Abstract
Background
Limited data exist describing metabolic syndrome (MetS) and Pathobiological Determinants of Atherosclerosis in Youth (PDAY) coronary arteries (CA) and abdominal aorta (AA) risk scores in youth with HIV in sub‐Saharan Africa.
Methods
This cross‐sectional analysis assessed MetS and PDAY CA and AA risk scores among youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV, and HIV‐seronegative youth. Elevated PDAY score was defined as ≥1. Cluster heat map analysis was used, and logistic regression models were fit to assess the association of HIV status with MetS and PDAY CA and AA risk scores separately after adjusting for covariates.
Results
We enrolled 237 youth with perinatally acquired HIV, 56 youth with non‐perinatally acquired HIV and 71 HIV‐seronegative youth; median (interquartile range = IQR) age was 18 (17, 20) years, 58% females. Youth with non‐perinatally acquired HIV had the highest proportion with MetS (34%), while HIV‐seronegative youth had 23%, and youth with perinatally acquired HIV 12%. Forty‐seven percent of youth with perinatally acquired HIV, 63% of youth with non‐perinatally acquired HIV and 41% of HIV‐seronegative youth had elevated PDAY CA score; 30% of youth with perinatally acquired HIV, 39% of youth with non‐perinatally acquired HIV and 23% of HIV‐seronegative youth had elevated PDAY AA score. A non‐overweight but hyperlipidaemic phenotype predominantly comprised of youth with perinatally acquired HIV was observed by cluster analysis. Youth with perinatally acquired HIV had lower adjusted odds of MetS compared with HIV‐seronegative youth (odds ratio = 0.35, 95% confidence interval: 0.16, 0.79) but HIV status (either youth with perinatally acquired HIV or youth with non‐perinatally acquired HIV vs. HIV‐seronegative) was not associated with an elevated PDAY CA or AA risk score.
Conclusion
Youth with perinatally acquired HIV have a lower odd of MetS, reflecting an overall non‐overweight, but hyperlipidaemic phenotype highlighting the need for further cardiometabolic research in this ageing population in South Africa.
Keywords: cardiometabolic, cardiovascular disease, metabolic syndrome, Pathobiological Determination of Atherosclerosis, perinatally acquired HIV
INTRODUCTION
The available knowledge about cardiometabolic complications in youth with perinatally acquired HIV is mainly from the Global North (United States and Europe). However, the vast majority of youth with perinatally acquired HIV reside in sub‐Saharan Africa, and focus on outcomes in this setting has been limited.
Metabolic syndrome (MetS) includes risk factors specific for cardiovascular disease (CVD), and the components include (1) abdominal obesity, (2) elevated triglycerides (TG), (3) low high‐density lipoprotein (HDL) cholesterol, (4) elevated blood pressure (BP), and (5) impaired fasting glycaemia (IFG) [1]. MetS is frequently described in adults living with HIV worldwide [2] and has been described in youth with HIV with varying prevalence ranging between 2% and 10% [3, 4], with most studies focused in Europe and North America.
Current cardiovascular disease risk scores have relied on adult cohorts for derivation, and generally estimate risk in adults aged 30 years and above [5]. Increasingly, data in non‐HIV populations have shown that the use of an aggregated CVD risk score such as the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) scoring system in adolescents and young adults may be useful in predicting important clinical outcomes, such as atherosclerotic cardiovascular disease later in life [6]. The PDAY risk score is based on post‐mortem measurement of atherosclerosis in the coronary arteries (CA) and abdominal aorta (AA), in 15‐ to 34‐year‐olds who died from accidental causes [7].
Studies such as the Randomized Trial to Prevent Vascular Events in HIV (REPRIEVE) trial [8] assessing interventions to prevent CVD in adults with HIV, where it is easier to define hard outcomes such as myocardial infarction or stroke, have moved forward.
Development of paediatric and young adult studies has however been hampered by a lack of agreement on clinically significant and relevant outcomes. There are few CVD prevention trials in young adults and no quality evidence for CVD prevention at a young age [9], despite the fact that targeting therapy early may be the best way to reduce CVD risk at a later age. Identifying young adults and adolescents with HIV at high risk of CVD may help to identify a population for prevention trials [10].
We have previously highlighted that youth with perinatally acquired HIV, in the Cape Town Adolescents Antiretroviral Cohort (CTAAC), have PDAY scores that reflect increased aggregate atherosclerotic risk [11].
The objective of this study was to describe the risk of CVD among youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV on antiretroviral therapy (ART), and seronegative youth using MetS and PDAY scores in the CTAAC‐Heart study, and to investigate HIV and demographic factors associated with the scores.
METHODS
Study population
The CTAAC‐Heart study is an observational study which enrolled youth with perinatally acquired HIV from the parent CTAAC, youth with non‐perinatally acquired HIV and HIV‐seronegative youth Participants were between 12–24 years of age, and enrolled from October 2020 to August 2022. Study design and methods have been previously described [12]. During this period, the dolutegravir (DTG) roll‐out was not fully in place. All youth with perinatally acquired HIV in CTAAC‐Heart study were on ART for ≥6 months, and youth with non‐perinatally acquired HIV for ≥1 month prior to enrolment. Exclusion criteria included pregnant individuals, those with congenital heart disease or pre‐existing renal disease (glomerular filtration rate < 30 mL/min), as well as those with known in utero HIV exposure among HIV‐seronegative youth.
Patient consent statement
All participants provided written informed consent and/or assent as appropriate. This study was approved by the Institutional Review Boards of the University of Cape Town, Northwestern University Feinberg School of Medicine, and the Ann & Robert H. Lurie Children's Hospital of Chicago.
Primary exposure of interest
The primary exposure of interest in this cross‐sectional analysis was HIV status and mode of acquisition (perinatal vs. non‐perinatal). HIV status was confirmed through previously collected data in the parent CTAAC study (perinatal HIV acquisition), medical record review (non‐perinatal HIV acquisition), and repeated HIV testing (HIV‐seronegativity).
