, brain damage, medication, and psychological factors.
Adipose tissue is the organ primarily affected in obesity. Adults of average weight have approximately 25 billion fat cells, and weight gain is associated with an increase in the size of these cells (i.e., hypertrophy) but not in their number (i.e., hyperplasia). Persons who have been obese since childhood usually have hyperplastic and hypertrophic obesity; in those who are 100% or more overweight, the cell number may exceed 150 billion. When weight is lost, it is solely by a decrease in fat cell size; fat cell number appears to be irreversible. As a result, when cell size in hyperplastic obesity is reduced to normal levels by dieting, persons may still have two to five times more fat cells than persons of average weight and, accordingly, a fat mass that is increased by the same amount. This hypercellularity is a major determinant of the elevated body weight (fat) set point of hyperplastic obese persons.
A small number of persons become obese as a result of brain damage, particularly from tumors (e.g., pharyngeomas), even when they are successfully removed, and from infections (e.g., viral infections). These lesions exert their influence through two broad anatomic systems that mediate hunger and satiety, the former with representation in the lateral hypothalamus and related structures and the latter in the ventromedial hypothalamus. Damage to the ventromedial hypothalamus and its related structures results in impaired satiety, increased food intake, and gain in body weight.
Some medications, such as steroid hormones, contribute to obesity and even to altered body fat distribution. The widespread use of psychotropic medication has significantly increased the prevalence of iatrogenic obesity. Three major classes of drugs are responsible: tricyclic and heterocyclic antidepressants, lithium, and antipsychotic medication.
As many as one-half of all patients on tricyclic antidepressants discontinue desperately needed treatment because of weight gain, which may reach 2 kg per month. A major reason for the popularity of the selective serotonin reuptake inhibitors is that they are associated with far less weight gain. Weight gain with lithium is particularly troublesome, because lithium is usually prescribed for bipolar (manic-depressive) patients on a long-term basis, and weight gains of 10 kg have been reported over 2 to 6 years. Traditional antipsychotic medication also often produces weight gain, and the newer “atypical” neuroleptics may cause even larger weight gains. Since many of these medications are given on a long-term basis, weight gain is a very serious problem. A list of the psychotropic agents that promote weight gain is shown in Table 36.3.
TABLE 36.3. ANTIDEPRESSANT AND ANIPSYCHOTIC DRUGS AND BODY WEIGHT
Obesity has traditionally been viewed as a disorder with strong psychological determinants, but such determinants are confined to two specific eating disorders, binge-eating disorder and the night-eating syndrome. Binge-eating disorder is characterized by the consumption of large amounts of food in a short period, together with a subjective sense of loss of control during the binge and distress after it. Unlike patients with bulimia nervosa, these patients do not engage in compensatory behaviors, such as vomiting or laxative abuse, to prevent the weight gain after a binge, and the binge accordingly contributes to excessive caloric intake. The disorder is a source of distress, and as many as 50% of obese binge eaters suffer from depression, compared with 5% of obese persons who do not binge. Binge-eating disorder increases in prevalence with increasing body weight and afflicts about 10% to 20% of persons entering treatment programs for obesity. It may interfere with efforts at weight control and be an indication for treatment over and above that for obesity.
The night-eating syndrome, characterized by morning anorexia, evening hyperphagia, and insomnia, appears to be a manifestation of an altered circadian rhythm, precipitated by stressful life circumstances. It increases in prevalence with increasing body weight and may afflict as many as 10% of persons entering treatment programs for obesity.
Obesity has been classified on the basis of the nature of the predominant type of adipose tissue as hypertrophic, hyperplastic, or both, and this classification is still recognized. Another classification is based on the age at onset: childhood-onset obesity is characterized by a greater genetic contribution, greater adipose tissue hyperplasia, and a greater body weight than adult-onset obesity.
Body fat distribution has become the basis for a generally accepted classification of obesity. This interest was aroused in the early 1980s by the finding that persons whose fat was located primarily in the upper part of the body suffered far higher mortality and morbidity from ischemic heart disease than persons whose fat was located primarily in the lower part of the body. Body fat distribution has been estimated by the waist-hip ratio, calculated from the waist circumference halfway between the lower rib margin and the iliac crest and the hip circumference at the level of the greater trochanter. Upper-body obesity is defined as a waist-hip ratio of more than 1.0 for men and 0.8 for women. More recently, waist circumference alone has been used, with a value of greater than 100 cm considered to represent upper-body obesity. Risk, however, is directly proportional to the extent of upper-body fat, independent of gender; the mortality and morbidity rates of men are a function of their greater upper-body obesity.
