Alterations in Lipid Profiles in HIV-Infected Patients Treated With Protease Inhibitor Therapy Are Not Influenced by Diet  ,
* Department of Public Health and Nutrition
Infection Unit, Tufts University School of Medicine, Boston, Massachusetts; Correspondence: Christine Wanke, MD, Tufts University School of Medicine, 150 Harrison Avenue, Jaharis 266, Boston, MA 02111. Electronic mail may be sent to christine.wanke{at}tufts.edu. Background: The use of protease inhibitor (PI) –based highly active antiretroviral therapy (HAART) in patients infected with human immunodeficiency virus (HIV) has been linked to alterations in lipid profiles. Methods: Longitudinal cohort study. Lipid profiles were evaluated pre- and post-PI therapy in 49 HIV-infected patients. Diet was also evaluated to determine any contribution to alterations in lipid levels. Results: Pre- and post-PI-based HAART samples were examined from 42 men and 7 women, mean age 40.6 years. Mean CD4 count pre-PI was 242 ± 205 cells/mm3; HIV RNA was 4.2 log10 copies/mL; body mass index (BMI) was 24.7 ± 3.6 kg/m2; body fat was 17.1 ± 9.1 kg by bioelectrical impedance analysis (BIA), dietary intake was 2654 ± 732 kcal/day. Post-PI samples were collected 5.7 ± 1.66 months after starting therapy. CD4 counts increased to 384 ± 323 cells/mm3 and RNA decreased 1 log post-PI. Post-PI BMI (25.2 kg/m2), percent body fat (17.4%) and caloric intake (2656 kcal/day) were unchanged. Pre-PI total cholesterol (TC) in men was 167.6 ± 42.4 mg/dL; TC increased (190.4 ± 47.9 mg/dL) post-PI (p < .0001). Pre-PI triglycerides (TG) were 154.5 ± 109.4 mg/dL; TG increased post-PI (266.1 ± 363.6 mg/dL, p < .03). Low-density lipoprotein (LDL) -C was 97.8 ± 31 mg/dL pre-PI and rose to 107.1 ± 34.7 mg/dL post-PI (p < .05). High density lipoproteins (HDLs) were below desired levels before initiation of PI therapy and remained low. Conclusions: PI therapy significantly alters lipid profiles in HIV-infected patients. Dietary intake did not contribute to changes in lipid profile. More longitudinal studies are needed to demonstrate whether these alterations contribute to additional cardiovascular risk.
A syndrome of body-shape changes and metabolic abnormalities has been seen
in patients with human immunodeficiency virus (HIV) since
1996.1,2
The body-shape changes (central fat deposition and peripheral fat atrophy) and
metabolic abnormalities (insulin resistance and hyperlipidemia) may occur
independently of each
other.3 The
body-shape changes have been hypothesized to be associated with the duration
of antiretroviral therapy, nucleoside reverse transcriptase inhibitors, and
protease inhibitor (PI) therapy, but these associations are controversial, and
changes may occur in their
absence.4,5
Abnormalities in lipid metabolism occur with regularity after the initiation
of PI therapy, particularly ritonavir, and the association of the development
of hyperlipidemia with the use of specific PI therapy seems quite
strong.6,7
However, abnormalities in lipids were present before any therapy became
available for suppressing the virus. Low cholesterol levels and elevations in
triglycerides were often documented before PI therapy in the early part of the
HIV epidemic. Potential mechanisms for these baseline abnormalities include
elevated circulating levels of tumor necrosis factor (TNF) and
interferon- In the current era, HIV-infected patients have developed elevations in both serum triglyceride and non-high density lipoprotein (HDL) cholesterol levels. The etiology of the changes in lipid profiles is not completely clear. Although there are data indicating the PIs themselves induce alterations in lipid metabolism, there has also been speculation that some of the lipid changes are related to regain of weight and health with suppression of the viral load. Genetic predisposition to abnormal lipid metabolism is also possible. To address the potential role of diet in changes in cholesterol and triglycerides in the initiation of PI therapy, we documented changes in serum concentrations of lipids and cholesterol before and after the initiation of PI-based highly active antiretroviral therapy (HAART) in a group of HIV-infected men and women.
Study Population We have followed a cohort of 425 HIV-infected adults in a study called Nutrition for Healthy Living (NFHL) that has been ongoing since 1995. Patients enrolled with data collection between February 1995 and December 1999 were included in this analysis. For this analysis, individuals were selected from the cohort who had not been taking PI therapy at the time they enrolled in the study but began PI therapy while they were in the cohort, and for whom blood at pre-PI visit and a post-PI visit was available for analysis. Individuals were also selected if they had a 3-day food record at each of the identified visits. Details of the cohort are published elsewhere.9 Participant confidentiality was preserved according to the stipulations of the institutional review board at Tufts–New England Medical Center.
