The Intersection of Inequitable Gender Norm Endorsement and HIV Stigma: Implications for HIV Care Engagement for Men in Ugandan Fishing Communities

Abstract

Men’s engagement in the HIV care continuum may be negatively affected by adherence to inequitable gender norms, which may be exacerbated by HIV stigma. This cross-sectional study with 300 male fisherfolk in Uganda examined the independent and interacting effects of inequitable gender norm endorsement and HIV stigma on men’s missed HIV care appointments and missed antiretroviral (ARV) doses. Greater gender inequitable norm endorsement was associated with increased odds of missed HIV clinic visits (Adjusted Odds Ratio [AOR)]: 1.44, 95% CI: 1.16–1.78) and a statistically significant interaction between internalized HIV stigma and inequitable gender norms on missed ARV doses was identified (AOR: 5.32, 95% CI: 2.60–10.86). Adherence to traditional gender norms reduces men’s HIV appointment attendance, and among men with high internalized stigma, increases the likelihood of poor treatment adherence. These findings point to the need for HIV interventions that reconfigure harmful gender norms with a focus on stigma reduction.

Introduction

Significant gender disparities exist in the HIV care continuum throughout sub-Saharan Africa (1). While women are more likely to be HIV infected (2), men on antiretroviral treatment (ART) have a 41% increased risk of death compared to women on ART (3), which has been attributed to worse engagement across all stages of HIV care among men (1, 4–6). This pattern is reflected in Uganda, where 7.6% of women vs. 4.7% of men are living with HIV (7), but men are more than twice as likely to die of HIV-related illness (8). This gender gap may be even wider in high HIV prevalence occupation-based populations, such as fisherfolk, miners, and truck drivers. These populations are faced with unique social and structural barriers to HIV care, such as frequent mobility, distance to services and transportation challenges, conflict between work and clinic hours, heavy substance use, and HIV stigma (9–11). In addition, masculine norms, or the culturally grounded expectations for men’s roles, behaviors, and relationships, are a recognized driver of men’s HIV care engagement globally (1), and may be especially relevant to men’s health seeking behavior in communities where men cluster for work.

From a social constructionist perspective, health beliefs and behaviors are considered a display of masculinities and femininities that reinforce the broader social structure of gender and power (12, 13). Thus, inequitable gender norms prevalent in Eastern and Southern Africa that reinforce beliefs that men should be strong, dominant over women, sexually successful, and rigid in their gender role fulfillment shape both women’s elevated risk for HIV and men’s engagement in HIV services. Qualitative studies point to the socialization of men to be physically and mentally strong, self-reliant, emotionally inexpressive, and to avoid femininity as contributing to men’s avoidance of HIV testing, delay in ART initiation, and poor treatment adherence (14, 15). Positive masculine norms have also been identified, such as the roles of provider and father, which can serve as motivation for engagement in the care continuum for some men (14, 15). Research with both Ugandan fisherfolk and miners report that men view HIV as damaging to their reputation as a man, referred to by Siu et al. as “dented masculinity” (16–18). However, for men who experience ART’s restorative effect on their strength, health, and masculine role fulfillment, masculine identities can be “resuscitated” and adapted to be more in line with care engagement as men accept their HIV status (16–18).

HIV stigma, or social discrediting and devaluation associated with HIV (19), is another recognized determinant of men’s HIV care engagement in transient work communities (10, 20), as well as in sub-Saharan Africa more broadly (21). The HIV stigma framework posits that people living with HIV experience stigma through the mechanisms of enacted HIV stigma, anticipated HIV stigma, and internalized HIV stigma (22). Enacted HIV stigma is the degree to which people living with HIV believe they have actually experienced prejudice and discrimination from others in their community. Anticipated HIV stigma is the extent to which people living with HIV expect that they will experience prejudice and discrimination from others in the future. Internalized HIV stigma describes negative beliefs and feelings associated with HIV/AIDS that people living with HIV feel about themselves. Prior research has linked these stigma mechanisms to men’s poor engagement in HIV care, including not testing for HIV (23), delayed linkage to care (24), greater drop-out (25), and poor ART adherence (26).

