Association of experienced and internalized stigma with self-disclosure of HIV status by youth living with HIV

Abstract

We examined patterns of disclosure among youth living with HIV (YLHIV) in Kenya, and the association between self-disclosure and antiretroviral therapy adherence, stigma, depression, resilience, and social support. Of 96 YLHIV, 78% were female, 33% were ages 14–18, and 40% acquired HIV perinatally. Sixty-three (66%) YLHIV had self-disclosed their HIV status; 67% to family and 43% to non-family members. Older YLHIV were 75% more likely to have self-disclosed than those 14–18 years. Of the 68 either married or ever sexually active, 45 (66%) did not disclose to their partners. Those who had self-disclosed were more likely to report internalized stigma (50% vs. 21%, prevalence ratio [PR] 2.3, 1.1–4.6), experienced stigma (26% vs. 3%, PR 11.0, 1.4–86), and elevated depressive symptoms (57% vs. 30%, PR 1.8, 1.0–3.1). The association with stigma was stronger with self-disclosure to family than non-family. Support should be provided to YLHIV during self-disclosure to mitigate psychosocial harms.

Introduction

HIV status self-disclosure, defined as a person autonomously sharing their HIV status with another person(1) may improve social support and adherence to antiretroviral therapy (ART).(2, 3) In addition, self-disclosure to sexual partners may have public health benefits, as it increases likelihood of condom use, safer sex practices, and preventing transmission.(4) Previous studies have described the benefits of self-disclosure to persons living with HIV (PLHIV), including building trust in romantic relationships, gaining access to greater interpersonal support for treatment adherence, improved access to care, and viral suppression.(3–6) Evidence on the association between self-disclosure and mental health outcomes is, however mixed. Some studies link self-disclosure with adverse outcomes including feelings of isolation,(7) anxiety,(4, 5) negative self-concept,(6) internalized self-blame,(8) and depressive symptoms.(8, 9) Other studies, however, have linked self-disclosure with improved psychological well-being and emotional relief,(10, 11) less isolation,(12) and fewer symptoms of mental illness,(13, 14) including lower depressive symptoms.(15)

Self-disclosure is challenging for youth living with HIV (YLHIV).(1, 7, 16) Fear of rejection and discrimination, lack of communication skills, and low self-efficacy to deal with negative reactions after disclosure are frequently cited as challenges.(5) Disclosure to sexual partners presents heightened concerns, including intimate partner violence, emotional harm, and social exclusion and ostracization.(7, 16) Establishing peer relationships during adolescence is a priority,(17) and disclosure presents a potential risk that could jeopardize those relationships.(18) Few quantitative studies in sub-Saharan Africa have reported prevalence of self-disclosure among youth. Estimates of prevalence of self-disclosure among YLHIV vary widely (11%−100%), depending on age, gender, geographic location and the relationship with the potential disclosure target.(1, 16, 19–21) Data on how disclosure to different people (family, friends or intimate partners) affects YLHIV’s mental health is scarce.

Adolescence is a period of rapid and large changes in social environment, physical state, and neurocognitive development.(22) It is also the peak age of onset of many mental illnesses. The reasons for elevated risk of mental illness in adolescence are incompletely understood, but are thought to be due to age-specific changes in cognitive capacity, risk-taking behavior, and relationships with family and peers.(23, 24) Youth living with HIV are at a higher risk of adverse mental health outcomes compared to their counterparts not living with HIV.(25, 26) Prevalence of depression(27) and anxiety(28) in this population is high. Anticipated stigma (expectation that one will be discriminated against due to one’s HIV status), internalized stigma (acceptance of negative characterization) and experienced stigma (lived experience of discrimination)(29, 30) among YLHIV are also associated with adverse mental health outcomes, such as negative self-concept, feelings of isolation, depression, and anxiety.(31, 32) Poor mental health is associated with poorer adherence to treatment.(26) Treatment adherence among YLHIV is poorer compared to adults, predisposing them to lower viral suppression, opportunistic infections and even mortality.(33) Protective factors against poor mental health include presence of good social support,(34) especially from family and peers, and development of resilience,(35) which have also been shown to positively impact treatment adherence.

