8th Annual Retrovirus Conference
Late Breakers
Chicago, Feb 4-8 2001

 

MACS Database on Lipodystrophy: prevalence of body changes & lab abnormalities (lipids, sugar)

L. Kingsley from the University of Pittsburgh reported on a cross-sectional study at Retrovirus (Feb 2001) of 868 men (331 HIV-, 537 HIV+) in the MACS cohort. Participants were examined at visit 31, which took place during April-Sept '99. The MACS cohort is a study in which gay HIV positive & negative men have been followed for many years. A cross-sectional study is one win which patients are looked at once and evaluated in this case for body changes & metabolic or lab abnormalities (cholesterol, triglycerides, insulin, etc). One of the goals for this study was to be able to put together a simple definition of what is lipodystrophy or body changes in HIV. They also wanted to estimate a prevalence of the lipodystrophy syndrome (how often it occurs) and identify patterns of metabolic or lab changes in HIV- vs HIV+ gay men.

Brief Study Summary:

In brief, Kingsley found that moderate and moderate severe fat wasting or loss in the periphery (lipoatrophy) and fat accumulation in the belly were associated with HAART use in these men. Also associated with HAART use was low HDL (good cholesterol) and elevated triglycerides (>400 mg/dl). Elevated sugar & insulin was more frequent among men on HAART and mono/combo, but not found in HIV-infected men who received no treatment. Perhaps the only surprising or interesting finding was that the MACS data saw the incidence of moderate and moderate/severe body changes rise in the first two years of HAART use and level off during the next 2 years, but the incidence of "any" body changes rose during the first two years and continued to rise during the second two years.

Although I don't think Kingsley defined what mono/combo therapy means I assume it means single or double NRTIs. 2% of people who were treatment-naïve reported moderate/severe body changes.

This study could not clearly characterize elevated total cholesterol because of the limitations of a cross-sectional study. The incidence of elevated cholesterol (>240 mg/dl) for individuals taking HAART (22%) compared to those receiving no treatment (16%) and those receiving mono/combo (11%), While the incidence of low cholesterol among HIV- was 27%, which could be due to an aging study population (47 years on average), the low incidence in those receiving no treatment or mono/combo could be an effect of having HIV (which tends to result in lower cholesterol). Clearly the incidence of elevated cholesterol increased for people taking HAART.

Perhaps, the most interesting piece of information was his finding that the incidence of moderate to severe lipodystrophy (body changes) rose sharply during the first 2 years after starting HAART. After two years, and for up to 4 years the incidence in moderate/severe lipodystrophy did not appear to rise. He felt this observation was possibly significant and said it needs further study. Observers at the Conference felt this suggested that body changes generally occur within 2 years after starting HAART, and if after 2 years a person has not experienced HAART they are not likely to experience the body changes. Still, in the graph of this information the incidence of "any" as opposed to moderate/severe or severe body changes did continue to climb after 2 years. So, I'm not convinced about the notion that body changes are only seen within the first two years, and a person will not experience them after 2 years if they hadn't seen them earlier. This needs further study before we assume lipodystrophy won't emerge for a person after 2 years on HAART.

He reported that certain body changes such as fat gain in the belly, dorsocervical fat pad (behind the neck--buffalo hump), and breasts were observed in a high percentage of HIV- MACS participants. Peripheral fat wasting (legs, arms, face), not central or belly fat accumulation was found to be more specific to HIV infected men. Regarding a definition, he concluded that a symptom combination of at least moderate central fat gain and at least moderate peripheral wasting works well to increase our specifity of lipodystrophy (defined by body changes) with HIV infection. Not very compelling information.

Conducting a cross-sectional study where you exam patients & ask question at just one time as opposed to following them over the course of years has a number of well known limitations. This may be reflected by the findings he reported that there were smaller than expected differences seen between HAART treated & HIV- gay men in total cholesterol, LDL (bad cholesterol) and triglycerides. He said this highlights the need to better account for age & the course of HIV infection. For example, early in HIV infection before HAART cholesterol was low for many individuals but obviously increased after starting HAART. He also concluded that his study finding of increased fasting glucose and elevated insulin appears to be due to ART.

METHODS OF ASSESSING THE SYNDROME

In evaluating patients at one visit, which again is what a cross-sectional study is, the investigators used standardized anthropometric measurements to look at height, weight, BMI (body mass index), arm, thigh, waist, hip, and WHR( waste-to-hip ratio). The used self-reported (by the patient or study participant) and examiner-reported body changes and graded them as mild, moderate & severe. They took fasting blood for metabolic or lab abnormalities: total cholesterol, HDL (good cholesterol), LDL (bad cholesterol), triglycerides, HBA1c, glucose, insulin, apollpoprotein A and B, and lipoprotein A. I recall Kingsley saying he is only reporting the data from examiner-reports because these reports are better than self-reporting. I'm not convinced that's true. This area is controversial with some studies & researchers suggesting self-reporting is reliable and comparable to examiner-reporting and others reporting that self- is not comparable & not as reliable as examiner or doctor reporting.

