Abstract
Background
Persons with human immunodeficiency virus (HIV) on long-term antiretroviral therapy (ART) have exaggerated accumulation of ectopic fat and cardiometabolic disease risk.
Objectives
The objective of this study was to characterize longitudinal changes in habitual diet and macronutrient oxidation in relation to changes in abdominal adipose tissue and ectopic fat sites using magnetic resonance and computed tomography imaging.
Methods
A prospective cohort of 49 males with HIV had comprehensive assessments at baseline, 52 wk (year 1), and 104 wk (year 2). Linear mixed effects models accounted for the correlation structure in the data and estimated effects over time. Fixed effects included baseline value of the outcome, age, body mass index [BMI (in kg/m2)], time since HIV diagnosis, and duration of ART.
Results
The consumption of simple carbohydrates decreased over time (total and added sugars, P = 0.03) concomitant with increased ratio of fatty acid to carbohydrate oxidation (P = 0.01). The amount of abdominal subcutaneous adipose tissue decreased 15% (P < 0.0001), and the amount of visceral adipose tissue (VAT) decreased 7% (P = 0.009) simultaneously with improved subcutaneous adipose tissue and VAT densities. Improvement in VAT density is associated with reduced fat fraction of the liver (r = –0.28, P = 0.04) and thighs (r = –0.41, P = 0.03), indicating overall reduced ectopic fat accumulation. However, pancreas density decreased (P = 0.03), and no statistically significant change was observed in skeletal muscle density (P = 0.16), suggesting tissue-specific impacts.
Conclusions
These findings support the relationship between dietary carbohydrate intake, fat oxidation rate, and fat mobilization to reduce ectopic lipid deposition in persons with HIV. Although modest changes in dietary intakes show potential for improving metabolic flexibility and body composition among individuals on long-term ART, some organs and tissues may not respond in tandem with other depots of ectopic fat.
Keywords: body composition, ectopic fat, obesity, HIV, imaging, carbohydrate
Introduction
More than half of the persons with HIV (PWH) are aged >50 y in the United States. A serious clinical concern is that as PWH age, they may become more susceptible to multiple comorbid conditions due to the combined effects of persistent HIV reservoirs, long-term exposure to antiretroviral therapy (ART), and higher rates of overweight and obesity. Notably, 40%–50% of PWH now have a BMI (in kg/m2) indicative of being overweight or obese [1,2]. Indeed, in a multisite study of >14,000 PWH, 22% transitioned from normal to overweight BMI category, and another 18% from the overweight to obese BMI category, within 3 y of ART initiation [2]. Although longitudinal studies show that the normal aging-associated changes in body morphometrics comprise a gain in body fat concomitant with a reduction in lean mass, PWH tend to gain more body fat and lose more lean tissue over time than individuals without HIV [3].
In this state of high adiposity, the plasticity of adipocytes in subcutaneous adipose tissue (SAT) becomes dysregulated, leading to reduced capacity to store excess energy in the form of triglycerides. Consequently, the gain in body fat is not distributed uniformly but is characterized by a disproportionate expansion of intra-abdominal or visceral adipose tissue (VAT) compared with SAT [4]. Further, alterations in adipocyte lipid handling promote the accumulation of fat in intra-abdominal organs such as the liver. The influx of free fatty acids from SAT and VAT to the liver is likely a consequence of accelerated lipolysis and intrahepatic re-esterification in the state of HIV [5].
However, abnormal ectopic fat deposition is not limited to the liver, but also occurs in other sites not typically associated with excess fat storage, such as the heart, kidneys, pancreas, and skeletal muscle. The exposure to greater amounts of circulating free fatty acids, as well as dietary fats, contributes to altered metabolic functions of these organs and tissues by inducing lipotoxicity and insulin resistance. Notably, greater adiposity, especially VAT, is a stronger predictor of insulin resistance than aging per se [6].
In the current environment, public perception and opinions on the health promotion and chronic disease benefits of dietary macronutrients favor limiting the consumption of dietary carbohydrates. Annual surveys conducted by the International Food Information Council (IFIC) Foundation show significant increases in consumers’ beliefs that carbohydrates, particularly sugars, are the primary driver of weight gain and obesity [7]. Consumer behavior may be further influenced by additions to the nutrition facts label, which now provides specific information on added sugar content. Notably, NHANES data shows that although total carbohydrate intake as a percentage of energy has not changed, the trend for intake of carbohydrates has shifted to more whole grains and less sugars [8]. Such trends in food consumption may also be influenced by changes in the food supply and purchasing capacity.
Although prior evidence demonstrates significant changes in body weight and overall body composition after initiating modern ART [9], the aim of this study was to quantify and characterize longitudinal changes in metabolically critical ectopic fat depots using cross-sectional magnetic resonance (MR) and computed tomography (CT) imaging in aging males on long-term ART. Of relevance is that the study was initiated immediately after the acute period of the SARS-CoV-19 pandemic, a period of substantial change in diet and lifestyle habits for many individuals [10,11]. Thus, we hypothesized that quantitative (amount) and qualitative (density) changes in ectopic fat deposition were associated with alterations in dietary intakes and/or macronutrient oxidation rates that would positively impact fat mobilization.
Methods
Study design and eligibility
A prospective cohort of males with HIV on long-term ART was recruited from the Infectious Diseases Clinic at the Department of Veterans Affairs Tennessee Valley Healthcare System (TVHS) during the period of December 2020 through December 2022. Potential participants were enrolled if they were on integrase strand transfer inhibitor (INSTI)-based ART ≥6 mo with viral suppression (HIV-1 RNA <50 copies/mL) and CD4+ count >350 cells/μL. Age 35–75 y and BMI ≥22 were additional criteria to foster inclusion of veterans with treated HIV at high risk of excess ectopic fat accumulation. Exclusion criteria included contraindication to MR imaging, cirrhosis, active hepatitis B/hepatitis C, inflammatory or rheumatologic disease, type 1 diabetes or type 2 diabetes, usage of tesamorelin, oral corticosteroids, hormone replacement therapy, dipeptidyl peptidase-4 inhibitors or glucagon-like peptide-1 receptor agonists, indication of illicit drug use (cocaine, methamphetamines, and unprescribed opiates) and excess alcohol consumption (>2 drinks/d). The study was conducted according to the principles of the Declaration of Helsinki, approved by the TVHS and Vanderbilt University Medical Center institutional review boards, and all participants provided written informed consent.