Few studies have been conducted in sub-Saharan Africa (SSA) on YLHIV mental health, self-disclosure, stigma and the associations between them; yet close to 90% of all YLHIV reside in SSA.(36). In this study, we examined patterns of disclosure among YLHIV in Nairobi, Kenya, and the association between self-disclosure and antiretroviral therapy (ART) adherence, stigma, depressive symptoms, resilience, and social support.

Methods

Study design and population

This was a cross-sectional study nested in the social media for ART adherence and retention in adolescents and young adults (Vijana-SMART) pilot study. The Vijana-SMART study developed and evaluated a WhatsApp-based support group for YLHIV in Kenya to improve retention in care and treatment adherence. Two sets of youth were sequentially enrolled in the study. The first group of 55 youth were enrolled and followed-up in a 6-month control period during which they continued to receive standard HIV care. The second group (some overlapping with the control group) were enrolled and included in a 6-month WhatsApp group in which they received pre-developed content on adolescent health and HIV care as well as peer support and facilitated discussions by a healthcare worker. Participants ages 14–24 were enrolled between December 2017 and April 2019 from Kayole II sub-County hospital’s HIV clinic, which serves a low-income area in Nairobi, Kenya. Eligible participants were YLHIV, aware of their HIV status and had access to a smartphone. Recruitment data from the parent study showed that approximately half of all YLHIV in this age range attending the clinic had access to a smartphone.(37)

Study procedures

Recruitment and enrollment

YLHIV were either recruited by study staff in the waiting area of the HIV clinic, referred by clinic staff after their visit, or referred through snowball sampling, where enrolled YLHIV were provided with flyers describing the study and referred their peers outside the clinic. Potential participants completed a screening questionnaire, administered by study staff using the mobile data collection platform, Open Data Kit (ODK), to establish eligibility.

Data collection and measures

Eligible YLHIV completed a questionnaire, administered by study staff using ODK, during enrollment. The questionnaire ascertained sociodemographic characteristics, age at which they learned their HIV status, and mode of HIV acquisition (with the following choices: got HIV as a baby, sexual transmission, blood transfusion, accidental or intentional needle injection, don’t know, and no answer). Participants were considered to have acquired HIV perinatally if they reported knowing they acquired HIV as a baby or if they did not know how they acquired HIV but were ≤14 years when they learned their HIV status or ≤14 years when they initiated ART.(38) They were considered to have acquired HIV behaviorally if they reported knowing they acquired HIV through sex, blood transfusion, or accidental or intentional needle injection, or if they did not know how they acquired HIV, but were >14 years old when they learned their HIV status or >14 years when they initiated ART.

Disclosure.

Participants were asked who knew their HIV status (any disclosure), and whom they had told about their status (self-disclosure) with choices of no one, father, mother, sibling, and other relative (categorized as family), friend, neighbor, clergy, teacher, support group member, other (categorized as non-family).(16)

ART adherence.

Participants answered 3 questions to rate their ART adherence on Likert scales: number of doses taken in last 1 month, how well they felt they did with taking medication, and how often they felt they took their medication as prescribed. The 3-item ART adherence Wilson scale has good psychometric properties.(39) To define ART adherence, each of 3 self-report measures was converted to a 100-point scale and a mean calculated, as outlined by Wilson et al.(39) Good adherence was defined as a score above 80 on the Wilson scale. Internal consistency for the current study was acceptable (Cronbach’s α = 0.75).

Stigma.