Of the 537 HIV+ (62%) patients, 98.7% have a known ART history: 14.5% none, 11.9% mono/combo therapy, and 73.6% HAART. The average age of men in the study was 47 years and this is important to bear in mind when looking at the lab abnormalities & body changes, because as men age body changes & lipid & glucose levels go up.

Any Examiner Reported Fat Change by HIV Status
Over half of the belly fat was reported to be moderate or severe. Because of the incidence of fat accumulation being higher in the HIV- participants Kingsley concluded that fat loss or depletion (lipoatrophy) in the face, arms, legs, etc were more representative of fat redistribution syndrome in HIV. But, this observation was probably made because of the average age of all study participants being 47. In younger men fat accumulation in the belly is not as likely to occur as when men get older. So, the belly fat accumulation and the breast changes may be associated more with age. Younger men with lipodystrophy experience belly fat accumulation. The percentages were presented in bar form by Kingsley, so my numbers are estimated based on visual observation by me.

 
HIV-
HIV+
p-value
Facial
2-3%
35%
<0.001
Arms
2-3%
25%
<0.001
Legs
2-3%
31%
<0.001
Buttocks
2-3%
32%
<0.001
Belly
26%
38%
<0.001
Fat Pad
  5%
10%
  0.008
Breasts
  9%
11%
  0.187

Body Changes Prevalence: 16% in HIV+
Kingsley reported 16% incidence or prevalence of moderate/severe lipodystrophy verseus <1% in the HIV- persons (p<0.001). This prevalence was based only on the composite definition of fat loss and central fat accumulation. When he cateogorized by any, mild, moderate and severe, the percentages were: for HIV+-- 40% any, 10%, mild, 10% moderate, 2% severe. And for HIV- the percentages were all 1% or less.

Body Changes Prevalence By Treatment Regimen: 20% on HAART vs 8% on mono/combo nukes
      --HAART results in more severe body changes compared to mono/combo nukes
The data shows that the incidence of "any" and mild body changes were relatively equal among MACS participants whether they had no treatment, mono-combo nukes, or HAART. Although those on HAART had slightly more of "any" body changes compared to those on mono-combo, and much more changes compared to people on no therapy (42% HAART vs 10% no therapy vs 40% mono/combo). But when looking at moderate and moderate/severe body changes, people on HAART had higher incidence or prevalence of body changes.

 
 None 
Mono/Combo
HAART
Moderate
1%
7%
14% (p=0.002)
Moderate/Severe
2%
8%
20% (p<0.001)

SERUM LIPIDS in MACS (fasting)

The incidence of elevated total cholesterol is high both in HIV- and in people on HAART (28% vs 23%). This may in part reflect the aging process among people who are HIV-. The incidence of elevated total cholesterol is much more for those receiving HAART compared to nukes alone or no treatment (23% vs 16% & 12%). This may reflect the fact that many HIV-infected people had low cholesterol in the earlier years of having HIV but cholesterol started going up as people started taking HAART. The incidence of HDL (good cholesterol) is obviously too low much more often in HIV+ people compared to HIV- (29% vs 6%), but is relatively equal no matter which treatment or no treatment. The comparable low HDL in all HIV-infected regardless of treatment may be because HIV-infection itself leads to reduced HDL. The incidence of elevated triglycerides goes from 3% for HIV- to 12% for people on HAART, but 0% for persons on no treatment & nukes alone. The incidence of elevated glucose increases from 3% in HIV- to 11% for persons on HAART but is 0% for HIV+ not receiving treatment or receiving just nukes. The incidence of elevated insulin goes from 0% for HIV+ not recieving treatment to 9% for people on nukes to 14% for people on HAART.

 HIV- 
HIV+
 
Antiretroviral Therapy
 None 
Mono/Combo
 HAART 
  Cholesterol
Median
211
191
193
207
% >240 mg/dL
28%
16%
12%
23%
  HDL
Median
48
41
43
41
% <35 mg/dL
 6%
23%
27%
29%
  LDL
Median
100
115
146
179
% >400 mg/dL
 3%
 0%
  0%
12%
  Glucose
Median
98
95
106
100
% >125 mg/dL
 3%
 0%
  0%
11%
  Hg A1c
Median
% >6.1%
 3%
 5%
  0%
   3.5%
  Insulin
Median
% >25 m/ml
   7.6%
 0%
  9%
14%
  Lipoprotein A
Median
% >30 mg/dl
25%
14%
27%
23%
Apollpoprotein A
Median
% >176 mg/dl
32%
19%
18%
23%
Apollpoprotein B
Median
% >109 mg/dl
  1.5%
 9%
  9%
  9%

 <  www.natap.org

All other articles >