Abstract
Background
Prospective investigations on the risk of cardiovascular disease among youth with perinatally acquired human immunodeficiency virus (PHIV) in sub-Saharan Africa are lacking.
Methods
A prospective observational cohort study was performed in 101 youth (aged 10–18 years) with PHIV and 97 who were human immunodeficiency virus (HIV) uninfected (HIV−), from 2017 to 2021 at the Joint Clinical Research Center in Uganda. Participants with PHIV were receiving antiretroviral therapy (ART) and had HIV-1 RNA levels ≤400 copies/mL. The common carotid artery intima-media thickness (IMT) and pulse wave velocity (PWV) were evaluated at baseline and at 96 weeks. Groups were compared using unpaired t-test, and potential predictors of IMT and PWV were assessed using quantile regression.
Results
Of the 198 participants recruited at baseline, 168 (89 with PHIV, 79 HIV−) had measurements at 96 weeks. The median age (interquartile range) age was 13 (11–15) years; 52% were female, and 85% had viral loads <50 copies/mL that remained undetectable at week 96. The baseline mean common carotid artery IMT was slightly higher in participants with PHIV compared with controls (P < .01), and PWV did not differ between groups (P = .08). At week 96, IMT decreased and PWV increased in the PHIV group (P ≤ .03); IMT increased in the HIV− group (P = .03), with no change in PWV (P = .92). In longitudinal analyses in those with PHIV, longer ART duration was associated with lower PWV (β = .008 [95% confidence interval, −.008 to .003]), and abacavir use with greater IMT (β = .043 [.012–.074]).
Conclusions
In healthy Ugandan youth with PHIV, virally suppressed by ART, the common carotid artery IMT did not progress over 2 years. Prolonged and early ART may prevent progression of subclinical vascular disease, while prolonged use of abacavir may increase it.
Keywords: cardiovascular disease, inflammation, immune activation, perinatally acquired HIV, complications
Children and adolescents with perinatally acquired human immunodeficiency virus (HIV) in Uganda and viral suppression show no evidence of worsening in subclinical atherosclerosis over time but did show slightly increased arterial stiffness over 2 years, compared with HIV-uninfected participants.
The World Health Organization estimates there are >2 million children with human immunodeficiency virus (HIV), mostly acquired via perinatal transmission [1]. The majority of individuals with perinatally acquired HIV (PHIV) are now heading into adolescence or adulthood [2, 3]. Widespread access to antiretroviral therapy (ART) has transformed HIV from a fatal condition into a chronic disease, leading to increases in non-AIDS events, such as cardiovascular disease (CVD).
While CVD has increased in the general population in sub-Saharan Africa [4], published data are still lacking on the burden of CVD in people with HIV there [5]. Limited available data on adults with HIV in this setting suggest an increased risk of early atherosclerosis and stroke. However, there have been no large cohort studies including HIV-uninfected (HIV−) patients in the ART era are lacking [5]. Youth with PHIV in sub-Saharan Africa have been particularly understudied. The limited studies on longitudinal outcomes and data outside infancy and early childhood are primarily from higher-resource settings. There are unique differences in HIV and ART exposure (in utero) and treatment during periods of critical development that may exacerbate comorbid conditions in this population compared with adults.
Our understanding of the trajectory of cardiometabolic complications in children and adolescents with HIV is therefore limited. Our group has previously described subclinical vascular disease seen in a cross-sectional analysis of children with PHIV and age- and sex-matched HIV− children in Uganda [6]. In the current study, we aimed to investigate the progression of subclinical vascular disease over 2 years by HIV status and explore the relationship between vascular disease measures and ART, metabolic parameters, and markers of inflammation, immune activation and gut integrity in this same cohort of children and adolescents with PHIV and an HIV− control group.
METHODS
Study Design
This was a 96-week prospective longitudinal cohort study of children and adolescents with PHIV and an HIV− comparison group enrolled at the Joint Clinical Research Center in Kampala, Uganda. Full study design details have been reported elsewhere [6, 7]. All participants were 10–18 years of age. Participants with PHIV have been receiving ART for ≥2 years with a stable regimen for at least the last 6 months (HIV-1 RNA levels <400 copies/mL). HIV− participants were tested during clinic visit to confirm their HIV-seronegative status with a rapid HIV test (HIV 1/2 STAT-PAK; Chembio; sensitivity, 99.7%, specificity, 99.9%). Participants with evidence of self-reported or documented acute infection (malaria, tuberculosis, helminthiasis, pneumonia, or meningitis) or malnutrition and diarrhea in the last 3 months were excluded, and were those with known diabetes or CVD and adolescents who were pregnant or intended to become pregnant. Participants were enrolled between September 2017 and March 2019.
Study Evaluations
Participants were seen at 48 and 96 weeks, and the study was completed in February 2021. At each visit, common carotid artery intima-media thickness (IMT) and pulse wave velocity (PWV) were recorded, along with family history, physical activity, and socioeconomic variables. Physical examinations were performed to determine Tanner staging [8]. Body mass index (BMI)–for–age z scores were determined using World Health Organization 2007 reference values. ART start and stop dates were abstracted from the medical charts and categorized by class for protease inhibitors (PIs), nucleotide reverse-transcriptase inhibitors, nonnucleoside reverse-transcriptase inhibitors (NNRTIs), and integrase strand transfer inhibitors.
