Stigma, depression, and poor expectations regarding aging among older adults with HIV in Brazil

Abstract

Introduction

HIV stigma negatively impacts multiple outcomes for persons living with HIV (PLHIV). Among older PLHIV, stigma may influence mental health and aging expectation. We aimed to evaluate factors related to HIV stigma and whether depression mediates its association with expectations regarding aging (ERA-12).

Design

Cross-sectional study of PLHIV ≥ 50 years, on antiretroviral therapy with suppressed viral load, enrolled in the Longitudinal Study of HIV & Aging in Brazil (ELEA-Brasil).

Methods

We assessed factors related to HIV stigma (12-item Berger, range:12–48, higher = higher stigma) and associated with ERA-12 (range: 0–100, lower = more negative aging expectation). Mediation analysis was performed to evaluate depression (Patient Health Questionnaire-9 [PHQ-9]; range:0–27, higher = higher symptoms) as a mediator of the association between stigma and ERA-12. Univariable and multivariable Poisson and linear regression models were applied, adjusted for demographics, time since HIV diagnosis, and substance use.

Results

We enrolled 702 PLHIV (median age 62 years; 64.9% men). Median and interquartile range (IQR) of HIV stigma, ERA-12 and PHQ-9 were 30 (25–35); 36 (22–47) and 3 (1–7) respectively. In adjusted analyses, older age (≥ 70 vs. 50–55 years) and shorter time since HIV diagnosis (10–19 years vs. ≥ 20 years) were significantly associated with lower HIV stigma. In adjusted models, higher HIV stigma (Coefficient = − 12.05 [95%CI − 17.84, − 6.26]) and lower education (Coefficient = − 7.60 [95%CI: − 13.40, − 1.81]) were associated with worse ERA-12. Depressive symptoms may reflect 25% of the overall effect of stigma on ERA-12.

Conclusion

We observed a high prevalence of HIV stigma, with depressive symptoms consistent with a potential mediating role in aging expectations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12981-026-00867-4.

Introduction

Older adults living with HIV (defined as ≥ 50 years of age) currently account for a third of all persons living with HIV (PLHIV) in Brazil [1] and may face distinct patterns and consequences of HIV stigmatization. Older PLHIV have been shown to hold more negative attitudes toward aging compared with older adults without HIV, and negative views of aging have been associated with higher likelihood of depression [2], which may subsequently lead to more geriatric health conditions such as frailty [3].

Stigma has been characterized as the attribution of negative labels to a particular person or a group [4]. Multiple sociodemographic characteristics may influence the occurrence and impact of stigma; moreover, health conditions may also be associated with stigmatizing experiences, particularly contagious diseases. HIV stigma has been extensively investigated due its relevance, and high detrimental impact on PLHIV [5, 6]. Since the onset of the HIV/AIDS epidemic, associations with men who have sex with men, gay or bisexual individuals, and people who inject drug were considered wrongful or immoral, which generated widespread stigmatization [7]. As a result, HIV is one of the few chronic conditions for which disclosure remains a significant challenge [8].

HIV stigma may lead to social isolation [9, 10] and may affect expectations and perceptions of aging. No prior research has directly addressed the association between HIV stigma and aging expectations among older PLHIV. Understanding the subject is an essential pillar for providing educational initiatives that could reduce the discounting, discrediting and discrimination against PLHIV. Here, we aimed to investigate factors associated with HIV stigma in Brazilian PLHIV aged 50 years older, and the association between HIV stigma, depression, and expectations of aging.