Study visits
Enrolled participants had 3 metabolic testing visits at the Vanderbilt Clinical Research Center: baseline, at 52 wk (year 1), and at 104 wk (year 2). At each visit, fasting blood samples were collected, anthropometrics were measured, CT and MR imaging were performed, energy expenditure and macronutrient oxidation were quantified, and diet assessments were administered (Figure 1).
FIGURE 1.
Study visit schema.
Clinical biomarkers
Serum concentrations of glucose, insulin, and c-peptide were collected in the 10–12 h fasted state and assayed at the Vanderbilt University Medical Center Hormone Assay & Analytical Services Core Laboratory. HOMA-IR score was calculated as: (fasting insulin mU/L × fasting glucose mg/dL)/405. Serum lipid profiles (total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides), hemoglobin A1c, C-reactive protein, CD4 and CD8 counts, and HIV-1 RNA were assayed at the TVHS Pathology Laboratory.
Anthropometrics and CT imaging
Height (centimeters), weight (kilograms), waist (centimeters at iliac crest), hip (centimeters), and mid-thigh (centimeters) circumferences were recorded as an average of 3 measurements. Noncontrast scans of the abdomen/pelvis were acquired on a Phillips multislice CT scanner. CT scans were analyzed for quantification of abdominal SAT, VAT, and skeletal muscle areas and densities using 1 slice landmarked at the axial cross-section of the third lumbar vertebra (L3). After conversion to Digital Imaging and Communications in Medicine (DICOM) format, images were segmented using an automated version of Slice-O-Matic software (version 4.3, TomoVision) [12]. Manual editing of tissue boundaries after automated segmentation was performed by trained technicians of the Vanderbilt Diet, Body Composition, and Human Metabolism Core to assure thorough separation and quantification of all tissue depots. Quantification of cross-sectional areas and densities for each tissue depot was based on established radio-attenuation thresholds in Hounsfield units (HUs) (SAT: –190 to –30 HU, VAT: –50 to –150 HU, skeletal muscle: –29 to +150 HU). Mean attenuation (HU) for total abdominal skeletal muscles (psoas, quadratus lumborum, erector spinae, lateral oblique, internal/external oblique, and rectus abdominis) was used to determine average abdominal skeletal muscle density (SMD), an indicator of ectopic lipid accumulation. Liver and pancreas density were quantified by a clinical radiologist (for liver density, 9 homogenous areas between the twelfth thoracic and the first lumbar vertebra were manually segmented and averaged; for pancreas density, 3 homogeneous areas within each of the pancreatic head, body, and tail regions were manually segmented and averaged).
MR imaging
Participants were scanned in a supine position on a whole body Philips 3T Ingenia CX system (Philips Healthcare) using the integrated quadrature body coil for transmit and receive. Two sets of 3D multi-gradient echo sequences were obtained, 1 through the abdomen and 1 through the thighs. The abdominal sequence acquired 6 echoes with a repetition time (TR)/first echo time (TE1)/Δ TE) = 5.7/1.01/0.73 ms and a flip angle of 3 degrees. It obtained 77 slices with a 3D field of view of 384 × 384 × 231 mm, an acquired matrix of 156 × 154 × 77, a reconstructed matrix of 192 × 192 × 77, and a slice thickness of 6 mm with 3 mm spacing. Reconstructed voxel size was 2 × 2 × 6 mm. Water-only images, fat-only images, and proton density fat fraction (PDFF) maps were reconstructed in DICOM format inline by the scanner computer using a custom algorithm that measured and corrected for exponential T2∗ signal decay while incorporating a multipeak spectral model. These images were analyzed offline using a custom 3D U-net convolutional neural network, which automatically segmented the liver boundaries on the water images. A trained image analyst, who was blinded to the PDFF maps and all clinical data, manually corrected each segmentation under the supervision of the study clinical radiologist. The corrected segmentations were propagated to the PDFF maps, and the mean liver PDFF was recorded. The thigh sequence acquired 6 echoes with a flip angle of 3 degrees. It obtained 25 slices with a 3D field of view of 220 × 220 × 200 mm, an acquired matrix of 148 × 148 × 25, a reconstructed matrix of 224 × 224 × 25, and a slice thickness of 8 mm with 8 mm spacing. Reconstructed voxel size was 0.982 × 0.982 × 8 mm. Water-only and fat-only images in DICOM format were reconstructed inline by the scanner computer using a Philips commercial algorithm and then analyzed as previously described [13].
Energy expenditure and macronutrient oxidation
Participant oral and written instructions for the prior 48 h were to avoid alcohol, rigorous physical activity and excess caffeine intake, to collect all urine output for 24 h prior to arrival at the Vanderbilt Clinical Research Center, and to consume a balanced mixed meal comprised of 35% fat, 47% carbohydrate, and 18% protein kilocalories for dinner on the night prior to resting energy expenditure (REE) and substrate oxidation rate measurements. Testing occurred early in the morning (07:00–08:00) in the 10–12 h fasted state. Measurements were obtained in standard thermoneutral conditions using a portable integrated metabolic cart system (ParvoMedics TrueOne 2400). The metabolic cart was calibrated to room air and a single gas tank prior to each use. The participant rested supine for 5 min to acclimate to the ventilated hood. Data were then collected for 20 min at steady state with average change in minute VO2 ≤10% and average change in respiratory quotient ≤5%. REE was calculated via the Weir equation [14], and substrate oxidation rates were calculated after adjustment for protein oxidation using 24-h urinary urea nitrogen output following the method of Frayn [15].
