Samuel Noh, PhD, Ilene Hyman, PhD, and Haile Fenta, HO, MPH, PhD

Culture Community and Health Studies Program

Centre for Addiction and Mental Health

and

Department of Psychiatry, University of Toronto

In Cooperation with

The Ethiopian Association in Toronto

April 2001

TABLE OF CONTENTS

Executive Summary viii

Introduction 1

There have been significant advances in understanding social and cultural determinants of health during the last three decades. These scientific advances in population health have been contained in the understanding of the "general population" of domain groups in North America. However, the research on immigrant health has been less extensive and less well integrated. In part, this lag in scientific progress is due to the lack of political awareness and the lack of systemic and stable support for research on immigrant and minority population health. This is particularly the case in Canada, as evidenced by the relative scarcity of scientific data on immigrant health compared to the United States. For example, although the need for structural reform tends to coincide with social and demographic changes, and while there have been occasional eruptions of controversy over health care systems in Canada, ensuing policies have centred primarily on meeting the needs of native-born Canadians. The well being of immigrants and their children has rarely provided a strong impetus for public concern or action.

This report is derived from a survey research that has been funded by the Centre of Excellence for Research in Immigration and Settlement (CERIS) in Toronto and Canadian Heritage. The investigators of this research project are grateful to CERIS and Canadian Heritage for its dedication in immigration and settlement research and its outreach to the academic and community researchers, which has fostered genuine collaborations between these groups. The present research project and this report are outcomes of an effective collaboration between the Ethiopian Association in Toronto and a scientific research team at the Centre for Addiction and Mental Health and University of Toronto. The shared interests and more than generous offerings from the community were the main causes of the success of the research project.

We thank the 342 participants who shared their hours with our trained interviewers. Particular appreciation is extended to Mr. Yonas Taddesse, who spent many days and nights translating the extensive interview schedule into the Amharic language, and to all interviewers, who had to bear with such onerous tasks during the training sessions as contacting interviewees, conducting interviews, and filing the completed interviews. We would like to recognize each interviewer individually: Mr. Wosen Yitna Beyene, Mrs. Elizabeth Guete, Dr. Selamawit Tessema, Dr. Eyob Hailu, Mr. Abai Mengesh, Mr. Wudneh Bayleyegn, Dr. Fatuma Abdi, Dr. Abebe Worku, Dr. Bahru Melese, Mr. Benyam Hailu, Mr. Yibabe Kassa, Mr. Nigusu Goshu, Mrs. Rebecca Tola, Mrs. Tsedenia Getaneh, Mrs. Helina Besrat, Mr. Leikun Teshome and Kiddist Tesfaye. Undoubtedly, the research project would not have been possible without the dedicated and generous support of these people.

Finally, we would like to express our gratitude to the members of the Ethiopian Community Advisory Committee for their professional insights and persistent guidance which made this project possible and successful. We are very certain that the contribution and dedication of all participating members of the Ethiopian community and the findings of this study will assist researchers, community service providers, and policy makers to better serve the needs of the community and to understand the health and well-being of individual Ethiopian immigrants and refugees.

April 2001

The present document outlines the preliminary findings from epidemiological survey research of Ethiopian adults living in the Greater Toronto Area in 1999-2000. The focus of the study was to determine the prevalence of such major mental disorders as depression, anxiety, Post Trauma Stress Disorder (PTSD) and somatization. The study also describes the utilization rates and patterns of diverse health care services. These issues are critical in all communities of new settlers. However, Ethiopians may be at increased risk for such mental disorders. Many Ethiopian migrants have been exposed to traumatic crisis and events in their home before emigration, and have been wandering for an extended number of years before receiving Canadian permanent residency. It was hoped that the present study would provide new and pivotal information in comprehending the extent and nature of mental health needs in the "exposed" community and in planning social and health care services for the victims of traumatic life experiences in the Ethiopian community as well as those immigrants and refugees in other communities with similar experiences.

This research is the first epidemiological community survey concentrating on the mental health needs in Ethiopian communities in North America. Although the scope of the study has been limited due to fiscal constraints, the findings of the study provide highly valuable information, where some results confirm public perceptions and other results contradict general stereotypes.

Results illustrated in this report are based on data derived from person to person interviews with 342 adult participants and a sound representation of adults (aged 18 years or older) of 5000 households identified by this research team.

• The study consisted of 203 (59.4%) men and 139 (40,6%) women, mostly young adults between 30 to 49 years of age (76.6%). While the age distribution appears to show an over representation of younger adults, in fact, our study sample is over represented by those aged 30 years or older, compared to a 1996 census on Ethiopians in Toronto.
• One half were currently married or living in common law, while more than one third had never been married. A considerable minority (one in ten) were previously married – divorced, separated, or widowed.
• An overwhelming majority spoke Amharic (86.5%), and were members of either the Ethiopian Orthodox Church (68.0%) or protestant churches (22.0%). These figures include the 16 percent of respondents who changed their religious affiliations since coming to Toronto, mostly to protestant.
• More than 95% completed high school education; one quarter (23.4%) had at least one university degree, and another 44.2% had college education.
• Eighty percent of the respondents used Ethiopian local languages at home; another 10 percent used both English and native languages. About 7% used English exclusively.

• A large majority left their homes in urban areas in Ethiopia (87.1%).
• More than half (53.8%) left their home for "political reasons." The figure for this reason was substantially higher among men (64.5%) than among women. Among women, "for a better life" was the most popular choice for primary reason for coming to Canada (57/1%); however, even among women nearly one in four (37.6%) said the primary motive was political.
• Most left Ethiopia during the early twenties. The mean age at emigration was 23 years.
• More than three-quarters (77.8%) stayed in other countries, for an average of four years. Most stayed in either neighboring countries (most in Kenya, Sudan, and Djibouti) or Southern Europe (most in Italy and Greece).
• Approximately 11 percent stayed in refugee camps, for an average of 25 months. Compared with women, a greater proportion of men (12.6%) had been in refugee camps for a significantly extended period (mean of 31 months).
• Nearly three quarters of the respondents arrived in Canada alone, either as a refugee (37.4%), refugee claimant (7.6%) or independent immigrant (29.8%). One quarter came under family immigrant class.
• Currently, respondents have been living in Canada for an average of nine years; nearly all were either Canadian citizens (80.7%) or landed immigrants (16.7%); nine respondents (2.6%) were refugee claimants.

• Almost 80 percent of respondents were employed.
• Of those employed, 84 percent were employed full time.
• Although men were slightly more likely to be at work, a substantial proportion of women were in the labor force (64 percent).
• Less than 6 percent were unemployed (being in labor market and actively seeking employment); 4% among men and 8% among women.
• However, almost 60% of men and 50% of women said they were "under-employed."
• These figures suggest Ethiopian immigrants’ and refugees’ willingness, or necessity, to work under any conditions, and their ability to find work.
• About 15 percent of the study sample were either attending school or out of labor force for other reasons.