Although the independent influence of both gender norms and HIV stigma on men’s HIV care engagement is well-recognized (1, 14, 15, 21), only a few qualitative studies and no quantitative studies to-date explicitly explore their synergistic effects in relation to HIV care engagement (27–29). The interconnection between these constructs, however, is evident across qualitative research focused on the role of masculine norms in HIV care engagement; a scoping review of 17 studies with this aim in sub-Saharan African settings identified the intersection between masculinity and HIV stigma as the most prevalent theme across the studies (15). This literature describes men’s fear (anticipated HIV stigma) and experience (enacted HIV stigma) of the negative social consequences of being seen at the HIV clinic or taking ART as contributing to men’s suboptimal clinic attendance and treatment adherence. The negative consequences men fear/experience most are intrinsically linked to their masculine roles and ideals. That is, being known as living with HIV is often viewed as threatening to men’s desire to be perceived by others as respected, strong, self-reliant, and sexually successful, and threatens men’s socially valued relationships with their partners, families, male peers and co-workers (30–32). Thus, HIV stigma may have a greater effect on HIV care engagement for men who adhere to traditional masculine norms, as they are likely to experience a larger negative impact of stigma on their reputation as a man, self-worth, and performance of masculine ideals.

This qualitative literature points to the potential interaction between HIV stigma and gender norm endorsement on HIV care engagement. However, no studies to our knowledge have quantitatively tested the interaction between these constructs as they relate to HIV care engagement. Moreover, few quantitative studies have examined the independent effect of gender norm endorsement on HIV care engagement outcomes. The Inequitable Gender Norms (IGN) subscale of the Gender Equitable Men (GEM) scale is a validated quantitative measure of norms endorsing gender inequity and traditional gender roles, which has been associated with men’s HIV risk behavior (33–35), violence perpetration (36–39), and HIV testing (40–42), but has been underutilized in understanding men’s clinic attendance and ART adherence (40).

To fill these gaps in the literature, the aims of this study were to: 1) examine the independent effects of inequitable gender norm endorsement and dimensions of HIV stigma, including anticipated, enacted, and internalized HIV stigma, on men’s missed HIV care appointments and missed antiretroviral (ARV) doses; and 2) examine the interaction between dimensions of HIV stigma and inequitable gender norms on these outcomes. We hypothesize that greater inequitable gender norms and HIV stigma will be associated with poor HIV care engagement, and HIV stigma will interact with inequitable gender norms to worsen men’s HIV care engagement. This study focuses on male fisherfolk on ART in Ugandan fishing communities along Lake Victoria. Characterized as HIV hotspots, HIV prevalence estimates range from 20–42% in these communities (34–36) compared to the national prevalence of 6.2% (7). Both HIV stigma (11) and masculine norms (16) have been identified as determinants of male fisherfolk’s engagement in HIV services in this setting, but have not been examined together. Understanding the independent and interacting effects of inequitable gender norm endorsement and HIV stigma on HIV care engagement in this population could inform the development of gender-tailored HIV services to engage and retain fisherfolk and men in other key populations in HIV care in sub-Saharan Africa.

Methods

Three hundred male fisherfolk living with HIV and on ART in Wakiso District, Uganda participated in this cross-sectional study. The overall purpose of the study was to understand men’s barriers and facilitators to HIV care engagement. Mildmay Uganda supported the study team in all recruitment activities. Mildmay Uganda is a non-governmental organization that supports the Ugandan Ministry of Health to provide free HIV testing and treatment in several districts in Uganda. Through purposive sampling, the study team identified a sample of patients attending Mildmay Uganda-supported outreach clinics and public health facilities serving large populations of fisherfolk. There were seven research sites in total surrounding Lake Victoria, including three landing sites (Kasenyi outreach site, Entebbe Hospital, Kigungu Health Center III) and four island sites on Bussi, Zzinga, and Kachanga Islands (Rapha Health Center, Zzinga Island outreach site, Bussi Health Center, Kachanga Island outreach site).

Recruitment and data collection occurred between October 2016 and March 2017. During HIV clinic days, clinic staff and a trained research assistant reviewed clinic records to pre-screen for eligible participants, and non-randomly approached potentially eligible men for further eligibility screening. Clinic staff also informed men of the study during routine reminder phone calls for upcoming or missed appointments. Patients were eligible for the study if they met the following criteria: male, fisherman or another occupation supporting the fishing industry, living with HIV, on ART for at least 6 months (to assess adherence), and at least 18 years of age. Of the 322 eligible men invited to participate, 22 declined for the following reasons: did not have time (n=14), did not want to disclose sensitive information (n=4), intoxicated (n=1), hearing impairment (n=1), wanted more compensation (n=1), wanted wife present at interview (n=1). For eligible men who agreed to participate, the research assistant obtained written informed consent and conducted the interviewer-administered questionnaire immediately upon enrollment in a private space in or outside of the clinic, or at another agreed-upon time and/or location if more convenient for participants. The research assistant entered responses in real-time using computer-assisted personal interviewing (CAPI) software. The interview lasted approximately 45 minutes. Institutional review boards at San Diego State University in the United States and the Makerere University School of Public Health in Uganda, as well as the Uganda National Council for Science and Technology, approved all study procedures.