Diet assessment
Dietary intakes were assessed by trained research dietitians at the Vanderbilt Diet, Body Composition, and Human Metabolism Core using a standardized protocol predicated on the USDA 24-h multipass interview methodology, which has been validated in adults with and without obesity [16,17]. Various sizes of measuring utensils (plates, bowls, cups, serving spoons, and other cutlery) were utilized to obtain a reliable estimation of portion sizes of foods and beverages consumed. Dietary data were directly entered into Nutrition Data System for Research software version 2020 (University of Minnesota Nutrition Coordinating Center), which provides interviewer prompts designed to capture brand names, food preparation methods, and potentially missing ingredients and to reduce recall bias. Food composition is sourced from a USDA food and nutrition database that contains ∼19,500 foods and provides values for 178 food components, including energy, macronutrients, and micronutrients.
Statistical analysis
A priori sample size analysis showed 80%–90% probability to detect a significant difference over time in VAT, SAT, pancreas, liver, or SMD of ≥5 HU with 45–50 participants. Data normality was determined by visual inspection of histograms. Measures were mostly symmetrically distributed, with log transformation performed due to deviation from normality for triglycerides, triglycerides/HDL ratio, and HOMA-IR score. Descriptive statistics were calculated using medians with 25th and 75th quartiles for continuous variables and counts and frequencies for categorical variables. Spearman’s correlation coefficients were generated for univariate analyses. Multivariate analyses were performed using linear mixed effects modeling to account for the correlation structure in the data and estimate effects over time. The fixed effects included the baseline value of the outcome variable, baseline age, baseline BMI, time since HIV diagnosis, and duration of INSTI therapy. The random effect was the study participant, which accounted for repeated measures for each participant. There was no missing data to adjust for. Statistical analyses were performed using SPSS version 29.0 (IBM) and R version 4.4.1 (R Foundation for Statistical Computing), with P values <0.05 considered statistically significant.
Results
Study participants
At enrollment, the 49 male participants had been diagnosed with HIV for 18.5 ± 9.3 y and were taking prescribed INSTI-based medications for the prior 4.4 ± 3.3 y. Almost half (46%) of participants self-identified as Black and 54% as White. Participant median age at baseline was 56 y, and median BMI was 28.7, with 12% meeting BMI criteria for normal weight, 49% for overweight, and 39% for obese. On average, participants were normoglycemic, normotensive, and had within normal range concentrations of total, LDL cholesterol, and HDL cholesterol (Table 1). However, other clinical biomarkers indicated moderately high states of hyperinsulinemia (median insulin concentration of 19.5 mIU/mL) and systemic inflammation (median C-reactive protein concentration of 3.5 mg/dL) at baseline.
TABLE 1.
Longitudinal changes in clinical biomarkers in males with treated HIV.
| Baseline | 52 wk (year 1) | 104 wk (year 2) | Unadj P value1 | Adj P value1 | |
|---|---|---|---|---|---|
| Age (y) | 56 (51, 63) | 57 (52, 64) | 58 (53, 65) | 0.03 | 0.03 |
| CD4 Count (cells/mm3) | 672 (478, 897) | 778 (582, 995) | 742 (598, 1010) | 0.03 | 0.04 |
| CD8 Count (cells/mm3) | 867 (593, 1164) | 941 (668, 1229) | 880 (618, 1045) | 0.14 | 0.35 |
| CD4/CD8 (ratio) | 0.86 (0.46, 1.21) | 0.92 (0.60, 1.19) | 0.96 (0.64, 1.14) | 0.09 | 0.10 |
| Systolic blood pressure (mmHg) | 130 (122, 140) | 132 (121, 140) | 135 (120, 143) | 0.40 | 0.56 |
| Diastolic blood pressure (mmHg) | 85 (79, 92) | 89 (80, 93) | 86 (76, 95) | 0.48 | 0.31 |
| Total cholesterol (mg/dL) | 167 (144, 196) | 173 (155, 193) | 156 (143, 176) | 0.02 | 0.11 |
| LDL cholesterol (mg/dL) | 108 (83, 133) | 108 (84, 125) | 102 (83, 115) | 0.12 | 0.63 |
| HDL cholesterol (mg/dL) | 38 (32, 47) | 39 (34, 48) | 38 (34, 44) | 0.91 | 0.64 |
| TG (mg/dL) | 123 (87, 162) | 101 (73, 175) | 112 (77, 142) | 0.05 | 0.004 |
| TG:HDL (ratio) | 3.30 (1.93, 5.37) | 2.65 (1.57, 5.65) | 3.04 (1.72, 4.30) | 0.16 | 0.007 |
| C-reactive protein (mg/dL) | 3.5 (2.0, 5.9) | 2.7 (1.4, 4.6) | 2.5 (1.0, 4.0) | 0.25 | 0.02 |
| Glucose (mg/dL) | 103 (95, 113) | 96 (92, 104) | 102 (98, 107) | 0.51 | 0.07 |
| HbA1c (%) | 5.6 (5.2, 5.9) | 5.5 (5.3, 5.7) | 5.5 (5.3, 5.8) | 0.31 | 0.62 |
| Insulin (mIU/mL) | 19.5 (12.4, 38.9) | 13.6 (8.7, 22.8) | 12.7 (8.1, 23.3) | 0.14 | 0.17 |
| HOMA-IR (score) | 4.93 (2.93, 10.81) | 4.62 (2.88, 11.53) | 3.53 (2.31, 7.42) | 0.11 | 0.19 |
Abbreviations: Adj P value, adjusted P value; BMI, body mass index; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; HIV, human immunodeficiency virus; HOMA-IR, homeostatic model for insulin resistance; INSTI, integrase strand transfer inhibitor; LDL, low-density lipoprotein; TG, triglyceride; Unadj P value, unadjusted P value; CD4.
