Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Aug 1.
Published in final edited form as: AIDS Care. 2020 Apr 26;32(8):1014–1022. doi: 10.1080/09540121.2020.1728213

Anticipated Stigma and Medication Adherence among People Living with HIV: The Mechanistic Roles of Medication Support and ART Self-efficacy

Chengbo Zeng a,b,*, Xiaoming Li a,b, Shan Qiao a,b, Xueying Yang a,b, Zhiyong Shen c, Yuejiao Zhou c
PMCID: PMC7368809  NIHMSID: NIHMS1560897  PMID: 32336130

Abstract

This study aimed to examine the relationship between anticipated stigma and medication adherence as well as the mechanistic roles of medication support and ART self-efficacy. Data were derived from the baseline assessment of a prospective cohort study in Guangxi, China. A total of 1,198 PLWH were recruited and assessed on their sociodemographic characteristics, medication adherence, anticipated stigma, medication support, and ART self-efficacy. Path analysis was used to examine the direct effect from anticipated stigma to medication adherence and indirect effects through medication support and ART self-efficacy. Path model revealed that the indirect effect from anticipated stigma to medication adherence was statistically significant while the direct effect was not significant. Anticipated stigma could influence medication adherence through ART self-efficacy but not through medication support. The serial mediating effect of medication support and ART self-efficacy on the relationship between anticipated stigma and medication adherence was significant. Anticipated stigma affects medication adherence among PLWH through its adverse impacts on medication support and ART self-efficacy. Tailored interventions promoting medication support and ART self-efficacy may alleviate the negative influence of anticipated stigma on medication adherence among PLWH. Additionally, policy efforts aiming to reduce stigma against PLWH and increasing medication support are warranted to improve medication adherence among PLWH.

Keywords: Anticipated stigma, Medication support, ART self-efficacy, Medication adherence, HIV, China

Introduction

With the availability of antiretroviral therapy (ART), the mortality of people living with HIV (PLWH) has been decreased, and their life expectancy has been increased (UNAIDS, 2018). PLWH can control their symptoms and disease progression if they sufficiently adhere to ART (Nieuwkerk & Oort, 2005). The general criteria for optimal medication adherence is the completion of no less than 90% of the prescription, which corresponds to the best treatment effectiveness (Fogarty et al., 2002; Press, Tyndall, Wood, Hogg, & Montaner, 2002; Raffa et al., 2008). However, many patients do not strictly adhere to ART. A recent review of 84 observational studies found that the average rate of ≥ 90% medication adherence was 62% (Ortego et al., 2011).

Multiple factors could influence medication adherence among PLWH. These factors include intrapersonal factors (e.g., age, self-efficacy), social/environmental factors (e.g., HIV-related stigma, social support), medication-related factors (e.g., access to medication, side effects), and healthcare delivery systems-related factors (e.g., patient-provider relationship, health education/information) (Langebeek et al., 2014; Reda & Biadgilign, 2012). Among these factors, HIV-related stigma is one of the well-investigated factors in medication adherence (Rao et al., 2012; Turan et al., 2017; Zhou, Li, Qiao, Shen, & Zhou, 2018). Therefore, to design effective intervention strategies and improve medication adherence, it is imperative to elucidate the underlying mechanisms between HIV-related stigma and medication adherence.

Stigma was traditionally defined as “a significantly discrediting trait” and further expanded as a “discrediting and tainting social label” which could have negative impacts on health outcomes (Alonzo & Reynolds, 1995; Goffman, 1963). In the case of PLWH, HIV-related stigma causes this population to lose social value and makes them vulnerable to health problems, such as depression, substance use, and suicidal behaviors (Bruce, Ramirez-Valles, & Campbell, 2008; Rao et al., 2012; Zeng et al., 2018). According to the Health Stigma Framework (HSF), PLWH may anticipate prejudice, stereotypes, and discrimination from others, which may further influence their health outcomes (Earnshaw & Chaudoir, 2009; Earnshaw, Smith, Chaudoir, Amico, & Copenhaver, 2013). Thus, exploring the mechanisms of anticipated stigma on medication adherence may have important implication in improving health outcomes among PLWH (Earnshaw & Chaudoir, 2009; Earnshaw et al., 2013; Turan et al., 2017).

Anticipated stigma refers to the degree to which PLWH expect that they will be stigmatized by others in the future (Earnshaw & Chaudoir, 2009; Earnshaw et al., 2013; Turan et al., 2017). PLWH who anticipate stigma from others would have less interactions with people around them and reduced opportunities for obtaining social support (Turan et al., 2017). Additionally, PLWH who anticipate stigma from the social environment would avoid to conduct healthcare behaviors, such as medication adherence and help-seeking behaviors (Turan et al., 2017). For instance, Turan and colleagues (2017) found that anticipated stigma was inversely associated with social support and medication adherence.

