Psychometric Validation of a Scale to Assess Culturally-Salient Aspects of HIV Stigma among Women Living with HIV in Botswana: Engaging “What Matters Most” to Resist Stigma

Abstract

Perceived stigma deters engagement in HIV care and is powerfully shaped by culture. Yet few stigma measures consider how cultural capabilities that signify “full personhood” could be engaged to resist stigma. By applying a theory conceptualizing how culturally-salient mechanisms can worsen or mitigate HIV stigma in relation to “what matters most” (WMM), we developed the WMM Cultural Stigma Scale for Women Living with HIV in Botswana (WMM-WLHIV-BW) and psychometrically evaluated it among 201 respondents with known and unknown HIV status. The two subscales, Cultural Factors Shape Stigma (CFSS) and Cultural Capabilities Protect against Stigma (CCPS) were reliable (both α = 0.90). Among WLHIV, the CFSS Subscale showed initial construct validity with depressive symptoms (r=.39, p=.005), similar to an established HIV stigma scale, whereas the CCPS Subscale showed initial construct validity with self-esteem (r=.32, p=.026) and social support number (r=.29, p=.047), suggesting that achieving local cultural capabilities mitigates stigma and is linked with positive psychosocial outcomes. This culturally-derived scale could help WLHIV in Botswana experience improved stigma-related outcomes.

Introduction

Stigma has been identified as a global health research priority, with growing urgency to better understand its etiology, mechanisms, and consequences, in order to develop effective measures and interventions for this “cultural disease” that impacts people worldwide [1,2]. Despite substantial scientific and medical advances in preventing and treating HIV, the lives of individuals and effectiveness of health programs continue to be negatively impacted by stigma and discrimination. HIV stigma has been associated with increased sexual risk behaviors, poorer mental health and social support, and suboptimal use of health and social services including antiretroviral treatment (ART) adherence [3–5].

Stigma encapsulates concepts ranging from the individual to the societal level. In his seminal text, Goffman (1963) described stigma as “an attribute that is significantly discrediting,” resulting in a “spoiled identity” for stigmatized individuals [6]. More recently, stigma has been conceptualized as an overarching concept interlinking the labeling of human differences with undesirable characteristics (stereotypes), resulting in cognitive separation, emotional reactions (e.g., shame), status loss, and discrimination that unfold in an unequal power context, and ultimately, lead to inequitable health outcomes [7,8]. Structural [2] and cultural [9] frameworks of stigma have also been presented, with the latter being our focus. Conceptualization and measurement of stigma at the individual level (i.e., when a stigmatized individual internalizes negative stereotypes that in turn have harmful effects), known as internalized stigma or self-stigma [10], has drawn particular interest in predicting, and (if addressed) potentially mitigating, harmful individual-level outcomes. Internalized stigma develops from perceived stigma [11], as individuals accept and then apply negative community characterizations of PLHIV to themselves. Internalization of negative community beliefs may lead individuals to withdraw from social support, engage in unhealthy behaviors (e.g., risky sex), and forego health services to avert devaluation and discrimination [2,12,13]. Both internalized HIV stigma [14,15] and perceived stigma [15] have emerged as targets for intervention [14,15] but culturally-salient aspects in relation to either stigma concept have received less direct attention. In assessing cultural impacts upon stigma, an initial focus on perceived stigma is appropriate, as it is the most natural stigma mechanism linking local beliefs and cultural norms [16] and internalized stigma [17].

Although the importance of culture is recognized in prominent conceptualizations of stigma [7,9,18], there is an acute need for frameworks that can conceptualize and measure culturally-salient aspects of stigma to reduce stigma and facilitate positive psychosocial outcomes. Conceptualizations of measurement in stigma for HIV [19] and across other health conditions [16] have provided essential advances. Some dimensions of stigma appear universal [20]; yet multiple qualitative findings enumerate culture’s important and distinctive role regarding how stigma is locally manifested, experienced, and resisted [21–25]. Although prior studies have adapted and validated existing scales in new cultural settings [26,27], and in some cases developed scales culturally-informed by qualitative data [28–30], to our knowledge there has not yet been a systematic effort to assess how culturally-salient mechanisms can interact with stigma to worsen or mitigate its effects within a cultural context. Efforts to identify local cultural capabilities that could ameliorate stigma and encourage positive psychosocial outcomes have only recently begun in other areas (i.e., mental illness stigma [31,32]). This study, for the first time, seeks to apply an influential theory of culture and stigma [9] in a setting where cultural factors are known to exert a powerful and sustained influence on HIV stigma [33–36].

Botswana has persistently had one of the highest HIV rates worldwide, with studies estimating a 20% overall HIV prevalence rate [37]. The persistence of HIV rates in Botswana has been attributed to multiple factors, including migration, the subordinate position of women, and stigma [34,36]. Botswana’s national response to the epidemic has been lauded, since government-funded ART has been provided since 2002. Despite evidence that HIV stigma decreased at least somewhat after government-funded treatment programs were introduced [38], HIV stigma remains pervasive and continues to reduce effectiveness of testing and treatment programs [39–41]. Even when treatment is accessed, the burden of stigma hinders quality of life and impairs social integration, requiring adjunctive psychosocial interventions to complement ART programs [39]. HIV stigma can most severely impact vulnerable populations in Botswana, such as women [36]. Women are forced to negotiate the risks of HIV against cultural imperatives to bear children, in a context that affords them limited power in defining the terms of sexual relationships and often assigns them blame for spreading HIV [33].

To identify and operationalize these cultural aspects of stigma in Botswana, we turn to an influential theory that indicates that stigma is felt most acutely when people are not able to achieve “full personhood” by participating in the activities that “matter most” (e.g., accumulation of status, helping others, relationships, health, holding a job, or preservation of the family lineage) in their culture [42]. This theory further proposes that stigma exerts its effects most powerfully by threatening participation in these daily activities that are “most at stake” and central to being seen as a full status person by cultural group members [9]. Stated plainly, stigma is felt most acutely when people are not able to participate in the activities that “matter most” and determine “personhood” in their culture. The “what matters most” (WMM) theory therefore provides a framework to consider how cultural conceptions of “what matters most” within an individual’s cultural group directly interact with stigma. For example, within Chinese-American groups that value preservation of the family lineage [43], the culturally-salient threat of genetic contamination aided prediction of public stigma toward people with mental illness by this group, when compared with European-American respondents [31]. In prior qualitative research utilizing this theory in Botswana [44], we found “what matters most” to be highly gendered; i.e., that “personhood,” and the core engagements that comprised it, varied distinctly by gender. For women in Botswana, the core value for “full womanhood” is achieved by being a mother. Additional markers of personhood for women included getting married and respecting one’s husband, bearing and raising children, and providing a strong, moral foundation for the family. Being identified with HIV can therefore threaten one’s status as a “good mother” due to perceived sexual promiscuity, prompting real fears of abandonment by a male partner and further endangering personhood.

