The Potential Role of Undetectable=Untransmittable (U=U) in Reducing HIV Stigma Among Sexual Minority Men in the US

Abstract

The Undetectable=Untransmittable (U=U) message and its scientific underpinnings have been widely suggested to reduce HIV stigma. However, misunderstanding and skepticism about U=U may prevent this destigmatizing potential from being fully realized. This cross-sectional study examined associations between U=U belief (belief that someone with a sustained undetectable viral load has zero risk of sexually transmitting HIV) and HIV stigma among US sexual minority men. Differences by serostatus and effects of brief informational messaging were also explored. The survey was completed online by 106 men living with HIV and 351 HIV-negative/status-unknown men (2019–2020). Participants were 18–83 years old (M[SD]=41[13.0]). Most were non-Hispanic White (70.0%) and gay (82.9%). Although nearly all participants (95.6%) were aware of U=U, only 41.1% believed U=U. A greater percentage of participants living with HIV (66.0%) believed U=U compared with HIV-negative/status-unknown participants (33.6%). Among participants living with HIV, U=U belief was not significantly associated with perceived, internalized, or experienced HIV stigma or with viral load prejudice (prejudice against people who have a detectable HIV viral load). Among HIV-negative/status-unknown participants, U=U belief was associated with less frequently enacted HIV discrimination, more positive feelings toward people with an undetectable viral load, and lower personal endorsement of stigmatizing beliefs. Brief informational messaging about U=U did not affect most stigma dimensions and did not favorably affect any. Interventions are needed to correct commonly held, outdated misconceptions about HIV transmission risk. Such initiatives must not only engage people living with HIV but also engage HIV-negative/status-unknown people to maximize the destigmatizing potential of U=U.

Introduction

The scientific evidence is clear: People living with HIV (PLWH) who have a sustained undetectable viral load cannot sexually transmit HIV to their partners [1–5]. Undetectable=Untransmittable (U=U) is a phrase coined by the Prevention Access Campaign to represent and communicate this concept as part of their global initiative to raise awareness [6]. Although awareness about U=U has indeed increased over time among sexual minority men (SMM)¹, widespread disbelief persists [8–13]. Belief in the concept of U=U is a necessary precursor to realizing many of the psychosocial benefits of U=U, one of which may be reducing HIV stigma. The current cross-sectional survey study examined associations between U=U belief and multiple dimensions of HIV stigma among a US sample of SMM, including both men living with HIV and HIV-negative/status-unknown men.

HIV stigma refers to social devaluation and discrediting of an individual or group based on their HIV-positive serostatus [14, 15]. HIV stigma is pervasive in the US and around the world [16–20]. For example, in a nationally representative sample of US adults surveyed in 2015, 18% reported being afraid to be around a person with HIV due to concerns about becoming infected [20]. Within a national probability sample of US PLWH interviewed in 2018–2019, over 40% agreed that most people think a person with HIV is disgusting [16]. The prevalence of HIV stigma and its adverse implications for the wellbeing of PLWH has led health officials to include HIV stigma reduction among the goals of national and global HIV strategies [21–23].

There are multiple cognitive, affective, and behavioral dimensions of HIV stigma [14, 24, 25]. Some dimensions are relevant to all people, irrespective of HIV status, and others are specific to PLWH or to HIV-negative/status-unknown people. A dimension of HIV stigma relevant to all people is perceived societal HIV stigma (“attributed stigma”), or an individual’s perception of negative beliefs about PLWH held by others in their community [25]. Dimensions of HIV stigma specific to PLWH include internalized HIV stigma (“self-stigma”), or personal endorsement of negative beliefs about HIV and associated self-devaluation; and experienced HIV discrimination, or unfair treatment received from others because of one’s HIV-positive status [14, 24]. Dimensions of HIV stigma specific to HIV-negative/status-unknown people include HIV prejudice, which can manifest as disgust, fear, and other negative feelings towards PLWH because of their HIV-positive status; personal HIV stigma, or individually held negative beliefs about PLWH; and enacted HIV discrimination, or unfair treatment of PLWH because of their HIV-positive status [14, 24, 25].

U=U has the potential to destigmatize HIV among both PLWH and HIV-negative/status-unknown people. PLWH’s knowledge of their untransmittable status can reduce internalized HIV stigma directly by assuaging prior concerns about posing a risk to others and restoring a sense of normalcy [26]. Additionally, reductions in community stigma—including negative perceptions, prejudice, and discrimination enacted against PLWH by the HIV-negative/status-unknown majority—can indirectly alleviate internalized stigma and directly reduce experienced discrimination among PLWH. A theorized driver of these various forms of stigma perpetrated against people with HIV is self-protection [27–31]. Among HIV-negative/status-unknown people, U=U has the potential to reduce the perceived threat of contagion associated with HIV and consequent motivation for self-protection, thereby averting the emotional and behavioral responses these conditions are theorized to evoke (e.g., disgust, avoidance) [29].

To date, research examining linkages between U=U and HIV stigma remains sparse [32–36]. Existing findings have generally corroborated theoretical associations between U=U and HIV stigma among PLWH [33, 34]. For example, in a study with over 30,000 US SMM with HIV, 82% reported that the U=U message made them feel better about their HIV status, and 79% believed that it would favorably impact HIV stigma [34]. Studies examining associations between U=U and HIV stigma among HIV-negative/status-unknown people have yielded mixed findings [32, 35]. For example, among 182 HIV-negative/status-unknown US SMM with pre-existing awareness of the U=U message, perceived accuracy of the U=U message (an indicator of U=U belief) was not significantly associated with perceived societal HIV stigma [32]. In a US sample of HIV-negative Black and Latinx heterosexual adults, awareness of the concept of U=U was not associated with anticipated HIV stigma (e.g., expected discrimination by others) in general but was associated with anticipated HIV stigma with respect to prospective romantic/sexual partners [35]. Although these and other early studies offer foundational insights, they are few in number and employ differing measures of U=U and HIV stigma, precluding clear consensus. A fuller picture of the empirical associations between U=U belief and HIV stigma—including cognitive, affective, and behavioral dimensions of HIV stigma—is needed to better understand the destigmatizing potential of the U=U message.