Social support, including medication-related social support, is an important facilitator for medication adherence (DiMatteo, 2004). Assistance and support from community/social network have been implicated in improving adherence to medical treatment by encouraging motivation and self-efficacy and giving practical assistance (DiMatteo, 2004). Indeed, perceived emotional and practical support from social network can improve medication adherence through self-efficacy, which is a key factor for initiating and maintaining a specific behavior (DiMatteo, 2004; Nafradi, Nakamoto, & Schulz, 2017; Zhang et al., 2016). For instance, using data from a longitudinal study, Simoni and colleagues (2006) found that social support was positively associated with greater spirituality and less negative affect, which in turn were related to adherence self-efficacy. Baseline adherence self-efficacy could predict medication adherence at 3-month follow-up among PLWH (Simoni, Frick, et al., 2006). Thus, social support, including medication support, might lead to better medication adherence through its impact on adherence self-efficacy. However, there is still a dearth of studies investigating the mechanisms between anticipated stigma and medication adherence through the serial mediating effect of social support and self-efficacy among PLWH.

To address the knowledge gaps, this study aimed to examine the relationship between anticipated stigma and medication adherence as well as the mechanistic roles of medication support and ART self-efficacy. We hypothesized that: a) anticipated stigma was inversely associated with medication adherence; b) both medication support and ART self-efficacy could mediate the relationship between anticipated stigma and medication adherence; and c) the serial mediating effect of medication support and ART self-efficacy on the relationship between anticipated stigma and medication adherence was statistically significant. Figure 1 shows the hypothetical model.

Figure 1.

Figure 1.

Open in a new tab

Hypothetical Model

Methods

Study sites and participants

Data was derived from the baseline survey of an ongoing prospective cohort study. The baseline survey was conducted from November 2017 to February 2018 in Guangxi, China. With the assistance and collaboration of Guangxi Center for Disease Control and Prevention (CDC), six hospitals/clinics in five cities with the largest number of HIV patients under care were selected as study sites.

The inclusion criteria for the cohort study were: 1) 18–60 years of age; 2) a confirmed diagnosis of HIV; and 3) having no plan to permanently relocate outside of the Guangxi province in the next 12 months. The exclusion criteria were: 1) linguistic, mental or physical inability to respond to assessment questions; and 2) currently incarcerated or institutionalized for drug use or commercial sex. With an overall refusal rate of about 5%, a total of 1,198 PLWH were recruited in this study and completed the baseline assessment.

Data collection

The interviewer-administered questionnaire was used for data collection. Medical staff or HIV case managers at the HIV clinics referred potential participants to research term. Local team members screened PLWH for eligibility and discussed the prospective participants about the benefits and risks of the study and invited them to participate. After obtaining their written informed content, the face-to-face interviews were conducted in private rooms of the clinics. Each participant received a gift (e.g., household items) equivalent to US$5.00 (1 USD≈6.5 Chinese Yuan at the time of the survey) at the completion of the questionnaire interview. The study protocol was approved by the Institutional Review Boards at both University of South Carolina in the United States and Guangxi CDC in China.

Measures

Sociodemographic characteristics

Participants provided information on their sociodemographic characteristics including age, gender (0=Female, 1=Male), ethnicity (0=non-Han, 1=Han), employment (0=Unemployed, 1=Part time, 2=Full time), levels of education (0=Illiteracy/primary school, 1=Middle school/high school, 2=Higher education), monthly income in Chinese currency (0=0~1,999 RMB, 1=2,000~3,999 RMB, 2=4,000 RMB~), and marital status (1=Married/cohabited, 0=Others).

Medication adherence

Due to the challenges of evaluating self-report ART adherence, a multiple-item approach was used to measure medication adherence (Simoni, Kurth, et al., 2006). Five measures were used to inquiry participants about their levels of medication adherence. An overall score was calculated according to the following procedure. First, participants reported the days of completing their prescribed ART doses within four specified timeframes (last three days, last weekend, last two weeks, and last month [30 days]). These responses were then converted into a percentage of prescribed doses and recoded into 0 (<90%) or 1 (≥ 90% of prescribed doses). Second, whether participants have ever missed a prescribed dose before (1=Yes, 0=No) was used as the fifth adherence measure. Finally, by summing scores of the above five measures, a composite score was calculated. The sum score ranged from 0 to 5, with higher value indicating higher medication adherence. Cronbach alpha for the five measures was 0.73. As the categorical measure of medication adherence has clinical implication, we also dichotomized this measure using 5 as the cutoff point. PLWH who scored 5 would be considered as optimal adherence to ART while those who scored less than 5 would be considered as suboptimal to ART.

