FRAM Study and Lipodystrophy: lipodystrophy study presented at Barcelona and an interview with the study author Carl Grunfeld
Reported by Jules Levin
Carl Grunfeld, MD, is a researcher and treating physician specializing in diabetes, endocrinology, lipodystrophy, and metabolic abnormalities. He treats HIV- and HIV+ patients in San Francisco at the VA Hospital.
The goal of this study is to find out what contributes to the body changes HIV+ individuals are experiencing. Grunfeld's reported findings from the FRAM Study are preliminary and based only an analysis of less than all the men in the study, and does not include an analysis of the women. But his report was very controversial at the conference and still is very controversial, because his findings are at odds with some key beliefs we generally have come to accept about lipodystrophy. This article is a report of the FRAM study findings presented by Grunfeld at Barcelona and our telephone discussion about this study, and the syndrome of body changes patients are experiencing.
It's important to bear in mind that this study is cross-sectional. That means that they did not follow patients over time to see how their body changed. The researchers took a one-time snapshot look at the patients. They asked the patients once to describe their perception of body changes they may have experienced (self-report). The researchers gave them one physical exam. And they performed one-time an objective test (MRI, DEXA) to measure the fat in various parts of the body including the belly, face, legs and arms. This means they did not follow patients over a long period of time and repeatedly ask them about body changes, have repeated physical exams, nor did they repeatedly perform objective tests to see if they found body changes over time. This would be called a longitudinal study.
Grunfeld found that on average HIV+ men reported and had greater fat loss in the belly than the HIV- men in the study. He found that in general HIV+ men had greater fat loss in all areas of the body (periphery and central depots including belly) than HIV- men in the study, and thus he suggests fat loss is the hallmark of lipodystrophy. In fact, he suggests that our currently used definition of lipodystrophy underestimates fat loss; he says fat loss is greater than we think. Grunfeld found the HIV+ men in the study were skinnier than the HIV- men. He found that in the upper trunk area as well, that HIV+ men had fat loss. Apparently, this also created controversy because people feel they see fat gain in the upper trunk. The upper trunk is the chest and shoulder areas. Even though he found that on average HIV+ men experienced fat loss in the belly he also found that some HIV+ men in the study do experience increased belly fat.
In our interview Grunfeld said:
"NO WHERE DID I SAY THAT THEY CAN'T BE SEEN TOGETHER IN THE SAME PERSON (fat gain in the belly plus fat loss in the face & other peripheral areas); THEY CAN. BUT THE DATA SAY THAT THE ASSOCIATION IS LESS THAN EXPECTED IF IT WERE RANDOM. Thus it is important to analyze the causes of subcutaneous fat loss separately from any form of central fat gain. They can't be pooled."
He also reported that there is no linkage between fat accumulation in the belly and fat loss in the periphery; it is not a "combined syndrome". As a result, he recommends that fat accumulation in the belly should not be part of the definition of a unified lipodystrophy syndrome. Fat accumulation in the belly has been part of what most people consider lipodystrophy. He also found that the HIV- men in this study had buffalo hump just as often as the HIV+ men (12% HIV- men vs 8% HIV+ men), so the presence of buffalo hump should not be included in definition of lipodystrophy. Based on the study data Grunfeld also recommends, that the Waist-to-Hip Ratio should not be used as a surrogate for central fat gain or VAT.
Again, this has created quite a stir. I've spoken with a number of community advocates and several researchers and doctors who mostly do not agree with Grunfeld's notion emanating from this study -- that fat gain in the belly is not a hallmark of lipodystrophy.
Grunfeld emphasized in his opening comments that this report at Barcelona by him is of preliminary data and only on men. It should not be extrapolated to women where the data analysis on women is not ready yet. There are about 350 women in the study. This preliminary analysis includes 357 of the 800 men in the study.
Grunfeld said in our interview that we not refer to this syndrome as lipodystrophy, but we should call it should call this syndrome lipoatrophy and/or fat gain in belly gain. We should not use the term lipodystrophy because that term suggests the two body changes, fat gain and fat loss, are a combined syndrome. He believes you can have one or the other or both, but again you can experience fat loss and no belly fat gain.
