Abstract
Objective
This study assessed the effect of obesity on metabolic and cardiovascular disease risk factors in HIV-infected adults on antiretroviral therapy (ART) with sustained virologic suppression.
Design
Observational, comparative cohort study with three group-matched arms: 35 non-obese and 35 obese HIV-infected persons on efavirenz, tenofovir, and emtricitabine with plasma HIV-1 RNA <50 copies/ml for >2 years, and 30 obese HIV-uninfected controls. Subjects did not have diabetes or known cardiovascular disease.
Methods
We compared glucose tolerance, serum lipids, brachial artery flow mediated dilation (FMD), carotid intima-media thickness (cIMT), and soluble inflammatory and vascular adhesion markers between non-obese and obese HIV-infected subjects, and between obese HIV-infected and HIV-uninfected subjects, using Wilcoxon rank sum tests and multivariate linear regression.
Results
The cohort was 52% male and 48% non-white. Non-obese and obese HIV-infected subjects did not differ by clinical or demographic characteristics. HIV-uninfected obese controls were younger than obese HIV-infected subjects and less likely to smoke (p≤0.03 for both). Among HIV-infected subjects, obesity was associated with greater insulin release, lower insulin sensitivity, and higher serum hsCRP, IL-6, and TNF-α receptor 1 levels (p<0.001), but similar lipid profiles, sCD14, sCD163, ICAM-1 and VCAM-1, and cIMT and FMD. In contrast, HIV-infected subjects had adverse lipid changes, and greater circulating ICAM-1, VCAM-1 and sCD14, compared to HIV-uninfected controls after adjusting for age and other factors.
Conclusions
Obesity impairs glucose metabolism and contributes to circulating hsCRP, IL-6, and TNF-α receptor 1 levels, but has few additive effects on dyslipidemia and endothelial activation, in HIV-infected adults on long-term ART.
Introduction
HIV-infected persons on long-term antiretroviral therapy (ART) are at increased risk of developing cardiovascular and metabolic disease as compared to HIV-uninfected individuals with otherwise similar risk profiles [1–5]. This observation has been attributed to HIV-related factors such as persistent systemic or vascular inflammation [6, 7], antiretroviral toxicity [8, 9], or changes in immune cell populations or function [10, 11]. However, epidemiologic studies have demonstrated heterogeneous effects of non-HIV related risk factors on the incidence of non-communicable diseases (NCDs) in the context of HIV infection. In particular, a higher body mass index (BMI) is associated with an increased risk of an incident diabetes mellitus diagnosis among both HIV-infected and HIV–uninfected persons, but the incremental effect of each unit increase in BMI on diabetes risk is disproportionately greater in HIV-infected persons as compared to HIV-uninfected persons [5, 8, 12]. In contrast, large epidemiologic studies have not found that a higher BMI increases the risk of incident cardiovascular events in HIV-infected persons [3, 4, 13]. Interpreting these findings on BMI and NCD incidence in the HIV-infected population is hampered by a paucity of clinical data on how body composition and ART-treated HIV infection interact to affect metabolic and cardiovascular parameters.
The prevalence of obesity (a BMI >30 kg/m2) among HIV-infected individuals in the United States is approaching parity with the general population and is particularly high among women and minorities [14–18]. As patients can now survive decades on ART, the identification of individuals at high risk for developing chronic comorbid medical conditions is increasingly important for clinical care. In this study we use a comparative cohort approach to first assess how obesity affects glucose tolerance, lipid profiles, vascular health, and systemic inflammation in HIV-infected adults on stable, long-term ART treatment, and second to assess how the presence of treated HIV-infection affects the same outcomes in obese individuals.
Methods
We enrolled 70 HIV-infected patients on ART from the Vanderbilt Comprehensive Care Clinic and 30 obese (BMI >30 kg/m2), uninfected controls between April 2013 and September 2014. The HIV-infected subjects were distributed equally between four BMI categories of <25.0, 25.0–29.9, 30.0–34.9, and ≥35.0 kg/m2. Within each BMI strata similar numbers of males and females and whites and non-whites were enrolled. All subjects were on efavirenz, tenofovir, and emtricitabine (i.e., the combination pill Atripla) for at least the 6 months prior to enrollment and had been on ART treatment with persistent HIV-1 RNA measurements <50 copies/ml for at least the previous 2 years. Additional inclusion criteria were CD4+ count >350 cells/µl at the time of enrollment, no use of any anti-diabetic or statin (i.e., HMG CoA reductase inhibitor) medication in the prior 6 months, no self-reported heavy alcohol (defined as >11 drinks/week) or cocaine/amphetamine use, no active infectious conditions aside from HIV, and no previously diagnosed diabetes, cardiovascular disease (CVD), or rheumatologic disease recorded in the medical record.
Thirty healthy volunteers with obesity were recruited from the community to serve as controls. Controls were distributed equally between the BMI categories of 30.0–34.5 and >35 kg/m2 and group matched by sex and race with the HIV-infected subjects. The uninfected controls had not received any anti-diabetic or statin medications in the prior 6 months, did not report alcohol or illicit drug abuse, and had no active infectious conditions or previously diagnosed diabetes, CVD, or rheumatologic disease by self-report.
Data on ART history and CD4+ count and viral load values were obtained from the medical record. Data on smoking was obtained by self report. All 100 subjects underwent a 3-hour assessment in the Vanderbilt Clinical Research Center after fasting overnight for at least 8 hours (all visits began between 8 and 11 am). Brachial artery reactive hyperemia and bilateral carotid intima-media thickness (IMT) was measured using a Philips iE33 ultrasound with L9–3 linear transducer prior to any other procedures. A 3-lead ECG was attached to chest in standard manner and the patient relaxed in a supine position for 10 minutes. A blood pressure cuff was placed on the right forearm 1cm below the antecubital fossa and the arm was extended away from body 90 degrees resting on an armboard at the height of the bed. An image of the brachial artery 3–10 cm above the antecubital fossa was acquired on the EKG “R” wave to measure the artery diameter. The cuff was inflated to 50 mmHg above systolic pressure for 5 minutes. After deflation, EKG-gated measurements were acquired from still-frame images at 30 seconds, 60 seconds, 90 seconds and 120 seconds. The flow mediated dilation (FMD) was calculated as the largest percent increase in vessel diameter after cuff deflation.