In addition to identifying distinct ways in which culture shapes stigma, the “what matters most” theory enables identification of the everyday cultural capabilities that signify full “personhood” in a local setting, thus protecting against HIV stigma. This framework enables discovery of the capabilities that are core to personhood and can help women strive to achieve them (e.g., preserve status as a “good mother” [45]), thus potentially improving self-esteem, community integration and other positive psychosocial outcomes, including diminished experience of HIV stigma. That is, if a woman with HIV in Botswana fulfills the capabilities of being a “good mother” by bearing and raising children in culturally-approved ways, this could facilitate acceptance by the community, and act to mitigate HIV stigma. Incorporating all societal perspectives is crucial, as men’s perceptions of women as “good” or “proper” and by extension, eligible romantic partners, plays a critical role in women’s achievement of “full womanhood” (i.e. men’s willingness to financially support women to care for children and therefore be a “good mother” [44]). By identifying constructs central to “womanhood”, this framework can thereby indicate how women with HIV could achieve social reintegration (not identified by prior HIV stigma scales [3,46]) and potentially be linked to positive outcomes such as empowerment and psychological well-being [47]. In a similar fashion, among a sample of primarily undocumented Chinese immigrants with psychosis, engaging in employment as a strategy to perpetuate the lineage (thereby achieving “what matters most”) appeared to ameliorate stigma and facilitate community acceptance within this cultural group even while living with mental illness [38]. Building from our prior qualitative work of “what matters most” for women in Botswana, we therefore developed a quantitative stigma measure to assess how culturally-based factors can both shape and protect against HIV stigma among women.

We applied a theory conceptualizing how culturally-salient mechanisms affected HIV stigma in relation to “what matters most” (WMM) for women to develop two subscales of the WMM Cultural Stigma Scale for Women Living with HIV in Botswana (WMM-WLHIV-BW): a) WMM Cultural Factors Shape Stigma (CFSS) and; b) WMM Cultural Capabilities Protect against Stigma (CCPS). We examined initial reliability (i.e., internal consistency and test-retest reliability) and construct validity of these two subscales, relative to an established HIV stigma scale and stigma-related psychosocial outcomes.

Methods

Development and validation of the two WMM-WLHIV-BW subscales occurred in three phases: (1) Formative qualitative research to identify key constructs comprising culturally-salient aspects that shape and protect against stigma; (2) Operationalization of new culturally-salient items for HIV stigma for women in urban Botswana that reflect these concepts; (3) Initial psychometric validation of the new culturally-salient HIV stigma scale focusing on these two distinct subscales.

(1). Formative qualitative research to identify key constructs comprising culturally-salient aspects of stigma

The ‘what matters most’ (WMM) approach to developing a culturally-salient stigma measure for women with HIV in Botswana followed established psychometric procedures whereby constructs for scale items were elicited in a comprehensive manner (i.e., content validity [48]) via exploratory qualitative work. Our study team conducted five focus groups (FGs; n=38 individuals) and additional in-depth interviews (IDIs; n=46) with: (a) Setswana-and English speaking individuals whose HIV-positive status was known at the Princess Marina Hospital Infectious Diseases Care Centre (PMH-IDCC), and; (b) Setswana- and English speaking members of the general community whose HIV status was unknown and who resided in Gaborone. To capture both men’s and women’s perspectives, genders were sampled equally across both groups. Individuals with both known and unknown HIV status were interviewed to obtain perspectives of stigmatized individuals as well as community members regarding the capabilities that ‘matter most’ to maintaining status as a ‘good woman’ in Setswana culture. A detailed description of data collection [21] and qualitative findings are described elsewhere [44]; a brief summary followed by the main qualitative themes that were operationalized into stigma items are described here. Analysis of the FG and IDI transcripts involved deductive qualitative analysis using the WMM framework with a two-step directed content analysis approach to first identify main categories and then the conceptual relationships between them [49]. Transcripts were double-coded over 18 months with weekly discussions enabling reconciliation of codes and development of a main codebook. This process resulted in a thematic inventory of cultural capabilities that defined WMM for urban women in Botswana, including codes identifying how WMM shaped women’s experience of HIV stigma and how cultural capabilities for women could protect against HIV stigma.

(2). Operationalization of new culturally-salient items for HIV stigma

Three team members who had conducted the original qualitative analyses (LHY, OBP, and cultural expert SR) constructed scale items based on key themes that emerged from the qualitative data regarding being viewed as a ‘good woman’ in relation to HIV stigma. The operationalized WMM concepts that shaped and protected against HIV stigma encompassed core themes of: a) religion, b) taking care of/providing for the family, c) achieving motherhood and successfully raising children, d) being a supportive wife, and e) being faithful to one’s partner (Table 1, Column 1). These concepts overlapped across the two subscales, while other WMM concepts emerged that only shaped or only protected individuals from HIV stigma (Table 1, Column 1). Based upon these two aspects of the WMM theory, scale items were constructed for two distinct subscales: i) How cultural factors shape HIV stigma among women (Cultural Factors Shape Stigma [CFSS] Subscale) and ii) How cultural capabilities protect against HIV stigma among women (Cultural Capabilities Protect against Stigma [CCPS] Subscale). The study team developed items based on Link and colleagues’ wording for perceived stigma [12], capturing what “most Batswana” would believe about a woman with HIV in Botswana, which reflected each operationalized WMM theme as illuminated by the qualitative data (Table 1, Columns 2 and 3). An initial pool of 75 items (2–5 alternatively phrased items per operationalized theme) represented the WMM items; 38 reflected the CFSS Subscale and 37 reflected the CCPS Subscale.

