Summary
Background
Altered adipose tissue biology plays a key role in the development of cardiovascular disease (CVD) and cancer and contributes to mortality. We assessed the association between obesity in people with HIV (PWH) at antiretroviral therapy (ART) initiation and the risk of clinical events.
Methods
In this prospective cohort study, we included ART-naïve adults PWH enrolled in the Italian Cohort Naive Antiretrovirals (Icona) Foundation study cohort in Italy across 96 sites between January 1997 and July 2025. Participants were included if they initiated ART, had body mass index (BMI) > 18.5 kg/m2, and were free from AIDS, CVD, and cancer at ART initiation. Participants were classified as having obesity, defined as BMI ≥ 30 kg/m2, or not having obesity, defined as BMI 18.5–29.9 kg/m2, at ART initiation. The primary outcome was the first occurrence of CVD, cancer, or death. We performed a standard survival analysis with time-fixed covariates at baseline with a composite outcome of CVD/cancer/death. We hypothesized that age may be an effect measure modifier: results are presented after stratification by age (young [18–30], middle aged [31–60] and older [>60 years]).
Findings
A total of 11,652 PWH (20.7% females, median age 38 years [Inter Quartile Range (IQR): 31, 46]) were included: 11,005 (94.4%) were people without obesity and 647 (5.6%) people with obesity. A total of 837 events were recorded: 122 CVD, 374 cancers and 341 deaths. By 15 years from ART initiation, the risk of CVD/cancer/death was 19.1% in people with obesity (95% Confidence Interval [CI]: 14.1–24.9%) vs. 12.4% in people without obesity (95% CI:11.3–13.4%, log-rank p = 0.0029). After controlling for confounding, the difference was attenuated (adjusted hazard ratio [aHR] 1.28 [95% CI:0.98–1.67], p = 0.073). Although we did not find evidence for interaction (p-value = 0.758) there was a trend for a larger effect among younger PWH: young (aHR 1.83 [95% CI 0.74, 4.56]), middle aged (aHR 1.49 [95% CI 1.11, 2.00]) and older (aHR 1.20 [95% CI 0.60, 2.41]).
Interpretation
Obesity at ART initiation may be associated with a clinically meaningful increase in the risk of CVD, cancer, and death. We found weak evidence of a possible age-related gradient, with larger effect size in PWH under 60 years of age, which warrants further investigation.
Funding
The Italian Cohort Naive Antiretrovirals (Icona) Foundation Study is supported by unrestricted grants from Gilead Sciences, ViiV Healthcare, Merck Sharpe & Dohme.
Keywords: HIV, Obesity, Cardiovascular disease, Cancer, Death
Research in context.
Evidence before this study
We searched PubMed from database inception to 3 July 2026 for studies published in English using the terms “HIV”, “obesity”, “body mass index”, “cardiovascular disease”, “cancer”, “mortality”, and “antiretroviral therapy”. A total of 14,245 studies were identified. Body weight abnormalities have long been recognized as clinically relevant in people with HIV (PWH). While underweight was historically linked to advanced disease and mortality, recent data suggest a growing epidemic of obesity among PWH. Adipose tissue plays a key role in chronic inflammation, immune activation, and viral persistence, potentially contributing to non-AIDS comorbidities. Observational studies in PWH have shown an association between body mass index (BMI) and diabetes, and cardiovascular disease (CVD), but causal link has been limited by confounding and study design. To date, no study has evaluated whether obesity at antiretroviral therapy (ART) initiation affects the long-term risk of CVD, cancer, or death in PWH.
Added value of this study
Our analysis of over 11,000 ART-naïve PWH receiving care in Italy identified an association between obesity at ART initiation and a higher subsequent risk of CVD, cancer, and death. Notably, this association was not consistent across age groups: the excess risk was more pronounced in younger and middle-aged individuals, suggesting potential age-related heterogeneity in the clinical consequences of obesity. Together with emerging evidence on the metabolic and anti-inflammatory benefits of novel weight-reducing therapies, these findings underscore the need to formally evaluate whether treating obesity at the time of ART initiation can improve long-term health outcomes in PWH. This may be particularly relevant at earlier stages of adult life.
Implications of all the available evidence
These findings provide new insight into the complex relationship between obesity and the development of age-related comorbidities in PWH, highlighting the importance of early metabolic risk assessment and prevention strategies at ART initiation. Obesity at ART initiation may be associated with a worse long-term clinical prognosis in PWH. We found weak evidence that there may be an age-related gradient, with higher risk among those younger than 60 years, which would be of clinical relevance if true and warrants further investigation.
Introduction
Body weight has long been a central concern in epidemiology and medical science in general. Obesity (defined as body mass index [BMI]>30) is a complex chronic disease associated with a markedly increased risk of serious clinical events. A causal link between obesity and the risk of morbidity, such as cardiovascular disease (CVD) and some types of cancers, has been reported in the general population.1,2 The recent Lancet Commission on obesity further emphasized that obesity should be understood as a chronic, relapsing disease characterized not only by excess adiposity but also by evidence of organ dysfunction or substantial health impairment attributable to adipose tissue dysfunction.3
Recently, obesity has fallen under the spotlight in HIV research, as it has grown to be an epidemic within the community of people with HIV (PWH).4 A single centre study conducted in the US found a prevalence of obesity of around 28% among PWH.5 However, the burden of obesity in PWH varies across settings and largely reflects the prevalence of obesity in those settings, underscoring the importance of contextual and population-level determinants.
The Copenhagen Comorbidity in HIV Infection Study reported that for a given BMI, PWH showed an independent almost two-fold higher risk of abdominal obesity compared to people without HIV,6 highlighting a disproportionate burden of central adiposity and altered fat distribution in this population. The impact of obesity has previously been reported in antiretroviral therapy (ART)- experienced PWH. Data from the D:A:D study found an association between BMI and the risk of diabetes during ART, but there was insufficient evidence to support an association with the risk of CVD.7
Several mechanisms of association between obesity and morbidity and/or death have been postulated. Altered adipose tissue biology is a well-recognized driver of systemic inflammation and immune dysfunction.8 Indeed, the co-existence of HIV infection and obesity seems to exponentially increase inflammation and create a substrate for the early onset of clinical complications during ART.9
Whether obesity at ART initiation influences long-term risk of major clinical outcomes in PWH remains unclear. In particular, it is unknown whether the association between baseline obesity and subsequent clinical events differ across age groups or by sex assigned at birth.
We aimed to assess the association between obesity at ART initiation and the risk of developing CVD, cancer, or death in PWH, and to investigate whether this association is modified by age and sex assigned at birth.
Methods
Study design
We included PWH enrolled in a prospective observational cohort of ART-naïve adults. Enrolment in the Italian Cohort Naive Antiretrovirals (Icona) Foundation study cohort began in 1st January 1997 and is still ongoing.10 More than 20,000 Italian PWH have been enrolled so far across 96 centres. Collected data includes laboratory parameters (CD4 count, HIV-RNA), dates of ART initiation and interruption, and clinical events.11,12
The inclusion criteria for the present study specified PWH enrolled prior to 31 July 2025 and at ART initiation (i) were free from AIDS and any CVD or cancer, and (ii) BMI > 18.5. Underweight (BMI < 18.5) PWH were excluded.