Methods

Study population

This cross-sectional study used baseline data collected between October, 2022, and June, 2023, as part of the observational Longitudinal Study of HIV & Aging in Brazil (ELEA-Brasil). Designed to examine the social, demographic, clinical, and immunologic determinants of aging outcomes among older PLHIV in Brazil. ELEA-Brazil prospectively enrolled adults living with HIV aged ≥ 50 years from clinical sites in three Brazilian cities (Centro de Referência e Treinamento HIV/AIDS (CRT-DST/AIDS) in São Paulo; Serviço de Extensão a Pessoas vivendo com HIV/AIDS do Hospital as Clínicas da Faculdade de Medicina de São Paulo (SEAP HCFMUSP) in São Paulo; Hospital Universitário Professor Edgard Santos (HUPES) in Salvador; and Instituto Nacional de Infectologia Evandro Chagas, Fundação Oswaldo Cruz (INI-Fiocruz) in Rio de Janeiro). Participants were included if were currently in use of antiretroviral therapy and HIV RNA < 200 copies/mL in the past 6 months. Potential participants were identified through their routinely scheduled appointments and were contacted either in person or via telephone. Stratified (age and sex) and systematic recruitment into ELEA-Brasil intended to enroll a study population reflective of each clinic’s patient populations. All participants received detailed explanation of the study procedures from a trained professional and provided informed consent. Study interviews were conducted in private, quiet rooms by trained personnel. All deidentified data were recorded through Research Electronic Data Capture (REDCap) [11] and prepared for data analysis at the data coordinating center at Vanderbilt University Medical Center.

Measurements

HIV stigma

We applied 12-item Short HIV Stigma scale [12] with 4-point Likert-type responses, translated and validated in Brazilian Portuguese [13], which comprise four distinct dimensions: 1. Personalized stigma (negative implications of revealing HIV status); 2. Disclosure concerns (strategies to maintain confidentiality); 3. Concerns about public attitudes (anticipated reactions from the general public); and 4. Negative self-image (emotional impact of feeling inferior to others due to HIV status). Total score was calculated by summing all items within the dimensions. Each dimension consisted of 3 questions, with scores ranging from 3 to 12, and the overall HIV stigma scale ranged from 12 (lowest stigma) to 48 (highest stigma) points. For statistical analysis, the HIV stigma scale was operationalized using two complementary approaches. First, consistent with previous studies, the scale was dichotomized at the 50th percentile to define higher versus lower stigma for descriptive purposes [14, 15]. Second, HIV stigma scale was categorized into quartiles and treated as an ordinal variable, allowing for a more nuanced interpretation of stigma burden across increasing levels. One important aspect is that this instrument does not explicitly account for individuals who have never disclosed their HIV status. The first three items refer to consequences of HIV status disclosure; therefore, these items could not be answered and remained blank for participants who reported never disclosing their HIV status. Because this missingness was not random but driven by non-disclosure, we assigned the maximum possible score to these items, based on the strong assumption that non-disclosure reflects higher HIV-related stigma [16–18].

Expectations regarding aging (ERA-12)

To address perception of aging, ERA-12 questionnaire was used [19]. It is composed by 12 questions, with 4 answer options, varying from definitely true to definitely false. ERA-12 comprises three dimensions: Physical Health Scale, Mental Health Scale and Cognitive Function Scale. Total scores range from 0 to 100 and lower scores indicate more negative views toward aging.

Patient health questionnaire-9 (PHQ-9)

To evaluate depression symptoms the present study applied PHQ-9 [20], validated in Brazilian Portuguese [21]. This scale was considered a categorized ordinal variable, ranging from 0 to 27 points, and divided into four groups: no depression (0–4); mild (5–9); moderate (10–14); moderately severe/severe depression (≥ 15).