Data are presented as median (25th, 75th percentiles). Data were analyzed using linear mixed effects models for outcome at end of year 2 (104 wk) adjusted for baseline value of outcome variable, baseline age, baseline BMI, time since HIV diagnosis, and duration of INSTI treatment.
Changes over time in anthropometrics and ectopic fat sites
Both body weight and BMI remained stable over time (Table 2). Multivariate analysis accounting for covariates showed a significant reduction in waist circumference along with an average reduction in abdominal SAT area of 15% (P < 0.0001). The reduction in abdominal SAT area occurred concomitantly with an average reduction in VAT area of 7% (P = 0.009). Notably, the CT-determined density of abdominal SAT and VAT increased (P < 0.001; P < 0.0001, respectively), indicating reduced ectopic lipid content over time. The increase in VAT density was associated with reduced HOMA-IR score (r = –0.52, P = 0.03).
TABLE 2.
Longitudinal changes in anthropometrics and ectopic fat depots in males with treated HIV.
| Baseline | 52 wk (year 1) | 104 wk (year 2) | Unadj P value∗ | Adj P value1 | |
|---|---|---|---|---|---|
| Height (cm) | 176 (172, 183) | 176 (173, 184) | 176 (172, 184) | 0.98 | 0.98 |
| Weight (kg) | 90.7 (82.2, 106.6) | 90.5 (83.7, 102.6) | 89.8 (82.6, 110.9) | 0.62 | 0.40 |
| BMI (kg/m2) | 28.7 (26.6, 32.1) | 28.6 (25.6, 31.0) | 29.6 (25.7, 34.3) | 0.42 | 0.45 |
| Waist circumference (cm) | 107 (98, 115) | 107 (97, 112) | 104 (95, 118) | 0.15 | 0.04 |
| Hip circumference (cm) | 109 (104, 111) | 110 (103, 111) | 103 (102, 116) | 0.29 | 0.05 |
| Waist:hip (ratio) | 0.99 (0.95, 1.04) | 0.97 (0.94, 1.01) | 0.96 (0.93, 1.02) | 0.27 | 0.004 |
| Thigh area (cm) | 50.83 (46.17, 57.33) | 49.45 (46.24, 56.21) | 49.25 (45.17, 58.16) | 0.02 | 0.001 |
| Thigh fat fraction (%) | 4.58 (2.66, 6.64) | 4.52 (2.22, 6.28) | 3.52 (1.86, 6.57) | 0.007 | <0.001 |
| SAT area (cm2) | 226.17 (158.58, 324.32) | 214.46 (153.49, 278.33) | 202.27 (117.39, 271.29) | 0.002 | <0.0001 |
| SAT density (HU) | –104.85 (–108.41, –102.22) | –105.19 (–108.11, –100.39) | –103.81 (–106.51, –96.99) | 0.01 | <0.001 |
| VAT area (cm2) | 245.58 (161.80, 277.69) | 218.84 (140.19, 296.69) | 209.48 (125.78, 289.26) | 0.35 | 0.009 |
| VAT density (HU) | –100.38 (–103.40, –96.57) | –99.64(–101.21, –97.63) | –96.82 (–98.89, –93.79) | 0.005 | <0.0001 |
| VAT area:SAT area (ratio) | 0.93 (0.73, 1.44) | 0.96 (0.78, 1.45) | 0.97 (0.53, 1.64) | 0.09 | 0.52 |
| Liver fat fraction (%) | 2.47 (0.00, 25.4) | 2.44 (0.10, 19.8) | 2.08 (0.10, 11.5) | 0.05 | 0.40 |
| Liver density (HU) | 55.67 (52.05, 59.94) | 55.76 (52.26, 60.67) | 56.72 (53.28, 60.46) | 0.28 | 0.16 |
| Pancreas density (HU) | 38.61 (30.24, 42.60) | 35.57 (24.02, 40.80) | 34.22 (28.28, 38.61) | 0.02 | 0.03 |
| Skeletal muscle area (cm2) | 193.13 (175.27, 210.79) | 191.87 (171.76, 212.56) | 201.93 (173.23, 228.04) | 0.37 | 0.40 |
| Skeletal muscle density (HU) | 38.84 (34.49, 42.14) | 37.12 (34.65, 40.77) | 38.59 (35.52, 42.16) | 0.57 | 0.16 |
Abbreviations: Adj P value, adjusted P value; BMI, body mass index; HIV, human immunodeficiency virus; HU, Hounsfield units; INSTI, integrase strand transfer inhibitor; SAT, subcutaneous adipose tissue; Unadj P value, unadjusted P value; VAT, visceral adipose tissue.
Data are presented as median (25th, 75th percentiles). Data analyzed using linear mixed effects models for outcome at end of year 2 (104 wk) adjusted for baseline value of outcome variable, baseline age, baseline BMI, time since HIV diagnosis, and duration of INSTI treatment.
Further, there was an average relative reduction in thigh muscle fat fraction of 27% (P < 0.001), and the decrease in thigh muscle fat was significantly associated with the reductions in abdominal SAT and VAT areas (r = 0.48, P = 0.04; r = 0.56, P = 0.01; respectively). Despite significant relationships between thigh and abdominal skeletal muscle areas as well as between abdominal SAT density, VAT density, and SMD (Figure 2), the mean increase in abdominal skeletal muscle area by 2% was not statistically significant, and no change was observed over time in the density of abdominal skeletal muscle. Although not statistically significant, average liver density improved by 3%. In contrast, pancreas density worsened (P = 0.03).
FIGURE 2.
Relationships between changes in size and density of ectopic fat depots. SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.
Changes in energy homeostasis and macronutrient oxidation
No significant change was detected in daily REE (P = 0.25, Table 3). Although the increase in total physical activity score was also not statistically significant, the average increase of 19% was significantly associated with the reduction observed in thigh fat fraction (r = –0.35, P = 0.04). Interestingly, there appeared to be a spontaneous change in the habitual diet – a significant increase in the intake of dietary fats (P < 0.001) occurred concomitantly with a significant reduction in the intake of simple carbohydrates (average total sugars intakes reduced from 120.9 ± 85.0 to 89.6 ± 70.9 g/d and average added sugars intakes reduced from 90.2 ± 77.5 to 59.7 ± 62.4 g/d). The reductions in total and added sugars were associated with the significant improvement observed in daily dietary energy density (calories per gram of food consumed, P = 0.02, Figure 3).