Right common carotid artery (CCA) and carotid bulb images were acquired using EKG gating after patients were placed supine on the bed without a pillow and with the head turned 45 degrees to the left. The IMT was measured in plaque free arterial segments at the carotid bulb, and 1 cm from the bulb, as the distance between the inner echogenic line representing the blood-intima interface and the outer echogenic line representing the media-adventia border. A similar procedure was used for measurement of the proximal internal carotid artery (ICA) IMT at 1 cm, and all measurements were repeated on the left side of the head (in this analysis, we report IMT values from the right side only).
After ultrasound assessments were complete, fasting blood samples were collected. High-sensitivity C-reactive protein (hsCRP), total cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglycerides, and glucose were measured by the Vanderbilt Clinical Chemistry Laboratory. Plasma levels of soluble CD14 and CD163, two surface markers released into circulation by activated macrophages, were measured using ELISA (R&D Systems, Minneapolis, MN). Serum levels of interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α) receptor 1, intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) were measured in duplicate using a multiple immunoassay (MesoScale, Rockville, MD). Fasting insulin was measured by radioimmunoassay.
After the collection of the fasting blood sample, subjects ingested a 75 gram oral glucose dose dissolved in 12 ounces of water. Plasma glucose was again measured at 90 and 120 minutes, and insulin at 90 minutes. Fasting beta cell function and insulin sensitivity were calculated using the Homeostasis Model Assessment 2 (HOMA2) equation (https://www.dtu.ox.ac.uk/homacalculator) [19]. The oral glucose insulin sensitivity [OGIS] index (a model developed to approximate a glucose clamp study) was calculated using fasting, 90 minute, and 120 minute glucose and insulin values [20].
Anthropometric measurements were performed in triplicate and averaged. A full body dual-energy x-ray absorptiometry (DEXA) scan was performed on all subjects to measure regional and total bone mass, lean mass, and fat mass (GE Lunar Prodigy). A software algorithm estimated visceral fat mass.
Statistical analyses
Demographic, clinical, and body composition characteristics were compared in a pairwise fashion by obesity status in the HIV infected group, and by HIV status in the obese group using Wilcoxon rank sum or chi-square tests. Medians and interquartile ranges were calculated for continuous variables and percentages for categorical variables.
To assess whether obesity alters metabolic and cardiovascular parameters in the context of long-term, treated HIV infection, we compared each outcome variable according to obesity status in the HIV-infected participants using Wilcoxon Rank Sum tests. A secondary analysis using multivariable linear regression was also performed to assess the relationship of DEXA percentage body fat with each outcome variable after adjusting for age, sex, race, CD4+ count, current smoking status, and the duration of ART treatment. DEXA percent body fat was used rather than BMI as it represents a more accurate measurement of total adiposity and, unlike DEXA total body fat, it is less affected by differences in height. The outcome variables were natural log transformed and CD4+ count was square root transformed. Sensitivity analyses adjusted for hepatitis C co-infection and pre-treatment CD4+ count.
To assess whether HIV status affected metabolic and cardiovascular parameters in the context of obesity, we limited our analysis to the 35 obese HIV-infected and 30 obese HIV-uninfected subjects and utilized multivariable linear regression models incorporating HIV status, age, sex, race, smoking status, DEXA body fat, and an interaction term between HIV and body fat. The outcome variables were natural log transformed. In models where the interaction term p-value was >0.1, the regression coefficient for HIV status was calculated after removing the interaction term from the model. Sensitivity analyses were performed which adjusted for hepatitis C co-infection and pre-treatment CD4+ count.
The selected metabolic and cardiovascular outcome variables were grouped into biologically rational categories (glucose metabolism, adipokines, plasma lipids, inflammation biomarkers, vascular adhesion biomarkers, and vascular ultrasound measurements) and represented planned comparisons. The analysis strategy was to assess the probability of an association between the exposure variable and members of each category, and no adjustments were made for multiple comparisons [21]. Analyses were conducted using SPSS 22.0.0 (IBM) and R Statistical Software (http://www.R-project.org).
Results
Comparison of HIV-infected Subjects by Obesity Status
The clinical and demographic characteristics of the HIV-infected subjects stratified by obesity status are shown in Table 1. Age, race, sex, smoking status, CD4+ count at enrollment and ART initiation, duration of ART treatment, and hepatitis C prevalence were similar between the HIV-infected non-obese and obese (p>0.05 for all comparisons).
Table 1.
Clinical characteristics and body composition of non-obese and obese HIV-infected subjects
| Variable | Non-obese (n=35) | Obese (n=35) | p-value |
|---|---|---|---|
| Age, median years (IQR) | 45 (38, 49) | 46 (39, 51) | 0.61 |
| Female, % | 14 (40%) | 16 (46%) | 0.41 |
| Non-white, % | 18 (51%) | 20 (57%) | 0.63 |
| BMI, median kg/m2 | 23.9 (21.9, 26.5) | 35.6 (33.0, 40.1) | <0.001 |
| Smoker, % | 12 (34%) | 13 (37%) | 0.80 |
| Hepatitis C, % | 4 (11%) | 4 (11%) | NA |
| CD4 at enrollment, cells/µl | 621 (504, 924) | 758 (605, 966) | 0.08 |
| CD4% at enrollment | 36 (30, 40) | 38 (32, 42) | 0.15 |
| CD4 at ART initiation, cells/µl | 250 (142, 307) | 262 (132, 400) | 0.59 |
| Duration of ART treatment, years | 6.05 (4.33, 9.88) | 6.65 (4.42, 11.15) | 0.58 |
| DEXA body composition measurements | |||
| Bone mass, kg | 2.73 (2.21, 3.07) | 2.93 (2.54, 3.39) | 0.03 |
| Total lean mass, kg | 48. 4 (41.0, 55.4) | 56.2 (48.9, 65.9) | <0.001 |
| Fat mass, kg | 21.2 (14.8, 27.7) | 46.4 (37.9, 52.2) | <0.001 |
Abbreviations: ART, antiretroviral therapy; BMI, body mass index; DEXA, dual-energy x-ray absorptiometry; IQR, interquartile range
Among the HIV-infected subjects, obesity was associated with increased HOMA 2 model β-cell insulin release, and lower HOMA 2 and OGIS 120 index insulin sensitivity (Table 2; p≤0.001 for all). There was no significant difference in percent glycosolated hemoglobin, which likely reflects compensation from the over 2-fold higher median fasting insulin levels in the obese (p<0.001). Obesity was closely associated with higher serum hsCRP, IL-6, and TNF-α receptor 1 levels (p<0.001), but not soluble CD14 or CD163. We observed no difference between obese and non-obese HIV-infected subjects in median plasma lipid levels, ICAM-1 and VCAM-1 levels, or measurements of carotid IMT or brachial artery FMD. Similarly, in the adjusted linear regression model, DEXA percent fat mass was associated with insulin resistance and hsCRP, IL-6, and TNF-α receptor 1 levels, but not plasma lipids, vascular adhesion molecules, or vascular ultrasound measurements (Table 3). These results were similar when the model was further adjusted for hepatitis C and pre-treatment CD4+ count (data not shown).
