| |
Prediction of Coronary Heart Disease Risk in HIV-Infected Patients with Fat
Redistribution
|
| |
| |
Clinical Infectious Diseases 2003;36:909-916
Colleen Hadigan,1,5 James B. Meigs,3,5 Peter W. F. Wilson,6,7 Ralph B. D'Agostino,6,7 Benjamin Davis,2,5 Nesli Basgoz,2,5 Paul E. Sax,4,5 and Steven Grinspoon1,5
1Program in Nutritional Metabolism, 2Infectious Disease Unit, and 3General Medicine Division, Massachusetts General Hospital, and 4Infectious Disease Unit, Brigham and Women's Hospital, 5Harvard Medical School, 6Boston University School of Medicine, and 7Framingham Heart Study, Boston, Massachusetts
Comment from Jules Levin: in brief, this study appears to find that fat redistribution, both lipoatrophy and fat accumulation in the belly, increase risk for heart disease compared to control subjects. Control subjects were selected from the Framingham Offspring Study, a population-based observational study of risk factors for cardiovascular disease. Data on control subjects were obtained from January 1991 through September 1995, during the fifth Framingham Offspring Study examination cycle. Trends in diet and obesity may have been different during 1991-1995 compared to now. You could make a case that perhaps HIV+ subjects today may not have increased risk compared to HIV-negative individuals selected today who in general might have increased risk for CHD due to increased obesity in our society. However, I donŐt think I would accept this latter premise.
"...An unanticipated finding of this study..our results suggest that subcutaneous fat loss in patients with HIV disease may substantially contribute to CHD risk.a nonsignificant increased proportion of smokers and greater age among subjects with lipoatrophy may have contributed to this finding.."
"authors conclude: HIV-infected patients with fat redistribution demonstrate an increase in 10-year CHD risk estimates, but this risk does not differ from that of a population with a similar central pattern of fat distribution. Nearly one-third of the HIV-infected patients evaluated in this study have a 10% risk of having CHD diagnosed in the next 10 years, as determined on the basis of the Framingham risk equation estimates. Further longitudinal follow-up studies are needed to determine the actual incidence of CHD and whether CHD risk reduction decreases morbidity in this population. Nonetheless, our data strongly suggest that modification of risk factors for CHD is warranted for HIV-infected patients with fat redistribution".
Recent reports have identified a metabolic syndrome among HIV-infected patients receiving combination antiretroviral therapy; the syndrome is characterized by fat redistribution (increased visceral fat and/or reduced subcutaneous fat), insulin resistance, dyslipidemia (increased triglyceride level and decreased high-density lipoprotein [HDL] cholesterol level), and hypertension. The etiology of these abnormalities is not known but may relate to specific effects of antiretroviral drugs on adipocyte metabolism with consequent metabolic abnormalities, and/or from direct effects of antiretroviral drugs on lipid or glucose metabolism. HIV infection may play a role independent of drug-related effects, but prior data do not demonstrate severe metabolic abnormalities in patients without fat redistribution. Preliminary reports suggest increased rates of cardiovascular disease among HIV-infected patients, but data from prospective, longitudinal cohort studies are not available.
The Framingham risk equation, which estimates 10-year coronary heart disease (CHD) risk, was developed on the basis of coronary events (including angina pectoris, myocardial infarction, and deaths due to CHD) that occurred during longitudinal follow-up of the Framingham cohort. The equation includes age, total cholesterol level, HDL cholesterol level, blood pressure, presence of diabetes, and smoking status, and it is a well-established tool to evaluate long-term CHD risk. In this study, the calculated 10-year CHD risk estimate in HIV-infected subjects who have fat redistribution is compared with the CHD risk estimate in sex-matched, age-matched, body mass index (BMI)matched, and, in a substudy, waist-to-hip ratio (WHR)matched subjects from the Framingham Offspring Study.
