Antiretroviral Therapy–Related Experiences Predict Mental Health Status Among People Living with HIV in China: A Prospective Observational Cohort Study

Background: Antiretroviral therapy (ART) coverage has increased sharply. This prospective observational cohort study investigated the association between experiences related to ART and mental health problems among a sample of people living with HIV in China. Methods: The participants were 400 people living with HIV who had started ART for the rst time in Guangzhou city. They were followed-up one-year after ART initiation. Probable depression and moderate/severe anxiety were measured at baseline and Month 12, while experiences related to ART were measured at Month 6. Univariate and multivariate logistic regressions were used to explore the predictors of mental health status. Results: Among the 300 participants (75.0%) who completed all three surveys, a signicant decline in prevalence of probable depression (23.0% at baseline vs. 14.0% at Month 12, P = 0.002) and moderate/severe anxiety (14.7% at baseline vs. 8.7% at Month 12, P = 0.023) was observed during the follow-up period. After adjustment for mental health status and potential confounders at baseline, a number of ART-related experiences at Month 6 were associated with probable depression and/or moderate/severe anxiety measured at Month 12. Improved physical health, relationships with sexual partners, and self-condence were associated with decreased mental health issues, while the side effects of ART, AIDS-related symptoms, and inconvenience in daily life due to ART use were associated with increased mental health issues. Conclusions: ART-related experiences were predictors of mental health problems, tailored mental health promotion interventions targeting these experiences are needed.

mental health problems have potentially severe clinical implications for PLWH. Evidence suggests that depression and anxiety are associated with low medical adherence [16], worse retention in HIV care [17,18], and poor HIV-related outcomes (e.g., quality of life) [19,20]. A meta-analysis has shown an overall prevalence of depression and anxiety of 33.6% and 28.4%, respectively, among PLWH on ART [21]. Mental health problems have been found to be more prevalent among PLWH in low-and middle-income countries [21]. In China, the prevalence of depression was reported as about 40% among PLWH on ART [22,23].
Given these ndings, the prevention and treatment of mental health problems may be an important element of HIV treatment and care.
A growing body of literature has demonstrated improved mental health status among PLWH after ART initiation. A longitudinal study in the United States showed that, compared to pre-treatment levels, the prevalence of depression decreased from 30-21% after one year on ART [24]. It is possible that the bene ts of ART are mediated via a reduction in in ammatory pathways that affect the risk of depression, such as the kynurenine pathway or tryptophan catabolism [25]. ART may also indirectly affect mental health status among PLWH by reducing internalized stigma, increasing community support and access to primary care and psychological support services, and enhancing control of other comorbidities (e.g., tuberculosis and diabetes) [26,27]. Previous studies have identi ed a number of factors that are associated with mental health problems among PLWH on ART, including gender [28], age [22], personal income level [29], marital status [22,28], living arrangements [30], insurance [24], comorbidities [28], sleep quality [23], perceived stigma and discrimination [23,28], and family support [23]. However, most of these studies were cross-sectional and could not identify the predictors of mental health problems after PLWH start taking ART.
ART requires lifelong treatment, which brings both positive and negative experiences for users. On the one hand, ART brings signi cant clinical bene ts for PLWH, such as improving physical health status, facilitating immune system recovery (e.g., higher CD4 cell counts), and slowing disease progression [31,32]. There is also evidence that viral suppression by ART greatly reduces the risk of transmitting HIV to sexual partners [1,2], potentially improving relationships between PLWH and their sexual partners. On the other hand, although advances in ART have largely reduced its side effects, recent studies have shown that about 25.0-53.3% of PLWH still experience severe side effects in their rst year after ART initiation [33,34]. Moreover, the daily regimen of ART is inconvenient for PLWH. For example, 25% of PLWH in eight high-income countries agreed that being tied to daily medication limited their day-to-day life, and 29% felt stressed by the need to take their medication at the right time every day [35]. These experiences related to ART may also in uence mental health among PLWH. Previous studies found that the side effects of ART, lowered CD4 cell counts, and AIDS-related symptoms after ART initiation were associated with depression and anxiety [22,29,30]. However, these studies were limited by their cross-sectional nature and the lack of inclusion of potentially relevant experiences of ART.
As far as we know, no longitudinal study has investigated ART-related experiences and how they predict mental health problems among PLWH. To address these gaps, this prospective cohort study measured ART-related experiences at Month 6 and mental health problems at Month 12 among a sample of PLWH in China who had initiated ART for the rst time, and investigated whether ART-related experiences (Month 6) could predict mental health problems (Month 12) after considering potential confounders.

