For primary care physicians, obesity is one of the most challenging problems confronted in office practice. The disorder is increasing in prevalence despite the efforts of both patients and physicians. Treatment requires a multimodality approach that addresses diet, physical activity, and behavioral issues. Medication and surgical approaches may be appropriate as well. This review outlines the evidence for each approach, suggests how primary care physicians can best help obese patients, and provides practical tips for weight loss.
Obesity is a rapidly growing epidemic worldwide, and it increases the risk of morbidity and mortality. In the United States, obesity may be responsible for as many as 300,000 deaths per year, although this figure is controversial. Direct medical costs in the United States (1999) are estimated to be $70 billion annually. Despite the expenditure of billions of dollars in weight-loss products, the epidemic is getting worse.
The most common definition of obesity is a body mass index (BMI) (calculated as weight in kilograms divided by the square of height in meters) of 30 kg/m2 or greater. Overweight is defined as a BMI between 25.0 and 29.9 kg/m2, and extreme obesity is defined as a BMI of 40 kg/m2 or greater. Approximately one third of Americans are overweight, one third are obese, and 4.5% have extreme obesity. Unfortunately, physicians often do not formally diagnose obesity when it exists (Bardia A, Holtan SG, Slezak JM, Thompson WG, unpublished data, 2006). Assessment of patients for obesity can be facilitated by having office staff enter the BMI into the patient’s record after measuring height and weight. Measuring waist and hip circumference should also be considered because these parameters affect outcome independent of the BMI. Physician diagnosis of obesity is important because a management plan is more likely to be formulated than when no diagnosis is made. This review focuses on optimizing that management plan.
Weight loss requires a sustained negative energy balance: energy output must exceed energy intake. We recommend a 2092-kJ (500-kcal) daily deficit, which can be expected to lead to a weight loss of 0.45 kg per week. Although it is theoretically possible to achieve this amount of weight loss through either reduced energy intake or increased energy output, a 2092-kJ daily deficit is best achieved and sustained by a combination of the two. This review focuses first on strategies for reducing energy intake (diets, drugs, and bariatric surgery), then discusses strategies for increasing energy output (exercise and nonexercise movement), and concludes with guidelines for how primary care physicians can help their obese patients successfully change their energy intake and physical activity level.
Many different diets have been advocated for weight loss, but there is little scientific evidence to recommend one diet over another. Dansinger et al compared 4 approaches, the Atkins (low carbohydrate), Zone (high protein, low carbohydrate), Ornish (very low fat), and Weight Watchers diets, and found no significant difference in weight loss at 1 year. Other studies have shown greater weight loss at 3 months with low-carbohydrate than with other diets but no significant difference at 1 year. While some small studies of low-glycemic index diets have shown a benefit others have found little effect beyond that of energy restriction. A high-protein diet may facilitate weight loss, but more and larger studies are needed to confirm this hypothesis. Calcium and dairy products have also been postulated to facilitate weight loss, but to date most randomized trials have not confirmed this theory.
One strategy for reducing energy intake is to reduce fat content. The Women’s Health Initiative Randomized Controlled Dietary Modification Trial randomized 19,541 women to a low-fat diet with increased amounts of fruits, vegetables, and whole grains and 29,294 women to a control diet. The study found a significant correlation between dietary fat reduction with increased fruit and vegetable consumption and weight loss. Further evidence for a reduced-fat approach comes from the National Weight Control Registry, a self-report registry of subjects who have lost at least 13.6 kg and maintained that loss for 1 to 5 years. These subjects reported that only 25% of their total energy intake came from fat, which is considerably lower than the national average of 37%. However, fat restriction without energy reduction will not result in weight loss.
Reducing fat in the diet reduces the energy density of the diet. Energy density is the ratio of energy provided (calories) by a food to its weight. Foods with low energy density such as fruits and vegetables provide considerable bulk with minimal energy intake (filling but not fattening). Thus, a pound (0.45 kg) of carrots, which have low energy density, has the same amount of energy value (calories) as an ounce (28 g) of peanuts, which have high energy density. A crossover study compared the effect of reducing portion size with the effect of reducing energy density on energy intake. When portion size was reduced by 25%, energy intake declined by 10%, but when energy density was reduced by 25%, energy intake declined by more than 20%. The group randomized to a diet of foods with reduced energy density consumed more food but less energy than the group whose diet consisted of reduced portion size. The subjects’ ratings of hunger and taste did not vary across the comparisons. Although more investigation is needed, there is some evidence that increasing the amount of low-energy density foods in the diet is effective for long-term weight management as well.