Table 2.
Comparison of metabolic and cardiovascular parameters between non-obese and obese HIV-infected subjects
| Outcome variable | Non-obese (n=35) | Obese (n=35) | p-value |
|---|---|---|---|
| Glucose metabolism assessments | |||
| HOMA2 beta-cell function, % | 100 (80, 128) | 173 (116, 211) | <0.001 |
| HOMA2 insulin sensitivity, % | 130 (74, 191) | 58 (41, 89) | <0.001 |
| OGIS 120, ml/min/m2 | 447 (401, 485) | 379 (328, 422) | <0.001 |
| Hemoglobin A1c, % | 5.1 (4.9, 5.5) | 5.2 (5.0, 5.6) | 0.14 |
| Insulin, µU/ml | 6.1 (4.2, 10.6) | 13.3 (8.6, 19.4) | <0.001 |
| Adipokine levels | |||
| Leptin, ng/ml | 9.2 (3.9, 13.4) | 30.4 (18.3, 42.6) | <0.001 |
| Adiponectin, µg/ml | 12.4 (6.7, 16.5) | 9.1 (5.2, 12.0) | 0.03 |
| Resistin, ng/ml | 17.5 (11.8, 23.3) | 20.0 (14.6, 26.2) | 0.28 |
| Plasma lipids | |||
| Total cholesterol, mg/dl | 174 (152, 203) | 177 (155, 200) | 0.59 |
| HDL, mg/dl | 46 (35, 64) | 44 (39, 49) | 0.28 |
| LDL, mg/dl | 101 (85, 122) | 111 (88, 129) | 0.50 |
| Triglycerides, mg/dl | 94 (66, 131) | 104 (85, 152) | 0.12 |
| Inflammation biomarkers | |||
| High sensitivity C-reactive protein, mg/l | 1.6 (0.7, 2.8) | 5.8 (2.0, 9.4) | <0.001 |
| Interleukin-6, pg/ml | 2.5 (1.6, 3.8) | 4.2 (2.9, 7.1) | <0.001 |
| Tumor necrosis factor-alpha receptor 1, ng/ml | 11.0 (9.5, 12.2) | 12.2 (10.2, 15.1) | 0.048 |
| Soluble CD163, ng/ml | 500 (413, 743) | 563 (417, 664) | 0.36 |
| Soluble CD14, µg/ml | 1.69 (1.44, 2.00) | 1.69 (1.50, 1.92) | 0.74 |
| Vascular adhesion biomarkers | |||
| Intercellular adhesion molecule 1, ng/ml | 513 (441, 663) | 557 (448, 617) | 0.55 |
| Vascular cell adhesion molecule 1, ng/ml | 583 (449, 630) | 587 (470, 665) | 0.83 |
| Vascular ultrasound measurements | |||
| Carotid bulb intima-media thickness (IMT), cm | 0.062 (0.054, 0.069) | 0.062 (0.057, 0.077) | 0.25 |
| Common carotid IMT, cm | 0.057 (0.052, 0.062) | 0.062 (0.052, 0.071) | 0.11 |
| Internal carotid IMT, cm | 0.056 (0.047, 0.067) | 0.053 (0.044, 0.071) | 0.97 |
| Brachial artery flow-mediated dilation, % | 9.0 (5.9, 11.6) | 8.4 (4.8, 10.6) | 0.31 |
Abbreviations: HLD, high-density lipoprotein; HOMA, Homeostasis Model Assessment; OGIS, oral glucose insulin sensitivity; LDL, low-density lipoprotein
Table 3.
Multivariable linear regression model for the effect of fat mass on metabolic and cardiovascular parameters among HIV-infected subjects on long-term ART (n=70)
| Outcome variable | Regression coefficient for percent body fat |
p-value |
|---|---|---|
| Glucose metabolism assessments | ||
| HOMA2 beta-cell function | 0.54 | <0.001 |
| HOMA2 insulin sensitivity | −0.81 | <0.001 |
| OGIS 120 | −0.48 | 0.001 |
| Hemoglobin A1c | 0.10 | 0.49 |
| Insulin | 0.79 | <0.001 |
| Adipokine levels | ||
| Leptin | 0.96 | <0.001 |
| Adiponectin | −0.21 | 0.13 |
| Resistin | 0.30 | 0.08 |
| Plasma lipids | ||
| Fasting total cholesterol | 0.20 | 0.22 |
| Fasting HDL | −0.19 | 0.18 |
| Fasting LDL | 0.31 | 0.05 |
| Fasting triglycerides | 0.26 | 0.09 |
| Inflammation biomarkers | ||
| High sensitivity C-reactive protein, mg/l | 0.53 | <0.001 |
| Interleukin-6, pg/ml | 0.45 | 0.001 |
| Tumor necrosis factor-alpha receptor 1, ng/ml | 0.32 | 0.01 |
| Soluble CD163 | 0.24 | 0.13 |
| Soluble CD14 | −0.12 | 0.42 |
| Vascular adhesion biomarkers | ||
| Intercellular adhesion molecule 1 (ICAM-1), ng/ml | 0.09 | 0.52 |
| Vascular cell adhesion molecule 1 (VCAM-1), ng/ml | 0.08 | 0.57 |
| Vascular ultrasound measurements | ||
| Carotid bulb intima-media thickness (IMT) | 0.26 | 0.08 |
| Common carotid IMT | 0.19 | 0.24 |
| Internal carotid IMT | 0.07 | 0.68 |
| Brachial artery flow-mediated dilation | −0.13 | 0.41 |
Multivariable model adjusted for age, sex, race, CD4 count (square root transformed), duration of ART treatment, smoking, and DEXA percent fat
Abbreviations: ART, antiretroviral therapy; DEXA, dual-energy x-ray absorptiometry; HLD, high-density lipoprotein; HOMA, Homeostasis Model Assessment; OGIS, oral glucose insulin sensitivity; LDL, low-density lipoprotein
Comparison of Obese Subjects by HIV Status
The clinical and demographic characteristics of the obese subjects stratified by HIV status are shown in Table 4. The obese HIV-infected had a higher median age compared to the obese controls, 46 versus 37 years (p=0.01), and a higher smoking and hepatitis C prevalence, but did not significantly differ according to sex or race, or total bone mass, lean mass, or fat mass. However, obese HIV+ subjects had a higher trunk-to-appendicular fat ratio, a predictor of cardiovascular disease, compared to controls (1.58 versus 1.32; p=0.05) and higher calculated visceral fat (1.97 versus 1.60 kg, p=0.04).