Cardiometabolic outcomes
Metabolic
Metabolic risk factors measured were: body mass index (BMI), waist/height ratio, fasting lipids [high density lipoprotein (HDL), low density lipoprotein (LDL), total cholesterol (TC)/HDL ratio, triglycerides] and homeostatic model assessment of insulin resistance (HOMA‐IR). Insulin resistance was defined as HOMA‐IR >3.16 [13, 14]. MetS was defined based on the International Diabetes Foundation (IDF) criteria [15]: abdominal obesity (waist circumference ≥90th percentile for 10–<16 years, and ≥94 cm or ≥80 cm for males and females ≥16 years, respectively [16, 17]), elevated TG (≥150 mg/dL), low HDL cholesterol (<40 mg/dL for 10–<16 years for both sexes and males ≥16 years, and < 50 mg/dL for females ≥16 years), high BP (systolic BP ≥130 mm Hg or diastolic BP ≥85 mm Hg), and raised fasting glucose (≥100 mg/dL). Blood samples were collected after a >8‐h fast.
Cardiovascular
PDAY CA and AA scores were calculated separately by summing variables associated with atherosclerosis as described previously [18]. Components of the PDAY score were defined as follows: non‐HDL cholesterol ≥130 mg/dL, HDL <40 mg/dL, hyperglycaemia (fasting plasma glucose ≥126 mg/dL), hypertension (BP ≥95th percentile for age, sex and height), obesity (BMI > 30 kg/m2), and cigarette smoking (>1 pack/day in the past 3 months). Elevated PDAY score was defined as ≥1.
Covariates
Clinical and sociodemographic variables were collected through record review and questionnaires. Anthropometric and blood pressure measurements were performed by trained research staff in a standardized fashion using calibrated scales, stadiometers, measuring tapes and sphygmomanometers.
Statistical analysis
Continuous variables were tested for normality using the Shapiro–Wilk test. For normally distributed variables, data were reported as means with standard deviations (SD), while medians with interquartile ranges (IQR) were used for non‐normally distributed variables. Categorical variables were summarized as frequencies and percentages. Comparisons of continuous variables were conducted using Kruskal–Wallis tests between youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV and HIV‐seronegative youth. Categorical outcomes were compared using Chi‐square or Fisher's exact tests, as appropriate. Hierarchical cluster analysis with heat maps was performed to visualize multivariate data and identify if potential subgroups with similar phenotypes emerged across the entire cohort. Logistic regression models were fit to assess the association of HIV status (youth with perinatally acquired HIV vs. youth with non‐perinatally acquired HIV vs. HIV‐seronegative) with MetS and with elevated PDAY CA and AA scores separately after adjusting for covariates determined a priori. We also conducted subgroup analyses in youth with HIV only to explore the association of integrase inhibitor (INSTI)‐based ART with elevated PDAY scores. In addition, among youth with HIV, we explored whether the association between INSTI and odds of elevated PDAY scores was different for males and females by including an interaction term for INSTI*female sex. Separate models were also fit stratified by sex. All statistical analyses were conducted in Stata, V18 (Statacorp®, College Station, TX, USA) and assumed a two‐sided type I error rate of 0.05.
RESULTS
Baseline characteristics
A total of 364 participants were included: 237 youth with perinatally acquired HIV, 56 youth with non‐perinatally acquired HIV and 71 seronegative youth. The median (IQR) age was 18 (17–20) years, and 58% were females, Table 1. HIV‐seronegative youths were slightly younger compared with those living with HIV, and there was a higher proportion of women in the youth with non‐perinatally acquired HIV. Eighty percent of participants were post‐pubertal. CVD risk factors including family history of CVD and tobacco use were similar between the groups. YPHIV had a higher proportion with a reported history of TB disease compared with youth with non‐perinatally acquired HIV and HIV‐seronegative youth (71% vs. 20% and 7%).
TABLE 1.
Characteristic of participants.
| Perinatally acquired HIV (N = 237) | Non‐perinatally acquired HIV (N = 56) | HIV‐seronegative (N = 71) | p Value | |
|---|---|---|---|---|
| Sociodemographic | ||||
| Age, year, median (IQR) | 18.66 (17.32–20.01) | 21.12 (19.23–22.32) | 17.19 (16.11–18.95) | <0.001 |
| Female, n (%) | 127 (53.59) | 48 (85.71) | 38 (53.52) | |
| Home running water, n (%) | 209 (88.19) | 38 (82.61) | 41 (57.75) | <0.001 |
| Family history | ||||
| Family history of cardiovascular disease, n (%) | 6 (2.65) | 0 (0.00) | 2 (2.82) | 0.55 |
| Family history of hypertension, n (%) | ||||
| Yes | 142 (62.83) | 17 (39.53) | 17 (23.94) | <0.001 |
| Family history of hyperlipidaemia, n (%) | 29 (12.24) | 11 (20.00) | 7 (9.86) | 0.21 |
| Tobacco use, n (%) | 29 (12.24) | 11 (20.00) | 7 (9.86) | 0.21 |
| History of tuberculosis, n (%) | 168 (70.89) | 11 (20.00) | 4 (7.27) | |
| Tanner stage ≥ 4 a , n (%) | 190 (97.94) | 54 (100.00) | 48 (94.12) | |
| Physical activity a , n (%) | ||||
| High | 20 (8.51) | 1 (1.79) | 15 (21.43) | 0.001 |
| Moderate | 32 (13.62) | 5 (8.93) | 12 (17.14) | |
| Low | 183 (77.87) | 50 (89.29) | 43 (61.43) | |
| HIV‐specific factors | ||||
| CD4, cells/mm3 a , n (%) | ||||
| <200 | 28 (11.97) | 3 (5.66) | ‐‐‐ | 0.41 |
| 200–499 | 80 (34.19) | 19 (35.85) | ‐‐‐ | |
| >500 | 126 (53.85) | 31 (58.49) | ‐‐‐ | |
| Viral load result (≤50 copies/mL) | 138 (58.97) | 46 (86.79) | <0.001 | |
| Viral load result (>50 copies/mL) | 96 (41.03) | 7 (13.21) | ‐‐‐ | |
| Age at ART initiation, years | 3.99 (1.81–6.98) | 18.78 (16.86–20.99) | ‐‐‐ | <0.001 |
| Current INSTI‐based ART | 79 (33.33) | 48 (85.71) | ‐‐‐ | <0.001 |
| Regiment categories, n (%) | ||||
| 2 NRTI + INSTI | 75 (31.65) | 48 (85.71) | ‐‐‐ | <0.001 |
| 2 NRTI + NNRTI | 64 (27.00) | 8 (14.29) | ‐‐‐ | |
| 2 NRTI + PI | 93 (39.24) | 0 (0.00) | ‐‐‐ | |
| Other | 5 (2.11) | 0 (0.00) | ‐‐‐ | |
Abbreviations: ART, antiretroviral therapy; INSTI, integrase strand transfer inhibitor; IQR, interquartile range; n (%), number in category (percentage); NNRTI, non‐nucleotide reverse transcriptase inhibitor; NRTI, nucleotide reverse transcriptase inhibitor; PI, protease inhibitor.