• About 65 percent of the sample experienced one or more forms of discrimination in Canada because of their racial or ethnic background.
• Close to 50% of the sample reported at least one incidence of subtle discrimination (being looked down upon, expressed resentment, unfair treatment); more than forty percent said they were treated rudely in public.
• Although the rates of overt discrimination were significantly lower compared to subtle forms of discrimination, one quarter of the sample said they received insulting remarks, and more than 10 percent were threatened because of their race, while 2 percent were physically abused or attacked.
• Substantially greater proportions of male respondents reported the experiences of discrimination.
• Despite the gender difference in reporting the incidents of discrimination, the reports on the setting in which discriminations were experienced varied little between men and women.
• Both men and women reported most experiences at work (44%) and in service sectors (30.7%). More than a quarter reported incidents "from the school system."
• Over 20% of male respondents and 3% to 14% of female respondents reported experiences of racial discrimination from government agencies, media, and police.

• The lifetime prevalence of depression (Major Depressive Disorder or MDD) was 9.8%. The lifetime prevalence of MDD was comparable to the rate found in Ontario, but three times higher than the rate reported for Ethiopia (3.2%) by Awas, Kebede, and Alem (1999).
• In the general populations of Ethiopia and Ontario, the prevalence was significantly higher for women than men. In our sample, we found little gender difference in depression. In fact, the rate was slightly higher for men (10.4%) than for women (8.4%). This unusual pattern was also found during the initial settlement period among the Southeast Asian refugees in the Vancouver area (Hou and Beiser, 2001). The pattern can be explained by the fact that more male than female Ethiopians had "refugee experiences" including temporary stay in other countries, living in refugee camps, and arriving as a refugee or refugee claimant.
• Indeed, in the present study, we found that the post-arrival (in Canada) incidence of MDD was more frequent in women (6.6%) than in men (5.4%).
• In addition, 3 percent and 5.8 percent of the current sample had generalized anxiety disorder and Post-traumatic Stress Disorder (PTSD), respectively. The PTSD was more prevalent among men (6.9%) than among women (4.3%), although the gender difference was not statistically significant.
• Major predictors of mental disorders were pre-migration factors (experiences of trauma and stay in refugee camps), post-migration life events and lack of social support.
• Data suggest that there is an urgent need for caring for those who had been forced to live in refugee camps and experienced trauma during the migration process.
• Results on psychosocial correlates of depression demonstrated that intervention programs concentrating on employment would be effective in easing psychological strain in the community. While unemployment was directly related to depression, it was also related to the increase of family and marital conflict, lower income, and other life events related to inadequate income.
• Living alone also constituted a risk for depression. Non-believers were also at high risk. But, social support for emotional and instrumental needs were effective in preventing mental disorders. It seems church based intervention for single persons may promote informal networks to meet the emotional needs.

• The majority of respondents (82.3%) had taken a drink of alcohol in their lifetime

• Among those who had taken a drink of alcohol in their lifetime, 85.6% had taken a drink within the past 12 months.

• Over 14% of respondents reported that they drank enough to get drunk in the last 12 months; and this proportion was higher for males than for females (18.4% among males and 5.5% among females).

• The prevalence of current drinkers in the study population was higher than the rate for the Ontario population (85.6% and 73%, respectively).

• However, the proportion of frequent drinkers (i.e., individuals who reported drinking every day or 4-6 times a week) was lower in the study sample compared to the Ontario population.

• Of those ever drunk, one-third (33.9%) reported that they started drinking more since they had left Ethiopia or after arriving at Canada. Male respondents were more likely than female respondents to report increased drinking (36% and 29.2%, respectively).

• Approximately 21% of the respondents reported smoking regularly or occasionally.

• Male respondents were more likely to be smokers compared to females (28.1% and 10.3%, respectively).

• Males were also significantly more likely to begin smoking daily at a younger age and to smoke more cigarettes per day than females. The overall prevalence of smoking was about the same as the Toronto population in 1996 (19%).

• Thirty eight percent of our study population had tried drugs in their lifetime. This proportion was significantly higher for males compared to females (53% and 15.6%, respectively).

• The most frequently tried drug was Chat (28%).

• Approximately 86% of Ethiopians used at least one form of medical service during the 12 months immediately preceding the interviews.

• Most of them had visits to family physicians.

• However, more than 16 percent of the sample made a visit to a hospital emergency room, and 5 percent were hospitalized during the last twelve months.

• Overall, women used the services more frequently than men.

• Regarding help-seeking for mental health / emotional problems, or stressful life events, 4.9 percent of males and 5.9 percent of females said they saw at least one medical professional, mostly family physicians and psychiatrists, for such problems during the past 12 months.

• A greater proportion of females (12.9%) than males (4.4) went to non-medical help for their emotional and other forms of life problems.

• Overall, the data suggested that 8.4% of male and 16.5% of female respondents felt a need for professional help, medical or non-medical, for their emotional problems.

• These utilization rates suggest that Ethiopians in Toronto are likely to use more medical services compared to the immigrant and refugee communities in Montreal including, Caribbean, Vietnamese, and Filipinos).

• The overall rate of service use in the Montreal sample was 78.5%, compared to the rate of 85.5% among the Toronto Ethiopian sample.

• The help seeking rates for mental health / emotional problems were 3.6%and 5.3% for the Montreal and Toronto study samples, respectively.

April 2001

1. Introduction

Recent epidemiological literature reports health advantage of foreign-born residents in the leading receiving countries - Australia, Canada and the United States (Hernandez and Charmy, 1999; Vega et al, 1999, Chen et al, 1996; Noh and Avison, 1996). However, such an overarching conclusion may mislead readers to believe that all immigrants and refugees enjoy better health compared to non-immigrant residents of Canada. In fact, considerable variations in health were found across diverse ethnic groups of foreign-born populations. Those with refugee status and previous experiences of refugee camps and traumatic events prior to arrival are more likely to demonstrate many forms of psychopathology.

Ethiopians in Toronto appear to be exposed to most critical risk factors. An analysis of the 1996 census of Toronto residents showed that such risk factors as poverty and unemployment were highest among Ethiopians, as well as Ghanaians and Afgahanistans. Are these rates of risk factors translated into exceeding rates of mental disorders and physical illnesses? Are they receiving adequate health care provisions? While these questions are critical in planning health care and social programs for newly arrived refugees and immigrants there has been no attempt to estimate the needs of mental health care within the community in distress.

The purpose of this study was to obtain reliable and valid assessments of mental health problems among adult refugees and immigrants from Ethiopia living in the Greater Toronto Area. Specifically, the objectives of the study included:

1. To improve our understanding of the nature and extent of mental health problems among Ethiopians in Toronto;
2. To describe the health risk behaviour, (alcohol, tobacco, and drug use), among Ethiopians in Toronto;
3. To identify the patterns of health care utilization among individuals with depression, anxiety, somatic disorders and stressful life events;
4. To examine the relationship between acculturation, other psychosocial variables, emotional problems as well as service utilization rates; and
5. To provide input for formal and informal help systems in order to improve the delivery of mental health services to Ethiopians in Toronto.

