Antiretroviral therapy (ART) has significantly increased the survival and quality of life of people living with HIV/AIDS. Since the introduction and widespread use of combination ART, referred to as highly active antiretroviral therapy (HAART), HIV-related mortality has been reduced from 50 to 80%.¹ However, the long-term use of ARTs has been associated with metabolic abnormalities, including increased serum lipids, glucose, and insulin resistance. The combination of these disorders leads to a highly atherogenic profile, increasing the risk of cardiovascular disease (CVD).²

The interaction between host factors, HIV, and HAART is strongly associated with the accumulation or loss of body fat in specific body sites,³ which has been identified as lipodystrophy.⁴ Lipodystrophy is characterized by fat redistribution, with subcutaneous fat loss (lipoatrophy), mainly in the face, limbs, and buttocks, or fat accumulation (lipohypertrophy) in the abdomen, breast or posterior neck, or a combination of both.⁴ The HIV-associated body fat redistribution in individuals receiving ART (HIV/HAART-associated lipodystrophy) by itself is associated to dyslipidemia and hypertriglyceridemia, low-HDL-cholesterol, reduced insulin sensitivity, and diabetes.⁵ Subjects with HIV/HAART-associated lipodystrophy may have increased Framingham risk scores and higher coronary calcium scores and thus are at increased risk of coronary heart disease.^{6 , 7}

The most common diagnostic tool reported in the literature to detect HIV/HAART-associated lipodystrophy is the self-reported body fat distribution.^{8 - 10} It is an accurate, reproducible and easy-to-implement method, adding almost no costs to the usual care of HIV/AIDS patients.^{10 - 12} Despite the common use of patients' self-reported methods to identify lipodystrophy, to our knowledge, there are no available data on the association of cardiovascular risk factors and HIV/HAART-associated lipodystrophy assessed exclusively by this method. Therefore, we assessed subjects attending an HIV/AIDS outpatient care center of a capital city in Brazil (a middle-income country) to determine the prevalence of self-reported HIV/HAART-associated lipodystrophy and possible independent associations between traditional modifiable cardiovascular risk factors and self-reported lipodystrophy.

A total of 183 patients met eligibility criteria and were included in this study, with a mean age of 39.3 ± 10.9 years. The prevalence of HIV/HAART-associated lipodystrophy in the overall sample was 52.5% (96/183). Most of the sample was male (77.6%), non-white (50.82%) and single (53.0%); sociodemographic characteristics of the sample as well as the prevalence of lipodystrophy by subgroups are described in Table 1 . There was a significantly higher percentage of female subjects with lipodystrophy ( p =0.003).

Table 1

Variables	Total		Lipodystrophy prevalence				p-value*
	n	%	n	%	PR	95% CI
Sex							0.003
Male	142	77.60	67	47.18	1.00
Female	41	22.40	29	70.73	1.50	1.15 – 1.95
Age							0.952
19-29 years	35	19.13	18	51.43	1.01	0.67 – 1.52
30-39 years	61	33.33	32	52.46	1.03	0.73 – 1.46
40-49 years	59	32.24	30	50.85	1.00
≥50 years	28	15.30	16	57.14	0.51	0.75 – 1.69
Skin color							0.598
White	90	49.18	49	54.44	1.07	0.82 – 1.42
Non-white	93	50.82	47	50.54	1.00
Schooling (years)							0.808
≤4 years	30	16.39	18	60.00	1.20	0.79 – 1.81
5-8 years	42	22.95	22	52.38	1.05	0.70 – 1.58
9-11 years	65	35.52	33	50.77	1.01	0.70 – 1.48
≥12 years	46	25.14	23	50.00	1.00
Marital Status							0.434
Single	97	53.01	50	51.55	1.07	0.76 – 1.52
Married	50	27.32	24	48.00	1.00
Widowed/Divorced	36	19.67	22	61.11	1.27	0.86 – 1.88
Monthly income							0.864
1st quartile	59	32.24	31	52.54	1.07	0.74 – 1.55
2nd quartile	35	19.13	18	51.43	1.05	0.68 – 1.61
3rd quartile	51	27.87	25	49.02	1.00
4th quartile	38	20.77	22	57.89	1.18	0.79 – 1.74

PR: prevalence ratio; 95% CI: 95% confidence interval. *Wald’s Test.

The distribution of the different levels of lipodystrophy is presented in Figure 1 . No differences were observed between males and females ( Table 1 ). Severe lipodystrophy was observed in more than half of the individuals (55.2%).

