The use of highly active antiretroviral therapy (HAART) has changed the prognosis of human immunodeficiency virus type 1 (HIV-1) infection, resulting in decreased mortality and improved quality of life (1). However, the use of these drugs is associated with the development of adverse effects, such as anemia, myopathy, and lipodystrophic syndrome (2). Since 1998, metabolic disorders have been described in HIV-1–infected individuals, especially in those patients on HAART, becoming an important clinical challenge in the global management of the disease (2- 3). Similarly, the nutritional aspects of HIV-1 infection have changed over the last years, because the severe and potentially lethal clinical problems related to progressive malnutrition have been replaced by subcutaneous fat atrophy, excess of visceral fat, and metabolic disorders, often in the context of undetectable viral load and immunological reconstitution. At the moment, weight gain, fat redistribution, obesity, and dyslipidemia, characterized by hypertriglyceridemia and/or hypercholesterolemia, are the new nutritional challenges of HIV/acquired immune deficiency syndrome (AIDS) individuals exposed to anti–HIV-1 therapy (4). The changes in lipid profile can be usually detected between the third and sixth months, but may develop even after a year (2- 3,5- 6).

Although its etiology is not completely elucidated, several factors may participate in the development of the HAART-associated dyslipidemia. Genetic and environmental factors may play important roles in triggering the process, as well as the current antiretroviral drugs and the HIV-1 virus itself (2- 3). The AIDS Clinical Trials Group has published guidelines for the management of dyslipidemia in HIV-1–infected people (3), which are mostly based on the recommendations of the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) guidelines regarding the intervention in the general population (7). Because there is little information on the efficacy and safety of pharmacological treatment of HAART-associated dyslipidemia in HIV-1–positive individuals, diet and physical exercise are considered first-choice treatment in current guidelines (3). In HIV-1–negative individuals, the effects of diet and exercise on lipid profile are well documented (8- 9). Although observational studies suggest that diet and exercise interventions are able to improve the lipid profile in HIV-1–positive patients (10- 11), recent randomized clinical trials do not corroborate these findings (12- 13). Moreover, a recent observational study suggests that the adherence to a Mediterranean diet does not influence serum lipid profiles during the first year of HAART (6). However, these are short-term studies, with reduced number of participants, and the use of long-term nutritional intervention in individuals who are started on HAART has not been assessed by randomized clinical trials. Therefore, the primary aim of this study was to assess the effects of nutritional intervention in total cholesterol and triglyceride levels in HIV-1–infected individuals who are started on HAART.

This was a randomized clinical trial conducted at a national reference center for the management of HIV-1–infected individuals in southern Brazil.

From March 2004 to April 2006, HIV-1–infected patients (diagnosed according to Centers for Disease Control revised criteria) (14), older than 17 years, without previous antiretroviral treatment, and with indication to begin HAART were consecutively enrolled at the HIV/AIDS outpatient clinic of the Hospital de Clínicas de Porto Alegre. Pregnant women; patients with active opportunistic disease, mental impairment, or diabetes mellitus; those being treated with drugs that cause changes in lipid profile or illicit drugs; and those who did not accept to participate or had logistic problems that precluded their participation in the protocol were not included. The study protocol was approved by the Ethics on Research Committee of Hospital de Clínicas de Porto Alegre and all participants signed an informed consent form.

During the initial assessment, all participants were instructed about nutritional status and lifestyle, with emphasis on the benefits of having a healthy diet and a 24-h dietary recall (24 h-DR) was used. Next, the individuals were randomized to intervention group or control group, on the basis of a random numbers table, generated by allocation sequence. Both groups were initiated on HAART therapy according to the recommendations of specialized medical team and the treatment was continued for 1 year. The intervention group received quarterly nutritional guidance from a registered dietician focused on dyslipidemia, based on the Phase II diet of the NCEP ATP III guidelines (7), and were followed up to the 12th month of treatment with antiretroviral therapy. The control group received the same baseline nutritional guidance, but there was no guided nutritional follow-up. Neither group received advice on physical activity. The 24 h-DR was used in both groups at the sixth month and at the end of the study (12th month). Lipid profile, body mass index (BMI), and waist-to-hip ratio (WHR) were assessed in both groups before HAART and every 3 months thereafter.