1. Background

Ethiopia is the oldest independent country in Africa that has never been under colonial rule. It is a land of great diversity in its topography, climate, people and languages. There are 80 cultural and linguistic groups that live together in Ethiopia. The major ethnic groups include Amhara, Oromo, Tigrean, Sidama, Shankilla, Gurage, Somali, and Afar. The dominant religions are Christianity (mainly Ethiopian Orthodox) and Islam. About 80% of the population live in rural areas. According to the 1995 estimate, the literacy rate was about 36%. Regarding healthcare, an estimate of 43% of the population has access to basic health services. In terms of mental health care services, there is only one psychiatric hospital and 10 psychiatrists in the country.

Until 1974, Ethiopia had been a peaceful country. The many cultural and religious groups had lived together peacefully for a long time. However, in 1974, after a long and peaceful reign, the Derg (military) led by Lt. Col. Mengistu Hailemariam, deposed Emperor Haile Selassie. The Derg stayed in power for 17 years between 1974 and 1991, a period of totalitarian rule, massive militarization, and the adoption of Communism. There were also border clashes with Somalia and a civil war with the province of Eritrea. Thousands of suspected enemies of the military government were tortured and killed and the major exodus of Ethiopian refugees began.

In May 1991, the Ethiopian People's Revolutionary Democratic Party, an ethnic rebel group, overthrew the Derg and formed a transitional government. In 1994, the Ethiopian People's Revolutionary Democratic Party joined by other ethnic groups to form a Federal Government, which adopted ethnic politics. This created ethnic tension. Opponents of ethnic politics were imprisoned, killed, and the whereabouts of some individuals was unknown, which led to the second wave of exodus of Ethiopian refugees to neighbouring countries, Europe and North America.

Consequently, Ethiopia has become one of the main source countries for immigrants and refugees in North America. During the last twenty-five years, over one million Ethiopians have been displaced within the country and an estimated 1.25 million Ethiopians fled to neighboring countries, such as the Sudan, Kenya, Djibouti and Yemen. A relatively smaller proportion of Ethiopians immigrated to Europe and North America (McSpadden & Moussa, 1993).

Between 1974 and September 1998, over 13,000 Ethiopians migrated to Ontario (CIC immigration report, 1974-1996; George & Mwarigha, 1999). This number does not include inland refugee claimants. Moreover, the mobility of Ethiopians from other provinces to Ontario is known to be very high. According to the Ethiopian Association in Toronto, the current Ethiopian population of Toronto is estimated to be 30, 000.

Mental health problems had been major concerns to the Ethiopian community in Toronto. Mental health concerns, particularly suicidal behaviour, were addressed at two community meetings hosted by the Ethiopian Association in Toronto and the Centre for Addiction and Mental Health (Clarke Division). Participants highlighted the urgent need to develop new approaches to the provision of mental health and social services in order to reduce existing barriers to care and strengthen the community’s own capacity to respond to the needs.

A Steering Committee composed of members of the Ethiopian Association and Culture Community and Health Studies (CCHS) was formed to further develop the current research project. A Community Advisory Committee, composed of 13 health and social professionals and community and religious leaders, was also formed to provide overall guidance and direction. The present project, Pathways and Barriers to Mental Health Care for Ethiopians in Toronto emerged as a result of the community meetings and orchestrated supports of the researchers at the CCHS. The project was awarded two years of funding from Canadian Heritage and CERIS. The first phase of the project included the development of a sampling frame and an interview schedule started in July 1998.

Traumatic events such as civil war and forced displacement may result in variety of psychological problems including depression, post-traumatic stress disorder and suicide (Boehnlein & Kinzie, 1995; McSpadden, 1987; Lin, 1986; Stein, 1986). Furthermore, resettlement in a new country often involves a period of significant re-adjustment and stress (Canadian Council on Multicultural Health, 1989; Wood, 1988; Canadian Task Force, 1988).

Areili and Ayche (1993) examined the mental health needs in Ethiopian immigrants in Israel. Their study reported alarming rates of psychopathology, including estimated point prevalence of 37.1% for anxiety disorders, 28.4% for depression, 29.4% for somatization, and 28.8% for sleep disorders. The authors noted that sleep disorders and nightmares persisted in many subjects even after five years of residence in Israel. Suicide rate was about six times higher among Ethiopian immigrants and refugees than the national rate of Israel (Arieli et. al., 1996). Dissociative disorders were also significantly higher among Ethiopian adolescents compared to the national rate (Ratzoni et al, 1993).

To date, we are aware of one US study on single male Ethiopians. Findings of the study indicated a high level of depression and suicide among Ethiopian single male refugees in California, Washington, and Nevada (McSpadden, 1987). No studies have examined these specific issues among Ethiopian newcomers to Canada. Although the differential patterns of migration, community diversity and resettlement processes limit the applicability of the Israeli and American findings to the Canadian context, it seems plausible to assume that Ethiopian immigrants in Toronto may also experience similar adjustment stress and mental distress.

Although migration has been implicated as a mental health risk factor, there is nothing inherent in the process of migration and settlement that jeopardizes mental health (Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees, 1988). Rather it is the contingencies surrounding the resettlement experience that determine the risk of developing a mental health problem. Beiser (1990) suggested that these contingencies include personal strengths, pre- and post-migration stresses, and the availability of family and community support. As illustrated in figure 1, pre-migration stresses such as catastrophic experiences, refugee camp internment and losses and post-migration stresses such as poverty, unemployment and separation from family -- frequent components of the refugee and resettlement process -- jeopardize mental health. Personal resources such as fluency in the host country language, ethnic pride and positive attitudes toward acculturation and social resources, such as family and ethnic community support and a positive reception by the host society, exert beneficial effects on mental health. The model also shows that the resources may buffer the adverse effects of the pre-migration and post-migration stresses. Socio-demographic characteristics such as age, gender, education and ethnicity affect the chances of being exposed to stressful situations as well as the availability of personal and social resources. In fact the model is similar to the stress process model that had been used in other studies of migration and mental health (Kuo, 1984; Kuo and Tsoi, 1986; Noh and Avison, 1996). The present study has also been designed based on the model.

Immigrants use fewer mental health services compared to non-immigrants (Canadian Task Force, 1988; Wells et al. 1989; Wierenga et al, 1996; Fitzgerald, 1996; Harris and Koeler, 1992; Marin et al, 1990). But it is important to note that health service utilization cannot be fully accounted for based on health status and health care needs. As illustrated in the classic model of health care utilization proposed by Andersen and Newman (1973), other broad and interrelated determinants of health service use include many predisposing factors and enabling factors.

Predisposing factors include gender, age and ethnicity (Portes et al., 1992). In addition, health care need is partly determined by cultural perceptions of illness. For example, immigrants from non-traditional source countries may hold beliefs that may differ from those of Western mental health professionals, especially about the causes and treatment of illness (Millet et al., 1996; Edman & Kameoka, 1997; Landrine & Klonoff, 1994), which is likely to influence help-seeking behaviour among immigrants.