Questionnaires also assessed anticipated stigma (1 question), experienced stigma (2 questions) and internalized stigma (3 questions), using 6 of the 10 items in the validated adolescents living with HIV stigma scale (ALHIV-SS) from South Africa.(30) Examples of questions were: ‘People in my community think that a person with HIV is disgusting’, ‘Sometimes having HIV makes people feel contaminated and dirty inside’, ‘I have been teased because of my HIV’. To define stigma, YLHIV who answered any questions affirmatively in each category of the stigma scale (anticipated, experienced, and internalized), were classified as having the outcome in that category. The internal consistency for the questions used to assess experienced and internalized stigma in this study was good (Cronbach’s α: 0.81 and 0.91 respectively). We used only one question to assess anticipated stigma and so could not assess internal consistency.

Depression.

Depressive symptoms were assessed using the patient health questionnaire (PHQ-9), which has previously been used among people living with HIV in Kenya.(41) Mild or higher elevation of depressive symptoms was defined as a score ≥5, while moderate or higher elevation of depressive symptoms was defined as a score ≥10 on PHQ-9. The internal consistency in the current study was also good (Cronbach’s α = 0.88)

Resilience.

Resilience was assessed using an abbreviated Connor-Davidson resilience scale,(42) where high resilience was defined as a score above the median on the resilience scale. The internal consistency for the current study was however moderate (Cronbach’s α = 0.65).

Social support.

Social support from family and non-family members was assessed using an abbreviated social support behaviors (SSB) scale.(43) High social support was defined as a score above the median on social support scores in both the family and non-family domains. The internal consistency for the social support scale used in this study was good (Cronbach’s α = 0.89).

Statistical analysis

We described sociodemographic characteristics, romantic relationship status, mode of HIV acquisition, HIV treatment and disclosure history using counts and proportions. We calculated the proportion of YLHIV who had any disclosure and self-disclosure categorized by age (14–18 and 19–24 years age bands) and non-exclusive categories of disclosure to family and non-family.

Analysis of determinants of self-disclosure

We assessed associations between age, gender, relationship status, mode of HIV acquisition, and duration on ART as exposure variables, and self-disclosure as the outcome variable using generalized linear models (GLM) with binomial family and log link, to provide prevalence ratios (PR) and 95% confidence intervals (95% CI). We adjusted for age in all models except where age was the exposure of interest, due to the established positive association between older age and self-disclosure.(13, 45) If models failed to converge, GLM with poisson family and log link was used.

Analysis of association between self-disclosure, clinical and behavioral outcomes

We assessed the association between self-disclosure as an exposure variable with seven behavioral outcomes: ART adherence, stigma (anticipated, experienced and internalized), depressive symptoms, resilience, and social support. We reported the counts and proportion of YLHIV with the outcomes and the 95% CI for each proportion.

We assessed the association between self-disclosure (exposure variable) and the 7 outcomes using GLM (binomial family and log link), to provide PR and 95% CI. We a priori adjusted for age and gender in all models due to previously established association between mental health,(15, 46) stigma, social support,(47) and age and gender among persons living with HIV. To determine the strength of association between various outcomes and disclosure to family versus non-family, we developed GLM models for each outcome with the exposure variable being self-disclosure with three mutually exclusive categories (no disclosure, only family disclosure, and only non-family disclosure). We excluded those who had disclosed to both family and non-family due to their low number (n=4). We reported age- and gender-adjusted PR and 95%CI. Data cleaning and analysis was done using R Studio (Version 1.1.456, 2009–2018).

Results

Sociodemographic characteristics

A total of 96 YLHIV were enrolled, of whom 76 (78%) were female and 64 (67%) were ages 19–24. All were on ART at enrollment, with 83% having initiated ART within the last 5 years. Forty percent had learned their HIV status by age 14, and were classified as perinatally acquired infections, and 68 (76%) reported being married, in a romantic relationship or sexually active (Table 1).