The goal of this study is to find out what contributes to fat gain in the belly, and fat loss in various fat depots. The study was undertaken to address several additional questions. Andrew Carr and other researchers felt the syndrome was combined: fat gain + fat loss. There was much disagreement about how many people had body changes. There was a huge disagreement about what syndrome looks like: is it fat loss, fat gain, both, where do the changes occur. etc. He also wanted to explore the relative importance of a physical exam or how people look compared to objective testing. Are you supposed to look at clinical syndrome, how people look? In our interview he talked about how some individuals may not appear to have lipoatrophy but it can be found when using objective tests.
Does lipoatrophy contribute to high lipids, glucose & insulin independent of drugs? Do high lipids cause fat loss or fat gain? Grunfeld said fat loss or fat gain can contribute to metabolic changes. But metabolic changes do not lead to fat changes. There is no precedent for that. No one has showed that the primary event of insulin resistance & diabetes or lipid abnormalities leads to fat loss or gain. There is no precedent for that either; with exception of someone who is very sick from diabetes.
Grunfeld said they have great patient records and charts: cd4 history, drug history, VL, weight history, age, duration of HIV, family history, DNA stored. In future study analysis Grunfeld plans to explore many questions. He will be looking at various potential causes for body changes. He will compare the effects on whites to African-Americans; the effects on women. An important analysis will be the effect of individual drugs. Does d4T, AZT, abacavir, nelfinavir or indinavir lead to a certain body change? If a specific drug predisposes a patient to a body change, perhaps one additional factor such as a large cd4 increase (immune reconsitution) from HAART may lead to significant lipoatrophy. From the patient charts, we will know who had a mother, father, brother or sister with diabetes. Suppose we can show that people on a specific PI with diabetes who had a family history of diabetes are more predisposed to lipoatrophy. Then perhaps we could pick certain HIV drugs for a regimen.
Grunfeld measured body composition on almost all study patients. So they can see what percentage of fat loss patient's had and perhaps correlate that with the therapy they were taking. For example, perhaps they can figure out if AZT or ddI causes 10% fat loss and ritonavir causes 15% fat loss. But Grunfeld believes that several factors are likely to be acting at once and causing body changes. He said not all individuals will have perceptable fat loss. Some people have mild lipoatrophy, which may not be detectable by visual observation or physical exam. "We want to know who are the bad cases. We should dissect this apart". Some patients who have lipoatrophy may not meet the currently accepted definition because the changes are not readily perceptible but can be seen only if objective tests are performed.
We also discussed the role HIV might play in contributing to body changes. Low cd4 count, changes in CD4 count (immune reconstittion from HAART), changes in viral load may all contribute to body changes. A rise in cd4 may play a small contributing role towards lipoatrophy, or a specific drug may also make a small contribution towards lipoatrophy, but perhaps you need two factors: cd4 increase+that drug to push you over the top in developing significant lipoatrophy.
Perhaps HIV or immune reconstitution speeds up metabolism, and changes the immune system in a way that leads to body changes. What role does HGH therapy play? It can reduce visceral fat, but it might also increase fat loss. So, patients should be careful about that. If you have a predisposition to fat loss HGH might push you over the edge in developing fat loss.
Diet, exercise. Grunfeld was very emphatic in recommending diet and exercise. He was most high on exercise. He recommends moderate exercise can play an important role in reducing visceral belly fat, in decreasing cholesterol, triglycerides, sugar, and in controlling diabetes. He likes moderate running and weight lifting together to achieve these goals. Diet alone won't raise HDL (good cholesterol), but exercise can. He also warned that exercise and weight loss may also increase fat loss in the face and other peripheral areas, so be careful.
Are body changes reversible? Grunfeld believes increased visceral fat in the belly is reversible;, he says he's pretty sure. But improving fat loss is possible but more difficult. We may be able to improve fat loss moderately but not completely or very much.
In order to continue this study Grunfeld is working with the NIH to submit a follow-up grant. I think he would like to continue following patients in the study with objective testing with full body MRIs to evaluate body changes; and he would also like to follow changes in sugar cholesterol, triglycerides and other metabolic measures.