Table 1:

Key qualitative themes, scale items, and sample quotes for WMM stigma scale

Key Qualitative Themes	Scale Item	Sample Quote
a. How Cultural Factors Shape HIV Stigma
Religion *	1. Most Batswana believe that a woman who has HIV is not a devout Christian.	“I’m a Christian. I read my Bible and I know who I am before God, before people. But those who, who are not in any religion, those who don’t understand themselves, they might go wayward.” “She [had HIV because she] wasn’t a Christian and did not use protection.”
Money	2. Most Batswana believe that a woman who has HIV got HIV because she was promiscuous to get money.	“Most people would think someone is a prostitute and they got it through sex… [they] choose to be a prostitutes, [because] they love money but they do not want to work.”
Being a supportive wife *	3. Most Batswana believe that a woman who has HIV will not be seen as a respectful wife.	“Once she has HIV then she is seen as, even though she is married, she wasn’t looking after herself [was not acting as a proper wife].”
Achieving motherhood and successfully raising children *	4. Most Batswana believe that a woman who has HIV is unable to take care of her children.	“A good or proper or real Motswana woman… takes care of her own children, especially bathing them, making sure they have food, and they go to school.”
Taking care of the home	5. Most Batswana believe that a woman who has HIV does not have the ability to properly take care of her home.	“We take it that a woman belongs to the home, so we wonder if she has HIV, there is no woman in her. She is a whore, there is no woman.”
Taking care of/providing for the family *	6. Most Batswana believe that a woman who has HIV is not a reliable foundation for her family.	“Women are the most…they are the ones who are blamed the most [for HIV] because they believe that a woman should be the one who is responsible for raising the kid and making sure that the kid is growing up properly.”
Having a good upbringing	7. Most Batswana believe that a woman who has HIV has not been raised by her family to be disciplined and respectful.	“We are from different family backgrounds, there are those families where people would respect and want to associate with like pastors, dikgosi, [and] there are those [who have a person with HIV] …[that are] very dysfunctional… a woman from a respected family will grow up disciplined and respectful.”
Having self-respect	8. Most Batswana believe that a woman who has HIV does not respect herself.	“l think a proper woman is the one who has self-respect.”
Being faithful to one’s partner *	9. Most Batswana believe that a woman who has HIV was promiscuous before marriage.	“Our mothers, they will tell you that, ‘Yah that’s why I told you, that you shouldn’t have kids before marriage, now you see, you have been running around with many boyfriends, now you are [HIV] positive. That’s the result of it.’”
Community standing	10. Most Batswana believe the community will gossip a lot about a woman who has HIV.	“There is a lot of gossip around. That’s a big thing… it’s usually about spreading… people’s [HIV] status once they know somebody or suspect somebody is HIV positive.”
b. How Culture Protects Against Stigma
Religion *	11. Most Batswana believe that a woman who has HIV who is a God-fearing woman will not feel ashamed about her HIV status.	“A [proper] woman, is hard working, fears God and gives family first priority.”
Achieving a supportive romantic partnership	12. Most Batswana believe a woman who has HIV who has a man who can provide for her will be respected by others.	“[For a real woman] it is important to be married… you are incomplete if you are … not married.”
Taking care of/providing for the family *	13. Most Batswana believe that an employed woman who has HIV who provides for the family should be respected.	“In the society people they look at how you are able to maintain and support your family and yourself, … so if you are HIV positive and you are working, they will respect you.”
Being faithful to one’s partner *	14. Most Batswana will not look down on a woman who has HIV who is loyal to her husband.	“A good or proper or real Motswana woman… she has to be very faithful [to her husband].”
Achieving motherhood and successfully raising children *	15. Most Batswana will not look down on a woman who has HIV who has children.	“I think the lady who is HIV positive, who has got kids is much better (off)” [than a childless HIV negative woman].
Achieving motherhood and successfully raising children *	16. Most Batswana respect a woman who has HIV who raises successful children.	“You would look at the kind of parents that raised her. A woman from a respected family will be raised to be disciplined and respectful. She will be exemplary to other children in the society.”
Being a supportive wife *	17. Most Batswana will respect a woman who has HIV who properly cares for her husband.	“In Botswana, most of the responsibility is just placed on the woman’s shoulders, the woman is the foundation, as a woman…you are the one who takes care of the husband.”
Caring for oneself	18. Most Batswana respect a woman who has HIV who takes care of herself and looks attractive.	“If I have HIV, you can discriminate against me because of the way I carry myself. You can not carry yourself properly and people will see that, but I take care of myself [so others will respect me].”
Treating others with respect (Botho)	19. Most Batswana would give Botho to a woman who has HIV who treats others with Botho.	“A person is a person through other people” (“motho ke motho ka batho”) (I.e., Botho signifies treating others with dignity)
Looking healthy	20. Most Batswana will respect a woman who has HIV who looks healthy and attractive.	“When they see you, you are still alive, you are still bubbly, you are still beautiful [even with HIV]... [the people] just like you.”

^*Indicates overlapping themes between subscales

Next, three bilingual, Batswana HIV clinicians who had multiple years of experience treating patients living with HIV reviewed the 75 items and phrasings to judge which were most culturally relevant (using a 1 to 3 rating scale; 1= not relevant; 3= most relevant). Using this feedback, the study team selected a subset of 40 items (21 CFSS items and 19 CCPS items) for further testing. The English version of this pilot 40-item scale was translated to and back-translated from Setswana, and then administered with other validation scales during the psychometric validation phase (below).

(3). Initial psychometric validation of new WMM culturally-salient stigma scale for HIV among women

To conduct initial psychometric analyses, the WMM Cultural Stigma Scale for Women with HIV in Botswana (WMM-WLHIV-BW) was administered to 201 participants in Gaborone in January 2019: (a) 102 Setswana- and English speaking individuals with known HIV-positive status at the PMH-IDCC, and; (b) 99 Setswana- and English speaking general community members with unknown HIV status in Gaborone. Men and women were sampled equally across both groups. Following consent procedures, four trained bilingual research assistants administered the structured questionnaire. A subset of respondents (n=28) were re-administered the WMM-WLHIV-BW one to two weeks after first administration. Ethical approval was obtained from the Botswana Ministry of Health and Wellness (PMH 5/79 282–1-2017), and the Institutional Review Boards of the University of Botswana (UBR/RES/IRB/GRAD/1617), University of Pennsylvania (823407), and New York University (IRB-FY2017–609).

Measures

In addition to the WMM-WLHIV-BW, the survey included socio-demographic characteristics, an established stigma measure (Berger HIV Stigma Scale [50]), and psychosocial constructs (i.e., depressive symptoms, self-esteem, social support) that have been previously associated with stigma among people living with HIV.