Study population and procedure
This prospective cohort study was conducted in Guangzhou, China, from June 2016 to May 2018. Faceto-face interviews were conducted at baseline (the rst day of ART admission), and at Month 6 and Month 12 after ART admission. The inclusion criteria were that the subjects were 1) aged 18 years or above, 2) HIV seropositive, 3) initiating ART for the rst time, and 4) willing to attend surveys at baseline, Month 6, and Month 12. Those who had major psychiatric illnesses (schizophrenia or bipolar disorder) or could not communicate with the interviewers were excluded.
The participants were recruited from the ART clinic of Guangzhou Eighth People's Hospital in Guangzhou. The Guangzhou Eighth People's Hospital was the rst ART centre in Guangdong province and has provided ART to over 10,000 PLWH since 2004 [36]. During the recruitment period (June 2016 to May 2017), trained nurses screened the new ART users for eligibility and referred those who met the criteria to the research team. The trained interviewers then recon rmed the participants' eligibility, explained the details of the study, and assured them that refusal would not affect their right to use any related services and that they could quit the study at any time without being questioned. After providing written informed consent, the participants completed a self-administered questionnaire which took about 15 minutes. They were invited to visit the hospital again to complete two other self-administered questionnaires after six and twelve months. On completion of each of the three surveys, the participants received a souvenir (worth 15 RMB or 2 USD) in appreciation of their time. Ethical approval for the study was obtained from the Institutional Review Board (IRB) of the School of Public Health, Sun Yat-sen University, Guangzhou, Mental health problems measured at baseline and Month 12 Depressive symptoms were measured by the validated Chinese version of the Patient Health Questionnaire (PHQ-9) [37,38]. The PHQ-9 has been widely used to screen for depression in different Chinese populations [37,38], including PLWH [39], and has acceptable reliability and validity. The scale contains nine items regarding the frequency of certain negative feelings over the past two weeks (response categories: 0 = never, 1 = sometimes, 2 = often, 3 = almost every day). A score of ≥ 10 was used to de ne probable depression [24,29].
Generalized anxiety disorder was measured by the validated Chinese version of the Generalized Anxiety Disorder Scale (GAD-7), which has shown acceptable reliability and validity in prior research [40]. In this study, a score of ≥ 10 was used to de ne moderate/severe anxiety [41].
Experiences related to ART measured at Month 6 Four items measured positive experiences related to ART. The participants were asked to report whether they had had the following experiences during the past six months: 1) a rise in CD4 cell count, 2) improvement in physical health, 3) improvement in relationships with sexual partners, and 4) regained self-con dence (response categories: 1 = yes, 0 = no). Negative experiences related to ART in the past six months were also measured by four items: 1) side effects of ART, 2) AIDS-related symptoms, 3) inconvenience of ART use in daily life, and 4) exposure of HIV-positive status due to ART use (response categories: 1 = yes, 0 = no).

Baseline characteristics
The participants were asked about their background characteristics, including age, sex, education, current marital status, monthly personal income, employment status, and city of permanent residence. HIV/AIDSrelated characteristics were also recorded, including AIDS-related symptoms, CD4 cell count at the most recent test, route of HIV transmission (sexual behaviour with same-sex partner, sexual behaviour with opposite-sex partner, intravenous drug use, mother-to-child transmission, blood transfusion, occupational exposure, or not sure), and ART regime.
Psychosocial status including social support and social stigma was also measured at baseline. Social support was measured by the validated Chinese version of the Multidimensional Scale of Perceived Social Support (MSPSS) [42,43]. The MSPSS contains 12 items and 3 dimensions including family, friends, and signi cant others. The score range for each dimension is 1-7, with a higher score indicating better support. We used the social stigma module of the Chinese Courtesy Stigma Scales (CSSSs) to measure the social stigma patients experienced [44]. The scale contains 13 items, with a score range of 13-52, a higher score representing greater stigma.

