The roads less traveled: Mapping some pathways on the global mental health research roadmap

Key Issue 1: Developing a language of global mental health

A literature search failed to reveal a formal definition of the term global mental health, indicating that there is an assumption that everyone engaging in the discourse must already know what is meant by the term. Currently there is no discussion delineating the field, or differentiating it from similar fields such as cultural psychiatry or cross-cultural psychology. Language shapes discourse and perception, and even a broad, tentative definition may help to establish the perimeter of the field. Reviewing published papers which used the phrase global mental health, one gathers that the term often refers to the overall mental health of an individual; for example, the Global Assessment of Functioning Scale in the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR; American Psychiatric Association [APA], 2000) provides a rating of overall functioning from 1 to 100. However, this is not how the term is used in GMH research, where it refers to the sum or aggregate of mental disorders in all countries. This usage, which relates to geographic scope, is the framework for GMH research. For example, the Lancet series of 2007 provides an impressive array of statistics and information, on mental disorders in different countries so that the reader obtains an overall profile of psychiatric illnesses across the globe.

There is, however, a third aspect of the term that must be confronted and acknowledged. The GMH discourse around mental health is typically produced by psychiatrists in countries high in economic resources, about people in countries low in resources. This definition relates less to scope and more to issues of power and perspective. It does not limit the GMH movement to low-income countries, but acknowledges that, in an attempt to increase equity of care and respite from suffering from mental disorders, more attention is paid to scaling up the availability of services in countries with fewer resources. ²

Currently, therefore, the term global mental health is used by psychiatrically oriented researchers in high-income countries to denote the mental ill health of people in countries around the world, with a particular focus on people living in low-income countries. This use takes for granted the application of notions of mental ill health across cultures. An alternative definition of global mental health, that acknowledges context, could be “the mental health and mental ill health of people across the globe, experienced and expressed in culturally distinct ways.” The term “mental” here refers to both intrapsychic and psychosocial processes. This definition explicitly acknowledges culture as a key component of mental health, which could help remind researchers to address culture explicitly in GMH research.

A second issue relating to language and definitions lies with reference to cultural groups. The field has yet to agree upon a clear terminology to refer to different cultural groups. The old practice of referring to cultures as “eastern” and “western” is clearly inadequate. The newer practice is to use income as the defining factor, thus delineating countries as “high-income” (HIC), “middle-income” (MIC), or “low-income” (LIC). However, these terms do not provide an adequate referent to culture, and in fact mask the reality that not all HICs participate in the discourse on mental health. The term HIC can refer to countries as disparate as Kuwait, Japan, and the USA, a grouping that makes little sense in the context of culture.

A new terminology is proposed and henceforth used in this paper, one that is based on the method used by economists to group together countries similar in economic power, such as “group of eight” (G8) and “group of 20” (G20; see Appendix). This new terminology groups together countries based on broadly similar cultural values and acknowledges the current power structure in funding and discourse in mental health. The 18 countries that tend to set the GMH agenda and have more funding and resources for their own programs and for programs in LICs are designated “HPI18,” to denote both the power level (high power) and the psychological perspective (individualism) of these cultures (Oyserman, Coon, & Kemmelmeier, 2002). All other countries are broadly designated “LPI” (low-power individualistic) or “LPCO” (low-power collectivistic), and other terms are formulated to designate each cultural group (see Appendix). This new terminology is currently being pilot-tested among mental health professionals from different countries and fields. It explicitly denotes the broad cultural orientation of two distinct groups across the globe.