TABLE 3.
Longitudinal changes in energy homeostasis in males with treated HIV.
| Baseline | 52 wk (year 1) | 104 wk (year 2) | Unadj P value∗ | Adj P value1 | |
|---|---|---|---|---|---|
| Energy intake (kcal) | 2136.2 (1580.3, 2926.9) | 1991.9 (1530.5, 2598.9) | 1844.7 (1430.3, 2192.6) | 0.24 | 0.05 |
| Energy density (kcal/gm food) | 0.81 (0.57, 1.09) | 0.72 (0.62, 1.01) | 0.68 (0.57, 0.87) | 0.05 | 0.02 |
| Dietary fat (% kcal) | 39.8 (31.7, 43.7) | 42.6 (33.8, 49.7) | 44.8 (36.3, 47.5) | 0.68 | <0.001 |
| Dietary carbohydrate (% kcal) | 43.2 (38.9, 49.7) | 40.4 (29.7, 47.6) | 41.6 (29.9, 47.9) | 0.44 | 0.12 |
| Dietary protein (% kcal) | 14.8 (10.5, 22.4) | 17.4 (13.7, 22.2) | 15.1 (12.5, 18.4) | 0.51 | 0.70 |
| Resting energy expenditure (kcal) | 1869.0 (1632.0, 2054.0) | 1891.5 (1657.0, 2160.8) | 1942.0 (1738.3, 22.92.3) | 0.39 | 0.25 |
| Fat oxidation (% REE kcal) | 44.4 (31.5, 60.3) | 53.4 (39.3, 61.0) | 50.6 (45.6, 63.8) | 0.05 | 0.02 |
| Carbohydrate oxidation (% REE kcal) | 34.8 (25.5, 50.1) | 29.8 (21.3, 44.7) | 30.1 (25.3, 41.8) | 0.08 | 0.06 |
| Protein oxidation (% REE kcal) | 14.8 (10.5, 22.4) | 14.7 (9.3, 21.6) | 13.5 (8.4, 23.8) | 0.36 | 0.44 |
| Respiratory quotient (VCO2/VO2) | 0.83 (0.83, 0.83) | 0.81 (0.82, 0.82) | 0.81 (0.81, 0.79) | 0.10 | 0.05 |
| Total physical activity score | 2.99 (1.50, 4.73) | 2.80 (1.58, 6.95) | 3.21 (1.63, 7.05) | 0.05 | 0.58 |
Abbreviations: Adj P value, adjusted P value; BMI, body mass index; HIV, human immunodeficiency virus; INSTI, integrase strand transfer inhibitor; REE, resting energy expenditure; Unadj P value, unadjusted P value; VCO2, volume of carbon dioxide exhaled; VO2, volume of oxygen inhaled.
Data are presented as median (25th, 75th percentiles). Data analyzed using linear mixed effects models for outcome at end of year 2 (104 wk) adjusted for baseline value of outcome variable, baseline age, baseline BMI, time since HIV diagnosis, and duration of INSTI treatment.
FIGURE 3.
Relationships between changes in energy balance factors and density of ectopic fat depots. HOMA-IR, homeostasis model assessment of insulin resistance; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.
The reductions in sugars are also associated with the significant increase in fatty acid oxidation, which showed an average increase of 8% (P = 0.02) whereas carbohydrate oxidation decreased on average by 6% (P = 0.05). In addition to the role of increased dietary fat and reduced dietary simple carbohydrate consumption, the improvement in fatty acid oxidation was consistent with the reduction in respiratory quotient over time. Not only were the reductions in total and added sugars intake significantly associated with the improvements in dietary energy density and fatty acid oxidation (Figure 3), but also with increased SAT density (total sugars: r = –0.43, P = 0.03; added sugars: r = –0.47, P = 0.02).
Discussion
We report here changes in anthropometrics and body morphometrics that occurred in the 2-y period from December 2020 through December 2022 in a cohort of males with HIV on contemporary INSTI-based ART. In contrast to the pathological changes in body composition favoring increased fat with reduced lean mass along with metabolic perturbations in lipid metabolism and insulin action that have been reported after the initiation of modern ART, we found significant reductions in the amounts of abdominal subcutaneous and VATs. Additionally, the densities of SAT and VAT improved along with decreased ectopic fat in the liver and thigh skeletal muscle. Despite no study-related or “lifestyle” intervention during this period, we also detected a significant reduction in average daily simple carbohydrate (total and added sugars) consumption. Previously, we reported similar spontaneous improvements in dietary intakes in adults without HIV during the same post-acute SARS-CoV-19 period [18]. Although it is possible that participants changed their behaviors due to greater awareness from interactions with the research team over the course of the study (i.e., the “Hawthorne effect”), we postulate these changes more likely reflect the influence of current scientific and public consensus regarding the detrimental role of excess consumption of dietary sugars on human health [19]. Guidelines and scientific statements from national and international health organizations now reflect the body of scientific evidence linking excess consumption of total and added sugars to a variety of cardiometabolic conditions, including obesity, diabetes, and coronary artery disease [[20], [21], [22], [23]]. Consequently, added sugar intake has declined in subgroups of the population [24]. It is also plausible that persons with a chronic disease such as HIV may have embraced healthier behaviors that might strengthen their immune system and improve overall health status to decrease susceptibility to COVID-19 infection.
