Our guest discussant today is Dr. Edward Saltzman who is an Instructor at the Tufts University School of Medicine and a member of the Division of Clinical Nutrition. Dr. Saltzman is Medical Director of the Obesity Consult Center at New England Medical Center and his research interests pertain to the control of energy metabolism.

JM

I would like to begin our discussion with Dr. Saltzman by asking him to define obesity and to tell us whether our definition of obesity has changed over the past decade.

Definition of Obesity

ES

Obesity is a condition that is easy to recognize but difficult for many practitioners to precisely define. Technically speaking, obesity is an excess of body fat. However, in the clinical situation, measuring body fat is rather difficult. This entails the use of specialized calipers or more detailed and involved methods, such as dual photon absorptiometry (DEXA). We routinely use a surrogate, such as body weight, for the measurement of body fat. Historically, a measure of body weight compared to a norm is what passed for the standard of obesity. The most common reference point for "ideal weight" has been the Metropolitan Life Insurance Company's Weight Tables.

A better measure of weight, however, is the body mass index (weight in kilograms divided by the height in meters squared). The BMI is a relatively height-independent measure of body weight. By convention, a body mass index of 30 or above is designated as obesity. For some of the large surveys done in the US and other countries, overweight is defined as a body mass index greater than 27.8 or 27.3 for men and women, respectively, since the co-morbidities that accompany obesity start to increase between the 27 to 30 range.(1) There is a good correlation between a body mass index and excess body fat and thus we feel comfortable with that definition.

Prevalence

JM

Thanks for clarifying the definitions, Ed. Another issue, which has been quite prominent in the news, is prevalence. I have seen several news reports suggesting that the prevalence of obesity in North American society is increasing. Is this really true, or is this increase simply related to changes in how we define obesity?

ES

There should be no doubt in anyone's mind that the prevalence of obesity is increasing in North America, Europe and even in some parts of the developing world.(2) Obesity has become so frequent in the US that it is now often called an "obesity epidemic." The prevalence of obesity in the US is approximately 30%. The prevalence of obesity in developing countries is also increasing at an alarming rate. In some minority populations, the prevalence of obesity approaches 50% in some age groups.

JM

I would also like to add that there are now many reports indicating an alarming increase in the prevalence of pediatric obesity as well.

Co-morbidities

There are a myriad of both psycho-social as well as medical complications of obesity. Some of the common ones include an increased prevalence of type II diabetes, hypertension, and osteoarthritis. Are there other co-morbidities associated with this condition that are frequently forgotten?

ES

There are two types of very common co-morbidities which are often overlooked by primary care practitioners. The first is sleep apnea. Many moderate and severely overweight individuals have some degree of obstructive sleep apnea. In some series, up to 40% of overweight individuals have this disorder. We always ask our patients about the common symptoms, such as daytime somnolence and disturbed sleep in the evening, but we also ask about more subtle changes, such as diminished cognitive capacity and mood disturbances. I think you will find a mild, smoldering type of depression is quite common in obese individuals and I strongly believe that disturbed sleep contributes to this in many persons.

This leads to the second under-diagnosed type of co morbidity -- the psycho-social. If your obese patients are not dysphoric prior to becoming obese, they are likely to suffer from dysphoria with weight gain. The dysphoria often does respond to various types of treatment, including psychotherapy and antidepressant medications. Importantly, effective treatment of the patient's underlying mood enhances the ability to treat the obesity and allows the patient to lose weight.

JM

Ed, I am surprised by the remarkable prevalence of sleep apnea in this disorder. Do you think there is a role for routine screening for sleep apnea, such as a formal sleep test, for individuals over a particular BMI?

ES

I think that we need to treat the patient and not just look for and treat the disorder. If a patient has significant symptoms consistent with sleep apnea, we should proceed with a sleep study. If a patient does have some symptoms that are consistent with sleep apnea, but is functioning well during the daytime and does not appear to have markedly disturbed sleep, I think that one can avoid formal diagnosis and treatment. One exception to this may be the obese patient with hypertension who also has symptoms of sleep apnea, since there appears to be a link between hypertension and sleep apnea. Even those patients who appear to be functioning well during the day but are hypertensive should be considered candidates for a formal sleep study.

Pharmacologic Treatments

JM

I know that our audience would be particularly interested in hearing about the remarkable attention that the lay press has paid to the use of new drugs in the treatment of obesity. Given the great concern over some of the potential side effects of these drugs, is there any role for drug treatment in the treatment of obesity that is left to us at the present time?