A third issue relating to language concerns how it is used to shape the GMH research agenda. As noted above, the current GMH roadmap appears to be directed by an individualistic psychiatric approach. Even the term “mental health,” in the way that it is currently used, reflects a bias towards intrapsychic, individual, dysfunctional processes. In the Lancet 2007 series, all data and statistics that outline the scope of the problem focus on the individual, with such indices as disability adjusted life years (DALYs), mortality, and morbidity. These are valuable data that provide justification for paying attention to global mental health; the need for these data is not in question. What is also apparent, however, is that there is no corresponding index for groups, such as “Disability-adjusted family/Community year.” The justification for such exclusion is the paucity of systematic data, but this only serves to perpetuate the practice of focusing solely on the individual. Most researchers in GMH would agree that community distress is greater than the sum of the distress of its individual members, particularly in collectivistic cultures. For example, Somasundaram (2007, 2010) uses the term “collective trauma” to describe how the destruction of homes and migration resulting from the armed conflict have eviscerated the Tamil communities in the Jaffna peninsula. GMH researchers must begin to conceive of ways in which such collective distress can be measured and indexed. For example, an index of “Disability-adjusted family/Community year” might focus on family or community well-being lost due to physical and psychological disability. Community stress and loss could be evaluated by comparing indices of a thriving community (via such indicators as expected economic stability, population growth, suicide rates, and substance use/abuse rates) against the indices of a distressed community.

The dominant psychiatric perspective is also apparent in the types of data and indices used to summarize resources for mental health across the globe. In one such paper (Jacob et al., 2007), the authors include hospital beds, psychiatrists, and psychiatric nurses. Composite data on the availability of psychologists and other mental health professionals are absent, apart from two case studies provided as insets in the text. Interestingly, when Saraceno et al. (2007) conducted a qualitative survey on the topic, respondents readily identified psychologists and other mental health professionals as resources. The Mental Health Atlas of the WHO also includes data on psychologists, social workers, and community mental health services around the world (WHO, 2005b), but these data are not included in its summary, and the atlas data are not easily digested in their current format. Papers on GMH interventions do acknowledge the contribution of personnel other than psychiatrists (e.g., Patel, Araya, et al., 2007). The GMH movement should consistently ensure the inclusion of all mental health professionals when mapping its agenda, or consider using a more limited term like “global psychiatry” to refer to the movement.

The current GMH research agenda could also be critiqued for its exclusive focus on the language of distress such as DALYs and morbidity while ignoring indices of resilience, recovery, and hardiness. Patel and Sartorius (2008) explain that the focus on mental illness over health was a practical necessity, as the editors of the Lancet series wished to include only data that were obtained through scientifically rigorous methods. However as with the lack of community indices, such exclusion only serves to perpetuate the cycle of neglecting more positive aspects of mental health. As the WHO constitution (1946) asserts, mental health includes not only absence of disease, but the presence of well-being. The WHO (2004) summary document on promoting positive mental health provides some excellent examples of steps and studies undertaken to prevent mental distress, but while it describes the concept of resilience, no studies explicitly examining the topic are included in this document. A GMH research agenda should include research on such variables as resilience and recovery, so that interventions promoting these positive processes can be developed and evaluated. Scattered publications indicate that some steps towards such an approach are being taken (e.g., Garber et al., 2009; Ramon, Shera, Healy, Lachman, & Noel, 2009), and the attempt to assess the entire continuum of mental health from distress to eustress via such instruments as the Mental Health Continuum (Lamers, Westerhof, Bohlmeijer, ten Klooster, & Keyes, 2011) is another promising step in this direction. However, the GMH agenda needs more systematic and programmatic efforts in this direction.

Key Issue 2: The etic–emic balance

The balance between etic and emic perspectives is a crucial issue for researchers and practitioners in the GMH movement, yet one that is largely missing in current discussions. In cross-cultural psychology, the term etic refers to an approach where behaviors are studied from outside a culture with the goal of finding broad (culture-general) patterns across cultures; emic refers to an approach where behaviors are studied from within a culture to understand the unique (culture-specific) aspects of that phenomenon (Pike, as cited in Berry, 1989; see for a more detailed explanation). A derived etic occurs when behaviors in two or more cultures can be shown to be functionally equivalent, and when “features … exist, not only within one culture, but also exist outside it (in the second culture)” (Berry, 1989, p. 727). Problems arise when a culture-specific syndrome in one culture is considered a universal or culture-general syndrome, and measures and interventions originally intended for that culture are then imported and imposed on other cultures, referred to as imposed etic (Berry, 1989) or pseudoetic (Triandis & Marín, 1983). When discourses around culture are sparse, there is a danger that such pseudoetic approaches will dominate and derail the GMH agenda.