Notably, the changes observed in dietary carbohydrate intakes and ectopic fat in multiple sites occurred simultaneously with an alteration in the oxidation rates of fats and carbohydrates as measured by indirect calorimetry conducted in the fasted state. In contrast to the present finding of increased fat oxidation, prior studies in PWH who have lipodystrophy show impaired fat oxidation when compared with healthy controls [[25], [26], [27], [28]]. Impaired fat oxidation in PWH has been associated with ectopic fat accumulation in the liver and skeletal muscle [25]. This may be explained, in part, by the presence of hyperinsulinemia in PWH, which can impair the metabolic flexibility that occurs in a normal physiological state. One characteristic of metabolic inflexibility is a lack of increased fat oxidation in fasting conditions.
In the postprandial state, dietary fat is preferentially stored in adipose tissue, and the partitioning of dietary fat between storage and oxidation is modulated by obesity, where preference for storage is greater. Thus, in the present cohort, where 88% of participants were overweight or obese based on BMI criteria, the increase in fat oxidation rate over time may be a function of the reductions in subcutaneous, visceral, liver, and thigh fat. Although not statistically significant, it is also plausible that the average increase in total daily physical activity, especially if participants were performing at least moderate-level physical activities, elicited increased energy demands that were met by increased fat compared with carbohydrate oxidation, as reflected by the reduction in the fat fraction of the thigh. At the same time, the reduction in total and added sugar intakes would contribute to reduced carbohydrate oxidation.
The accumulation of fat in the pancreas, an organ crucial for glucose metabolism and insulin secretion, is associated with an increased risk of metabolic syndrome, prediabetes, and diabetes [29]. We observed a worsening of pancreas density, indicating greater accumulation of pancreatic ectopic fat over time, which suggests that this organ may not respond in tandem with other ectopic depots. A growing body of evidence shows a relationship between fat accumulation in the pancreas and the liver, and some evidence suggests fatty pancreas is independently associated with metabolic dysfunction-associated steatotic liver disease and a driver of metabolic dysfunction-associated steatotic liver disease progression. However, other data show no relationship between reductions in liver and pancreatic fat in the setting of weight loss [30]. Consistent with weight loss data, the increase in ectopic fat in the pancreas in the present cohort occurred in the context of a 2.7% reduction in liver fat fraction (although not a statistically significant decrease). It is noteworthy that 25% of the cohort had hepatic steatosis at baseline based on having liver PDFF ≥5%, the clinical threshold used to identify hepatic steatosis [31].
Further supporting the concept of a tissue-specific nature of ectopic fat mobilization, we observed no change in SMD. Preserved intermuscular adipose tissue in the setting of SAT loss has been observed in prior studies [32]. Of concern is that PWH are at increased risk of sarcopenia (a state of reduced skeletal muscle mass, strength, and physical function) at an earlier age than HIV-negative persons [33]. Long-term use of ART and higher BMI appear to be risk factors. We previously showed that both myosteatotic-type (i.e., accumulation of intermuscular and intramyocellular fat) and sarcopenic-type obesity are prevalent in PWH [34]. It is now understood that myosteatosis is a distinct degenerative condition from sarcopenia, possibly functioning synergistically with sarcopenia or preceding and promoting the development of sarcopenia [35]. Preclinical research indicates that mitochondrial dysfunction, impairing the ability to oxidize fatty acids, fosters the development of myosteatosis [36]. Initially, in a state of excess fatty acids, oxidation capacity would be increased to restrict lipid storage in muscle; however, in a state of chronic overload, the mitochondria become less functional and the rate of fatty acid oxidation declines. Clearly, the increase in fat oxidation rate observed during indirect calorimetry, which likely reflects more immediate dietary fatty acid intake, was inadequate for the mobilization of intermuscular adipose tissue in the present cohort.
Strengths of our study include a cohort of middle-aged PWH with elevated BMI on long-term contemporary INSTI-based ART regimens, which reflects the many aging individuals with HIV in the United States. The employment of CT and MR imaging at multiple time-points enabled assessment of multiple ectopic fat depots. Further, these data were combined with rigorous measurements of dietary intake, energy expenditure, and macronutrient oxidation - variables most often not included in translational studies. Limitations include a small sample size with inclusion only of males, which was due to the demographics of the local TVHS veteran population. Additionally, we do not know if the trajectory of changes observed continues beyond the 2-y observation period. Being observational data, we cannot draw conclusions regarding the direction of relationships. For example, reduced SAT, VAT, and liver fat may increase fat oxidation rate, or conversely. Finally, despite the rigorous methodology employed, self-report of physical activity and dietary intake has the potential for recall bias. Although 24 h recall is the diet assessment method with the lowest level of reporting variability, underreporting of dietary energy intake is well-established in all assessment methods [37]. It is also possible that participants’ reporting was influenced by familiarity with the assessment methodology over time (i.e., response bias).
In conclusion, we observed that males on long-term ART had spontaneous reductions in SAT and VAT mass, improvements in SAT and VAT densities, and reduced liver and thigh fat fractions over 2 y. A possible explanation for the improvements in body morphometrics observed is the concomitant increase in fatty acid oxidation rate detected via indirect calorimetry. Simultaneously, we found that this cohort of males had reduced dietary intakes of simple carbohydrates (total and added sugars). Although these findings were surprising and not entirely consistent with prior evidence from PWH, they suggest that long-term improvements much needed in the body composition of chronically treated PWH are possible with limited intervention. The finding of a site-specific coordinated pattern of changes in ectopic fat sites, occurring simultaneously with an alteration in the oxidation rates of fats and carbohydrates, is an important contribution to the field of characterizing and improving metabolic health in persons with chronic disease.
Author contributions
The authors’ responsibilities were as follows – JRK, HJS: concept and design; AW, MER, KL, JW, HE, MB, JRK, HJS: data acquisition, analysis, and interpretation; RF, FY, HJS: statistical analysis; JRK, AW, HJS: drafting of manuscript; JRK, HJS: critical revision of the manuscript for intellectual content; HJS, JRK: obtained funding; HJS: has primary responsibility for the final content; and all authors: read and approved the final manuscript.
Data availability
Data can be made available upon reasonable request to the corresponding author.