In addition, it is important to consider the possibility that rather than eradicating HIV stigma, the U=U message and belief therein instead shifts the burden of HIV stigma to PLWH who do not or cannot attain an undetectable viral load [37, 38]. Advances in medicine, including the evolution of highly effective treatments and the accumulation of evidence for U=U, have fostered the development of subclasses of HIV status based on viral load detectability, to which social value has been differentially ascribed [26, 39, 40]. An undetectable status has been characterized as an achievement and a source of comfort, acceptance, and resilience relative to a detectable status. Accordingly, a detectable status has been experienced as feeling “dirty” or “dangerous” and has served as the basis for social rejection [26, 41–43]. For example, HIV-negative sexual minority men, including those on PrEP, have previously expressed less openness to “bridging the serodivide” (having sex with PLWH) when their prospective sex partners who are living with HIV have a detectable rather than undetectable viral load [39, 44]. The idealization of “achieving” an undetectable viral load and the common framing of “being undetectable” as an expected and desired identity symbolic of social responsibility simultaneously casts PLWH who are not virally suppressed as deviant, infectious, and immoral [37, 38]. The slogan Viral Load Does Not Equal Value (V≠V) has been advanced out of concern that the focus on undetectability could contribute to the devaluation of PLWH who have a detectable viral load [45, 46]. In raising this specter, community members and other stakeholders have asserted that HIV treatment (and thus viral load undetectability) is not universally accessible—largely due to structural determinants of health and other inequities—and, even when accessible, is a personal choice that should be respected regardless of outcome [45–47]. To our knowledge, the linkage between U=U belief and stigma related to viral load detectability has yet to be examined quantitatively.

As the U=U movement continues to expand worldwide, it is valuable to consider how specific facets of the messaging itself (e.g., content, delivery) may affect stigma or moderate the association between U=U belief and stigma. Several studies offer early promise that messaging can favorably affect U=U belief and HIV stigma [48, 49]. For example, one study conducted with SMM in the Southern US reported increases in understanding of U=U and acceptance of the message following a U=U social media campaign [48]. Another study with HIV-negative/status-unknown heterosexuals in Japan found that providing information about U=U in an online survey helped to reduce HIV stigma [49]. Other studies have shown limited effects. Research led by Coyne and colleagues [50, 51] that investigated the effects of U=U message framing on U=U belief and HIV stigma found no evidence that risk- vs. protective-framed messages affected either outcome [51] or that evidence-based vs. opinion-based vs. unrelated health messaging affected U=U belief [50]. We are unaware of prior research that has examined how U=U messaging might influence (moderate) associations between pre-existing beliefs about U=U and stigma.

Study Objectives and Hypotheses

The primary objective of the current study was to examine associations between U=U belief and multiple dimensions of stigma among US SMM. Specifically, among men living with HIV, we examined associations between U=U belief and perceived societal HIV stigma, internalized HIV stigma, experienced HIV discrimination, and—among those reporting an undetectable viral load—viral load prejudice (i.e., prejudice against PLWH who have a detectable viral load). Among HIV-negative/status-unknown men, we examined associations between U=U belief and perceived societal HIV stigma, enacted HIV discrimination, feelings toward PLWH with a detectable viral load, feelings toward PLWH with an undetectable viral load, personal HIV stigma, and viral load prejudice. Based on prior empirical findings [26, 34], we hypothesized that U=U belief would be associated with less perceived societal HIV stigma, lower internalized HIV stigma, and less frequently experienced HIV discrimination among men living with HIV. We did not make other specific hypotheses about the associations between U=U belief and stigma among either serostatus group given limited and/or inconsistent supporting literature.

A second objective of the study was to examine the relevance of HIV status to U=U belief and the two forms of stigma measured among both participants living with HIV and HIV-negative/status-unknown participants (i.e., perceived societal HIV stigma and viral load prejudice), including whether HIV status was directly associated with stigma or moderated associations between U=U belief and stigma. Although these analyses were largely exploratory, we hypothesized based on prior empirical studies [11–13] that a higher percentage of participants living with HIV vs. HIV-negative/status-unknown participants would believe U=U.

A third, exploratory objective of the study, which builds on previous work examining U=U messaging effects [48–51], was to assess whether brief informational messaging about U=U impacted stigma or moderated associations between U=U belief and stigma.

Methods

Some of the methods described here have been reported in other articles for which data were collected as part of the same study [44, 52].

Participants and Procedures

All study procedures were reviewed and approved by the George Washington University Office of Human Research (IRB #NCR191092). Participants were recruited from across the US to participate as part of a larger online survey study in 2019–2020. Nearly all (95.6%) were recruited via dating apps and social media. Sixteen (3.5%) were referred by a friend or partner; three (0.7%) were recruited through in-person or online referral by HIV service providers in Washington, DC; and one (0.2%) was recruited from another, unspecified source. To be eligible, individuals needed to be able to read and answer questions in English, be 18 years of age or older, identify as a “man” or “transgender man,” and report having sex with men. Following the initial online screening and consent process, participants completed the online survey and were compensated via entry into a US$250 gift card lottery.

Measures

Background Characteristics

Sociodemographic variables assessed included age, race/ethnicity, sexual orientation, education, and annual household income. We also evaluated prior awareness of U=U with a single question: “Have you ever heard of ‘Undetectable = Untransmittable’ or ‘U=U’?” Corresponding response options of “Yes,” “No,” and “I don’t know” were recoded as (1) Yes or (0) No/Don’t know. We determined HIV status by asking, “What is your HIV status?,” with response options of “Living with HIV (HIV-positive),” “HIV-negative,” or “I don’t know.” Participants who reported that they were living with HIV were asked, “Was your most recent viral load detectable or undetectable?,” with response options of “Undetectable (too low to be detected),” “Detectable,” and “I don’t know/remember.”

U=U Belief

U=U belief was assessed among all participants with a single question: “If someone with HIV has been virally suppressed (has an undetectable viral load) for 6 months or more, what do you think the risk of that person sexually transmitting the virus to an HIV-negative partner is?” Response options included: “0% (they definitely will NOT sexually transmit HIV),” “1–25% (low risk),” “26–50% (moderate risk),” “51–75% (high risk),” “76–99% (very high risk),” and “100% (they definitely WILL sexually transmit HIV).” The “0% (they definitely will NOT sexually transmit HIV)” response option was recoded as (1) U=U Belief. All other response options were recoded as (0) U=U Disbelief given that all of them indicated some level of perceived risk, which is contrary to the concept of U=U.