Although the waist circumference is now the most widely used clinical measure of body fat distribution, a major refinement has been introduced by imaging techniques, which have shown that essentially all of the risk of upper-body obesity is conveyed by the visceral fat depot within the abdominal wall. This finding has greatly expanded our understanding of the complications of obesity and has provided a rationale for the metabolic cascade that mediates many of these complications. This rationale suggests that visceral fat, particularly under the influence of steroid hormones, gives rise to an increased free-fatty-acid flux, which leads to decreased hepatic insulin clearance, hyperinsulinemia, insulin resistance, hyperlipidemia, hypertension, and, eventually, cardiovascular disease.
The serious health hazards of obesity have received increasing attention as part of the movement to view obesity as a disease rather than as a biologic variant. McGinnis and Foege estimated that 280,000 deaths per year in the United States are attributable to “overnutrition,” making it second only to smoking as a cause of death. Some of the disorders through which this influence is exerted are described in the following sections.
Non-insulin-dependent diabetes (type II) is strongly associated with obesity; 70% of all type II diabetics are overweight or obese, and the prevalence of diabetes increases with advancing age and greater body weight. It occurs in fewer than 1% of young persons of normal weight, rising to 9% of older obese men and 13% of older obese women and to even higher rates in persons with more severe forms of obesity. Risk of diabetes is also amplified by a family history of the disorder and by upper-body-fat distribution. The good news is the remarkable improvement in insulin sensitivity and related abnormalities that is produced by weight reduction.
CORONARY HEART DISEASE
Obesity is responsible for several coronary risk factors and is an independent risk factor for coronary heart disease; compared with other risk factors, however, it is not a particularly strong one. Upper-body obesity is a strong risk factor for coronary heart disease, independent of the overall level of obesity.
Hypertension is one of the indirect measures by which obesity contributes to coronary heart disease. The prevalence of hypertension increases with greater body weight gains, with upper-body-fat distribution, and with advancing age. Weight reduction controls the disorder; the blood pressure values of most obese hypertensive patients fall to normal levels with weight losses of 10 kg. Such weight losses also permit a large percentage of obese hypertensive patients who are receiving antihypertensive medication to cut down or discontinue this medication.
Thyroid problems, which have been invoked in the past as contributing to obesity, are rarely significant. Adrenocortical hyperactivity, which gives rise to the characteristic fat distribution of Cushing’s syndrome, is usually readily recognized. Excessive amounts of estrogen, converted from androstenedione and testosterone in adipose tissue, give rise to menstrual irregularities, dysfunctional uterine bleeding, and endometrial carcinoma.
SLEEP APNEA AND THE PICKWICKIAN SYNDROME
Sleep apnea is a greatly underdiagnosed disorder that is characterized by brief periods during sleep when breathing ceases, occurring as often as hundreds of times each night. The resulting impairment in sleep gives rise to daytime somnolence, automobile accidents, cardiovascular mortality and morbidity, and premature death.
Obesity is the most important cause of sleep apnea, and weight reduction is a highly effective therapy. Recognition of sleep apnea requires first becoming aware of the possibility and then enlisting the help of someone who sleeps with the patient. Alerted to the possibility, anyone sleeping with the patient has little difficulty in making a provisional diagnosis that can readily be confirmed by study in a sleep laboratory. The severity of this disorder, the ease of diagnosis, and the effectiveness of treatment cry out for vigorous inquiry into the possibility of sleep apnea in all obese patients.
The obesity-hyperventilation syndrome, also called the pickwickian syndrome, is a sleep-related disorder of greater severity, which is characterized by daytime hypercapnia and cor pulmonale and by a greater risk of premature death.
Gallstones occur in about 30% of obese women, compared with 10% of nonobese women. Unlike other obesity-related disorders, weight reduction can exacerbate the problem, presumably because of supersaturation of the bile with cholesterol.
Obesity is strongly associated with joint symptoms; arthritis of one knee is six times more common and arthritis of both knees is 15 times more common among obese than among nonobese persons. In addition to its association with arthritis of the weight-bearing joints, obesity is also associated with arthritis of non-weight-bearing joints, such as the fingers. Weight loss is remarkably effective in controlling symptoms and in significantly lowering the risk of osteoarthritis.