Table 1:

Sociodemographic and HIV-related factors associated with self-disclosure for youth living with HIV

Characteristic	Overall N=96 n (%)	Self-disclosed (N=63) *n (%)	Unadjusted PR (95% CI)	Age adjusted PR (95% CI)	Wald statistic, P Value***
Age
14–18	32 (33)	14 (44)	Reference
19–24	64 (67)	49 (77)	1.75(1.15–2.65)		7.00, 0.008
Gender
Male	20 (21)	12 (60)	Reference	Reference
Female	76 (79)	51 (67)	1.12(0.76–1.65)	0.94(0.49–1.82)	0.03, 0.86
^Romantic relationship status					^^2.00, 0.36
Never in a relationship	21 (24)	7 (33)	Reference	Reference
Ever in a relationship	44 (46)	30 (68)	2.05(0.90–4.66)	1.86(0.71–4.86)	1.60, 0.20
Married	24 (25)	20 (83)	2.50(1.06–5.91)	2.20(0.74–6.54)	2.00, 0.16
^Duration on ART					^^0.57, 0.75
< 1 year	26 (29)	17 (65)	Reference	Reference
1–5 years	48 (54)	36 (75)	1.15(0.83–1.59)	1.03(0.57–1.86)	0.01, 0.93
5–10 years	15 (17)	7 (47)	0.71(0.39–1.31)	0.75(0.31–1.81)	0.41, 0.52
**Mode of HIV acquisition
Perinatal	36 (38)	18 (50)	Reference	Reference
Behavioral	60 (62)	45 (75)	1.50(1.05–2.15)	1.15(0.56–2.34)	0.15, 0.70

^*Row percentages

^{^}lower N due to missing data

^**Based on reported age of acquisition

^***Age-adjusted (except age predictor)

PR – Prevalence ratio

^{^^}p value from Global Wald test

Bolded: Association is statistically significant at p<0.05

Disclosure characteristics

Of 96 YLHIV, 87 (91%) reported that someone knew their HIV status, with 59 (61%) reporting only a family member, 19 (20%) only a non-family member, and 9 (9%) both family and non-family knew. Disclosure was more common in younger participants; all 32 YLHIV age 14–18 reported someone knew their HIV status, in contrast with 55 (86%) YLHIV ages 19–24. Sixty-three (66%) YLHIV had self-disclosed their HIV status, 42 (67%) to a family member, and 27 (43%) to non-family. While disclosure was higher in younger adolescents, self-disclosure to anyone was higher in older youth: only 14/32 (44%) YLHIV 14–18 years had self-disclosed, compared to 49/64 (77%) YLHIV ages 19–24 (PR 1.75, 95%CI: 1.15–2.65). Only 10 (31%) of participants ages 14–18 had self-disclosed to family, while 32 (50%) of participants ages 19–24 had self-disclosed to family. Further, 4 (13%) of those ages 14–18 had self-disclosed to non-family, while 23 (36%) of participants ages 19–24 had self-disclosed to non-family. Of 68 YLHIV who reported being married, in a romantic relationship or sexually active, 45 (66%) did not disclose to their partners (Table 1).

Description of outcomes and their association with self-disclosure

Seventy YLHIV (74% [95%CI: 64–83]) reported good adherence (≥80% score on the Wilson scale). Fifty-one (58% [47–68]) reported experiencing anticipated stigma, 17 (18% [11–28]) experienced stigma, and 38 (40% [30–51]) internalized stigma. Forty-six YLHIV (48% [38–58]) had mild or higher elevation in depressive symptoms (PHQ-9 score ≥5) and 19 (20% [13–29]) had moderate or higher elevation in depressive symptoms (PHQ-9 score ≥10). Thirty-six YLHIV (38% [29–49]) had above-median resilience scores and 30 (31% [22–42]) above-median social support scores.