In summary, long-term weight loss and weight maintenance require a reduction in energy intake. We believe this is best achieved by a combination of reducing total fat intake, reducing portion size, reducing energy density, and increasing fruit and vegetable intake. However, in the absence of behavior modification and continued input from health professionals, diets are ineffective in the long term. Counseling patients regarding behavior modification that will successfully reduce fat and energy intake is addressed subsequently in this article.
The role of medications in weight loss is controversial, and their effectiveness appears to be limited. First, the amount of weight lost with use of drugs is small (as discussed subsequently). Second, the long-term safety of weight-loss drugs is not established, and the occurrence of adverse effects (such as the cardiac valve abnormalities associated with fenfluramine) suggests that this is an important consideration. Finally, when weight-loss drugs are discontinued, weight is regained. Because no weight-loss drug has been approved by the Food and Drug Administration (FDA) for use for more than 2 years, drugs represent a short-term solution to a long-term problem with only modest benefit and with unclear risk. The decision to prescribe medication hinges on whether the patient will become more or less motivated to make the long-term changes in eating and activity that are necessary to lose weight and keep it off. Until safety data are available, physicians should refrain from prescribing medications for weight loss for durations longer than those approved by the FDA.
Phentermine, benzphetamine, and phendimetrazine are FDA approved for short-term use (12 weeks) for weight loss. Meta-analysis suggests a short-term weight loss of 3.5 kg with phentermine, but no long-term data are available. Until long-term studies demonstrate the effectiveness and safety of these drugs, they are best avoided.
Sibutramine is a norepinephrine and serotonin reuptake inhibitor that is thought to lead to reduced food intake. In a meta-analysis of 29 trials, sibutramine use resulted in a weight loss of 4.5 kg at 1 year. A more recent study found that sibutramine use alone resulted in a weight loss of 5 kg at 1 year compared with 6.7 kg with lifestyle modification alone and 12.1 kg with sibutramine plus lifestyle modification. Blood pressure and heart rate increase modestly with sibutramine use and should be monitored. Palpitations may also occur. Sibutramine should not be used by patients with cardiovascular disease, heart failure, or arrhythmias or by patients taking selective serotonin reuptake inhibitors or monoamine oxidase inhibitors. No studies lasting longer than 2 years have been reported, and sibutramine is approved by the FDA for 1-year use.
Topiramate and rimonabant are associated with weight loss but are not yet approved by the FDA for this purpose. Topiramate, which modulates γ-aminobutyric acid receptors, was associated with a 6% weight loss in a meta-analysis (24-week data with several studies reported in abstract form) but was also associated with a significant incidence of adverse central nervous system effects (especially paresthesias and loss of taste). Rimonabant is a cannabinoid-1 receptor blocker not yet approved for use in the United States. Two large randomized trials have demonstrated efficacy, reporting a 6-kg weight loss at 1 year. Because both studies had a considerable number of dropouts, the reported weight loss may be optimistic. The incidence of psychiatric problems (depression and anxiety) in the patients taking rimonabant was significantly higher (approximately double) than that seen in the placebo groups, a finding that warrants further investigation.
Impairment of Energy Absorption
Orlistat prevents absorption of a portion of the energy fraction from ingested fats. Many randomized trials of orlistat have been conducted, and most have shown greater weight loss with orlistat than with placebo. However, meta-analysis suggests that the mean weight loss is only 2.89 kg. When patients are switched from a weight-loss to a weight-maintenance diet, they regain less weight if they continue orlistat than if they switch to placebo. The studies in the meta-analysis showed a reduction in low-density lipoprotein cholesterol, glucose, insulin, and hemoglobin A1c (in diabetics) with orlistat, which occurred both as a result of weight loss and independent of weight loss (probably because the orlistat group absorbed less fat). Orlistat improves glucose tolerance and reduces the rate of progression to impaired glucose tolerance and diabetes. Orlistat improves alanine transaminase levels and steatosis in patients with nonalcoholic fatty liver disease independent of weight loss. Gastrointestinal adverse effects such as diarrhea and oily stool are common (especially if the diet is high in fat) but usually subside with time, and serious adverse events appear to be rare. Orlistat costs more than $100 a month and may not be covered by insurance. It is best suited for patients who can comply with a low-fat diet, are at high risk of developing diabetes, and can afford the medication. The longest published study of orlistat use is 4 years, and the drug is FDA approved for 2-year use. Approval for over-the-counter formulations (half dose) is expected soon.
Combined treatment with multiple weight-loss drugs has received little evaluation to date and cannot be recommended at this time. One study that combined orlistat and sibutramine showed no greater weight loss with the combination regimen than with sibutramine alone.