WMM-WLHIV-BW response options for scale items were: 1=strongly disagree, 2=disagree, 3=agree, 4=strongly agree; thus, higher Cultural Factors Shape Stigma (CFSS) Subscale scores indicate higher endorsement of culturally-salient forms of stigma, and higher Cultural Capabilities Protect against Stigma (CCPS) Subscale scores indicate greater potential resistance to stigma (i.e., participants agreed that a woman who achieved the cultural capabilities associated with being a “good mother” in Botswana would experience reduced HIV stigma). Because these are two distinct constructs, the subscales are not combined into an overall scale score. For the final 10-item versions of each subscale (below), the total score for each ranged from 10 to 40.

Socio-demographic Characteristics

Socio-demographic variables included age, sex, relationship status, education level, and employment status. For individuals with unknown HIV status, we assessed their self-reported HIV testing history and intention (ever tested, tested in last 12 months, willingness to test). These measures were used to characterize the samples.

Stigma Measure

The established HIV stigma measure used to assess construct validity was the 40-item Berger HIV Stigma Scale [50] which captures multiple key aspects of stigma (i.e., personalized stigma, disclosure concerns, negative self-image, and concern with public attitudes towards people living with HIV) in a single sum score. Because items assess stigma among individuals with HIV, this scale was only administered to the sample with known HIV-positive status.

Associated Psychosocial Constructs

Measures hypothesized to correlate with stigma among women with known HIV-positive status, but to be attenuated among individuals who do not share an HIV-positive status (i.e., the sample with unknown HIV status; see “Analyses”), included depressive symptoms, self-esteem, and social support. Depressive symptoms were assessed using the 20-item Center for Epidemiological Studies Depression Scale (CES-D) (0=rarely or none of the time to 3=most or all of the time) [51] as a continuous sum score (range 0–60). Self-esteem was assessed using the 10-item Rosenberg Self-Esteem Scale (0=strongly disagree to 3=strongly agree) as a continuous sum score (range 0–30) [52]. Social support was assessed using the brief 6-item version of the Sarason Social Support Scale, which asks participants to list the number of social support contacts per item (count) and their level of social support satisfaction from those contacts (1=very dissatisfied to 6=very satisfied) and then calculates the mean number of contacts (range 0–9) and level of satisfaction (range 0–6) [53].

Statistical Data Analysis

First, we computed descriptive statistics for socio-demographic characteristics, stigma measures, and associated psychosocial constructs. To test for differences between individuals with known and unknown HIV status, we used the Wilcoxon rank-sum test for continuous variables and Pearson’s chi-squared test for categorical variables. Second, we conducted an exploratory factor analysis (EFA) to assess the number of potential subscales and whether items would align with the two theoretical subscales (Cultural Factors Shape Stigma [CFSS] and Cultural Capabilities Protect against Stigma [CCPS]). We included all 40 items using a maximum likelihood estimator and oblimin rotation with the entire sample. We included all respondents in the EFA because WMM is conceptually shared by all members of a local cultural group, whether they are known to have HIV or not. For the subscales, we used the standard cutoff of an eigenvalue greater than one, and to confirm that individual items only loaded onto one factor, factor loadings greater than 0.30. We further used the EFA results to reduce scale length while maintaining content validity. Within each identified qualitative WMM theme, we selected the item with the highest factor loading to reduce the scale from 40 to 20 items (with one exception; we selected two items for the theme of “Achieving motherhood and successfully raising children” for the CCPS Subscale due to its salience in the qualitative data and to maintain each subscale at 10 items). We then re-ran the EFA to assess the final 20-item version. Lastly, for the final 20-item version, we assessed initial reliability (internal consistency, test-retest reliability) and construct validity (with an established HIV stigma scale and associated psychosocial constructs; see below) for the two WMM-WLHIV-BW subscales. For internal consistency, we examined Cronbach’s alpha with a goal of α ≥ 0.80. For test-retest reliability, we compared responses from a subset of participants (n=28) who completed the scale twice, one to two weeks apart, using the Wilcoxon signed-rank test.

To assess construct validity, we hypothesized convergent validity of the two WMM stigma subscales with: 1) the Berger HIV Stigma Scale; and 2) associated psychosocial constructs (depressive symptoms, self-esteem, social support), using Spearman’s rank correlation coefficient for ordinal and continuous data [54]. Since the WMM subscales focus on how WMM interacts with stigma for women with HIV, tests for convergent validity were conducted separately for women with known HIV-positive status (n=51) and individuals who did not share the HIV-positive status (i.e., men and women with unknown HIV status; n=99) as it was hypothesized the correlations would be attenuated in the latter group for whom stigma was not directly salient. Based on meta-analyses of correlates with HIV stigma [3,5]), among the sample of women with known HIV-positive status in which perceived stigma is viewed as being salient, we proposed that: (1) the CFSS Subscale would show positive and significant associations with a validated stigma measure (Berger HIV Stigma Scale) [50]; (2) greater scores on this subscale would be positively associated with negative psychosocial outcomes including worse self-esteem [52], greater depressive symptoms [51], and reduced social support [53]; and (3) in contrast, greater scores on the CCPS Subscale would be inversely associated with negative psychosocial outcomes, in that showing increased beliefs that HIV stigma could be ameliorated by achieving the local cultural capabilities core to ‘womanhood’ would be associated with greater self-esteem, fewer depressive symptoms, and increased social support; and finally that among the sample of individuals with unknown HIV status, (4) both WMM-WLHIV-BW subscales will show attenuated associations with the psychosocial outcomes identified above (i.e., self-esteem, depressive symptoms, and social support), because even if this sample perceived culturally-based HIV stigma, this would presumably not be as salient to (i.e., strongly associated with) their psychosocial outcomes. Psychometric analyses were conducted using Stata 15.1 [55] and significance testing assessed at the p<0.05 level (two-tailed).

Results

Socio-demographic Characteristics

We surveyed 201 respondents: 102 individuals with known HIV-positive status ascertained from the HIV clinic and 99 individuals with unknown HIV status residing in urban Gaborone (Table 2). Overall, the average age was 44.2 years (SD=13.4) and 49.8% were male. Nearly half (48.8%) indicated they were in a long-term relationship (marriage and/or living with partner). About two-thirds (68.2%) had not completed high school and nearly two-thirds (60.5%) were not employed full-time. These demographics are comparable to those reported in a recent, nationally representative household survey [56]. We did not ask individuals with unknown HIV status directly about their HIV status, but nearly all respondents (97.0%) reported having taken an HIV test at least once. Slightly fewer (81.8%) had taken an HIV test in the past twelve months, but almost all (98.0%) said they would be willing to do so. All respondents identified themselves as Black Africans and citizens of Botswana.