We found significant age- and gender-adjusted associations between self-disclosure and prevalence of internalized and experienced stigma. Thirty-one (50%) YLHIV who had self-disclosed reported internalized stigma, compared with 7 (21%) who had not self-disclosed (aPR 2.3, 1.1–4.6). Similarly, 16 (26%) YLHIV who had self-disclosed reported experienced stigma, compared with 1 (3%) who had not self-disclosed (aPR 11.0, 1.4–86.0). There was a trend for mildly elevated depressive symptoms among YLHIV who had self-disclosed, 36 (57%) versus 10 (30%) with elevated depressive symptoms, (aPR 1.8, 1.0–3.1) though the association with moderate to severe depressive symptoms was weaker (24% vs. 12%, aPR 1.6, 0.6–4.6). No associations were found between self-disclosure and ART adherence, anticipated stigma, resilience, and social support. (Figure 1).

Figure 1.

Association between self-disclosure and outcomes: adherence, mental health and social support. aPR Age and gender adjusted prevalence ratios, W Wald statistic, 95% CI 95% confidence interval.

Association between self-disclosure and outcomes

Of 63 YLHIV who had self-disclosed, 36 (57%) had self-disclosed only to family, 21 (33%) to only non-family, while 6 (10%) had self-disclosed to both family and non-family. Those who had self-disclosed to only family were 13 times more likely to report experienced stigma compared to those who had not self-disclosed to anyone (aPR 13, 1.7–105.0). They were also 2.5 times more likely to report internalized stigma than those who had not self-disclosed (aPR 2.5, 1.0–6.0). No significant associations were detected between self-disclosure to non-family and mental health outcomes (Table 2).

Table 2:

Association between self-disclosure to family and non-family members, and outcomes (adherence, mental health and social support)

Outcomes	No self-disclosure (N=33) n (%)	Self-disclosure to family (N=36) n (%)	*aPR (95%CI)	Wald statistic, p value	Self-disclosure to non-family (N=21) n (%)	*aPR (95%CI)	Wald statistic, p value
Good adherence, N=88	20 (61)	27 (75)	1.1 (0.6–1.9)	0.05, 0.83	18 (83)	1.2 (0.6–2.3)	0.23, 0.63
Anticipated stigma, N=83	18 (58)	21 (55)	0.9 (0.4–1.7)	0.14, 0.71	12 (63)	1.0 (0.5–2.2)	<0.01, 0.98
Experienced stigma, N=87	1 (3)	11 (32)	13 (1.7–105)	6.00, 0.01	3 (14)	6.4 (0.6–64)	2.50, 0.11
Internalized stigma, N=89	7 (21)	19 (54)	2.5 (1.0–6.0)	4.00, 0.05	8 (38)	1.8 (0.6–5.1)	1.20, 0.28
Depression, N=90	10 (30)	20 (56)	1.7 (0.8–3.7)	1.70, 0.19	11 (52)	1.5 (0.6–3.8)	0.91, 0.34
High resilience, N=87	9 (29)	15 (46)	1.2 (0.5–2.9)	0.26, 0.61	9 (41)	1.2 (0.5–3.2)	0.15, 0.70
High social support, N=90	7 (21)	15 (42)	1.9 (0.8–4.9)	1.90, 0.16	4 (19)	0.8 (0.2–2.8)	0.14, 0.71

^*Age and gender adjusted prevalence ratios for outcomes following self-disclosure to family versus no self-disclosure, and self-disclosure to non-family versus no self-disclosure

aPR – Adjusted prevalence ratio

Bolded: Association is statistically significant (p<0.05)

Discussion

This study enrolled majority female, older adolescents and young adults living with HIV, on ART, mostly initiated in the last 5 years. Most YLHIV reported that someone, more commonly a family member, knew their HIV status. The majority also reported having self-disclosed to someone, more commonly a family member. We reported a higher prevalence of self-disclosure among YLHIV (66%), than previous studies in Kenya (48%),(20) Uganda (38%−56%),(16, 19) and South Africa (33%−38%).(21, 48) The finding from the current study that younger adolescents, likely having acquired HIV perinatally, had lower rates of self-disclosure, is consistent with findings from studies in Thailand(49) and the United States(13, 45) that specifically studied adolescents who acquired HIV perinatally. Our study, consistent with several studies around the world,(14, 49–51) reported higher prevalence of self-disclosure to family members than to other persons. A few studies however have previously reported a lower prevalence of self-disclosure to family than friends among young men.(52–54) We did not observe differences by gender in our cohort. The inclination to disclose more to family members may be because youth spend more time with family and trust family members more than friends. This provides an opportunity for implementing family-focused interventions.