Enabling factors including cognitive and financial resources as well as cultural, linguistic and systemic factors mediate the paths between predisposing factors and actual help-seeking outcomes (Aponte & Barnes, 1995; Woodward et al., 1992; Berkanovic & Reeder, 1973; Suchman, 1964; Alvidrez et al., 1996). For example, cultural factors such as differences in explanatory models of beliefs and cultural idioms or modalities used to describe distress (e.g. "somatization") can influence the type of care sought. Personal and family perceptions of mental health problems as highly stigmatizing may discourage immigrants from using the mental health care system (Lin et al., 1978; Lam & Kavangh, 1996). For many minority groups, the family is the primary source of treatment especially for mental health problems and outside help is sought only after all family resources have been exhausted (Flaskerud, 1984). Preferences for alternative remedies and modes of treatment seeking, including religious leaders, elders and folk healers are also well-documented (Snowden, 1996)

Other studies have discussed the negative impact of racism and distrust of majority institutions on entry into the health care system and during the subsequent treatment process (Aponte & Barnes, 1995). These factors would play important roles in determining the extent and pattern of Ethiopian immigrant and refugee use of mental health services and other medical care system.

Part of the problem may stem from a mismatch between what service providers feel they are offering and what potential clients perceive as their needs and the appropriate resources to meet those needs. Presently, health and social services in Canada respond to ethnocultural diversity with three types of delivery models: 1) Multicultural mainstream services that enhance the cultural appropriateness of their services by providing cultural sensitivity training, recruiting bilingual/bicultural workers and engaging in outreach efforts, 2) Parallel services provided by ethnocultural organizations to their own communities, and 3) Multicultural services provided by independent agencies developed to serve the needs of broad ethnic categories rather than specific ethnic or cultural groups.

None of these models is without fault. Barriers such as inadequate interpreter services, Euro-centric orientation, cultural insensitivity, a lack of awareness or utilization of existing social networks, continue to limit the accessibility of mainstream services to ethnic communities. Although parallel service may be more culturally sensitive and effective in terms of their links to the community, they are typically underfunded, staffed by non-professional volunteers and incapable of providing specialized services (e.g. mental health) to community members. Though multicultural services often employ professionals from ethnocultural communities, and recognize and address problems common to several groups (e.g. racism), they often function under the assumption that "immigrants", as diverse as Africans or Southeast Asians, form coherent groups whose needs can be met through one organization (Matsuoka & Sorenson, 1991).

It must also be recognized that immigrants’ perceptions of mental health problems and their attitudes towards the health care system continue to evolve, as newcomers acculturate to Canada (Wells et al, 1989; Atkinson & Gim, 1989). To date, empirical investigators report a deteriorating process. While newly arriving immigrants demonstrate more resilient and healthier attitudes and behaviours, as they assimilate into American and Canadian way of life, they adjust such poor health behaviour as smoking, alcohol and drug abuse, and high-risk sexual behaviours. There has been no study that documents how immigrants and refugees in Toronto modify their help-seeking and health care behaviours as they progress in the acculturation process.

The study design was modeled on a research study conducted by Kirmayer et. al, Sir Mortimer B. at Davis Jewish General Hospital/Division of Social and Transcultural Psychiatry, McGill University. The purpose of the study was to investigate pathways and barriers to mental health care among immigrants from Vietnam, the Philippines and the Caribbean as well as Canadian born Anglophone and Francophone subjects.

The selection of appropriate samples from new immigrant and refugee communities is challenging. Sampling frame for these populations are rarely available. Developing a new sampling frame is costly and time consuming, and requires active initiations and co-operations of many community organisations and their leaders.

Given that there had been no systematic effort within the Ethiopian Community to register all households, we combined a range of procedures to select our sample. First, all the ethnic, religious, political, and social organizations of Ethiopians in Toronto were identified and membership lists were obtained from each of these organizations. This task was relatively easy. Second, a list of Ethiopian specific names was compiled using the Bell Canada telephone directory. One of the limitations of this approach was the existence of Ethiopian Moslems with non-Ethiopian specific names, such as Mohamed. These names were listed and matched with membership lists obtained from the Ethiopian Moslem organizations. The membership lists and the Ethiopian specific names from the telephone directory were combined to form our sampling frame. Using these approaches, 5000 Ethiopian households, or approximately 12,500 individual members were assembled.

Upon completion of the sampling frame, 400 households were selected using a simple random sampling method. One person, who was eighteen years of age or over at the time of the interview, and who resided in Canada for at least one year was, randomly selected from each household. When the individual, who was selected from each household agreed to participate in the survey, he / she would sign a consent form and be interviewed. A total of 342 individuals were interviewed.

2. Development of Interview Schedule

Questionnaire development was informed by the conceptual models of immigration, mental health and service utilisation adopted by this study as well as by the experiences of colleagues in the McGill "Pathways and Barriers to Mental Health Care" study. Several of their measures were modified to reflect the experiences of the Ethiopian community in Toronto and other measures, such as migration history, post-traumatic stress disorder, social support and health risk behaviours, were added based on our models. The final questionnaire included questions on socio-demographic characteristics, migration history, symptoms of distress, recent life events, post-traumatic stress disorder, racial discrimination, service utilization, acculturation, social support services, and health risk behaviour.

The questionnaire was translated into Amharic. It was presented to members of the Community Advisory Committee to ensure that (1) the information collected was pertinent to mental health issues in the Ethiopian community and (2) questions were phrased in a culturally appropriate manner. The questionnaire was also pre-tested with members of the Ethiopian community in both English and Amaharic.

The main outcomes for the study were mental health and health care utilization. The following measures of mental health were included:

• Somatic Symptoms - Diagnostic Interview Schedule Somatization Disorder module (Swartz, Hughes, George et al, 1986) was supplemented with items found to be common in previous studies of immigrant groups (e.g., Westermeyer et al, 1989).

• Recent Life Events - A 14-item scale of stressful life events was developed at the Enquête Santé Québec, was supplemented with events specifically relevant to ethnic minorities.

• Post-Traumatic Stress Disorder - The 19-item scale of Post-Traumatic Stress Disorder, devised by Kinzie, Sack, Angell & Manson (1986) for a study of Cambodian refugee youth, was adopted for use in this study. Diagnostic criteria (DSM-IV) were applied to the responses to the scale items to make diagnosis for PTSD.
• Psychiatric Diagnoses – Two modules – anxiety and depression – of the Composite International Diagnostic Interview (CIDI; Wittchen et al, 1991) was used in the study. The CIDI was developed to identify psychiatric cases according to DSM-III-R criteria in diverse social and cultural setting and population. As such, the CIDI was the most appropriate survey instrument available for diagnostic assessments in large samples of ethnic or cultural groups.