Table 2.

Demographic characteristics (n=201)

	Overall				HIV patients (IDCC)				General community
	n (%)	Mean (±SD)	Median (IQR)	Range	n (%)	Mean (±SD)	Median (IQR)	Range	n (%)	Mean (±SD)	Median (IQR)	Range	p-value
Age	201	44.2 (±13.4)	44 (35–54)	18–78	102	47.4 (±10.2)	46 (42–55)	20–74	99	40.8 (±15.4)	38 (28–51)	18–78	<0.001
Sex – male	100 (49.8)				51 (50.0)				49 (49.5)				0.943
Relationship status – married	45 (22.4)				32 (31.4)				13 (13.1)				0.002
Education – less than secondary school	137 (68.2)				75 (73.5)				62 (62.6)				0.097
Employment – full-time	79 (39.5)				49 (48.0)				30 (30.6)				0.012
HIV testing
Ever tested									96 (97.0)
Past 12 months									81 (81.8)
Willing to test									97 (98.0)
Depression (CES-D)	198	22.7 (±15.6)	20 (8–36)	0–54	101	27.1 (±16.0)	28 (12–42)	0–54	97	18.2 (±13.9)	17 (6–28)	0–51	<0.001
Self-Esteem (Rosenberg)	200	19.3 (±3.9)	19 (17–22)	9–29	102	20.1 (±4.3)	19.5 (17–23)	9–29	98	18.4 (±3.2)	18 (17–20)	10–27	0.004
Social Support Number (Sarason)	201	2.1 (±1.3)	1.8 (1.2–2.7)	0–9	102	2.3 (±1.1)	2 (1.5–3.0)	0.67–6	99	1.9 (±1.4)	1.5 (1–2.5)	0–9	<0.001
Social Support Satisfaction (Sarason)	198	5.9 (±0.4)	6 (6–6)	3.4–6	102	5.9 (±0.3)	6 (6–6)	4.67–6	96	5.8 (±0.4)	6 (5.8–6)	3.4–6	0.006
HIV Stigma (Berger)					99	102.1 (±21.5)	103 (89–118)	48–160
Cultural Factors Shape Subscale	199	25.1 (±6.0)	24 (21–30)	10–40	101	26.4 (±6.9)	28 (22–31)	10–40	98	23.7 (±4.6)	22 (21–27)	12–36	<0.001
Cultural Capabilities Protect Subscale	200	30.1 (±4.6)	30 (28–31)	16–40	101	31.4 (±5.0)	30 (29–35)	18–40	99	28.8 (±3.8)	30 (28–30)	16–40	<0.001

^*Due to missing data, n may not add to the sample totals; p-values were calculated using the Wilcoxon rank-sum test for continuous variables and Pearson’s chi-squared test for categorical variables.

Individuals with known and unknown HIV status differed on a number of factors. On average, individuals with known HIV-positive status were older (average age 47.4 years vs. 40.8 years, p<0.001), were more likely to be married (31.4% vs. 13.1%, p=0.002), and to be employed (48.0% vs. 30.6%, p=0.012). Regarding stigma measures, individuals with known HIV-positive status showed significant differences in the WMM-WLHIV-BW subscales compared to those with unknown HIV status, including greater endorsement of both the CFSS Subscale (26.4 (6.9) vs. 23.7 (4.6), p<0.001) and the CCPS Subscale (31.4 (5.0) vs. 28.8 (3.8), p<0.001). Regarding associated psychosocial constructs, as expected, individuals with known HIV-positive status reported higher scores than those with unknown HIV status on depressive symptoms (27.1 (16.0) vs. 18.2 (13.9), p<0.001). Somewhat unexpectedly, individuals with known HIV-positive status reported higher self-esteem (20.1 (4.3) vs. 18.4 (3.2), p=0.004), and higher social support both in number (2.3 (1.1) vs. 1.9 (1.4), p<0.001) and satisfaction (5.9 (0.3) vs. 5.8 (0.5), p=0.006) when compared with individuals with unknown HIV status.

Exploratory Factor Analysis (EFA)

An initial EFA using all 40 items yielded two factors that aligned with the two subscales (“CFSS” and “CCPS”) with eigenvalues higher than 1.0, accounting for about half the variance each (Factor 1: eigenvalue=9.59, proportion=0.52; Factor 2: eigenvalue: 9.26, proportion=0.50) indicating a two-factor solution. To preserve content validity, the highest loading item within each qualitative WMM theme was used to select the final 20 items (10 items per subscale). In the final 20-item version, a subsequent EFA further supported the two subscales with a two-factor solution with eigenvalues higher than 1.0, accounting for about half the variance each (Factor 1: eigenvalue=5.18; proportion=0.51; Factor 2: eigenvalue: 5.16; proportion=0.51). Factor loadings for individual items onto a single subscale were all greater than 0.30 and nearly all were greater than 0.50 (loading range, Factor 1: 0.46–0.87, Factor 2: 0.38–0.86), indicating that factors uniquely loaded onto one of the two subscales (factor loadings typically <0.10 for the other subscale). See Table 3.

Table 3.

WMM Stigma item means, percent endorsement by response category, factor loadings, and internal consistency reliability (n=199)