Measures

The questionnaire included socio-demographic items and other variables included in our analysis as hypothesized covariates. We used a previously published analysis that assessed multi-level determinants of clinic attendance and ART adherence with the present sample to guide the selection of covariates (11). Covariates include the following: age (continuous), research site (i.e., location of study recruitment, landing site vs. island), marital status (not married/separated vs. married/living together), education (dichotomized for analysis into any vs. none based on distribution), monthly income (continuous), travel time to clinic in minutes (continuous), ART regimen or the number of ARVs prescribed (once daily vs. twice daily), months since ART initiation (continuous), any drinking in the prior 30 days (yes/no).

We measured endorsement of inequitable gender norms through the Inequitable Gender Norms (IGN) scale, a sub-scale of the Gender Equitable Men (GEM) scale (43, 44). The IGN scale has been validated with populations in Uganda, Tanzania, and Ghana (36, 44). The 15-item scale assesses men’s agreement with norms endorsing gender inequity, such as “a man should have the final word about decisions in the home” and “men are always ready to have sex,” with responses ranging from “agree” to “do not agree” (3-point scale). A total continuous score was computed by averaging the items (Cronbach’s alpha = 0.67).

Using Earnshaw’s HIV stigma framework measures (22), we assessed HIV stigma with 6-items on enacted stigma, 6-items on anticipated stigma, and 6-items on internalized stigma. Enacted stigma items measure the perceived degree of prejudice and discrimination experienced by people living with HIV due to their HIV status (e.g., “Family members have avoided you”). Anticipated stigma items measure the expectation among people living with HIV that they will experience prejudice and discrimination in the future (e.g., “Healthcare workers will avoid touching you”). Internalized stigma items assess endorsement of negative beliefs and feelings associated with HIV about themselves (“Having HIV makes you feel like you are a bad person”). Response options were measured on a 5-point scale to assess participants’ level of agreement with statements about enacted and anticipated HIV stigma (0 “strongly disagree” to 4 “strongly agree”) and the frequency of experiencing enacted HIV stigma (0 “never” to 4 “often”) (Cronbach’s alphas: enacted=0.65, anticipated=0.65, internalized=0.79). We computed the average score for each subscale, used as continuous variables in analysis.

To collect the first engagement outcome, missed HIV clinic visits, a research assistant extracted data from participants’ HIV clinic records, which the research assistant confirmed with self-report. If discrepancies arose, the research assistant recorded self-report. Data included the number of clinic visits participants were scheduled for, and the number of those visits participants did not attend within a month of the scheduled visit, in the prior 12 months (or since enrolling in care for those on ART for less than 12 months). The Adult AIDS Clinical Trial Group (AACTG) scale was used to measure the second outcome, self-reported adherence to ARVs (45). The AACTG includes recall questions about ARVs missed for the previous four days prior to the interview and items measuring reasons for non-adherence. It has demonstrated construct validity in Uganda and similar settings (46).

Data analysis approach

In SPSS version 25, we used descriptive statistics and frequencies to describe the sample. For monthly income, months on ART, and minutes traveled to the clinic, outliers (values greater than 3 standard deviations [SD] outside of the mean) (47) were identified and transformed using winsorization (i.e., truncated to the highest value within 3 SDs of the mean). Generalized linear modeling tested bivariate associations between covariates and HIV stigma variables and inequitable gender norm endorsement; we present betas (β) and standard errors (SE). Through generalized linear modeling analyses with a binomial distribution and logit link in the form of events within trials, we examined bivariate and multivariable models with the two engagement outcomes: missed HIV clinic visits and missed ARVs. Controlling for covariates identified in bivariate analyses (p < 0.05), we developed separate multivariable models to assess the association between dimensions of HIV stigma and inequitable gender norms with the two engagement outcomes, missed clinic appointments and missed ARVs, and testing interactions between dimensions of HIV stigma and inequitable gender norms. The presented bivariate models include odds ratios (ORs) and multivariable models include adjusted odds ratios (AORs) and 95% confidence intervals, and include only interactions that remained statistically significant (p < 0.05).