Numbers in these cells do not total the denominators due to missing data.
Among youth with perinatally acquired HIV versus youth with non‐perinatally acquired HIV, respectively: 59% versus 87% had viral suppression (<50 copies/mL), 53% versus 58% had CD4 >500 cells/mm3, and 33% versus 86% were on an INSTI‐based ART.
Metabolic risk factors
Youth with perinatally acquired HIV had the lowest BMI (21.7 vs. 26.1 and 22.3 kg/m2) and waist/height ratio (0.45 vs. 0.50 and 0.47) but the highest HOMA (2.64 vs. 2.36 and 2.14) and triglycerides (65 vs. 63 and 55 mg/dL) compared with youth with non‐perinatally acquired HIV and HIV‐seronegative youth. BMI was highest in the youth with non‐perinatally acquired HIV group, and 27% of youth with perinatally acquired HIV, 59% of youth with non‐perinatally acquired HIV, and 36% of the HIV‐seronegative participants were overweight or obese (Table 2). Similar trends were seen for participants below 19 years for BMI‐for‐age z‐scores (Table S2). HOMA (p < 0.001) and LDL (p = 0.035) were highest among obese participants (Table S1).
TABLE 2.
Frequencies of MetS components and cardiometabolic risk factors by HIV status.
| Perinatally acquired HIV (N = 237) | Non‐perinatally acquired HIV (N = 56) | HIV‐seronegative (N = 71) | p Value | |
|---|---|---|---|---|
| MetS components | ||||
| Central obesity, n (%) | ||||
| No | 211 (89.03) | 37 (66.07) | 55 (77.46) | <0.001 |
| Yes | 26 (10.97) | 19 (33.93) | 16 (22.54) | |
| Raised TG, n (%) | ||||
| No | 230 (98.71) | 56 (100.00) | 70 (98.59) | 0.69 |
| Yes | 3 (1.29) | 0 (0.00) | 1 (1.41) | |
| Reduced HDL, n (%) | ||||
| No | 131 (56.22) | 20 (35.71) | 37 (52.11) | 0.022 |
| Yes | 102 (43.78) | 36 (64.29) | 34 (47.89) | |
| Raised BP, n (%) | ||||
| No | 228 (96.61) | 51 (91.07) | 69 (97.18) | 0.14 |
| Yes | 8 (3.39) | 5 (8.93) | 2 (2.82) | |
| Raised fasting glucose, n (%) | ||||
| No | 231 (97.47) | 56 (100.00) | 68 (97.14) | 0.47 |
| Yes | 6 (2.53) | 0 (0.00) | 2 (2.86) | |
| Proportion of participants meeting MetS, n (%) | ||||
| No | 206 (87.66) | 37 (66.07) | 55 (77.46) | <0.001 |
| Yes | 29 (12.34) | 19 (33.93) | 16 (22.54) | |
| Cardiometabolic risk factors | ||||
| BMI, median (IQR) | 21.73 (19.63–25.19) | 26.10 (22.33–30.81) | 22.32 (19.75–28.68) | <0.001 |
| BMI categories, n (%) | ||||
| BMI < 25 | 173 (73.00) | 23 (41.07) | 45 (63.38) | <0.001 |
| BMI ≥ 25 | 64 (27.00) | 33 (58.93) | 26 (36.62) | |
| TG, median (IQR) | 64.60 (51.33–84.07) | 62.39 (45.58–80.97) | 54.87 (45.13–67.26) | <0.001 |
| TC, median (IQR) | 140.15 (116.60–159.85) | 123.75 (112.74–146.33) | 130.89 (110.42–155.98) | 0.038 |
| LDL, median (IQR) | 76.83 (61.00–93.82) | 66.41 (53.47–84.56) | 70.66 (54.44–91.12) | 0.058 |
| HOMA‐IR, n (%) | ||||
| No | 135 (60.81) | 36 (64.29) | 42 (67.74) | 0.59 |
| Yes | 87 (39.19) | 20 (35.71) | 20 (32.26) | |
Note: MetS was defined based on the International Diabetes Foundation (IDF) criteria [15]: abdominal obesity (waist circumference ≥90th percentile for 10–<16 years and ≥94 cm or ≥80 cm for males and females ≥16 years, respectively [16, 17]), elevated TG (≥150 mg/dL), low HDL cholesterol (<40 mg/dL for 10–<16 years for both sexes and males ≥16 years, and <50 mg/dL for females ≥16 years), high BP (systolic BP ≥130 mm Hg or diastolic BP ≥85 mm Hg), and raised fasting glucose (≥100 mg/dL). Blood samples were collected after a >8‐h fast.