Anticipated stigma

A 9-item scale from the Health Stigma Framework (HSF) was adapted to measure anticipated stigma in this study (Earnshaw et al., 2013). The anticipated stigma scale assesses the expectation of HIV stigma from family members, community, and healthcare providers. The sample items were “family members will avoid touching me”, “community managers will refuse to provide me with social services”, and “healthcare providers will treat me with less respect”. Item response options were from “definitely not (1)” to “definitely (5)”. The sum score was used as a composite score ranging from 9 to 45, with a higher score indicating a higher level of anticipated stigma. The internal consistency estimate (Cronbach alpha) for this scale was 0.92.

Medication support

Medication support was assessed using an 11-item scale that was developed in the cohort study. PLWH was asked whether there was someone helping them adhere to ART in the recent one month. Sample items were “remind you of adhering to ART” and “discuss ART with you”. Items were scored from “never (1)” to “almost always (5)”. The sum score was used as a composite score ranging from 11 to 55, with a higher score indicating more medication support. This scale showed a good internal consistency (Cronbach alpha=0.96).

ART self-efficacy

ART self-efficacy was assessed using a 10-item scale adapted from the Self-efficacy for Appropriate Medication Use Scale (SEAMS) (Risser, Jacobson & Kripalani, 2007; Zhang et al., 2016). PLWH were asked “how confident are you that you can take your medicines correctly under the following scenarios”? Sample items were “when you take several different medicines each day” and “when you are away from home”. Item response options were from “not confident (1)” to “confident (3)”. The sum score of these ten items was used as a composite score ranging from 10 to 30, with a higher score indicating a higher level of ART self-efficacy. The Cronbach alpha of this scale was 0.96.

Statistical analysis

First, descriptive statistics were reported on sociodemographic characteristics and variables of interest. Mean (standard deviation or SD) was used to describe continuous variables, and frequencies (percentages) were used to describe categorical variables.

Second, bivariate analyses were performed to examine the relationship between medication adherence and sociodemographic characteristics, anticipated stigma, medication support, and ART self-efficacy using spearman correlation analyses for continuous variables and Wilcoxon rank-sums tests or Kruskal-Wallis tests for categorical variables.

Third, adjusting for covariates that were significantly associated with medication adherence in bivariate analyses, path model analysis was conducted to examine hypothetical model. Valente and colleagues (2016) demonstrated that when the variables of interest were not normally distributed, bias-corrected bootstrap method had the largest statistical power to detect significance. As medication adherence was not normally distributed, the direct and indirect effects from anticipated stigma to medication adherence were examined using bias-corrected bootstrap procedure based 1,000 bootstrap samples (Wang & Wang, 2012). Bias-corrected confidence intervals for the direct and indirect paths were reported (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002; Valente et al., 2016). As a default approach to examine mediating effect in Mplus software, delta z score was used to evaluate the mediating effects of medication support and ART self-efficacy (Muthen & Muthen, 2017). Maximum likelihood method (estimator=ML in Mplus) was used to estimate the parameters. Missing values for predictors were handled using Full Information Maximum Likelihood (FIML) Method. The same analysis was applied to examine the model with dichotomized measure of medication adherence, and weighted least square mean and variance method (estimator=WLSMV in Mplus) was used to estimate the model parameters.

Multiple indices were used to evaluate goodness of fit of the path model. These indices included χ2/df, Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR), and Weighted Root Mean Square Residual (WRMR). χ2/df < 3, CFI>0.95, RMSEA≤0.06, SRMR≤0.08, and WRMR≤1.00 indicate a good model fit (Wang & Wang, 2012).

Descriptive statistics, bivariate analyses, and correlation analyses were performed using SAS software version 9.4 (SAS Institute, Inc., Cary, NC). Path model and mediation analyses were performed using Mplus version 7.0 (Muthen & Muthen, Los Angeles, CA).

Results

Descriptive statistics

Among the 1198 PLWH, the mean age of them was 39.0 (SD=9.19) years old (Table 1). More than half of the individuals were male (64.36%, 771/1198), Han ethnicity (65.08%, 779/1197), full time employed (61.65%, 733/1189), having middle school/high school level of education (58.93%, 706/1198), and married/cohabited (56.24%, 667/1186). Most of the participants were having monthly income less than 4,000 RMB (79.83%, 954/1195). The mean level of medication adherence was 4.68 (SD=0.78). There were 80.3% (962/1198) of the participants having optimal medication adherence.

Table 1.