Who are the subjects in this study?
The HIV+ patients were selected by random selection from the database of each clinic where they attended in 1999. Recruitment of men was from 6/00 to 1/02. Less than 20% of those contacted declined study. This is a one time cross-sectional study. In other words, patients were looked at once, a snapshot in time. The patients were selected from the entire clinic database and not from frequent clinic attendees of typical study volunteers.
The HIV- or control patients were age-matched with the HIV+ men and were selected from the CARDIA study, a representative sample of young adults, stratified across Caucasian and African-American, gender, age (at the time 18-30), and education. They are now 33-45 years old since they were recruited 15 years ago.
At the Birmingham site, selection was from a list of residential telephone numbers for the city. At the Oakland site, selection was from a computer database of members of the largest HMO in California. For FRAM, they mainly selected CARDIA participants enrolled five years ago in a substudy of visceral fat. Less than 1% had HIV. The prevalence of obesity in the control patients is similar to that in the larger NHAHNES database that is representative of the US population.
Grunfeld compared HIV+ men & women selected randomly from patient database in each clinic site to HIV negatives from a proportion of patients in the CARDIA study. Grunfeld said they are a representative sample of US population studying cardiovascular disease. He said prevalence of obesity should be representative of average men and women in US in terms of fat or obesity, from NHAHNES database, which is supposed to represent prevalence of obesity in US population.
HIV+ & HIV- men are 33-45 years; HIV+ men CD4s average 380 (range: 6-1500); VL 35,000 average (range: <400-750,000). HIV+ men: 77% MSM, 10% IVDU. 14% no ART, 83% on NRTI, 55% PI, 37% NNRTI. 50-55% white, 33% AA in HIV+ vs 45% AA in HIV-.
Grunfeld reported that in this study fat loss was found to be main problem in body changes as measured by objective test & self-report &researcher exam. Belly fat gain was not on average a problem when comparing HIV positives to negatives by self-report, researcher observation & objective test, although some patient & researcher observation reported fat gain in belly.
Self-Report & Researcher Observation
The areas studied in the self-report and visual inspection by physical exam by researcher included in the periphery- cheeks (next to the nose), face shape, buttocks, legs, arms; in the central or belly area- waist size, abdominal fat, neck, upper back, chest.
It is generally presumed in HIV lipodystrophy that there is a decrease in peripheral fat accompanied by an increase in central fat, but Grunfeld found otherwise in this study. This study also differed from many pervious ones in that questions and exam were graded in both directions (increased and decreased) at all sites.
HIV+ men on the average reported more peripheral fat loss in all sites whereas on the average HIV- men reported an increase in fat; and this was statistically significant in each of the peripheral sites.
HIV+ on average reported a prevalence of peripheral fat loss of 30% in each site of fat loss. Some HIV- negative men, about 5%, reported fat loss at peripheral sites.
Fewer HIV+ men reported increases in central fat in the belly and in other areas (neck, chest) than HIV-. About 40% of HIV+ men reported increases in waist size. Significantly more HIV- men reported increases in waist size, about 60%. There was no difference in reported abdominal fat between HIV+ and HIV- men.
There is some controversy and discussion raised by doctors and researchers about the control group. Perhaps there was something unusual about the control group where you might find more belly fat reported by the men.
Of more interest, said Grunfeld, more HIV+ men reported decreases in waist and most other central fat depots (neck, chest, upper back) than did HIV- men. About 20% of HIV+ men reported decreased belly fat and about 15% of HIV- men reported decrease in waist (p=0.02). About 15% of HIV+ men reported decreased abdominal fat while about 10% of HIV-men reported this (p=0.13).
Summing up self-report and exam, Grunfeld said peripheral fat loss is more common in HIV+ than in HIV- and fat increase in belly & other fat depots is on average not more but perhaps less in HIV+ men than HIV- men. Therefore, he concluded that the data from this study does not support including fat gain in the belly in a definition of lipodystrophy syndrome. Further, he suggests that his data suggests that perhaps less central fat me be characteristic of lipodystrophy.