It is probably not a coincidence that attention to culture is lacking in the prevailing discourses on GMH. As noted in what follows, the majority of professionals currently engaged in developing the GMH agenda are either raised, and/or trained in HPI18 cultures. There is a natural human tendency to believe that one’s own experiences are normative, and to overestimate how similar others are to oneself; this is known in social psychology as social projection, or the false consensus bias (Marks & Miller, 1987; Ross, Greene, & House, 1977). This tendency is likely to be stronger when one belongs to a dominant group, since one is not required to challenge this assumption. As a demonstration of the lack of attention paid to culture in research agendas I conducted a search of the PsychINFO database in April 2012. Using the search term “research agenda” generated 2,521 publications in books and peer-reviewed journals; adding the search term “culture” reduced this number to just 174 references—less than 10% of the overall number of publications. It must not be forgotten that when systems of nosology such as the DSM were being developed, the majority of patients whose symptoms were being observed and documented were Caucasian. Thus, psychiatrists and other mental health researchers in HPI18 countries have inherited a nosology developed in and used for a specific culture, but are now using it rather unquestioningly among people very different from Caucasian Americans. With some notable exceptions (e.g., Kirmayer, 2006; Kleinman, 1977, 1995; Summerfield, 2008), it is almost always mental health researchers and psychiatrists from non-HP118 cultures who raise concerns about the lack of attention paid to culture, and unfortunately these voices appear to be going unheard.

The etic–emic issue is particularly salient in assessment and diagnosis, because mental health by definition involves perceptions and cognitions, and these are developed and shaped through culture (Alarcón et al., 2002; Kirmayer, 2006; Kleinman, 1977, 1995; Miller & Fernando, 2008). Despite the rhetoric on the importance of culture, even large-scale global surveys often either fail to assess and/or account for cultural beliefs about mental health, or do not address them in their publications. An example is the World Mental Health Survey Initiative, begun in 1998 and sponsored by the Assessment, Classification, and Epidemiology group of the WHO (World Health Organization Mental Health Consortium, 2004). In this multisite, multicountry study, researchers intentionally planned to address cultural variations in assessment and diagnosis, via face-to-face structured diagnostic interviews. Of the more than 500 subsequent publications, however, fewer than 25 explicitly examine the cross-cultural validity of the assessment interview and format used in the study. This is not to devalue the noteworthy findings of the many studies conducted under this initiative, but only to point out that such studies provide an excellent—but thus far missed—opportunity for researchers to fully engage in the difficult task of validating mental health constructs in different cultural contexts.

One way that researchers attempt to demonstrate the validity of assessments is to translate, and sometimes back-translate, checklists and measures that have been developed and validated in HPI18 cultures and then use them in LPCO cultures. Psychometric analyses of scores on these measures (e.g., internal consistency, test–retest reliability) are then taken as partial indications of validity. More methodologically rigorous research includes a second measure or clinical interview against which the assessment measure is compared. However, using a measure developed and validated in the United States in Uganda or Sri Lanka and then comparing scores on this measure to another measure or diagnostic criteria developed and validated in the United States cannot be considered sufficient evidence of ecological validity.

Clearly, accurate translation does not ensure that context and meaning have been accounted for (Miller & Fernando, 2008; van Ommeren, 2003), and context and meaning are essential to manifestations of psychological processes (Kirmayer & Young, 1999; Kleinman, 1995). Additionally, some turns of phrase in the English language are difficult to translate to other languages. For example, a popular measure of depression and anxiety is the Hopkins Symptom Checklist (Derogatis, Lipman, Rickles, Uhlenhuth, & Covi, 1974), which includes the item “Feeling blue” (distinct from the item “Feeling sad”). This expression is related to a culture-bound experience in the United States that is extremely difficult to translate. In our own work with youth in Sri Lanka (Fernando, Miller, & Berger, 2010), two items on an assessment of posttraumatic stress disorder (PTSD; APA, 2000) were particularly difficult for youth to understand. For example, the youngsters could not comprehend how one could possibly be “unable to have loving feelings” for those around them (a PTSD symptom of numbing), under any circumstance. Another concern with the use of such measures is that they are often administered in large-scale self-report surveys and, unlike participants in HPI18 countries, people from traditional cultures may not understand how to respond to such surveys. It is imperative that GMH researchers pilot-test their measures with small groups of representative respondents, to assess not only the theoretical equivalence of meaning in survey items, but also to assess what the respondents understand about the items (comprehensibility) and to assess the method of administration itself.