In addition, blood samples were obtained after an 8-hour fast. Blood was processed and plasma, serum, and peripheral blood mononuclear cells were cryopreserved at –80°C for batch shipment to University Hospitals Cleveland Medical Center in Cleveland, Ohio.
Cellular Markers of Monocyte and T-Cell Activation
Monocyte and T cells were phenotyped using flow cytometry, as described elsewhere [7]. CD4+ and CD8+ T-cell activation was defined as coexpression of CD38 and HLA-DR. Monocyte subset proportions were determined by the relative expression of CD14 and CD16.
Inflammation, Soluble Immune Activation, and Gut Markers
Zonulin (Promocell), BDG (Mybiosource), intestinal fatty acid binding protein (I-FABP), soluble CD14, soluble CD163, soluble tumor necrosis factor receptor I, high-sensitivity C-reactive protein, interleukin 6 (R&D Systems), and oxidized low-density lipoprotein (ALPCO and Mercodia) were measured by means of enzyme-linked immunosorbent assay. The intra-assay variability ranged between 4% and 8%, and the interassay variability was <10% for all markers. All assays were done in a laboratory at Ohio State University in Columbus under the supervision of Dr. Funderburg. Laboratory personnel were blinded to group assignments.
Subclinical Vascular Disease
To minimize differences in measurements, an experienced ultrasonographer performed all IMT and PWV measurements at the Joint Clinical Research Center, following published guidelines [9] and as described elsewhere [6], for all visits. Briefly, B mode ultrasonongraphy of the carotid arteries was performed using a Philips iU22 system with 12–3-MHz broadband linear array probe (Philips). Five-second cine loops of the distal common carotid artery were obtained in longitudinal views at 3 angles (anterior, lateral, and posterior) bilaterally. An experienced reader (D. L.), blinded to HIV status, measured the IMT offline throughout the study, using semiautomated edge detection software (Medical Imaging Applications). Measurements (in millimeters) obtained from the right and left sides were then averaged and reported as a single mean-mean IMT (hereafter mean IMT) and mean-maximum IMT (hereafter max IMT).
PWV, in meters per second, was measured by means of carotid to femoral applanation tonometry (Vicorder; SMT Medical). In brief, the carotid to femoral path length was measured, and the tonometer was applied to the carotid artery and femoral artery in sequence to obtain waveforms in relationship to the electrocardiographic tracing. An average of 3 measurements was used for analyses. Higher-velocity measurements correspond to greater arterial stiffness.
Statistical Analysis
All variables were compared between groups using Wilcoxon rank sum or Fisher exact tests, as appropriate. We presented max IMT and PWV data using box plots. The primary outcome variables of interest in this study were PWV and IMT measurements. We used Spearman correlation to study correlations between outcomes and predictor variables of interest. Based on the correlational analysis results, we included potential risk factors with P < .05 in the regression analyses. We fitted generalized estimating equations with unstructured correlations on the 2 longitudinal outcome variables over 96 weeks. The generalized estimating equation models included terms for observation time, HIV status, and an interaction term, HIV status × time.
After the interaction analysis, we estimated effects of HIV status at each time point of the study by combining the coefficients of HIV status and the interaction term. Furthermore, to assess the longitudinal changes in the IMT (or in the PWV) over the study period in PHIV, unadjusted and adjusted models were fitted for each CVD measure separately. We adjusted for age, sex, BMI, and CD4+ T-cell count at baseline. Again, we restricted our analysis to the PHIV group only to assess the association of IMT and PWV with ART, ART duration, and exposure to thymidine analogues. Models were fit for each drug separately, owing to drug substitutions during the study; we focused our analysis on ART exposure at time of the baseline visit by class (PIs and integrase strand transfer inhibitors) and on ART known to be associated with cardiometabolic complications (abacavir and tenofovir), and we then adjusted for age, sex, BMI, and CD4+ T-cell count at baseline. All statistical analyses were performed using Stata 15.0 and R 4.0.5 software.
RESULTS
Participant Characteristics
Of the 198 participants recruited at baseline (101 in the PHIV and 97 in the HIV− group), 168 (89 and 79, respectively) had measurements at 96 weeks. Reasons for dropout before 96 weeks included loss to follow-up (n = 10), fear/inability to come to the clinic during the first wave of the coronavirus disease 2019 pandemic (n = 10), relocation (n = 9), and pregnancy (n = 1). The median age at enrollment was 13 years ,and 52% of participants were female. At baseline, participants with PHIV had higher waist-to-hip ratio, homeostatic assessment of insulin resistance, and worse lipid profiles (P ≤ .03).
At baseline, all participants were Tanner stage III or below, at week 96; 31% of participants were identified as Tanner stage IV or V. There was no difference between participants in BMI at baseline and week 96. The BMI increase was slightly more pronounced in participants with PHIV than in the HIV− group (median increase, 1.21 [interquartile range (IQR), 1.48–1.58] vs 0.96 [0.97–1.2], respectively, with BMI calculated as weight in kilograms divided by height in meters squared).
At baseline, 85% of participants with PHIV had a viral load <50 copies/mL, which remained undetectable at week 96. At baseline 72% were on a NNRTI-containing regimen, 28% were on lopinavir-ritonavir and 2 participants were on dolutegravir; 71% had a history of thymidine exposure (zidovudine, stavudine, or didanosine). During the 96-week study period, 53% of participants had drug substitutions, 85% of whom switched to lamivudine, tenofovir, and dolutegravir for regimen optimization, based on the Ugandan Ministry of Health HIV Guidelines [10].