On average, those with peripheral atrophy by self-report and exam had less central fat. Grunfeld said that these data do not support the linkage between peripheral fat and central fat gain.
40% of HIV+ men had fat loss (mild, moderate, or severe) at 1 or more sites. 27% at 2 or more sites.15% at 3 or more sites. 5% HIV- reported fat loss at 1-2 sites.
Patients with central fat gain had less fat loss in periphery than patients with no central fat gain. Grunfeld added that some patients with central fat gain also had peripheral atrophy. The data suggest that peripheral fat loss is not linked to central fat gain. Patients reporting central fat loss had an incidence of peripheral fat loss higher than patients with no central fat loss. Therefore, patients with central fat gain are less likely to have peripheral fat loss, implying no linkage. The data from this study find there is not a "combined syndrome" of fat loss in periphery and central fat gain. Rather, central fat loss is linked to peripheral fat lipoatrophy.
Body Composition - Objective Testing
HIV- men had much more limb fat, by DEXA, than the HIV+ men. Eventually Grunfeld will show us full body MRI results on the study patients. This will present better understanding of DEXA, because he believes MRI is better at detecting changes than DEXA. MRI of the face is being done but he is concerned its not accurate enough because you need to do more MRI slices than is being done to get a better evaluation. But he believes MRI could be used for the face. And he believes there are other ways to measure the face.
Those with peripheral lipoatrophy by self-report & exam also had less limb fat, as expected, as measured by DEXA than HIV+ men without lipoatrophy. But it's important to note that those without clinical lipoatrophy had on average significantly less limb fat than the HIV- men. But there were HIV+ men with more limb fat.
They used MRI to look at lower subcutaneous trunk fat (belly fat). HIV+ men had on average less fat than HIV- men. And again some men HIV+ men had more fat than other HIV+ men. HIV+ men with clinical lipoatrophy had less fat in the subcutaneous section of the abdomen than those without clinical lipoatrophy. Again on average HIV+ men without lipoatrophy had less fat in the subcutaneous section of the abdomen than HIV- men.
These data suggest that in general HIV+ men were skinnier than the HIV- men. The results were similar but not as strong when looking at the upper trunk. Still, there is less fat in HIV+ men without lipoatrophy than HIV- men.
HIV+ appear to lose more fat in legs > arms > lower trunk s.c. > upper trunk s.c.
VAT By Total Body MRI
HIV+ had on average significantly less fat than HIV- men in VAT. But he added the data finds there are a lot of patients with increased VAT among the HIV+, just as there are in the HIV- men. VAT was significantly the same whether patient had LA or not, suggesting that patients with LA do not necessarily develop a compensatory fat gain in belly.
Grunfeld said he is analyzing in detail what factors are associated with the level of VAT one has; but younger patients have less fat than older patients, and this is true among the HIV- men as well.
Grunfeld mentioned that study limitations included that this is one time cross sectional look so we don't know how patients got to their current body composition, and where the trends are going. Whether those with HIV-infection might have lower VAT if they had not developed decreased SAT (subcutaneous lipoatrophy), we can't determine. However, the data indicate that the clinical syndrome (patient self-report and exam) of HIV-associated lipoatrophy is NOT associated with increased fat. Because of the body composition abnormalities found in HIV+ men in this study, Grunfeld said we cannot use Waist-to-Hip Ratio (WHR) as a surrogate for increased VAT or central obesity.
--Self-report, physical exam and objective measurements (including many more that time does not permit to show) clearly demonstrate that lipodystrophy is an HIV-specific syndrome while it is not for HIV- men in this study.
--However, self-report, exam, and objective measurements do not support a compensatory (linkage) central fat accumulation.
--HIV+ men without clinical lipoatrophy have less fat than HIV- men in this study.
--The clinical (self-report, exam) syndrome of lipoatrophy underestimates the degree of fat loss in HIV+ men.
--Therefore, we should ask not merely what causes the clinical syndrome of lipoatrophy, but we should also ask:
What causes decreased fat in HIV infection?
Future studies need direct measurement of fat.