Emic considerations are also lacking in current diagnostic systems. The psychiatric disorders listed in the DSM-IV-TR (APA, 2000) could be considered disorders on a pseudoetic continuum, having been developed through observation and treatment of mostly Caucasian Americans in the United States. Although it may be reasonable to assume that syndromes such as depression, bipolar disorders, schizophrenia, and substance abuse disorders (the disorders focused on in the Lancet 2007 series) are culture-general, no large-scale phenomenological studies (as opposed to epidemiological studies) have been undertaken to test this assumption in different cultures. ³ The pseudoetic bias also may be operating when GMH researchers fail to pay more attention to somatic presentations of psychological distress. Although about a third of somatic symptoms presented in primary-care settings around the world cannot be medically explained (Prince et al., 2007), somatic syndromes are given scant attention in GMH research. The dualistic perspective in HPI18 research that splits symptoms by “mind” and “body” may lead some GMH researchers to consider somatic symptoms as “associated symptoms” rather than the core of a psychological disorder. Conversely, somatic symptoms may be mistakenly diagnosed as psychopathology. In a key study on depression sponsored by the WHO, Simon, VonKorff, Piccinelli, Fullerton, and Ormel (1999) found that a large range (45% to 95%) of patients diagnosed with depression in several countries reported only somatic symptoms, and a smaller but significant proportion (11%) denied experiencing psychological symptoms of depression when directly questioned. The data could not address whether the denial was due to fear of stigma or accurately represented reality, but such phenomenological studies hold much promise for a nascent GMH movement.

The diagnosis of PTSD (APA, 2000) has come under particular criticism for being pseudoetic. There is now a robust body of research indicating that the PTSD model, though easy to utilize via checklists and clinical interviews, may not be the most culturally valid or useful approach for working with people in traditional cultures postdisasters (Fernando, 2008; Miller & Rasmussen, 2010). Furthermore, the traumatic events themselves may not be what continues to cause “PTSD” in people posttrauma (for example, daily stressful events may also cause symptoms of PTSD [Fernando et al., 2010]); and psychosocial approaches, such as religious rituals and reconnecting to one’s social and family roles, may be more effective than the decontextualized cognitive-behavioral treatments developed in HPI18 countries (Fernando, 2005). In an emic study of trauma among the highland Quechua of Peru, Pederson, Kienzler, and Gamarra (2010) found that the particular sadness and anxiety-like symptoms reported by the group were due to extreme adversity resulting from “persistent social inequalities, social exclusion and a recent history of political violence” (2010, p. 279). The context-specific trauma syndromes experienced by this group are unlikely to be addressed by cognitive-behavioral or other purely psychological interventions.

Because of the controversies associated with the PTSD construct, GMH researchers have taken care to focus on those disorders considered common to all cultures: mood disorders, schizophrenia, and substance-abuse disorders. Yet even in the case of these seemingly etic disorders, it is possible that we may be missing core features of the disorder that do not fit neatly into the symptoms listed in current dominant diagnostic categories or checklists (Fernando, 2008). Additionally, assessment of posttrauma reactions should not be abandoned simply because of diagnostic controversy, particularly given the number of large-scale, high-impact disasters that regularly occur around the world. Rather, qualitative, ethnographic descriptions of mental disorders should first be obtained in different cultures. These descriptions can be used to develop measures using the indigenous language of the community. If some of the symptoms of that syndrome represent ones that are recognizable in other cultures, these can be grouped together in one part of the assessment, while culturally specific symptoms of that same disorder can be included in a second part, for use in that particular culture. The culturally similar symptoms can then be compared across cultures. In this way, both etic and emic expressions of mental and psychosocial health and distress can be addressed.