Results

Table 1 includes details of participant characteristics. The average age of the sample was 36.9 years old (standard deviation [SD]=8.6), with half of men married and living with their partner (49%) and two thirds of men reporting no schooling (67.3%). Men traveled an average of 42.8 minutes (SD=46.7) to get to the HIV clinic, and had been on ART, on average, for nearly 2 years (mean=23.99 months, SD=17.02). Overall, scores were normally distributed for anticipated HIV stigma, which had the highest endorsement (anticipated: mean=0.86, SD=0.53), while enacted stigma (mean=0.62, SD=0.48) and internalized stigma (mean=0.61, SD=0.53) scores skewed towards low-moderate endorsement (22). Inequitable gender norms were normally distributed, indicating overall moderate endorsement, comparable to other studies with men in sub-Saharan Africa (36, 44), with the mean score of 1.69 (SD=0.48). In the prior year, men had missed 39% of their scheduled HIV clinic appointments (SD=28.8%). Men reported taking 87.8% (SD=21.5%) of their ARV pills over the 4-day recall period, making 69% of the sample adherent (optimal adherence defined as 95% of pills taken as prescribed), and about a third of the sample sub-optimally adherent (31%).

Table 1.

Participant characteristics, Men in Ugandan fishing communities, 2016–2017 (N=300)

	n (%)/Mean	SD	Range
Dimensions of HIV Stigma & Inequitable Gender Norm Endorsement
Internalized HIV stigma (0–4 scale)	0.61	0.53	0–2.17
Anticipated HIV stigma (0–4 scale)	0.86	0.53	0–2.83
Enacted HIV stigma (0–4 scale)	0.62	0.48	0–3.00
Inequitable gender attitudes (0–4 scale)	1.69	0.48	0–3.00
Socio-demographics
Age	36.90	8.60	20.00–70.00
Research site
Island	175 (58.30%)
Landing site	125 (41.70%)
Marital status
Never married	30 (10.00%)
Divorced	78 (26.00%)
Widowed	16 (5.30%)
Married and separated	29 (9.70%)
Married and living together	147 (49.00%)
Education
No schooling	202 (67.30%)
Primary level	84 (28.00%)
Secondary level	14 (4.70%)
Occupation
Fishermen	246 (82.00%)
Fish seller, cleaner, dryer	35 (11.67%)
Boat operator, repairer, loader	19 (6.33%)
HIV treatment and care engagement
ART regimen
Once daily regimen	260 (86.70%)
Twice daily regimen	40 (13.30%)
Sub-optimal adherence (< 95% of pills taken as prescribed)
Yes	93 (31.00%)
No	207 (69.00%)
Number of missed clinic appointments (prior year)	2.11	1.59	0.00–6.00
Number of scheduled clinic appointments (prior year)	5.52	0.96	1.00–8.00
Proportion of missed clinic appointments (prior year)	39.00%	28.80%	0.00–100.00%
ART adherence (% of prescribed pills taken)	87.70%	21.50%	0.00–100.00%
Monthly income (converted to USD)	62.20	29.70	4.10–160.40
Travel time to clinic (minutes)	42.80	46.70	2.00–360.00
Months since ART initiation	23.99	17.02	2.00–113.00

See Table 2 for detailed statistics on correlates of inequitable gender norm endorsement. Higher scores on the internalized HIV stigma scale was associated with greater endorsement of inequitable gender norms (β = 0.10, SE = 0.05, p = 0.04). Inequitable gender norm endorsement was also greater among men of younger age (β = −0.01, SE = 0.04, p < = 0.04) and those with any schooling compared to no schooling (β = 0.12, SE = 0.06, p = 0.04).

Table 2.

Results of bivariate regression models testing correlates of inequitable gender norm endorsement, Men in Ugandan fishing communities, 2016–2017 (N=300)

	Inequitable Gender Norms
	β	SE	p
Dimensions of HIV stigma
Internalized HIV stigma	0.10	0.05	0.04
Anticipated HIV stigma	0.08	0.05	0.13
Enacted HIV stigma	0.01	0.06	0.81
Covariates
Age	−0.01	0.04	0.04
Research site
Landing site	0.02	0.06	0.73
Island (ref)
Marital status
Not married or separated	−0.08	0.06	0.13
Married, living together (ref)
Education
Any schooling	0.12	0.06	0.04
No schooling (ref)
Monthly income	−0.02	0.01	0.19
Travel time to clinic	0.00	0.00	0.07
ART regimen
Twice daily regiment	−0.15	0.08	0.06
Once daily regimen (ref)
Months since ART initiation	−0.02	0.01	0.06
Any drinking (prior 30 days)
Yes	−0.02	0.06	0.79
No (ref)

Note: β = unstandardized beta; SE = standard error; models were generalized linear regression; Monthly income: 1 unit = 50,000 UGX; Travel time to clinic: 1 unit = 30 minutes; Months since ART initiation: 1 unit = 6 months.