The primary analysis is the composite outcome of CVD (myocardial infarction, coronary revascularization, cerebral hemorrhage and cerebral ischemia), cancer or death, as has been reported in meta-analyses of Mendelian randomization studies in the general population.13 In line with previous studies, we excluded non-melanoma skin cancers from the endpoint.14
The secondary endpoints were: i) CVD + cancer (censoring persons’ follow-up at time of death), ii) CVD + BMI-related cancers (hepatocellular, breast, colorectal, pancreatic, kidney, esophageal, gallbladder and corpus uteri) + death.15
Participants were grouped according to their BMI measured at ART initiation in people with obesity (≥30 kg/m2) vs. people without obesity (18.5–29.9 kg/m2).
Ethics
The institutional review boards of all participating centres approved the Icona Foundation Study. Each participant signed an informed consent according to committees’ ethical standards and the Helsinki Declaration (October 2013). Patient engagement was sought at the annual Icona Foundation Study meeting (June 2025). The latest amendment of the Icona Foundation Study was approved centrally in July 2024 (Lazio Area 4 Territorial Ethics Committee approval 158 no. 83–2024). Reporting of the study follows the strengthening of the reporting of observational studies in epidemiology (STROBE) guidelines.
Statistical analysis
Main participants’ characteristics at ART initiation were described overall and stratified by BMI exposure groups (people with obesity vs. people without obesity).
Individuals with overweight were included in the reference group because BMI ≥30 kg/m2 represents the established clinical threshold for obesity and associated adipose tissue dysfunction. In addition, previous HIV cohort studies have not consistently demonstrated excess risk among individuals with overweight compared with those of normal weight. A sensitivity analysis using three BMI categories (normal weight, overweight, and obesity) was conducted to explore potential differences across strata.6
Hypothesis testing was conducted by means of chi-square test for categorical variables and Mann–Whitney test for numerical variables. We also described the prevalence of obesity at ART initiation over time.
We performed a standard survival analysis with time-fixed covariates at baseline with the primary composite outcome described above. Follow-up started at ART initiation and ended at the first occurrence of the outcome of interest, last available clinical visit, or administrative censoring, whichever occurred first. Specifically, we performed unweighted and weighted (controlled for baseline confounding) Kaplan–Meier (KM) curves (up to 15 years follow-up). KM curves were compared with the log-rank test. Weighted Kaplan–Meier curves were generated to provide a graphical representation of cumulative outcome risk after accounting for measured baseline differences between exposure groups. We used inverse probability of treatment weighting based on the estimated probability of obesity at ART initiation conditional on baseline covariates included in the adjusted Cox model: year of ART initiation, age, sex assigned at birth, mode of HIV acquisition, HIV-RNA and CD4 count at ART initiation, hepatitis co-infection, smoking, diabetes, level of education, and employment status. Covariate balance before and after weighting was assessed using absolute standardised mean differences, with values < 0.1 considered indicative of good balance. We did not use inverse probability of censoring weight as the analysis essentially uses cause-specific hazards. A standard proportional hazard Cox regression analysis conditioned on baseline covariates was conducted. We tabulated unadjusted and adjusted hazard ratios (HR) of the exposure. Adjusted HR (aHR) was controlled for by a set of confounders including a list of a priori variables identified as important predictors of outcome, exposure or both: year of ART initiation, age, sex assigned at birth, mode of HIV acquisition, HIV-RNA and CD4 count at ART initiation, hepatitis co-infection, smoking, diabetes, level of education and employment status. Blood pressure and lipid levels were not included in the primary adjustment set because they may represent intermediate factors in the pathway linking obesity to cardiovascular events and mortality.
We performed a sensitivity analysis using a 3-way categorization of the exposure of normal weight (BMI 18.5–24.9) and PWH with overweight into distinct groups. As an additional sensitivity analysis for the CVD/cancer endpoint, we fitted Fine–Gray sub-distribution hazard models treating death as a competing event. Models were adjusted for the same baseline covariates as the main Cox regression models.
We also hypothesised at the outset that the effect of obesity on the risk of developing the clinical endpoint may differ according to sex and age. To evaluate effect-measure modification we stratified age using the cut-off of young (18–30), middle aged (31–60) and older (>60 years) and formally tested the interaction between exposure and age fitted as continuous. The age groups were set to maximize the statistical power of the proportions of PWH in the cohort at ART initiation (a very low proportion of participants aged 65+ were people with obesity). Because the analysis had low power to detect interactions, regardless of the p-value, we chose to present the results after stratification by age group. A specular analysis was conducted to investigate effect measure modification by sex assigned at birth.
All statistical analyses were performed using SAS (version 9.4, SAS Institute, Cary, NC, USA) and Stata (StataCorp. 2023. Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC).
Role of funding source
The Icona Foundation Study is supported by unrestricted grants from Gilead Sciences, ViiV Healthcare, Merck Sharpe & Dohme. The funders had no involvement in study design, data collection, data analyses, data interpretation, or the writing of the report.
Results
A total of 11,652 PWH (20.7% females, median age 38 years [Inter Quartile Range (IQR): 31, 46]) were included: 647 (5.6%) were people with obesity and 11,005 (94.4%) were people without obesity (8157 normal weight; 2848 overweight). Compared to PWH with obesity, PWH without obesity were younger (38 [IQR: 31, 45] vs. 41 [IQR: 35, 50] years, p < 0.001) and more frequently men who have sex with men (44.8% vs. 31.7%, p < 0.001). At ART initiation, PWH without obesity also had lower baseline CD4 count (349 [IQR: 221, 494] vs. 389 [IQR: 249, 556] cells/mm3, p < 0.001) and lower prevalence of diabetes (1.5% vs. 6.8%, p < 0.001) (Tables 1 and 2). Supplementary Fig. S1 shows absolute standardised mean difference in factors comparing PWH with and without obesity before and after weighting. The prevalence and distribution of obesity at ART initiation over time is depicted in Supplementary Fig. S2. This figure shows that the prevalence of obesity has remained stable and close to the average estimate of 6% over time.
Table 1.