The median duration of ART was 10 years, with 8 participants starting ART at age <1 year and the rest between the ages of 2 and 10 years. All participants with PHIV were receiving cotrimoxazole, and no participants were receiving tuberculosis medications. Those with PHIV had higher monocyte and T-cell activation at baseline, along with higher soluble CD14 levels (P = .01) and elevated frequencies of nonclassic monocytes (P < .001). The details of the baseline characteristics are in Table 1.
Table 1.
Participant Characteristics at Baseline
| Characteristic | Median Value (IQR)a | P Value | |
|---|---|---|---|
| PHIV Group (n = 101) |
HIV− Group (n = 96) |
||
| Demographic variables | |||
| ȃAge, y | 12.93 (11.53–14.71) | 12.67 (11.08–14.33) | .24 |
| ȃFemale sex, no. (%) | 54 (53) | 50 (52) | .96 |
| ȃSmoking history, no. (%) | 0 | 0 | … |
| HIV variables | |||
| ȃViral load <20 copies/mL, no. (%) | 84 (86) | … | … |
| ȃCD4+ T-cell count nadir, cells/µL | 619.50 (333–1097) | … | … |
| ȃCD4+ T-cell count, cells/µL | 988 (631–1310) | … | |
| ȃCD4+ T-cell proportion, % | 34.50 (27–41) | … | … |
| ȃART duration, y | 9.88 (7.61–11.08) | … | … |
| ȃNRTI treatment, no. (%) | |||
| ȃȃAbacavir | 42 (47) | … | … |
| ȃȃLamivudine | 1 (1) | … | … |
| ȃȃTenofovir | 12 (14) | … | … |
| ȃȃZidovudine | 33 (37) | … | … |
| ȃNevirapine | 18 (27) | … | … |
| ȃEfavirenz | 46 (44) | … | … |
| ȃLopinavir-ritonavir | 27 (28) | … | … |
| ȃLifetime exposure to thymidine analogue | 72 (71) | … | … |
| Socioeconomic characteristics, no. (%) | |||
| ȃFood insecurity | 25 (25) | 35 (37) | .047 |
| ȃElectricity | 22 (22) | 11 (12) | .08 |
| ȃLiving in extreme poverty | 38 (40) | 52 (54) | .05 |
| CVD and metabolic parameters | |||
| ȃWaist-hip ratio | 0.87 (0.83–0.90) | 0.85 (0.82–0.89) | .02b |
| ȃBMIc | 17.69 (15.82–19.10) | 18.02 (16.09–19.46) | .34 |
| ȃBMI-for-age z score | −0.57 (−1.29 to −.01) | −0.33 (−1.04 to .36) | .07 |
| ȃBlood pressure, mm Hg | |||
| ȃȃSystolic | 107 (101–114) | 115 (106–121) | <.01b |
| ȃȃDiastolic | 65 (60–71) | 67 (62–73) | .07 |
| ȃHOMA-IR | 1.25 (0.80–2.05) | 1.00 (0.74–1.58) | .02b |
| ȃTotal cholesterol, mg/dL | 152 (134–172) | 148 (131–170) | .49 |
| ȃLDL, mg/dL | 84 (68–104) | 83 (70–105) | .61 |
| ȃVLDL, mg/dL | 19 (13–23) | 16 (12–20) | .03b |
| ȃTriglycerides, mg/dL | 93 (66–115) | 82 (61–102) | .03b |
| ȃAverage PWV, m/s | 5.73 (5.40–6.33) | 6.02 (5.48–6.57) | .08 |
| Soluble markers of systemic inflammation and monocyte activation | |||
| ȃhsCRP, ng/mL | 515.92 (185.69–1619.31) | 398.60 (126.77–1145.82) | .16 |
| ȃIL-6, pg/mL | 1.13 (0.77–2.12) | 1.23 (0.84–1.86) | .93 |
| ȃsTNFRI, pg/mL | 905.46 (736.92–1034.54) | 959.90 (811.10–1107.97) | .02b |
| ȃsCD14, pg/mL | 2110.70 (1759.38–2568.92) | 1675.39 (1445.09–1956.71) | <.01b |
| ȃsCD163, pg/mL | 589.60 (404.24–735.95) | 684.97 (495.51–854.47) | .01b |
| Cellular monocyte and lymphocyte activation, % | |||
| ȃCD14+CD16− | 0.68 (0.59–0.78) | 0.71 (0.65–0.77) | .63 |
| ȃCD14+CD16+ | 0.23 (0.15–0.30) | 0.23 (0.17–0.27) | .89 |
| ȃCD14dimCD16+ | 0.08 (0.06–0.11) | 0.07 (0.04–0.09) | .02b |
| ȃCD4+CD38+HLA-DR+ | 4.10 (2.89–5.28) | 3.58 (2.54–4.66) | .050b |
| ȃCD4+CD38+ | 11.57 (10.18–13.19) | 11.66 (9.81–13.30) | .83 |
| ȃCD4+HLA-DR+ | 11.25 (9.59–12.67) | 9.24 (7.58–10.19) | <.001b |
| ȃCD4+CD38+HLA-DR+ | 8.93 (6.71–14.69) | 8.29 (5.75–12.79) | .15 |
| ȃCD8+CD38+ | 8.06 (6.77–10.25) | 7.66 (6.03–9.46) | .14 |
| ȃCD8+HLA-DR+ | 8.11 (7.25–9.43) | 7.16 (6.24–7.91) | <.001b |
Abbreviations: ART, antiretroviral therapy; BMI, body mass index; CVD, cardiovascular disease; HIV−, human immunodeficiency virus (HIV) uninfected; HOMA-IR, homeostatic assessment of insulin resistance; hsCRP, high-sensitivity C-reactive protein; IL-6, interleukin 6; IQR, interquartile range; LDL, low-density lipoprotein; NRTI, nucleoside reverse-transcriptase inhibitor; PHIV, perinatally acquired HIV; PWV, pulse wave velocity; sCD14 and sCD163, soluble CD14 and soluble CD163; sTNFRI, soluble tumor necrosis factor receptor I; VLDL, very low-density lipoprotein.