In bivariate models predicting clinic attendance, greater anticipated stigma was associated with more missed HIV clinic visits in the prior year (OR: 1.23, 95% CI: 1.02–1.49, p = 0.03). Men with greater inequitable gender norm endorsement also reported greater missed HIV clinic visits (OR: 1.47, 95% CI: 1.19–1.81, p < 0.001). In addition, we identified the following variables as statistically significant correlates of missed HIV clinic visits in the prior year: younger age (OR: 0.97, 95% CI: 0.96–0.99, p < 0.001), being recruited from a landing site rather than an island (OR: 1.74, 95% CI: 1.42–2.13, p < 0.001), being unmarried/separated compared to being married/living with partner (OR: 1.30, 95% CI: 1.06–1.58, p = 0.01), greater travel time to the HIV clinic (OR: 1.00, 95% CI: 1.00–1.01, p = 0.004), being on the once daily regimen compared to twice daily (OR: 0.72, 95% CI: 0.53–0.96, p = 0.03), less time on ART (OR: 0.99, 95% CI: 0.98–0.99, p = 0.005), and reporting any drinking in the prior 30 days (OR: 1.30, 95% CI: 1.07–1.59, p = 0.01).

In bivariate models predicting non-adherence to ART, men reporting greater anticipated HIV stigma (OR: 2.17, 95% CI: 1.62–2.89, p < 0.001) and greater enacted HIV stigma (OR: 1.77, 95% CI: 1.29–2.42, p < 0.001) reported more missed ARVs doses. Other statistically significant bivariate associations identified with missed ARVs include: younger age (OR: 0.96, 95% CI: 0.94–0.98, p < 0.001), any schooling compared to no schooling (OR: 1.58, 95% CI: 1.12.−2.22, p = 0.01), greater travel time to the HIV clinic (OR: 1.00, 95% CI: 1.00–1.01, p = 0.01), being on the once daily regimen compared to twice daily, (OR: 0.37, 95% CI: 0.22–0.62, p < 0.001), less time on ART (OR: 0.99, 95% CI: 0.98–1.00, p = 0.04), any drinking in the prior 30 days (OR: 1.59, 95% CI: 1.14–2.23, p = 0.007). See Table 3 for detailed statistics.

Table 3.

Results of bivariate regression models testing the association between covariates, dimension of HIV stigma, and inequitable gender norms with missed HIV clinic visits and missed antiretroviral (ARV) doses, Men in Ugandan fishing communities, 2016–2017 (N=300)

	Missed HIV Clinic Visits			Missed ARVs
	OR (95% CI)	x2	p	OR (95% CI)	x2	p
Dimensions of HIV Stigma & Inequitable Gender Norm Endorsement
Internalized HIV stigma	1.14 (0.95–1.37)	1.85	0.17	1.10 (0.81–1.51)	0.37	0.57
Anticipated HIV stigma	1.23 (1.02–1.49)	4.71	0.03	2.17 (1.62–2.89)	27.32	<0.001
Enacted HIV stigma	1.16 (0.95–1.42)	2.00	0.16	1.77 (1.29–2.42)	12.74	<0.001
Inequitable gender norms	1.47 (1.19–1.81)	13.03	<0.001	1.25 (0.89–1.76)	1.68	0.19
Covariates
Age	0.97 (0.96–0.99)	18.36	<0.001	0.96 (0.94–0.98)	12.99	<0.001
Research site
Landing site	1.74 (1.42–2.13)	29.03	<0.001	1.08 (0.77–1.52)	0.21	0.65
Island (ref)
Marital status
Not married or separated	1.30 (1.06–1.58)	6.61	0.01	1.05 (0.75–1.46)	0.07	0.79
Married, living together (ref)
Education
Any schooling	1.14 (0.92–1.40)	1.46	0.23	1.58 (1.12–2.22)	6.75	0.01
No schooling (ref)
Monthly income	0.98 (0.94–1.03)	0.57	0.45	0.92 (0.84–1.00)	3.86	0.05
Travel time to clinic	1.00 (1.00–1.01)	8.52	0.004	1.00 (1.00–1.01)	6.33	0.01
ART regimen
Twice daily regiment	0.72 (0.53–0.96)	4.95	0.03	0.37 (0.22–0.62)	14.28	<0.001
Once daily regimen (ref)
Months since ART initiation	0.99 (0.98–0.99)	7.89	0.005	0.99 (0.98–1.00)	4.09	0.04
Any drinking (prior 30 days)
Yes	1.30 (1.07–1.59)	6.65	0.01	1.59 (1.14–2.23)	7.31	0.007
No (ref)

Note: OR = odds ratio; 95% CI = 95% confidence interval; x² = Wald Chi Square; ref = reference group; missed HIV clinic visits and missed ARVs were tested using generalized logistic modeling analyses with a binomial distribution and logit link; Monthly income: 1 unit = 50,000 UGX; Travel time to clinic: 1 unit = 30 minutes; 1 unit = 6 months; Months since ART initiation: 1 unit = 6 months.