Main characteristics of people with HIV (PWH) at antiretroviral therapy (ART) initiation, stratified by exposure group (PWH with obesity and without obesity): Sociodemographic characteristics and lifestyle behaviours.
| Characteristics | PWH without obesity |
PWH with obesity |
p-valuea | Total |
|---|---|---|---|---|
| N = 11,005 | N = 647 | N = 11,652 | ||
| Baseline characteristics | ||||
| Sex at birth, n (%) | <0.001 | |||
| Female | 2241 (20.4%) | 171 (26.4%) | 2412 (20.7%) | |
| Male | 8764 (79.6%) | 476 (73.6%) | ||
| Age, years | <0.001 | |||
| Median (IQR) | 38 (31, 45) | 41 (35, 50) | 38 (31, 46) | |
| Nationality, n (%) | 0.126 | |||
| Not born in Italy | 1999 (18.2%) | 133 (20.6%) | 2132 (18.3%) | |
| Highest education level, n (%) | <0.001 | |||
| Primary school | 628 (5.7%) | 47 (7.3%) | 675 (5.8%) | |
| Middle school | 2436 (22.1%) | 145 (22.4%) | 2581 (22.2%) | |
| High school | 3361 (30.5%) | 181 (28.0%) | 3542 (30.4%) | |
| University | 1377 (12.5%) | 50 (7.7%) | 1427 (12.2%) | |
| Other/Unknown | 3203 (29.1%) | 224 (34.6%) | 3427 (29.4%) | |
| Employment, n (%) | <0.001 | |||
| Unemployed | 1591 (14.5%) | 91 (14.1%) | 1682 (14.4%) | |
| Employed | 5025 (45.7%) | 276 (42.7%) | 5301 (45.5%) | |
| Self-employed | 1732 (15.7%) | 97 (15.0%) | 1829 (15.7%) | |
| Occasional | 347 (3.2%) | 24 (3.7%) | 371 (3.2%) | |
| Student | 395 (3.6%) | 8 (1.2%) | 403 (3.5%) | |
| Retired | 250 (2.3%) | 20 (3.1%) | 270 (2.3%) | |
| Housewife | 333 (3.0%) | 26 (4.0%) | 359 (3.1%) | |
| Other/unknown | 1312 (11.9%) | 103 (15.9%) | 1415 (12.1%) | |
| Mode of HIV acquisition, n (%) | <0.001 | |||
| PWID | 1685 (15.3%) | 86 (13.3%) | 1771 (15.2%) | |
| MSM | 4926 (44.8%) | 205 (31.7%) | 5131 (44.0%) | |
| Heterosexual contacts | 3995 (36.3%) | 332 (51.3%) | 4327 (37.1%) | |
| Other/Unknown | 399 (3.6%) | 24 (3.7%) | 423 (3.6%) | |
| Hepatitis co-infection, n (%) | 0.434 | |||
| HCV-Ab | ||||
| No | 5726 (52%) | 353 (54.6%) | 6079 (52.2%) | |
| Yes | 1328 (12.1%) | 69 (10.7%) | 1397 (12%) | |
| Not tested | 3951 (35.9%) | 225 (34.8%) | 4176 (35.8%) | |
| HBsAg | 0.247 | |||
| No | 6916 (62.8%) | 422 (65.2%) | 7338 (63) | |
| Yes | 58 (0.5%) | 4 (0.6%) | 62 (0.5%) | |
| Not tested | 4031 (36.6%) | 221 (34.2%) | 4252 (36.5%) | |
| Psychiatric diagnosis, n (%) | 0.279 | |||
| Yes | 326 (3.0%) | 24 (3.7%) | 350 (3.0%) | |
| Diabetes, n (%) | <0.001 | |||
| Yes | 165 (1.5%) | 44 (6.8%) | 209 (1.8%) | |
| Active injecting drug use, n (%) | 0.356 | |||
| Yes | 389 (4.2%) | 27 (5.0%) | 416 (4.2%) | |
| Alcohol use, n (%) | 0.816 | |||
| Abstain | 2135 (19.4%) | 133 (20.6%) | 2268 (19.5%) | |
| Social use | 1651 (15.0%) | 95 (14.7%) | 1746 (15.0%) | |
| Abuse | 492 (4.5%) | 32 (4.9%) | 524 (4.5%) | |
| Unknown | 6727 (61.1%) | 387 (59.8%) | 7114 (61.1%) | |
| Smoking, n (%) | 0.001 | |||
| No | 3179 (28.9%) | 230 (35.5%) | 3409 (29.3%) | |
| Yes | 3008 (27.3%) | 160 (24.7%) | 3168 (27.2%) | |
| Characteristics at ART initiation | ||||
| Time from HIV diagnosis to ART initiation, months | 0.268 | |||
| Median (IQR) | 3 (1, 42) | 4 (1, 46) | 3 (1, 42) | |
| Calendar year of ART initiation | 0.002 | |||
| Median (IQR) | 2013 (2004, 2017) | 2014 (2008, 2018) | 2013 (2004, 2017) | |
| Type of first ART regimen, n (%) | 0.033 | |||
| 3TC + DTG | 311 (2.8%) | 32 (4.9%) | 343 (2.9%) | |
| PI/r | 2980 (27.1%) | 172 (26.6%) | 3152 (27.1%) | |
| NNRTI | 2764 (25.1%) | 153 (23.6%) | 2917 (25.0%) | |
| INSTI | 3063 (27.8%) | 173 (26.7%) | 3236 (27.8%) | |
| Other | 1887 (17.1%) | 117 (18.1%) | 2004 (17.2%) | |
| CD4 count, cells/mmc | <0.001 | |||
| Median (IQR) | 349 (221, 494) | 389 (249, 556) | 351 (223, 498) | |
| 0–200 cells/mm3, n (%) | 2314 (21.9%) | 113 (18.1%) | 2427 (21.7%) | |
| Viral load, log10 copies/mL | <0.001 | |||
| Median (IQR) | 4.71 (4.15, 5.24) | 4.58 (3.95, 5.04) | 4.71 (4.14, 5.23) | |
| Follow-up, months | 0.634 | |||
| Median (IQR) | 84 (32, 135) | 83 (32, 132) | 84 (32, 135) |
PWH, people with HIV; With obesity, BMI >30; without obesity, BMI 18.5–29.9; n, number; IQR, Inter Quartile Range; PWID, people who inject drugs; MSM, men who have sex with men; 3TC, lamivudine; DTG, dolutegravir; PI/r, boosted protease inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; INSTI, integrase inhibitors; BMI, body mass index.
Chi-square or Kruskal–Wallis test as appropriate.
Table 2.