Clinical Data Collection Body Composition. At each study visit, body composition data were gathered by anthropometrics and bioelectric impedance analysis. For weight/height/body mass index (BMI), subjects were weighed on a digital scale fully dressed but without shoes and with removal of heavy clothing or objects. Weights were measured in the morning and before eating or drinking (12-hour fast). Height was measured without shoes using a stadiometer. Subjects stood on both feet, with heels, scapulae, and buttocks against the back wall, and the head held in a horizontal plane. BMI is calculated as weight (in kg) divided by height (in m) squared. Subscapular, suprailiac, and triceps skinfolds were measured each in triplicate with Lange Skin-fold Calipers (Beta Technology Incorporated, Cambridge, MD).10 Bioelectrical impedance analysis (BIA) measures the impedance or opposition to the flow of an electric current through the body fluids contained mainly in the lean and fat tissue. In this study, the single-frequency BIA was performed using a Quantum handheld analyzer (RJL Systems, Clinton Township, MI). The subjects were instructed not to have taken any alcohol or vigorous exercise for 48 hours before testing. The measurements were taken on subjects at least 1 hour after the last water intake and within 30 minutes of voiding. The subjects were clothed with the exception of the right shoe and sock and were lying supine with limbs not touching their bodies. Gel electrodes were placed on the skin of the wrist, metacarpal region, ankle, and metatarsal region on the right side of the body. Lean body mass and fat estimates were derived using the equation of Lukaski et al.11
Dietary Assessment. A self-administered, 3-day food record was
averaged to estimate daily nutrient intake. The participants were given a food
scale (Sunbeam Corporation, Mississauga, Ontario, Canada) and a ruler to
document portion sizes. Careful instruction was given to the participants on
the details of completing the food diary for accuracy. The dietary data from
the food records were analyzed using the Nutrition Data System (NDS) software
(Nutrition Coordinating Center, University of Minnesota, Version 2.92). The
estimated dietary data represent nutrient values from diet plus
vitamin/mineral supplements. Dietary variables compared included the
following: total calories and fat; carbohydrate; protein; alcohol;
cholesterol; total saturated fatty acid; monounsaturated fatty acid;
polyunsaturated fatty acid; Lipoprotein Analyses. Archived heparinized plasma from 49 participants who had samples obtained before and after initiation of PI therapy were assayed for lipoprotein analyses. Plasma total cholesterol, triglycerides, and HDL-cholesterol concentrations were measured, using standard enzymatic methods and reagents obtained from Abbott Diagnostics (Irving, TX) essentially as previously described.12 Serum LDL cholesterol levels were measured directly using kits from Equal Diagnostics (Exton, PA), and lipoprotein(a) (Lp(a)) was measured by mass, using Macra kits13 (Wampole Laboratories, Cranbury, NJ). Among-run precision studies indicate coefficients of variation of <2% for total cholesterol, <3% for triglycerides, <4% for LDL cholesterol, and <5% for HDL cholesterol and Lp(a). Lipid assays are standardized through the Centers for Disease Control Lipid Standardization Program.
Statistical Analysis
The characteristics of the 49 subjects are shown in Table 1. Eighty-six percent (42/49) of the group was men, and the mean age of the cohort was 40.6 years. Patients had been receiving PI therapy for a mean of 5.7 months at the time of the follow-up lipid analysis. Patients demonstrated a significant response to the initiation of PI-based HAART, as the mean CD4 count rose from 242 to 384 cells/mm3 and the mean viral load decreased from 4.2 log10 to 3.2 log10 pre- and post-PI. The patients were well nourished, with a mean BMI of 24.7 kg/m2 at the pre-PI visit and 25.2 kg/m2 at the post-PI visit. Body fat by BIA was 17.1% at the pre-PI visit and 17.4% at the post-PI visit. Triceps skinfolds as a measure of peripheral fat atrophy were unchanged with PI therapy. Thirty-seven percent of the group answered yes to a question about smoking >3 cigarettes per day.
The lipid profiles of the 42 men in the group are presented in Table 2. Mean serum total cholesterol and LDL cholesterol levels were significantly lower in the study group at the pre-PI visit than at the post-PI visit (p < .0001). Mean serum triglyceride levels at the pre-PI visit were above normal and rose significantly at the post-PI visit (p < .03), HDL cholesterol levels did not change after the introduction of PI therapy.
Table 3 presents the lipid profiles of the 7 women in the study group. The total cholesterol and LDL cholesterol changes followed the same pattern as the men, were low at the pre-PI study visit and rose significantly at the post-PI study visit (p = .002 and .002, respectively). The mean triglyceride level was high at the pre-PI visit but did not increase further during the study period (p = .784). As in men, HDL cholesterol was low at the start of the study but did not change over the study period (p = .583).
Table 4 presents the characteristics of the anti-retroviral therapy the participants were taking during the study period. Fifty-nine percent of the participants took Indinavir as their PI-based regimen, followed by 29% taking nelfinavir, 14% taking ritonavir, and 4% taking saquinavir.