In the multivariate model for missed HIV clinic visits (see Table 4), no interactions were identified between inequitable gender norms and HIV stigma variables. However, a positive main effect was observed between inequitable gender norm endorsement and missed HIV clinic visits (AOR: 1.44, 95% CI: 1.16–1.78, p = 0.001). Anticipated HIV stigma remained only marginally associated (AOR: 1.20, 95% CI: 0.99–1.45, p = 0.07) with missed HIV clinic visits. Other covariates that remained statistically significant in the final model included younger age (AOR: 0.98, 95% CI=0.97–0.99, p = 0.001), being recruited from a landing site compared to an island (AOR: 1.65, 95% CI=1.35–2.03, p < 0.001), and not being married/living with a partner (AOR: 1.35, 95% CI=1.10–1.65, p = 0.004).

Table 4.

Multivariate model testing the association between anticipated HIV stigma and endorsement of inequitable gender norms with missed HIV clinic visits, Men in Ugandan fishing communities, 2016–2017 (N=300)

	AOR (95% CI)	x2	p
Inequitable gender attitudes	1.44 (1.16–1.78)	10.87	0.001
Anticipated HIV stigma	1.20 (0.99–1.45)	3.36	0.07
Age	0.98 (0.97–0.99)	10.52	0.001
Research site
Landing site	1.65 (1.35–2.03)	22.92	<0.001
Island (ref)
Marital status
Not married or separated	1.35 (1.10–1.65)	8.23	0.004
Married and living together (ref)

Note: AOR = adjusted odds ratio; 95% CI = 95% confidence interval; x² = Wald Chi Square; Travel time to clinic: 1 unit = 30 minutes

In the multivariate analyses for missed ARVs, a statistically significant interaction was observed between internalized HIV stigma and inequitable gender norm endorsement on missed ARV doses (AOR: 5.32, 95% CI: 2.60–10.86, p < 0.001). That is, the effect of gender inequitable norms on greater non-adherence increases with greater internalized HIV stigma. A positive main effect was identified between anticipated HIV stigma and missed ARV doses (AOR: 2.02, 95% CI=1.50–2.72, p < 0.001). Covariates that remained statistically associated with missed ARVs in the final model included younger age (AOR: 0.97, 95% CI=0.95–0.99, p = 0.02) and once daily regimen compared to twice daily (AOR: 0.43, 95% CI=0.25–0.75, p = 0.003). See Table 5 for detailed statistics and Figure 1 for a depiction of the observed interaction.

Table 5.

Multivariate model testing the interaction between endorsement of gender inequitable norms and internalized HIV stigma on missed antiretroviral (ARV) doses, Men in Ugandan fishing communities, 2016–2017 (N=300)

	AOR (95% CI)	x2	p
Inequitable gender norms × internalized HIV stigma	5.32 (2.60–10.86)	21.02	<0.001
Anticipated HIV stigma	2.02 (1.50–2.72)	21.52	<0.001
Internalized HIV stigma	0.05 (0.01–0.19)	19.70	<0.001
Inequitable gender norms	0.40 (0.23–0.68)	11.22	0.001
Age	0.97 (0.95–0.99)	5.76	0.02
Number of pills prescribed
2 pills/day	0.43 (0.25–0.75)	9.13	0.003
1 pill/day (ref)

Note: AOR = adjusted odds ratio; 95% CI = 95% confidence interval; x² = Wald Chi Square

Figure 1.

Mean predicted value of number of missed ARVs by gender norm endorsement and internalized HIV stigma. Note: The graph compares men with gender inequitable attitudes scores 1.5 standard deviation below the mean vs. 1.5 standard deviation above the mean.

Discussion

This study sought to understand the independent and interacting associations between inequitable gender norm endorsement and HIV stigma on HIV clinic attendance and ART adherence among men on ART. These research questions were examined among male fisherfolk in Ugandan fishing communities, a most-at-risk and priority population for HIV test-and-treat initiatives in Uganda. The study’s findings extends the current literature by quantitatively assessing the effect of inequitable gender norm endorsement on HIV clinic attendance and ART adherence, and by examining the interaction between dimensions of HIV stigma with inequitable gender norms – adding support and further insight into previous qualitative inquiries. In this study, fisherfolk endorsing more inequitable gender norms reported greater internalized HIV stigma, were less likely to attend HIV clinic appointments, and inequitable gender norm endorsement negatively affected ART adherence for men with high internalized HIV stigma.