Socio-demographic characteristics and bivariate analyses

Variables N(%) Missing
value(%)		 Medication adherence
(Mean±SD)		 p-value
Total 1198 (100.0) - 4.68 (0.78) -
Age(Mean±SD) 39.0 (9.19) 0 (0.00) 0.02a 0.409
Gender 0 (0.00) 0.858b
 Male 771 (64.36) 4.68 (0.81)
 Female 427 (35.64) 4.68 (0.76)
Ethnicity 1 (0.00) 0.552b
 Han 779 (65.08) 4.69 (0.77)
 Non-Han 418 (34.92) 4.67 (0.81)
Employment 9 (0.75) 0.072c
 Unemployed 221 (18.59) 4.63 (0.84)
 Part time 235 (19.76) 4.60 (0.90)
 Full time 733 (61.65) 4.73 (0.72)
Levels of education 0 (0.00) 0.032c
 Illiteracy /primary school 253 (21.12) 4.59 (0.86)
 Middle school /high school 706 (58.93) 4.71 (0.74)
 Higher education 239 (19.95) 4.69 (0.82)
Monthly income 3 (0.00) 0.119c
 0~1,999 RMB 369 (30.88) 4.63 (0.82)
 2,000~3,999 RMB 585 (48.95) 4.71 (0.74)
 4,000~ RMB 241 (20.17) 4.70 (0.80)
Marital status 12 (0.00) 0.139b
 Married/cohabited 667 (56.24) 4.66 (0.79)
 Others 519 (43.76) 4.71 (0.77)
Open in a new tab

Note: SD: Standard deviation

a:

Spearman correlation coefficient

b:

Wilcoxon rank sum test

c:

Kruskal-Wallis test.

Bivariate analyses

Bivariate analyses between sociodemographic characteristics and medication adherence are shown in Table 1. Levels of education was positively associated with medication adherence (p=0.032). Anticipated stigma was negatively associated with medication support (r=−0.18, p<0.001) and ART self-efficacy (r=−0.22, p<0.001). ART self-efficacy was positively related to medication adherence (r=0.17, p<0.001) (Table 2).

Table 2.

Correlation matrix

1 2 3 4
1. Anticipated stigma 1.00
2. Medication support −0.18*** 1.00
3. ART self-efficacy −0.22*** 0.17*** 1.00
4. Medication adherence −0.01 0.04 0.17*** 1.00
Missing value (%) 0 (0.00) 4 (0.01) 0 (0.00) 0 (0.00)
Mean (SD) 23.13 (7.75) 22.55 (11.39) 24.44 (5.02) 4.68 (0.78)
Open in a new tab

Note: SD: Standard deviation

***:

p<0.001.

Path analysis and path coefficients

After adjusting for levels of education, the final model with continuous measure of medication adherence showed a good model fit (χ2/df =4.06, CFI=0.98, RMSEA=0.05, SRMR=0.01). The model accounted for 2.9% (p=0.005) of the variance in term of medication adherence while 3.2% (p=0.004) for medication support and 7.1% (p<0.001) for ART self-efficacy. Path model reveals that anticipated stigma was negatively associated with medication support (std.β=−0.179, p<0.001) and ART self-efficacy (std.β=−0.194, p<0.001) but not significantly associated with medication adherence (std.β=0.026, p=0.411). Medication support was positively related to ART self-efficacy (std.β=0.128, p<0.001) but not to medication adherence (std.β=0.014, p=0.557). ART self-efficacy was positively related to medication adherence (std.β=0.166, p<0.001).

For the model with dichotomized medication adherence, it also showed a good model fit (χ2/df =3.54, CFI=0.98, RMSEA=0.05, WRMR=0.50). This model accounted for 5.5% of the variance in term of medication adherence while 3.3% (p=0.001) for medication support and 7.2% (p<0.001) for ART self-efficacy. Table 3 shows the path coefficients, and Figure 2 shows the final path model with continuous measure of medication adherence.

Table 3.

Path coefficients

Paths β std.β 95% C.I. S.E. p-value
Medication adherence (continuous variable)
AS---»MS −0.264 −0.179 −0.241~−0.124 0.047 <0.001
MS---»ARTS 0.057 0.128 0.073~0.183 0.012 <0.001
AS---»ARTS −0.126 −0.194 −0.248~−0.138 0.018 <0.001
ARTS---»MED 0.026 0.166 0.107~0.230 0.005 <0.001
MS---»MED 0.001 0.014 −0.035~0.060 0.002 0.557
AS---»MED 0.003 0.026 −0.037~0.085 0.003 0.411
Medication adherence (dichotomized variable)
AS---»MS −0.265 −0.181 −0.357~−0.175 0.039 <0.001
MS---»ARTS 0.057 0.128 0.034~0.081 0.013 <0.001
AS---»ARTS −0.126 −0.194 −0.160~−0.088 0.020 <0.001
ARTS---»MED 0.043 0.216 0.027~0.059 0.008 <0.001
MS---»MED 0.001 0.017 −0.005~0.010 0.004 0.699
AS---»MED 0.005 0.035 −0.006~0.016 0.006 0.423
Open in a new tab

Note: AS: Anticipated stigma; MS: Medication support; ARTS: ART self-efficacy; MED: Medication adherence.

Figure 2.

Figure 2.

Open in a new tab

Final Path Model

Note: ***: p<0.001. Standardized path coefficients are presented in the final path model. Levels of education is adjusted as covariates in this model. In figure 2, medication adherence is a continuous outcome.