Covariates

Covariates were selected a priori based on literature [18] and a conceptual framework focusing on sociodemographic and clinical factors consistently associated with HIV stigma, depression, and aging-related outcomes, particularly those reflecting potential intersectionality with other forms of stigma. Although variables such as social support and income are relevant, these data were not included and are acknowledged as limitations. Substance use was included given its clinical relevance [14] and was operationalized conservatively to capture more harmful exposure. The following variables were accounted in the analysis as potential confounders: age (categorized as 50–55 years, 56–60 years, 61–65 years, 66–70 years, > 70 years old), sex assigned at birth (male/female), time since HIV diagnosis (< 5, 5–9, 10–19, ≥ 20 years), race (White vs. non-white, including Black, Pardo [admixed Black and White], Asian, Indigenous, other), sexual orientation (heterosexual vs. non-heterosexual, including homosexual, bisexual, pansexual, asexual, other), regular substance use (daily alcohol/tobacco/cannabis or any use of other illicit substances), and years of formal education (< 8; 8–11; ≥ 12 years).

Statistical analysis

Demographic characteristics, health-related behaviors and depressive symptoms of the study population were presented as frequencies, percentages, means/medians, and standard deviations (SD) or interquartile ranges (IQR), as appropriate.

We applied Poisson regression to evaluate whether covariates such as age, sex, time since HIV diagnosis, race, sexual orientation, and level of education were associated with a binary HIV stigma outcome (defined as ≥ 50th percentile). Factors associated with HIV stigma were described as percentages, prevalence ratio (PR) and adjusted prevalence ratio (aPR) with their respective 95% confidence intervals (CI). Sensitivity analyses were performed using continuous and dichotomized HIV stigma measures.

We next explored the association between HIV stigma and expectations regarding aging using ERA-12 score as an outcome, identifying a priori depression as a potential mediator, all treated as continuous variables. We employed the product of coefficients method, estimating three linear regression models with robust standard errors: (1) the total effect of HIV stigma on ERA-12; (2) the effect of HIV stigma on depression; and (3) the direct effect of HIV stigma on ERA-12 controlling for depression. Bootstrap resampling with 5000 iterations was used to estimate 95% confidence intervals for the indirect effect. The proportion mediated was calculated as the indirect effect divided by the total effect, multiplied by 100. The direct, indirect and total effect were described as coefficients, with their respective 95% CI and p-value. Mediation analysis relies on strong assumptions, including absence of unmeasured confounding and correct temporal ordering between variables. We addressed confounding by adjusting for sexual orientation, education, race, time since HIV diagnosis, sex, age, and regular substance use.

Factors associated to the expectations regarding aging were described through beta coefficients and adjusted beta coefficients with 95% confidence intervals.

Finally, we conducted separate adjusted linear regression analyses to assess the association between each of the four domains of the HIV Stigma score (personalized stigma, disclosure concerns, concerns about public attitudes, and negative self-image) and ERA-12 score. This analysis aimed to verify the consistency of results across all four domains of HIV stigma. We performed diagnostics checks for multivariable model including residual-by-predictor plots, residual plots by categorical variables, and tests for homoscedasticity. There was no evidence of collinearity or important violations of model assumptions. To handle missing data, for participants who reported never disclosing their HIV status, the first three items of the HIV Stigma Scale were not answered, and we treated non-disclosure as indicative of a higher stigma burden and imputed the highest possible score for these disclosure-dependent items. To assess the robustness of this assumption, we conducted sensitivity analyses using two alternative linear regression models examining the association between HIV stigma and expectations regarding aging (ERA-12): one model including stigma scores with imputed values for non-disclosure, and a second model in which disclosure-dependent items were left missing for participants who never disclosed their HIV status. For all other questionnaires, we performed single imputation for missing survey responses when the individual question missingness within an instrument was < 20% for a single participant’s survey; imputation was performed using MICE package in R with regression models that only included the participant’s other responses on that survey as predictors of the missing value.

The statistical package Stata (StataCorp. College Station, TX: StataCorp LP, version 15.1 or higher) was used for all analyses, with a significance level of 0.05.

Sensitive data, including those related to stigma and mental health issues, were obtained by trained investigators in private interviews. We referred for applicable care all participants for whom unattended health conditions were identified during the study, including newly recognized mental health issues.