	Item	Percent endorsement by category							Factor loadings
	Mean (SD)	SD		D		A		SA	I		II
Cultural Factors Shape
1. Most Batswana believe that a woman with HIV is not a devout Christian.	2.5 (0.8)	9.1%		43.2%		35.7%		12.1%	0.46		0.09
2. Most Batswana believe that a woman with HIV got HIV because she was promiscuous to get money.	2.7 (0.9)	8.0%		35.5%		35.0%		21.5%	0.69		0.08
3. Most Batswana believe that a woman with HIV will not be seen as a respectful wife.	2.6 (0.8)	8.0%		35.0%		41.5%		15.5%	0.73		0.05
4. Most Batswana believe that a woman with HIV is unable to take care of her children.	2.3 (0.8)	15.0%		48.5%		31.0%		5.5%	0.67		−0.12
5. Most Batswana believe that a woman with HIV does not have the ability to properly take care of her home.	2.3 (0.8)	12.5%		53.0%		27.5%		7.0%	0.58		−0.16
6. Most Batswana believe that a woman with HIV is not a reliable foundation for her family.	2.4 (0.8)	10.0%		44.0%		39.0%		7.0%	0.87		−0.04
7. Most Batswana believe that a woman with HIV has not been raised by her family to be disciplined and respectful.	2.3 (0.8)	12.0%		50.5%		29.0%		8.5%	0.81		0.00
8. Most Batswana believe that a woman with HIV does not respect herself.	2.4 (0.8)	12.5%		46.0%		32.0%		9.5%	0.83		−0.02
9. Most Batswana believe that a woman with HIV was promiscuous before marriage.	2.6 (0.9)	9.0%		37.5%		34.5%		19.0%	0.80		0.05
10. Most Batswana believe the community will gossip a lot about a woman with HIV.	2.9 (0.8)	5.5%		21.0%		53.0%		20.5%	0.49		0.11
Cultural Capacities Protect
11. Most Batswana believe that a woman with HIV who is a God-fearing woman will not feel ashamed about her HIV status.	2.9 (0.6)	2.5%		16.0%		67.0%		14.5%	0.02		0.38
12. Most Batswana believe a woman with HIV who has a man who can provide for her will be respected by others.	2.8 (0.7)	5.5%		21.5%		60.5%		12.5%	0.08		0.42
13. Most Batswana believe that an employed woman with HIV who provides for the family should be respected.	3.0 (0.7)	3.5%		13.5%		67.5%		15.5%	0.03		0.44
14. Most Batswana will not look down on a woman with HIV who is loyal to her husband.	3.0 (0.7)	3.0%	14.5%		62.0%		20.5%			−0.04	0.78
15. Most Batswana will not look down on a woman with HIV who has children.	2.9 (0.7)	2.5%	18.5%		63.0%		16.0%			−0.09	0.66
16. Most Batswana respect a woman with HIV who raises successful children.	3.1 (0.6)	1.0%	10.5%		67.5%		21.0%			0.05	0.81
17. Most Batswana will respect a woman with HIV who properly cares for her husband.	3.1 (0.6)	2.0%	9.5%		67.5%		21.0%			−0.05	0.83
18. Most Batswana respect a woman with HIV who takes care of herself and looks attractive.	3.1 (0.6)	2.5%	9.5%		65.5%		22.5%			−0.02	0.86
19. Most Batswana would give Botho to a woman with HIV who treats others with Botho.	3.2 (0.6)	1.5%	5.5%		63.5%		29.5%			0.06	0.73
20. Most Batswana will respect a woman with HIV who looks healthy and attractive.	3.1 (0.6)	1.5%	10%		67.5%		21%			0.04	0.86
		Cronbach’s Alpha – α								0.90	0.90

Response categories: SD = strongly disagree, D = disagree, A = agree, SA = strongly agree; exploratory factor analysis specifying 2 factors and oblique rotation; internal consistency reliability (α) for each subscale based on items in bold

Scale Reliability

For the final 20-item version of the WMM-WLHIV-BW, Cronbach’s alpha for each WMM subscale was 0.90. Test-retest reliability (n=28) was also very high for the CFSS Subscale (p=0.92) and CCPS Subscale (p=0.85). We found no differences by gender (man vs. woman) or sample group (individuals with known HIV-positive status vs. individuals with unknown HIV status).

Scale Validity

To assess initial construct validity, we first assessed convergent validity among women with known HIV-positive status. Among this group, the CFSS Subscale showed a positive, trend association with the Berger HIV Stigma Scale (ρ=0.24, p=0.095). Conversely, the CCPS Subscale showed a negative, trend association (ρ=−0.24, p=0.093) (Table 4). Regarding psychosocial constructs, the CFSS Subscale and the CCPS Subscale showed some distinctive patterns (Table 4). The Berger HIV Stigma Scale was positively, moderately and significantly associated with depressive symptoms (CES-D, ρ=0.57, p<0.001) and negatively, modestly and significantly associated with self-esteem (Rosenberg, ρ=−0.42, p=0.003) and social support number (Sarason, ρ=−0.34, p=0.015). In comparison, the CFSS Subscale was only significantly correlated with depressive symptoms (CES-D, ρ=0.39, p=0.005). However, the CCPS Subscale showed significant, modest positive correlations with self-esteem (Rosenberg, ρ=0.32, p=0.026) and social support number (Sarason, ρ=0.29, p=0.047). That is, as endorsement of the CCPS Subscale increased, endorsement of self-esteem and social support increased.

Table 4.

Construct validity of “WMM Cultural Stigma Scale for Women Living with HIV in Botswana” Subscales in relation to an established measure of stigma and associated psychosocial outcomes

	Women with known HIV status (n=49)			Men and women in the community (n=92)
	Cultural Factors Shape Subscale	Cultural Capabilities Protect Subscale	Berger HIV Stigma Scale1	Cultural Factors Shape Subscale	Cultural Capabilities Protect Subscale
Cultural Factors Shape Subscale	1			1
Cultural Capabilities Protect Subscale	0.12 p=0.423	1		0.22* p=0.034	1
Berger HIV Stigma Scale1	0.24 p=0.095	−0.24 p=0.093	1
Depression Scale (CES-D)	0.39* p=0.005	−0.02 p=0.896	0.57* p<0.001	0.10 p=0.308	−0.06 p=0.561
Self-Esteem Scale (Rosenberg)	−0.02 p=0.868	0.32* p=0.026	−0.42* p=0.003	0.05 p=0.662	−0.13 p=0.188
Social Support Number (Sarason)	0.02 p=0.912	0.29* p=0.047	−0.34* p=0.015	0.19 p=0.058	−0.01 p=0.936
Social Support Satisfaction (Sarason)	−0.01 p=0.932	−0.005 p=0.973	0.02 p=0.907	−0.07 p=0.487	0.08 p=0.429

Calculated using Spearman’s rank correlation coefficient (ρ)

^*p-value < 0.05

¹only individuals with known HIV status

When next examining convergent validity of the WMM-WLHIV-BW subscales among community members with unknown HIV status, as expected there were minimal and non-significant correlations observed between each subscale and depressive symptoms, self-esteem, social support number, and social support satisfaction (all p values >0.05; Table 4).

