The Educational Commission for Foreign Medical Graduates (ECFMG) (1) defines an international medical graduate (IMG) as "A physician who received his or her basic medical degree or qualification from a medical school located outside the United States and Canada." Historically, IMGs have significantly contributed to the U.S. physician work force across all specialties, particularly psychiatry. IMGs comprise between 23% and 28% of U.S. physicians (2). Roughly 41% of psychiatric residents are IMGs (3), which is especially impressive in light of widespread physician shortages (4). IMGs serve the underserved, both as trainees and practitioners, and disproportionately treat immigrant and minority cohorts (5).

Residency is a time of great emotional and psychological distress. Workload, specialty choice, practice setting, patient characteristics, personality type, work-life balance, and dealing with death and suffering are all factors that may influence physician distress (6—8).

Resident stress and its antecedents have been approached with an implicit assumption of homogeneity of residents, despite the immense diversity among residents in country of origin, first language, country of undergraduate medical education, and so on (9—12). Surprisingly, the distinction between an IMG and an American medical graduate has rarely been discussed, and the effects of acculturative stress on IMGs have mostly been neglected (9).

As a subset of the immigrant community, IMGs share certain common characteristics. IMGs confront, to varying extents, the twin stressors of migration—acculturative stress and changes in social support. IMGs from non-English-speaking countries especially have difficulties adjusting to their new social environment largely due to unfamiliarity with U.S. customs, norms, and values and, consequently, have difficulty effectively interacting with their coworkers and patients. Striving for balance between personal and academic demands and the subjective loss of social status that accompanies this cross-cultural transition intensifies the burden (13).

Prevalence estimates of psychopathology in psychiatric residents range from 4% to 22% (14). There is some evidence that serious psychological disturbances (including substance abuse) may be an important reason for dismissal of psychiatric residents (15). However, the available data do not distinguish between American medical graduates and IMGs.

Understanding migration and its antecedents are important in identifying and addressing the psychological needs of IMGs. A strong association exists between both social support and acculturative stress and psychological well-being. Improved psychological health can assist in successful cultural adaptation and, eventually, in the optimal integration of IMG residents into the U.S. physician workforce.

Social support, defined here as provision of psychological and material resources intended to benefit a person's ability to cope with stress (16), is divided into instrumental, informational, emotional, and appraisal subtypes (17—19). Research supports a positive correlation between social support and mental health (20). Smaller social networks, fewer relationships, and lower perceived adequacy of social support have been linked to depressive symptoms (21).

Acculturation, defined here as psychosocial adjustment and adaptation to a new culture for people from another culture (21—23), is a multidimensional process of change that occurs when persons of differing cultural groups come into continuous contact and depends on how a person deals with two fundamental issues: immersion in an ethnic society other than the dominant society and adoption of the dominant society.

Some attention has been focused on the barriers within the American medical system that IMGs face while applying to residency training programs, obtaining licenses, and considering practice options (24, 26). Factors that influence an IMG's mental health (possibly compounding those barriers) have not been examined. This study was intended to determine the degree to which social support and acculturation affected the overall mental health of IMGs enrolled in psychiatric residencies across the United States accredited by the Accreditation Council for Graduate Medical Education (ACGME).

This study was approved by the Review Board for the Protection of Human Subjects at the University of Texas in Houston. A 55-item, self-report questionnaire was assembled online by putting together validated and psychometrically established instruments for mental health, social support, and acculturation. Only IMGs who had a first language other than English and were in a U.S. psychiatric residency training program were enrolled.

A link to the study was e-mailed to all residency directors through the American Association of Directors of Psychiatric Residency Training (AADPRT) 2008 member directory. Participation was voluntary and anonymous. Before the actual survey, the web site presented an information screen that explained the study's purpose, directions for completing the survey, and stated that consent was implied by survey completion.

Demographic information was obtained, including respondents' age, gender, and level of residency training (postgraduate years 1—4). The 12-item version of the Interpersonal Support Evaluation List, a multidimensional inventory used to evaluate the effect of perceived availability of social support resources on health and well-being (27), was used to assess social support. The Interpersonal Support Evaluation List was composed of items that measure belonging or companionship support, appraisal support (described as emotional support or someone to talk to about problems), and instrumental support (the availability of tangible material aid). For each Interpersonal Support Evaluation List subscale, higher scores indicate higher levels of perceived appraisal support, belonging support, or tangible support.

Mental health was measured through the Kessler Psychological Distress Scale (28, 29). The scale was designed to be sensitive around the threshold for the clinically significant range of nonspecific distress to maximize the ability to discriminate cases of serious mental illness from nonserious cases. Serious mental illness has been defined as any DSM-IV disorder lasting 12 months, other than a substance use disorder; with a Global Assessment of Functioning (GAF) score of less than 60. The Kessler Psychological Distress Scale is composed of six Likert-type questions (with a possible range from 6 to 30); scores less than 19 on this scale indicate serious mental illness.

Acculturation was measured through Stephenson's Multigroup Acculturation Scale (30), a 32-item, Likert-type scale composed of 17 "ethnic society immersion" items and 15 "dominant society immersion" items. This instrument has been tested with diverse groups (31) and includes questions corresponding to domains of language, interaction, media, and food. Certain survey items were used as proxies for excluding respondents who may not have undergone acculturation (e.g., a U.S. citizen born and raised in the United States who completed medical school in the Caribbean) or whose first language was English.