The study’s findings add support to a large body of research reporting detrimental effects of inequitable gender norms on men’s health behavior and outcomes. Specific to HIV risk in sub-Saharan Africa, research links inequitable gender norms to unprotected sex, multiple partners, substance use in the context of sex, and intimate partner violence (33–39). More recent inquiries in the region quantitatively link inequitable gender norms to less HIV testing (40–42), but the effect of inequitable gender norms on engagement in the HIV care continuum after men are linked to care is less studied (40). In this study, men endorsing more inequitable gender norms had a greater likelihood of missing their HIV clinic appointments. Men with greater anticipated HIV stigma reported greater missed HIV clinic visits. While the fear of being seen at the HIV clinic is a known deterrent to male clinic attendance across settings (21), this association may be particularly relevant in occupation-based communities, where the risk of being seen by one’s coworkers is high.

Prior qualitative studies in sub-Saharan Africa provide insight as to why men with more traditional gender attitudes may be more likely to avoid HIV care, including the tendency for men to feminize health and HIV services and not want to be seen in “female spaces” (48–51). Men with inequitable gender attitudes may also be more resistant to assuming the patient role, associating it with weakness and vulnerability (52–54). In addition, an unwillingness to change male-sanctioned risk behaviors associated with inequitable gender norms, such as sex with multiple partners, unprotected sex, and alcohol use, can contribute to men’s avoidance of HIV services (49, 52, 55). A study with men in Ugandan mining towns, which have similar characteristics of fishing communities, found one barrier to HIV testing and treatment uptake was the fear that being seen at the HIV clinic would undermine opportunities for sex in the context of competition for partners (56). Men in the present study who reported greater anticipated HIV stigma were more likely to report missed HIV clinic appointments. However, we did not identify interactions between inequitable gender norms and any of the HIV stigma dimensions on clinic attendance. It is possible that these null findings are a result of having recruited the sample directly from the HIV clinic. Future studies should test the interacting effects of HIV stigma and gender norms on clinic attendance among a community-based sample of men, for whom stigma and gender norms may be an especially pertinent barrier.

ART adherence was compromised in this study for men with high endorsement of inequitable gender norms and high internalized HIV stigma, as demonstrated by the observed interaction between these constructs. Thus, inequitable gender norms and internalized HIV stigma have a synergistic effect on ART adherence. We did not identify any other interactions between inequitable gender norms and enacted or anticipated HIV stigma. Prior research has demonstrated internalized stigma can reduce ART adherence (57–60). It is theorized that internalized HIV stigma may have a more direct effect on ART adherence than other interpersonal forms of HIV stigma because of its greater implications for psychological and mental health outcomes, making it more proximal to behavioral and clinical outcomes (22, 58, 61). Future research should explore indirect pathways in which internalized HIV stigma and inequitable gender norms may lead to non-adherence, which could include depression or other compromised psychological processes, such as adaptive coping and social support (59, 62–64). Nevertheless, the present study extends the current literature by demonstrating that for men with high inequitable gender norms endorsement and high internalized HIV stigma, ART adherence was negatively affected. These men may view HIV as particularly damaging to masculine roles/ideals, which when compounded by internalized HIV stigma, is to the determinant of men’s treatment adherence.

Implications

This study’s findings have implications for the use of gender-transformative interventions to address men’s gendered barriers to HIV care engagement. Gender-transformative interventions broadly challenge harmful masculine norms and promote gender equality as an approach to improve men and women’s healthy behaviors and outcomes. Systematic reviews report that gender-transformative programs that engage individuals or groups of men and women to adopt gender equitable attitudes have been successful at reducing IPV incidence, sexual risk, and gender inequitable attitudes in African settings (65, 66). Emerging evidence supports their application for engaging men in HIV testing and care (67–70), but more research is needed that tests this intervention approach’s effect more specifically on ART adherence, retention in HIV care, and viral load suppression among people living with HIV.

The findings of the present study suggests that integrating evidence-based stigma reduction strategies into gender-transformative interventions would strengthen this approach, and vice versa. This approach might be especially relevant for engaging young men in HIV care, as younger men in the present sample were more likely to miss HIV clinic visits and ARV doses, and were more likely to endorse inequitable gender norms. Research points to multi-level strategies that combine individual-level with community or structural-level components as potentially more effective than single-level interventions for both gender transformative interventions (65, 71) and stigma reduction interventions (72). Both gender inequitable attitudes and HIV stigma are thought to stem from broader social norms embedded in community and institutional structures (73–75), further justifying a focus on multi-level interventions. At the interpersonal-level, increasing social support through couples, family, and peer-based approaches has reduced HIV stigma in a range of settings (76–78). One peer-based study in Uganda recruited model men called “Emanzis” to lead peer groups through a 10-session curriculum designed to transform gender norms and motivate men to engage in family planning and HIV services (68). Although the study saw only small changes in gender attitudes, men in the intervention reported more health facility visits, HIV testing, and condom use (68).

