Weight gain and waist-to-height ratio in people living with HIV after switching to integrase strand transfer inhibitors in Guatemala
DW Ortíz1, HE Marroquin1, L Larson2, KB Franco1, A Spec2, JR Melendez1, R Pinzón1, AJ Samayoa1, C Mejia-Chew2, JA O'Halloran2
1Unidad de Atención Integral de VIH e Infecciones Crónicas “Dr. Carlos Rodolfo Mejía Villatoro”, Hospital Roosevelt, Guatemala; 2Department of Medicine, Division of Infectious Diseases, Washington University in Saint Louis
Objectives: Ever since the inclusion of integrase strand transfer inhibitors (INSTI) as first-line therapy for HIV, low-to-middle income countries began simplifying antiretroviral therapy (ART) regimens to INSTI regimens to reduce costs and improve treatment availability. However, undesired weight gain and increased waist adiposity associated with INSTI use remains a concern. To date there are no studies relating these outcomes to INSTI use in Guatemala or comparing their effect with older regimens. Our objective is to compare weight gain in PLWH switched to INSTI with PLWH that did not switch treatment and to determine risk factors associated with significant weight gain and high waist-to-height ratio (WTH) in PLWH receiving care in Hospital Roosevelt in Guatemala.
Methods: Data from a prospective cohort of PLWH enrolled between July and March 2020 was analysed. We included all PLWH over the age of 18 with over 6 months of ART use and with a 6-month follow-up. Participants included in the switch group were prescribed INSTI for at least 6 months. Normality was assessed with K-S and Shapiro Wilk test. Independent samples t-test was used to determine mean difference (MD) in the weight change experienced between groups. Demographics, lifestyle, HIV parameters, anthropometrics and treatment history were compared based on signifi-cant weight gain and high WTH ratio using univariable logistic regression. Significant weight gain was defined as weight at follow-up over 5% from weight at enrolment and high WTH ratio was defined as WTH ratio above the cutoff for gender (0.579 for men, 0.587 for women).
Results: Of 166 participants enrolled, 55.4% were males and 14% had indigenous background. High WTH ratio was found in 33.7% participants and 20.5% gained over 5% of their weight at enrolment. Median follow-up time was 6.2 months. Forty-one participants (24%) switched to INSTI-based ART. Mean weight change was +1.90 kg in the switch-group and +1.21 kg in the no-switch group without difference among groups (MD=0.065, 95% CI: -1.39, 0.094; P=0.086). In the univariable analysis, alcohol use and smoking at follow-up were associated with significant weight gain. Female gender and low physical activity were associated with high WTH ratio (Table 1). Switching to INSTI-based regimens was not significantly different based on high WTH ratio or significant weight gain.
(Abstract O05)
Table 1. Univariable logistic regression of 166 people living with HTV that report significant weight gain and high waist-to-height ratio after six month follow up at a large HIV clinic in Guatemala
Discussion: Our findings are consistent with other studies that report weight gain over 1 kg after 6 months on INSTI-based regimens. However, we found no association that weight gain experienced in the switch group was different than in the no-switch group. Demographic and lifestyle factors were significantly associated with outcomes, which further implicate their influence in body composition regardless of ART. Further studies and longer follow-up times are needed to determine the long-term INSTI effect in our cohort.
Abstract O06
Antiviral Therapy 2020; 25 Suppl 1:A9
Persistence of abnormal central fat distribution in adults with HIV acquired from early childhood
SJ Sahagun1, T Yeramosu1, JB Purdy2, J Reynolds2, CM Hadigan1
1National Institute of Allergy and Infectious Diseases, NIH, Bethesda, MD; 2National Institutes of Health, Clinical Center, Bethesda, MD
Objective: Antiretroviral (ARV)-associated effects on body composition and obesity remain a significant concern among people living with HIV (PLWHIV). In adults infected from birth or in early life, little is known of the long-term effects of extensive ARV exposure on body composition and metabolism. This study explores the changes in body fat distribution and associated metabolic factors in relation to ARV exposure.