Results

This study enrolled 702 older adults living with HIV. Clinical and demographic characteristics of the individuals are shown in Table 1. Most participants self-identified as Pardo or Black race (53.3%) and heterosexual orientation (57.3%). The most prevalent marital status in our study population was single (42.4%). Regular substance use was noted in 35.5% of participants, with tobacco being the most used substance (18.5%). Regarding clinical measurements, we found that more than half of all individuals reported no symptoms of depression, while a quarter had mild depression, 9% had moderate depression, and 7% had moderately severe/severe depression symptoms. Median HIV stigma score was 30 (IQR: 25–35). Overall, 116 of all 702 individuals (16.5%) reported having never disclosed their HIV status to anyone. Considering the four HIV stigma scale dimensions, disclosure concerns had the highest median score (median = 10, IQR 9–12), followed by concerns about public attitudes (median = 9, IQR 7–10), negative self-image (median = 6, IQR 5–7) and personalized stigma (median = 5, IQR 3–9). The median ERA-12 score was 36, with an interquartile range of 22–47.

Table 1.

Demographic and clinical characteristics of ELEA participants

	Total N = 702
Age, years (IQR)	62 (56–68)
50–55, n (%)	147 (20.9)
56–60, n (%)	157 (22.4)
61–65, n (%)	115 (16.4)
66–70, n (%)	168 (23.9)
> 70, n (%)	115 (16.4)
Time since HIV diagnosis, years (IQR)	23 (14–27)
< 5 years, n (%)	42 (6.0)
5–9 years, n (%)	58 (8.3)
10–19 years, n (%)	152 (21.8)
≥ 20 years, n (%)	447 (63.9)
Sex, n (%)
Male	455 (64.8)
Female	247 (35.2)
Race/skin color, n (%)
White	317 (45.2)
Pardo/Mixed	219 (31.2)
Black	155 (22.1)
Other	11 (1.5)
Sexual orientation, n (%)
Heterosexual	402 (57.3)
Homosexual	237 (33.7)
Other/ No information	63 (9.0)
Marital status, n (%)
Single	298 (42.4)
Married/stable union	222 (31.6)
Separated/divorced/widowed	182 (26.0)
Education, n (%)
< 8 years	149 (21.2)
8–11 years	275 (39.2)
≥ 12 years	278 (39.6)
Regular substance use, n (%)	152 (35.5)
Daily alcohol	30 (4.5)
Daily tobacco	105 (18.5)
Daily cannabis	22 (4.8)
Any use of other drugs	42 (7.5)
Depression symptoms
Median PHQ-9 (IQR)	3 (1–7)
No depression (PHQ-9 score = 0–4), n (%)	410 (58.4)
Mild depression (PHQ-9 score = 5–9), n (%)	179 (25.5)
Moderate depression (PHQ-9 score = 10–14), n (%)	66 (9.4)
Moderately severe/severe (PHQ-9 score = ≥ 15), n (%)	47 (6.7)
HIV Stigma Score, n (%) ‡
Median Stigma Score (IQR)	30 (25–35)
< 25, n (%)	172 (24.5)
25–29, n (%)	157 (22.3)
30–34, n (%)	196 (27.9)
≥ 35, n (%)	177 (25.2)
Median ERA-12 (IQR)	36 (22–47)

‡Higher score denotes higher HIV-related stigma, assessed by 12-item Short HIV Stigma scale, values corresponding to quartiles

IQR = interquartile range; PHQ-9 = Patient Health Questionnaire-9; ERA-12 = Expectations Regarding Aging questionnaire

The results of the analysis of variables associated with higher HIV stigma (≥ 50th percentile = score ≥ 30) are shown in Table 2. The unadjusted prevalence ratios were statistically significant in the following groups: female compared to male, individuals of non-white race compared to white race, those with heterosexual orientation compared to non-heterosexual, individuals with < 8 years of education compared to those with ≥ 12 years, and those with a shorter duration of HIV infection compared to more than 20 years living with HIV. However, in the adjusted analysis, the associations were generally attenuated and no longer statistically significant, except for the comparison between 10 and 19 years since HIV diagnosis versus ≥ 20 years (higher stigma scores) and PLHIV aged ≥ 65 years versus 50–55 years.