Main characteristics of people with HIV (PWH) at antiretroviral therapy (ART) initiation, stratified by exposure group (PWH with obesity and without obesity): clinical characteristics.
| Characteristics | Total |
PWH without obesity |
PWH with obesity |
p-valuea |
|---|---|---|---|---|
| N = 11,652 | N = 11,005 | N = 647 | ||
| Baseline characteristics | ||||
| Hepatitis co-infection, n (%) | 0.434 | |||
| HCV-Ab | ||||
| No | 6079 (52.2%) | 5726 (52%) | 353 (54.6%) | |
| Yes | 1397 (12%) | 1328 (12.1%) | 69 (10.7%) | |
| Not tested | 4176 (35.8%) | 3951 (35.9%) | 225 (34.8%) | |
| HBsAg | 0.247 | |||
| No | 7338 (63%) | 6916 (62.8%) | 422 (65.2%) | |
| Yes | 62 (0.5%) | 58 (0.5%) | 4 (0.6%) | |
| Not tested | 4252 (36.5%) | 4031 (36.6%) | 221 (34.2%) | |
| Psychiatric diagnosis, n (%) | 0.279 | |||
| Yes | 350 (3.0%) | 326 (3.0%) | 24 (3.7%) | |
| Diabetes, n (%) | <0.001 | |||
| Yes | 209 (1.8%) | 165 (1.5%) | 44 (6.8%) | |
| Characteristics at ART initiation | ||||
| Time from HIV diagnosis to ART initiation, months | 0.268 | |||
| Median (IQR) | 3 (1, 42) | 3 (1, 42) | 4 (1, 46) | |
| Calendar year of ART initiation | 0.002 | |||
| Median (IQR) | 2013 (2004, 2017) | 2013 (2004, 2017) | 2014 (2008, 2018) | |
| Type of first ART regimen, n (%) | 0.033 | |||
| 3TC + DTG | 343 (2.9%) | 311 (2.8%) | 32 (4.9%) | |
| PI/r | 3152 (27.1%) | 2980 (27.1%) | 172 (26.6%) | |
| NNRTI | 2917 (25.0%) | 2764 (25.1%) | 153 (23.6%) | |
| INSTI | 3236 (27.8%) | 3063 (27.8%) | 173 (26.7%) | |
| Other | 2004 (17.2%) | 1887 (17.1%) | 117 (18.1%) | |
| CD4 count, cells/mmc | <0.001 | |||
| Median (IQR) | 351 (223, 498) | 349 (221, 494) | 389 (249, 556) | |
| 0–200 cells/mm3, n (%) | 2427 (21.7%) | 2314 (21.9%) | 113 (18.1%) | |
| Viral load, log10 copies/mL | <0.001 | |||
| Median (IQR) | 4.71 (4.14, 5.23) | 4.71 (4.15, 5.24) | 4.58 (3.95, 5.04) | |
| Follow-up, months | 0.634 | |||
| Median (IQR) | 84 (32, 135) | 84 (32, 135) | 83 (32, 132) |
With obesity, BMI >30; without obesity, BMI 18.5–29.9; n, number; IQR, Inter Quartile Range; 3TC, lamivudine; DTG, dolutegravir; PI/r, boosted protease inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; INSTI, integrase inhibitors.
Chi-square test for categorical variables and Wilcoxon rank-sum test for numerical variables.
A total of 837 events were recorded: 122 CVD, 374 cancers and 341 deaths. Table 3 shows the breakdown of the components of the main composite outcome by exposure group. Individual cancer diagnoses and type of CVD events are reported in Supplementary Table S1. The breakdown of the components of the main composite outcome by age strata in people with obesity and people without obesity is reported in Supplementary Table S2.
Table 3.
Clinical events observed throughout follow-up.
| Clinical events, n (%) | PWH without obesity | PWH with obesity | Total | p value |
|---|---|---|---|---|
| CVD | 112 (14.5%) | 10 (15.6%) | 122 (14.6%) | 0.794 |
| Cancer | 348 (45.0%) | 26 (40.6%) | 374 (44.7%) | |
| BMI related | 114 (14.7%) | 5 (7.8%) | 119 (14.2%) | |
| Non-BMI related | 234 (30.3%) | 21 (32.8%) | 255 (30.5%) | |
| Death | 313 (40.5%) | 28 (43.8%) | 341 (40.7%) | |
| Total | 773 (100%) | 64 (100%) | 837 (100%) |
With obesity, Body mass index (BMI) > 30; without obesity, BMI 18.5–29.9; CVD, cardiovascular disease.
Main composite endpoint analysis
The KM analysis estimated that up to 15 years from ART initiation, the risk of CVD, cancer or death was higher in people with obesity (19.1%; 95% Confidence Interval [CI]: 14.1–24.9%) vs. people without obesity (12.4%; 95% CI:11.3–13.4%, log-rank p = 0.0029) (Fig. 1A). Median follow-up, calculated as the arithmetic median of the survival time was 84 months (IQR 32–135). In the unadjusted Cox model, obesity at ART initiation was associated with a higher risk of CVD/cancer/death (HR 1.47, 95% CI 1.14–1.90, p = 0.003). However, after controlling for confounding, the difference in risk was attenuated, as shown by the weighted KM results (Fig. 1B). In the Cox regression model, after adjustment for baseline confounders, the point estimate remained compatible with a clinically relevant increase in risk (aHR 1.28, 95% CI 0.98–1.67, p = 0.073) (Table 4). Results were similar, although the magnitude of the association was attenuated, when we analysed the endpoint considering only comorbidities (aHR 1.08 [95% CI 0.75, 1.55], p = 0.67) and after restricting cancers to BMI-related cancers (aHR 1.17 [95% CI 0.86, 1.60], p = 0.31) (Supplementary Table S3).
Fig. 1.
Unweighted (A) and weighted (B) Kaplan–Meier curves of the main outcome: risk of cardiovascular disease (CVD), cancer or death according to the exposure. p values were calculated using the log-rank test. Numbers below the curves indicate the number of participants at risk at each time point. ART, antiretroviral therapy; CVD, cardiovascular disease.
Table 4.
Risk of cardiovascular disease (CVD)/cancer/death by exposure group (Hazard Ratio (HR)–standard Cox proportional hazard regression models): overall risk.
| Unadjusted HR (95% CI) |
Adjusted HR (95% CI) |
|||
|---|---|---|---|---|
| p value | p value | |||
| CVD/cancer/death | ||||
| PWH without obesity | 1 | 1 | ||
| PWH with obesity | 1.47 (1.14, 1.90) | 0.003 | 1.28 (0.98, 1.67) | 0.073 |
Multivariable models included year of antiretroviral therapy (ART) initiation, age, sex assigned at birth, mode of HIV acquisition, HIV-RNA and CD4 at ART, hepatitis co-infection, smoking, diabetes, level of education and employment status as covariates.
Sensitivity analysis
In the sensitivity analysis using three BMI categories (normal weight [reference group], overweight and obesity], overweight was not associated with the composite endpoint after adjustment for baseline confounders (aHR 0.92 [95% CI 0.78–1.08], p = 0.307). The adjusted estimate for obesity vs. normal weight was consistent with that of the primary analysis comparing obesity with non-obesity (aHR 1.26, 95% CI 0.96–1.66, p = 0.098) (Supplementary Table S3).
Effect measure modification by age and sex
Overall, there was no formal evidence for an interaction with age (p = 0.758). However, KM curves stratified by age suggested that the risk of developing a clinical event associated with obesity at ART initiation was higher among the young and middle aged compared to the older PWH (Supplementary Fig. S3A–C). These findings persisted after adjustment for confounding factors, with a trend toward larger effect sizes in younger participants (aHR 1.83 [95% CI 0.74–4.56]), more modest estimates in middle-aged individuals (aHR 1.49 [95% CI 1.11–2.00]), and smaller estimates in older participants (aHR 1.20 [95% CI 0.60–2.41]) (Table 5). However, the limited number of events within each stratum reduced the precision of these estimates.
Table 5.