At the community-level, community-wide communication campaigns or broader network-approaches to engage men in HIV care could apply a gender transformative lens to HIV stigma reduction, building on previous research that separately demonstrates success in community-level normative change related to gender equity (79) and HIV stigma (77). Figueroa et al. (80) used community dialogues in Mozambique to change gender and sexual norms for HIV prevention, a potentially replicable and scalable approach to altering community norms, while increasing social support protective against HIV stigma; the study reported reduced inequitable gender attitudes and HIV stigma among men in the intervention (80).

Other structural-level gender-transformative interventions target economic HIV risk factors, such as microfinance or cash transfer programs to reduce women’s vulnerability to HIV by improving their economic independence (81). Economic vulnerability is a known driver of sexual risk and alcohol use in Ugandan fishing communities (82, 83). Therefore, similar livelihood or savings-led economic strengthening interventions could address structural barriers to HIV care engagement in this setting (83), while reducing HIV stigma. Such approaches have been shown to reduce internalized HIV stigma associated with HIV diagnosis when paired with ART programs (21, 72, 84, 85).

Limitations

This study adds quantitative evidence to a largely qualitative literature linking gender norm endorsement, HIV stigma, and HIV engagement outcomes. However, the cross-sectional design of this study limits inference of a causal relationship between the constructs examined, which likely have bidirectional relationships. We found men with greater endorsement of inequitable gender norms also reported greater internalized HIV stigma. This finding provides some preliminary support that gender norms shape men’s experience of stigma – pointing to the potential mediating role of HIV stigma, in addition to moderation. Future longitudinal research can extend this study’s findings by further delineating the pathways between the constructs examined, including testing HIV stigma as a mediator between inequitable gender norms and men’s health seeking behavior. Longitudinal study designs can help to establish temporality and further delineate the pathways in which these constructs are interconnected and influence HIV care engagement.

The use of self-reported HIV care engagement outcomes are subject to recall and social desirability bias, and would be improved through biological measures of ART adherence or viral load. We used clinic record data to collect participants’ clinic attendance, which research assistants confirmed with self-report. However, we did not systematically measure the degree of agreement between self-report and clinic records. The measurement of HIV stigma and gender norm endorsement is also limited to the individual-level; future studies would strengthen the literature using community-level measures of these constructs.

The generalizability of the findings reported are limited to male fisherfolk on HIV treatment in Uganda. We included men on ART for at least 6 months to allow us to measure retention in care and ART adherence. Therefore, we did not capture men who are known HIV positive, but not on HIV treatment. It is possible that selection bias resulting from this eligibility criterion could have influenced the interaction results. However, the scores on internalized HIV stigma in this sample are comparable to that reported in cohort studies of individuals on ART in Uganda (87, 88). Therefore, the findings may be generalizable to others on treatment.

Further, this sample of men was a convenience sample recruited directly from the HIV clinic using non-random, purposive sampling. Men engaged in HIV care and on HIV treatment, as well as men agreeing to participate in this study, may be less likely to endorse gender inequitable norms and HIV stigma than those not on ART or not agreeing to participate in this study. Future research should explore the relationship of gender norms and HIV stigma in relation to care engagement among men lost to follow-up, and community samples of men in relation to HIV testing. Research on this topic with men not found at the clinic could yield a larger the effect size if they endorse more inequitable gender norms and HIV stigma than those in the present sample. It could also uncover interactions with the other dimensions of stigma measured (i.e., anticipated and enacted), which may be more important for men avoiding care.

Conclusions

The findings from this study add support for the role of inequitable gender norm endorsement in men’s HIV clinic attendance, as well for the synergistic effect of inequitable gender norms and internalized HIV stigma on ART adherence among Ugandan fisherfolk living with HIV. The findings suggest gender-transformative interventions that alter inequitable gender norms could help improve men’s engagement in HIV care, and would be strengthened through the integration of strategies that simultaneously reduce HIV stigma. Future research should continue to explore the intersection between inequitable gender norms and HIV stigma, while testing multi-level strategies to reduce gender gaps in the HIV care continuum for men in high-risk, transient work communities and more broadly.