Figure 1

When we assessed the prevalence of HIV/HAART-associated lipodystrophy by HIV-related clinical and laboratorial variables, a higher prevalence of lipodystrophy (66.7% - PR 1.88; 95% CI 1.12 – 3.14; p =0.025) was observed among those with a time of diagnosis of HIV infection between 1 and 3 years. No other HIV-related clinical or laboratorial data evaluated were associated with lipodystrophy. All patients were using NRTI ( Table 2 ).

Table 2

Variables	Total		Lipodystrophy prevalence				p-value*
	n	%	n	%	PR	95% CI
CD4 lymphocyte count							0.564
≤ 350 cells/mm3	64	36.36	36	56.25	1.09	0.82 – 1.44
> 350 cells/mm3	112	63.64	58	51.79	1.00
Time of HIV infection diagnosis							0.025
< 1 year	31	18.24	11	35.48	1.00
1 - 3 years	51	30.00	34	66.67	1.88	1.12 – 3.14
> 3 years	88	51.76	45	51.14	1.44	0.86 – 2.42
ART time							0.131
< 1 year	58	35.58	25	43.10	1.00
1 - 3 years	41	25.15	26	63.41	1.88	1.01 – 2.14
> 3 years	64	39.26	36	56.25	1.31	0.90 – 1.88
Drug classes
NRTI							-
Yes	173	100.00	93	53.76		-
No	0	0.00	0	0.00
NNRTI							0.436
Yes	133	76.88	69	51.88	1.00
No	40	23.12	24	60.00	1.16	0.85 – 1.56
PI							0.799
Yes	47	27.17	26	55.32	1.04	0.77 – 1.41
No	126	72.83	67	53.17	1.00

PR: prevalence ratio; 95% CI: 95% confidence interval; ART: antiretroviral therapy. *Wald’s Test. NRTI: nucleoside reverse transcriptase inhibitor; NNRTI: non-nucleoside reverse transcriptase inhibitor; PI: protease inhibitor

The prevalence of HIV/HAART-associated lipodystrophy was assessed by the presence of cardiovascular risk factors. In this analysis, lipodystrophy was associated with a family history of CVD (PR 1.70; 95% CI 1.22 – 2.39; p =0.002) and increased/greatly increased WC (PR 1.39; 95% CI 1.06 – 1.82; p =0.018), as shown in Table 3 .

Table 3

Variables	Total		Lipodystrophy prevalence				p-value*
	n	%	n	%	PR	CI 95%
Sedentary lifestyle							0.087
No	105	57.38	61	58.10	1.29	0.96 – 1.74
Yes	78	42.62	35	44.87	1.00
Smoking							0.646
No	89	48.63	44	49.44	1.00
Yes	49	26.78	26	53.06	1.07	0.77 – 1.50
Former smoker	45	24.59	26	57.78	1.17	0.84 – 1.62
Family history of CVD							0.002
No	173	96.11	87	50.29	1.00
Yes	7	3.89	6	85.71	1.70	1.22 – 2.39
Hypertension							0.230
No	166	90.71	85	51.20	1.00
Yes	17	17	11	64.71	1.26	0.86 – 1.85
Diabetes							0.905
No	177	96.72	93	52.54	1.05	0.46 – 2.37
Yes	6	3.28	3	50.00	1.00
Dyslipidemia							0.132
No	65	35.52	29	44.62	1.00
Yes	118	64.48	67	56.78	1.27	0.93 – 1.74
Waist circumference							0.018
Normal	123	71.93			1.00
Increased or very increased	48	28.07			1.39	1.06 -1.82
Overweight							0.306
No	122	71.35	62	50.82	1.00
Yes	49	28.65	29	59.18	1.16	0.87 – 1.56

PR: prevalence ratio; 95% CI: 95% confidence interval. CVD: cardiovascular disease. *Wald’s Test.

After including the clinical and laboratorial HIV-related variables, and the cardiovascular risk factors in the model, female sex, 1-3 years of HIV infection diagnosis and a positive family history of CVD remained associated with lipodystrophy in the multiple regression analysis ( Table 4 ).

Table 4

Variables	Lipodystrophy		p-value*
	aPR	95% CI
Sex			0.003
Male	1.00
Female	1.49	1.15 – 1.95
Time of HIV infection diagnosis			0.002
< 1 year	1.00
1 - 3 years	1.83	1.09 – 3.08
> 3 years	1.39	0.83 – 2.33
Family history of CVD			<0.001
No	1.00
Yes	1.62	1.11 – 2.36

_aPR: adjusted prevalence ratio; 95% CI: 95% confidence interval; CVD: cardiovascular disease. *Wald’s Test.