Stigma

Stigma Among Participants Living with HIV.

We measured four dimensions of HIV and viral load stigma among participants living with HIV: perceived societal HIV stigma, internalized HIV stigma, experienced HIV discrimination, and viral load prejudice.

Perceived societal HIV stigma was measured with Visser et al.’s (2008) Attributed Stigma Scale [25], which includes 12 statements about how “most people” think or feel about people with HIV. Sample items include: “Most people think less of someone because they have HIV” and “Most people are afraid to be around people with HIV.” Participants responded by answering (1) “Agree” or (0) “Disagree.” We calculated an overall scale score from the 12 items as a mean ranging from 0 to 1. Higher scores indicated more perceived societal stigma. Cronbach’s alpha of the 12-item perceived societal stigma measure in the current sample was 0.91.

Internalized HIV stigma was measured using Visser et al.’s (2008) Internalised Stigma Scale [25], which captures the personal beliefs held by a person with HIV about their HIV status and parallels the Attributed Stigma Scale with respect to item foci and response options. Sample items include: “I think less of myself because I have HIV” and “People are right to be afraid of me because I have HIV.” We calculated an overall scale score from the 12 items as a mean ranging from 0 to 1. Higher scores indicated greater internalized stigma. Cronbach’s alpha of the 12-item internalized stigma measure in the current sample was 0.75.

Experienced HIV discrimination was assessed using Eaton et al.’s (2020) HIV Microaggressions Scale [53], which includes 13 items describing subtle acts of discrimination related to HIV status. Sample items include: “You heard about someone being outed about their HIV status” and “Someone’s body language showed you that they were bothered because of your HIV status.” Participants rated the frequency of each act occurring over the past month on a four-point scale ranging from (1) Never to (4) Often. We calculated an overall scale score from the 13 items as a mean ranging from 1 to 4. Higher scores indicated more frequently experienced discrimination over the past month. Cronbach’s alpha of the 13-item experienced HIV discrimination measure for the current sample was 0.88.

Finally, viral load prejudice was assessed using a single item that was adapted from Eaton et al.’s (2020) HIV Microaggressions Scale [53]: “You were bothered because someone had a detectable viral load.” Participants rated the frequency of being bothered over the past month on a four-point scale ranging from (1) Never to (4) Often. “Rarely,” “Sometimes,” and “Often” response options were recoded as (1) Any Viral Load Prejudice, and “Never” response options were recoded as (0) No Viral Load Prejudice. The viral load prejudice measure was only administered to the subset of participants living with HIV who reported an undetectable viral load.

Stigma Among HIV-Negative/Status-Unknown Participants.

We measured six dimensions of HIV and viral load stigma among HIV-negative and status-unknown participants: perceived societal HIV stigma, enacted HIV discrimination, feelings toward PLWH with a detectable viral load, feelings toward PLWH with an undetectable viral load, personal HIV stigma, and viral load prejudice.

The perceived societal HIV stigma measure administered to the HIV-negative/status-unknown participants was identical to the one administered to participants living with HIV. Cronbach’s alpha of the 12-item perceived societal HIV stigma measure in the HIV-negative/status-unknown sample was 0.90.

Enacted HIV discrimination was assessed using five items adapted from Eaton et al.’s (2020) HIV Microaggressions Scale [53] to reflect unfair treatment of PLWH because of their HIV status. Items captured external behaviors that were directly or indirectly discriminatory. Eight of the original 13 scale items (e.g., “You saw an HIV positive person portrayed negatively in the media”) could not be directly adapted to reflect an individually enacted microaggression. Sample items include: “You outed someone for their HIV status” and “You did not include someone with HIV in a group event because of his/her HIV status.” Participants rated the frequency of each act occurring over the past month on a four-point scale ranging from (1) Never to (4) Often. We calculated an overall scale score from the five items as a mean ranging from 1 to 4. Higher scores indicated more frequently enacted discrimination over the past month. Cronbach’s alpha of the 5-item enacted HIV discrimination measure for the current sample was 0.60.

Feelings towards PLWH who had detectable and undetectable viral loads were assessed independently from one another as single items. Specifically, “HIV-positive people who are NOT virally suppressed (have detectable levels of virus)” and “HIV-positive people who are virally suppressed (have undetectable levels of virus)” were included in a list of 18 different groups, and participants were asked to rate their general feelings toward each of the groups on a nine-point response scale ranging from (1) Extremely Negative to (9) Extremely Positive. Other groups on the list included PrEP users and non-users, condom users and non-users, and multiple groups not defined by their HIV or prevention status (e.g., elderly people, politically liberal people, gay/lesbian/bisexual people, rich people). The order in which the groups were listed was randomized across participants.

Personal HIV stigma was measured with Visser et al.’s (2008) Personal Stigma Scale [25], which captures HIV-negative/status-unknown people’s beliefs about people with HIV and parallels the Attributed Stigma Scale and Internalised Stigma Scale with respect to item foci and response options. Sample items include: “I think less of someone because they have HIV” and “I feel afraid to be around people with HIV.” We calculated an overall scale score from the 12 items as a mean ranging from 0 to 1. Higher scores indicated greater personal stigma. Cronbach’s alpha of the 12-item personal HIV stigma measure for the current sample was 0.79.

Finally, the single-item viral load prejudice measure administered to the HIV-negative/status-unknown participants was identical to the one administered to participants living with HIV.

U=U Messaging (Exploratory Moderator)

After completing measures of background characteristics (including U=U awareness) and U=U belief, and prior to completing the stigma measures, participants were randomized to view one of three brief informational messages delivered in the form of a quote attributed to the leader of the Division of HIV/AIDS Prevention of the US Centers for Disease Control and Prevention (CDC). The first message, which communicated the concept of U=U, was adapted from a letter issued by the CDC in 2017 [54]. Adaptations of the text of the original letter included adding “completely” in front of “prevents,” deleting “effectively” where it originally appeared in front of “no risk,” and capitalizing “no risk”:

Antiretroviral therapy (ART) preserves the health of people living with HIV... When ART results in viral suppression, defined as less than 200 copies/ml or undetectable levels, it completely prevents sexual HIV transmission... This means that people who take ART daily as prescribed and achieve and maintain an undetectable viral load have NO RISK of sexually transmitting the virus to an HIV-negative partner.

There were two comparison messages, the first of which expressed the original Treatment as Prevention concept that viral suppression lowered risk; it was identical to the first message except that “completely” was replaced with “helps to” and “NO RISK” was replaced by “LOW RISK.”² The second comparison message omitted any information about transmission risk, replacing “completely… partner” with “helps keep people with HIV healthy,” essentially reiterating the first part of all three messages stating that ART preserves the health of PLWH.

An item that inquired about the level of HIV sexual transmission risk conveyed in the message was embedded as an attention/manipulation check toward the end of the survey, deliberately separated from the message by multiple measures. Participants were debriefed about the experiment following completion of all survey measures. Specifically, they were informed that key pieces of information in the message presented had been modified from the original source for the purposes of the study, and they were provided with an explanation of U=U and links to the original message source (CDC letter) and additional information about U=U.

In the present study, we explored whether brief informational messaging about U=U affected stigma. We report more proximal outcomes of this experiment, including participants’ open-response reactions to the messages and perceptions of message accuracy, elsewhere [52].

Analysis

Frequencies, means, and standard deviations were calculated to describe the sample and variables of interest. Chi-square tests (or Fisher’s two-sided exact tests if expected cell counts were prohibitive) were performed to test for differences in U=U awareness and U=U belief by serostatus group. Correlations, linear regressions, and logistic regressions were conducted to examine associations between U=U belief and stigma. Linear and logistic regressions were also performed to test associations of HIV status with the two forms of stigma measured among both participants living with HIV and HIV-negative/status-unknown participants (i.e., perceived societal stigma and viral load prejudice), including partial, conditional, and interaction effects of HIV status and U=U belief. Likewise, linear and logistic regressions were performed to test associations of U=U messaging (represented by two dummy-coded variables) with all forms of stigma, including partial, conditional, and interaction effects of U=U messaging and U=U belief. For the two forms of stigma measured among both participants living with HIV and HIV-negative/status-unknown participants, three-way U=U messaging × HIV status × U=U belief interactions were tested. Regression analyses involving U=U messaging were repeated restricting the analytic sample to participants who had passed the attention/manipulation check (i.e., correctly responded to the attention/manipulation check item based on the randomly assigned message they viewed). All regression analyses were adjusted for age, race/ethnicity, sexual orientation, and income.

Regression analyses involving participants living with HIV were repeated with the subsample of participants who reported their viral load to be undetectable to ensure findings remained consistent. There were too few participants who reported a detectable viral load to repeat analyses with that subsample.

Results

Sample Description

A total of 457 participants completed the survey, including 106 participants living with HIV, 333 participants who were HIV-negative, and 18 participants who did not know their HIV status. Of the participants living with HIV, 90.6% reported their most recent viral load status to be undetectable.

Sociodemographic characteristics of the sample are presented in Table 1. One participant (0.2%) identified as a “transgender man,” and the remainder of the sample (99.8%) identified as a “man.” (Note: Because of the wording of available response options, the exact percentage of participants who were cisgender cannot be inferred.) Participants ranged in age from 18 to 83, with a mean age of 41 (SD = 13.0). Most participants were non-Hispanic White (70.0%), identified as gay (82.9%), and had received a bachelor’s degree or higher education (62.6%). The majority (75.0%) earned an annual household income of at least US$30,000, and 30.0% earned an annual household income of at least US$90,000.

Table 1.

Sample Characteristics

	Full Sample n (%)	Participants Living with HIV n (%)	HIV-Negative/Status-Unkown Participants n (%)
Gender a
Man	456 (99.8)	106 (100.0)	350 (99.7)
Transgender Man	1 (0.2)	0 (0.0)	1 (0.3)
Age
18 – 25	55 (12.0)	6 (5.7)	49 (14.0)
26 – 35	126 (27.6)	22 (20.8)	104 (29.6)
36 – 45	101 (22.1)	19 (17.9)	82 (23.4)
46 – 55	93 (20.4)	30 (28.3)	63 (17.9)
56 or Older	82 (17.9)	29 (27.4)	53 (15.1)
Race/Ethnicity
Hispanic/Latino/x	59 (12.9)	14 (13.2)	45 (12.8)
Non-Hispanic White	320 (70.0)	70 (66.0)	250 (71.2)
Non-Hispanic Black	27 (5.9)	9 (8.5)	18 (5.1)
Non-Hispanic Other	51 (11.2)	13 (12.3)	38 (10.8)
Sexual Orientation
Gay	379 (82.9)	101 (95.3)	278 (79.2)
Bisexual	54 (11.8)	2 (1.9)	52 (14.8)
Queer	11 (2.4)	3 (2.8)	8 (2.3)
Pansexual	10 (2.2)	0 (0.0)	10 (2.8)
Asexual	1 (0.2)	0 (0.0)	1 (0.3)
Heterosexual	1 (0.2)	0 (0.0)	1 (0.3)
Other	1 (0.2)	0 (0.0)	1 (0.3)
Education
High School or Less	38 (8.3)	9 (8.5)	29 (8.3)
Some College or Associate Degree	133 (29.1)	29 (27.4)	104 (29.6)
Bachelor’s Degree	160 (35.0)	41 (38.7)	119 (33.9)
Master’s, Professional, or Doctoral Degree	126 (27.6)	27 (25.5)	99 (28.2)
Annual Household Income b
Less than US$10,000	33 (7.2)	9 (8.5)	24 (6.9)
US$10,000 – US$29,999	81 (17.8)	22 (20.8)	59 (16.9)
US$30,000 – US$49,999	93 (20.4)	18 (17.0)	75 (21.4)
US$50,000 – US$69,999	52 (11.4)	11 (10.4)	41 (11.7)
US$70,000 – US$89,999	60 (13.2)	17 (16.0)	43 (12.3)
US$90,000 or Higher	137 (30.0)	29 (27.4)	108 (30.9)
Total	457 (100.0)	106 (100.0)	351 (100.0)

^aResponse options included “woman,” “man,” “transgender woman,” “transgender man,” and “other.” The item was administered as part of eligiblity screening, and only participants identifying as a “man” or “transgender man” were eligible to participate.

^bDue to missing data, n = 456 for total sample and 350 for HIV-negative/status-unknown subsample (denominators adjusted accordingly)

Figure 1 presents U=U awareness and belief among the total sample and stratified by HIV status. Nearly all (95.6%) of the participants reported having heard of U=U prior to the study, but considerably fewer (41.1%) reported believing it. Although the percentage of participants who were aware of U=U did not significantly differ between the two serostatus groups (Fisher’s exact test p = .056), a higher percentage of participants living with HIV vs. HIV-negative/status unknown participants believed U=U, X² (1, N = 457) = 35.34, p < .001. Among participants who perceived some level of sexual transmission risk when a PLWH has an undetectable HIV viral load (i.e., U=U disbelief), 77.7% reported “1–25% (low risk),” 12.3% reported “26–50% (moderate risk),” 5.6% reported “51–75% (high risk),” 3.7% reported “76–99% (very high risk),” and 0.7% reported “100% (they definitely WILL sexually transmit HIV).”

Figure 1. U=U Awareness and Belief.

Awareness was high among participants, irrespective of HIV status. Belief in U=U was significantly higher among participants living with HIV compared with HIV-negative/status-unknown participants.

Viral load prejudice was reported by a minority of participants: 25.0% of participants living with HIV who had an undetectable viral load and 27.6% of HIV-negative/status-unknown participants reported being bothered by a PLWH having a detectable viral load sometime in the past month. (The measure was not administered to participants living with HIV who reported a detectable viral load.) Figure 2 displays HIV-negative/status-unknown participants’ ratings of their feelings towards PLWH who had a detectable viral load, PLWH who had an undetectable viral load, and—for context—PrEP users, PrEP nonusers, condom users, and condom non-users. Most participants (65.5%) rated their feelings toward PLWH who had an undetectable viral load more positively than they rated their feelings toward PLWH who had a detectable viral load. Whereas 15% reported negative feelings towards PLWH who had an undetectable viral load, 51% reported negative feelings towards PLWH who had a detectable viral load. Additional descriptive characteristics of stigma measures are presented in Tables 2 and 3.

Figure 2. Variability in Feelings Towards People Based on Their Viral Load Status and Prevention Practices.

HIV-negative/status unknown participants rated their general feelings toward the different groups on a scale ranging from (1) Extremely Negative to (9) Extremely Positive. Whereas 15% reported negative feelings towards people living with HIV who had undetectable viral loads, 51% reported negative feelings towards people living with HIV who had detectable viral loads.

PLWH = People Living With HIV

Table 2.

Characteristics and Bivariate Correlations of Main Measures

Participants Living with HIVa
Variable		Range of Possible Scores	M (SD)	r
				1	2	3	4	5
1	U=U belief	0 or 1	-	-
2	Perceived societal HIV stigma	0–1	0.52 (0.34)	.08	-
3	Internalized HIV stigma	0–1	0.32 (0.23)	−.13	.69**	-
4	Experienced HIV discrimination	1–4	2.37 (0.66)	−.11	.49**	.50**	-
5	Viral load prejudice	0–1	-	.09	.08	.18	.21*	-
HIV-Negative/Status-Unknown Participantsb
Variable		Range of Possible Scores	M (SD)	r
				1	2	3	4	5	6	7
1	U=U belief	0 or 1	-	-
2	Perceived societal HIV stigma	0–1	0.42 (0.33)	−.10	-
3	Enacted HIV discrimination	1–4	1.45 (0.49)	−.21**	.17**	-
4	Feelings toward people with detectable viral load (positive)	1–9	4.42 (1.96)	.09	−.19**	−.15**	-
5	Feelings toward people with undetectable viral load (positive)	1–9	6.29 (1.89)	.20**	−.19**	−.19**	.32**	-
6	Personal HIV stigma	1–4	0.06 (0.13)	−.18**	.24**	.36**	−.15**	−.33**	-
7	Viral load prejudice	0–1	-	−.10	.14**	.35**	−.36**	−.21**	.33**	-

^an = 106 for Variables 1–4; n = 96 for Variable 5 because measure only administered to participants reporting an undetectable viral load

^bn = 348–351 for all variables (range due to missing data)

^*p < .05

^**p < .01

Table 3.

Effect of U=U Belief on Multiple Dimensions of Stigma

Participants Living with HIVa
Modelb	Stigma Outcome	U=U Belief M (SD)	U=U Disbelief M (SD)	b	SE	β	p
1	Perceived societal HIV stigma	0.54 (0.33)	0.48 (0.34)	.07	.07	.10	.327
2	Internalized HIV stigma	0.29 (0.20)	0.36 (0.27)	−.06	.05	−.12	.241
3	Experienced HIV discrimination	2.32 (0.68)	2.48 (0.62)	−.15	.14	−.11	.265
		U=U Belief n (%)	U=U Disbelief n (%)	aOR	95% CI		p
4	Viral load prejudice	18 (27.7)	6 (19.4)	1.6	0.57 – 4.69		.358
HIV-Negative/Status-Unknown Participantsc
Modela	Stigma Outcome	U=U Belief M (SD)	U=U Disbelief M (SD)	b	SE	β	p
1	Perceived societal HIV stigma	0.38 (0.32)	0.44 (0.33)	−.06	.04	−.09	.086
2	Enacted HIV discrimination	1.30 (0.43)	1.52 (0.50)	−.22	.05	−.21	<.001
3	Positive feelings toward people with detectable viral load	4.68 (1.90)	4.29 (1.98)	.36	.22	.09	.102
4	Positive feelings toward people with undetectable viral load	6.82 (1.81)	6.02 (1.87)	.77	.21	.19	<.001
5	Personal HIV stigma	0.03 (0.07)	0.08 (0.15)	−.05	.02	−.17	.001
		U=U Belief n (%)	U=U Disbelief n (%)	aOR	95% CI		p
6	Viral load prejudice	25 (21.2)	72 (30.9)	0.61	0.36 – 1.02		.061

^an = 106 for Models 1–3; n = 96 for Model 4 because analytic sample was restricted to participants who reported an undetectable viral load

^bAll models adjusted for age, race/ethnicity, sexual orientation, and income

^cn = 348–350 for Models 5–10 (range due to missing data)

Associations Between U=U Belief and Stigma

Table 2 shows means, standard deviations, and bivariate correlations for U=U belief and the multiple dimensions of stigma measured. Among participants living with HIV, U=U belief was not significantly correlated with any of the four dimensions of stigma measured. Among HIV-negative/status-unknown participants, U=U belief was significantly correlated with three of the six stigma measures. Specifically, U=U belief was correlated with less frequently enacted HIV discrimination, more positive feelings toward PLWH who had undetectable viral load, and lower personal HIV stigma.

Table 3 summarizes the results of the adjusted linear and logistic regressions. Consistent with the bivariate correlations, U=U belief was not associated with any dimension of stigma among participants living with HIV. U=U belief was significantly associated with the same three stigma measures among HIV-negative/status-unknown participants as in the bivariate correlation analyses. Specifically, U=U belief was significantly associated with less frequently enacted HIV discrimination, more positive feelings toward people with an undetectable viral load, and lower personal HIV stigma.

Implications of HIV Status

Table 4 presents the effects of HIV status on the two dimensions of stigma measured in both participants living with HIV and HIV-negative/status-unknown participants (i.e., perceived societal HIV stigma and viral load prejudice) including partial, conditional, and interaction effects of HIV status and U=U belief. There was a significant effect of HIV status on perceived societal HIV stigma, with PLWH reporting greater perceived stigma (M [SD] = 0.52 [0.34]) than HIV-negative/status-unknown participant (M [SD] = 0.42 [0.33]). There was no significant effect of HIV status on viral load prejudice. There was no significant interaction effect relative to either of the two forms of stigma, indicating that the association between U=U belief and these stigma outcomes did not significantly vary by HIV status.

Table 4.

Effects of HIV Status and U=U Belief on the Two Dimensions of Stigma Measured Among All Participants (Both Serostatus Groups)

	Perceived Societal HIV Stigmaa,b
	Model 1				Model 2				Model 3
Variable	b	SE	β	p	b	SE	β	p	b	SE	β	p
HIV Status (ref = negative/status unknown)	0.12	0.04	0.15	.002	0.13	0.04	0.16	.001	0.05	0.06	0.07	.377
U=U Belief (ref = disbelief)	-	-	-	-	−0.03	0.03	−0.05	.308	−0.06	0.04	−0.09	.088
HIV Status × U=U Belief	-	-	-	-	-	-	-	-	0.13	0.08	0.14	.092
	Viral Load Prejudicea,b
	Model 1				Model 2				Model 3
Variable	aOR	95% CI		p	aOR	95% CI		p	aOR	95% CI		p
HIV Status (ref = negative/status unknown)	0.98	0.57 – 1.68		.941	0.75	0.47 – 1.17		.205	0.61	0.36 – 1.03		.062
U=U Belief (ref = disbelief)	-	-		-	1.08	0.62 – 1.89		.794	0.60	0.23 – 1.54		.285
HIV Status × U=U Belief	-	-		-	-	-		-	2.66	0.83 – 8.57		.101

^aAll models adjusted for age, race/ethnicity, sexual orientation, and income

^bn = 456 for models predicting perceived societal stigma; n = 446 for models predicting viral load prejudice because analytic sample was restricted to participants living with HIV who reported an undetectable viral load (and HIV-negative/status-unknown participants)

Effects of U=U Messaging (Experimental Manipulation)

Multivariable regression analyses were conducted to examine the association of U=U messaging with all stigma outcomes. In initial analyses examining associations separately within each HIV status group, messaging was not significantly associated with any of the stigma outcomes except one: viral load prejudice. Specifically, among participants living with HIV who had an undetectable viral load, those who received the message lacking any information about transmission risk had lower odds of reporting viral load prejudice (10.3%) than those who received the U=U message (37.5%), aOR (95% CI) = .20 (.05 - .78), p = .021. Moderation analyses confirmed that this association of U=U messaging with viral load prejudice was significant among participants living with HIV, aOR (95% CI) = .20 (.05 - .86), p = .031, but not among HIV-negative/status-unknown participants. All other two-way and three-way moderation analyses were non-significant with one exception: Among HIV-negative/status unknown participants, the negative association between U=U belief and enacted HIV discrimination was stronger among participants who received the message that described viral suppression as lowering transmission risk compared with participants who received the message that described it as eliminating transmission risk (i.e., the U=U message), b = −0.29, SE = .13, β = −.19, p = .027. Collectively, findings of the experiment indicated that brief informational messaging about U=U had minimal effects on stigma or the association between U=U belief and stigma, and the two effects that emerged were unfavorable.

When analyses were repeated restricting the analytic sample to participants who passed the attention/manipulation check (n = 299), the same pattern of results emerged. Of note, the association between messaging and viral load prejudice among participants living with HIV and the corresponding two-way interaction between U=U messaging and HIV status could not be tested due to the distribution of the data in the restricted dataset.

Replication of Results with Subsample of Participants Reporting an Undetectable Viral Load

We repeated all multivariable regression analyses pertaining to participants living with HIV restricting our analytic sample to the subsample who reported their viral load level to be undetectable (n = 96 [90.6%]). The same pattern of significant findings emerged with one exception: There was a significant interaction between U=U belief and HIV status relative to perceived societal HIV stigma, b = 0.18, SE = 0.08, β = 0.20, p = .023. However, probing the interaction revealed no significant simple effects. Thus, results with the subsample of participants living with HIV who reported an undetectable viral load were essentially consistent with those of the full sample of participants living with HIV.

Discussion

Despite mounting awareness about U=U among SMM, disbelief persists. Our results underscore the need to address U=U disbelief in future public health initiatives and to target such initiatives to an audience inclusive of both PLWH and HIV-negative/status-unknown people. The vast majority of our participants were aware of U=U (i.e., had heard of it), but many did not fully believe it. Contrary to our hypotheses, U=U belief was not significantly associated with any dimension of stigma among participants living with HIV. However, U=U belief was associated with multiple dimensions of stigma among HIV-negative/status-unknown participants. Compared with their disbelieving counterparts, HIV-negative/status-unknown men who believed U=U reported enacting HIV discrimination against PLWH less frequently, expressed more positive feelings toward PLWH whose viral load was undetectable, and endorsed less personal HIV stigma.

Analyses examining the relevance of HIV status revealed that, consistent with our hypothesis, participants living with HIV were more likely to believe U=U compared with their HIV-negative/status-unknown counterparts. Participants living with HIV also perceived greater societal HIV stigma. Associations between U=U belief and the two dimensions of stigma measured among both participants living with HIV and HIV-negative/status-unknown participants did not vary by HIV status. Additionally, brief informational messaging about U=U had minimal effects on stigma or the association between U=U belief and stigma, and the two effects that emerged were unfavorable.

The very high level of awareness in our study sample (95.6%) exceeded levels previously reported in other samples of SMM nationally and internationally [11, 12]. For the many people who are already aware of U=U, focused efforts are now needed to bridge the gap between awareness and belief, a gap documented in other studies as well [8–12]. Maintaining U=U awareness-raising efforts to reach those yet to be informed is also important. Although high levels of awareness were reported among our and other study samples of SMM, SMM who use dating apps and social media (the primary recruitment source for our study) and who choose to participate in online survey research are not representative of the full SMM community [55], nor do they represent others outside of the SMM community who could benefit from knowledge of U=U [56]. Several studies have suggested that greater social marginalization is associated with lower awareness of U=U [8, 57] and less confidence in U=U as a prevention strategy [10], rendering broader saturation of the U=U message to be an ethical imperative. Results of the current study corroborating the destigmatizing potential of the U=U message among HIV-negative/status-unknown people reinforce this imperative.

The lack of significant associations between U=U belief and stigma among PLWH was unexpected. Several considerations should be borne in mind when interpreting this finding. First, the sample size was small (n = 106) and much smaller than the HIV-negative/status-unknown sample (n = 351), conferring lower external validity and less statistical power to detect significant effects. Additionally, fewer dimensions of stigma were examined in this group (4 vs. 6). Given previous literature linking U=U to lower stigma among larger samples of PLWH [33, 34] as well as qualitative accounts of U=U fostering a sense of normalcy [26, 36], we caution against interpreting our results as evidence for the absence of a relationship between U=U belief and stigma among PLWH. Other psychosocial and physical correlates of U=U message exposure among PLWH, including mental health, sexual health, and viral suppression [33], highlight the continued importance of U=U messaging for this group.

Accordingly, findings of this study support a status-neutral approach to HIV messaging and programming that incorporates U=U and reaches both PLWH and HIV-negative/status-unknown people. A status-neutral approach uses the same process of HIV engagement for all people, irrespective of HIV status, and conceptualizes prevention and treatment within an integrated framework [58]. A status-neutral approach aims to normalize and destigmatize HIV prevention and treatment by presenting prevention and treatment as parallel processes, emphasizing similarity over difference between PLWH and HIV-negative/status-unknown people.

In light of concerns that U=U messaging and the corresponding emphasis on undetectability may contribute to the othering and devaluation of PLWH who are not virally suppressed [37, 38, 45, 46], we investigated stigma related to having a detectable HIV viral load. About one in four participants in our study reported that they were bothered by someone having a detectable viral load in the past month. Furthermore, over 50% of HIV-negative/status-unknown participants reported feeling negatively toward PLWH who have a detectable viral load compared with only 15% reporting negative feelings toward PLWH who have an undetectable viral load. The pervasiveness of viral load prejudice in our sample is concerning and illuminates the biomedical stratification of PLWH, with value differentially ascribed based on viral load detectability [26, 37, 39, 40]. Our findings highlight the need for careful consideration of how information about U=U is communicated and received and suggests that concurrent promotion of the Viral Load Does Not Equal Value (V≠V) message and associated principles alongside U=U messaging may be advantageous [38, 45, 46]. Authors of a recent systematic review reporting no documented HIV transmissions among PLWH with viral loads below 600 copies/mL and “an incredibly rare occurrence of possible transmissions” among PLWH with viral loads between 600 and 1,000 copies/mL recommended public health messaging about the “almost zero” sexual transmission risk associated with viral loads below 1,000 copies/mL as a means of destigmatization [59]. Although such messaging could broaden the range of PLWH who experience less social devaluation because of their viral load status, it does not eliminate the stigma faced by PLWH whose viral load is detectable above 1,000 copies/mL. Also, it could contribute to further social stratification of PLWH (e.g., undetectable; <1,000 copies/mL detected; ≥1,000 copies/mL detected).

A benefit of the current biomedical era is the flexibility it affords those with access to multiple prevention strategies to pick and choose according to their own personal preference. SMM have indeed reported diverse practices with respect to relying on U=U, PrEP, and/or condoms for prevention [41, 60], consistent with the diversity of prevention practices enacted in earlier years of the HIV epidemic based on available options at the time [61]. However, our study found participants’ feelings toward people varied significantly based on those practices, with high levels of negativity expressed not only toward PLWH who had a detectable viral load but also toward people who did not use condoms. HIV messaging that encourages individual choice and respect for decisional differences might help to address this negativity.

Our exploratory analysis examining whether messaging about HIV transmission risk affected stigma or the association of U=U belief with stigma largely yielded non-significant findings, suggesting that exposing participants to brief informational messaging about U=U did not substantively affect stigma. Manipulation of other information or elements of messaging may be more impactful. The two effects that did emerge, neither of which favored U=U messaging over the two comparison conditions, related to dimensions of stigma assessed with respect to the preceding month (i.e., being bothered by someone having a detectable viral load status and enacting discrimination against PLWH). Therefore, message exposure could not have caused the occurrences of stigma that participants reported, which would have preceded exposure, but message exposure may have primed recall of the stigma occurrences or enhanced participants’ comfort reporting them.