Table 2.

Prevalence, unadjusted and adjusted prevalence ratios estimating the associations between independent variables and HIV Stigma Score ≥ 30

Variable	HIV Stigma Score ≥ 30*
	Yes	No	Unadjusted estimates		Adjusted estimates
	%	%	PR	95% CI	aPR	95% CI
Age, years
50–55	54.4	45.6	Ref.	–	Ref.	–
56–60	56.7	43.3	1.04	0.85–1.27	1.08	0.85–1.37
61–65	59.1	40.9	1.09	0.88–1.34	0.99	0.76–1.31
66–70	47.0	53.0	0.86	0.69–1.07	0.74	0.55–0.99
≥ 70	49.6	50.4	0.91	0.72–1.15	0.63	0.43–0.91
Sex
Male	49.0	51.0	Ref.	–	Ref.	–
Female	60.7	39.3	1.24	1.08–1.42	1.02	0.81–1.28
Race
White	45.7	54.3	Ref.	–	Ref.	–
Non-white	59.2	40.8	1.29	1.12–1.50	1.07	0.87–1.31
Sexual orientation
Non-heterosexual	43.3	56.7	Ref.	–	Ref.	–
Heterosexual	60.5	39.5	1.39	1.20–1.62	1.18	0.92–1.50
Education
≥ 12 years	46.8	53.2	Ref.	–	Ref.	–
8–11 years	52.4	47.6	1.12	0.94–1.32	1.00	0.79–1.25
< 8 years	66.4	33.6	1.42	1.20–1.68	1.17	0.90–1.53
Regular substance use&
No	51.1	48.9	Ref.	–	Ref.	–
Yes	50.0	50.0	0.98	0.80–1.19	0.96	0.79–1.17
Time since HIV diagnosis
≥ 20 years	46.7	53.2	Ref.	–	Ref.	–
10–19 years	63.8	36.2	1.37	1.17–1.60	1.27	1.03–1.57
5–9 years	63.8	36.2	1.37	1.10–1.70	1.09	0.79–1.52
< 5 years	66.7	33.3	1.43	1.13–1.81	1.15	0.82–1.61

*HIV Stigma Score was treated as a binary outcome for clinical interpretation (≥ 50th percentile = ≥ 30)

^& Regular substance use defined as daily alcohol/tobacco/cannabis or any use of other illicit substances

95% CI = 95% confidence intervals; PR Prevalence ratio; aPR Adjusted prevalence ratio; Ref Reference; Values in bold indicate statistical significance (p < 0.05)

Table 3 shows the results of analyses investigating the association between HIV stigma (quartiles defined as ≤ 25, 26–29, 30–34, and ≥ 35) and expectations of aging. In models adjusted for age, sex, race, sexual orientation, and substance use, only fewer years of education and higher HIV stigma scores were associated with worse expectation of aging. In mediation analysis, we found that the indirect effect of a 1-unit increases in the HIV stigma score through depression results in an average decrease of 0.15 in ERA-12 score. This indirect pathway, consistent with depression as a mediator, accounted for 25% of the overall association between HIV stigma and perceptions of aging (Fig. 1).

Table 3.