ES

Yes, Joel, there is a place for drug treatment in obesity.(3) Let me first say that drug treatment for obesity is an adjunct to the basics of obesity treatment -- dietary modification, increase in physical activity and lifestyle changes. What medications add is the ability to better achieve some of the goals set forth in those basic treatments.

As of February 1998, the currently available medications include the adrenergic, anorectic agents, such as phentermine, which work on the central nervous system to decrease appetite and food intake. A new drug, sibutramine, has been approved but it is not yet commercially available. This agent is a selective serotonin re-uptake inhibitor, as well as a noradrenergic agent, which also acts on the central nervous system to decrease food intake and increase satiety. Sibutramine combines both the serotonergic, as well as noradrenergic aspects, of the previous fen-phen combination. Sibutramine promotes a 5-10% decrease in body weight, significantly greater than placebo, when used over a 6-12 month period.(4) This is fairly consistent with the benefit conferred by other drug regimens which are no longer available. This 5-10% decrease in body weight may seem small, compared to the total degree of overweight, but really a 5-10% decrease in body weight confers dramatic improvement in some co-morbidities and, therefore, this is a reasonable goal.

There is another drug, tetrahydrolipstatin (Orlistat^®), a pancreatic lipase inhibitor, which has not yet been approved by the FDA. Orlistat^® contributes to weight loss not by acting on the central nervous system, but by promoting the excretion of approximately 30% of ingested fat. The weight loss attributable to Orlistat^® is largely that of the excreted fat, although given the side effects that can be experienced if OrlistatAfAE?Ac€A!Af‚A?® is taken with a high fat diet, this can be viewed as the "Antabuse" for a high fat diet. Common side effects of Orlistat^® with the high fat diet include: oily bowel evacuation and spotting.

JM

Let's return, for a moment, to the central nervous system acting drugs. Can you discuss whether the grave concerns over valvular heart disease apply to either phentermine or subutramine and, if not, what are the side effects that physicians should be aware of, if they are going to prescribe either of these drugs? Lastly, maybe you could return to Orlistat^® and discuss some of the other concerns about potential side effects of that drug.

ES

As most practitioners are aware, the fenfluramine derivatives, both the d,l form and the d (dex) form, were withdrawn from the market due to concerns about valvular heart disease. More recent reports appear to suggest that the incidence of valvular lesions with these agents may be less than the 30% prediction based on initial reports. Nonetheless, this research is yet to be completed. The use of phentermine, as a single agent, is not associated with valvular heart disease, at least thus far. However, the use of phentermine with Prozac^®, the so-called "phen-pro" combination, is not recommended because the combination of phentermine with the serotonergic agent, Prozac^®, may mimic the effects of the fen-phen combination. While there is no evidence to state that this is so, there is no evidence yet to refute this proposition.

The anorectic agents, as a class, appear to be associated with a small risk of primary pulmonary hypertension. This disorder is estimated to occur in approximately one in 20,000 individuals who take these drugs. Because of the rarity of this disorder, we have little information about specific agents, such as phentermine, or the newly approved agent, sibutramine. However, it would be prudent to consider all these agents as possibly inducing pulmonary hypertension until proven otherwise. The new agent, sibutramine, has not been associated with either valve disease or pulmonary hypertension, thus far. Because it is a combination serotonergic and noradrenergic agent, when a larger number of patients are exposed to sibutramine the rare occurrence of these disorders may emerge. Some believe that since sibutramine is a serotonin reuptake inhibitor and it does not increase neurotransmitter secretion, valve disease and pulmonary hypertension may not occur.

The common side effects of sibutramine reflect both its noradrenergic as well as its serotonergic profile. Dry mouth and constipation are reported and there may be some insomnia or headache. Of most concern would be increases in blood pressure or heart rate. While the average increase in blood pressure is just a few millimeters, some patients may experience a rise greater than 15 millimeters. For this reason, uncontrolled hypertensive patients should not receive this medication and all patients, especially hypertensive patients, should have frequent monitoring of blood pressure. Use in ischemic heart or cerebrovascular disease may not be worth the potential gains.

The side effects of Orlistat^® are almost entirely attributable to its mechanism of action -- blocking the digestion and absorption of fat in the intestinal lumen. While the occurrence of these side effects is common when patients first start taking Orlistat^®, it appears that patients quickly learn the degree to which they can tolerate a small amount of fat in the diet and adhere to that thereafter. The other complications of OrlistatAfAE?Ac€A!Af‚A?® theoretically result from decreases in absorption of some of the fat soluble micronutrients and, as such, the manufacturer recommends the routine use of a multivitamin supplement. In the pre-marketing trials of Orlistat^®, small (and probably clinically insignificant) decreases in the levels of some of the fat soluble vitamins were seen and were stable over a period of over one year. However, the long term effects of OrlistatAfAE?Ac€A!Af‚A?® on some of the more obscure or unmeasurable micronutrients remain unknown.