The current GMH agenda is also biased towards assessing intrapsychic processes of mental health. It is certainly true that mental processes occur within an individual, and that individual internal experiences deserve attention. However, how an individual experiences and expresses his/her mental health may be different in different cultures (Fernando, 2005; Kirmayer, 2006; Kirmayer & Young, 1999). In individualistic cultures, the focus may be on what the individual is feeling and thinking. In collectivistic cultures, it may be just as important, if not more so, to consider the person in the context of her/his family and community; that is, to consider psychosocial functioning (Fernando, 2008). Religious beliefs are also intimately associated with psychosocial functioning and mental health in collectivistic cultures (Deo, 2003; Hussein, 2009), and culturally informed discussions on such issues need to be included in a GMH agenda if it is to be authentically “global” in outlook. Inability to perform one’s religious or filial duties may not be an outcome of one’s symptoms of depression, as the current DSM model considers it, but may in fact be one of the symptoms (Fernando, 2008). While such a formulation may seem foreign to the current intrapsychic-oriented framework of mental health diagnosis, large-scale phenomenological studies of mental disorders such as depression have yet to be conducted in many countries around the world. Therefore one cannot definitively set aside the argument that psychosocial symptoms can be part of the presentation of mental distress in some cultures. Without paying some attention to emic formulations of psychological distress, the scaling up of programs and resources called for by GMH researchers (Horton, 2007; Lancet Global Mental Health Group, 2007) may only result in more studies that perpetuate the same diagnostic biases and category fallacies that Kleinman (1977) warned against over 30 years ago.

The etic–emic issue concerns not only assessment and diagnosis but training as well. Indeed, training is probably the single most important factor in ensuring that GMH researchers become both aware of and skilled in assessing cultural elements in the experience and expression of mental distress. Ironically, the field of psychiatry has identified as a primary need in GMH the availability of well-trained psychiatrists (Saxena, Thornicroft, Knapp, & Whiteford, 2007; WHO & World Psychiatric Association, 2005), while others have identified the need for more trained mental health professionals in general (Adams, Daniels, & Compagni, 2009; Beinecke et al., 2009; Kleinman, 2003; Shah & Beinecke, 2009). Kleinman (2003) has argued that raising a generation of GMH researchers is key to a successful GMH agenda. The irony is that researchers in resource-rich HPI18 countries who are identifying the need may be using and importing training models that are heavily biased towards their own dominant perspective on mental health. Thus, psychiatrists trained in China, Sri Lanka, and India initially may have ideas about how to ensure cultural sensitivity in research, assessment, diagnosis, and intervention in mental health, but the training they receive, using textbooks and methods developed in countries such as the UK and the USA, tends to erode these ideas over time. Consequently, “local” researchers trained in these models may miss significant associated or even core symptoms of a disorder in their culture when that disorder has already been delineated with other symptoms in the dominant diagnostic model. In this way, the “myth of etic psychological disorders” is perpetuated.

This unintended consequence of training and collaboration is disheartening to observe. I have reviewed several papers on PTSD conducted by teams comprised of researchers from outside and within Sri Lanka in which the issues of meaning and context of PTSD are given little, if any, attention. When this neglect was raised as an issue during the review process, the authors simply added it as a limitation in the Discussion section of their papers and recommended that future research attend to this issue; the editors considered this an acceptable revision. For researchers who believe that paying attention to culture improves the scientific endeavor, it is frustrating to observe how local mental health researchers become blind to the effects of culture on mental health when they team up with GMH researchers from HPI18 countries who consider culture to be a peripheral issue. There are of course some training models that are more sensitive to culture, like the program at McGill University (Kirmayer, Rousseau, Corin, & Groleau, 2008), but these are limited to specialized areas such as cultural psychiatry. The value of assessing culture may be taught to trainees in different fields of mental health, but its effect is not evident in the majority of papers currently published in GMH research.

An added advantage of paying attention to emic issues is that it may in turn provide insight into psychological disorders in HPI18 cultures. For example, Sri Lankan Sinhalese use the term hitha kadaavetuna (literally translated “[my] mind broke”) to describe being overwhelmed by grief or trauma (Fernando, 2009). When examining the meaning that this term is intended to convey, the best translation for this would be “My will was broken,” as the word “mind” here can be equivalent to the word “will” or “soul” in non-Buddhist cultures. Conversely, the term “hitha hadaagaththa” (literally “[I] made my mind”) is best understood as “I rebuilt my mind,” connoting “healing” or “mending” of the will. ⁴ Interestingly, depressed and grief-stricken people in the United States also often use the word “broken” to describe their condition, and recovery from depression is often described as a “clearing of the mind” from its fog of depression. Perhaps feelings of “being broken,” experienced as will-break or heartbreak, are etic experiences of some types of depression around the world.