Unadjusted and adjusted beta (β) coefficients associated with higher (more positive) expectation regarding aging

Variable	Expectations regarding aging score (ERA-12)*
	Unadjusted estimates		Adjusted estimates
	β coefficient	95% CI	β coefficient	95% CI
Age, years
50–55	Ref.		Ref.
56–60	− 1.82	− 6.25, 2.61	− 1.75	− 7.25, 3.75
61–65	− 3.43	− 8.14, 1.27	− 3.15	− 8.81, 2.51
66–70	0.16	− 4.22, 4.54	− 1.12	− 6.77, 4.53
≥ 70	− 0.94	− 5.86, 3.97	− 2.40	− 8.20, 3.40
Sex
Male	Ref.		Ref.
Female	0.06	− 3.15, 3.27	1.46	− 3.66, 6.58
Race
White	Ref.		Ref.
Non-white	− 2.55	− 5.60, 0.49	0.77	− 3.25, 4.78
Sexual orientation
Non-heterosexual	Ref.		Ref.
Heterosexual	0.95	− 2.05, 3.96	3.66	− 1.44, 8.76
Education
≥ 12 years	Ref.		Ref.
8–11 years	− 4.41	− 7.76, − 1.05	− 6.03	− 10.40, − 1.65
< 8 years	− 7.93	− 11.96, − 3.90	− 7.60	− 13.40, − 1.81
Time since HIV diagnosis
≥ 20 years	Ref.		Ref.
10–19 years	− 1.39	− 8.59, 3.80	1.57	− 2.91, 6.06
5–9 years	2.20	− 2.90, 7.29	3.85	− 2.39, 10.08
< 5 years	− 2.39	− 8.60, 3.80	2.87	− 4.19, 9.93
Regular substance use&
No	Ref.		Ref.
Yes	1.42	− 2.51, 5.36	0.85	− 3.16, 4.85
HIV Stigma Score‡
< 25	Ref.		Ref.
25–29	− 7.53	− 11.78, − 3,28	− 6.29	− 11.73, − 0.84
30–34	− 7.23	− 11.47, − 3.00	− 7.29	− 12.53, − 2.04
≥ 35	− 11.14	− 15.58, − 6.69	− 12.05	− 17.84, − 6.26

^*Positive coefficients indicate higher ERA-12 scores and more positive perception of aging

^‡ Higher score denotes higher HIV-related stigma, assessed by 12-item Short HIV Stigma scale, values corresponding to quartiles

^& Regular substance use defined as daily alcohol/tobacco/cannabis or any use of other illicit substances

95% CI = 95% confidence intervals; Ref = reference; Values in bold indicate statistical significance (p < 0.05)

Fig. 1.

Effect of HIV stigma score on expectation of aging, mediated by depression. HIV stigma score was assessed by 12-item Short HIV Stigma scale. Aging expectations was defined using Expectations regarding aging (ERA-12). Depression symptoms were evaluated through Patient Health Questionnaire-9 (PHQ-9). Coeff = Coefficient; 95% CI = 95% confidence intervals; p = p-value

Figure 2 provides a forest plot with coefficients and 95% confidence intervals obtained in the multivariable analysis of the four domains of HIV stigma scale (personalized stigma, disclosure concerns, concerns about public attitudes, negative self-image) and perception of aging, adjusted for age, sex, race, sexual orientation, substance use, and time since HIV diagnosis. All HIV stigma domains were inversely associated with poorer perception of aging.

Fig. 2.

Forest plot with adjusted coefficients of the association between the four domains of HIV stigma score (12-item Berger scale) and aging expectations (ERA-12) from stratified analyses (reference: first quartile of each domain score). CI Confidence interval, Q1 First quartile and reference (< 25%); Q2 Second quartile (25–50%); Q3 Third quartile (51–75%); Q4 Fourth quartile (≥ 75%)

Sensitivity analyses comparing imputation methods for the HIV stigma scale and alternative modeling approaches (continuous vs. dichotomized) for the mentioned variable are presented in Supplementary Tables S1 and S2, with no statistically significant differences observed. In another sensitivity analysis adjusting for study site in analysis, we did not observe significant differences, suggesting minimal between-site heterogeneity.