We compared the 10-year coronary heart disease (CHD) risk estimates for 91 HIV-infected men and women with fat redistribution with the risk estimates for distribution
IV-infected patients without fat redistribution did not have a greater CHD risk estimate than did control subjects. In addition, the CHD risk estimate was greatest in HIV-infected patients who had primary lipoatrophy, compared with those who had either lipohypertrophy or mixed fat redistribution. Therefore, although CHD risk is increased in HIV-infected patients with fat redistribution, the pattern of fat distribution and sex are potential important components in determining the risk in this population.
See detailed results below.
HIV-infected patients with lipodystrophy. Ninety-one consecutive HIV-infected subjects (65 men and 26 women) who reported recent changes in body fat distribution were prospectively evaluated from December 1998 through November 1999 at the Clinical Research Center of the Massachusetts Institute of Technology (Cambridge, MA). Subjects were recruited using community-based advertisements seeking HIV-infected patients with fat redistribution or were referred by their physicians for evaluation of fat redistribution. Subjects were screened by telephone and asked whether they had experienced any of the following changes: (1) loss of fat in the face, (2) increased fat under the chin or back of the neck, (3) increased abdominal girth, (4) increased chest or breast fat, or (5) loss of fat in the arms or legs. Subjects who identified a change in fat distribution in 1 body area were invited to participate, and fat redistribution was confirmed by physical examination of all subjects.
Subjects were excluded from the study if they had changed antiretroviral medications 6 weeks before presentation; had a history of diabetes mellitus or previous treatment with an antidiabetic agent; reported use of testosterone, estrogen, growth hormone, or other steroids in the 6 months before presentation; were active alcohol or substance abusers; or were not aged 1860 years.
Demographic characteristics and 10-year CHD risk estimates for HIV-infected subjects with fat redistribution and for control subjects are shown in table 1: HIV+ men or women had bigger waists and were smaller hips, thighs, or arms as measured by body circumfrance; waist to hip ratio was bigger in HIV+; total cholesterol was higher; HDL (good) cholesterol was lower; . HIV-infected subjects with fat redistribution had a duration of infection of 7.2 ± 0.5 years, received 4.7 ± 0.3 years of antiretroviral therapy, and had a mean HIV RNA level of 6572 ± 2511 copies/mL. The 10-year CHD risk estimate was significantly increased among HIV-infected patients with fat redistribution, compared with that for control subjects from the Framingham Offspring Study: 7.4% ± 0.6% versus 5.3% ± 0.3%, for all subjects (P = .002); 9.0% ± 0.7% versus 6.5% ± 0.3%, for men (P = .001); and 3.4% ± 0.8% versus 2.2% ± 0.3%, for women (P = .19). The percentage of HIV-infected subjects with a 10-year CHD risk estimate of 10% was substantially elevated for HIV-infected subjects, compared with the percentage of Framingham control subjects: 29.1% vs. 12.8%, for all subjects (P = .001); 38.7% vs. 17.4%, for men (P = .001); and 4.2% vs. 1.3%, for women (P = .37). HIV-infected patients without evidence of fat redistribution (age, 39.4 ± 0.9 years; BMI, 24.0 ± 0.4 kg/m2; WHR, 0.90 ± 0.01; duration of HIV infection, 7.7 ± 0.7 years; duration of antiretroviral therapy, 3.0 ± 0.6 years; and HIV RNA level, 23,175 ± 9927 copies/mL) did not demonstrate elevated 10-year CHD risk estimate, compared with matched Framingham control subjects (10-year CHD risk estimate, 4.1% ± 0.7% vs. 3.3% ± 0.3%, respectively; P = .27).