Data represent median values with IQRs unless otherwise identified as no. (%).
Significant at P < .05.
BMI calculated as weight in kilograms divided by height in meters squared.
Vascular Measures
IMT and PWV measurements at baseline and week 96 are highlighted in Figure 1. At baseline, PWV did not differ between groups (median [IQR], 5.73 [5.40–6.33] m/s for the PHIV and 6.02 [5.48–6.57] m/s for the HIV− group; P = .08). At week 96, PWV had increased in the PHIV group by a median of 0.3 m/s (to 6.03 [5.67–6.57] m/s; P = .01) but did not change significantly in the HIV− group (5.97 [5.57–6.40] m/s; P = .92).
Figure 1.
Distributions of pulse wave velocity (PWV) and maximum intima-media thickness (max IMT) at baseline and week 96 between human immunodeficiency virus (HIV)–uninfected (HIV−) and perinatally acquired HIV (PHIV) groups. The median PWV (interquartile range) at baseline was 5.73 (5.40–6.33) m/s in the PHIV and 6.02 (5.48–6.57) m/s in the HIV− group; the median PWV at week 96, 6.03 (5.67–6.57) and 5.97 (5.5– 6.40) m/s, respectively. The median max IMT (interquartile range) at baseline was 0.63 (0.59–0.68) mm in the PHIV and 0.60 (0.57–0.63) mm in the HIV− group; the median max IMT at week 96, 0.61 (0.58–0.64) and 0.62 (0.59–0.66) mm, respectively.
The median IMTs (IQRs) were slightly larger at baseline in participants with PHIV than in controls (mean IMT, 0.53 [0.49–0.55] vs 0.51 [0.47–0.54] mm, respectively [P = .06]; max IMT, 0.63 [0.59–0.68] vs 0.60 [0.57–0.63] mm [P < .01]). At 96 weeks, the median IMTs (IQRs) were decreased in the PHIV group (mean IMT, 0.52 [0.48–0.54] mm [P = .45]; max IMT, 0.61 [0.58–0.64] mm [P = .03]) and increased in the HIV− group (mean IMT, 0.52 [0.49–0.55] mm [P = .11]; max IMT, 0.62 [0.59–0.66] mm [P = .03]). There were no differences in IMT or PWV at week 96 between participants with PHIV who switched ART regimens during the study and those who remained on the same regimen (P = .61 and P = .58, respectively).
Factors Associated With Vascular Measures
In univariate analyses for all participants, age (r = 0.45), BMI (r = 0.28), and systolic (r = 0.29, and diastolic (r = 0.16) blood pressure were all correlated with higher PWV at 96 weeks (all P ≤ .04). Higher BMI was correlated with lower mean IMT and max IMT at 96 weeks (r = −0.22–0.26; P ≤ .004). on– high-density lipoprotein cholesterol was not correlated with PWV or IMT (P ≥ .08), and HOMA-IR was correlated with lower IMT (r = −0.19; P = .01).
In participants with PHIV, the higher CD4+ T-cell proportion at baseline was correlated with lower PWV at week 96 (r = 0.33; P = .002). Among inflammatory and immune markers, only higher activated CD8+ T-cell counts at baseline were correlated with higher IMT at 96 weeks in the PHIV groups (r = 0.25; P = .02). None of the inflammatory markers were correlated with any vascular measures in HIV− controls.
In the longitudinal analysis (see Table 2 and Figure 2A) for PWV, we found that the interaction term between HIV status and study duration (“time”) was not statistically significant at a 5% level. The 2 fitted lines crossed, however, indicating that there may be a potential interaction (P = .12). We further estimated the effects of HIV status at each time point by combining the coefficients of all relevant regression parameters. The results of a combination of coefficients at each study time point show that HIV status is not associated with PWV (−0.15 [95% confidence interval (CI), −.37 to .0607 m/s] [P = .15] for PHIV compared with HIV− group] at baseline and 0.07 m/s [95% CI, −.162 to .2957m/s] [P = .57] at week 96). The progression of PWV between the groups is in the opposite direction (Figure 2A). The main effects of HIV status did not have a significant association with PWV over the 96-week study period (−0.37 m/s [95% CI, −.817 to .06507 m/s]; P = .10) either. Age and BMI remained associated with faster PWV.
Table 2.