Table 4.
Clinical characteristics and body composition in obese HIV-infected subjects versus obese controls
| Obese HIV-infected (n=35) |
Obese HIV uninfected controls (n=30) |
p-value | |
|---|---|---|---|
| Age, median years (IQR) | 46 (39, 51) | 37 (28, 44) | 0.01 |
| Female, % | 16 (46%) | 18 (60%) | 0.25 |
| Non-white, % | 20 (57%) | 11 (37%) | 0.10 |
| BMI, median kg/m2 | 35.6 (33.0, 40.1) | 35.8 (31.3, 41.0) | 0.49 |
| Smoker, % | 13 (37%) | 4 (13%) | 0.03 |
| Hepatitis C, % | 4 (11%) | 0 | 0.06 |
| Anthropometric measurements | |||
| Waist circumference, cm | 122 (109, 130) | 114 (99, 129) | 0.80 |
| Waist-to-hip ratio | 1.01 (0.95, 1.06) | 0.94 (0.83, 1.03) | 0.24 |
| DEXA body composition measurements | |||
| Bone mass, kg | 2.93 (2.54, 3.39) | 3.05 (2.66, 3.52) | 0.73 |
| Total lean mass, kg | 56.2 (48.9, 65.9) | 57.0 (46.1, 70.0) | 0.93 |
| Fat mass, kg | 46.4 (37.9, 52.2) | 44.1 (36.0, 53.0) | 0.78 |
| Fat mass % | 42.7 (37.4, 48.5) | 42.0 (38.9, 47.1) | 0.67 |
| Limb fat mass, kg | 16.7 (14.4, 20.7) | 18.4 (14.9, 22.7) | 0.31 |
| Limb fat % | 16.2 (12.3, 19.8) | 17.5 (14.8, 21.2) | 0.24 |
| Trunk fat mass, kg | 27.1 (22.3, 31.9) | 24.8 (19.4, 32.2) | 0.25 |
| Trunk fat % | 25.6 (21.5, 28.5) | 23.7 (20.7, 26.5) | 0.12 |
| Trunk-to-appendicular fat ratio | 1.58 (1.40, 1.99) | 1.32 (1.03, 1.75) | 0.05 |
| Calculated visceral fat, kg* | 1.97 (1.51, 2.91) | 1.60 (0.94, 2.74) | 0.04 |
Abbreviations: BMI, body mass index; DEXA, dual-energy x-ray absorptiometry; IQR, interquartile range
Calculated from DEXA scan data by GE Lunar Prodigy software
In contrast to the metabolic findings in the HIV-infected subjects, HIV status was found to be associated with adverse changes in lipid profiles and higher soluble endothelial adhesion molecules in obese subjects in a regression model adjusted for age and other covariates (Table 5). HIV status was associated with higher total cholesterol, LDL, and triglycerides, and lower HDL, and higher plasma ICAM-1 and VCAM-1. HIV-infected subjects also had significantly higher sCD14, a marker of monocyte activation. We observed no significant differences in β-cell insulin release, insulin sensitivity, serum hsCRP, IL-6, TNF-α receptor 1, or sCD163 levels, or ultrasound measurements of carotid IMT or brachial artery FMD. These results were similar when the model was further adjusted for hepatitis C co-infection.
Table 5.
Relationship of HIV status to metabolic and cardiovascular parameters among obese subjects (35 HIV-infected, 30 HIV-uninfected)
| Outcome variable | Regression coefficient for HIV status* |
p-value |
|---|---|---|
| Glucose metabolism assessments | ||
| HOMA2 beta-cell function | −0.04 | 0.79 |
| HOMA2 insulin sensitivity | 0.13 | 0.31 |
| OGIS 120 | −0.06 | 0.69 |
| Hemoglobin A1c | −0.16 | 0.21 |
| Insulin | −0.12 | 0.34 |
| Adipokine levels | ||
| Leptin | −0.08 | 0.35 |
| Adiponectin | −0.01 | 0.96 |
| Resistin | −0.40 | 0.004 |
| Plasma lipids | ||
| Total cholesterol | 0.11 | 0.03 |
| HDL | −0.13 | 0.04 |
| LDL | 0.08 | <0.01 |
| Triglycerides | 0.20 | 0.03 |
| Inflammation biomarkers | ||
| High sensitivity C-reactive protein, mg/l | −0.01 | 0.92 |
| Interleukin-6, pg/ml | 0.04 | 0.70 |
| Tumor necrosis factor-alpha receptor 1, ng/ml | 0.03 | 0.81 |
| Soluble CD163 | 0.13 | 0.33 |
| Soluble CD14 | 0.54 | <0.01 |
| Vascular adhesion biomarkers | ||
| Intercellular adhesion molecule 1 (ICAM-1), ng/ml | 0.33 | 0.01 |
| Vascular cell adhesion molecule 1 (VCAM-1), ng/ml | 0.31 | 0.02 |
| Vascular ultrasound measurements | ||
| Carotid bulb intima-media thickness (IMT) | −0.09 | 0.42 |
| Common carotid IMT | −0.04 | 0.74 |
| Internal carotid IMT | −0.18 | 0.19 |
| Brachial artery flow-mediated dilation | 0.05 | 0.76 |
Multivariable model adjusted for age, sex, race, HIV status, smoking status, DEXA total fat mass, and interaction term for DEXA fat mass and HIV status
In models where the interaction term p-value was >0.1, the regression coefficient for HIV status is calculated after removing the interaction term from the model
Abbreviations: DEXA, dual-energy x-ray absorptiometry; HLD, high-density lipoprotein; HOMA, Homeostasis Model Assessment; OGIS, oral glucose insulin sensitivity; LDL, low-density lipoprotein
The absolute values of the cardiovascular and metabolic parameters in obese HIV+ subjects and obese controls are compared using Wilcoxon rank sum tests in the Supplementary Table. The variables which differ significantly according to HIV status are similar to the adjusted regression model with the exception of lipids, which we attribute to the lack of adjustment for age and smoking.