The waist circumference was measured mid-way between the lowest rib margin and the iliac crest, and the hip circumference was measured as the widest measure over the buttocks. The height and weight were measured using a mechanical anthropometric scale. BMI was calculated by dividing the weight by the square height (BMI = weight (kg) / [height(cm)]²). All anthropometric measures were obtained by an investigator (R.K.L.) who was unaware of patient allocation.

Diet programs were planned individually, on the basis of nutritional needs, socioeconomic status, and the dietary habits of each person. Diets were designed according to a total amount of calories in order to maintain or reduce weight, as needed, with 25% of the total calories in lipids (5% saturated fatty acids, 10% monounsaturated, and 10% polyunsaturated), 15% in proteins (50% high biologic value), and 60% in carbohydrates, including 30 g per day of fiber and 200 mg of dietary cholesterol. In addition, the volunteers were instructed not to eat anything containing trans fatty acids (9). The diet plan included a detailed description of the types of foods per group, with possible replacements, times, and daily amounts in home measures, besides a list of not-allowed types of foods. Regarding dietary counseling, patients received instructions about the selection and amounts of foods, as well as cooking techniques. The directions were provided after dietary history and were mainly composed of information on the adjustment of total fat intake, reduction of saturated fat, cholesterol, and trans fatty acids, as well as increase of fibers, especially of the soluble type. Information on food intake was obtained using the 24 h-DR method, through which the individuals provided detailed reports on food and beverages consumed during the 24 h before the interview. Questions were asked about amounts and home-measured portion sizes. In order to increase the reliability of the measurements reported, a manual of photographic record for dietary inquiries was used as a support material (15). Home measures were converted into grams and milliliters for quantitative analysis of the consumed energy and nutrients through the Nutritional Support Program (version 1.5) of the Health Informatics Center of Escola Paulista de Medicina (São Paulo, Brazil). Food and preparations that were apart from the list provided by the program were included through complementary tables (16- 17), or nutritional information on the labels of the manufactured products. In addition, the total calories, carbohydrates, proteins, lipids, fibers, cholesterol, saturated fatty acids, and monounsaturated and polyunsaturated acids were selected for analysis according to the recommendations of the guidelines of the NCEP ATP III (7) and of the AIDS Clinical Trials Group (2).

Lipid profile included measurements of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C) after a 12-h fast. TC and TG levels were measured with colorimetric enzymatic methods and HDL-C was measured by the homogeneous colorimetric enzymatic method (Hitachi 917, Roche Diagnostics GmbH, Mannheim, Germany). LDL-C was calculated using the Friedewald Formula if TG levels were <400 mg/dl (18). Dyslipidemia was defined as plasma levels of fasting TG ≥150 mg/dl and/or fasting total TC ≥200 mg/dl and/or LDL-C ≥130 mg/dl (9).

Initially, a sample size was estimated in 88 individuals for each group, considering an increase of 20% in the TG levels without nutritional intervention in persons who start HAART (19), with a statistical power of 80% for an α = 5%. A planned interim analysis was performed when 60 patients had concluded 6 months of follow-up, for reassessment of the statistical power. In this interim analysis, it was found that a sample size of 26 individuals in each group would have enough power to detect a clinically significant effect of the intervention. Therefore, recruitment was stopped. Data were analyzed on the Statistical Package for Social Sciences (version 16.0, SPSS, Chicago, Illinois). Continuous variables are expressed as mean ± SD and categorical variables are expressed as proportions. After randomization, groups were compared by means of the Student t test for continuous variables and the chi-square or the Fisher exact tests for categorical variables. The measurements taken at different time intervals were compared by linear mixed models (20), which were fitted considering metabolic and lipid parameters as dependent variables to evaluate the effect of group allocation, adjusting for time effect (interaction term of time · group). All analyses were performed by intention to treat, and a p value <0.05 was considered statistically significant.

Of the 105 screened individuals, 90 (86%) were considered eligible and were included in the study, but 7 were excluded from the final analysis due to nonadherence to HAART, confirmed by a viral load >5 copies/ml (5 in the control group) (Figure 1). All 83 individuals reached the sixth month of the study. After this period, 1 woman in the intervention group became pregnant, and 2 individuals were excluded from the control group (1 death from cranial fracture due to seizures and 1 loss to follow-up). Statins were prescribed to 2 individuals in this period in the control group.