Longitudinal Analysis for Pulse Wave Velocity and Maximum Intima-Media Thickness in All Participantsa
| PWV | Max IMT | |||
|---|---|---|---|---|
| Estimate, β (95% CI) | P Value | Estimate, β (95% CI) | P Value | |
| Time | −.062 (−.265 to .139) | .54 | .008 (−.015 to .032) | .48 |
| HIV status (reference: HIV−) | −.376 (−.817 to .065) | .10 | .086 (.035–.137) | .001 |
| HIV status × time | .221 (−.596 to .502) | .12 | −.046 (−.079 to −.013) | .006 |
| Age (y)b | .115 (.071–.161) | <.001 | −.003 (−.007 to .002) | .24 |
| Female sexb | −.133 (−.305 to .038) | .13 | −.008 (−.026 to .009) | .36 |
| BMIb | .047 (.008–.085) | .02 | −.003 (−.007 to .001) | .17 |
Abbreviations: BMI, body mass index; CI, confidence interval; HIV, human immunodeficiency virus; HIV−, HIV uninfected; max IMT, mean-maximum intima-media thickness; PWV, pulse wave velocity.
Analysis performed using generalized estimating equations.
Age, sex, and BMI at baseline, with BMI calculated as weight in kilograms divided by height in meters squared.
Figure 2.
A, Predictive margins of average pulse wave velocity (PWV) in meters per second with 95% confidence intervals (CIs). PWV estimates are displayed by human immunodeficiency virus (HIV) serostatus over 96 weeks; estimates were derived from a generalized estimating equations (GEE) model with PWV as the outcome and a combination of coefficients (HIV status plus interaction term) as the predictor of interest. Bars represent 95% CIs. In the perinatally acquired HIV (PHIV) group, β = −.15 (95% CI, −.37 to .06; P = .15) at baseline and β = .07 m/s (95% CI, −.162 to .295; P = .57) at week 96. B, Predictive margins of mean-maximum intima-media thickness (max IMT), with 95% CIs. IMT regression analysis is displayed, with estimates derived from a GEE model with PWV as the outcome and a combination of coefficients (HIV status plus interaction term) as the predictor of interest. The PHIV group has approximately 0.04-mm (95% CI, .017–.063-mm; P < .01) higher max IMT at baseline, and no difference at week 96 (β = −.01 [95% CI, .031–.019]; P = .64). Abbreviation: HIV−, HIV uninfected.
Longitudinal analysis of max IMTs (Table 2 and Figure 2B) demonstrated that there is a statistically significant interaction between HIV status and time (P = .006). The 2 fitted lines for the 2 groups of participants slope in opposite directions. Further analysis revealed that HIV status was associated with larger max IMT at baseline, with approximately 0.04-mm (95% CI, .017–.063 mm; P < .01) higher max IMTs in the PHIV group compared with HIV− controls. HIV status, however, was not associated with IMT at week 96 (−0.01 mm [95% CI –.031 to .019 mm]; P = .64).
In analyses restricted to participants with PHIV, use of abacavir at the baseline visit was associated with slower PWV and higher IMT over 96 weeks (Table 3). Use of tenofovir at baseline was associated with lower IMT over 96 weeks, but use of PIs at baseline was not associated with either PWV or IMT. Lifetime exposure to thymidine analogues was not associated with IMT or PWV (P > .07), but longer ART duration was associated with lower PWV over 96 weeks (β = −.008 [95% CI, −.015 to −.002]; P = .01). In adjusted longitudinal models, the association between abacavir use and higher IMT over 96 weeks was dampened (Table 3).
Table 3.
Longitudinal Analysis for Pulse Wave Velocity and Maximum Intima-Media Thickness in Participants With Perinatally Acquired Human Immunodeficiency Virusa
| Exposure | Unadjusted Modelb | Adjusted Modelc | ||
|---|---|---|---|---|
| Estimate, β (95% CI) | P Value | Estimate, β (95% CI) | P Value | |
| PWV | ||||
| ȃAbacavir | −.281 (−.560 to −.002) | .048 | −.116 (−.435 to .202) | .47 |
| ȃTDF | .195 (−.062 to .452) | .14 | … | … |
| ȃPI s | .167 (−.138 to .473) | .28 | … | … |
| ȃINSTIs | .103 (−.182 to .387) | .48 | … | … |
| ȃThymidine exposure | .331 (−.035 to .697) | .08 | … | … |
| ȃART duration | −.008 (−.008 to .003) | .01 | … | … |
| Mean-maximum IMT | ||||
| ȃAbacavir | .043 (.012–.074) | .007 | .034 (−.002 to .071) | .07 |
| ȃTDF | −.035 (−.064 to −.006) | .02 | −.002 (−.040 to .037) | .93 |
| ȃPIs | −.002 (−.034 to .030) | .89 | … | … |
| ȃINSTIs | .017 (−.008 to .041) | .18 | … | … |
| ȃThymidine exposure | −.027 (−.068 to −.013) | .18 | … | … |
| ȃART duration | −.0001 (−.0009 to .0006) | .70 | … | … |
Abbreviations: ART, antiretroviral therapy; CI, confidence interval; IMT, intima-media thickness; INSTIs, integrase strand transfer inhibitors; PIs, protease inhibitors; PWV, pulse wave velocity; TDF, tenofovir.
Analysis performed using generalized estimating equations.
The model was run for each drug separately.
Adjusted for sex and for age (in years), body mass index (calculated as weight in kilograms divided by height in meters squared), and CD4+ T-cell count at baseline.