This study identified a high proportion of YLHIV (66%) who were married or had been in romantic relationships or sexually active, but had not disclosed their HIV status (no difference by sex), which is consistent with existing literature.(55, 56) As demonstrated in previous studies, non-disclosure by youth who are sexually active presents challenges to engagement in safer sex practices.(21, 57, 58) Elevated rates of virologic non-suppression and low condom use in this age group mean this has important population-level implications for prevention of onward transmission.(59) This highlights that YLWH are a critical group to target for assisted disclosure,(60) as well as strategies to reduce transmission such as condom promotion and ART adherence support to achieve undetectable (untransmittable) viral load.(61)

Younger adolescents (14–18) were less likely to have self-disclosed than their older counterparts, which may be explained by a larger proportion having acquired HIV perinatally(62) so their families already knew their status. Self-disclosure to non-family was also lower in younger adolescents, which may simply be a function of having had less opportunity to disclose due to their age. However, disclosure to friends and teachers has been found to be challenging to younger adolescents in previous studies due to anticipated stigma and lack of self-efficacy to disclose.(56) Teachers especially may be important support persons for treatment adherence by adolescents in boarding schools.(63) For those who acquired HIV perinatally, there are also additional barriers to self-disclosure including implications for disclosure of parental HIV status. (56) Targeting this group for skills development for disclosure could improve their disclosure outcomes.

Previous studies have demonstrated moderately high levels of HIV-related stigma among YLHIV.(16) Anticipated stigma in particular is a major reason for HIV status non-disclosure in this population.(52, 56) Our data showed that over half of YLHIV anticipated stigma from their community, with similar levels in those who had and had not self-disclosed. This could mean that for YLHIV, their perception of discrimination by their families and the wider community due to their HIV status does not improve after self-disclosure. In fact, a higher proportion of YLHIV who had self-disclosed reported experienced and internalized stigma than those who had not self-disclosed, consistent with previous studies in similar populations.(5) We further observed that disclosure to family was associated more strongly with experienced and internalized stigma than disclosure to non-family. YLHIV’s experiences of stigma following self-disclosure may heighten anticipated and internalized stigma, which may prevent future disclosure. These findings highlight the importance of engaging a more holistic treatment approach for YLHIV that includes information to their close contacts on how they can support the youth and their treatment. Interventions to reduce stigma in the community(64) are also needed to complement efforts to reduce HIV-related stigma in individual YLHIV and their families.

Our observation that self-disclosure was associated with a trend for elevated depressive symptoms highlights the interpersonal and psychological risks associated with self-disclosure for YLHIV. While this association was not strong, the finding is supported by results of previous studies.(5) Further, in our study there was no difference in perceived social support among those who had self-disclosed and those who had not, in contrast to results from previous studies that demonstrated an association between social support and self-disclosure.(4, 5, 65) Social support may be an important mechanism through which self-disclosure impacts adherence, experience of stigma and depression. Those who disclose their status and receive support may be more likely to adhere to treatment, experience less stigma and have fewer depressive symptoms.(14, 66, 67) The lack of association between social support and self-disclosure in our data may explain the lack of association between self-disclosure and adherence, and the observed association with higher stigma and depression. Mental health support, including screening, treatment, and prevention of depressive illnesses is required for YLHIV, especially when disclosing to others.(68) These results should be viewed within the context of a high prevalence of mental health problems among Kenyan youth, especially YLHIV,(40) and their communities.(69) HIV is also significantly stigmatized among communities in Kenya, with misconceptions persisting about its modes of transmission and its association with sexual immorality.(70) The individual challenge of disclosure is further worsened by poor access to psychosocial support and mental health care,(71) and, for some, gender-based violence following disclosure.(72) Clinicians working with YLHIV need to help YLHIV navigate interrelated issues of HIV stigma and depression and how these factors could influence decisions and impact self-disclosure.