Abbreviations: BMI, body mass index; BP, blood pressure; HDL, high‐density lipoprotein; HOMA‐IR, homeostatic model assessment of insulin resistance; IQR, interquartile range; LDL, low density lipoprotein; MetS, metabolic syndrome; TC, total cholesterol; TG, triglycerides.
Frequencies of MetS components by HIV status are listed in Table 2. Proportion of participants meeting the definition for MetS included 12% in youth with perinatally acquired HIV; 34% in youth with non‐perinatally acquired HIV; and 23% in HIV‐seronegative youth. Youth with non‐perinatally acquired HIV were more likely to have central obesity and reduced HDL (p ≤ 0.022).
In both adjusted and unadjusted analyses, the odds of MetS were lower in youth with perinatally acquired HIV when compared with either HIV‐seronegative youth or youth with non‐perinatally acquired HIV. Odds were not different between youth with non‐perinatally acquired HIV and HIV‐seronegative youth (Table 3).
TABLE 3.
Adjusted odds ratios of MetS.
| Odds ratio | 95% CI | |
|---|---|---|
| Entire cohort a | ||
| HIV status | ||
| PHIV | 0.351* | [0.156, 0.789] |
| NPHIV | 0.950 | [0.346, 2.610] |
| HIV‐seronegative | Ref | |
| Model (youth with HIV) b | ||
| HIV status | ||
| PHIV | 0.137* | [0.019, 0.991] |
| NPHIV | Ref | |
Abbreviations: CI, confidence interval; MetS, metabolic syndrome; NPHIV, non‐perinatally acquired HIV; PHIV, perinatally acquired HIV.
Model 1: Adjusted for age, sex, access to tap water and family history of cardiometabolic disease.
Model 2: Adjusted for age, sex, access to tap water and family history of cardiometabolic disease, viral load >50 copies/mL, duration on ART and INSTI use.
p < 0.05.
Factors associated with metabolic risk factors
Hierarchical cluster analyses were used to visualize relationships between HIV status and metabolic risk factors (Figure 1). Three distinct clusters were identified: Cluster 1: a mix of youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV and HIV‐seronegative youth with normal metabolic characteristics; Cluster 2: a mix of youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV and HIV‐seronegative youth with high BMI, waist/height ratio, lipids and HOMA; and Cluster 3: primarily composed of youth with perinatally acquired HIV with low BMI and waist/height ratio as well as high triglycerides and TC:HDL ratio.
FIGURE 1.
Heat map of hierarchical clustering by metabolic risk factors and HIV status. Hierarchical cluster analysis with heat mapping was performed to visualize data on cardiometabolic risk factors (HDL, BMI, Waist:height ratio, LDL, Chol:HDL, HOMA, and triglycerides) and identify if potential subgroups with similar phenotypes of these risk factors emerged across the entire cohort. Three distinct clusters were identified: Cluster 1: Mix of YPHIV, YNPHIV and HIV‐seronegative youth with normal metabolic characteristics. Cluster 2: Mix of YPHIV, YNPHIV and HIV‐seronegative youth with high BMI, waist/height ratio, lipids and HOMA. Cluster 3: Primarily composed of YPHIV with low BMI and waist/height ratio as well as high triglycerides and TC:HDL ratio. BMI, body mass index; HDL, high‐density lipoprotein; HOMA, homeostatic model assessment; LDL, low density lipoprotein; TC, total cholesterol; YNPHIV, youth with non‐perinatally acquired HIV; YPHIV, youth with perinatally acquired HIV.
In regression models (Table 3), youth with perinatally acquired HIV had lower odds of MetS compared with HIV‐seronegative youth [aOR = 0.35 (95% CI 0.16, 0.79)] and females had a higher odds of MetS [aOR = 13.80 (95% CI 4.79, 39.66)].
Cardiovascular disease risk and PDAY
PDAY characteristics by HIV status for the subgroup with PDAY score ≥1 are highlighted in Table 4 and overall in Table S3. Overall, 48% of participants had a PDAY CA score ≥1 and 30% had a PDAY AA score ≥1. Participants with an elevated PDAY CA score included: 47% of youth with perinatally acquired HIV, 62% youth with non‐perinatally acquired HIV and 41% of seronegative youth. Participants with an elevated PDAY AA score included: 30% of youth with perinatally acquired HIV, 39% of youth with non‐perinatally acquired HIV and 22% of HIV‐seronegative youth. Overall, elevated scores were associated primarily with low levels of HDL cholesterol (PDAY CA score only), smoking, hypertension and high obesity rates among females. PDAY components for the subgroup with PDAY ≥1 were not significantly different between groups except for females with BMI >30 in the youth with nonperinatally acquired HIV group (Table 4).
TABLE 4.
PDAY components by HIV status for those with PDAY ≥1.