Risk of cardiovascular disease (CVD)/cancer/death by exposure group (Hazard Ratio (HR)–standard Cox proportional hazard regression models): risk stratified by age group.
| Unadjusted HR (95% CI) | Adjusted HR (95% CI) | Interaction p-value | |
|---|---|---|---|
| CVD/cancer/death by age strata | |||
| Young (18–30 years) | |||
| PWH without obesity | 1 | 1 | |
| PWH with obesity | 1.76 (0.71, 4.35) | 1.83 (0.74, 4.56) | |
| Middle age (31–60 years) | p = 0.76 | ||
| PWH without obesity | 1 | 1 | |
| PWH with obesity | 1.37 (1.03, 1.83) | 1.49 (1.11, 2.00) | |
| Older (> 60 years) | |||
| PWH without obesity | 1 | 1 | |
| PWH with obesity | 1.16 (0.58, 2.30) | 1.20 (0.60, 2.41) | |
Multivariable model included year of ART initiation, sex assigned at birth, mode of HIV acquisition, HIV-RNA and CD4 at ART, hepatitis co-infection, smoking, diabetes, level of education and employment status as covariates.
Finally, data carried no formal evidence for an interaction between obesity and sex assigned at birth after controlling for confounding with similar magnitude of aHR among males (aHR 1.38 [95% CI 1.04, 1.84]) and females (aHR 1.42 [95% CI 0.77, 2.62], p = 0.93, Supplementary Fig. S4A and B and Table 6).
Table 6.
Risk of cardiovascular disease (CVD)/cancer/death by exposure group (Hazard Ratio (HR)–standard Cox proportional hazard regression models): risk stratified by sex at birth.
| Unadjusted HR (95% CI) | Adjusted HR (95% CI) | Interaction p-value | |
|---|---|---|---|
| CVD/cancer/death by sex at birth | |||
| Male | |||
| PWH without obesity | 1 | 1 | |
| PWH with obesity | 1.59 (1.20, 2.10) | 1.38 (1.04, 1.84) | |
| Female | p = 0.93 | ||
| PWH without obesity | 1 | 1 | |
| PWH with obesity | 1.10 (0.60, 2.02) | 1.42 (0.77, 2.62) | |
Multivariable models included year of ART initiation, age, mode of HIV acquisition, HIV-RNA and CD4 at ART, hepatitis co-infection, smoking, diabetes, level of education and employment status as covariates.
PWH, people with HIV; ART, antiretroviral treatment; HR, hazard ratio; CI, confidence interval.
Discussion
Our analysis shows that, when compared to PWH without obesity, PWH with obesity at ART initiation have a 28% higher risk of developing CVD, cancer or death after controlling for confounding. Although the magnitude of the association was attenuated after multivariable adjustment, its consistency across analyses and overall size suggests a clinically meaningful increase in risk. Of note, the point estimates were larger among young and middle aged PWH with obesity. In particular, we found weak evidence of a possible age-related gradient, with larger effect size in PWH under 60 years of age. Indeed, the interaction test did not formally provide statistical evidence of effect modification, and the estimates were uncertain due to the low number of events per age strata. Nevertheless, if confirmed in other studies, this finding has potential clinical relevance, identifying PWH under the age of 60 who have obesity at ART initiation as a group at higher risk of developing serious clinical outcomes.
At baseline, the prevalence of obesity in our cohort was 5.6% and remained stable over time. These data are consistent with national prevalence estimates in the general population in Italy, suggesting that obesity in PWH is not HIV-specific. The observed prevalence of obesity appears lower than expected, although it should be interpreted in the context of estimates from ART-naïve populations. Indeed, the discrepancy when comparing with higher prevalences reported in other studies, may be explained by differences in duration of ART exposure or by other factors, such as population-specific characteristics including diet or genetic background. In ART-naïve or early ART, other European/Mediterranean cohorts have reported similar obesity prevalence to our study, with approximately 5.0% in AMACS in Greece16 and 6.4% in the CAPOTA study in Southern Spain.17 A Brazilian cohort reported an estimate of 7.9% but by contrast,18 North American cohorts have reported substantially higher baseline obesity prevalence, ranging between 15% and 20%.19,20 These differences probably reflect geographical variation in background obesity prevalence, case-mix, calendar period, baseline immune status, and BMI ascertainment methods. Therefore, our estimate appears plausible for an Italian ART-naïve cohort, but in countries with higher baseline obesity prevalence, the population-attributable burden of obesity-related complications in PWH may be substantially greater.
In the general population, both overweight and obesity are associated with an increasing risk of all-cause mortality.21 In the D:A:D study including PWH, the highest risk of all-cause mortality was among those with obesity or underweight.7 Despite differences in the selection of study populations and timing of BMI assessment, our study confirmed obesity as an important determinant of long-term outcome.
Despite a lack of significance in the interaction test our adjusted estimates suggest that the risk of developing CVD, cancer or death varied by age with a trend for a larger effect size of obesity in younger age groups. When ordered by magnitude of effect, the association was strongest among individuals aged 18–30 years, intermediate among those aged 31–60 years, and weakest among those older than 60 years. The higher risk among young PWH with obesity may be due to the early development of co-morbidities (premature ageing) observed in PWH compared to people without HIV, and obesity may further accelerate this process.22 The inflammatory effect of obesity becomes less dominant with increasing age due to age-related immune modifications; the so-called immunosenescence. This may explain the smaller difference in risk in PWH with or without obesity among the older PWH.23 Although age-specific estimates may reflect a true biological effect, age-dependent biases may have also played a role, including differential misclassification of adiposity (e.g., BMI poorly reflects metabolic risk in older adults due to sarcopenia) as well as selective survival (older participants with obesity may represent a selected group more resilient to severe clinical events) and competing risks (older people are at higher risk of dying for other causes).24
Our study did not find any evidence that sex at birth was an effect measure modifier for the association of interest. It was shown that sex-related differences in adipose tissue distribution, hormonal milieu, and metabolic regulation can lead to distinct obesity trajectories and cardiometabolic consequences. In the general population, women accumulate more subcutaneous adipose tissue, while men show greater visceral fat deposition, conferring higher cardiometabolic risk.25 BMI is likely to be unable to capture differences in adipose tissue distribution. General data on the interaction between sex at birth and obesity for predicting clinical outcome are conflicting. The Women’s Interagency HIV Study did not find an increased risk of mortality among women classified as overweight or with obesity.26 A cross-sectional study showed that higher BMI and excessive ectopic fat burden among PWH on ART were associated with circulating markers of systemic inflammation.27
Following the adoption of integrase inhibitors as the preferred ART regimen, several studies have reported weight gain during treatment.28 In light of our findings suggesting an association between obesity at ART initiation and poorer clinical outcomes in PWH, clinicians should be encouraged to address obesity from the time of starting therapy.