Discussion
In HIV-infected adults on long-term, non-protease inhibitor-based ART and without previously diagnosed metabolic disease or CVD, obesity was associated with increased insulin resistance and systemic inflammation, but obesity did not appear to adversely affect HbA1c or key cardiovascular risk factors including lipid profiles, circulating vascular endothelial adhesion molecules, or measurements of carotid IMT or brachial artery reactivity. Of note, the greater serum insulin levels and calculated insulin resistance in our obese HIV-infected subjects was not accompanied by higher HbA1c values. This indicates an ability to compensate for insulin resistance with greater insulin secretion to maintain glucose homeostasis, but represents a condition more likely to progress to clinical diabetes.
Among obese subjects with similar total and regional adiposity, HIV infection was not associated with significant differences in insulin secretion, calculated insulin resistance, or plasma inflammatory markers (except soluble CD14), but HIV status was associated with adverse changes in plasma lipids and vascular adhesion molecules. We interpret our findings as an indication that greater adipose tissue stores contribute to insulin resistance and circulating cytokines in HIV-infected persons on ART, but any additive effects of obesity on changes in lipids or ICAM-1 and VCAM-1 in HIV-infected persons is masked by the adverse effects of HIV infection alone on these parameters.
Glucose metabolism
Our finding that progressive adiposity is accompanied by lower glucose tolerance in HIV patients is consistent with prior epidemiologic analyses of diabetes in this population. In a multi-country study of 33,000 subjects in the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) cohort, a BMI >30 kg/m2 was associated with a 4.5-fold higher risk of incident diabetes compared to those with a BMI 18–26 kg/m2, which was independent of cumulative exposure to stavudine, zidovudine, didanosine, and other ART agents known to cause alterations in fat partitioning and adipocyte energy metabolism [8]. A longitudinal study of 1046 HIV-infected French adults found the incidence of diabetes rose in a stepwise fashion with higher BMI strata, again independent of exposure to individual ART medications [12].
While prior studies have reported HIV-infected patients to have significantly higher rates of incident diabetes compared HIV-uninfected persons, we did not observe an association between HIV status and glucose metabolism in the obese subjects [2, 5]. We attribute this to two factors. First, many of the prior studies included patients with substantial cumulative exposure to older ART agents, such as stavudine and zidovudine, which are associated with greater glucose intolerance as compared to newer regimens [2, 8]. Second, we postulate that among our obese subjects the effects of excess adiposity on glucose metabolism may have masked any impact from HIV infection or our selected ART regimen. Prior studies have shown efavirenz treatment is associated with higher blood glucose levels, and a difference in glucose tolerance may have been apparent if we had compared non-obese HIV-infected to non-obese uninfected persons [22, 23].
Lipid profiles and vascular health
We did not observe an adverse impact of obesity on plasma lipids, ICAM-1 or VCAM-1, or ultrasound measurements of carotid IMT or brachial FMD in the HIV-infected subjects. In contrast, a recent study of HIV-infected young women in the Adolescent Trials Network found dyslipidemia was more prevalent at higher BMI, though the median age of subjects was approximately 20 years younger than our cohort [24]. However, our results were in accordance with prior clinical studies showing BMI is not associated with carotid IMT or brachial artery FMD [25, 26]. Our finding that obesity does not adversely affect several clinical risk factors for cardiovascular events is in accordance with large epidemiology studies which found no increased risk of myocardial infarction among higher BMI patients [4, 13].
In contrast, we found treated HIV infection was associated with adverse changes in plasma lipid profiles and endothelial activation, but not carotid IMT or brachial artery FMD, in our obese subjects after adjusting for age and other covariates. HIV infection, despite effective virologic suppression, is an independent risk factor for myocardial infarction and other CVD events, which has been linked to adverse effects on lipid profiles, platelet activation, inflammation, and endothelial function [3, 4]. We observed higher total cholesterol, LDL, and triglycerides, and lower HDL, in the obese HIV-infected subjects compared to HIV-uninfected, but the relative contributions of efavirenz exposure, which is shown to cause more lipid elevations compared to other ART agents, versus HIV infection to these findings deserves further investigation [27–30]. Higher plasma levels of VCAM-1 and ICAM-1 were also associated with HIV infection, which has been reported in prior studies of HIV patients and can persist for years despite effective ART treatment [31–33]. The lack of an association between HIV status and carotid IMT may have been due to our considerably smaller sample size as compared to prior studies [25, 26, 34].
Biomarkers of systemic inflammation
Several markers of systemic inflammation were significantly associated with obesity, but among the obese patients, hsCRP, IL-6, and TNF-α receptor 1 levels did not differ by HIV status. This was unexpected, and we postulate that adipocyte-derived IL-6 and TNF-α masked additional innate immune system activation due to HIV infection. Circulating hsCRP, IL-6, and TNF-α receptor 1 increased with total fat mass, and higher plasma levels of these biomarkers in obese HIV patients have been previously reported by our group and others [35, 36]. We postulate this finding reflects constitutive cytokine production by hypertrophied adipocytes and adipose-resident immune cells (primarily macrophages and, to a lesser extent, lymphocytes) as reported from in vitro studies [37–40]. Increased adipose tissue mass is primarily due to adipocyte hypertrophy rather than hyperplasia, and interval increases in adipocyte size result in disproportionate increases in IL-6 and TNF-α expression [37–39]; it is estimated that adipose tissue-derived IL-6 constitutes up to 35% of circulating levels in obese individuals and serves as a major signaling pathway for CRP production [40].