Given that the GMH movement is still in its early stages of development, an argument might be made that focusing on the emic rather than on cultural similarities in mental health could derail this emergent movement or slow its momentum. Indeed, Saraceno et al. (2007) identify the differing and sometimes competing perspectives of mental health advocates as a barrier to improving mental health services, because lack of consensus and a clear, simple message might discourage donors from funding these projects. However, this threat should not outweigh the strength and opportunities presented by the nascent stage of the GMH movement. The movement has the opportunity, early in its journey, to incorporate some guiding principles that can be used to form consensus in the field. Indeed, as noted throughout this paper, there are efforts being made by researchers to incorporate culture into studies on mental health. By making these efforts the rule rather than the exception, the young but healthy GMH movement can only become stronger and more robust.

There is room for optimism that the GMH movement will embrace community-oriented, emic research. A few edited books and journal articles include studies on culturally relevant assessment, diagnosis and treatment, providing empirical and phenomenological information on mental disorders around the world (e.g., Bhui & Bhugra, 2007; Incayawar, Wintrob, Bouchard, & Bartocci, 2009; Ingleby, 2005; Salerno et al., 2009). Mixed-methods research is also increasing in frequency, with researchers using qualitative and quantitative methods to address the phenomenology of psychological disorders. For example in Northern Uganda and Rwanda, Betancourt and colleagues successfully used qualitative methods to determine the validity of existing measures of psychological syndromes, modifying the measures as needed and developing and statistically validating culturally relevant scales when the existing scales left a gap in knowledge (Betancourt et al., 2009; Betancourt et al., 2011). These studies provide ample evidence that emic-oriented research can be conducted using rigorous scientific methodology. It is hoped that the GMH movement will move further in this direction, and the pseudoetic pathway eventually will lapse into disuse, remaining only as historic trail.

Key Issue 3: Identifying stakeholders

Stakeholders drive agendas, and each stakeholder in the GMH endeavor wields some power in shaping its agenda. In one of the most helpful papers in the Lancet series, Saraceno et al. (2007) identify the absence of a consistent message as one of the barriers to the GMH research agenda. This may be due to conflicting agendas among stakeholders, which could result in a fragmented message to the public. At a minimum, primary stakeholders in the GMH research agenda include: researchers; funders; the funders’ stakeholders (e.g., federal/national organizations, donors); the local communities being studied; organizations and liaisons in the countries where the research/interventions are being conducted; organizations interested in the research questions and obtained data; consumer groups; and, in intervention research, clinical service providers. Secondary stakeholders include: the researchers’ institutions, colleagues/coworkers and supervisees; journal editors and publishers that disseminate findings; and the research community that digests the findings.

Local communities are crucial stakeholders in GMH research, but they may believe themselves to have the least amount of power. If these communities lack leaders who represent their interests and voices, they may be unable to articulate their needs. Communities may also perceive that voicing such needs will result in further fragmentation of the research endeavor. Researchers should therefore build into their research studies some mechanism for eliciting and responding to the mental health needs of LPCO communities, and attempt to marry those needs to the research questions that interest the researcher. If the researcher is willing to be educated by the community where the research is being conducted, she/he may gain valuable insights into the way in which mental health is experienced and expressed in that group, and those insights could inform ecologically and culturally valid evaluation and intervention programs. This would be a useful corollary to Saraceno et al.s’ (2007) call for GMH researchers to use nonformal community resources in research.