DISCUSSION
In this longitudinal cohort of virally suppressed youth with PHIV in Uganda, we did not find evidence of worsening in subclinical atherosclerosis over time. Our data provide evidence that prolonged ART may be protective against arterial stiffness over time. Interestingly, consistent with observations in adults with HIV, we found that abacavir use may increase the risk of subclinical atherosclerosis in youth.
Our findings are similar to those in the only other study assessing IMT longitudinally in PHIV youth [11]. In this study by Ross et al, performed in the United States, the participants were slightly younger than our, with a median age of 10 years; 80% were virally suppressed, and 50% were on abacavir and 47% on an NNRTI at baseline. The method used to measure the common carotid artery was similar to that used in our study, but the transducer was different. In that study, participants with PHIV had higher IMTs at baseline than HIV− controls, but at 48 weeks IMTs were similar between the groups. As in our Ugandan youth, the IMTs in those with PHIV decreased more than in the HIV− participants.
Unlike findings in US adults [12–15], which have highlighted greater IMT progression in adults with HIV, data from sub-Saharan Africa have shown no increase in subclinical atherosclerosis in adults with HIV compared with HIV− adults [16, 17] or a negative association between HIV infection and IMT [18]. Potential hypotheses for these findings among African adults with HIV include lower rates of established risk factors for CVD among African adults and increased access to primary care. IMT increases with age in European children [19], and we hypothesize that several factors may limit the natural progression of IMT in those with PHIV, including (1) treatment with cotrimoxazole, which mitigates mortality risk and improves health outcomes, likely owing to its anti-inflammatory properties [20], and (2) regular access to primary care for PHIV, which may mitigate HIV and socioeconomic risk factors that increase CVD [21].
We found that youth with PHIV have lower average PWV at baseline than their HIV− counterparts, but the velocity did not change significantly over time. Earlier cross-sectional studies have found conflicting results in PWV between HIV-infected and HIV− children or youth. These participants varied in age, viremia, and ART status from those in our study [22, 23]. To our knowledge, no studies in sub-Saharan Africa have compared arterial stiffness in children/youth with PHIV with that in HIV− controls, and no longitudinal studies have been performed in any settings among children/youth with PHIV. When comparing our findings to reference values for PWV in healthy European and North African children and teenagers, the median PWV for PHIV in our study falls between 90% and 95% at both time points, according to age, height, and sex [24]. In addition, age and height are closely correlated with PWV in healthy children, which we also found in our analyses. Therefore, in our cohort of young participants, a third of whom went through puberty during the study, it is unclear whether a trend toward an increase in PWV is associated with the same CVD risk seen in adults, in whom a 1-m/s increase in PWV conferred an increased risk of cardiovascular events, cardiovascular deaths, and all-cause deaths [25].
We found that prolonged ART duration was associated with lower arterial stiffness, which is consistent with findings in virally suppressed children in Mozambique [26] and youth in the United States [22]. In contrast to findings in Ethiopian children [27], we did not find any associations between NNRTIs or PIs and PWV or IMT [27]. Abacavir use was associated with higher IMT, but the strength and statistical significance of this association was dampened after adjustment for demographic variables and CD4+ T-cell counts. Abacavir use has been associated with an increased risk of CVD use in large observational studies among adults with HIV [28–30]. However, this risk was not substantiated in a US Food and Drug Administration meta-analysis, which demonstrated no short-term effect of abacavir on myocardial infarction among low-risk individuals [31].
Although we initially found a correlation between activated CD8+ T cells and IMT, this association was not significant after adjustment for known confounders. We did not find any association between subclinical vascular disease and other markers of systemic inflammation monocyte and T-cell activation. This may suggest that, unlike in adults, T-cell activation or inflammation may not necessarily be linked to complications [32, 33, 34].
Our study is strengthened by the longitudinal design, long-standing viral suppression in children/youth with PHIV, and age- and sex-matched comparators from the same geographic region. Our study is limited by the fact that there are no validation data for IMT and PWV in sub-Saharan Africa. Our ability to see an association between IMT and inflammatory markers may be due to assay variability and the small sample size. Owing to the design, we could not eliminate potential residual confounding. We focused on virally suppressed urban Ugandan youth with PHIV who have been enrolled in routine care, so our findings may not be applicable to all with PHIV. In addition, a little more half of our participants with PHIV had an ART regimen switch during the study period.
In summary, our understanding of the trajectory of cardiometabolic complications in PHIV is lacking. As African youth with PHIV are aging and entering adulthood, this population deserves special consideration and longitudinal studies of prevalence, incidence, and pathogenesis are needed to determine how CVD- and HIV-related risk factors will translate into CVD outcomes. Our study and the small, albeit growing, body of literature on cardiometabolic complications in this population highlight the findings that risk factors and pathogenesis differ from complications seen in adults with HIV and that early and prolonged ART may be protective.
Contributor Information
Sahera Dirajlal-Fargo, Department of Pediatrics, University Hospitals Cleveland Medical Center, Cleveland, Ohio, USA; Rainbow Babies and Children’s Hospital, Cleveland, Ohio, USA; Case Western Reserve University, Cleveland, Ohio, USA.
Chenya Zhao, Case Western Reserve University, Cleveland, Ohio, USA.
Danielle Labbato, Department of Pediatrics, University Hospitals Cleveland Medical Center, Cleveland, Ohio, USA.
Abdus Sattar, Case Western Reserve University, Cleveland, Ohio, USA.