Discussion

This cross-sectional study of 702 Brazilian adults living with HIV aged ≥ 50 years found that individuals with higher HIV stigma scores and lower levels of education may have a less positive view of aging. Higher HIV stigma scores were observed among women, non-white individuals, heterosexual individuals, those with lower education, and those with shorter duration of HIV diagnosis. Our findings also showed high prevalence of any depression symptom and that depression symptoms may account for 25% of the relationship between HIV stigma and aging expectations. Our study is novel in its examination of how HIV stigma is associated with positive or negative expectation regarding aging in a large cohort of older PLHIV.

Even in the current era, HIV stigma continues to have a significant impact on adults aging with HIV. Various studies have highlighted the ongoing challenges in disclosing HIV status, with individuals often anticipating feelings of rejection and discrimination. Our study demonstrated the barrier in disclosing HIV status through the frequent unanswered questions at HIV stigma scale assuming previous disclosure. In unadjusted analyses, a range of sociodemographic factors, such as female sex, non-White race, heterosexual orientation, lower education, and shorter time since HIV diagnosis, were associated with higher stigma. Emlet et al. [22] reported inconsistent findings across their previous studies regarding the characteristics most associated with higher HIV stigma, particularly sex and sexual orientation. Sandelowski et al. [23] highlighted that women living with HIV, positioned at the intersection of key axes of difference—including sex, race, social class, and vulnerability to HIV exposure—encounter significant barriers in securing a stable social identity and achieving a sense of belonging within their social communities. Aligned with this, Costa et al. [24] found a high prevalence of HIV stigma among Brazilian women, with approximately 70% reporting experiences of stigma linked to their serological status. The author suggests that sexism and sex-based violence, both prevalent issues in Brazil, may contribute to these findings. Recent work by Kokorelias et al. [25] reinforces the complex and intersectional nature of aging with HIV, particularly among women. Their findings demonstrate that HIV stigma in older adults does not occur in isolation, but is deeply intertwined with gender, race, socioeconomic position, and broader social inequities.

In multivariable analyses, several demographic and personal characteristics were not associated with HIV stigma as observed in univariate analyses. In this context, older age emerged as an associated factor only in the multivariable model. This finding may be justified by a stronger sense of group identity, specially related to the HIV epidemic period, along with the perception of HIV as a chronic condition like other long-term health issues. One of the possible explanations highlighted in a recent meta-analysis [18] for older individuals experiencing less stigma is the knowledge they have acquired about their HIV-related clinical condition over the years of follow-up in specialized care services. We hypothesized that the divergence between unadjusted and adjusted findings is related to complex intersectional relationships that could not be encountered by this study. It is important to mention that the intersectionality between HIV stigma and other existing inequities such as sexism, ageism, racism, homophobia and transphobia creates a new level of inequality negatively that impacts overall health [26] and it is not always possible to address in quantitative studies. Unfortunately, our cohort only included a validated measure of HIV-related stigma, and no additional stigma constructs were systematically assessed. Therefore, we were unable to directly evaluate intersectional stigma.

Expectations of aging is a distinctive gerontological concept that refers to an individual's perceptions of their future physical, mental and cognitive health and serve as fundamental elements in the context of successful aging [27]. In the general population, negative future self-views is associated with a decline in overall well-being [28, 29] and, according to Lin et al. [30], low ERA-12 score may be linked to negative aging stereotypes and ageism. Among PLHIV, data on this topic remain limited. Few studies evaluated the self-perception of aging and only one qualitative study involving 35 participants has explored its relationship with HIV stigma [31]. The association between chronological age and expectations regarding aging did not follow a strictly linear pattern in our cohort, which likely reflects the complex and multifactorial nature of aging perceptions among older PLHIV [10, 32]. Although the magnitude varied across HIV stigma subdomains, associations were conceptually consistent and showed the same overall direction.