Discussion

We applied an influential theory of culture and stigma to develop and provide initial validation of an HIV stigma scale assessing culturally-salient dynamics in Botswana. Prior formative qualitative research indicated strong, distinct gender roles in a culture where women typically have a subordinate position. The intersectional identities [57,58] of being a woman and living with HIV lead to a distinct lived experience of stigma within Setswana culture that must be considered simultaneously. Given this, in this study we developed and initially validated a scale focused on cultural aspects of stigma for women living with HIV. The 20-item WMM-WLHIV-BW consisted of two distinct subscales confirmed by the factor analytic findings, representing culturally-distinct aspects that appear to shape, and protect against, HIV stigma. Each subscale’s initial reliability was strong for both internal consistency and test-retest reliability. Developing items via formative qualitative research and review by local experts provided content validity. Each WMM-WLHIV-BW subscale showed distinct initial construct validity with an established stigma scale and psychosocial outcomes (i.e., depression, self-esteem, and social support). Among women with known HIV-positive status, the Berger HIV Stigma Scale was correlated with greater depressive symptoms as well as lower self-esteem and fewer individuals providing social support. In comparison, in addition to a trend-level, positive correlation with the Berger HIV Stigma Scale, the CFSS Subscale was only correlated with greater depressive symptoms, thus evidencing somewhat weaker construct validity when compared with the Berger HIV Stigma Scale. These overall findings are consistent with a meta-analysis of correlates of HIV stigma, which found moderate associations between HIV stigma and psychosocial outcomes including depression [3].

Conversely and as hypothesized, the CCPS Subscale showed distinct, initial construct validity compared to both the Berger HIV Stigma Scale and the CFSS Subscale among women with known HIV-positive status. Specifically, the CCPS Subscale showed significant positive correlations with psychosocial outcomes (i.e., self-esteem and social support number). That is, as respondents showed increased beliefs that HIV stigma could be mitigated by achieving the local cultural capabilities central to “womanhood,” they also endorsed greater positive self-esteem and more individuals providing social support. Despite many of the same themes being captured by the two subscales, given their differing pattern of correlations with these psychosocial constructs, their minimal association with each other (ρ=0.12, p=0.423), and their distinct conceptual underpinnings (i.e., for the CCPS Subscale, that attaining WMM signals the achievement of “personhood” in one’s local community—see below), these subscales should be treated separately in future use. As hypothesized, no correlations were found between the two WMM-WLHIV-BW subscales and psychosocial outcomes among the respondents with unknown HIV status, as perceived stigma would not be expected to affect people who are not known to have HIV. In sum, our analyses in these initial validation samples indicate that the CFSS Subscale may function similarly to other HIV stigma scales but that the CCPS Subscale, guided by theory in which the cultural capabilities “most valued” by a cultural group could facilitate social reintegration and thereby lessen HIV stigma, shows initial protective properties in relation to psychosocial outcomes.

Strengths and Implications

It has been stated that the core components of stigma are sufficiently consistent so that generic stigma scales may be reliably used across settings [20,59], but it has also been argued that universal measures may insufficiently account for culturally-meaningful aspects of stigma that may enhance prediction of important outcomes [32]. Our initial study illustrates how measures that incorporate culturally-meaningful aspects can operationalize valuable psychometric concepts neglected by many existing measurements. Regarding the CCPS Subscale, the degree to which women with known HIV-positive status endorsed that achieving capabilities core to “womanhood” were linked with mitigation of stigma was associated with higher levels of self-esteem and greater number of social support contacts, relationships not captured by traditional stigma scales like the Berger HIV Stigma Scale. By measuring the achievement of cultural capabilities core to “womanhood,” this framework identified everyday engagements that could facilitate social reintegration and could be linked with other positive outcomes (e.g., empowerment) [47]. The CCPS further aligns with recent developments in measuring resilience among people living with HIV [60,61]. Although resilience while living with HIV is meant to capture the ability to overcome HIV-related adversity (including, but not limited to, stigma) [60], operationalization of cultural concepts core to “personhood” into quantitative measures, and their linkages to mitigating the experience of HIV stigma specifically, remains an underdeveloped area.

Although qualitative studies have identified factors associated with resistance to HIV stigma in sub-Saharan Africa [22,62], the CCPS Subscale is, to our knowledge, one of the first quantitative measures to define the everyday cultural engagements that signal “full personhood” that could facilitate social integration into a local context, and could be considered alongside stigma reduction efforts. The items comprising the CCPS Subscale parallel prior qualitative findings in South Africa that individuals are empowered to resist HIV stigma when they are able to participate in “social roles that have value” [62]. Although measures of constructs that signal resistance to HIV stigma have yet to be developed, we draw comparisons to a subscale intended to measure resistance to mental illness stigma [47]. The “stigma resistance” subscale of the Internalized Stigma of Mental Illness Scale (ISMI) has limited correlation to other ISMI stigma-related subscales and has been interpreted as capturing an independent construct. Likewise, and as noted above, the CCPS Subscale, despite substantial overlap in thematic content of items, shows minimal correlations with the CFSS Subscale (and the Berger HIV Stigma Scale) among women with known HIV-positive status. Like the CCPS Subscale, the stigma resistance subscale of the ISMI showed positive associations with measures of self-esteem and social support [47]. These similar patterns between our newly-developed subscale that assesses how aspects of “womanhood” might mitigate HIV stigma and a separate stigma resistance subscale indicate that these may assess distinct constructs in relation to traditional stigma domains. Further, while requiring future testing, the CCPS Subscale, in assessing constructs core to “personhood,” could be linked with other positive psychosocial outcomes, in addition to lessening the experience of stigma.

Identifying local cultural capabilities that define ‘personhood’, in tandem with other stigma-reduction strategies, could offer important advances in intervening with stigma. We recognize that unequal gender norms have been linked with negative health consequences [63,64] and that any potential adverse effects should be considered prior to recommending that women strictly adhere to culturally-sanctioned versions of womanhood. Ultimately, societal change is needed to decrease harmful rigidity in gender roles and reduce inequitable power dynamics between men and women. Nonetheless, a desire to achieve these notions of ‘womanhood’ were expressed by many women in our formative qualitative work. We believe that achieving the capabilities that “matter most” can be valuable when these culturally-aligned goals are participant-identified (i.e., that many women themselves wished to attain these capabilities, and felt challenged in doing so due to stigma) and accordingly, achieving these capabilities could provide potential new alternatives for culturally-salient interventions to reduce these women’s experience of stigma. Qualitative studies in South Africa have indicated that positive and supportive experiences following diagnosis with HIV can foster development of new identities and relationships that buffer against stigma [22,62]. Specifically, participating in meaningful social roles (e.g., helping others, supporting family) in the local setting enabled rebuilding of one’s self-image and resistance to stigma [62]. Our results suggest that interventions targeting perceived stigma can potentially be enhanced by supporting the achievement of the local cultural capabilities core to ‘womanhood’ (e.g., motherhood, respecting one’s husband, caring for the family) as an option, if desired by participants, among women with HIV in urban Botswana.