Mediation analysis

Mediation analysis found that while the direct effect from anticipated stigma to medication adherence was not statistically significant (std.β=0.026, delta z=0.823, p=0.411), the indirect effect between them through medication support and ART self-efficacy was significant (std.β=−0.039, delta z=−3.999, p<0.001). Anticipated stigma was negatively associated with ART self-efficacy, which in turn was positively related to medication adherence (std.β=−0.032, delta z=−3.810, p<0.001). Although the indirect path from anticipated stigma to medication adherence through medication support was not statistically significant (std.β=−0.002, delta z=−0.582, p=0.561), the serial mediating effect from anticipated stigma to medication adherence through medication support and ART self-efficacy was significant (std.β=−0.004, delta z=−2.875, p=0.004). The same results were found in the model with dichotomized measure of medication adherence (Table 4).

Table 4.

Mediation analysis

Effect β std.β 95% C.I. S.E. p-value
Medication adherence (continuous variable)
Total effect −0.001 −0.013 −0.007~0.005 0.003 0.673
Indirect effect −0.004 −0.039 −0.006~−0.002 0.001 <0.001
 Path 1 −0.003 −0.032 −0.005~−0.002 0.001 <0.001
 Path 2 0.000 −0.002 −0.001~0.001 0.000 0.561
 Path 3 0.000 −0.004 −0.001~0.000 0.000 0.004
Direct effect 0.003 0.026 −0.004~0.009 0.003 0.411
Medication adherence (dichotomized variable)
Total effect −0.002 −0.015 −0.012~0.009 0.006 0.735
Indirect effect −0.006 −0.050 −0.010~−0.003 0.002 <0.001
 Path 1 −0.005 −0.042 −0.008~−0.003 0.001 <0.001
 Path 2 0.000 −0.003 −0.003~0.001 0.001 0.700
 Path 3 −0.001 −0.005 −0.001~0.000 0.000 0.003
Direct effect 0.005 0.035 −0.006~0.016 0.006 0.423
Open in a new tab

Note: Path 1: Anticipated stigma---»ART self-efficacy---»Medication adherence.

Path 2: Anticipated stigma---»Medication support---»Medication adherence.

Path 3: Anticipated stigma---»Medication support---» ART self-efficacy ---»Medication adherence.

Total effect=indirect effect + direct effect.

Discussion

Using a large representative sample of PLWH from Guangxi, China, this study examined the mechanisms among anticipated stigma, medication support, ART self-efficacy, and medication adherence in PLWH. Path model revealed that anticipated stigma could affect medication adherence through the indirect effects of medication support and ART self-efficacy while the direct effect between anticipated stigma and medication adherence was not statistically significant. The serial mediating effect from anticipated stigma to medication adherence via medication support and ART self-efficacy was significant. To the best of our knowledge, this was one of the first studies investigating the mechanism of anticipated stigma on medication adherence among PLWH.

Although indirect relationship was found, anticipated stigma was not directly associated with medication adherence, which was partially aligned with previous studies which found that HIV stigma could influence medication adherence directly and indirectly (Rao et al., 2012; Shrestha, Altice & Copenhaver, 2019; Turan et al., 2017). One of the possible reasons for the insignificant relationship between these two concepts may be the ceiling effect of medication adherence (Stirratt et al., 2015). Due to the “Four Frees and One Care” policy, PLWH in China can obtain free ART treatment and are required to have regular clinical check-ups every three or six months (Sun et al., 2010). In addition, case managers at local healthcare centers will provide PLWH with treatment-related support (e.g., reminders of ART and health education) (Ma et al., 2015; Shao, 2006). Thus, although PLWH still anticipate stigma in the future, most of them reported relatively good adherence to ART in this study. Another possible reason may be that anticipated stigma, as a chronic stress, could influence medication adherence indirectly through reducing the interaction between PLWH and others and their access of treatment-related resource (e.g., health information, health education) (Earnshaw & Chaudoir, 2009; Earnshaw et al., 2013; Turan et al., 2017).

The data in the current study revealed an insignificant total effect from anticipated stigma to medication adherence, although the indirect effect was significant. The total effect, as the sum of direct and indirect effects, might not be significant when the direct effect was insignificant and the indirect effect was not strong enough to dominate the total effect. There were two possible explanations for the small indirect effect between anticipated stigma and medication adherence. First, there might be some unmeasured and uncontrolled confounders (e.g., avoidance coping) that affected the associations among medication support, ART self-efficacy, and medication adherence (Loeys, Moerkerke & Vansteelandt, 2015; Turan, Fazeli, Raper, Mugavero & Johnson, 2016). Therefore, the indirect effect could not capture the effects of confounders and might be underestimated. Second, as the indirect effect is equal to the product of the paths from anticipated stigma to mediators (i.e., medication support and ART self-efficacy) and the paths from mediators to medication adherence, the indirect effect will decrease with the increasing number of mediators (Loeys, Moerkerke & Vansteelandt, 2015).

