One of the most common and distressing side effects associated with combination therapy today concerns visible changes in body composition and appearance. Although some aspects of this phenomenon were reported in earlier years of the epidemic, the frequency of reports has greatly increased since the beginning of the protease inhibitor era in 1996.

These changes in appearance are brought about by several different forms of redistribution of body fat -- fat moving away from the face, arms, butt and legs, and then collecting around the gut (central obesity), at the base of the neck (buffalo hump) and/or in the breasts (breast enlargement). Less frequently seen are accumulations of lumpy fat tissue, similar in appearance to cellulite, on the back below the shoulder blades. Associated with these changes, and possibly causally related, are striking increases in a number of blood test markers, such as triglycerides and cholesterol level, and insulin resistance. These blood changes have sometimes been linked to increased risk of heart disease and to the development of diabetes.

The cause of these changes is unknown or at least uncertain. Is it a consequence of HIV disease progression? Is the virus itself interfering with the way the body processes fat and protein? If so, why is it more common today than before? Are these changes a result of protease inhibitor therapy, as is commonly thought? If so, why have they occurred in some people who haven't used protease inhibitors, or any anti-HIV therapy at all? Are some of the drugs used today more or less likely than others to contribute to these problems? Are there any treatments that can prevent or treat these changes? What is the long-term consequence of these changes, and what percentage of HIV infected people are experiencing them?

Sadly, all these questions currently remain unanswered. Changes in body composition and all laboratory (e.g. blood work results) abnormalities associated with these changes have not been clearly defined, but are currently grouped together and referred to as "lipodystrophy" (pronounced, lip-oh-dis-troh-fee).

Can Lipodystrophy Be Treated?

While lipodystrophy clearly affects an increasing number of people, there have been no controlled studies investigating how to treat the condition. Reports of individual successes cannot be considered predictive of whether the treatments will work for others.

Three separate reports claim some success in treating lipodystrophy associated with the use of protease inhibitor-containing regimens. Those reports occurred as a result of switching people off protease inhibitors and substituting a non-nucleoside reverse transcriptase inhibitor (NNRTI) (see "Drug Identification Chart"). Physicians reported some reductions in central obesity (fat around the gut); however, not everyone had a return of peripheral fat (fat in the arms and legs). Reductions were also reported in lipid levels (cholesterol and triglycerides) and a reversal of insulin resistance (which is associated with diabetes).

One potential area of concern from a study showed that 10% of the group who switched to a NNRTI-class drug had increases in HIV levels. This could be just coincidence. It is possible these people would have had an increase in HIV levels if they stayed on their protease inhibitors. Nevertheless, this observation does cause some concern. Despite these seemingly positive findings, the number of people involved was very small and some individual physicians have reported contradictory results in their own medical practices.

A physician in New York has observed that growth hormone (Serostim®) had some effect in reducing central obesity and buffalo hump in a few patients. But Serostim® had no effect on facial and limb wasting or on decreasing lipid levels. One person had to stop growth hormone therapy because of side effects and had a rapid return of central obesity. It was again reduced when the growth hormone was restarted. Considering the extremely high cost of Serostim® and the apparent need for continuous therapy, it is hard to consider this approach particularly hopeful.

A number of people have reported success with using liposuction technology to remove disfiguring buffalo humps; however, there is every reason to fear that such humps will slowly grow back over time, since the underlying cause has not changed. Liposuction is not recommended for treatment of central obesity, since the fat is stored deeply behind the abdominal muscles and cannot be easily removed. Some women have resorted to breast reduction surgery when excessive growth lead to pain and difficulty walking. Surgical solutions such as these should only be considered in the most extreme cases, especially since their long-term success is unknown.

Several small studies looked at the use of specific drugs to treat some of the laboratory abnormalities associated mainly with the use of protease inhibitor therapy. One study showed that the combination of gemfibrozil (Lopid®) and atorvastatin (Lipitor®) lowered lipid levels to the normal range in about 50% of people. Another study showed that metformin (Glucophage®) reduced central obesity and insulin resistance but also led to an average 2kg weight loss. Finally, one other study showed that troglitazone (Rezulin®) reduced glucose levels but had no effect on lipid levels.

Many of these therapies have potential drug interactions with protease inhibitors. Doctors and patients who consider experimenting with these approaches to manage lab abnormalities should talk to a pharmacist about possible drug interactions and any necessary dose adjustments to therapies.