Funding
This work was supported by the Department of Veterans Affairs Merit Award (I01CX001930 to HJS and JRK), the National Center for Advancing Translational Sciences (CTSA award UL1TR002243), and the Tennessee Center for AIDS Research (P30 AI110527).
Conflict of interest
JRK has served as an advisor and received grant support from Merck & Co and Gilead Sciences.
All other authors report no conflicts of interest.
Acknowledgments
We thank the Veterans from the Tennessee Valley Healthcare System who participated in this study.
References
- 1.Centers for Disease Control and Prevention (CDC) State-specific prevalence of obesity among adults—United States, 2005. MMWR. 2006;55(36):985–988. [PubMed] [Google Scholar]
- 2.Koethe J.R., Jenkins C.A., Lau B., Shepherd B.E., Justice A.C., Tate J.P., et al. Rising obesity prevalence and weight gain among adults starting antiretroviral therapy in the United States and Canada. AIDS Res. Hum. Retrovir. 2016;32(1):50–58. doi: 10.1089/aid.2015.0147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Grant P.M., Kitch D., McComsey G.A., Collier A.C., Bartali B., Koletar S.L., et al. Long-term body composition changes in antiretroviral-treated HIV-infected individuals. AIDS. 2016;30(18):2805–2813. doi: 10.1097/QAD.0000000000001248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kuk J.L., Saunders T.J., Davidson L.E., Ross R. Age-related changes in total and regional fat distribution. Ageing Res. Rev. 2009;8(4):339–348. doi: 10.1016/j.arr.2009.06.001. [DOI] [PubMed] [Google Scholar]
- 5.Magkos F., Mantzoros C.S. Body fat redistribution and metabolic abnormalities in HIV-infected patients on highly active antiretroviral therapy: novel insights into pathophysiology and emerging opportunities for treatment. Metabolism. 2011;60(6):749–753. doi: 10.1016/j.metabol.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Basu R., Breda E., Oberg A.L., Powell C.C., Dalla Man C., Basu A., et al. Mechanisms of the age-associated deterioration in glucose tolerance: contribution of alterations in insulin secretion, action, and clearance. Diabetes. 2003;52(7):1738–1748. doi: 10.2337/diabetes.52.7.1738. [DOI] [PubMed] [Google Scholar]
- 7.Webster A.D., Sollid K. 2018. Is everyone really on a low-carbohydrate diet? Consumer Perceptions of Carbohydrates and Sugars. [Google Scholar]
- 8.Wu P., Zhang L., Zhao Y., Xu M., Tang Q., Chen G.C., et al. Adherence to the 2015-2020 dietary guidelines for Americans compared with the Mediterranean diet in relation to risk of prediabetes: results from NHANES 2007-2016. Nutrients. 2023;15(16):3546. doi: 10.3390/nu15163546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.sa S. Sriniva., Grinspoon S.K. Metabolic and body composition effects of newer antiretrovirals in HIV-infected patients. Eur. J Endocrinol. 2014;170(5):R185–R202. doi: 10.1530/EJE-13-0967. [DOI] [PubMed] [Google Scholar]
- 10.Pandey V., Mohan R., Kumar A., Gangadevi P., Kurien N. The impact of the COVID-19 outbreak on lifestyle-related behavior among the general population. Cureus. 2023;15(9) doi: 10.7759/cureus.45756. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mignogna C., Costanzo S., Ghulam A., Cerletti C., Donati M.B., de Gaetano G., et al. Impact of nationwide lockdowns resulting from the first wave of the COVID-19 pandemic on food intake, eating behaviors, and diet quality: a systematic review. Adv. Nutr. 2022;13(2):388–423. doi: 10.1093/advances/nmab130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chung H., Cobzas D., Lieffers J.R., Birdsell L., Baracos V. Automated segmentation of muscle and adipose tissue on CT images for human body composition analysis. SPIE Med. Imag. 2009 [Google Scholar]
- 13.Silver H.J., Niswender K.D., Kullberg J., Berglund J., Johansson L., Bruvold M., et al. Comparison of gross body fat-water magnetic resonance imaging at 3 Tesla to dual-energy x-ray absorptiometry in obese women. Obesity (Silver Spring) 2013;21(4):765–774. doi: 10.1002/oby.20287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Weir J.B. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol. 1949;109(1–2):1–9. doi: 10.1113/jphysiol.1949.sp004363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Frayn K.N. Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl. Physiol. Respir. Environ. Exerc. Physiol. 1983;55(2):628–634. doi: 10.1152/jappl.1983.55.2.628. [DOI] [PubMed] [Google Scholar]
- 16.Conway J.M., Ingwersen L.A., Moshfegh A.J. Accuracy of dietary recall using the USDA five-step multiple-pass method in men: an observational validation study. J Am. Diet Assoc. 2004;104(4):595–603. doi: 10.1016/j.jada.2004.01.007. [DOI] [PubMed] [Google Scholar]
- 17.Conway J.M., Ingwersen L.A., Vinyard B.T., Moshfegh A.J. Effectiveness of the US Department of Agriculture 5-step multiple-pass method in assessing food intake in obese and nonobese women. Am. J Clin. Nutr. 2003;77(5):1171–1178. doi: 10.1093/ajcn/77.5.1171. [DOI] [PubMed] [Google Scholar]