Anticipated stigma was indirectly related to medication adherence through ART self-efficacy but not through medication support. However, this study found that the indirect path between anticipated stigma and medication adherence via ART self-efficacy and the serial mediating effect from anticipated stigma to medication adherence through medication support and ART self-efficacy (“anticipated stigma→medication support→ART self-efficacy→medication adherence”) were significant. These findings emphasized that ART self-efficacy, as a key factor for behavioral change, played an important mechanistic role on the relationship between anticipated stigma and medication adherence under the scenario of adequate medication support and high adherence to ART (Zhang et al., 2016). These results were aligned with previous findings which found that self-efficacy could mediate the relationship between stigma and medication adherence among PLWH (Diiorio et al., 2009). Additionally, as medication support could improve ART self-efficacy and medication adherence of PLWH, both medication support and ART self-efficacy might mediate the negative relationship between anticipated stigma and medication adherence.

The data in the current study suggested that to increase medication adherence among PLWH, multilevel interventions reducing anticipated stigma and improving medication support and ART self-efficacy are needed. At the societal level, structural efforts should be made to promote public health education of HIV-related knowledge to HIV uninfected population, improve awareness, and reduce stigma towards PLWH (Tse & Huang, 2017; Wu et al., 2008). Policy efforts should be made at the societal level to maintain and/or improve both social support and medication support towards PLWH, to improve their psychological well-being (e.g., less perceived stigma and more self-efficacy) and adherence behavior. Although the “Four Frees and One Care” policy and case management services in China can provide policy and practical support for getting access and adhering to ART, these efforts have been constrained in rural areas due to the limited funding resource, inadequate manpower, and relatively low capacity for health service delivery (Ma et al., 2015). To promote and sustain policy effort and case management services in rural China, adequate governmental funding, planned team building, and professional skill training are needed (Ma et al., 2015). At the individual level, resilience-based intervention approach are needed to help PLWH cope with stigma, obtain medication support, and improve ART self-efficacy.

Although the current study is novel in exploring the mechanism of anticipated stigma underlying medication adherence, there are still some limitations to be acknowledged. First, self-report measures of medication adherence were employed in this study, and self-report bias might exist. Due to the self-report bias, the levels of medication adherence in this study might be overestimated (Stirratt et al., 2015). Future research with more objective assessments (e.g., biomarkers and device-based measurement) would benefit from less self-report bias. Second, the causal relationships cannot be warranted with cross-sectional data. Future longitudinal studies are needed to confirm the findings of this study. Third, the sample was representative for the rural PLWH in the current study sites, but cautions should be given when generalizing the findings from this study to other rural/urban areas in China as well as other developing and developed countries.

Conclusion

The results of this study shed light on the mechanistic roles of medication support and ART self-efficacy on the relationship between anticipated stigma and medication adherence among PLWH. Though medication support could not mediate the negative pathway between anticipated stigma and medication adherence, the serial mediating effect between these two variables through medication support and ART self-efficacy revealed that medication support could improve ART self-efficacy and alleviate the negative influence of stigma on adherence behavior among PLWH. To intervene the negative impact of anticipated stigma on medication adherence, intervention targeting policy and psychosocial factors to improve medication support and ART self-efficacy are warranted.

Acknowledgements

This study was supported by the National Institute of Health (NIH) Research Grant R01MH112376 by National Institute of Mental Health. The authors wish to thank all participants who gave of their time for the current study, and the reviewers for their helpful comments.