Statistical Analysis
The baseline characteristics of those who completed all follow-up surveys and those who did not were compared using a chi-square test (for categorical variables), independent samples t-test (for continuous variables), or non-parametric test (for ranked variables). Further analysis was performed among those who completed all three surveys: baseline, Month 6, and Month 12 (n = 300).
The difference in mental health problems between baseline and Month 12 was tested by the pairedsample chi-square test. Associations between baseline characteristics and probable depression and moderate/severe anxiety measured at Month 12 were described by the odds ratio (OR) obtained from univariate logistic regression models. Multivariate stepwise logistic regression models were used to identify baseline predictors that independently affected mental health status at Month 12. To test whether ART-related experiences measured at Month 6 could predict future mental health problems (probable depression and moderate/severe anxiety measured at Month 12), both univariate and adjusted logistic regression models were applied, adjusted for baseline variables with P < 0.1, plus baseline mental health status. R software version 4.0.2 was used for the data analysis, and P < 0.05 was considered statistically signi cant.

Baseline characteristics of participants
Of the 400 participants, most were male (92.2%), more than 25 years old (79.6%), currently single (55.8%), and non-permanent residents of Guangzhou (77.7%). About 60% of the participants reported that they became infected through sexual behaviour with a same-sex partner. Over 80% used a combination of tenofovir, lamivudine, and efavirenz (84.0%) as their ART regimen. At baseline, the prevalence of probable depression and moderate/severe anxiety was 26.0% and 15.0%, respectively.
Those who did not complete both follow-up surveys were more likely to be older, have a lower or no xed monthly personal income, have probable depression at baseline, and not have been infected through sexual behaviour with a same-sex partner (Table 1). Experiences related to ART measured at Month 6 Among participants who completed all three surveys (n = 300), most reported a rise in CD4 cell count (90.0%), improvement in physical health status (80.3%), and regained self-con dence (74.7%). Some also reported improvement in relationships with sexual partners since ART initiation (21.7%). On the negative side, 76.3% experienced inconvenience in daily life due to ART use and 50.7% suffered side effects of ART. Some participants also reported AIDS-related symptoms (18.3%) and exposure of their HIV-positive status due to ART use (22.7%).
Predictors of probable depression and moderate/severe anxiety measured at Month 12 Participants who were female, received less support from friends and signi cant others, perceived greater social stigma, and had probable depression and moderate/severe anxiety at baseline were more likely than others to have probable depression at Month 12. Participants who were younger, had lower monthly personal income, received less support from friends, and perceived greater social stigma at baseline were more likely to have moderate/severe anxiety at Month 12 (Table 2). Stepwise regression models showed that sex, support from family and signi cant others, and probable depression at baseline affected depression at Month 12 independently. Support from friends and social stigma were independent predictors of moderate/severe anxiety at Month 12.  (Table 3). -- † 0.05 < P < 0.1; * P < 0.05; ** P < 0.01; *** P < 0.001 --: Not signi cant 1 OR a : Adjusted odds ratios, adjustment for mental health status measured at baseline (probable depression and moderate/severe anxiety) and baseline background variables with P < 0.1 (sex, monthly personal income, friend support, signi cant other support, and social stigma).
2 OR a : Adjusted odds ratios, adjustment for mental health status measured at baseline (probable depression and moderate/severe anxiety) and baseline background variables with P < 0.1 (age group, sex, monthly personal income, friend support, signi cant other support, and social stigma).