Our findings raise the question of whether hsCRP and IL-6 levels can reliably predict adverse CVD outcomes and mortality as described in prior studies of predominantly non-obese populations [6, 41, 42]. The link between soluble inflammatory biomarkers and adverse health outcomes reported in prior studies likely reflects the production of these cytokines at sites of tissue inflammation and damage. However, it is unclear whether the same cytokines originating from adipocytes, particularly IL-6 and TNF-α, would contribute substantially to these end-organ pathogenic processes. Of note, sCD14, a soluble monocyte receptor not produced by adipocytes, was significantly higher in our subjects with HIV and did not appear to be affected by fat mass [43]. Higher circulating sCD14 is associated with mortality and disease progression in HIV patients, and may have more utility for predicting health outcomes among obese patients than adipocyte-derived cytokines [44, 45].
Strengths of this study included a uniform ART treatment regimen in all HIV-infected subjects, the exclusion of subjects on medications to treat metabolic or cardiovascular diseases (aside from antihypertensives), a minimum 2-year period of virologic suppression to allow the effects of plasma viremia to fade, a required minimum CD4+ T-cell count (>350 CD4+ cells/µl), and approximately equal distribution of subjects by sex and race (white and non-white). Obese control subjects were matched to HIV-infected subjects by age, race and BMI, and met the same criteria for excluded medications.
While the study sample size was relatively small, we observed several statistically significant findings between the HIV-infected groups and the obese groups which were relatively consistent within each category (e.g., glucose metabolism, lipids, and inflammatory cytokines). This suggests the lack of a detectable difference, when present, was not clearly due to inadequate power. The sample size was calculated to provide 90% power to detect an association between body composition and serum IL-6 levels (based on previously reported cytokine levels in a similar cohort) [35], and 80% power to detect a 37% difference in IL-6 levels between obese HIV-infected and –uninfected subjects. Second, the use of DEXA imaging provided less accurate quantification of visceral fat than CT or MRI. Third, the cross-sectional design prevented assessments of causality or variability of our endpoints over time. Fourth, the age of the HIV-uninfected controls was lower than obese HIV-infected subjects, which precluded direct comparisons of the groups and necessitated the use of multivariable models. Lastly, the regimen of efavirenz, tenofovir, and emtricitabine was selected due to the lower reported effects on lipid parameters and the widespread use if this regimen in the US and worldwide, but additional studies are needed before extrapolating our results to patients on a protease inhibitor or the increasingly common integrase inhibitors.
The health outcomes of obese HIV-infected individuals are increasingly relevant to clinical care as the prevalence of obesity in the HIV population approaches parity with the general population in many areas of the United States and Europe [14–18, 46]. Long-term weight loss maintenance is a major challenge for overweight and obese persons, and in HIV-infected individuals glucose and lipid abnormalities may persist despite short-term weight loss programs [47, 48]. While further long-term data are needed on the reversibility of obesity-associated cardiometabolic risk-factors after weight loss, our findings suggest that the prevention of excessive weight gain and obesity is critical to preventing insulin resistance in HIV patients, while CVD risk reduction remains important for all HIV-infected individuals irrespective of body composition.
Supplementary Material
Acknowledgements
The authors thank the participants in the Adiposity and Immune Activation Cohort study.
Funding support:
This work was supported by NIAID [grant numbers K23 100700 and K24 AI65298], the NIH-funded Vanderbilt Clinical and Translational Science award from NCRR/NIH [grant number UL1 RR024975-01], and the NIH-funded Tennessee Center for AIDS Research [grant number P30 AI110527]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Samaras K, Gan SK, Peake PW, Carr A, Campbell LV. Proinflammatory markers, insulin sensitivity, and cardiometabolic risk factors in treated HIV infection. Obesity (Silver Spring) 2009;17:53–59. doi: 10.1038/oby.2008.500. [DOI] [PubMed] [Google Scholar]
- 2.Brown TT, Cole SR, Li X, Kingsley LA, Palella FJ, Riddler SA, et al. Antiretroviral therapy and the prevalence and incidence of diabetes mellitus in the multicenter AIDS cohort study. Arch Intern Med. 2005;165:1179–1184. doi: 10.1001/archinte.165.10.1179. [DOI] [PubMed] [Google Scholar]
- 3.Womack JA, Chang CC, So-Armah KA, Alcorn C, Baker JV, Brown ST, et al. HIV infection and cardiovascular disease in women. J Am Heart Assoc. 2014;3:e001035. doi: 10.1161/JAHA.114.001035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Freiberg MS, Chang CC, Kuller LH, Skanderson M, Lowy E, Kraemer KL, et al. HIV infection and the risk of acute myocardial infarction. JAMA Intern Med. 2013;173:614–622. doi: 10.1001/jamainternmed.2013.3728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Butt AA, McGinnis K, Rodriguez-Barradas MC, Crystal S, Simberkoff M, Goetz MB, et al. HIV infection and the risk of diabetes mellitus. AIDS. 2009;23:1227–1234. doi: 10.1097/QAD.0b013e32832bd7af. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Duprez DA, Neuhaus J, Kuller LH, Tracy R, Belloso W, De Wit S, et al. Inflammation, coagulation and cardiovascular disease in HIV-infected individuals. PLoS One. 2012;7:e44454. doi: 10.1371/journal.pone.0044454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Brown TT, Tassiopoulos K, Bosch RJ, Shikuma C, McComsey GA. Association between systemic inflammation and incident diabetes in HIV-infected patients after initiation of antiretroviral therapy. Diabetes Care. 2010;33:2244–2249. doi: 10.2337/dc10-0633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.De Wit S, Sabin CA, Weber R, Worm SW, Reiss P, Cazanave C, et al. Incidence and risk factors for new-onset diabetes in HIV-infected patients: the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study. Diabetes Care. 2008;31:1224–1229. doi: 10.2337/dc07-2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sabin CA, Worm SW, Weber R, Reiss P, El-Sadr W, Dabis F, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration. Lancet. 2008;371:1417–1426. doi: 10.1016/S0140-6736(08)60423-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Helleberg M, Kronborg G, Ullum H, Ryder LP, Obel N, Gerstoft J. Course and Clinical Significance of CD8+ T-Cell Counts in a Large Cohort of HIV-Infected Individuals. J Infect Dis. 2015;211:1726–1734. doi: 10.1093/infdis/jiu669. [DOI] [PubMed] [Google Scholar]
- 11.Serrano-Villar S, Sainz T, Lee SA, Hunt PW, Sinclair E, Shacklett BL, et al. HIV-infected individuals with low CD4/CD8 ratio despite effective antiretroviral therapy exhibit altered T cell subsets, heightened CD8+ T cell activation, and increased risk of non-AIDS morbidity and mortality. PLoS Pathog. 2014;10:e1004078. doi: 10.1371/journal.ppat.1004078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Capeau J, Bouteloup V, Katlama C, Bastard JP, Guiyedi V, Salmon-Ceron D, et al. Ten-year diabetes incidence in 1046 HIV-infected patients started on a combination antiretroviral treatment. AIDS. 2012;26:303–314. doi: 10.1097/QAD.0b013e32834e8776. [DOI] [PubMed] [Google Scholar]
- 13.Friis-Moller N, Reiss P, Sabin CA, Weber R, Monforte A, El-Sadr W, et al. Class of antiretroviral drugs and the risk of myocardial infarction. N Engl J Med. 2007;356:1723–1735. doi: 10.1056/NEJMoa062744. [DOI] [PubMed] [Google Scholar]
- 14.Crum-Cianflone N, Roediger MP, Eberly L, Headd M, Marconi V, Ganesan A, et al. Increasing rates of obesity among HIV-infected persons during the HIV epidemic. PLoS One. 2010;5:e10106. doi: 10.1371/journal.pone.0010106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Amorosa V, Synnestvedt M, Gross R, Friedman H, MacGregor RR, Gudonis D, et al. A tale of 2 epidemics: the intersection between obesity and HIV infection in Philadelphia. J Acquir Immune Defic Syndr. 2005;39:557–561. [PubMed] [Google Scholar]
- 16.Tedaldi EM, Brooks JT, Weidle PJ, Richardson JT, Baker RK, Buchacz K, et al. Increased body mass index does not alter response to initial highly active antiretroviral therapy in HIV-1-infected patients. J Acquir Immune Defic Syndr. 2006;43:35–41. doi: 10.1097/01.qai.0000234084.11291.d4. [DOI] [PubMed] [Google Scholar]
- 17.Buchacz K, Baker RK, Palella FJ, Jr, Shaw L, Patel P, Lichtenstein KA, et al. Disparities in prevalence of key chronic diseases by gender and race/ethnicity among antiretroviral-treated HIV-infected adults in the US. Antivir Ther. 2013;18:65–75. doi: 10.3851/IMP2450. [DOI] [PubMed] [Google Scholar]
- 18.Koethe JR, Jenkins CA, Lau B, Shepherd BE, Justice AC, Tate JP, et al. Rising Obesity Prevalence and Weight Gain Among Adults Starting Antiretroviral Therapy in the United States and Canada. AIDS Res Hum Retroviruses. 2015 doi: 10.1089/aid.2015.0147. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Levy JC, Matthews DR, Hermans MP. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998;21:2191–2192. doi: 10.2337/diacare.21.12.2191. [DOI] [PubMed] [Google Scholar]
- 20.Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care. 2001;24:539–548. doi: 10.2337/diacare.24.3.539. [DOI] [PubMed] [Google Scholar]
- 21.Savitz DA, Olshan AF. Multiple comparisons and related issues in the interpretation of epidemiologic data. Am J Epidemiol. 1995;142:904–908. doi: 10.1093/oxfordjournals.aje.a117737. [DOI] [PubMed] [Google Scholar]
- 22.Erlandson KM, Kitch D, Tierney C, Sax PE, Daar ES, Melbourne KM, et al. Impact of randomized antiretroviral therapy initiation on glucose metabolism. AIDS. 2014;28:1451–1461. doi: 10.1097/QAD.0000000000000266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lennox JL, Dejesus E, Berger DS, Lazzarin A, Pollard RB, Ramalho Madruga JV, et al. Raltegravir versus Efavirenz regimens in treatment-naive HIV-1-infected patients: 96-week efficacy, durability, subgroup, safety, and metabolic analyses. J Acquir Immune Defic Syndr. 2010;55:39–48. doi: 10.1097/QAI.0b013e3181da1287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mulligan K, Harris DR, Monte D, Stoszek S, Emmanuel P, Hardin DS, et al. Obesity and dyslipidemia in behaviorally HIV-infected young women: Adolescent Trials Network study 021. Clin Infect Dis. 2010;50:106–114. doi: 10.1086/648728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Mangili A, Polak JF, Skinner SC, Gerrior J, Sheehan H, Harrington A, et al. HIV infection and progression of carotid and coronary atherosclerosis: the CARE study. J Acquir Immune Defic Syndr. 2011;58:148–153. doi: 10.1097/QAI.0b013e31822d4993. [DOI] [PubMed] [Google Scholar]
- 26.Hsue PY, Ordovas K, Lee T, Reddy G, Gotway M, Schnell A, et al. Carotid intima-media thickness among human immunodeficiency virus-infected patients without coronary calcium. Am J Cardiol. 2012;109:742–747. doi: 10.1016/j.amjcard.2011.10.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Quercia R, Roberts J, Martin-Carpenter L, Zala C. Comparative changes of lipid levels in treatment-naive, HIV-1-infected adults treated with dolutegravir vs efavirenz, raltegravir, and ritonavir-boosted darunavir-based regimens over 48 weeks. Clin Drug Investig. 2015;35:211–219. doi: 10.1007/s40261-014-0266-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Riddler SA, Smit E, Cole SR, Li R, Chmiel JS, Dobs A, et al. Impact of HIV infection and HAART on serum lipids in men. JAMA. 2003;289:2978–2982. doi: 10.1001/jama.289.22.2978. [DOI] [PubMed] [Google Scholar]