• Health-Risk Behaviours - Use of alcohol, tobacco, drugs (adapted from the OHS, NPHS and Refugee Resettlement Project (RRP) surveys)

Measures of health service utilisation included:

• Health Care Utilization (In-patient and Out-patient) - Adapted from Bland, Orn & Newman (1988), and modified to include community sources of primary and mental health care as well as alternative care, use of traditional medicine, religious counselling.
• Help-Seeking - For each type of mental health problem (including somatic symptoms, depression, and life events), subjects were asked whether they sought for help and where they found help. In addition, if respondents did not receive any help for the reported problems, they were asked to provide reasons for not seeking help.

Explanatory variables were measured using the following instruments/scales.

• Self-Perceived Barriers to Mental Health Care – To measure perceived barriers in receiving professional mental health services, items were drawn from the Ontario Health Survey (Lin and Goering, 1992), and supplemented by items used in a study of Asian Americans (Takeuchi et al., 1992). A few scale items were derived from the information obtained from focus group participants of Vietnamese and Afro-Caribbean participants.

• Socio-Demographic Information – Obtained data included information on age, marital status, migration history, ethnicity, household composition, language, education, employment, religion, and neighbourhood.
• Acculturation – Measures of acculturation included the acculturation mode scales developed by Berry and his colleagues (Berry et al, 1986) and the self-reported Ethnic Behaviour scale (Clement et al, 1991;1992;1993).

• Social Support – Social support was assessed in three dimensions -- advice and information, instrumental and emotional support. On each, sources of the support (relations to subjects and ethnicity of the support providers), as well as subjective ratings of satisfaction with the received support were measures (adapted from Wolchik et al, 1989).

• Perceived Racial Discrimination - Frequency of experiencing different types of discriminatory behaviour and emotional reactions were assessed using the scales and items derived from Noh et al (1999).

Interviewers were recruited in the Ethiopian Community mostly using word-of-mouth. All interviewers received a four-day training program. During the first two-day session, interviewers were provided with general information about the research project and the content of the questionnaire was reviewed. Interviewers were required to perform mock interviews with their family members or friends in order to familiarise themselves with the questionnaire format. During the second 2-day session, they were trained on how to administer the Composite International Diagnostic Interview (CIDI) questionnaire, a diagnostic tool designed to collect data on mental health problems such as, depression and general anxiety disorder. All interviewers were provided with an interviewer's manual to be used as a guide for interviewing. During the data collection process, they received ongoing supervision to correct errors.

Retaining interviewers was a challenge. For various reasons, many of the interviewers left the project before the survey was completed. Second and third waves of recruitment and training of interviewers took place. A total of 19 interviewers were trained during the course of the study.

Coding and data-entry were performed by a research co-ordinator and a research assistant. Random spot-checking of completed questionnaires was done before the data was entered. Inconsistencies and missing values were brought to the attention of the Research Coordinator and clarified with the interviewers, when necessary. Data were entered and cleaned using SPSS software. Variable frequencies were used to identify outliers.

The CIDI 2.1 Computer Diagnostic Package was used to enter, verify and score the mental health data. Using a transfer program, the raw data was converted into an SPSS readable form. A diagnostic scoring algorithm allowed us to compute diagnoses according to the 10^th revision of the international Classification of Diseases (WHO, 1990) and DSMIV.

Data analysis was performed using SPSS software. Both univariate and multivariate logistic regression analyses were employed

1. Demographic Characteristics

Data describing demographic characteristics of the sample are summarized in Table 1.1. Approximately 60% of the study sample were males. Age of the respondents ranged from18 to 59 years with a mean of 35.3 years. However, a majority (56.8%) was young adults aged between 30 and 39 years. Most of the respondents (55.3%) were married or living with someone as though married. Approximately 9% had a non-Ethiopian partner or spouse. The average household size was 2.5 persons. Nearly one quarter (23.4%) of respondents had completed university, while less than 5% had not completed high school. The proportion of males with higher education was significantly higher than females (32.5% and 10.1%, respectively).

The majority (86.5%) of the respondents claimed Amharic as their first language, followed by Tigrinia (9.7%) and other languages including Guragigna and Aderigna. In terms of their religious affiliation, the majority of study respondents were members of the Ethiopian Orthodox Church (68%) and 'Protestant' (22%), followed by 'Roman Catholic' (5.3%). Only a small proportion was 'Moslem' (2.3%). About 74% of the respondents stated that religion was very important to them. A notable 16.8% of study respondents reported a change in religious affiliation since their arrival in Canada. The changes were more frequent among female respondents (21.2%) compared to male respondents (13.8). These figures reflect the visible evangelical (protestant) movement in immigrants and refugees from Africa.

Table 1.2 compares socio-demographic characteristics of the study sample with 1996 census information. Two reference groups are provided: Ethiopian immigrants in Toronto and the Toronto population as a whole. Among the factors that limited the comparability of these populations were differences in the year of data collection and the populations' age distributions. While the current study collected data on adults (18 years or older) in 1999, the census data were based on the total population during the year of 1996. Age specific data on the 1996 census have not been made available to the public. It was also observed that 9.2% of the 1996 census population of Ethiopian immigrants to Toronto was comprised of individuals with a mother tongue of English, French, Italian, Arabic and Hindi, who were, most likely, non-Ethiopian.

Differences could be observed in educational attainment between the two groups. A larger proportion of our sample had completed secondary school and had a university degree compared to the census Ethiopian population. This was possibly related to the fact that many Ethiopians attended higher education institutions since their arrival in Canada. That a greater proportion of Ethiopians was not in the labour force in 1996 census compared to our sample may reflect the higher proportions of Ethiopians in school at the time of census.

2. Migration and Pre-Migration Factors

Table 2 presents information on migration history of study respondents. The majority of study respondents lived in big cities prior to coming to Canada (87.1%). When asked to identify their primary motives for migrating to Canada, "political reasons" was most often identified by males (64.5%) followed by "for a better life" (29%). Among females, the most frequent response was "for a better life" (57.1%), closely followed by "political reasons" (37.6%). Overall, the two reasons were the choices for 97% of the total sample. The average age of respondents was 22.9 years when they left Ethiopia and 26.1 years when they arrived in Canada. More than three quarters of respondents (77.8%) had lived in neighbouring countries (primarily Kenya, Sudan and Djibouti) or in European countries (primarily Italy and Greece) for an average duration of 4.1 years before coming to Canada. Approximately 11% had spent time in refugee camps for an average duration of 25 months.

It appears that the exile and migration processes had been more distressful for men than for women. As shown in Table 2, more male respondents compared to female respondents reported the experiences of staying in a refugee camp, for a significantly extended period (31.3 months for men compared to 7.1 months for women), and had traumatic experiences (27.1% of men compared to 12.4% of women).

Most respondents (71.6%) came to Canada alone. Approximately 71% of respondents had family members or friends already living in Canada when they arrived. Of these, nearly 80% responded that they had family members or friends who could have helped them when they arrived. On average, females were significantly younger than males when they arrived in Canada (24.5 years and 27.2 years, respectively). The immigration status of the majority of respondents upon arrival in Canada was refugee (37.4%) or refugee claimant (7.6%). However, by 1999-2000, respondents had been living in Canada for an average of 9 years, and most of them (80.7%) had obtained Canadian citizenship.