| Perinatally acquired HIV | Non‐perinatally acquired HIV | HIV‐seronegative | p Value | |
|---|---|---|---|---|
| Coronary artery PDAY | ||||
| Elevated coronary artery PDAY score (≥1) | N = 112 | N = 35 | N = 29 | |
| Non‐HDL cholesterol, mg/dL, n (%) | ||||
| <130 | 88 (79.28) | 32 (91.43) | 27 (93.10) | 0.38 |
| 130–159 | 19 (17.12) | 3 (8.57) | 1 (3.45) | |
| 160–189 | 3 (2.70) | 0 (0.00) | 1 (3.45) | |
| >220 | 1 (0.90) | 0 (0.00) | 0 (0.00) | |
| HDL cholesterol, mg/dL, n (%) | ||||
| <40 | 67 (60.36) | 21 (60.00) | 16 (55.17) | 0.61 |
| 40–59 | 40 (36.04) | 13 (37.14) | 10 (34.48) | |
| ≥60 | 4 (3.60) | 1 (2.86) | 3 (10.34) | |
| Smoking | ||||
| No | 89 (79.46) | 23 (67.65) | 23 (79.31) | 0.34 |
| Yes | 23 (20.54) | 11 (32.35) | 6 (20.69) | |
| Blood pressure, mm Hg, n (%) | ||||
| Not hypertensive | 83 (74.11) | 23 (65.71) | 19 (65.52) | 0.49 |
| Hypertensive (systolic > 120|diastolic > 80) | 29 (25.89) | 12 (34.29) | 10 (34.48) | |
| Body mass index (obesity), kg/m2, n (%) | ||||
| Male, n (%) | ||||
| BMI ≤ 30 | 49 (94.23) | 6 (100.00) | 15 (88.24) | 0.547 |
| BMI > 30 | 3 (5.77) | 0 (0.00) | 2 (11.76) | |
| Female, n (%) | ||||
| BMI ≤ 30 | 48 (80.00) | 13 (44.83) | 7 (58.33) | 0.003 |
| BMI > 30 | 12 (20.00) | 16 (55.17) | 5 (41.67) | |
| Hyperglycaemia, mg/dL | ||||
| ≥126 mg/dL | 0 | 0 | 0 | |
| Abdominal aorta PDAY | ||||
| Elevated abdominal aorta PDAY score (≥1) | N = 71 | N = 22 | N = 16 | |
| Non‐HDL cholesterol, mg/dL, n (%) | ||||
| <130 | 47 (67.14) | 19 (86.36) | 14 (87.50) | 0.38 |
| 130–159 | 19 (27.14) | 3 (13.64) | 1 (6.25) | |
| 160–189 | 3 (4.29) | 0 (0.00) | 1 (6.25) | |
| >220 | 1 (1.43) | 0 (0.00) | 0 (0.00) | |
| HDL cholesterol, mg/dL, n (%) | ||||
| <40 | 21 (30.00) | 8 (36.36) | 3 (18.75) | 0.59 |
| 40–59 | 39 (55.71) | 13 (59.09) | 10 (62.50) | |
| ≥60 | 10 (14.29) | 1 (4.55) | 3 (18.75) | |
| Smoking, n (%) | ||||
| No | 42 (59.15) | 11 (50.00) | 9 (56.25) | 0.75 |
| Yes | 29 (40.85) | 11 (50.00) | 7 (43.75) | |
| Blood pressure, mm Hg, n (%) | ||||
| Not hypertensive | 42 (59.15) | 10 (45.45) | 6 (37.50) | 0.21 |
| Hypertensive (systolic > 120|diastolic > 80) | 29 (40.85) | 12 (54.55) | 10 (62.50) | |
| Body mass index (obesity), kg/m2, n (%) | ||||
| Male | ||||
| BMI ≤ 30 | 41 (100.00) | 5 (100.00) | 13 (92.86) | 0.317 |
| BMI > 30 | 0 (0.00) | 0 (0.00) | 1 (7.14) | |
| Female | ||||
| BMI ≤ 30 | 24 (80.00) | 5 (29.41) | 1 (50.00) | 0.001 |
| BMI > 30 | 6 (20.00) | 12 (70.59) | 1 (50.00) | |
| Hyperglycaemia, mg/dL, n (%) | ||||
| ≥126 mg/dL | 0 | 0 | 0 | |
Abbreviations: BMI, body mass index; HDL, high‐density lipoprotein; PDAY, Pathobiological Determinants of Atherosclerosis in Youth.
Factors associated with cardiovascular disease risk factors and PDAY
Adjusted odds ratios (aOR) for PDAY are highlighted in Table 5. HIV status was not associated with an elevated CA [aOR 1.050 (95% CI 0.572, 1.927) for youth with perinatally acquired HIV vs. HIV‐seronegative and aOR 2.285 (95% CI 0.934, 5.587) for youth with non‐perinatally acquired HIV vs. HIV‐seronegative] or PDAY AA score [aOR 1.045 (95% CI 0.509, 2.143) for youth with perinatally acquired HIV vs. HIV‐seronegative and aOR 1.752 (95% CI 0.655, 4.682) for youth with non‐perinatally acquired HIV vs. HIV‐seronegative]. In youth with HIV, INSTI was associated with higher odds of elevated CA [aOR 2.287 (95% CI 1.308, 4)] and AA [aOR 2.001 (95% CI 1.092, 3.667)] PDAY score. In sex‐stratified analyses, the association between INSTI use and elevated PDAY CA score persisted in females [aOR 2.729 (95% CI 1.323, 5.629)] but not in males [aOR 2.006 (95% CI 0.783, 5.135)], and not for PDAY AA (Table 5). However, we found that the interaction term between sex and being on INSTI in females among youth living with HIV was not significant [aOR 1.763 (95% CI 0.612, 5.078)].
TABLE 5.
Adjusted odds ratios for factors associated with elevated in PDAY score (≥1).
| Exposure of interest | PDAY score CA | PDAY score AA | ||
|---|---|---|---|---|
| aOR | CI | aOR | CI | |
| Model (entire cohort) a | ||||
| YPHIV | 1.05 | [0.57, 1.93] | 1.04 | [0.501, 2.14] |
| YNPHIV | 2.28 | [0.93, 5.97] | 1.75 | [0.65, 4.68] |
| HIV‐seronegative | Ref | ‐‐ | Ref | ‐‐ |
| Model (youth with HIV) b | ||||
| On INSTI | 2.29** | [1.31, 4] | 2.00* | [1.09, 3.67] |
| Model (female youth with HIV) c | ||||
| On INSTI | 2.723** | [1.32, 5.623] | 2.02 | [0.88, 4.63] |
| Model (male youth with HIV) c | ||||
| On INSTI | 2.01 | [0.78, 5.13] | 2.14 | [0.83, 5.48] |
| Model (youth with HIV) c | ||||
| On INSTI | 1.63 | [0.701, 3.77] | 1.85 | [0.78, 4.40] |
| On INSTI female sex* | 1.76 | [0.61, 5.08] | 1.15 | [0.37, 3.57] |
Model 1: Adjusted for age, sex, access to tap water and family history of cardiometabolic disease.