Methods: Body composition of young adults who acquired HIV early in life (n=70; perinatal or transfusion-acquired) was cross-sectionally compared to that of age- and sex-matched healthy controls (n=47). As part of a natural history cohort study, longitudinal body composition data were available for a subset of PLWHIV (n=40). Regional body fat distribution, particularly trunk-limb fat ratio, was obtained using dual energy X-ray absorptiometry (DEXA). Waist-hip ratio (WHR), BMI, fasting glucose, insulin, and lipids, total CD4, viral RNA and detailed ARV exposure history were obtained.
Results: Although PLWHIV had similar BMI relative to controls, WHR, trunk-limb fat ratio, HOMA-IR and triglycerides were significantly greater in PLWHIV (Table 1). WHR was positively correlated with cumulative exposure to PIs (r=0.275; P=0.02) as well as NRTIs and NNRTIs (P≤0.04). The specific ARV agents significantly correlated with WHR were tenofovir DF (r=0.358; P=0.002) and nevirapine (r=0.275; P=0.02). Trunk-limb fat ratio was positively correlated with cumulative years on nelfinavir (r=0.346; P=0.003) and NRTIs (r=0.314; P=0.008), specifically stavu-dine (r=0.424; P=0.0003) and tenofovir DF (r=0.238; P<0.05). Neither WHR nor trunk-limb fat ratio were related to cumulative exposure to INSTIs.
(Abstract O06)
Table 1. Clinical Characteristics and Body Composition of PLWHIV and Healthy Controls
Note: Statistics are reported as mean ± standard deviation, except as indicated.
Significantly different from controls at p<.05
Significantly different from baseline at p<.05
Longitudinal analysis revealed similar CD4 T-cell count and rates of viral suppression over a median follow-up of 7 years (Table 1). Significant increases were observed in WHR, trunk-limb fat ratio, percentage body fat and percentage trunk fat. Interestingly, the rates of overweight (28% versus 53%) and obesity (12.5% versus 25%) doubled and increases in BMI correlated with longer exposure to certain ARVs during follow-up. Particularly, changes in BMI were positively correlated with longer exposure to abacavir (r=0.395; P=0.01), lamivudine (r=0.546; P=0.0003) and raltegravir (r=0.329; P=0.04). Change in trunk-limb fat ratio was positively correlated with longer exposure to stavudine (r=0.398; P=0.01) and didanosine (r=0.385; P=0.01) but was inversely correlated with emtricitabine (r=-0.326; P=0.04). Change in WHR was not related to ARV exposure during follow-up.
Conclusion: This study presents strong evidence for sustained alterations in body fat distribution with persistent or worsening central adiposity in young adults with life-long HIV. Not only did the effects of ARV agents previously recognized to contribute to lipodys-trophy persist in adulthood, but more contemporary ARVs were associated with weight gain. As this cohort ages, further evaluation of life-long ARV use and metabolic risk factors are warranted to mitigate the risks of cardiovascular disease and the health consequences of obesity.
Abstract O07
Antiviral Therapy 2020; 25 Suppl 1:A11
Greater increases over time in BMI, waist circumference and subcutaneous and visceral fat area among men with HIV
JE Lake1, S Langan2, J Sun2, S Adrian3, WS Post2, LA Kingsley4, FJ Palella Jr5, M Budoff6, TT Brown2, KM Erlandson3
1UTHealth, Houston, TX, USA; 2Johns Hopkins University, Baltimore, MD, USA; 3University of Colorado Anschutz Medical Center, Aurora, CO, USA; 4University of Pittsburgh, Pittsburgh, PA, USA; 5Northwestern University, Chicago, IL, USA; 6Harbor-UCLA Medical Center, Torrance, CA, USA
Objectives: Adipose tissue (AT) abnormalities are common in people with HIV (HIV+) and are associated with metabolic disease. AT area and density may provide complimentary information. In the traditional obesity paradigm, increases in AT area are associated with decreases in density due to lipid engorgement of adipocytes. We assessed longitudinal changes in AT area and density by HIV serostatus among men in the Multicenter AIDS Cohort Study (MACS).
Methods: Men who completed MACS sub-studies CVD2 (2011–2013, 40–70 years old, no known cardiovascular disease) and CVD3 (2015–2017) underwent measurement of visceral (VAT) and subcutaneous (SAT) AT area (cm2) and density (Hounsfield Units [HU]) from single-slice L4-L5 abdominal CT scans. Wilcoxon rank sum tests compared between-group parameters.