3. Post-Migration Experiences

3.1 Education and Employment

Table 3.1 presents information on education achievement and employment. Over 89% of the total study respondents had attended school in Canada, the majority (49.3%) for more than 2 years. Approximately 23% of respondents were still in school at the time of the survey.

The majority of study respondents were currently working either full-time or part-time (78.8%). The proportion of males who were currently working (89.0%) was higher than the proportion of females (64.0%). Among the respondents who were employed, 83.9% were working full-time. However the mean number of months worked in the last year was only 9.4, suggesting a high turnover rate. Moreover, more than one half (54.3%) did not feel that they were working at a job that reflected their ability, i.e. underemployed. Among the respondents who were not currently working, the majority was students, and only 5.6% were currently unemployed and looking for a job. Most respondents (75.4%) had worked prior to coming to Canada.

3.2 Life Events

Respondents were asked questions about stressful life events experienced in the 12 months prior to the interview. Results are presented in Table 3.2. Almost 50% of the study respondents reported no life events during the last 12 months, 36.1% had experienced one to two events, and 14.1% had experienced three or more events. The most frequently experienced life event by males and females was illness or death in the family' (30.3%, 20.4%, respectively). This was followed in frequency by 'trouble with housing', 'difficulties at school/work', 'financial problems' and 'concerns with children'. Males were more likely to experience 'trouble with police', although the proportion of the population reporting this was low (7.9%). Overall, 42.7% of study respondents had sought help from individuals or agencies for recent life events. The most commonly reported reason for not seeking help was because it was thought to be unnecessary.

3.3 Discrimination

Personal experiences of being fairly or unfairly treated in Canada have significant implications. In our survey we asked respondents whether and how often they felt they were unfairly treated because of their racial background. Table 3.3 presents information on the proportion of study respondents who experienced racial discrimination in this way in Canada. Almost two thirds of respondents had experienced discrimination and this proportion was significantly higher among males compared to females (74.9% and 48.9%, respectively).

The most common forms of discrimination experienced by males and females were examined. Among males, 60.6% had experienced 'being looked down on', 60.7% had experienced 'resentment', 50.3% 'being treated unfairly', 53.2% 'rudeness', 35.5% 'insulting remarks', 25.3% 'threats' and 3% 'physical abuse'. Females also reported these forms of discrimination, though less frequently. Overall 31.6% of study respondents reported more than 4 forms of discrimination. The proportion of males who experienced racial discrimination was higher compared to that of female respondents.

A scale was also created to measure discrimination based on 7 items with coded frequencies never (0), sometimes (1) and often (2). The scale ranged 0-14. The overall mean discrimination score was 2.5. Males had higher mean discrimination score compared to females (3.0 and 1.6, respectively).

The common places where discrimination was experienced were: on the job (44.3%), in service areas (30.7%), in the general community (30.5%), from the school system (25.6%), from government agencies or departments (18.7%), through the media (16.5%), from the police (15.5%), and on private property (2.9%).

4. Personal Resources

Personal differences influence the ability to cope with stress. Familiarity with and the ability to use an official language, ethnic interaction, and ethnic identification are found to be resourceful in promoting and protecting the mental health of immigrants and refugees. In this study, personal resources include English language proficiency, acculturation level and ethnic pride.

As illustrated in Table 4.1, 80% of respondents reported that they spoke an Ethiopian language only, 10.5% spoke an Ethiopian language as well as English, 7.4% spoke only English and 2.1% spoke other languages most often at home. The language proficiency scale was calculated based on subjective ratings of respondents' abilities to speak, read and write Amharic (as a second language) and English. The mean score was 11.4 for Amharic and 9.8 for English languages.

2. Acculturation and Identity

Berry, Trimble & Olmedo (1986) and Berry (1997) suggested a useful typology of acculturation. Based on this work, four questions were adapted to measure acculturation status; "How much do you consider your way of life in Canada as being Ethiopian?" "How much do you consider your way of life in Canada as being Canadian". Respondents were then classified "separated", "assimilated", "integrated" or "marginalized" depending on their response patterns. The distribution of Ethiopian respondents by acculturation status is presented in Table 4.1. The majority of respondents were categorized as assimilated (37.0%). A higher proportion of males was assimilated compared to females. On the other hand, females were more likely to be separated; possibly they arrived in Canada later and less likely to be employed.

Respondents were also asked questions about their ethnic identity and how they felt about being in Canada. Ethnic identity was explored using the Multi-group Ethnic Identity Measure (Phinney 1992). The MEIM was designed to assess different dimensions of identity: affirmation and belonging and ethnic behaviours. Ethiopian and Canadian identity scales were constructed using the responses to the identity question items presented in Table 4.2. Factor analysis was used to identify the critical items from the MEIM measure to be included in the scales. Reliability testing was then used to select the best items. The Ethiopian identity scale consisted of 4 items: extent of feeling Ethiopian, sense of belonging to Ethiopia, pride, feel good about Ethiopian background) (Alpha = .77). The Canadian identity scale consisted of the corresponding 4 Canadian identity items (Alpha = .84). For both male and female respondents, self-identification with Ethiopian heritage was significantly greater than the level of Canadian identity (Table 4.1).

1. Ethnic Enclave

Information on neighbourhood composition is presented in Table 5. It is apparent that the majority of study respondents lived in neighbourhoods where several (36.6%) or many (26.3%) Ethiopian households were present. Furthermore, most of the study respondents reported that they knew their neighbours very well (46%) or well (35.8%).

2. Social Support

A set of descriptive results on informational, instrumental and emotional supports are presented in Table 5. On average, Ethiopian respondents felt that they could rely on two persons for social support, if they were in need. Overall, the most common sources of social support were friends (66.2%), religious leaders (10%), family members (8.8%), family doctors (5.4%), the Ethiopian Association (2.4%) and 'others' including counsellors, neighbours, and non-Ethiopian organizations (7.3%). On a scale of 1 to 10, the average level of satisfaction with social support was 8.5. The lowest satisfaction with social support was reported for instrumental support.

6.1 Somatic Symptoms

Respondents were asked about their experiences during the last 12 months. The checklist included 16 physical symptoms (Table 6.1). The most frequently reported symptoms were backaches, headaches and forgetfulness. Twenty percent or more of the sample responded positively to these symptoms. These symptoms were more prevalent among females (22.3% - 29.5%) than among male subjects (16.7% - 21.2%). Following these symptoms, abdominal pain, leg or arm pain, insomnia, periods of weakness, and fatigue were experienced by more than 10 percent of the sample. In fact somatic symptoms were prevalent among the Ethiopian immigrants. More than 63% of the sample reported as having experienced at least one of the 16 symptoms. Only 37 percent had no somatic symptoms. Although most had one or two symptoms experienced (42.4%), more than one in five (20.8%) reported four or more symptoms. It is interesting to note that 5.3% of the sample said that they had physical symptoms for which doctors could not provide a specific diagnosis.