(Table 1) shows the demographic, anthropometric, and biochemical characteristics, as well as the food intake and HAART prescribed at baseline. The groups presented similar characteristics, with predominance of women. Therapeutic regimens consisted of 2 nucleoside analog reverse transcriptase inhibitors plus 1 protease inhibitor (PI), with or without low-dose ritonavir (19%) or non-nucleoside reverse transcriptase inhibitor (NNRTI) (81%), according to Brazilian guidelines (21). The most commonly prescribed HAART regimen was the combination of zidovudine, lamivudine, and efavirenz. During the study, 9 individuals needed to switch therapy due to toxicity (5 on intervention group). However, all antiretroviral changes occurred in the first 3 months of the study (diet: 66 ± 33 days; control: 67 ± 29 days) and were not related to dyslipidemia (Table 2).

(Table 3) shows the estimate of the usual intake of calories and macronutrients that was reported in the 24 h-DR at baseline, at 6 months and at 12 months of the study. The nutritional intervention resulted in the reduction of total calories, cholesterol, and percentage of lipid and saturated fatty acid intake, as well as increased percentage of carbohydrate and fiber consumption. The percentage of protein and polyunsaturated fatty acids remained stable throughout the study in both groups. However, there was a significant reduction in the percentage of monounsaturated fatty acid intake in the group that received nutritional intervention.

(Table 4) shows the results on anthropometric measures. BMI remained stable in the intervention group, while in the control group there was a significant increase after 12 months of nutritional follow-up (p < 0.001) (Figure 2A). The WHR presented a slight increase in both groups, with a significantly smaller increment in the diet group.

The results of lipid profile are shown on (Table 3) and (Figure 2). The nutritional follow-up maintained TC plasma levels (Figure 2B) and LDL-C levels (Figure 2C) stable in the intervention group, with significant increments in the control group (p < 0.001). TG levels decreased in the intervention group, while in the control group these levels had a significant increase (p < 0.001) (Figure 2D). After 6 months, TC, LDL-C, and TG levels were significantly different between diet and control group, after adjustment for time effect (interaction term p < 0.001). There was a similar increase in HDL-C in both groups throughout the study.

At the end of the 1-year period, hypercholesterolemia was diagnosed in 14 (39%) individuals in the control group and only in 3 (7%) in the intervention group (p=0.001). Hypertriglyceridemia was found in 20 (51%) and in 7 (16%) individuals, respectively (p = 0.001). Overall, 9 (21%) individuals who received nutritional intervention developed alterations in lipid profile compatible with dyslipidemia, compared with 26 (68%) in the control group (p < 0.001) (Table 5).

In this 1-year randomized trial, we demonstrate, for the first time, that a diet based on the NCEP ATP III guidelines (7) was able to prevent dyslipidemia induced by HAART in patients who were previously naive to these medications. The nutritional guidelines used in this study focused on the quality of the lipids consumed, with the purpose of reducing the saturated and trans fats and increasing monounsaturated and polyunsaturated fatty acids and fibers, as well as decreasing daily cholesterol intake. Our study demonstrates that this nutritional intervention controlled the changes in lipid profile in HIV-1–infected individuals who started HAART. In the group that received dietary intervention, mean plasma levels of total cholesterol, triglycerides, LDL-C, and BMI remained stable during the 12-month follow-up, while the control group presented the expected increment.

Our findings also confirm that individuals who start HAART are at risk to develop dyslipidemia (almost 70% in the control group of our study) according to revised NCEP ATP III criteria (even with initial dietary guidance) (7), and it could be avoided by the beneficial effects of nutritional intervention as done in our study. Fitch et al. (13) randomized a small group HIV-1–infected individuals with metabolic syndrome to an intensive lifestyle modification program, which included dietary intervention, or to a control group. After 6 months, despite improvement in anthropometric variables, blood pressure, and physical activity, lipid profile was not improved by the lifestyle modification program. Our group has also failed to demonstrate the effect of exercise training on the lipid profile of HIV-1–infected individuals (12). Both of these trials (12- 13) evaluated patients on chronic HAART, while in the present study only patients naive from HAART were included. These findings raise the hypothesis that, for a dietary intervention to be effective in controlling dyslipidemia in HIV-1–infected individuals, it should be started early.