Limitations

This study has some limitations. Generalizability is limited due to the urban setting and only including participants who had access to smartphones. While the snowball sampling strategy ensured that an adequate sample was recruited, it may also have led to overrepresentation of groups of related people in the population and thus reduced the representativeness of the participants to the general population of YLHIV. Our cross-sectional analysis was unable to determine causal relationships and our small sample size may have limited our ability to detect associations between self-disclosure and some mental health outcomes. In addition, we did not collect information on time since disclosure. There is likely a process of recovery from the initial disruption of disclosure, which we were unable to account for in our analysis. We also did not ask participants how they perceived self-disclosure affected them, how stigma affected their ART adherence, or the duration of their romantic relationships, which would have provided additional context to these findings. Due to questionnaire length limitations, we employed abbreviated versions of validated scales to assess stigma, resilience and social support that have not previously been used in YLHIV. However, the questions used showed good psychometric properties. Lastly, the classification of YLHIV as having acquired HIV perinatally or behaviorally using age cut-offs may have resulted in misclassification. However, being one of very few studies that has studied multiple mental health outcomes and disclosure patterns in Kenya and the region, it provides useful data to develop hypotheses for larger longitudinal studies.

Implications and future directions

In summary, our data suggest that the benefits of self-disclosure as a public health initiative to reduce onward transmission may require more nuanced consideration.(73, 74) Since the rollout of the ‘U=U’ campaign,(75) global awareness has grown that suppressive ART essentially eliminates risk of onward transmission.(61) This knowledge combined with our observations that self-disclosure carries risks of adverse psychosocial impact may create confusion in YLHIV about the benefits of disclosure. Guidelines from the World Health Organization(76) and Kenya’s local guidelines(77) on adolescent-friendly care of YLHIV recognize the potential harms of disclosure and advocate for additional support to YLHIV to develop self-efficacy to self-disclose and mitigate psychosocial harm. In addition to providing disclosure support to youth, there is a need for clear messaging and frank discussion with YLHIV considering disclosure about the balance between potential benefits like increased social support and trust, and potential harms like increased stigma and depression, contextualized in the YLHIV’s individual circumstances, including their virologic suppression and sexual relationship characteristics. To balance the known benefits of disclosure to support adherence with potential risks of stigma, we propose that disclosure to key support people be considered if it is deemed crucial for adherence. Moreover, discussions between HCW and youth about disclosure should be individualized and integrated with mental health support.

More research is needed to understand the impact of disclosure, including qualitative studies exploring youth’s individual perspectives on the costs and benefits of self-disclosure, longitudinal studies to understand if and how youth recover from the negative impacts of disclosure, and epidemiologic and modeling studies to assess what public health benefit there is to promoting disclosure. Moreover, studies are needed that develop care models that are informed by youth’s needs as they consider disclosure.

Conclusion

In this study, self-disclosure was more common among older youth, and more commonly to a family member. A high proportion of YLHIV reported anticipated and internalized stigma and mildly elevated depressive symptoms, while a moderate number reported experienced stigma, poor adherence, high resilience and high social support. We found significant associations between self-disclosure and elevated internalized stigma and experienced stigma, and a trend for association with elevated depression symptoms. Self-disclosure, while encouraged, should be carefully considered for individual YLHIV, with stigma reduction, screening for depression, and adherence support being integrated with disclosure support interventions. Our findings that youth experience stigma from families to whom they disclose rather than support is concerning. This highlights the need for HIV programs and HCW to prioritize individual- and community-level stigma reduction interventions, especially those targeting important support persons for YLHIV such as families and friends, to mitigate the negative psychological impact of disclosure.