6.2  Depression

Approximately 10% of the study population met diagnostic criteria for depressive disorder (Table 6.2). The lifetime prevalence of depression was slightly higher in males compared to females (10.4% and 8.8%, respectively). By comparison, the lifetime prevalence of depression among the population 18-50 years according to data from the Ontario Mental Health Supplement study (1990) was 8.9% (Elizabeth Lin, Center for Addiction and Mental Health, personal communication). In 1996/7 the prevalence of depression in the previous 12 months among the Ontario population aged twelve and older (measured using a subset of questions from the CIDI) was 4%.

Twenty-two persons developed depression since they left Ethiopia, indicating a post-migration incidence rate of 5.9%.

Approximately 3% of the study population met the criteria for a diagnosis of generalized anxiety disorder (Table 6.2). There was no statistically significant difference in the lifetime prevalence of general anxiety disorder between males and females. In 1990, the lifetime prevalence of general anxiety disorder in the Ontario population (18-50 years) was 2.0% (Elizabeth Lin, Center for Addiction and Mental Health, personal communication).

Many immigrants and refugees leave their country carrying with them memories of torture, brutal war, escape, and concentration camp experiences. Individuals with these types of horrible experiences are at increased risk of developing post-traumatic stress disorder after they get asylum in resettlement countries. Information of the prevalence of PTSD among Ethiopian immigrants in Toronto is presented in Tables 6.2. Approximately 6% of the respondents in our sample met diagnostic criteria for PTSD. This proportion was higher among males compared to females (6.9% and 4.3%, respectively).

Study respondents were asked a series of questions about their alcohol consumption. Results are presented in Table7.1. The majority of study respondents had taken a drink of alcohol in their lifetime (82.3%) and this proportion was significantly higher for males than for females. Among those who had taken a drink of alcohol in their lifetime, 85.6% had taken drink within the past 12 months. The majority (68%) drank once or twice a month or less.

Drinking enough at one time to get drunk is considered to be a symptom of an alcohol problem. In our sample, 14.4% of respondents reported that they drank enough to get drunk in the last 12 months. This proportion was significantly higher among males (18.4%) compared to females (5.5%).

The prevalence of current drinkers was higher in our sample compared to the 1996/1997 Health Survey estimates for the Ontario population (85.6% and 73%, respectively). However the proportion of frequent drinkers (i.e., individuals who reported drinking every day or 4-6 times a week) was lower in the study sample compared to the Ontario population.

Respondents were also asked whether their drinking behaviour had changed since leaving their country of birth. Out of those that drank, one-third (33.9%) reported that they started drinking more since they had left Ethiopia or arrived at Canada. Male respondents were more likely than female respondents to report increased drinking (36% and 29.2%, respectively).

The most common reason provided for drinking was 'to be sociable' (76%) followed by 'to feel good’ or ‘get into the party mood' (49.6%). Approximately 8% reported that they drank to overcome stress or when they feel sad, lonely or depressed.

Data on cigarette smoking are presented in Table 7.2. Approximately one fifth of study respondents (20.9%) reported regular or occasional smoking. Male respondents were more likely to be smokers compared to females (28.1% and 10.3%, respectively). Among current smokers, males were significantly more likely to begin smoking daily at a younger age and to smoke more cigarettes per day than females. The overall prevalence of smoking (regular and occasional) was approximately the same as the Toronto population as a whole in 1996 (19%). In Toronto, the proportion of males who smoked (21%) was also higher than females (17%).

Information on drug use is presented in Table 7.3. Overall, 38% of the study population had tried drugs in their lifetime. This proportion was significantly higher for males compared to females (53% and 15.6%, respectively). The most frequently tried drug was Chat (28%).

A high proportion of the study population (68.9%) sought help for somatic symptoms (Table 8.1). The major sources of help were family physicians (75%). A small proportion of the sample had other sources of help including, physiotherapists (12.5%), general hospital (7.6%) and other sources (4.9%), for example dentists and chiropractors. Female respondents were more likely to seek help from physiotherapists compared to males.

Table 8.1 also presents information on the main reasons that study respondents did not seek help for somatic symptoms. The most common reasons were, 'the problem was not serious enough' (49.2%), followed by 'no solution' (19.7%), 'could handle on own' (18%). Respondents were also asked whether they would consult any one for help if they develop one or more somatic symptoms. The vast majority of them (96.8%) said they would consult a professional.

Study respondents were asked questions about health care service utilization during the last 12 months preceding the interview. As shown in Table 8.2, 85.5% of Ethiopians used at least one form of medical services during the 12 months immediately preceding the interviews. Undoubtedly, most had visits to family physicians. However, more than 16 percent of the sample made a visit to a hospital emergency room, and 5 percent were hospitalized during the last twelve months. Overall, women used the services more frequently than men.

When the respondents were asked specifically about help seeking for emotional problems, or stressful life events, 4.9 percent of men and 5.9 percent of women said they saw at least one medical professional, mostly family physicians and psychiatrists, for such problems during the past 12 months. It is interesting to note that about an equal proportion of men (4.4%) and an even greater proportion of women (12.9%) went to non-medical help for their emotional and other forms of life problems. These non-medical help providers include the Ethiopian community councillors and religious leaders. A small number (n= 4) went to traditional healers for help. Together, the data suggested that 8.4% of male and 16.5% of female respondents felt a need for professional help, medical or non-medical, for their emotional problems.

These utilization rates suggest that Ethiopians in Toronto tend to use more medical services compared to the members of other immigrant and refugee communities in Montreal including, Caribbean, Vietnamese, and Filipinos. The McGill study, "Pathways and Barriers to Care" used the same measurement procedures as the present study. Overall rate of service use in the Montreal sample was 78.5%, compared to the rate of 85.5% among the Toronto Ethiopian sample. The help seeking rates for emotional problems were 3.6% and 5.3% for the Montreal and Toronto study samples, respectively.

This section expands the perspectives on depression and presents an overview of the results from initial analyses of bivariate associations of depression with three sets of psychosocial factors - demographic characteristics, pre-migration backgrounds, and post-migration experiences.

In table 9.1, crude prevalence rates of depression and confidence intervals are reported across demographic groupings. Although the results showed a greater prevalence of depression among young adults (18-24) than older adults, the effect size was not large enough to reach statistical significance. Neither have we found statistical evidence for the effects of widely recognized risk factors including female gender, currently unmarried status, lower educational level and unemployment. Given that these are well-established epidemiological risk factors for depression, the current findings are unusual.

There seem to be two explanations for the unusual findings. First, statistical significance test, (i.e., power) is determined by sample size and the levels of measurements. The data shown in Table 9.1 were based on clinical diagnosis, and as expected, only a few cases (n=33) were classified as depressed. The current study has a limited power to detect small or moderate effects. With this in mind, we may observe in Table 9.1 that the trends suggest higher prevalence of depression among younger adults, currently unmarried, less well educated and currently not employed. These trends are consistent with the well-standardized epidemiological patterns of social risks of depression.