- 18.Silver H.J., Olson D., Mayfield D., Wright P., Nian H., Mashayekhi M., et al. Effect of the glucagon-like peptide-1 receptor agonist liraglutide, compared to caloric restriction, on appetite, dietary intake, body fat distribution and cardiometabolic biomarkers: a randomized trial in adults with obesity and prediabetes. Diabetes Obes. Metab. 2023;25(8):2340–2350. doi: 10.1111/dom.15113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Huang Y., Chen Z., Chen B., Li J., Yuan X., Li J., et al. Dietary sugar consumption and health: umbrella review. BMJ. 2023;381 doi: 10.1136/bmj-2022-071609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lichtenstein A.H., Appel L.J., Vadiveloo M., Hu F.B., Kris-Etherton P.M., Rebholz C.M., et al. 2021 Dietary guidance to improve cardiovascular health: a scientific statement from the American Heart Association. Circulation. 2021;144(23):e472–e487. doi: 10.1161/CIR.0000000000001031. [DOI] [PubMed] [Google Scholar]
- 21.WHO . WHO guideline; Geneva: 2023. Carbohydrate intake for adults and children.https://www.who.int/publications/i/item/9789240073593 Available from: [PubMed] [Google Scholar]
- 22.U.S. Department of Health and Human Services and U.S. Department of Agriculture . 8th Edition. 2015. 2015–2020 Dietary Guidelines for Americans.https://odphp.health.gov/sites/default/files/2019-09/2015-2020_Dietary_Guidelines.pdf Available from: [Google Scholar]
- 23.Government of Canada . Health Canada; Ottawa: 2020. Canada’s Food Guide. Resources for health professionals and policy makers. [Google Scholar]
- 24.DiFrancesco L., Fulgoni V.L., 3rd, Gaine P.C., Scott M.O., Ricciuto L. Trends in added sugars intake and sources among U.S. adults using the National Health and Nutrition Examination Survey (NHANES) 2001-2018. Front Nutr. 2022;9 doi: 10.3389/fnut.2022.897952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Luzi L., Perseghin G., Tambussi G., Meneghini E., Scifo P., Pagliato E., et al. Intramyocellular lipid accumulation and reduced whole body lipid oxidation in HIV lipodystrophy. Am. J Physiol. Endocrinol. Metab. 2003;284(2):E274–E280. doi: 10.1152/ajpendo.00391.2001. [DOI] [PubMed] [Google Scholar]
- 26.Korach M., Leclercq P., Péronnet F., Leverve X. Metabolic response to a C-glucose load in human immunodeficiency virus patients before and after antiprotease therapy. Metabolism. 2002;51(3):307–313. doi: 10.1053/meta.2002.30505. [DOI] [PubMed] [Google Scholar]
- 27.Kosmiski L.A., Kuritzkes D.R., Sharp T.A., Hamilton J.T., Lichtenstein K.A., Mosca C.L., et al. Total energy expenditure and carbohydrate oxidation are increased in the human immunodeficiency virus lipodystrophy syndrome. Metabolism. 2003;52(5):620–625. doi: 10.1053/meta.2003.50103. [DOI] [PubMed] [Google Scholar]
- 28.Vassimon H.S., de Paula F.J., Machado A.A., Monteiro J.P., Jordão A.A., Jr. Hypermetabolism and altered substrate oxidation in HIV-infected patients with lipodystrophy. Nutrition. 2012;28(9):912–916. doi: 10.1016/j.nut.2011.12.010. [DOI] [PubMed] [Google Scholar]
- 29.Wagner R., Eckstein S.S., Yamazaki H., Gerst F., Machann J., Jaghutriz B.A., et al. Metabolic implications of pancreatic fat accumulation. Nat. Rev. Endocrinol. 2022;18(1):43–54. doi: 10.1038/s41574-021-00573-3. [DOI] [PubMed] [Google Scholar]
- 30.Gaborit B., Abdesselam I., Kober F., Jacquier A., Ronsin O., Emungania O., et al. Ectopic fat storage in the pancreas using 1H-MRS: importance of diabetic status and modulation with bariatric surgery-induced weight loss. Int. J Obes. (Lond.). 2015;39(3):480–487. doi: 10.1038/ijo.2014.126. [DOI] [PubMed] [Google Scholar]
- 31.Cunha G.M., Thai T.T., Hamilton G., Covarrubias Y., Schlein A., Middleton M.S., et al. Accuracy of common proton density fat fraction thresholds for magnitude- and complex-based chemical shift-encoded MRI for assessing hepatic steatosis in patients with obesity. Abdom. Radiol. (NY). 2020;45(3):661–671. doi: 10.1007/s00261-019-02350-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Dodell G.B., Kotler D.P., Engelson E.S., Ionescu G., Gimelshteyn Y., Pollack A., et al. Intermuscular and subcutaneous adipose tissue distributions differ in HIV+ versus HIV-men and women. Int. J Body Compos. Res. 2009;7(2):73–78. [PMC free article] [PubMed] [Google Scholar]
- 33.SeyedAlinaghi S., Ghayomzadeh M., Mirzapour P., Maroufi S.F., Pashaei Z., Ali Z., et al. A systematic review of sarcopenia prevalence and associated factors in people living with human immunodeficiency virus. J Cachexia. Sarcopenia. Muscle. 2023;14(3):1168–1182. doi: 10.1002/jcsm.13212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Swartz A.Z., Robles M.E., Park S., Esfandiari H., Bradshaw M., Koethe J.R., et al. Cardiometabolic characteristics of obesity phenotypes in persons with HIV. Open Forum Infect. Dis. 2024;11(7) doi: 10.1093/ofid/ofae376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ahn H., Kim D.W., Ko Y., Ha J., Shin Y.B., Lee J., et al. Updated systematic review and meta-analysis on diagnostic issues and the prognostic impact of myosteatosis: a new paradigm beyond sarcopenia. Ageing Res. Rev. 2021;70 doi: 10.1016/j.arr.2021.101398. [DOI] [PubMed] [Google Scholar]
- 36.Gumucio J.P., Qasawa A.H., Ferrara P.J., Malik A.N., Funai K., McDonagh B., et al. Reduced mitochondrial lipid oxidation leads to fat accumulation in myosteatosis. FASEB J. 2019;33(7):7863–7881. doi: 10.1096/fj.201802457RR. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Burrows T.L., Ho Y.Y., Rollo M.E., Collins C.E. Validity of dietary assessment methods when compared to the method of doubly labeled water: a systematic review in adults. Front Endocrinol. (Lausanne). 2019;10:850. doi: 10.3389/fendo.2019.00850. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data can be made available upon reasonable request to the corresponding author.