Statistical Package for Social Sciences (SPSS) version 17.0 was used for data analysis. Level of significance was set at 0.05. Any survey with more than 10% missing values was excluded from analysis. Maximum likelihood was used to replace random missing values. Descriptive statistics, such as frequencies and means, helped identify participant characteristics and evaluate assumptions for regression analysis. Multiple regression analyses predicted the variance of mental health attributable to acculturation, social support, age, gender, and postgraduate training year. All possible interaction terms were tested, but only significant interaction terms are reported.

Descriptive analysis indicated that most respondents were between ages 26 and 30 (n=45, 41.7%). The remainder were most frequently between ages 31—35 (n=36, 33.3%) and 36—40 (n=13, 12%), while 11 respondents (10.2%) were older than 40. Respondents (n=108) were composed of 53 men (49.07%) and 55 women (50.9%). Thirty-four respondents (31.5%) were in their first postgraduate year (PGY-1), 25 in PGY-2 (23.1%), 23 in PGY-3 (21.3%), 18 in PGY-4 (16.7%), and eight in PGY-5 (7.4%).

Table 1 shows the reliability coefficients for the study instruments. Established norms for the interpretation of the reliability coefficient state that a value of 0.70 suffices for early stages of research (32). Most items were found to be reliable at that level.

Figure 1 shows the normal distribution of the mental health scores. Possible scores on the Kessler Psychological Distress Scale ranged from 6 to 30. Respondents' mean mental health score was 25.31 (SD=3.74), which is considerably higher than the cutoff used for serious mental illness when mental health is measured by the Kessler Scale (a score less than 19). More than 92% of participants scored higher than 19 on the survey items related to mental health.

Table 2 shows the means and standard deviations for the social support and acculturation scores. Stepwise multiple regression was used to pair the significant independent variables. The general purpose of multiple regression is to uncover the relationship between several independent (predictor) variables and a dependent (criterion) variable. The best model (Table 3) found one construct of acculturation (dominant society immersion, or DSI), the appraisal construct of social support, and the postgraduate training year as significantly paired with mental health. The coefficient of determination, R², used in statistical models for the prediction of future outcomes, is the proportion of variability in a dataset that is accounted for by the statistical model. This model has an adjusted R² of 0.241 (Table 3). Together, these constructs can explain 24.1% of the variability in mental health of the respondents. Ethnic society immersion, age, gender, and two constructs of social support (tangible and belonging) were not significant predictors.

This study explored the variables associated with the mental health of international medical graduates (IMGs) pursuing psychiatric residencies. Although most respondents had very high mental health scores, 7.4% of respondents scored less than 19, the cutoff used for serious mental illness. Preexisting or current mental health issues and other concurrent psychosocial stressors and/or life events may have contributed to these scores.

The study adds to the body of literature on trainees in residency programs and helps draw attention to the paucity of research on IMGs' mental health (10). That migration and the accompanying acculturative stress may be related to adverse mental health outcomes is an old idea (33). Despite an exhaustive literature review, we did not find any studies exploring the link between acculturation and social support with mental health among IMGs. One study looked at acculturation's effect on burnout in IMG residents (10).

The potential implications of this study are many, both for psychiatry training programs and trainees. The appraisal component of social support was found to be a significant predictor. IMG residents who have higher emotional social support are more likely to have increased mental health (Table 3). As a corollary, an IMG with inadequate emotional support would have lower mental health scores. This positive correlation of social support with mental health is in line with contemporary literature on social support, which visualizes social support as either a buffer for dealing with acute crises or a normative component for healthy growth and development (34).

In contrast to previous research, where mental health is higher with integration than assimilation of the migrant (35), this study found that dominant society immersion was a mental health predictor. This finding may relate to the uniqueness of our study population relative to other migrant groups (e.g., frequent physician-patient interpersonal encounters, academic credentials, higher perceived need, and external pressures to speak English fluently). If this finding is replicated, future interventions targeting acculturation in this group may have to be designed.

Postgraduate training year was also a significant predictor, perhaps because as IMGs advance through residency, they may get used to the rigors of training. Also, it is logical to assume that social networks and a greater degree of acculturation develop with the passage of time. It is worthwhile for training directors and faculty to be aware of this association.

Interventions that help improve the social support and acculturation of IMGs could potentially positively affect the academic performance of IMGs in that language is integral to both acculturation and academic performance. IMGs from non-English-speaking countries may profit from language training focusing particularly on pronunciation, slang, and accent reduction (36). They could also potentially benefit from curricula in North American history and culture. Of course, this acquisition of knowledge specific to U.S. culture would increase the academic burden on an IMG, but it is justifiable if it would help smooth and facilitate the process of acculturation.

Mentoring relationships to deal with personal and professional problems during training can help buttress the appraisal social support of IMGs. Residency programs could also institute buddy systems in which an upper-level trainee is paired with an intern to help facilitate the intern's transition to the field. In 2006, the Educational Commission for Foreign Medical Graduates instituted a new "Acculturation Program" that assisted IMGs with the transition to working and living in the United States. The program attempts to pair new residents with advisers; offers introductory videos discussing the U.S. health care system; and, more importantly, highlights the health care jargon typically used (37). This is a step in the right direction, but as our data suggest, much work is left to be done.

This study found that dominant society immersion, the appraisal component of social support, and postgraduate training year were significant predictors of mental health in a study population composed of international medical graduates pursuing psychiatric residency training. Graded by the standards proposed by the Institute of Medicine (37), this study can be categorized as having limited or insufficient evidence of association. More research needs to be done to provide evidence for causality. The results of this study are limited by the study design, including participant recall and self-selection bias. The small sample size, convenience sampling, lack of similar studies in the research literature, and the failure to exclude residents who have a previous or current history of mental illness or who may have experienced significant life events during their residency further limit the generalizability.

At the time of submission, the authors reported no competing interests.