With the introduction of glucagon-like peptide (GLP)-1 receptor-agonists, the objective of reducing excessive weight seems more clinically achievable.29 In the general population, short term data suggest that GLP-1 receptor-agonists reduce CVD event risk.30 Data in PWH are limited, but it has been hypothesized that a positive effect on metabolic-dysfunction-associated steatohepatitis may contribute to reducing CVD and cancer risks.31 A randomized double-blinded study in PWH showed that once-weekly semaglutide reduces visceral fat by 30.6%. Sub-analysis data suggest that inflammation decreases, offering a further protective effect.32 Our results raise the hypothesis that earlier identification and treatment of obesity, particularly among younger PWH, could modify long-term risk trajectories. Nevertheless, access to GLP-1 receptor agonists remains uneven, and reimbursement is often restricted to individuals with diabetes, raising concerns about socioeconomic disparities in preventive care. Compared to people without HIV, PWH are at higher risk of CVD and several cancers, mostly due to chronic low-grade inflammation.33 In the Icona cohort, soluble markers of inflammation are not routinely collected, thus we could not include them in our analysis, and therefore further studies are needed.
Our study has several limitations. In observational studies, BMI as an exposure has been the target of controversies in epidemiology, mainly in establishing associations or causal links between BMI (or a change in BMI) and clinical outcome.2 The relatively low prevalence of obesity at ART initiation in our cohort limited the precision of subgroup analyses and restrict generalisability to populations with higher baseline obesity prevalence or larger studies. Smoking status and alcohol use had a high proportion of missing or unknown values which may have introduced bias in our “missing indicator” analysis. Several factors considered common causes of obesity and risk of clinical progression are not typically measured in HIV studies (i.e., diet, physical activity, asymptomatic clinical diseases and complex genetic factors or interactions of genetic and environmental factors) so unmeasured confounding cannot be ruled out. Nevertheless, in sophisticated analyses in which unmeasured confounding is expected to be minimal, such as Mendelian randomisation studies, a causal link between BMI and clinical risk has been reported.13 Besides the issue of unmeasured confounding, there was also a potential issue with information bias and misclassification of the outcome as some events may not have been precisely documented (e.g., for cancers). However, a large effort has recently been carried out by Icona investigators to improve the quality of data.12 Also, our analysis was likely underpowered for single events (CVD, cancer or death). Reassuringly, secondary analysis of alternative end points, including CVD and cancer with death as non-informative censoring or treated as a competing risk, reported similar results to those of our primary analyses. With regards to the exposure, BMI was the sole anthropometric measure available. Incorporating measures of central obesity or body composition would likely provide a more precise risk stratification. Despite the intrinsic limitations of BMI, as we showed that obesity at ART-initiation may be a marker of increased long-term clinical vulnerability, our study provides actionable information for early risk stratification in routine HIV care. Finally, regarding potential mechanisms underlying our findings, it is known that PWH are at increased risk of CVD and several cancers, largely driven by chronic low-grade inflammation.30 However, in the Icona cohort, soluble markers of inflammation are not routinely collected and therefore could not be included in our analysis.
In conclusion, our analysis suggests that obesity at ART initiation may be associated with a worse long-term clinical prognosis in PWH. Obesity at ART initiation may not only be a peripheral metabolic comorbidity but also an early marker of long-term clinical vulnerability. We found weak evidence that there may be an age-related gradient, with higher risk among those younger than 60 years, which would be of clinical relevance if true and warrants further investigation.
Contributors
CM conceptualisation, writing—original draft, and writing—review & editing, A Gi writing—original draft, and writing—review & editing, GL investigation and writing—review & editing, ADB investigation and writing—review & editing, AGo investigation and writing—review & editing, GO investigation and writing—review & editing, EQR investigation and writing—review & editing, AC investigation and writing—review & editing, GMad investigation and writing—review & editing, VM investigation and writing—review & editing, RR investigation and writing—review & editing, SLC investigation and writing—review & editing, GMar investigation and writing—review & editing, JC writing—original draft, and writing—review & editing, AdM investigation and writing—review & editing, GG investigation and writing—review & editing and ACL accessed and verified data and formal analysis and methodology. ACL and AGi accessed and verified the underlying data. Icona Foundation Study Group investigation. All authors read and approved the final version of the manuscript.
Data sharing statement
The datasets generated during the current study are not publicly available because they contain sensitive data to be treated under data protection laws and regulations. Appropriate agreement of data sharing can be arranged after a reasonable request to the corresponding author.
Declaration of interests
CM has received research grants from Gilead Sciences, Speaker honoraria from Gilead Sciences, ViiV Healthcare, MSD, Johnson & Johnson, travel grants from Gilead.
AGi received consultancy fees from ViiV Healthcare, Gilead Sciences, MSD and Janssen travel grant from Gilead Sciences and research grant from ViiV and Gilead.
GL received fees for consultations from ViiV Healthcare, Insmed and Pfizer srl.
ADB served as a consultant or attended advisory board meetings for Gilead Sciences, ViiV Healthcare, MSD, and received research grants from Gilead Sciences and ViiV Healthcare.
AGo served as consultant and received research grants from Janssen, ViiV Healthcare, MSD, BMS, ABBVIE, Gilead Sciences, Novartis, Pfizer, Astellas, Astrazeneca, Angelini, Moderna, Novavax, Shionogi.
GO has nothing to declare.
EQR received travel Grant from Gilead Sciences and ViiV Healthcare, received a research grant from Gilead Sciences.
GMad has received consultancy and/or speakers’ fees from Gilead Sciences, Janssen, Merck Sharp & Dohme, ViiV and Advanz Pharma.
AC received consultant fees from Gilead and ViiV Healthcare.
GMar has received consultancy and/or speakers’ fees from Gilead Sciences, Merck Sharp & Dohme, ViiV and Advanz Pharma.
VM has served as a paid consultant for Gilead Sciences and ViiV Healthcare and has received institutional research funding from these companies.
RR received consultancy fees from ViiV Healthcare, Gilead Sciences, MSD and Johnson&Johnson, travel and research grant from Gilead Sciences.
SLC received research grant from Gilead Sciences, ViiV Healthcare, MSD.
GM is Advisory board member for Gilead Sciences, Viiv Healthcare and received travel grants from Gilead Sciences.
JC has nothing to declare.
GG received research grant and speaker honorarium from Gilead Sciences, ViiV Healthcare, MSD, Jansen and attended advisory boards of Gilead, ViiV and MERCK.
AdM has nothing to declare.
ACL declared research grant or contract for his institute (UCL) by Icona Foundation, and by European Union's Horizon2020: Grant Agreement No 101046016 “EuCARE: European Cohorts of Patients and Schools to Advance.
Acknowledgements
The abstract has been presented as a poster abstract at the Conference on Retroviruses and Opportunistic Infections February 22–25 2026, Denver, Colorado, USA. Section (n.651).
The Icona Foundation Study is supported by unrestricted grants from Gilead Sciences, ViiV Healthcare, Merck Sharpe & Dohme.
Footnotes
Supplementary data related to this article can be found at https://doi.org/10.1016/j.eclinm.2026.104150.