In a relatively large group of patients living with HIV/AIDS receiving HAART, the prevalence of HIV/HAART-associated lipodystrophy assessed exclusively by patient self-report was 52.5%. Most of these patients (55.2%) reported their lipodystrophy as severe, as opposed to mild or moderate. HIV/HAART-associated lipodystrophy was independently associated with female sex, 1-3 years of HIV infection diagnosis and a positive family history of CVD.

No association was found between self-reported HIV/HAART-associated lipodystrophy with traditional modifiable cardiovascular risk factors. In contrast, marked elevation of cardiovascular risk factors in HIV-infected patients with fat redistribution was previously reported by Hadigan et al .,²⁷ Some differences between the two studies need to be discussed. In the study by Hadigan et al .,²⁷ the population was older, had a longer time of HIV infection diagnosis and longer ART exposure. Additionally, and most importantly, the presence of lipodystrophy was determined by examiners rather than self-report. A difference in time course, considering ART prescriptions, between the two studies needs to be highlighted as well. In the last decade efforts have been made to reduce the use of drugs that are more strongly related to lipodystrophy such as stavudine and zidovudine and increase prescriptions of better tolerated drugs,¹² changing the lipodystrophy patterns over time.^{28 , 29} Individuals in the study by Hadigan et al.,²⁷ were assessed in 2000, therefore not subject to this new prescription pattern, while our study was conducted in 2009.

The decision to use a self-report method alone instead of using evaluations performed by healthcare providers or combined methods relied on the fact that body changes related to HIV/HAART-associated lipodystrophy are more likely to be noticed by the patients themselves and their families, then by third parties.³⁰ This situation is particularly relevant in the HIV/AIDS clinics of Brazil and other developing countries, where the healthcare provided is mainly public, with high staff rotation in different work shifts. Therefore, the patient is rarely followed by the same physician throughout time. Moreover, the self-report method plays an important role in lipodystrophy diagnosing, since it is validated, low-cost for implementation, and widely applicable, especially in low-income countries.¹⁰

This study shared similarities with other studies regarding HIV/HAART lipodystrophy. The prevalence of HIV/HAART-associated lipodystrophy can range from 9% to 83%,³¹ and depends on the criteria adopted for the diagnosis and on the characteristics of the studied population. Studies where lipodystrophy was assessed, either by a self-report method alone or by a combination of self-report method and observer evaluation, found a prevalence ranging from 45.9% to 64.3%,^{11 , 12 , 32 , 33} which is similar to the 52.5% found in the present study.

Other similarities in the clinical characteristics of our population compared with other studies on HIV/HAART-associated lipodystrophy were found, such as a higher prevalence of lipodystrophy in females and patients living with HIV for a relatively longer time.^{34 - 37} The higher prevalence of lipodystrophy among women might be explained by the higher body fat percentage in females when compared to males,³⁸ enabling a more noticeable fat redistribution related to the HAART. Additionally, women may have a better self-perception of body changes³⁹ than men.

Lipodystrophy prevalence was higher in patients living with HIV for 1 to 3 years. Since there is a dose-response relationship between the time of ART and lipodystrophy,⁴⁰ one would expect that the longer the HIV infection time, the higher the prevalence of lipodystrophy. Since the time of diagnosis does not necessarily coincide with treatment initiation, and ART time was not shown to be associated with lipodystrophy prevalence in our analysis, further investigations addressing exclusively the initiation time of ART are necessary.

Our study has some limitations. Despite the fact that lipodystrophy represents a continuum of fat redistribution over time, the study had a cross-sectional design, which allowed this temporal perspective. No control group was used for comparison. Also, no other method for the diagnosis of lipodystrophy was used for comparison with the self-reported method.

This study provides an important contribution to the knowledge of HIV/HAART-associated lipodystrophy identified by patient self-report. To our knowledge, this is the first time that this diagnostic approach was assessed focusing on associations between lipodystrophy and traditional cardiovascular risk factors. Future studies using the same methodology in different clinical contexts are necessary to continue developing this accessible, accurate, reproducible and easy-to-implement tool.^{10 - 13}

HIV/HAART-associated lipodystrophy assessed by patient self-report was not associated with traditional modifiable cardiovascular risk factors in HIV/AIDS patients attending an outpatient care center in a capital city of a middle-income country.