Model 2: Adjusted for age, sex, access to tap water and family history of cardiometabolic disease, viral load > 50 copies/mL and duration on ART.
Models 3, 4 and 5: Adjusted for age, access to tap water, family history of cardiometabolic disease, viral load > 50 copies/mL and duration on ART.
Abbreviations: AA, abdominal aorta; aOR, adjusted odds ratio; CA, coronary arteries; CI, confidence interval; INSTI, integrase strand transfer inhibitor; PDAY, Pathobiological Determinants of Atherosclerosis in Youth; YNPHIV, youth with non‐perinatally acquired HIV; YPHIV, youth with perinatally acquired HIV.
p < 0.05;
p < 0.01.
DISCUSSION
Our study adds important insights into the possible long‐term risk of CVD in youth with and without perinatally acquired HIV in sub‐Saharan Africa. This is the first study to examine MetS and PDAY score in youth with perinatally acquired HIV, youth with non‐perinatally acquired HIV and HIV‐seronegative youth in South Africa. There appears to be a non‐overweight but hyperlipidaemic and metabolically unhealthy phenotype among youth with perinatally acquired HIV in South Africa.
We observed similar rates of MetS in youth with perinatally acquired HIV in our study (12%) to other studies in European and US cohorts (1%–13%) youth with perinatally acquired HIV [3, 4]. However, both youth with non‐perinatally acquired HIV (34%) and HIV‐seronegative youth (23%) had higher rates of MetS than those previously reported in African youth [19, 20]. Lower rates of central obesity in youth with perinatally acquired HIV likely contribute to the lower rates of MetS in this population, as central obesity is a key criteria for MetS. Furthermore, MetS in children is not well defined, and there is no current consensus on its definition. For example, in the IDF criteria, the US definition of elevated fasting glucose is used for IFG, rather than the European definition of elevated fasting insulin. One study describes the MetS prevalence in South African adolescents ranging from 3% to 6% [21]. A meta‐analysis of 297 studies reported the overall prevalence of MetS in African children was 13% [19] and a worldwide review of 36 studies estimates the paediatric prevalence to range from 1% to 22% [22].
Use of different definitions may lead to different classifications for MetS, and focusing on cardiometabolic risk factor clustering may be more clinically meaningful than MetS [23]. We found that the risk of MetS was not different between youth with non‐perinatally acquired HIV and seronegative youth, while the risk was lower in youth with perinatally acquired HIV compared with seronegative youth. Among youth with perinatally acquired HIV, a low BMI and waist/height ratio with dyslipidaemia profile observed in our study is similar to the ‘thin and hypercholesterolaemic’ phenotype reported in previous youth with perinatally acquired HIV studies [4]. However, overall this is a cohort with low risk of dyslipidaemia, and although youth with perinatally acquired HIV have higher lipid levels compared with the other groups, the values for total cholesterol, triglycerides and LDLs are within the normal range as highlighted in Table 2. Central obesity appears to be the major risk factor for MetS overall for all groups.
In addition, we found that female sex, regardless of HIV status, was associated with increased odds of MetS, consistent with previous African studies [24]. Possible aetiologies include a high prevalence of obesity among African women, especially in the urban setting (26% of females in our study compared with 3% of males had a BMI ≥30), low levels of adiponectin reported in African women, a higher pro‐inflammatory environment that has been described in females with HIV in sub‐Saharan Africa, and decreased physical activity reported in South African women [25, 26].
The PDAY risk score measured in adolescence has been shown to predict future carotid intima media thickness measured in adulthood in several large studies in HIV‐seronegative youth, including the Cardiovascular Risk in Young Finns study [27]. In the Coronary Artery Risk Development in Young Adults (CARDIA) study, PDAY risk score, measured in over 3000 HIV‐seronegative youth between 18 and 30 years, was strongly associated with the presence and intensity of subclinical coronary atherosclerosis measured 25 years later by coronary artery calcium on CT scan [28], with 53% of participants having an elevated PDAY AA and 28% having an elevated PDAY CA (28%) at 25 years of follow‐up [18]. Contrary to CARDIA, we found that overall participants had a higher prevalence of elevated PDAY CA score (48%) versus AA score (30%). We have previously shown in CTAAC, lower rates of PDAY CA (30%) and AA (18%) scores in youth with perinatally acquired HIV [11] compared with data in this current manuscript (47% for CA and 30% for AA in youth with perinatally acquired HIV). The higher prevalence of PDAY score may be because of the older age of the participants in the current study (17–20 compared with 15–18 years old) with almost all participants post‐pubertal (80%). In addition, more participants in the current study met the definition for hypertension (6% in younger CTAAC vs. 26% in this study) and more females were obese (16% in younger CTAAC vs. 20% in this study). In the Adolescent Master Protocol study of the Paediatric HIV/AIDS Cohort Study (PHACS), youth with perinatally acquired HIV in the United States also had a prevalence of elevated PDAY CA score almost double that of an elevated PDAY AA (48.5% for CA and 23.6% for AA) [29], which is similar to our study (47% for CA and 30% for AA in youth with perinatally acquired HIV). Only one other study has assessed PDAY in South African youth with perinatally acquired HIV; the combined data from the Children with HIV Early antiRetroviral (CHER) and International Maternal Paediatric Adolescent AIDS Clinical Trials Network P1060 clinical trials (median age 16 years) highlighted high rates for elevated PDAY CA (55%) and AA (59%) scores in youth with perinatally acquired HIV [30]. Similarly to our findings, PDAY score in the PHACS and CHER studies was attributed to low levels of HDL cholesterol. Cardiovascular risk prediction functions are known to underestimate CVD risk in adults living with HIV [31, 32], and this may also be the case in youth with HIV. This may be due to a discordance between systemic biomarkers included in risk factor‐based scores and anatomy, as people living with HIV have accelerated development of fatty streaks and atherosclerosis in arteries.