JM

Ed, given the array of potential side effects that you have spoken about, do you think primary care physicians should be prescribing these drugs or do you feel these drugs should only be administered by someone with particular expertise in the treatment of obesity?

ES

Successful treatment of obesity requires attention to diet, physical activity, and lifestyle factors. Most of a physician's time is spent addressing these factors. If a primary care practitioner feels comfortable adding a medication to existing treatment, then I see no problem with them doing so. The safety profile of these medications, at present, is such that I believe primary care providers can successfully use them. I would emphasize, however, that if your plan is to prescribe a medication to a patient and not provide the other services yourself or by referral to other individuals in your practice, then drug therapy is likely not to succeed.

Other Treatment Modalities

JM

Ed, I am certain that you want to emphasize the need for other modalities in obesity treatment. Would you care to discuss these in a bit of detail?

ES

As I said before, the cornerstones of treatment for overweight are long term dietary changes and increased physical activity. These changes are really best brought about by a comprehensive plan to address lifestyle issues. This sort of treatment can be undertaken by almost anyone, whether a physician, a registered dietitian, or a psychologist. What it takes, really, is paying attention to all the factors associated with an individual's given lifestyle that promote overeating.

Bariatric Surgery

In addition to the basic treatments noted above, we also, at times, recommend people for bariatric surgery. The most commonly done surgery is the roux-en-y gastric bypass. This is a treatment option for those individuals who are greater than 200% of ideal weight or whose body mass index is greater than 40 kg/m.(2) For individuals in this weight class, this is the single treatment most likely to result in a decrease to a healthy weight and maintenance of this healthy weight. Obviously, such a dramatic change in one's capacity to eat, via surgical intervention, is not for everyone.

Genetic Determinants

JM

Ed, as is true of most fields of medicine, there is increasing attention to genetic determinants of obesity. Could you explain for us where this field sits in regard to this issue?

ES

Large epidemiologic studies, as well as studies in families and in identical twins, demonstrate that the heritability of body mass index is probably in the neighborhood of 30-40%. The knowledge that body composition has a significant genetic determinant doesn't help us, however, with an approach to individual patients. Clearly, there will be families in which obesity is very common. However, what enables one member of that family to lose weight will not necessarily work for the others.

There are a number of factors that are probably under genetic control that are thought to influence body weight. These include the protein, leptin, as well as other hormones such as neuropeptide Y. Leptin is a protein produced in fat cells that communicates with the brain and which decreases energy intake in animal models such as the ob/ob mouse. An absence of this protein in animals results in severe obesity. To date, in humans, there have been extremely few individuals found to be deficient in this protein as a result of genetic deletions. The leptin story may be analogous to that of type II diabetes, in that some overweight individuals produce leptin but are insensitive to its actions. Partly because of this rationale, trials are now underway to look at pharmacologic doses of leptin in overweight individuals to see if this is an effective way to reduce body weight. The bottom line about leptin and its effect on humans is still unknown.

Prevention

JM

There is a growing paradigm which applies to nearly all chronic illnesses -- that it is more important to prevent an illness than treat it. In this regard, could you comment a bit on some of the effective ways that we might think about preventing obesity?

ES

I think prevention is a critically important concept because treatment of obesity is so difficult. The current treatment regimens are often ineffective and long term weight control is extremely difficult for most individuals. Prevention, therefore, should constitute an important role in the primary care provider's office. Individuals who start to gain weight (even as little as 5-10 pounds) should be counseled to try and control their weight. It is much easier to keep those individuals from gaining further weight than it will be to reduce their weight once they are 30-40 lbs. overweight. Given that prevention may be the most effective approach, it is very surprising that there are few large scale trials in the primary practice setting to document that the weight gain observed in a number of individuals can be stemmed at an early point. Nonetheless, I think it is very important to try and do this in your individual practices.

This same strategy should also apply to pediatric populations. There are now substantial data demonstrating that childhood obesity is associated with subsequent adult obesity, as well as subsequent health problems in adulthood. For this reason, it is important to instill in children, both by example and by teaching, the elements of a healthy lifestyle, including appropriate diet and appropriate levels of physical activity.

JM

I would like to thank you very much for all your insights and comments regarding this very difficult and frustrating condition that all of us, as physicians, will be dealing with on an increasing basis in the decades to come.