Although obesity has been viewed as a result of deep psychological problems, such problems of obese persons are now understood to be a consequence of obesity and of the stigma with which it is associated. In addition to the eating disorders described previously, an emotional problem specific to obesity is disparagement of the body image. Persons with this disturbance feel that their bodies are grotesque and loathsome. The problem most commonly affects middle-class white women, among whom the prevalence of obesity is low and the sanctions against it very high. It is confined to those who have been obese since childhood, in whom neurosis is closely related to their obesity.
PHYSICAL EXAMINATION AND LABORATORY FINDINGS
Before weight reduction, significantly obese individuals should undergo a physical examination and routine blood tests (biochemical profile) to rule out the neuroendocrine and other abnormalities described previously. Decreases in medications, such as antihypertensive and antidiabetic agents, that are anticipated with weight loss should be reviewed.
With significantly obese individuals, it is not necessary to assess the body composition with specialized techniques, such as hydrostatic weighing, skinfold measurement, and bioelectric impedance. The patients can be presumed to have excess fat. Such assessment may be useful with athletic, mildly overweight individuals, who may have a surfeit of lean rather than fat tissue. Body composition assessment is commonly available at health clubs and sports medicine clinics.
The resting metabolic rate should be measured for individuals who report weight gain on low-energy intakes. Objective confirmation of low-energy requirements is often reassuring to patients and indicates the desirability of increasing physical activity rather than severely restricting caloric intake. Repudiating the reported low-energy requirements may reveal problems of distortion and denial. The composition of the patient’s diet should be assessed by a 3-day food diary, recorded just before the office visit. The objective is to determine the fat content of the diet, the extent to which the patient eats at regular times, and the presence of binge eating.
After completing the medical evaluation, the physician should summarize the findings for the patient and then invite questions. They should then discuss the need for weight loss and the goals and methods of treatment.
NEED FOR WEIGHT LOSS
Independent of the degree of obesity, the need for weight loss is greater in overweight persons who have upper-body obesity (particularly increased visceral fat), persons who have one or more obesity-related complications (hypertension, diabetes, hyperlipidemia), those who are younger (20 to 45 years of age), males, and individuals with a family history of an obesity-related illness. The greater the number of these risk factors, the more pressing the need for weight reduction. The absence of these factors minimizes the need for treatment. A goal of weight stability, rather than weight reduction, would seem appropriate for a 60-year-old woman with lower-body obesity who is 15 kg overweight but otherwise in excellent health.
GOALS OF TREATMENT
Persons with health complications from obesity should be encouraged to lose weight—but how much weight loss is necessary? Historically, overweight individuals have been told to reduce to “ideal” weight, as defined by height-weight tables. Obese individuals, however, cannot reach this goal, which has been superseded by a weight-loss goal of as little as 5% to 10% of their initial weight. Reductions of this magnitude are sufficient to improve or control weight-related complications, including hypertension, diabetes, and hyperlipidemia.
METHODS OF TREATMENT
The review by the Institute of Medicine of the National Academy of Sciences divided methods of treatment into three categories: do-it-yourself programs that include diet books and self-help approaches, such as Overeaters Anonymous; nonclinical (commercial) programs, such as Weight Watchers and Jenny Craig; and clinical programs that provide medical care and more aggressive measures, such as very low calorie diets (VLCDs) and drug therapy.
A primary task of physicians is to manage the health complications of their obese patients. They may also choose to manage their patients’ weight-control efforts. Such efforts are most successful with highly motivated, mildly obese individuals who have become obese as adults. Significantly obese persons usually require greater structure and support. The physician should help such patients evaluate the many available treatments and select the most appropriate one. Thereafter, the physician’s task is to monitor changes in patients’ health and to encourage them to attend treatment sessions and improve their eating and exercise habits.
An overview of current treatments for obesity is provided in Fig. 36.3. The model combines a classification based on body weight with a stepped-care approach and efforts to identify the patient’s specific treatment needs. The more obese the individual, the more structured and possibly aggressive is the therapy required. We think all significantly obese individuals should be treated initially by a 1,200 to 1,800 kcal/day diet of conventional foods combined with a program of behavior modification. Such treatment is provided by Weight Watchers to over two million people a week, at a cost of as little as $12 a week. More intensive and costly versions of this behavioral approach are frequently offered, with better results, by hospital- and university-based programs. If the patient has already received this kind of treatment, the next intervention in the stepped-care model should be considered. The following sections describe behavior therapy, VLCDs, pharmacotherapy, and surgery. Results of nonclinical (commercial and residential) programs are not discussed, because of a lack of adequate data.