Table 5 shows the dietary
intake by 3-day food records for the 49 men and women included in this study.
Diet records were taken from the same visits at which the blood samples for
lipid analysis were drawn. There was no statistically significant difference
in caloric intake or in the quality of diet when examined before and after the
initiation of PI therapy. Specifically, there was no difference in the percent
of calories from fat, saturated fat, or polyunsaturated fat, nor was there a
difference in
Table 6 shows the micronutrient intake of the 49 participants. Of the 8 selected vitamins and minerals analyzed, only vitamin E intake was statistically different before and after initiation of PI therapy. In this group of patients, vitamin E intake decreased after the initiation of PI.
Elevations in serum triglyceride levels have been noted in HIV-infected patients since the beginning of the HIV epidemic, although total cholesterol levels were often quite low in the early history of the epidemic. More recent studies have suggested that an increase in serum cholesterol and triglyceride concentrations occurs after the initiation of PI therapy.14–17 Our data concur with the literature in showing that total cholesterol and LDL cholesterol levels rise in HIV-infected patients when they begin PI-based HAART. The increase in serum concentrations of total and LDL cholesterol occurred in both men and women. Triglyceride levels increased further in men after the initiation of PI therapy (266.1 mg/dL), as they had been high (154.5 mg/dL) before the initiation of PI therapy. After the initiation of PI therapy, mean triglyceride levels for HIV-infected men were well above 200 mg/dL, the point at which intervention should be considered. It is important to note that these increases in cholesterol and triglycerides occurred in a population of HIV-infected patients who were well nourished (normal BMI), with normal body composition at the start of PI therapy. There was no significant change in weight, BMI, body composition by BIA, or anthropometry as a measure of body shape after the initiation of PI therapy in this group of patients. Although no change was detected in body composition and the patients were not obese, there is a limitation in our ability to determine normality in waist and hip measures, as they were only available on few subjects. Although body-shape changes, as seen in HIV-infected patients with lipodystrophy, may take longer to develop, there was no evidence of these body-shape changes in this study population over the 5.7 month follow-up period. In addition, BIA is limited in its ability to determine body composition in patients who may have regional changes. There was also no change in caloric intake after the initiation of PI therapy that would suggest a dietary reason for increases in serum cholesterol and triglyceride levels. In evaluating cardiovascular risk for patients infected with HIV, it will be important to determine whether these increases in cholesterol and triglycerides are progressive over time or if they plateau at roughly the same level as their HIV-negative counterparts. Notably, many of the dietary intake parameters, although not changed from before the initiation of PI therapy, are potentially atherogenic. Recommendations suggest that dietary intake of cholesterol be <300 mg/day; this population was consuming close to 400 mg/day. Recommendations suggest that fat intake should be <30% of total calories; these patients had been consuming approximately 35% of their calories as fat, and the mean intake of saturated fat was higher than the recommended <10% of total intake.18,19 Recommendations suggest that both monounsaturated and polyunsaturated fatty acid be approximately 10% of total calories, respectively; these patients had been consuming very low levels of polyunsaturated fatty acids (5.9%) and more than normal amounts of monounsaturated fatty acids (13.2%). Mean fiber intakes were below the recommended 20–35 g/day.18 Dietary intake of all of the selected micronutrients was above the recommended level according to the daily reference intake (DRI).20,21 These data suggest that many of the participants were using vitamin/mineral supplements. The number of patients in this study is small, but the observations may reflect issues that are pertinent to HIV-infected patients being treated with PI-based therapy. Overall, there are modifiable risks for cardiovascular disease in this population. Thirty-seven percent of the study group acknowledged that they were smokers, so attempts to address smoking as a cardiovascular risk factor in this population will become increasingly important. Other factors that may have contributed to the development of the abnormal lipid parameters include genetics or family history, inactivity, or direct impact of the PIs on the synthesis and clearance of lipids. Although there were no differences in dietary intake before and after initiation of PI therapy, the dietary data collected from these patients suggests that there is room for dietary education to promote general and cardiovascular health in this patient population and to minimize effects of the drug therapy on cardiovascular risk factors. The CD4 count and viral load response seen in these patients suggest that the patients were adherent to their PI-based HAART regimens. As dietary changes or weight changes did not play a role in the development of lipid changes in this group of HIV-infected patients, it is likely that the changes in lipids on HAART are related to the direct effect of the agents in the ART regimen. This work was supported by DK45734-10, M01R-R00054, 1K24AI055293, HL074819, 1P30AI42853, and 1R01AI46381.
Nutrition in Clinical Practice, Vol. 20, No. 6,
668-673 (2005)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
, leading to adipose tissue lipolysis and hepatic
lipogenesis.
Top
Methods
Results
-3 fatty acids; simple sugars; starch;
total fiber; pectin; vitamins A, C, E, B12; selenium; zinc; iron;
and folate. 
2 test. Pearson's correlation
coefficients were used to assess the relationship between dietary intake and
serum lipids.