Contributor Information
Cristina Mussini, Email: cristina.mussini@unimore.it.
Italian Cohort Naive Antiretrovirals (Icona) Foundation study for the Icona Foundation Study Group:
Clara Abeli, Loredana Alessio, Andrea Antinori, Spinello Antinori, Matteo Augello, Alessandra Bandera, Valentina Barocci, Alessandro Bartoloni, Matteo Bassetti, Giuliana Battagin, Sebastiano Bazzichetto, Sabrina Blanchi, Nicoletta Bobbio, Serena Irene Bonelli, Paolo Bonfanti, Valeria Bono, Stefano Bonora, Beatrice Borchi, Luca Bortolani, Giorgio Bozzi, Beatrice Bragato, Elena Bruzzesi, Andrea Calcagno, Guido Calleri, Valeria Calvino, Marta Camici, Marcella Capozzi, Nicolo Capra, Stefania Carrara, Giorgia Carrozzo, Antonio Cascio, Antonella Castagna, Alice Castellaccio, Annamaria Cattelan, Francesca Ceccherini-Silberstein, Benedetto Maurizio Celesia, Massimo Cernuschi, Adriana Cervo, Luchino Chessa, Paola Cinque, Pietro Colletti, Laura Comi, Nicola Coppola, Romina Corsini, Maria Luisa Cosmaro, Maria Vittoria Cossu, Cecilia Costa, Andrea Costantini, Marco Cotrufo, Stefania Cretella, Gabriella D’Ettorre, Sarah Dal Zoppo, Giuseppe De Socio, Andrea De Vito, Cosmo Del Borgo, Vincenzo Leonardo Del Negro, Serena Dell’Isola, Eugenia Di Brino, Giovanni Di Filippo, Cinzia Di Giuli, Giovanni Di Perri, Elke Maria Erne, Vincenzo Esposito, Massimiliano Fabbiani, Katia Falasca, Iuri Fanti, Massimo Farinella, Emanuele Focà, Rosa Fontana Del Vecchio, Agostino Fortunato, Antonina Franco, Francesco Maria Fusco, Marisa Fusto, Roberta Gagliardini, Ivan Gentile, Anna Maria Geretti, Andrea Giacometti, Nicola Gianotti, Laura Gianserra, Enrico Girardi, Silvia Graziano, Roberto Gulminetti, Gianacarlo Iaiani, Valentina Iannone, Chiara Iaria, Alessia Lai, Silvia Lamonica, Simone Lanini, Alessandra Latini, Claudia Lazzaretti, Nicolò Lentini, Miriam Lichtner, Riccardo Lolatto, Franco Maggiolo, Vincenzo Malagnino, Lisa Malincarne, Giulia Mancarella, Silvia Marchiori, Raffaella Marocco, Salvatore Martini, Ilaria Mastrorosa, Domitilla Meloni, Marianna Menozzi, Barbara Menzaghi, Daniela Messeri, Ivana Mezzaroma, Giulia Micheli, Eugenio Milano, Salvatore Minniti, Giulia Moi, Maria Cristina Moioli, Chiara Molteni, Annalisa Mondi, Francesca Montagnani, Davide Moschese, Camilla Muccini, Marco Muccio, Arianna Narducci, Stefano Nicolè, Silvia Nozza, Giuseppe Nunnari, Angelo Pan, Giustino Parruti, Roberta Pastorino, Jessica Paulicelli, Giovanni Pellicanò, Carlo Federico Perno, Annalisa Perziano, Francesco Pes, Stefania Piconi, Angela Pieri, Carmela Pinnetti, Maria Maddalena Plazzi, Nicoletta Policek, Maria Cristina Poliseno, Emanuele Pontali, Gianluca Prota, Massimo Puoti, Michela Rigoni, Marco Rivano Capparuccia, Viviana Rizzo, Alessandra Rodano’, Ashley Roen, Roberta Rovito, Stefano Rusconi, Daria Russo, Ylenia Russotto, Hanieh Sadeghi, Caterina Sagnelli, Valentina Sala, Nadia Sangiovanni, Maria Mercedes Santoro, Carmen Santoro, Annalisa Saracino, Loredana Sarmati, Daniela Segala, Federica Sozio, Vincenzo Spagnuolo, Nicola Squillace, Giulio Starnini, Christof Stingone, Barbara Suligoi, Valentina Svicher, Lucia Taramasso, Carlo Tascini, Alessandro Tavelli, Camilla Tincati, Carlo Torti, Marcello Trizzino, Silvia Truffa, Claudia Ucciferri, Daniela Valenti, Jacopo Vecchiet, Alessandra Vergori, Pierluigi Viale, Rosaria Viglietti, Donatella Vincenti, and Alberto Zuppiroli
Appendix A. Supplementary data
References
- 1.Lauby-Secretan B., Scoccianti C., Loomis D., Grosse Y., Bianchini F., Straif K., International Agency for Research on Cancer Handbook Working Group Body fatness and Cancer--Viewpoint of the IARC Working Group. N Engl J Med. 2016;375:794–798. doi: 10.1056/NEJMsr1606602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Britton K.A., Massaro J.M., Murabito J.M., Kreger B.E., Hoffmann U., Fox C.S. Body fat distribution, incident cardiovascular disease, cancer, and all-cause mortality. J Am Coll Cardiol. 2013;62:921–925. doi: 10.1016/j.jacc.2013.06.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rubino F., Cummings D.E., Eckel R.H., et al. Definition and diagnostic criteria of clinical obesity. Lancet Diabetes Endocrinol. 2025;13(3):221–262. doi: 10.1016/S2213-8587(24)00316-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Koethe J.R., Jenkins C.A., Lau B., et al. North American AIDS cohort Collaboration on research and design (NA-ACCORD). Rising obesity prevalence and weight gain among adults starting antiretroviral therapy in the United States and Canada. AIDS Res Hum Retroviruses. 2016;32:50–58. doi: 10.1089/aid.2015.0147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Becofsky K.M., Wing E.J., Wing R.R., Richards K.E., Gillani F.S. Obesity prevalence and related risk of comorbidities among HIV+ patients attending a New England ambulatory centre. Obes Sci Pract. 2016;2:123–127. doi: 10.1002/osp4.38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gelpi M., Afzal S., Lundgren J., et al. Higher risk of abdominal obesity, elevated low-density lipoprotein cholesterol, and hypertriglyceridemia, but not of hypertension, in people living with Human Immunodeficiency Virus (HIV): results from the copenhagen comorbidity in HIV infection study. Clin Infect Dis. 2018;67:579–586. doi: 10.1093/cid/ciy146. [DOI] [PubMed] [Google Scholar]
- 7.Petoumenos K., Kuwanda L., Ryom L., et al. D:A:D Study Group Effect of changes in body mass index on the risk of cardiovascular disease and diabetes mellitus in HIV-positive individuals: results from the D:A:D study. J Acquir Immune Defic Syndr. 2021;86:579–586. doi: 10.1097/QAI.0000000000002603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Koethe J.R., Hulgan T., Niswender K. Adipose tissue and immune function: a review of evidence relevant to HIV infection. J Infect Dis. 2013;208:1194–1201. doi: 10.1093/infdis/jit324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bourgeois C., Gorwood J., Olivo A., et al. Contribution of adipose tissue to the chronic immune activation and inflammation associated with HIV infection and its treatment. Front Immunol. 2021;12 doi: 10.3389/fimmu.2021.670566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.d'Arminio Monforte A., Cozzi–Lepri A., Rezza G., et al. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naive patients. I.CO.N.A. study Group. Italian Cohort of Antiretroviral-Naive patients. AIDS. 2000;14:499–507. doi: 10.1097/00002030-200003310-00005. [DOI] [PubMed] [Google Scholar]