Results: At CVD2, HIV+ (n=201) and HIV-negative (HIV-, n=126) men had median age 52 and 57 years, BMI 26 kg/m2 (both groups), 35% and 28% were Black, and 12% and 7% were Hispanic, respectively. HIV+ men had current median CD4+ T-lymphocyte count 642 cells/ml, 88% HIV-1 RNA <50 copies/ml, 10 years on highly active antiretroviral therapy (HAART, 89% with historical thymidine analogue NRTI exposure). Current ART use was 51% PI-, 43% NNRTI- and 16% INSTI-based. VAT area (HIV- 137 cm2, HIV+ 158 cm2; P=0.16) and density (HIV- −86 HU, HIV+ −87 HU; P=0.30) were similar by HIV serostatus, but HIV- men had greater SAT area (209 versus 170 cm2; P=0.0001) and less dense SAT (−96 versus −93 HU; P=0.04). Over a median of 4.8 years, HIV+ men experienced small but significant increases in BMI and waist circumference, with a trend towards greater SAT and VAT area gain than HIV- men. However, changes in VAT and SAT density were similar for HIV+ and HIV- men (Table 1). Concordance in directionality of changes in SAT and VAT area (HIV+ 71%, HIV- 77%) and density (92% for both) changes were high, suggesting frequency of redistribution from subcutaneous to visceral depots was low and not significantly more common among HIV+ men.
(Abstract O07)
Anthropometric Change Between CVD2 and CVD3 by HIV Serostatus
HIV+
HIV-
P value
BMI (kg/m2)
0.1 (-0.7, 1,4)
-0.1 (-1.2, 0.8)
0.02
Waist circumference (cm)
3.2 (0.1, 7.5)
1.5 (-1.6, 5.3)
0.007
SAT ares (cm2)
12.4 (-5.4, 37.7)
4.9 (-18.8, 35.5)
0.09
SAT density (HU)
-2.5 (-1.4, 7.1)
-3.3 (-1.7, 7.1)
0.79
VAT area (cm2)
18.9 (-14.1, 48.8)
8.6 (-27.3, 39.0)
0.09
VAT density (HU)
-4.3 (-0.4, 0.5)
-3.0 (-2.2, 8.1)
0.20
Conclusions: Although HIV+ men were younger and had less SAT at study start, HIV+ men had larger increases in BMI, waist circumference, SAT area and VAT area than HIV- men over a similar time interval. Changes in VAT and SAT density did not differ by HIV serostatus, with all men generally experiencing decreases in AT density, as expected accompanying adi-pocyte hypertrophy during weight gain.
Abstract O08
Antiviral Therapy 2020; 25 Suppl 1:A12
Expansion of CD4+ T effector memory CD45RA+ (TEMRA) cells is associated with incident diabetes in veterans with HIV
SB Bailin1, S Kundu2, K So-Armah3, MF Doyle4, RP Tracy4, M Wellons5, CN Wanjalla1, AL Landay6, S Mallal1,7, AC Justice8,9, MS Freiberg2,10, JR Koethe1,10
1Division of Infectious Diseases, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USA; 2Division of Cardiovascular Medicine, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USA; 3Boston University School of Medicine, Boston, Massachusetts, USA; 4Department of Pathology and Laboratory Medicine, University of Vermont, Burlington, Vermont, USA; 5Divison of Diabetes, Endocrinology, and Metabolism, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USA; 6Department of Internal Medicine, Rush University Medical Center, Chicago, IL, USA; 7Center for Translational Immunology and Infectious Diseases, Vanderbilt University Medical Center, Nashville, Tennessee, USA; 8Veterans Affairs Connecticut Healthcare System, West Haven, Connecticut, USA; 9Department of Internal Medicine, Yale School of Medicine, West Haven, Connecticut, USA; 10Veterans Affairs Tennessee Valley Healthcare System, Nashville, Tennessee, USA
Objectives: Differences in the relative distribution of T-cell subsets have been linked to a higher prevalence of cardiovascular and metabolic diseases in the general population. Given the persistent immune activation associated with HIV infection, we assessed whether T-cell subsets were associated with incident diabetes in HIV-positive and HIV-negative veterans.