Boosted protease inhibitors were a strong predictor of PDAY CA and AA in PHACS. We initially found that INSTI was a predictor of PDAY scores in females; however, when including an interaction term between sex and current INSTI use, the relationship between PDAY scores and INSTI no longer remained, suggesting that the effect of INSTI on PDAY scores is not dependent on sex and our initial findings are likely secondary to sample size variations in sex between the groups. These are exploratory analyses and our study may not be adequately powered to detect interactions. INSTIs have not been consistently found to increase the risk of CVD [33], although some data are conflicting, and further investigations are warranted in youth with perinatally acquired HIV in sub‐Saharan Africa who have recently transitioned to INSTIs.
We are limited by the cross‐sectional nature of our analysis and small sample size for some of the groups, especially the HIV‐seronegative youth which may have hampered our ability to investigate cardiometabolic health differences. There was a larger proportion of females in youth with non‐perinatally acquired HIV compared with youth with perinatally acquired HIV and HIV‐seronegative youth which reflects the current HIV epidemiology in South Africa. To account for this, we adjusted for females in both MetS and PDAY models, performed sensitivity analyses restricted to females, and included the interaction term INSTI*female to specifically test if the relationship between PDAY and INSTI was dependent on sex. We did not control for multiple comparisons, so results should be interpreted cautiously.
In conclusion, youth with perinatally acquired HIV and youth with non‐perinatally acquired HIV on antiretroviral therapy do not appear to have higher aggregate cardiometabolic risk based on MetS and PDAY criteria compared with HIV‐seronegative youth in South Africa. Youth with perinatally acquired HIV appear to have reduced odds for MetS. The ‘non‐overweight but metabolically unhealthy phenotype’ among youth with perinatally acquired HIV in South Africa warrants further investigation. While risk stratification by traditional clinical factors may not reveal increased aggregate risk for atherosclerosis or MetS, this population may benefit from other strategies to assess their long‐term cardiometabolic disease risk, especially as youth are ageing into adulthood and face a double burden of disease with high rates of HIV and CVD.
AUTHOR CONTRIBUTIONS
JJ, HJZ and NABN conceptualized the study. SD‐F wrote the first draft of the manuscript and took primary responsibility for coordinating revisions and finalizing the manuscript. NAA‐A, NJ and EC collected and cleaned data. MN and LM analysed the data, and LBB and LM helped edit the final versions of Methods section. LM and HJZ gave overall input on the structure of the paper. All authors reviewed the manuscript and approved the final version.
FUNDING INFORMATION
This work was supported by the National Heart, Lung, and Blood Institute (NHLBI) R01HL151287, the National Institute of Child Health and Human Development (R01HD074051) and the South Africa Medical Research Council (SA‐MRC).
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
Supporting information
Data S1. Supporting information.
ACKNOWLEDGEMENTS
We would like to thank all the study participants and staff at the University of Cape Town and Red Cross War Memorial Children's Hospital in Cape Town, without whom this study would not have been possible.
Dirajlal‐Fargo S, Nyathi M, Sun S, et al. Metabolic syndrome and Pathobiological Determination of Atherosclerosis in Youth risk score in adolescents with and without perinatally acquired HIV in the Cape Town Adolescent and Antiretroviral Cohort (CTAAC)‐Heart study. HIV Med. 2025;26(8):1239‐1250. doi: 10.1111/hiv.70049
These data were presented in part at the International Paediatric HIV Workshop, in July 2024, in Munich, Germany.
DATA AVAILABILITY STATEMENT
Data is not publicly available but may be made available upon request if needed.
REFERENCES
- 1. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) . Expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA. 2001;285(19):2486‐2497. [DOI] [PubMed] [Google Scholar]
- 2. Jacobson DL, Tang AM, Spiegelman D, et al. Incidence of metabolic syndrome in a cohort of HIV‐infected adults and prevalence relative to the US population (National Health and Nutrition Examination Survey). J Acquir Immune Defic Syndr. 2006;43(4):458‐466. [DOI] [PubMed] [Google Scholar]
- 3. Espiau M, Yeste D, Noguera‐Julian A, et al. Metabolic syndrome in children and adolescents living with HIV. Pediatr Infect Dis J. 2016;35(6):e171‐e176. [DOI] [PubMed] [Google Scholar]
- 4. Shiau S, Yu W, Jacobson DL, et al. Components of metabolic syndrome associated with lower neurocognitive performance in youth with perinatally acquired HIV and youth who are HIV‐exposed uninfected. J Neurovirol. 2021;27(5):702‐715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Khan SS, Matsushita K, Sang Y, et al. Development and validation of the American Heart Association's PREVENT equations. Circulation. 2024;149(6):430‐449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gidding SS, Colangelo LA, Nwabuo CC, et al. PDAY risk score predicts cardiovascular events in young adults: the CARDIA study. Eur Heart J. 2022;43(30):2892‐2900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. McGill HC Jr, McMahan CA, Gidding SS. Preventing heart disease in the 21st century: implications of the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) study. Circulation. 2008;117(9):1216‐1227. [DOI] [PubMed] [Google Scholar]