This is the first study to report that higher HIV stigma is independently associated with worse expectations of aging, highlighting an important intersection of stigma and stereotypes. Another revealing finding of the present study is the evidence that one of the processes across which stigma is associated with worse expectation of aging is through depression symptoms. According to our data, approximately 25% of the association between HIV stigma and ERA-12 scores was consistent with an indirect pathway through depressive symptoms, a finding that may reflect the relevance of depression screening and management among older PLHIV and should be interpreted as hypothesis-generating. Other studies have also observed depression as an important mediator linking HIV stigma to other health outcomes. Earnshaw et al. [33] observed that depressive symptoms acted as a partial mediator in the association between HIV stigma and substance use. HIV stigma and depression likely have a bidirectional relationship since the avoidance behavior in reveling HIV status can lead to consequences such as social isolation and loneliness, which directly affect mental health [6, 34].

While novel in its investigation of HIV stigma, aging expectations, and depression among older PLHIV, our study has several limitations to consider. The cross-sectional design raises the possibility of reverse causation, especially considering the relationship between stigma, depression and aging expectations. The mediation analysis relies on strong assumptions that cannot be fully satisfied in a cross-sectional design, including temporal ordering and the absence of unmeasured confounding; therefore, temporality cannot be established and residual confounding may persist. In addition, important contextual variables relevant to the study aims were not available, including social support, income, and measures of stigma beyond HIV-specific stigma, which may limit the interpretation of our findings. It is also possible that the study has information bias due as the questionnaires were performed by an interviewer, perhaps resulting in underreporting of stigma and depression symptoms, in particular. Although the instruments used in the present study have been extensively validated in the literature, the inherent limitations of quantitative surveys limit appreciation of the subjectivity and complexity of HIV stigma and aging expectation. Translated versions may face cultural limitations, and its ability to accurately reflect such a complex issue is inherently constrained. The first dimension of the HIV stigma scale (personalized stigma) assumes that the serological status has already been disclosed, addressing consequences for personal, intimate, and familial relationships. However, as we have observed, HIV diagnosis often remains confidential. We attempted to address this limitation by imputation of values during the analysis. Finally, while the ERA-12 scale has been validated in the general population, has been rarely utilized among PLHIV. This may limit its ability to fully capture the unique aspects of aging in this population and may potentially underestimate their expectation regarding aging. Despite these limitations, the study benefits from a large sample size and data from three cities in different macro-regions, enhancing its epidemiological relevance.

Conclusion

Until now, no study evaluated the relationship between HIV stigma and expectations regarding aging in older adults living with HIV. Our findings identify factors associated with stigma experiences and more negative aging expectations. Longitudinal studies are needed to clarify the temporal dynamics of mediation through depression and the impact on clinical outcomes.

Author contributions

ELEA-Brasil team conceived of the presented idea. MBA and JLC analyzed the data. MBA wrote the manuscript with support from JLC. All authors provided critical feedback and helped shape the research, analysis and manuscript.

Funding

This work was supported by the NIH-funded Longitudinal Study of HIV & Aging in Brazil (ELEA-Brasil) (R01AG071439). Study data were collected and managed using REDCap electronic data capture tools hosted at Vanderbilt University Medical Center (VUMC) (UL1 TR000445). TST was financed in part by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq#311871/2021-6), and Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ # E-26/201.270/2022).

Data availability

The datasets generated and analyzed during the current study are not publicly available. Collaboration in data analysis and publication is welcomed through the submission of research proposals and concept sheets, which should be sent to ELEA investigators for review. Additional information can be obtained by emailing ELEA@vumc.org.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Vanderbilt University, the Brazilian National Ethics Committee (CONEP, protocol number 55131622.1.1001.5262), and the Institutional Review Board of Record at each participating site. All participants provided informed consent before any study procedures, and all identifiable information were maintained confidentially throughout the study. Prior to inclusion, participants underwent an assessment conducted by a qualified healthcare professional to determine their ability to understand the nature and consequences of the research. All participants included in the study were able to provide consent independently.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.