Limitations and future directions

Due to our study’s cross-sectional design, causality cannot be ascertained; it is possible that psychosocial outcomes could have shaped endorsement of the WMM-WLHIV-BW items. Additionally, the scale was psychometrically validated in a relatively small, convenience sample that was diverse both socio-demographically and by HIV status. A convenience sample of participants with known HIV-positive status was recruited from a specialized clinic that primarily serves patients with complicated cases of HIV (e.g., higher rates of resistance, often resulting from a longer history of treatment and accompanying ART non-adherence). Multiple studies in Sub-Saharan Africa have found attenuation in the level of HIV stigma as time passes following initial diagnosis [65–67]; despite this, the group with known HIV-positive status still showed significantly higher scores on the CFSS Subscale when compared with the group with unknown HIV status, suggesting that culturally-meaningful aspects of stigma may persist in a group with known HIV-positive status even after a long duration of illness. Likely resulting from convenience sampling, the respondents with known HIV-positive status were also on average older, employed, and married. Although no demographic factors have shown consistent correlations with internalized stigma in sub-Saharan Africa [68], these demographic differences may account for the unexpectedly higher levels of social support and self-esteem that were found among the respondents with known HIV-positive status when compared with their counterparts with unknown HIV status. By recruiting via a clinic setting, we captured only respondents currently engaged in care; however, people living with HIV in Botswana most severely impacted by stigma are likely missed given their higher likelihood of avoiding care. Future studies utilizing the WMM framework in the Botswana context would benefit from ascertaining larger, representative samples of individuals with known and unknown HIV status. Finally, while some scale items, particularly in the CCPS scale, appear to capture co-occurring concepts, these items reflect key intersectional aspects of WMM for women with HIV within Setswana culture.

In this initial validation of the WMM-WLHIV-BW, we sampled both men and women with known and unknown HIV-positive status because WMM, even while gendered, is conceptually shared by all members of a cultural group. However, since this scale focused on the effects of HIV stigma among women, we restricted convergent validity analyses to women living with HIV. Examining what matters most among men and devising a separate scale, using the same processes outlined here, to assess the effects for men living with HIV is an important next step. Additional qualitative findings by our team found that WMM for men in Botswana centered on fulfilling male responsibilities of being a capable provider within the family and maintaining social status in the wider community, that being identified with HIV could threaten the ability to meet these goals, and that achieving WMM via employment, relationships, and physical health could protect against stigma [69].

Measuring culturally-meaningful stigma constructs can elicit participant-identified perspectives that in certain circumstances could be incorporated alongside stigma reduction interventions. Building on the conceptualization that stigma can exert powerful effects by threatening what is “most at stake” in a local cultural context [9], interventions that encourage participants to bolster concepts operationalized by the WMM-WLHIV-BW may potentially facilitate social reintegration and reduce culturally-meaningful aspects of perceived stigma. Future research can aim to develop and validate measures of internalized stigma grounded in culturally-salient concepts to better assess how cultural aspects of perceived stigma become internalized [70]. Additionally, interventions such as psychoeducation and cognitive therapy-based practices to reduce perceived stigma [71,72] can aim to reduce harmful, culturally-salient perceived stigma before it is internalized by individuals with HIV. If appropriate, efforts can also be made to promote participation in the activities that “matter most” as signified by the CCPS Subscale. By supporting women to achieve the local cultural capabilities central to ‘womanhood’ in Setswana culture, thereby potentially bolstering social integration in one’s community and reducing the experience of stigma, interventions incorporating WMM present a promising alternative towards potentially retaining (or restoring) status as a “good woman” to ameliorate the adverse psychosocial burdens of living with HIV.

Conclusions

Because psychometric measurement of the cultural impacts on HIV stigma has not been assessed systematically, we apply a formulation that stigma threatens ‘what matters most’ among women in urban Botswana to provide a new, replicable, and methodologically rigorous means to operationalize cultural constructs that may be related to psychosocial outcomes. In addition to the WMM-WLHIV-BW subscales showing robust initial reliability and construct validity, the CCPS Subscale adds to existing stigma measures by assessing culturally-meaningful aspects of “personhood” that may also mitigate the experience of HIV stigma. Because this subscale appears distinct from established HIV stigma measures, it can be used alongside traditional measures of HIV stigma to capture culturally-derived components of “personhood” that may confer additional protection against stigma. While we advocate for the incorporation of WMM into measurement (and depending on circumstances, stigma interventions), we also believe that simultaneous efforts to induce societal change to address inequitable gender norms should occur; as women’s accepted gender roles gradually change, WMM (and its measurement, and integration into stigma interventions) would correspondingly also change. Nonetheless, establishing initial psychometric properties of a culturally-salient HIV stigma measure signifies an advance in capturing important cultural variation in stigma. Our scales could be administered in other SSA cultures where women may share similar core everyday engagements [73] to examine to what extent these concepts are relevant. This would likely require local adaptation of at least some items for each setting (thus yielding new, culturally-salient items); however, at least some core items are likely to apply to the new setting(s). This would enable quantitative comparability of these original items and illuminate culturally-salient forms of stigma that may capture important experiences of stigma regionally [32].

In closing, this type of intensive research effort may be needed to help address “the last mile” for key and vulnerable populations (e.g., adolescent girls and young women) who remain especially susceptible to HIV infection [74]. The need for this type of approach is epitomized in Botswana, where stigma remains a potent barrier to accessing otherwise effective and free ART [40]. This and similar types of approaches to capture important experiences of intersectionality could help to identify and address culturally-meaningful types of stigma in these vulnerable groups. We seek to encourage further development of theoretically-driven, culturally-salient stigma measures to help mitigate crucial barriers to accessing and implementing HIV services in Botswana, sub-Saharan Africa, and other contexts.

Data/Code Availability

Data will be made available based upon written request to the two co-first authors, Dr. Yang and Mr. Ho-Foster.