References

  1. Alonzo AA, & Reynolds NR (1995). Stigma, HIV and AIDS: An Exploration and Elaboration of A Stigma Trajectory. Soc Sci Med, 41(3), 303–315. doi.org/10.1016/0277-9536(94)00384-6. [DOI] [PubMed] [Google Scholar]
  2. Bruce D, Ramirez-Valles J, & Campbell RT (2008). Stigmatization, Substance Use, and Sexual Risk Behavior among Latino Gay and Bisexual Men and Transgender Persons. Journal of Drug Issues, 38(1), 235–260. doi: 10.1177/002204260803800111. [DOI] [Google Scholar]
  3. DiMatteo MR (2004). Social Support and Patient Adherence to Medical Treatment: A Meta-Analysis. Health Psychol, 23(2), 207–218. doi: 10.1037/0278-6133.23.2.207. [DOI] [PubMed] [Google Scholar]
  4. Diiorio C, McCarty F, Depadilla L, Resnicow K, Holstad M, Yeager K, Sharma SM, Morisky DE & Lundberg B (2009). Adherence to Antiretroviral Medication Regimens: A Test of A Psychosocial Model. AIDS Behav, 13(1):10–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Goffman E (1963). Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice Hall; 1963. [Google Scholar]
  6. Earnshaw VA, & Chaudoir SR (2009). From Conceptualizing to Measuring HIV Stigma: A Review of HIV Stigma Mechanism Measures. AIDS Behav, 13(6), 1160–1177. doi: 10.1007/s10461-009-9593-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Earnshaw VA, Smith LR, Chaudoir SR, Amico KR, & Copenhaver MM (2013). HIV Stigma Mechanisms and Well-being among PLWH: A Test of the HIV Stigma Framework. AIDS Behav, 17(5), 1785–1795. doi: 10.1007/s10461-013-0437-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fogarty L, Roter D, Larson S, Burke J, Gillespie J, & Levy R (2002). Patient Adherence to HIV Medication Regimens: A Review of Published and Abstract Reports. Patient Educ Couns, 46(2), 93–108. doi.org/10.1016/S0738-3991(01)00219-1. [DOI] [PubMed] [Google Scholar]
  9. Langebeek N, Gisolf EH, Reiss P, Vervoort SC, Hafsteinsdottir TB, Richter C, … Nieuwkerk PT (2014). Predictors and Correlates of Adherence to Combination Antiretroviral Therapy (ART) for Chronic HIV Infection: A Meta-Analysis. BMC Med, 12, 142. doi: 10.1186/PREACCEPT-1453408941291432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Loeys T, Moerkerke B & Vansteelandt S (2015) A Cautionary Note on the Power of the test for the Indirect Effect in Mediation Analysis. Front Psychol, 5, 1549. doi: 10.3389/fpsyg.2014.01549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ma F, Lv F, Xu P, Zhang D, Meng S, Ju L, … Wu Z. (2015). Task Shifting of HIV/AIDS Case Management to Community Health Service Centers in urban China: A Qualitative Policy Analysis. BMC Health Serv Res, 15, 253. doi: 10.1186/s12913-015-0924-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. MacKinnon DP, Lockwood CM, Hoffman JM, West SG, & Sheets V (2002). A Comparison of Methods to Test Mediation and other Intervening Variable Effects. Psychol Methods, 7(1), 83–104. doi.org/10.1037/1082-989X.7.1.83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Muthen L, & Muthen B (2017). Mplus User’s Guide. Eighth Edition. Los Angeles, CA: Muthen & Muthen. [Google Scholar]
  14. Nafradi L, Nakamoto K, & Schulz PJ (2017). Is Patient Empowerment the Key to Promote Adherence? A Systematic Review of the Relationship between Self-efficacy, Health Locus of Control and Medication Adherence. PLoS One, 12(10), e0186458. doi: 10.1371/journal.pone.0186458. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Nieuwkerk PT, & Oort FJ (2005). Self-reported Adherence to Antiretroviral Therapy for HIV-1 Infection and Virologic Treatment Response: A Meta-Analysis. J Acquir Immune Defic Syndr, 38(4), 445–448. doi: 10.1097/01.qai.0000147522.34369.12. [DOI] [PubMed] [Google Scholar]
  16. Ortego C, Huedo-Medina TB, Llorca J, Sevilla L, Santos P, Rodriguez E, … Vejo J. (2011). Adherence to Highly Active Antiretroviral Therapy (HAART): A Meta-Analysis. AIDS Behav, 15(7), 1381–1396. doi: 10.1007/s10461-011-9942-x. [DOI] [PubMed] [Google Scholar]
  17. Press N, Tyndall MW, Wood E, Hogg RS, & Montaner JS (2002). Virologic and Immunologic Response, Clinical Progression, and Highly Active Antiretroviral Therapy Adherence. J Acquir Immune Defic Syndr, 31 Suppl, 3, S112–117. [DOI] [PubMed] [Google Scholar]
  18. Raffa JD, Tossonian HK, Grebely J, Petkau AJ, DeVlaming S, & Conway B (2008). Intermediate Highly Active Antiretroviral Therapy Adherence Thresholds and Empirical Models for the Development of Drug Resistance Mutations. J Acquir Immune Defic Syndr, 47(3), 397–399. doi: 10.1097/QAI.0b013e31815b0d35. [DOI] [PubMed] [Google Scholar]