Editors, publishers, and other scientists who disseminate the findings of GMH researchers also have a stake in the research endeavor. Unfortunately, only about 10% of published papers in mental health are authored by researchers living and working in non-HPI18 cultures (Patel, 2006; Saxena, Paraje, Sharan, Karam, & Sadana, 2006). One disheartening analysis found that of 11,501 clinical trials published on a range of psychological disorders, fewer than 1% were identified as being from LICs, and less than 10% were from lower middle-income countries (Patel, Araya, et al., 2007). The politics and above all the economics of publishing may influence which types of studies are published and accorded more scientific credence. Researchers invested in cultural conceptualizations of mental health tend to favor qualitative studies, but the scientific readership may not consider these to be as valuable as quantitative studies using evidence-based methods (Kirmayer, 2006). These qualitative empirical studies and concept papers are therefore less likely to be published in high-impact journals, whose editors are justifiably invested in the perceived scientific merits and readership rates of the journal. Yet papers written by researchers intimately associated with their culture are likely to contain useful emic information on mental health. While there has been a welcome increase in the number of studies published from LPCO cultures, and open-access online journals are providing more platforms for diverse studies, the 10/90 divide in higher impact journals is still too vast, and needs to be addressed. Editors, publishers, and others can collaborate with researchers from non-HPI18 cultures to increase their publications, and develop a rich body of data that can inform the GMH research agenda.

Key Issue 4: Tying the goals of research to stakeholders

Stakeholders have agendas, and this is no different in the GMH research endeavor. When working collaboratively, these agendas can be mutually beneficial, but competing agendas may derail the GMH vehicle. Although there are numerous theories (e.g., Kirmayer, 2006) and some publications in investigative journalism (e.g., Watters, 2010) about intentional and unintentional ways in which research interests and funding interests may collide, at present there are no systematic data on how and whether the different agendas of stakeholders might undermine the integrity of GMH research. One recent study is a promising first step (Sharan et al., 2009). In a large-scale study of the research priorities and stakeholder goals in 113 countries, the authors found that, overall, stakeholders’ priorities were congruent with those of the researcher. The findings of the study may be biased, since data could only be analyzed on the surveys that were returned; those who did not respond to the survey may have been the very stakeholders whose agendas were incongruent. Nonetheless, this is an important first step in examining the congruence of stakeholder priorities in the GMH research agenda. There need to be more such systematic reviews of stakeholder goals, particularly in low-income communities where power differentials between stakeholders are greater.

A GMH researcher may be unduly influenced by some stakeholders because of the practical reality of needing funding to maintain a program of research. If conventional research that uses pseudoetic approaches and measures is more likely to receive funding, the researcher must balance the need for continued funding against the goal of conducting an ecologically sound study. In instances where it becomes difficult to find funding for a qualitative study needed to develop emic measures of mental health, a researcher may be forced to move directly to the quantitative study that is more likely to be funded. Thus, competing goals of stakeholders may sabotage a truly global mental health research agenda.

Researchers working in a culture different from their own often must rely on community organizations as liaisons and translators to carry out some parts of the study. These community organizations also have a stake in the research endeavor. Language confers power, and in many LPCO countries where local community-members do not speak the languages of the HPI18 researchers, knowledge of the researchers’ (HPI18) language is a prized resource. While most community liaisons and translators work for the benefit of the community, some may use their skills and resources to benefit themselves or their families rather than the community as a whole. These “language brokers” may place themselves between the researcher and the local community in ways that make it difficult for the researcher to obtain an accurate community profile. Thus, researchers who rely on community liaisons and translators must pay attention to the relationship between these workers and their community. By spending time in the community prior to conducting the research, researchers can observe whether the community liaisons/translators have the trust of the community they serve.

Key Issue 5: Evaluation of benefit

Ultimately, GMH research should benefit the communities to which it brings international attention. As many mental health professionals rightly point out, there is great suffering among people in communities across the globe (Desjarlais, Eisenberg, Good, & Kleinman, 1995; Kleinman, 2009; Shah & Beinecke, 2009). Rigorous, ecologically valid research could help to alleviate this suffering via sound phenomenological and epidemiological studies and culturally sensitive interventions with high internal, external, and ecological validity. However, if there are competing goals among the multiple stakeholders, or the goals of the community are incongruent with those of a powerful stakeholder, it is unlikely that the community will benefit.