Framingham control subjects were then matched with HIV-infected subjects according to sex, age, BMI, and WHR, which revealed that the 10-year CHD risk estimate was not significantly elevated among HIV-infected patients with fat redistribution, compared with Framingham control subjects (7.6% ± 0.6% vs. 7.6% ± 0.4%; P = .9). Thirty percent of HIV-infected patients with fat redistribution had a 10-year CHD risk estimate of 10%, compared with 28% of the WHR-matched control subjects (P = .7). Matching by WHR did not change the results of the comparison between HIV-infected subjects without fat redistribution and Framingham control subjects; there was no difference in 10-year CHD risk estimates between the 2 groups (4.1% ± 0.7% vs. 3.4% ± 0.4% for 30 HIV-infected subjects vs. 58 control subjects; P = .37).
Lipoatrophy vs Belly Fat
Interestingly, in subanalyses of HIV-infected subjects with fat redistribution, 10-year CHD risk estimates differed on the basis of pattern of fat redistribution (e.g., primary lipoatrophy vs. abdominal lipohypertrophy vs. mixed lipodystrophy). Anthropometric measurements confirmed that thigh circumference and arm circumference were lowest among patients with lipoatrophy, whereas waist circumference was highest among patients with abdominal lipohypertrophy.
Subjects with primary lipoatrophy demonstrated the highest 10-year CHD risk estimate (9.2% ± 1.8%, compared with the abdominal lipohypertrophy group, 4.3% ± 0.7%; compared with the mixed lipodystrophy group, 7.6% ± 0.8%; P = .043, for overall comparison between the groups; P < .05, for the primary lipoatrophy group vs. the abdominal lipohypertrophy group; and P < .05, for the mixed lipodystrophy group vs. the abdominal lipohypertrophy group). Subjects with lipoatrophy (n=15) were 4 yrs older (43 vs 39) than subjects with lipohypertrophy (fat belly) (n=15), had less BMI (23 vs 29), and had smaller waist, hip, thigh, arm, and waist hip-ratio as measured by body circumference. The difference in cholesterol was small, 227 mg/dl in the lipoatrophy group vs 215 in the fat belly group, and HDL was 31 mg/dl in lipoatrophy group vs 38 mg/dl in fat belly group. Fasting glucose was the same, 92 mg/sl, in both groups. There were more current smokers in the lipoatrophy group (40% vs 26%). Blood pressure was about the same in both groups. As stated above the 10-year CHD risk estimate was 9.2% (lipoatrophy) vs 4.3% (fat belly) vs 7.6% (subjects with mixed lipodystrophy). The percent with >10% risk was 43% in lipoatrophy group vs 0% in fat belly group vs 33% in ixed lipodystrophy group.
Discussion By Authors
The mechanisms of fat redistribution in HIV-infected patients are not known and may relate to stimulatory effects of antiretroviral drugs on adipogenesis in certain fat depots, with effects that increase lipolysis and/or apoptosis in other depots. Increased visceral adiposity and reduced amounts of subcutaneous fat have been shown to independently predict hyperinsulinemia in this population and have been associated with other metabolic abnormalities, including hypertension and dyslipidemia, as well as increased cardiovascular disease risk in HIV-uninfected patients. In this study, when control subjects were matched to HIV-infected subjects according to WHR, the 10-year CHD risk estimate in the control subjects increased and was no longer different from that of the HIV-infected patients, suggesting important potential influences of central adiposity. In addition, recent data suggest that certain antiretroviral medications can have direct effects on glucose and lipid metabolism, and these direct and indirect effects may simultaneously contribute to increased CHD risk.
We used the Framingham risk equation to determine 10-year CHD risk estimates. This equation was developed on the basis of coronary events that occurred during longitudinal follow-up of the Framingham cohort. The equation uses age, total cholesterol level, HDL cholesterol level, blood pressure, presence of diabetes, and smoking status data, and it is a well-established tool to evaluate CHD risk. Of note, the Framingham risk equation does not include WHR or a measure of central adiposity in the calculation of risk. Therefore, WHR matching would not, a priori, be expected to result in a matched CHD risk estimate. However, in this study, selection of control subjects with a similar degree of central adiposity brought the 10-year CHD risk estimate for control subjects to a level similar to that for HIV-infected subjects with fat redistribution.
|
|
| |
| |
|
|
|