Christine Karungi, Joint Clinical Research Centre, Kampala, Uganda.
Chris T Longenecker, University of Washington, Seattle, Washington, USA.
Rashidah Nazzinda, Joint Clinical Research Centre, Kampala, Uganda.
Nicholas Funderburg, Ohio State University School of Health and Rehabilitation Sciences, Columbus, Ohio, USA.
Cissy Kityo, Joint Clinical Research Centre, Kampala, Uganda.
Victor Musiime, Joint Clinical Research Centre, Kampala, Uganda; Makerere University, Kampala, Uganda.
Grace A McComsey, Department of Pediatrics, University Hospitals Cleveland Medical Center, Cleveland, Ohio, USA; Rainbow Babies and Children’s Hospital, Cleveland, Ohio, USA; Case Western Reserve University, Cleveland, Ohio, USA.
Notes
Acknowledgments. The authors thank the patients and their parents who participated in this research.
Author contributions. S. D. F. and G. A. M. designed the study and obtained funding. R. N., C. K., and V. M. oversaw study evaluations and monitoring. C. Z. and A. S. provided statistical support. N. F. performed the biomarker assays and flow cytometry. C. T. L. provided technical assistance with the vascular measures. D. L. performed all measurements of intima-media thickness. S. D. F. wrote the first draft of the manuscript. All authors contributed to data analysis and reviewed the manuscript for intellectual content.
Disclaimer. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the University Hospitals Cleveland Medical Center Clinical Research Center or the National Institutes of Health.
Financial support. This work was supported by the Eunice Kennedy Shriver National Institute of Child Health (grant K23HD088295–01A1 to S. D. F.); the National Institutes of Health (NIH) (grant K23HD088295-01A1 to S. D. F.’s institution); CTSC and the University Hospitals Cleveland Medical Center Clinical Research Center (grant UL1TR002548 to S. D. F.’s institution); the National Center for Advancing Translational Sciences, NIH, and the NIH Roadmap for Medical Research (grant UL1TR002548 to the University Hospitals Cleveland Medical Center Clinical Research Center and the Clinical and Translational Science Collaborative of Cleveland; and the National Heart Lung and Blood Institute (grant K23 HL123314 to C. T. L.).
References
- 1. The World Health Organization . Fact sheet HIV/AIDS. 2019. Available at: who.int/teams/global-hiv-hepatitis-and-stis-programmes/hiv/treatment/treatment-and-care-in-children-and-adolescents#:~:text=An%20estimated%201.8%20million%20children,testing%20and%20treatment%20is%20essential. Accessed 7 September 2022.
- 2. Centers for Disease Control and Prevention . HIV surveillance report. 2017. Available at: https://www.cdc.gov/hiv/library/reports/hiv-surveillance.html-. Accessed 7 September 2022.
- 3. Flynn PM, Abrams EJ. Growing up with perinatal HIV. AIDS 2019; 33:597–603. [DOI] [PubMed] [Google Scholar]
- 4. Gersh BJ, Sliwa K, Mayosi BM, Yusuf S. Novel therapeutic concepts: the epidemic of cardiovascular disease in the developing world: global implications. Eur Heart J 2010; 31:642–8. [DOI] [PubMed] [Google Scholar]
- 5. Hyle EP, Mayosi BM, Middelkoop K, et al. The association between HIV and atherosclerotic cardiovascular disease in sub-Saharan Africa: a systematic review. BMC Public Health 2017; 17:954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Dirajlal-Fargo S, Albar Z, Bowman E, et al. Subclinical vascular disease in children with human immunodeficiency virus in Uganda is associated with intestinal barrier dysfunction. Clin Infect Dis 2020; 71:3025–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Dirajlal-Fargo S, Albar Z, Bowman E, et al. Increased monocyte and T-cell activation in treated HIV+ Ugandan children: associations with gut alteration and HIV factors. AIDS 2020; 34:1009–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Williams PL, Abzug MJ, Jacobson DL, et al. Pubertal onset in children with perinatal HIV infection in the era of combination antiretroviral treatment. AIDS 2013; 27:1959–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Dalla Pozza R, Ehringer-Schetitska D, Fritsch P, Jokinen E, Petropoulos A, Oberhoffer R. Intima media thickness measurement in children: a statement from the Association for European Cardiology (AEPC) Working Group on Cardiovascular Prevention endorsed by the Association for European Paediatric Cardiology. Atherosclerosis 2015; 238:380–7. [DOI] [PubMed] [Google Scholar]
- 10. Ugandan Ministry of Health . Consolidated guidelines for the prevention and treatment of HIV and AIDS in Uganda. 2020. Available at: https://differentiatedservicedelivery.org/Portals/0/adam/Content/HvpzRP5yUUSdpCe2m0KMdQ/File/Uganda_Consolidated%20HIV%20and%20AIDS%20Guidelines%202020%20June%2030th.pdf-. Accessed 7 September 2022.