- 11.Mussini C., Giacomelli A., Quiros-Roldan E., et al. Risk of clinical events in virologically suppressed people with HIV switching to a two-drug regimen vs. remaining on a three-drug regimen: a target trial emulation. EClinicalMedicine. 2025;86 doi: 10.1016/j.eclinm.2025.103368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Giacomelli A., Lanini S., De Vito A., et al. All-cause and cause-specific mortality in people with HIV in Italy in 1997-2022: data from the icona cohort. Open Forum Infect Dis. 2025;12 doi: 10.1093/ofid/ofaf455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Larsson S.C., Burgess S. Causal role of high body mass index in multiple chronic diseases: a systematic review and meta-analysis of Mendelian randomization studies. BMC Med. 2021;19:320. doi: 10.1186/s12916-021-02188-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Piselli P., Tavelli A., Cimaglia C., et al. Icona Foundation Study Group Cancer incidence in people with HIV in Italy: comparison of the ICONA COHORT with general population data. Int J Cancer. 2025;157:1142–1153. doi: 10.1002/ijc.35493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Achhra A.C., Sabin C., Ryom L., et al. D:A:D Study Group Body mass index and the risk of serious Non-AIDS events and all-cause mortality in treated HIV-positive individuals: D: A: d cohort analysis. J Acquir Immune Defic Syndr. 2018;78:579–588. doi: 10.1097/QAI.0000000000001722. [DOI] [PubMed] [Google Scholar]
- 16.Pantazis N., Papastamopoulos V., Antoniadou A., et al. Changes in body mass index after initiation of antiretroviral treatment: differences by class of core drug. Viruses. 2022;14(8):1677. doi: 10.3390/v14081677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gómez-Ayerbe C., Palacios R., Mayorga M., et al. Weight changes after first-line antiretroviral initiation in a cohort of HIV-positive patients in Southern Spain (CAPOTA study) Int J STD AIDS. 2022;33(13):1119–1123. doi: 10.1177/09564624221125356. [DOI] [PubMed] [Google Scholar]
- 18.Bakal D.R., Coelho L.E., Luz P.M., et al. Obesity following ART initiation is common and influenced by both traditional and HIV-/ART-specific risk factors. J Antimicrob Chemother. 2018;73(8):2177–2185. doi: 10.1093/jac/dky145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Koethe J.R., Jenkins C.A., Lau B., et al. Rising obesity prevalence and weight gain among adults starting antiretroviral therapy in the United States and Canada. AIDS Res Hum Retroviruses. 2016;32(1):50–58. doi: 10.1089/aid.2015.0147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lam J.O., Leyden W.A., Alexeeff S., et al. Changes in body mass index over time in people with and without HIV infection. Open Forum Infect Dis. 2024;11(2) doi: 10.1093/ofid/ofad611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen Y., Ma L., Han Z., Xiong P. The global burden of disease attributable to high body mass index in 204 countries and territories: findings from 1990 to 2019 and predictions to 2035. Diabetes Obes Metab. 2024;26:3998–4010. doi: 10.1111/dom.15748. [DOI] [PubMed] [Google Scholar]
- 22.Guaraldi G., Orlando G., Zona S., et al. Premature age-related comorbidities among HIV-infected persons compared with the general population. Clin Infect Dis. 2011;53:1120–1126. doi: 10.1093/cid/cir627. [DOI] [PubMed] [Google Scholar]
- 23.Nasi M., De Biasi S., Gibellini L., et al. Ageing and inflammation in patients with HIV infection. Clin Exp Immunol. 2017;187:44–52. doi: 10.1111/cei.12814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dramé M., Godaert L. The obesity paradox and mortality in older adults: a systematic review. Nutrients. 2023;15:1780. doi: 10.3390/nu15071780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Palmer B.F., Clegg D.J. The sexual dimorphism of obesity. Mol Cell Endocrinol. 2015;402:113–119. doi: 10.1016/j.mce.2014.11.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sharma A., Hoover D.R., Shi Q., et al. Relationship between body mass index and mortality in HIV-infected HAART users in the women's interagency HIV study. PLoS One. 2015;10 doi: 10.1371/journal.pone.0143740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Chen M., Hung C.L., Yun C.H., Webel A.R., Longenecker C.T. Sex differences in the association of fat and inflammation among people with treated HIV infection. Pathog Immun. 2019;4:163–179. doi: 10.20411/pai.v4i1.304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ryan P., Blanco J.L., Masia M., et al. PASO-DOBLE Study Group Maintenance therapy with dolutegravir and lamivudine versus bictegravir, emtricitabine, and tenofovir alafenamide in people with HIV (PASO-DOBLE): 48-week results from a randomised, multicentre, open-label, non-inferiority trial. Lancet HIV. 2025;12:e473–e484. doi: 10.1016/S2352-3018(25)00105-5. [DOI] [PubMed] [Google Scholar]
- 29.Zhang S., Niu S., An S., Cai X., Lao X. Efficacy and safety of semaglutide in non-diabetic adults with overweight or obesity: a meta-analysis of randomized controlled trials. Eur J Pharmacol. 2026;1015 doi: 10.1016/j.ejphar.2026.178587. [DOI] [PubMed] [Google Scholar]
- 30.Marso S.P., Daniels G.H., Brown-Frandsen K., et al. LEADER Steering Committee. LEADER Trial Investigators Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2016;375:311–322. doi: 10.1056/NEJMoa1603827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lake JE, Kitch D, Kantor A, et al. Abstract 757 in Programs and Abstracts Conference on Retroviruses and Opportunistic Infections. CA; San Francisco: March 9–12, 2025. Semaglutide improves steatohepatitis in people with HIV: the SLIM LIVER study. [Google Scholar]
- 32.Funderburg N.T., Ross Eckard A., Wu Q., et al. The effects of semaglutide on inflammation and immune activation in HIV-associated lipohypertrophy. Open Forum Infect Dis. 2025;12 doi: 10.1093/ofid/ofaf152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hart B.B., Nordell A.D., Okulicz J.F., et al. INSIGHT SMART and ESPRIT Groups Inflammation-related morbidity and mortality among HIV-Positive adults: how extensive is it? J Acquir Immune Defic Syndr. 2018;77:1–7. doi: 10.1097/QAI.0000000000001554. [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.