- 29.Rockstroh JK, Lennox JL, Dejesus E, Saag MS, Lazzarin A, Wan H, et al. Long-term treatment with raltegravir or efavirenz combined with tenofovir/emtricitabine for treatment-naive human immunodeficiency virus-1-infected patients: 156-week results from STARTMRK. Clin Infect Dis. 2011;53:807–816. doi: 10.1093/cid/cir510. [DOI] [PubMed] [Google Scholar]
- 30.Tebas P, Sension M, Arribas J, Duiculescu D, Florence E, Hung CC, et al. Lipid levels and changes in body fat distribution in treatment-naive, HIV-1-Infected adults treated with rilpivirine or Efavirenz for 96 weeks in the ECHO and THRIVE trials. Clin Infect Dis. 2014;59:425–434. doi: 10.1093/cid/ciu234. [DOI] [PubMed] [Google Scholar]
- 31.de Larranaga GF, Bocassi AR, Puga LM, Alonso BS, Benetucci JA. Endothelial markers and HIV infection in the era of highly active antiretroviral treatment. Thromb Res. 2003;110:93–98. doi: 10.1016/s0049-3848(03)00291-3. [DOI] [PubMed] [Google Scholar]
- 32.Calza L, Pocaterra D, Pavoni M, Colangeli V, Manfredi R, Verucchi G, et al. Plasma levels of VCAM-1, ICAM-1, E-Selectin, and P-Selectin in 99 HIV-positive patients versus 51 HIV-negative healthy controls. J Acquir Immune Defic Syndr. 2009;50:430–432. doi: 10.1097/QAI.0b013e31819a292c. [DOI] [PubMed] [Google Scholar]
- 33.Ronsholt FF, Ullum H, Katzenstein TL, Gerstoft J, Ostrowski SR. Persistent inflammation and endothelial activation in HIV-1 infected patients after 12 years of antiretroviral therapy. PLoS One. 2013;8:e65182. doi: 10.1371/journal.pone.0065182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hsue PY, Scherzer R, Hunt PW, Schnell A, Bolger AF, Kalapus SC, et al. Carotid Intima-Media Thickness Progression in HIV-Infected Adults Occurs Preferentially at the Carotid Bifurcation and Is Predicted by Inflammation. J Am Heart Assoc. 2012;1 doi: 10.1161/JAHA.111.000422. Epub April 12, 2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Koethe JR, Dee K, Bian A, Shintani A, Turner M, Bebawy S, et al. Circulating interleukin-6, soluble CD14, and other inflammation biomarker levels differ between obese and nonobese HIV-infected adults on antiretroviral therapy. AIDS Res Hum Retroviruses. 2013;29:1091–1025. doi: 10.1089/aid.2013.0016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Conley LJ, Bush TJ, Rupert AW, Sereti I, Patel P, Brooks JT, et al. Obesity is associated with greater inflammation and monocyte activation among HIV-infected adults receiving antiretroviral therapy. AIDS. 2015 doi: 10.1097/QAD.0000000000000817. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 37.Bastard JP, Lagathu C, Caron M, Capeau J. Point-counterpoint: Interleukin-6 does/does not have a beneficial role in insulin sensitivity and glucose homeostasis. J Appl Physiol. 2007;102:821–822. doi: 10.1152/japplphysiol.01353.2006. [DOI] [PubMed] [Google Scholar]
- 38.Skurk T, Alberti-Huber C, Herder C, Hauner H. Relationship between adipocyte size and adipokine expression and secretion. J Clin Endocrinol Metab. 2007;92:1023–1033. doi: 10.1210/jc.2006-1055. [DOI] [PubMed] [Google Scholar]
- 39.Dandona P, Weinstock R, Thusu K, Abdel-Rahman E, Aljada A, Wadden T. Tumor necrosis factor-alpha in sera of obese patients: fall with weight loss. J Clin Endocrinol Metab. 1998;83:2907–2910. doi: 10.1210/jcem.83.8.5026. [DOI] [PubMed] [Google Scholar]
- 40.Mohamed-Ali V, Goodrick S, Rawesh A, Katz DR, Miles JM, Yudkin JS, et al. Subcutaneous adipose tissue releases interleukin-6, but not tumor necrosis factor-alpha, in vivo. J Clin Endocrinol Metab. 1997;82:4196–4200. doi: 10.1210/jcem.82.12.4450. [DOI] [PubMed] [Google Scholar]
- 41.Kuller LH, Tracy R, Belloso W, De Wit S, Drummond F, Lane HC, et al. Inflammatory and coagulation biomarkers and mortality in patients with HIV infection. PLoS Med. 2008;5:e203. doi: 10.1371/journal.pmed.0050203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Nordell AD, McKenna M, Borges AH, Duprez D, Neuhaus J, Neaton JD, et al. Severity of cardiovascular disease outcomes among patients with HIV is related to markers of inflammation and coagulation. J Am Heart Assoc. 2014;3:e000844. doi: 10.1161/JAHA.114.000844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Shive CL, Jiang W, Anthony DD, Lederman MM. Soluble CD14 is a nonspecific marker of monocyte activation. AIDS. 2015;29:1263–1265. doi: 10.1097/QAD.0000000000000735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Krastinova E, Lecuroux C, Leroy C, Seng R, Cabie A, Rami A, et al. High Soluble CD14 Levels at Primary HIV-1 Infection Predict More Rapid Disease Progression. J Infect Dis. 2015 doi: 10.1093/infdis/jiv145. Epub March 6, 2015. [DOI] [PubMed] [Google Scholar]
- 45.Sandler NG, Wand H, Roque A, Law M, Nason MC, Nixon DE, et al. Plasma Levels of Soluble CD14 Independently Predict Mortality in HIV Infection. J Infect Dis. 2011;203:780–790. doi: 10.1093/infdis/jiq118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Hasse B, Iff M, Ledergerber B, Calmy A, Schmid P, Hauser C, et al. Obesity Trends and Body Mass Index Changes After Starting Antiretroviral Treatment: The Swiss HIV Cohort Study. Open Forum Infect Dis. 2014;1 doi: 10.1093/ofid/ofu040. ofu040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Wing RR, Phelan S. Long-term weight loss maintenance. Am J Clin Nutr. 2005;82:222S–225S. doi: 10.1093/ajcn/82.1.222S. [DOI] [PubMed] [Google Scholar]
- 48.Engelson ES, Agin D, Kenya S, Werber-Zion G, Luty B, Albu JB, et al. Body composition and metabolic effects of a diet and exercise weight loss regimen on obese, HIV-infected women. Metabolism. 2006;55:1327–1336. doi: 10.1016/j.metabol.2006.05.018. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.