Methods: We performed flow cytometry and functional assays on peripheral blood mononuclear cells collected from HIV-positive and HIV-negative participants in the Veterans Aging Cohort Study between 2005 and 2007 to characterize CD4+ and CD8+ memory (central, effector and effector RA+ [TEMRA]), CD57+, CD28−, and TH1, TH2, and TH17 CD4+ T-cells. We used two definitions of TEMRA cells: CD4+CD45RA+CD28−CD57+ and CD4+CD45RA+CD27− (Table 1). Cases of incident diabetes were identified by two-physician chart adjudication. Individuals were followed until the onset of diabetes, death or censored on 9/30/2015. We assessed HIV-positive and HIV-negative veterans separately using Cox proportional hazards models with incident diabetes as the outcome and T-cell subset as the main exposure, adjusted for age, race, time updated body mass index, cytomegalovirus serostatus, hepatitis C virus serostatus, alcohol use, high-density lipoprotein, low-density lipoprotein, total cholesterol, circulating inflammatory markers (interleukin-6, d-dimer and soluble CD14) and viral load and antiretroviral therapy use (HIV-positive only). We report the hazard ratio (HR) for incident diabetes per standard deviation increment in the T-cell subset.
(Abstract O08)
Table 1. Cox proportional hazards models, stratified by HIV status, assessing the relationship of baseline T cell subsets and incident diabetes adjusted for age, race, cytomegalovirus serostatus, viral load and antiretroviral therapy use (HIV-positive only), high-density lipoprotein, low-density lipoprotein, total cholesterol, time updated body mass index, hepatitis C virus, history of alcohol abuse, and circulating concentrations of interleukin-6, D-dimer, and soluble CD14. CI, confidence interval; SD, standard deviation
HIV-negative (N = 578)
HIV-positive (N = 1259)
Hazard Ratio per SD increment (CI)
P value
Hazard Ratio per SD increment (CI)
P value
CD4+ T cell subset
CD4+CD45RA+CD28CD57+ (TFMRA)
1.06 [0.77,1.44]
0.73
1.16 [1.00,1.34]
0.05
CD4+CD45RA+CD27− (TEMRA)
1.10 [0.93,1.30]
0.27
1.20 [1.04,1.38]
0.01
CD4+CD28−
1.03 [0.76,1.40]
0.86
1.16 [0.99,1.36]
0.06
CD4+CD57+CD28−
1.01 [0.87,1.18]
0.85
0.96 [0.84,1.09]
0.52
CD8+ T cell subset
CD8+CD45RA+CD28−CD57+ (TEMRA)
0.88 [0.73,1.07]
0.21
1.08 [0.91,1.29]
0.37
CD8+CD45RA+CD27− (TEMRA)
0.82 [0.67,1.01]
0.06
1.13 [0.96,1.33]
0.15
CD8+CD28−
0.87 [0.70,1.08]
0.22
0.99 [0.83,1.18]
0.90
CD8+CD57+CD28−
1.04 [0.89,1.22]
0.60
1.00 [0.86,1.16]
0.98
Results: A total of 1,259 HIV-positive and 578 HIV-negative individuals were diabetes free at baseline (date of blood collection) and there were 238 incident diabetes events (133 [10.6%] in HIV-positive and 105 [18.2%] in HIV-negative) over a median follow-up time of 8.6 years. Persons with HIV had a median age of 51 years, were 69% Black, 97% male, 65% virologically suppressed and 84% on antiretroviral therapy. HIV-negative individuals had a median age of 52 years, were 68% Black and 90% male. In the fully adjusted model, the risk of incident diabetes increased with higher baseline proportion of CD4+ TEMRA cells using both definitions (above) in HIV-positive persons only (HR 1.16 [1.00,1.34]; P=0.05 and HR 1.20 [1.04,1.38]; P=0.01; Table 1). The diabetes risk associated with CD4+ TEMRA was present even after adjustment for CMV serostatus, a known cause of TEMRA cell inflation. No other T-cell subsets were significantly associated with incident diabetes in HIV-positive or HIV-negative individuals.
Conclusions: In an observational cohort of veterans with and without HIV, higher baseline CD4+ TEMRA cells was associated with an increased risk of incident diabetes in HIV-positive individuals. HIV infection appears to drive expansion of terminally differentiated CD4+ memory T-cells that may predispose to the development of diabetes. Further studies in this cohort will use unsupervised analysis techniques to further characterize T-cells associated with diabetes and other comorbidities.