Our findings are consistent with previous uncontrolled studies. Barrios and coworkers (22) found a significant reduction in TC and TG in a large cohort of patients who had good compliance to a low-fat diet for 6 months. In the retrospective analysis of Batterham and coworkers (10), there were also significant reductions in TC and TG after changes in the diet. In the present study, the dietary intervention was also able to prevent the increase in LDL-C associated with HAART, probably due to the reduced intake of saturated fatty acids and cholesterol. In isoenergetic conditions, it is expected that 1% reduction in saturated fat intake and 100 mg/day of cholesterol may cause a reduction of 1 to 3% in LDL-C levels (23). The present study is the first randomized, controlled trial to isolate the effects of dietary intervention in HIV-1–infected individuals who are started on HAART. Previous controlled trials evaluated the effects of diet associated with other interventions, including physical activity (11- 12), and the use of drugs such as statins (24) and fibrates (25) or supplementation with fatty acids (26).

In the present study, both groups had significant and similar increments in HDL-C with HAART. Studies performed before the HAART era associated HIV infection with low HDL-C levels (27). More advanced stages of HIV-1 infection and lower CD4 lymphocyte count have been associated with lower HDL-C levels (28), and it has been shown that HDL-C levels may remain low despite HAART (29). It has also been shown that when NNRTI are used in the beginning of the therapy or as a replacement for a PI, they can increase the HDL-C levels (13% and 27%, respectively) (30- 31). Therefore, it is possible that the viremic control induced by HAART, as well as the frequent use of efavirenz, might have contributed to the increase in HDL-C levels in both groups.

Our patients presented mean BMI and WHR within the limits of normality at baseline. It is well known that patients who are started on HAART may present body changes such as accumulation (lipohypertrophy) and/or with loss (lipoatrophy) in body fat (32) and previous cohort studies in the HAART era have consistently found an increase in body weight after the diagnosis of HIV (33), but the mechanisms responsible for the change have not been elucidated (32). After 1 year, the dietary intervention prevented some of the anthropometric changes associated with the lipodystrophic syndrome, by preventing part of the weight gain and fat redistribution. This was most likely related to the reduction in caloric intake in the group who followed the diet, while the control group increased the caloric intake (Table 3). Therefore, our findings support the hypothesis that these patients may gain more weight in an attempt to obscure their diagnosis (33). Currently, there is no consensus on which is the method that better reflects the real food intake of humans (34). Despite the known limitations of the 24 h-DR (34), it was sensitive enough to detect the effects of our dietary intervention and can be easily reproduced in other clinical settings.

Several theories have been proposed to explain the association between HAART and dyslipidemia, although none of them explains all aspects (32). It is important to notice that some drugs are more likely to produce lipid changes, mainly stavudine (nucleoside reverse transcriptase inhibitor), efavirenz (NNRTI), and PIs (with the possible exception of atazanavir, which is the PI less associated to dyslipidemia) (35- 36). Nevertheless, as shown in (Table 2), the groups were comparable regarding HAART regimen. Overall, 23% of the individuals were using 1 PI in their regimen; however, 10% were on atazanavir as a part of HAART, which is known to be less associated with the development of dyslipidemia (35- 36).

Strengths of this study include its randomized design, its intention-to-treat analysis, a high compliance to the intervention, and the large number of women included. Because we have previously shown that aerobic exercise training had no significant effect on blood lipids in HIV-1–positive individuals on HAART (14), the present trial was focused on the isolated effects of dietary intervention. Therefore, our findings do not exclude the possibility that a more complete lifestyle intervention could have resulted in different outcomes. No advice on physical activity was given to the patients and we also did not evaluate physical activity during the trial. However, the possibility of cointervention, with increase in physical activity in the group that received dietary advice, is unlikely. The focus on HIV-1–infected patients who were naive of HAART allowed us to demonstrate a clear effect of dietary intervention in preventing dyslipidemia in this setting; however, it left unanswered the question of whether the same intervention can revert dyslipidemia in patients who are already taking HAART, and this should be addressed by future controlled trials. Likewise, the 1-year follow-up in a cohort of HIV-1–infected individuals who are at low risk for cardiovascular events cannot ascertain if the control of dyslipidemia might have any impact in prognosis. However, if we use the results of the meta-analysis involving more than 90,000 HIV-negative individuals (37), which showed that for each 39 mg/dl reduction in LDL-C levels there is a 23% reduction in the rate of cardiovascular events, we could estimate a potential reduction of 15% in the rate of cardiovascular events in our HIV-1–infected patients with the use of dietary intervention.

Our study showed the efficacy of nutritional intervention in the prevention of changes in lipid profile of HIV-positive individuals who start HAART. These findings indicate that nutritional intervention should be considered for all patients who start HAART. Future studies should be conducted to evaluate the impact of this intervention on clinical outcomes associated with lipid profile.