- 11. Ross AC, Storer N, Ann O’Riordan M, Dogra V, McComsey GA. Longitudinal changes in carotid intima-media thickness and cardiovascular risk factors in human immunodeficiency virus-infected children and young adults compared with healthy controls. Pediatr Infect Dis J 2010; 29:634–8. [DOI] [PubMed] [Google Scholar]
- 12. Hanna DB, Post WS, Deal JA, et al. HIV infection is associated with progression of subclinical carotid atherosclerosis. Clin Infect Dis 2015; 61:640–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Volpe GE, Tang AM, Polak JF, Mangili A, Skinner SC, Wanke CA. Progression of carotid intima-media thickness and coronary artery calcium over 6 years in an HIV-infected cohort. J Acquir Immune Defic Syndr 2013; 64:51–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hileman CO, Longenecker CT, Carman TL, McComsey GA. C-reactive protein predicts 96-week carotid intima media thickness progression in HIV-infected adults naive to antiretroviral therapy. J Acquir Immune Defic Syndr 2014; 65:340–4. [DOI] [PubMed] [Google Scholar]
- 15. Baker JV, Henry WK, Patel P, et al. Progression of carotid intima-media thickness in a contemporary human immunodeficiency virus cohort. Clin Infect Dis 2011; 53:826–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Mosepele M, Hemphill LC, Moloi W, et al. Pre-clinical carotid atherosclerosis and sCD163 among virally suppressed HIV patients in Botswana compared with uninfected controls. PLoS One 2017; 12:e0179994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Siedner MJ, Bibangambah P, Kim JH, et al. Treated HIV infection and progression of carotid atherosclerosis in rural Uganda: a prospective observational cohort study. J Am Heart Assoc 2021; 10:e019994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Nonterah EA, Boua PR, Klipstein-Grobusch K, et al. Classical cardiovascular risk factors and HIV are associated with carotid intima-media thickness in adults from sub-Saharan Africa: findings from H3Africa AWI-gen study. J Am Heart Assoc 2019; 8:e011506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Torkar AD, Plesnik E, Groselj U, Battelino T, Kotnik P. Carotid intima-media thickness in healthy children and adolescents: normative data and systematic literature review. Front Cardiovasc Med 2020; 7:597768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Bourke CD, Gough EK, Pimundu G, et al. Cotrimoxazole reduces systemic inflammation in HIV infection by altering the gut microbiome and immune activation. Sci Transl Med 2019; 11:486. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Miller GE, Chen E, Parker KJ. Psychological stress in childhood and susceptibility to the chronic diseases of aging: moving toward a model of behavioral and biological mechanisms. Psychol Bull 2011; 137:959–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Eckard AR, Raggi P, Ruff JH, et al. Arterial stiffness in HIV-infected youth and associations with HIV-related variables. Virulence 2017; 8:1265–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kuilder JS, Idris NS, Grobbee DE, et al. Association between human immunodeficiency virus infection and arterial stiffness in children. Eur J Prev Cardiol 2020; 24:480–8. [DOI] [PubMed] [Google Scholar]
- 24. Reusz GS, Cseprekal O, Temmar M, et al. Reference values of pulse wave velocity in healthy children and teenagers. Hypertension 2010; 56:217–24. [DOI] [PubMed] [Google Scholar]
- 25. Vlachopoulos C, Aznaouridis K, Stefanadis C. Prediction of cardiovascular events and all-cause mortality with arterial stiffness: a systematic review and meta-analysis. J Am Coll Cardiol 2010; 55:1318–27. [DOI] [PubMed] [Google Scholar]
- 26. Dobe IS, Mocumbi AO, Majid N, Ayele B, Browne SH, Innes S. Earlier antiretroviral initiation is independently associated with better arterial stiffness in children living with perinatally acquired HIV with sustained viral suppression in Mozambique. South Afr J HIV Med 2021; 22:1282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Gleason RL Jr, Caulk AW, Seifu D, et al. Efavirenz and ritonavir-boosted lopinavir use exhibited elevated markers of atherosclerosis across age groups in people living with HIV in Ethiopia. J Biomech 2016; 49:2584–92. [DOI] [PubMed] [Google Scholar]
- 28. Sabin CA, Worm SW, Weber R, 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; 371:1417–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Marcus JL, Neugebauer RS, Leyden WA, et al. Use of abacavir and risk of cardiovascular disease among HIV-infected individuals. J Acquir Immune Defic Syndr 2016; 71:413–9. [DOI] [PubMed] [Google Scholar]
- 30. Elion RA, Althoff KN, Zhang J, et al. Recent abacavir use increases risk of type 1 and type 2 myocardial infarctions among adults with HIV. J Acquir Immune Defic Syndr 2018; 78:62–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Ding X, Andraca-Carrera E, Cooper C, et al. No association of abacavir use with myocardial infarction: findings of an FDA meta-analysis. J Acquir Immune Defic Syndr 2012; 61:441–7. [DOI] [PubMed] [Google Scholar]
- 32. Sainz T, Alvarez-Fuente M, Navarro ML, et al. Subclinical atherosclerosis and markers of immune activation in HIV-infected children and adolescents: the CaroVIH Study. J Acquir Immune Defic Syndr 2014; 65:42–9. [DOI] [PubMed] [Google Scholar]
- 33. Kapetanovic S, Aaron L, Montepiedra G, Burchett SK, Kovacs A. T-cell activation and neurodevelopmental outcomes in perinatally HIV-infected children. AIDS 2012; 26:959–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Romeiro JR, Pinto JA, Silva ML, Eloi-Santos SM. Further evidence that the expression of CD38 and HLA-DR+ in CD8+ lymphocytes does not correlate to disease progression in HIV-1 vertically infected children. J Int Assoc Physicians AIDS Care (Chic) 2012; 11:164–8. [DOI] [PubMed] [Google Scholar]