Discussion
Our research is one of the few longitudinal studies to investigate the association between ART-related experiences and mental health problems among PLWH who have recently started ART. In our study, the prevalence of probable depression and moderate/severe anxiety both at baseline and at Month 12 was lower than that reported by meta-analyses targeting PLWH in China [11,45]. This discrepancy may be attributable to the broad range of measurements used in different studies and to differences in the populations studied (e.g., gender identity, age, region, treatment status) [22,23,46]. In studies using the same measurements as in our study, the prevalence of depression was 39.3% among newly diagnosed PLWH [47], and the prevalence of anxiety was 64.3% among those who had newly initiated ART [48]. Similar to previous longitudinal studies [24,49], our study showed a signi cant decline in the prevalence of mental health problems at Month 12 compared to pre-treatment. Our study also showed that PLWH who had probable depression when starting ART were more likely to drop out. Among those who remained in the cohort, baseline mental health status was a signi cant predictor of mental health problems one year after starting ART. Given the negative impact of mental health problems on PLWH [50], screening for mental health problems and related care is greatly warranted in ART clinics.
This study provides healthcare workers with insights for promoting mental health among PLWH who are starting ART, especially regarding segmentation. According to social marketing approaches, careful segmentation improves the effectiveness of health promotion programs [51]. Similar to previous studies, females are found to be at higher risk of mental health problems after receiving ART [50,52]. As female PLWH may suffer higher levels of internalized stigma and experience more social stigma than male patients [50,52], more attention should be focused on female PLWH in future mental health promotion programs. Furthermore, PLWH without xed incomes or with relatively low incomes should be prioritized in future programs, as they not only tend to drop out of programs but also to have a higher prevalence of mental health problems one year after receiving ART. Although ART is provided for free to all PLWH in China [53], PLWH still need to pay for compulsory physical examinations every three months. This nancial burden may become a stressor, especially for those with lower incomes.
Similar to the ndings of cross-sectional studies [12,54], we found that stronger social support and lower perceived social stigma at baseline were protective factors for mental health problems. A study among Chinese PLWH also showed that family and social networks, and the trust/intensity of relationships with family members and others played important roles in mental wellbeing [55]. In addition, studies have shown that living in a trusting social environment with lower stigma toward PLWH alleviates daily stressors and protects against mental health problems [55][56][57]. On the one hand, strategies to encourage and teach PLWH to communicate with their family/friends for support are needed [58]. On the other hand, public health education programs aimed at reducing discrimination against PLWH may help provide PLWH on ART with more social support to promote their mental health.
Our ndings highlight the in uence of ART-related experiences on mental health problems among PLWH, as a number of positive/negative experiences signi cantly or marginally signi cantly predicted probable depression and/or moderate/severe anxiety after controlling for potential baseline confounders.
Consistent with previous ndings, side effects of ART may trigger depressive and anxiety symptoms [22,23,30]. About half of our participants experienced some side effects at Month 6. These side effects bring physical discomfort to PLWH on ART, interrupting their normal daily and social functioning and reducing quality of life [34]. Thus, the development of effective interventions to facilitate self-management of side effects is greatly needed. Participants with AIDS-related symptoms were more likely to develop probable depression, in line with previous studies [29,30]. A review has reported a signi cant association between depressive symptoms and HIV progression [59]. The signi cant association may be due to the chronic impact of HIV on immune and disease-related parameters, which in turn may lead to depressive symptoms among PLWH on ART [60]. We found that about three quarters of the participants reported inconvenience in daily life due to ART use, and this was a predictor of moderate/severe anxiety. Taking ART every day may be a reminder of their HIV status and make them feel stressed [35]. Scaling up the single tablet regimen and reducing dose frequency could thus help to improve mental health among PLWH on ART.
Relatively few participants reported an improvement in their sex-partner relationship, but for those who did, it was protective against depressive and anxiety symptoms. Health communication messages should therefore be disseminated to PLWH and their sexual partners emphasizing that viral suppression by ART greatly reduces the risk of transmitting HIV through sex [1,2]. This may remove concerns about resuming sexual behaviour and improve intimacy between sexual partners. Most participants regained selfcon dence during ART, and this was also protective against depressive symptoms. Studies have shown that ART makes people feel ready to return to normal life [34]. Our ndings demonstrate the clinical bene ts of the new "treat all" policy, as more than 80% of PLWH experienced improvements in physical health, which contributed to improved mental health, and nearly 90% experienced a rise in CD4 cell count.
However, our study has some limitations. First, as participants were recruited by non-probabilistic sampling, the ndings may not be generalizable to all PLWH in China. Second, there may have been attrition bias, as those more likely to drop out of the study were older, had a lower/un xed income, were not infected through homosexual behaviour, and had depression at baseline. The prevalence of mental health problems at Month 12 may therefore be underestimated. Third, selection bias may exist, as we could not obtain the characteristics of those who refused to join the study. Moreover, the results were selfreported and thus subject to social desirability bias.

Conclusions
In conclusion, compared to the pre-treatment period, a signi cant decline in the prevalence of probable depression and moderate/severe anxiety was observed among Chinese PLWH one year after receiving ART. Experiencing ART side effects, AIDS-related symptoms, and inconvenience in daily life due to ART use were risk factors for mental health problems, while improvements in physical health, relationships with sexual partners, and self-con dence were found to be protective. Tailored mental health promotion interventions targeting these experiences are needed.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests