In the summer of 2011, The Fenway Institute of Boston and the University of Pittsburgh Center for Lesbian, Gay, Bisexual and Transgender (LGBT) Health Research convened a meeting of many of the thought leaders in the field of Gay and Bisexual men’s health to discuss how resilience among gay and bisexual men, and other men who have sex with men (henceforth referred to simply as gay and bisexual men), could inform HIV prevention interventions. The idea behind the meeting was to outline ways in which resiliencies could be used to move the field forward towards the abatement or elimination of health disparities within this population. The primary goals of this meeting were to raise interest in the study of resilience as a way to improve health outcomes among gay and bisexual men, to learn from community-based programs that were already using resilience-based approaches in their work, and to begin conceptualizing how researchers could adapt and use a resilience-based framework in their study of HIV and other health disparities in gay and bisexual men. We also critically reviewed deficit-based approach to health, focusing on assumptions underlying these methods.

Relative to the field of prevention research, community-based organizations for gay and bisexual men have better recognized that focusing on strengths, resilience, and other protective factors can bring about positive individual and community-level results. However, few community programs have had the resources to test or evaluate their work. To begin to bridge this divide, the meeting organizers brought together leaders in research as well as key informants from resilience-based community programs including Manifest Love, the Harlem United Community AIDS Center (www.harlemunited.org), Easton Mountain (www.eastonmountain.com), Life Lube (www.lifelube.org), and others who have already been using strength-based approaches to promote health among gay and bisexual men. After a day of sharing methods, challenges, questions and results from research studies and community organizations, the meeting participants worked together to identify a set of resiliencies that are common among many gay and bisexual men. These resiliencies were then used to create an initial theoretical model for how resilience can inform prevention/promotion interventions. By basing health promotion efforts on theoretical models that incorporate resilience-based approaches that enhance healthful behaviors, programs could be designed that more effectively lower health risks than those based on deficit models alone, an advance that could lower multiple health disparities among gay and bisexual men.

The Roots of Resilience Theory

The incorporation of resilience theory into research investigating health disparities among marginalized populations is not new. Over the past 40 years, developmental psychologists have strived to conceptualize, define and measure resilience to better understand health behaviors and outcomes among many at-risk populations [1–3]. Much of this prior work could be adapted be applied to gay and bisexual men. It is therefore unnecessary to start from scratch when thinking about how to incorporate resilience into theories of gay and bisexual men’s health.

Though many definitions of resilience have been proposed, that offered by Fergus, et al. [4], is a particularly useful and comprehensive definition: “the process of overcoming the negative effects of risk exposure, coping successfully with traumatic experiences, and avoiding the negative trajectories associated with risk”. Most useful definitions of resilience, including the aforementioned, have two core components [2, 4–6] (1) Positive adaption in the face of adversity and risk. The experience of adversity is a crucial component. Resilience theory does not attempt to learn from those with no risk exposure as that would not be generalizable to those most in need of interventions. Without adversity to overcome, there is no resilience. (2) Resilience is a process. Though some have conceptualized resilience as a trait, it appears to better represent reality when conceived of as a process that develops and changes over time. This view maintains that persons are not born with a resilience attribute or set of resilience attributes, nor do they, develop a single skill that allows them to be resilient to every adverse situation or context. Rather, it proposes that individuals can become resilient over time, to specific risks, developing the necessary protective factors as they are needed.

Evidence for Resilience among Gay and Bisexual Men

A rich literature on the psychosocial, mental and physical health of gay and bisexual men has expanded over the past 30 years, initially instigated by the burgeoning AIDS epidemic, and continues to grow. This increase in reports demonstrates a clear acknowledgement that gay and bisexual men, as a group, are an important part of any public health approach to ending the domestic HIV epidemic. However, much of this body of evidence has used the traditional epidemiological paradigm in which the goal is to identify risk factors that lead to deficits, disparities and disease and work toward removing these risks. Thus, the field of public health has traditionally been better at identifying associations with disease than in defining protective factors for health.

Nonetheless, many of these deficit-based investigations have identified implicit evidence for resilience among gay and bisexual men. Table 1 [7] shows results from the original investigation of syndemics (co-occurring psychosocial health problems, such as substance abuse, violence victimization, depression and childhood sexual abuse) among gay and bisexual men. This study found that the health conditions investigated (polydrug use, depression, childhood sexual abuse, and partner violence) were highly inter-correlated, and that as the number of conditions increased within the individual, so did odds of recent engagement in high risk sex, and with the exception of those with 3 or 4 conditions, HIV prevalence also increased. These findings may be somewhat intuitive. Individuals who are dealing with co-occurring health problems may not have the resources or fortitude to focus on their own personal HIV prevention strategies. However, somewhat hidden in this table is evidence that great resiliencies exist among gay and bisexual men. About 75 % of men who were dealing with one or more health issues, were still able to avoid high risk sexual behaviors and to remain HIV uninfected. This suggests that men who are struggling with multiple psychosocial health problems have resources that help them to remain sexually safe and to remain HIV negative.

Table 1 Example of results from a defect based analysis [7]
Full size table

There is nothing particularly unique about this example. This sort of phenomenon can be seen in most accounts of gay men’s health. Even in reports that have shown very high levels of negative health outcomes and disparities among gay and bisexual men, these outcomes have been among a minority of men. Evidence for strengths and resilience among gay and bisexual men is widespread in both scientific literature and historical accounts of gay culture. For instance, in an investigation of tobacco use among gay and bisexual men, Greenwood et al. [8] found that a greater proportion of gay men reported cessation of tobacco use than reported current daily tobacco use (26.9 vs. 25.7 %), indicating a widespread trend towards health promotion and recovery from an addiction that is very difficult to resolve. Gay and bisexual men have been shown to overcome internalized homophobia over the life-course despite existing in a context that promotes this form of self-loathing [9]. Most gay and bisexual men have also managed to avoid problematic drug use despite widespread use of recreational drugs generally perceived to be addictive [10, 11]. Most notable, however, is that over the past 40 years, gay men along with other members of the LGBT community have made an impressive and largely effective bid for multiple civil rights (e.g., marriage, adoption, equal access to health care, etc.) all while facing community-wide devastation from the HIV epidemic.

A Critique of Relying Only on a Deficit Approach to Public Health Research Among Gay and Bisexual Men

Despite the fact that the evidence for resiliencies among gay and bisexual men quickly becomes apparent once one starts to look for them, very few resiliencies have been included a priori in underlying theories upon which intervention research efforts are based. Much of the information that has been gleaned about how strengths and protective factors work to promote health among gay and bisexual men has been detected as an artifact of deficit-based investigation. Few studies have intentionally examined how sexual and gender minority communities use strengths to promote health. For example, does the process of coming out and the skills built as that process unfolds contribute to subsequent successful health behaviors? Does the process of homophobia management (the social understanding of where and where it is safe to be open about one’s sexuality) promote safety in otherwise unsafe situations? Does the ability to form affirming social structures like families and religious institutions help us to secure the resources that support health across the life course? Does the community’s history of activism and community building, which is passed down from generation to generation outside of the context of the traditional family structure, help protect communities from poor emotional and mental health outcomes?

These questions and many others have yet to be adequately addressed in public health research. Rather, health research focusing on gay and bisexual men has centered on a deficit-based approach, focusing on HIV prevention and not on life-affirming sexuality, to try to improve the health of gay and bisexual men by decreasing risk behaviors. This approach follows the traditional epidemiological process of first identifying risk factors for a negative health outcome and then attempting to diminish said outcome by eliminating those risk factors. While this approach has led to important advances in public health for gay and bisexual men, it fails to fully explain patterns of health and illness among gay and bisexual men.

Focusing on a deficit-based approach to gay and bisexual men health makes intuitive sense. The epidemiological evidence is very clear; there IS a health crisis among gay and bisexual men, associated with unprecedented levels of HIV and other STDs. A focus on what has gone wrong to create the crisis makes sense. But can a focus on what has gone wrong yield the most potent interventions? What would interventions look like if they focused learning from men who were doing well or men whose risk practices resolved? There is a need to understand men who may be exposed to risky environments, but are resilient to that exposure.

An overwhelming focus on deficits among gay and bisexual men predisposes prevention efforts to ignore the strong body of evidence for resilience that exists in this population [8, 12–14]; evidence that may be useful in designing interventions to address health disparities. This suggests that learning how strengths evolve could improve prevention efforts by capitalizing on the skills and resources that already exist among gay and bisexual men.

Limitations of Deficit-Based Analyses

Aside from the fact that deficit-based analyses of gay men’s health cannot explain all of the behavioral patterns observed in this population (i.e., the evidence for resiliencies among gay and bisexual men), several assumptions of deficit-based approaches are likely to prove problematic in designing health promotion programs among gay men. These include:

  1. 1)

    The trajectories that generate risk may be governed by different variables than the trajectories that produce safety. If it is the case that different variables govern resiliency than vulnerability, analyses that “flip” (that is, focus on the obverse of) the variables from risk factor analyses may not produce useful insights into intervention design. For example, if we find low levels of social support to be a risk factor for engaging in risky sex, then the assumption would be that high levels of social support could be a protective factor. However, only a structured intervention would likely demonstrate whether enhanced social support among high-risk persons in would enhance their ability to engage in safer sex behaviors. If research programs can intentionally look for potential pathways to resilience, they may find crucial variables that cannot be derived from exclusively deficit-based models. One implication of the idea that understanding resilience may necessitate a look at entirely different variables not derived from deficit based theories is that findings from studies that focus on vulnerability may not be generalizable to men who rely on resiliencies to stay safe over time. If the assumptions that underlie prevention programs are not generalizable, then prevention programs may be initiated that lack credibility and/or seem culturally inappropriate.

  2. 2)

    In addition to producing interventions that may not be meaningful for all gay and bisexual men, deficit-based interventions can produce uninviting interventions. Interventions that implicitly inform men that they don’t understand how dangerous HIV is, that they have few condom negotiation skills, that they have too many sexual partners, that they have poor community building skills, and/or that they are not making wise life choices may not be likely to draw large audiences. On the other hand, interventions that seek to transfer strengths that other men in the community are using to stay safe and healthy, even in risky environments, may prove more useful and inviting.

  3. 3)

    Interventions designed to address deficits may have shorter half-lives for efficacy than interventions that are designed to build resiliencies. That is, a focus on deficits alone does little to help men cope with ongoing exposure to risky environments [15, 16]. A focus on building resiliencies, on the other hand, help men gain skills that are useful in avoiding risks that are likely to be sustainable over longer periods of time. It may be that that one of the reasons that interventions designed to help gay and bisexual men avoid risks for dangerous health outcomes like HIV do not have lifelong effects is that deficit -based interventions focus on initiation of behavior change, but may not be able to help men access strengths that they need to maintain that behavior change over the long haul. Interventions that build resiliencies may help men access naturally occurring social and other types of reinforcements that occur in day-to-day life, in contrast to deficit-based interventions that stress avoidance of negative consequences.

  4. 4)

    Deficit-based approaches have been primarily based on consideration of individual factors and have not seamlessly led to appreciation of community-level strengths. A focus on resilience in the community might guide community level interventions. To the degree that resilience, although manifest in individuals, depends on social and community factors to develop and maintain it, interventions directly modifying social and community (rather than individual) variables may foster resilience. For example, it may be that gay social organizations, such as car clubs or sporting leagues, help individuals develop a balanced view of themselves and the gay community that promotes resilience in ways that are not found in those whose experiences have been relatively exclusively focused on the sexual aspects of gay culture. If so, promoting such organizations or broadening the scope of sexually-oriented venues (e.g., bars, on-line sites) might have impact.

  5. 5)

    Finally, deficit-based approaches can help us diagnosis what is wrong, but not how to fix it. Gay and bisexual men currently exist in a social context where there is great adversity, particularly in terms of social marginalization. This adversity potentiates risk; a pattern that is in not unique to gay and bisexual men. People who feel they are pariahs may be disinclined to focus on changing their behaviors. But, if ways can be found to provide gay and bisexual men with assets, skills, and resources to cope with adversity and to avoid risk, population-level reductions in health disparities might be manifested.

In a meta-analysis of the efficacy of HIV prevention interventions focused on gay and bisexual men, Herbst et al. [17] found that these interventions resulted in a 23 % reduction in the odds of engaging in unprotected anal intercourse and a 61 % increase in condom use during anal sex. This suggests that current prevention paradigms are effectively addressing some degree of risk. As important as these contributions to the health of gay and bisexual men have been, the question can still be raised as to how this pioneering work can be improved to make these interventions even more efficacious. Resilience-based approaches to health disparities research among gay and bisexual men may well contribute to improving intervention effect sizes and improving health even further by identifying new variables and mechanisms for health promotion that can be incorporated into interventions that already have been shown to be efficacious. Moreover, resilience theory is holistic, so focused interventions could also decrease high rates of STDs, substance use, and affective disorders among gay and bisexual men, which are highly prevalent and contribute to HIV risk taking.

Towards a Theory of Resilience Among Gay and Bisexual Men

One of the main questions addressed at the meeting was how to integrate resilience theories with the current body of research on gay and bisexual men’s health, with a particular focus on HIV prevention. Because so little thought had been previously given to defining resiliencies that exist among gay and bisexual men, the participants of the meeting proposed a set of variables that might be important to explain patterns of health among gay and bisexual men. These variables were qualitatively collapsed to create a list of overarching categories at the individual, dyadic, family, community levels. This initial sorting was then reviewed by the organizers of the meeting. The review included an assessment of the categorization of the variables, as well as the addition of any domains deemed theoretically important, with a particular focus on identifying modifiable factors. Table 2 presents a summary of the list of potential variables proposed during the meeting that could influence health outcomes among gay and bisexual men that are not typically measured. Incorporating even a small number of these variables into data collection efforts may well improve our ability to explain variance in health risk profiles and health outcomes among gay and bisexual men.

Table 2 Theoretically driven measures of resiliencies in gay and bisexual men
Full size table

Conceptualizing and Testing Causal Pathways for Hypothesized Resiliencies

The hypothesized resiliency variables in Table 2 vary in many dimensions, including level of influence (individual v. dyadic, for example) and degree to which they reflect internal psychological processes or externally verifiable behaviors. Therefore, it seems likely that they will also vary in how they would fit into any conceptual model attempting to present causal pathways for resiliencies. As a corollary, the methods to test theoretical statements about these variables would vary according to their role in a conceptual model.

Some variables would conform to a compensatory model, or a main effects model. In this model, protective factors are viewed as positively associated with or predictive of positive health outcomes. These factors exist and operate directly on outcomes. A clear example would be that a low level of HIV prevalence, such as was seen in many locales early in the epidemic, is associated with low levels of transmission regardless of any other factor. More to the point of this discussion, it is well documented that experiences of adversity are associated with increases in co-occurring psychosocial health problems (syndemics) [18], and syndemics are in turn associated with HIV risk [19], HIV prevalence [7], and HIV incidence [20]. As demonstrated in Fig. 1 model 1, protective factors may be directly associated with reduced odds of both syndemics and HIV infection.

Fig. 1
figure 1

Conceptualizing pathways in which protective factors promote health and resilience

Full size image

Other variables would operate within a protective model, or interaction model (Fig. 1, model 2). The protective model is a basic moderation model where protective factors are looked at as having a buffering effect on the relationship between the adversity (IV) and the outcomes (DV). For example, in individuals with a high level of a protective factor, such as social support, risk factors won’t necessarily translate into negative health outcomes, whereas for individuals without that protective factor, or low levels of that protective factor, we will see the relationship between increasing levels of risk and increasing levels of negative health outcomes.

The final model is the challenge model, or the quadratic model. In this model there is a curvilinear relationship between risk and negative health outcomes. In individuals with high or low levels of risk we will see the positive relationship between risk and outcomes, but in individuals with moderate levels of risk, that relationship dissolves. The logic in this model is that individuals who experience moderate levels of risk have been exposed to just enough risk so that they have learned how to cope with or avoid the associated outcome, but not so much so that they can no longer cope. There is little evidence of protective factors among gay and bisexual men that fall into the challenge model; however, this may be because it is the most difficult to model and does not follow as obvious of patterns as the other two. However, it could be that this model is particularly salient for gay and bisexual men and other marginalized populations as there is a constant, institutionalized level of adversity that accompanies.

Variables listed in Table 2 might operate in conformance with the predictions of one, both, or all three of these models, depending on time, stage of individual development, stage epidemic development, or other factors yet to be understood. Further, where models reveal moderating variables, many questions about mediating variables will arise in order to allow for more full explanations of how resiliency operates. Elaboration of complex models to explain resilience remains an area ripe for additional research investigations.

Incorporating a Resilience Framework into Research Among Gay and Bisexual Men: We Can Do Better

The traditional “deficits-based” approach to studying health among gay and bisexual men has led to substantial results in reducing behavioral risk reductions for HIV transmission. There is less than half of the number of new HIV infections annually among gay and bisexual men than at the height of the epidemic in the pre-HAART era. Nonetheless, over the past decade, more than 50,000 US residents became newly infected each year with an increasing proportion of new infections occurring among men who have sex with men [21]. In fact, in 2009, 61 % of new HIV infections were among men or transgender women who have sex with men [22]. Clearly, the risk for HIV remains a significant threat to the health of this population, particularly among black gay and bisexual men. Despite these horrifying statistics, the majority of gay and bisexual men remain HIV negative. Deficit-based analyses cannot explain why so many gay and bisexual men remain negative despite the fact that they are subject to adversity and marginalization. This suggests that we need to refine our theoretical understandings of patterns of health and illness among gay and bisexual men that take into account the many strengths found in this population. This effort is of more than scientific interest, however, in that improved theoretical understandings of gay men’s health will provide a guide to designing more effective health promotion efforts than those that are now currently available. Resilience-based investigations may provide the information needed to increase the efficacy and effectiveness of prevention efforts that ultimately will enable us to lower health disparities among gay and bisexual men.

Several essential steps are needed to successfully incorporate resilience-based inquiry into gay and bisexual men’s health promotion. The first step in this process is to develop a testable theory that incorporates both the effects of resiliencies and vulnerabilities in explaining patterns of health and illness among gay and bisexual men. This will necessitate conceptualizing and measuring resiliencies in ways that move beyond simply flipping (that is, focusing on the obverse of) deficit variables. The attempt to build a theory that takes both resiliencies and vulnerabilities into account might best start with qualitative methods and using iterative approaches to identify new variables, some of which may turn out to be those variables proposed in Table 2. One way that this qualitative work might begin would be to conduct a set of life history interviews among men who have exhibited life trajectories that cannot be explained through deficit-based theories (i.e., men who have few, if any, significant health problems and who are thriving in terms of social connection and in terms of health despite exposure to adversity). Such work would be analogous to basic biomedical studies that sought to identify how long-term non-progressors, or “elite controllers,” keep HIV infection under control. From a social and behavioral perspective, interviews focused on determining naturally occurring resiliencies might help identify new theoretical pathways that explain the production of health better than do theories that focus on vulnerabilities alone. New variables, in addition to those elements we already know to be important in predicting health outcomes among gay and bisexual men, will captures new variance that will serve as a framework for making interventions more efficacious.

We note that the variables identified in resilience-based investigations that will be most useful in intervention design are those that transcend individual strengths. That is, many studies of resiliency among non-gay and bisexual men populations have found personality traits or qualities of one’s family of origin to be important in predicting positive health outcomes. Although these same conclusions may also be true of gay and bisexual men, they are very hard to change on the individual level, and are therefore not amenable to incorporation into intervention design. However, strategies that men may have developed that confer health benefits (e.g., strategies for family building, better connection to community, health monitoring, etc.) may be translatable to other men as part of intervention design. Once we are in a position to propose a resiliency-based model to explain both health and illness among gay and bisexual men, we might wish to test that model using quantitative approaches. Ultimately, such tests will need to use longitudinal methods since cross-sectional approaches will be of limited utility in capturing the underlying processes that help an individual develop resiliencies.

The real proof of the theory will occur when we move toward rigorous testing of resilience based theories in intervention research. We anticipate that these tests will need to use RCT designs to determine if these approaches add to the efficacy of the current set of interventions now in the field. However evaluation of resiliency-based approaches should also test effectiveness given that interventions that are based on strengths may not only account for additional variance in health outcomes but may very well increase participation and reach to individuals most at risk.

We note that as HIV prevention efforts shift towards combination prevention where behavioral efforts are combined with biomedical approaches to HIV prevention, understanding resiliencies will also be an essential tool to translating efficacious biomedical strategies into real world effectiveness [23–25]. The impressive strides that have been made recently in the effort to prevent HIV through pre-exposure prophylaxis (PrEP) [26, 27], post-exposure prophylaxis (PEP) [28, 29], and treatment as prevention [30, 31] will be useless if there is not uptake by those most at risk. Real world effectiveness of biomedical prevention methods will be greatly served if we are able to understand and promote resilience factors that contribute to uptake and adherence.

Creating interventions designed to promote resilience will in some cases be exceedingly difficult as some of the factors associated with resilience stem from growing up in a supportive environment. Creating affirming environments for all young gay and bisexual men would entail major social and structural interventions. However, the current environment is creating test conditions to compare gay and bisexual men growing up in more and less supportive environments. Hatzebuehler [32] found that Oregon youth who grew up in counties that had an LGBT supportive environment were less likely to report youth suicide than those in less tolerant counties. Additionally, he found that sexual minority individuals in states without anti-marriage equality amendments fared significantly better in mental health outcomes than individuals in environments where these homophobic policies on the ballot [33]. Given the rapid changes in social acceptance of homosexuality and same sex families, social and behavioral scientists should carefully evaluate the intended and unintended consequences of marriage equality and increasing civil rights in some jurisdictions.

This paper has laid out an ambitious agenda, but one that is essential if we are to address the multiple health disparities that exist among gay and bisexual men and other sexual minorities. While initial gains in understanding how to promote health in this population have been important, they have, in general not had sufficient potency to serve as effective public health tools. We believe that the reason is not that the programs have been poorly fielded, that organizations that fielded these initial interventions did not understand the community or that the community did not see the value of these efforts. Rather, we believe that an important part of the reason that these interventions did not result in impressive potency is that the theory underlying interventions was limited by an overemphasis on the vulnerabilities that drive health disparities among gay and bisexual men and by taking only limited advantage of the many resiliencies that function to promote health among gay men. We hope that as progress is made in increasing the potency of interventions that are available to reduce HIV risk among gay and bisexual men, that there will be a move to address other health disparities that exist in this population. We expect that this work will draw heavily from the work that was done among gay and bisexual men in HIV prevention and hope that a greater focus on resiliencies will become an important tool in resolving health disparities in this and other sexual minority populations.