- 8. Grinspoon SK, Fitch KV, Zanni MV, et al. Pitavastatin to prevent cardiovascular disease in HIV infection. N Engl J Med. 2023;389(8):687‐699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gooding HC, Gidding SS, Moran AE, et al. Challenges and opportunities for the prevention and treatment of cardiovascular disease among young adults: report from a National Heart, Lung, and Blood Institute Working Group. J Am Heart Assoc. 2020;9(19):e016115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Navar AM, Fine LJ, Ambrosius WT, et al. Earlier treatment in adults with high lifetime risk of cardiovascular diseases: what prevention trials are feasible and could change clinical practice? Report of a National Heart, Lung, and Blood Institute (NHLBI) Workshop. Am J Prev Cardiol. 2022;12:100430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mahtab S, Frigati LJ, Ntusi NAB, et al. The determinants of elevated Pathobiological Determination of Atherosclerosis in youth risk score in perinatally HIV‐infected adolescents in South Africa. J Acquir Immune Defic Syndr. 2024;95(1):82‐89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Jao J, Zar HJ, Morné K, et al. Subclinical myocardial fibrosis in South African youth with HIV: results from the CTAAC‐heart study. Open Forum Infect Dis. 2024;11(10):ofae555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. de Lamas C, Kalén A, Anguita‐Ruiz A, et al. Progression of metabolic syndrome and associated cardiometabolic risk factors from prepuberty to puberty in children: the PUBMEP study. Front Endocrinol. 2022;13:1082684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Keskin M, Kurtoglu S, Kendirci M, Atabek ME, Yazici C. Homeostasis model assessment is more reliable than the fasting glucose/insulin ratio and quantitative insulin sensitivity check index for assessing insulin resistance among obese children and adolescents. Pediatrics. 2005;115(4):e500‐e503. [DOI] [PubMed] [Google Scholar]
- 15. Zimmet P, Alberti KG, Kaufman F, et al. The metabolic syndrome in children and adolescents – an IDF consensus report. Pediatr Diabetes. 2007;8(5):299‐306. [DOI] [PubMed] [Google Scholar]
- 16. Xi B, Zong X, Kelishadi R, et al. International waist circumference percentile cutoffs for central obesity in children and adolescents aged 6 to 18 years. J Clin Endocrinol Metab. 2020;105(4):e1569‐e1583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Fernández JR, Redden DT, Pietrobelli A, Allison DB. Waist circumference percentiles in nationally representative samples of African‐American, European‐American, and Mexican‐American children and adolescents. J Pediatr. 2004;145(4):439‐444. [DOI] [PubMed] [Google Scholar]
- 18. Gidding SS, Rana JS, Prendergast C, et al. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) risk score in young adults predicts coronary artery and abdominal aorta calcium in middle age: the CARDIA study. Circulation. 2016;133(2):139‐146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Bowo‐Ngandji A, Kenmoe S, Ebogo‐Belobo JT, et al. Prevalence of the metabolic syndrome in African populations: a systematic review and meta‐analysis. PLoS One. 2023;18(7):e0289155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Bitew ZW, Alemu A, Ayele EG, Tenaw Z, Alebel A, Worku T. Metabolic syndrome among children and adolescents in low and middle income countries: a systematic review and meta‐analysis. Diabetol Metab Syndr. 2020;12(1):93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Sekokotla MA, Goswami N, Sewani‐Rusike CR, Iputo JE, Nkeh‐Chungag BN. Prevalence of metabolic syndrome in adolescents living in Mthatha, South Africa. Ther Clin Risk Manag. 2017;13:131‐137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Tailor AM, Peeters PH, Norat T, Vineis P, Romaguera D. An update on the prevalence of the metabolic syndrome in children and adolescents. Int J Pediatr Obes. 2010;5(3):202‐213. [DOI] [PubMed] [Google Scholar]
- 23. Magge SN, Goodman E, Armstrong SC. The metabolic syndrome in children and adolescents: shifting the focus to Cardiometabolic risk factor clustering. Pediatrics. 2017;140(2):e20171603. [DOI] [PubMed] [Google Scholar]
- 24. Gradidge PJ, Crowther NJ. Review: metabolic syndrome in black south African women. Ethn Dis. 2017;27(2):189‐200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Cook I, Alberts M, Lambert EV. Influence of cut‐points on patterns of accelerometry‐measured free‐living physical activity in rural and urban black South African women. J Phys Act Health. 2012;9(2):300‐310. [DOI] [PubMed] [Google Scholar]
- 26. Adediran O, Akintunde AA, Edo AE, Opadijo OG, Araoye AM. Impact of urbanization and gender on frequency of metabolic syndrome among native Abuja settlers in Nigeria. J Cardiovasc Dis Res. 2012;3(3):191‐196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. McMahan CA, Gidding SS, Viikari JS, et al. Association of Pathobiologic Determinants of Atherosclerosis in Youth risk score and 15‐year change in risk score with carotid artery intima‐media thickness in young adults (from the Cardiovascular Risk in Young Finns Study). Am J Cardiol. 2007;100(7):1124‐1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Gidding SS, Rana JS, Prendergast C, et al. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) risk score in young adults predicts coronary artery and abdominal aorta calcium in middle age. Circulation. 2016;133(2):139‐146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Patel K, Wang J, Jacobson DL, et al. Aggregate risk of cardiovascular disease among adolescents perinatally infected with the human immunodeficiency virus. Circulation. 2014;129(11):1204‐1212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Greybe L, Barnabas S, Cotton M, Rose P, Rabie H, Frigati L. Cardiometabolic risk profiles of adolescents living with perinatally acquired HIV in South Africa. Pediatr Infect Dis J. 2024;43(7):669‐674. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Triant VA, Perez J, Regan S, et al. Cardiovascular risk prediction functions underestimate risk in HIV infection. Circulation. 2018;137(21):2203‐2214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Feinstein MJ, Nance RM, Drozd DR, et al. Assessing and refining myocardial infarction risk estimation among patients with human immunodeficiency virus: a study by the centers for AIDS research network of integrated clinical systems. JAMA Cardiol. 2017;2(2):155‐162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Surial B, Chammartin F, Damas J, et al. Impact of integrase inhibitors on cardiovascular disease events in people with human immunodeficiency virus starting antiretroviral therapy. Clin Infect Dis. 2023;77(5):729‐737. [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.
Supplementary Materials
Data S1. Supporting information.
Data Availability Statement
Data is not publicly available but may be made available upon request if needed.