There are several beneficiaries of large-scale research studies. These studies certainly benefit the researcher, whose credibility with funders and policy makers is enhanced. The researcher’s institution also gains prestige by association. The journals that publish the findings of the researcher stand to gain in higher impact-factor scores and subscription rates. The research community benefits from the additional information provided on global mental health. The funders and funders’ stakeholders are pleased because the funds were used in a study that receives international recognition.

Whether the local community benefits from much of the GMH research, however, is still in question (Tuhiwai Smith, 1999). Many researchers list the gathering of new information or the validation of an intervention as the main “benefit” to the community. However, it is uncertain whether the local communities would readily identify these outcomes as benefits. In an insightful and somewhat stark analysis of current research methods, Tuhiwai Smith (1999) denounces research that benefits all stakeholders except its participants. A GMH research agenda that is community-focused should build in some form of postresearch assessment to evaluate whether the research benefited the community it engaged. GMH researchers must be willing to listen to the voices of the community and adjust their research to include aspects that will result in tangible benefit to the community. A culturally sensitive research agenda requires listening to the voices of the community, and responding in substantive ways.

Conclusion

This paper highlights some concerns important in shaping a culturally responsive research agenda for global mental health, and suggests some pathways on a roadmap for a culturally informed GMH movement. Researchers must define GMH in language that prioritizes culture; grapple with the etic–emic tension in assessment, diagnosis, intervention, and training; pay greater attention to variables such as psychosocial functioning, resilience, and recovery; work on reducing the 10/90 divide in research publications; be prepared to include cultural phenomena as core aspects of mental health rather than as peripheral data; ensure congruence of stakeholder goals; prioritize the goals of the community; and, most importantly, include outcomes that communities identify as beneficial.

The onus for negotiating the competing agendas of stakeholders lies with the researcher. The researcher should develop an approach that is sensitive to the historical, sociopolitical, and psychological contexts and current lived realities of the communities in which they work. If GMH researchers truly have the local community’s interests at heart, they can engage in specific actions to bring together the different agendas of stakeholders. Researchers can secure small initial grants to fund emic, culturally authentic studies to develop ecologically valid assessments. The findings from such studies can in turn be used to educate other funders and the larger research community about the lived experience of the local communities struggling with psychosocial dysfunction and/or demonstrating resilience. Studies should include outcomes considered valuable both to the larger research community and the local communities where the research is conducted. Similarly, researchers can use their expertise and resources to educate LPCO and LPI community leaders about the research endeavor, and how best to leverage community resources (e.g., through participation in appropriate research studies) to meet the needs of the community.

This inclusion of the community in the research endeavor would increase the scientific merit of the research, and could be therapeutic. There is robust evidence to indicate that controllability and empowerment increase resilience in individuals and communities (e.g., Alexander, Langer, Newman, Chandler, & Davies, 1989). Such an approach also responds to the urgent call by those in the GMH movement to increase capacity in local communities (Adams et al., 2009; Beinecke et al., 2009; Horton, 2007), since GMH researchers could train local community members as research assistants, and, in turn, could learn from their community collaborators about the mental health of the community. Researchers could also use ecologically valid findings from their studies to push for policies that would increase resilience in communities and alleviate psychosocial and intrapsychic distress. When a researcher brings his/her own expertise to the research endeavor, yet is willing to be led by the community in developing research questions, the resulting positive synergistic effects can produce truly valuable results.

While the major responsibility for conducting culturally responsive research lies with the researcher, it is imperative that other stakeholders support these efforts of the researcher. Funders could make available more funding for innovative, qualitative studies that elucidate the cultural formulations of specific mental/psychosocial disorders as well as processes of well-being. Editors of high-impact journals that generally publish more quantitative studies could make space for these qualitative studies. Research and academic institutions could also provide more infrastructure support and culturally informed training for researchers engaged in GMH research. Task forces, which include LPCO community members and leaders, could be organized periodically to engage in self-reflection of the field, which could evaluate the cultural biases of present trends in GMH research. Empowered local communities could organize as a body to encourage research questions and studies they believe would best illuminate and address the mental health questions that most impact their communities. The knowledge gained from these local communities could even be used to reformulate diagnosis and intervention in HPI18 cultures. Such bidirectional pathways can only mutually benefit of GMH researchers and the communities they study.