  19. Rao D, Feldman BJ, Fredericksen RJ, Crane PK, Simoni JM, Kitahata MM, & Crane HM (2012). A Structural Equation Model of HIV-related Stigma, Depressive Symptoms, and Medication Adherence. AIDS Behav, 16(3), 711–716. doi: 10.1007/s10461-011-9915-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Reda AA, & Biadgilign S (2012). Determinants of Adherence to Antiretroviral Therapy among HIV-infected Patients in Africa. AIDS Res Treat, 2012, 574656. doi.org/10.1155/2012/574656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Risser J, Jacobson TA, & Kripalani S (2007). Development and Psychometric Evaluation of the Self-effiacy for Appropairte Medication Use Scale (SEAMS) in Low-literacy patients with Chronic Disease. J Nurs Meas, 15(3): 203–219. doi: 10.1891/106137407783095757. [DOI] [PubMed] [Google Scholar]
  22. Shao Y (2006). AIDS Epidemic at Age 25 and Control Efforts in China. Retrovirology, 3: 87. doi: Artn 8710.1186/1742-4690-3-87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Shrestha R, Altice FL, & Copenhaver MM (2019). HIV-related Stigma, Motivation to Adhere to Antiretroviral Therapy, and Medicaion Adherence among HIV-Positive Methadone-Maintained Patients. J Acquir Immune Defic Syndr, 80(2): 166–173. doi: 10.1097/QAI.0000000000001891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Simoni JM, Frick PA, & Huang B (2006). A Longitudinal Evaluation of A Social Support Model of Medication Adherence among HIV-positive Men and Women on Antiretroviral Therapy. Health Psychol, 25(1), 74–81. doi: 10.1037/0278-6133.25.1.74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Simoni JM, Kurth AE, Pearson CR, Pantalone DW, Merrill JO, & Frick PA (2006). Self-report Measures of Antiretroviral Therapy Adherence: A Review with Recommendations for HIV Research and Clinical Management. AIDS Behav, 10(3), 227–245. doi: 10.1007/s10461-006-9078-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Stirratt MJ, Dunbar-Jacob J, Crane HM, Simoni JM, Czajkowski S, Hilliard ME, … Nilsen WJ (2015). Self-report Measures of Medication Adherence Behavior: Recommendations on Pptimal Use. Translational Behavioral Medicine, 5(4), 470–482. doi: 10.1007/s13142-015-0315-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sun X, Lu F, Wu Z, Poundstone K, Zeng G, Xu P, … Liau A. (2010). Evolution of Information-driven HIV/AIDS Policies in China. Int J Epidemiol, 39 Suppl 2, ii4–13. doi: 10.1093/ije/dyq217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tse WF, & Huang WL (2017). Stigma among HIV/AIDS Patients in China. HIV & AIDS Review, 16(1), 11–17. doi: 10.5114/hivar.2017.65921. [DOI] [Google Scholar]
  29. Turan B, Fazeli P, Raper J, Mugavero M & Johnson M (2016). Social Support and Moment-to-moment Changes in Treatment Self-efficacy in Men Living with HIV: Psychosocial Moderators and Clinical Outcomes. Health Psychol, 35(10), 1126–1134. doi: 10.1037/hea0000356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Turan B, Budhwani H, Fazeli PL, Browning WR, Raper JL, Mugavero MJ, & Turan JM (2017). How Does Stigma Affect People Living with HIV? The Mediating Roles of Internalized and Anticipated HIV Stigma in the Effects of Perceived Community Stigma on Health and Psychosocial Outcomes. AIDS Behav, 21(1), 283–291. doi: 10.1007/s10461-016-1451-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. UNAIDS (2018). Country Progress Report-China. Retrieved from https://www.unaids.org/sites/default/files/country/documents/CHN_2018_countryreport.pdf
  32. Valente WJ, Gonzalez O, Mioevic M, & MacKinnon DP (2016). A Note on Testing Mediated Effects in Structural Equation Models: Reconciling Past and Current Research on the Performance of the Test of Joint Significance. Educ Psychol Meas, 76 (6): 889–991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang J, & Wang X (2012). Structural Equation Modeling: Applications using Mplus. New York, NY: John Wiley. [Google Scholar]
  34. Wu S, Li L, Wu Z, Liang LJ, Cao H, Yan Z, & Li J (2008). A Brief HIV Stigma Reduction Intervention for Service Providers in China. AIDS Patient Care STDS, 22(6), 513–520. doi: 10.1089/apc.2007.0198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Zeng C, Li L, Hong YA, Zhang H, Babbitt AW, Liu C, … Cai W. (2018). A Structural Equation Model of Perceived and Internalized stigma, Depression, and Suicidal Status among People Living with HIV/AIDS. BMC Public Health, 18(1), 138. doi: 10.1186/s12889-018-5053-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Zhang L, Li X, Lin Z, Jacques-Tiura AJ, Xu J, Zhou Y, … Stanton B. (2016). Side Effects, Adherence Self-efficacy, and Adherence to Antiretroviral Treatment: A Mediation Analysis in a Chinese sample. AIDS Care, 28(7), 919–926. doi: 10.1080/09540121.2015.1124984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zhou G, Li X, Qiao S, Shen Z, & Zhou Y (2018). Influence of Side Effects on ART Adherence Among PLWH in China: The Moderator Role of ART-Related Knowledge. AIDS Behav, 22(3), 961–970. doi: 10.1007/s10461-017-1791-9. [DOI] [PubMed] [Google Scholar]

RESOURCES