Our findings add to the sparse but growing body of research examining the effects of U=U messaging on U=U belief and HIV stigma [48–51]. Further research could enhance future messaging and educational interventions related to U=U. Previous work establishing a community-engaged method of U=U message development [62], identifying sociodemographic and behavioral factors associated with U=U belief [63], and evaluating other forms of HIV-related messaging (e.g., PrEP) [64–66] may also be instructive in the process of message development. Messaging strategies will need to account for the wide variability in recipients’ baseline HIV knowledge and beliefs. For example, whereas most participants in our study who did not believe U=U nonetheless believed that the risk of HIV sexual transmission when a PLWH has an undetectable viral load was low, a minority of participants believed transmission risk was very high or even 100%.

We also recommend future investigation of the mechanisms underlying identified associations between U=U belief and HIV stigma. Stigma theories commonly posit self-protection to be a driver of stigma related to HIV and other infectious diseases [28, 29, 31]. According to the disease-avoidance model of stigma, association of a person with a disease label (e.g., HIV) triggers access to semantic knowledge about that disease that can evoke self-protective emotional and behavioral responses (e.g., fear, disgust, and avoidance) that stigmatize the person with the illness [29]. By extension, U=U may avert such stigmatizing responses by correcting outdated misconceptions about HIV contagion and associated appraisals of threat. Future research could explore this and other psychological pathways via which U=U belief may operate, including the role that other marginalized characteristics (e.g., sexual minority identity, use of injection drugs) and associated stigma may play.

Limitations

Our study has limitations. The size of our sample (n = 457), particularly our PLWH subsample (n = 106), was relatively small, limiting generalizability and increasing the likelihood of statistical error.

All measures were self-reported and therefore subject to response biases. For instance, it is possible that participants underreported HIV stigma due to social desirability bias. In assessing viral load stigma, we used only single-item measures to capture viral load prejudice and feelings toward PLWH with a detectable viral load. We recommend developing and validating more comprehensive scales to assess these constructs and other dimensions of stigma related to viral load detectability in the future.

Likewise, our single-item measure of U=U awareness reflected only whether participants had heard of U=U and did not capture their level of engagement with the U=U message. The Elaboration Likelihood Model posits that message recipients who engage with a message via a central processing route that entails deeper, more effortful contemplation of message content can be persuaded by logic and facts, which can result in enduring attitudinal change. By contrast, those who respond via a peripheral processing route, allocating less time and effort to message consideration, may rely on superficial cues, and any resulting attitudinal change is likely to be short-lived [67, 68]. Thus, although nearly all participants in our study reported exposure to the U=U message, such exposure might have had a more pronounced and lasting impact on U=U belief among those who engaged with the message more deeply (e.g., by discussing it with a provider or serodifferent partner) than among those whose engagement was relatively limited (e.g., a quick glance at a billboard or bus stop advertisement), which could help to explain the observed gap between U=U awareness and belief. Future research could manipulate or assess level of engagement/processing as a moderator of the effect of U=U awareness on U=U belief.

We developed and administered our survey prior to the World Health Organization’s 2023 specification of three discrete categories of viral load status—unsuppressed (detectable virus greater than 1,000 copies/mL), suppressed (detectable virus less than or equal to 1,000 copies/mL), and undetectable (too few copies/mL of virus to be detectable) [69], and survey wording does not fully reflect this categorization system. In our survey, viral load suppression and undetectability were treated synonymously in the stimulus message quoted from a 2017 letter issued by the CDC (“viral suppression, defined as less than 200 copies/ml or undetectable levels”); the U=U belief item (“virally suppressed [has an undetectable viral load]”); the item assessing feelings toward PLWH who have a detectable viral load (“NOT virally suppressed [have detectable levels of virus]”; and the item assessing feelings toward PLWH who have an undetectable viral load (“virally suppressed [have undetectable levels of virus]”). Additionally, participant viral load and viral load prejudice items referred to (un)detectability without specification of viral suppression. Further research is warranted to explore nuances in transmission risk beliefs and stigma associated with unsuppressed and suppressed subcategories of viral load detectability. For example, if messaging is modified to reflect this three-category system, characterizing unsuppressed as “at risk”, suppressed as “almost zero risk,” and undetectable as “zero risk” of transmission [69], it would be worthwhile to explore whether PLWH who are suppressed experience lower stigma and greater acceptance comparable to those who are undetectable, or whether the enduring association with possible transmission risk limits access to such privilege.

The cross-sectional design of our study limits our ability to infer causality. Although we conceptualized U=U belief as predictive of HIV stigma throughout the study, the association between these two variables may be bidirectional. That is, not only is it the case that U=U belief could reduce HIV stigma, but someone with low HIV stigma may be more receptive to the U=U message and inclined to believe it. Indeed, recent work has illuminated HIV stigma as a barrier to accepting and relying upon U=U to prevent transmission for both PLWH and HIV-negative people [70, 71]. By this logic, efforts to spread the U=U message may be complemented by initiatives that enhance openness to such messaging by targeting HIV stigma reduction directly. Employing a longitudinal design in future research related to U=U belief and stigma would enable better understanding of the directional nature of the association.

Conclusions

Our study results suggest that U=U belief may reduce HIV stigma among HIV-negative/status-unknown people and, when considered in combination with past research indicating the favorable effects of U=U for PLWH [33, 34], demonstrate the importance of U=U knowledge reaching both PLWH and HIV-negative/status-unknown people to maximize its destigmatizing impact. In addition to public health educational initiatives aimed at the general population, focused training on these topics for healthcare providers that enhances their understanding and encourages accurate communication about U=U with all patients could help to propagate the message [72, 73]. Broadly disseminating U=U knowledge and instilling U=U belief are critical steps for engaging both PLWH and HIV-negative/status-unknown people in U=U promotional efforts as part of the larger goal of ending HIV stigma, a burden that has primarily been shouldered by PLWH [74]. Effectively addressing HIV stigma also necessitates concomitant consideration of its disproportionate impact on PLWH who have a detectable viral load and face intersectional forms of stigma based on other marginalized social statuses.

Availability of Data and Material

Data are available from the lead author (SKC) upon request.