It is difficult to conduct studies to treat lipodystrophy because of the need for an agreed upon case definition. An Australian group is likely to have some new results by the middle of this year that may give us a better definition for lipodystrophy. A large US study will start shortly to look at the prevalence of lipodystrophy. That study may also provide us a better working definition to use in other studies as well as in clinics.

To date, much of the effort to study lipodystrophy has been driven by pressure from AIDS activists. Even though the problem has been evident for nearly three years, it has taken a very long time for pharmaceutical companies and the research community in general to take positive steps toward further understanding it. It appears that community pressure has once again been helpful since several new studies will soon be underway, by both government and industry.

"When lab marker changes reach what appear to be life threatening levels, physicians should act appropriately, just as they would any other time these tests show alarming results.

Many groups are studying the cause of lipodystrophy. Glaxo Wellcome scientists have recently reported that they have conducted laboratory studies which seem to have identified two possible causes of fat redistribution. Both mechanisms are similar to, but not exactly the same as, that posed by the Australian group previously. The Glaxo theory suggests that the problem might be caused by one mechanism by ritonavir, saquinavir, and nelfinavir, and/or a somewhat different mechanism related to indinavir. For the moment, they believe that their own protease inhibitor, amprenavir, poses neither problem, but only time will tell if this is accurate or merely self-reporting.

Unfortunately, in the meantime, there is very little information available for people on how to diagnose and treat this syndrome. For most people, fat redistribution does not become physically dangerous. However, triglyceride and cholesterol levels can become so severely elevated that many physicians worry about increased risks of heart disease and other serious conditions. Careful and frequent monitoring of lab tests, along with regular physical examinations, should become part of the medical routine for people on combination therapies. When lab marker changes reach what appear to be life threatening levels, physicians should act appropriately, just as they would any other time these tests show alarming results.

It is unclear just how widespread these problems are, with various groups reporting incidence levels ranging from roughly 15% to as high as 75%. This widespread difference probably reflects variations in the underlying definition used for lipodystrophy. There is as yet no reason to suspect that these problems will affect everyone, but there is certainly enough evidence to suggest that a serious problem exists which demands greater attention than it has been given.

What Is "Lipodystrophy"?

Currently there is not an official definition for lipodystrophy, which makes it difficult to study in any systematic way. The following "working" definition lumps together the thinking of a number of groups. Much of this information may prove useful for doctors and their patients in monitoring for possible early signs of lipodystrophy.

Noted changes in body composition (self-reported by a patient, diagnosed or confirmed by a physician with the aid of tests such as magnetic resonance imaging (MRI), comparative photographs and measurements, or made apparent review of medical records). These changes include at least one of the following:

  • Peripheral fat wasting (a significant decrease in fat from the face, arms, butt and legs), resulting in a thin or drawn look;

  • Central obesity (truncal obesity or protease paunch). This build-up of fat behind the stomach muscle is generally not "mushy" but rather fairly firm/hard; this fat is different than the normal increases in fat associated with aging;

  • Development of a dorsal fat pad (buffalo hump), a build-up of fat at the base of the back of the neck between the shoulders;

  • Breast enlargement. This is a build up of fat in the breasts and has been observed in both women and men.

The changes in body composition are commonly accompanied by abnormal lab results including at least two of the following:

  • Elevated triglyceride levels, >2mmol/l

  • Elevated cholesterol, >5.5mmol/l

  • Elevated C-peptide, >2.5mmol/l

  • Glucose intolerance

  • Impaired fasting glucose, 6.0-7.0mmol/l

  • Impaired glucose intolerance, 7.8-11.1mmol/l

  • Diabetes mellitus, >7.0mmol/l fasting or >11.1mmol/l non-fasting

Because other HIV-related conditions might effect body composition, if an individual has had a serious HIV-related condition (e.g. an AIDS-defining condition) within the past three months, it can't be certain that changes in body composition are a result of lipodystrophy, but might be due to other infections.

We know that certain therapies (anabolic steroids) also affect body composition. If changes in body composition occur while someone is taking these therapies, it is unclear if the changes are HIV-related, related to anti-HIV therapies and/or related to the anabolic steroids.

Project Inform Website

If you have an internet connection to the world wide web and are looking for HIV/AIDS treatment information, log onto Project Inform's HIV/AIDS Treatment Website at:

Back to the Project Inform Perspective April 1999 contents page.