FIGURE 36.3. A conceptual scheme showing a three-stage process for selecting treatment. The first step, the classification decision, divides individuals into four levels, according to their percentage of overweight. These levels dictate which of the five steps would be reasonable in the second stage, the stepped-care decision, which indicates that the least intensive, costly, and risky approach will be selected from among the options. The third stage, the matching decision, is used to make the final selection of a program and is based on client and program variables. The dashed lines with arrows between the classification and stepped-care stages show the lowest level of beneficial treatment, but more intensive treatment is usually necessary for persons at the specified weight level. (From Brownell KD, Wadden TA. The heterogeneity of obesity: fitting treatment to individuals. Behav Ther 1991;22:162, with permission.)
As applied to obesity, behavior therapy refers to a set of principles and techniques for the modification of eating and exercise habits. This approach is goal oriented; the objectives and methods of treatment are clearly specified in weekly homework assignments that patients discuss with their counselors each week. The techniques are used to facilitate patients’ adherence to any one of a number of dietary regimens. Patients are usually asked to consume their customary foods but to reduce their caloric intake by 500 to 700 kcal/day and their consumption of fat to no more than 30% of total calories. In the initial weeks, patients keep daily records of the types and amounts of foods that they eat. Later, record keeping is expanded to include information about times, places, and feelings associated with eating.
Patients are asked to increase their physical activity by a variety of changes in lifestyle, such as walking more, using stairs rather than escalators, and reducing their dependence on energy-saving devices, for example, extension phones and remote-control devices. Most patients also adopt a structured exercise program, such as walking or swimming, but are cautioned not to make the program heroic or punishing. The behavioral approach is most effective when delivered in groups of 10 to 12 persons in which participants discuss their weekly homework assignments. Individual treatment may also yield adequate weight loss, although it does not usually provide the emotional support of group care. A favorable outcome is facilitated in either case by the use of a structured treatment manual, such as Brownell’s LEARN Program for Weight Control.
Patients treated by group behavior therapy lose an average of 8 kg in 15 to 20 weeks, equal to a loss of 0.5 kg (1 lb) a week. Longer therapy produces larger losses, but they rarely exceed 15 kg, even when treatment lasts as long as 52 weeks. When patients stop treatment, they regain approximately one-third of their weight loss in the year after therapy, with increasing regain over time. Weight regain can be minimized, if not prevented, if patients maintain frequent, regular contact with their providers after weight loss. Contact can take the form of group meetings, telephone calls, or postcards. Increased physical activity is also a key component of long-term weight control. Persons who engage in regular physical activity for approximately 2 to 3 hours a week are the most likely to maintain reduced weights.
Very Low Calorie Diets
VLCDs, providing 400 to 800 kcal daily, are reserved for patients 30% or more overweight who have failed to lose weight satisfactorily with conventional methods. Candidates for VLCDs should be referred to multidisciplinary centers that provide medical supervision and a comprehensive program of behavior change. In such programs, patients lose an average of 20 to 25 kg and maintain approximately 60% of the weight loss 1 year later. Better results are achieved by combining the diet with a sound exercise program.
We stand at the threshold of a major new emphasis on the treatment of obesity with medication. This development arises in part out of the growing awareness of limitations of the behavioral programs and in part out of a radical revision of our views on medication that had led to its virtual abandonment in recent years. These views had two sources. The first was the revelation, two decades ago, of the serious abuses of amphetamines by unscrupulous so-called diet doctors. The result was the banning of amphetamines and a concurrent disinclination on the part of physicians to use other safe and effective medications. The second cause for the abandonment of medication was the widespread and mistaken belief that tolerance develops to the effects of appetite-suppressant medication. As a result, state medical boards proscribed their use for periods of longer than 3 months.
This belief was mistaken; appetite-suppressant medication maintains its effectiveness for as long as it is used. As soon as it is discontinued, however, weight, which had been kept down by its use, rapidly rebounds to pretreatment levels. These newer findings reverse the old rationale for treatment and suggest that appetite-suppressant medication should be used on a long-term basis or not at all. Clinical trials have strongly supported this rationale. They have produced weight losses of 10% of body weight within 6 months, and these losses have been generally well maintained for as long as medication is continued; when it is discontinued, weight is promptly regained. Changes in the regulations for use of appetite-suppressant medication are already under way, and it seems likely that long-term use will become standard practice. Other, older medications have proved safe and effective. These agents and their dosages are listed in Table 36.4. Most are noradrenergic, and one of them, phenylpropanolamine, is available over the counter.