Second, the exceptional life contexts and personal experiences of immigrants and refugees often alter the pattern of social risks for mental health. For example, as presented earlier in this report, male subjects had more dangerous and traumatic experiences during exile (see the results section 2, Migration and Pre-migration Factors). In addition, such structural barriers as denial of foreign-trained credentials (academic and professional) and previous work and professional experiences place men and women at risk for depression during the initial settlement period (Hou and Beiser, in press). Among Asian immigrant women, Noh et al (1992) found increased risk for depression amidst the employed, compared to the unemployed. In this study we also found (Table 9.1) elevated rates of depression among men and those who completed university education. Although speculative, the data seem to suggest relatively greater exposure to stress and perceived deprivations among university educated men.

9.2. Pre-migration Correlates of Depression

As shown in Table 9.2, age at emigration did not seem to be related to depression. However, the context of emigration appeared, strongly and significantly, to influence the mental health of immigrants and refugees. The results indicated that the experiences of living in refugee camps and trauma during migration increased the crude rates of depression by 300 % to 400%. Data on the motive for emigration also showed a trend that suggested the salience of political situations in Ethiopia and surrounding countries. Those who said they left Ethiopia for political reasons showed a relatively higher rate of depression, although the result was not statistically significant. The current data seem to suggest the enduring psychological effects of stressful and traumatic events experienced during the exile and migration.

9.3. Post-migration Correlates of Depression

As shown in table 9.3, the recent experience of stressful life events was a significant risk factor for depression. More specifically, those who reported more than two life events were three times more likely to be depressed than those who reported no major life event. However, perceived discrimination (racial or ethnic) was not significantly related to depression, although depressed subjects reported a slightly higher level of discrimination.

From the information shown in Table 9.4, we find significant effects of emotional social support from the neighborhood. Those who were happy with emotional social support were less likely to be depressed, whereas other measures of social support, either availability or satisfaction, showed little effects. Except for emotional support, it seems as if the social resources are extremely limited, and perhaps, the same resources (e.g., community organizations and services) are used by both the depressed and non-depressed; both are not satisfied with the amount and quality of the supports. Although admittedly speculative, the data may suggest an urgent need for establishing more systematic and professional services.

Depression was substantially more prevalent among those who lived in neighborhoods where Ethiopian immigrants and refugees were evident. Given this finding, it is surprising to observe the lack of mental health relevance of all measures of ethnic identity, both Ethiopian and Canadian. The data may suggest that the residential neighborhoods can be chosen independent of a person’s identification with Ethiopian culture and people. Financial conditions may play the most critical role in neighborhood selection among Ethiopians.

Our ethnographic data were convincing in showing the critical role of religion in the Toronto Ethiopian community. Traditionally, a large majority of Ethiopians are affiliated with Ethiopian Orthodox Church and Muslim. However, there has been a considerable conversion to evangelical (protestant) churches in Toronto. Although we found no statistical data suggesting the mental health effect of religion, depression seemed to be more prevalent among those who believe that religion is not important. But, this trend was not statistically significant.

The preceding section reported data on prevalence of depression and relevant risk factors and potential protective factors. Such information as prevalence and cross-sectional correlates is essential for planning and developing intervention programs. However, scientific inquiries of etiologic processes and personal and social consequences of mental illness require more information. For one, the analysis requires data on the incidents of new cases.

Of the 33 depression cases, 20 were new cases, i.e., post-migration onset of the illness. The figure represents a crude rate of 5.9% of incidence. As shown in Tale 10.1, the incidence of depression was not different for men and women. Age was not related to the incidents. As presented in table 10.2, two etiologic factors were pre-migration experiences of refugee camp and trauma. Those refugees who were exposed to these risk situations were about three times more likely to develop post-migration depression compared to those who had not experienced the risk situations. The primary motive for migration has also significant influence on the occurrence of depression. Other risk factors, including age at emigration, and family separations were not significantly related to the onset of the illness.

Socio-demographic characteristics were not related to the incidence of depression (Table 10.1). However, the experience of stressful life events was a significant risk factor. Those who experienced more stressful life events had major depression compared with those who did not experience any incidence of recent life events. For example, the incidence rate of depression was about four times higher among those who reported more than 2 life events compared with those who reported no incidence of life events (Table 10.3a).

In Table 10.3b, we find life events and chronic strains that are strongly related to the onset of depression. Financial strain, marital problems, difficulties in school, family illness or death, trouble with policy, arguments in the family, and language problems, were the most important life problems that were related to depression. The cases of arguments in the family and language difficulty were not statistically significant.

Financial strain, school difficulties, and language problems are typical settlement challenges for all immigrants and refugees. Data also reflect the problems that black immigrants have with police (Wortley, 2000). Presumably, these settlement difficulties are the major contributors to marital and family conflicts and illness in the family.

While the pre-migration experiences of trauma and refugee camps and settlement problems contributed to the development of depression, as reported in Table 10.4, social support and ethnic identity were related to decreased rates of depression. The Ethiopian identity score was significantly lower among the depressed respondents compared to non-depressed respondents. The mean Canadian identity score was also lower among the depressed compared to the non-depressed. However, this difference was not statistically significant.

There was no statistical significance difference in the incidence of depression between respondents who accepted the Canadian way of life (assimilation or integration) and those who did not (marginalization or separation). Neither did we find significant difference in the amount of support between the depressed and non-depressed groups. However, satisfactions with emotional and instrumental supports were important for the prevention of depression. The mean score of emotional support satisfaction was significantly lower in the depressed group compared to the non-depressed group. Rate of depression also varied by neighbourhood composition. Depression rates were slightly higher among individuals who had several or many Ethiopian neighbours (7.7%) compared to individuals who had no or few Ethiopian neighbours (3.3%).

Theoretical considerations and the results of the univariate logistic analyses were used to identify critical factors associated with the incidence of depression. Twelve variables were identified for which the 95% confidence intervals for the odds ratios did not include 1 or barely do. These included: recent life events, pre-migration trauma, primary motives for migration, emotional support satisfaction, refugee camp experience, Ethiopian identity, racial discrimination, household size, Ethiopian neighbourhood, and current employment status (Table 10.5).

Both backward and forward stepwise logistic regression method was employed to select the most important predictors of depression. The results of the stepwise selection suggested that only recent life events, motives for migration, Ethiopian identity score and the interaction of pre-migration trauma and racial discrimination had statistically significant importance in the model (Table 10.6). The contribution of the remaining variables added very little to the regression model.

Therefore, our final regression model included recent life events, motives for migration, Ethiopian identity, pre-migration traumatic experience, racial discrimination and an interaction of pre-migration traumatic experience and discrimination (Table 10.7). As this final regression model suggested, recent life events, primary motives for migration, and Ethiopian identity score were strong predictors of depression. Pre-migration traumatic experiences and racial discrimination did not significantly predict the occurrence of depression. However, their interaction significantly influenced the incidence of depression.

The preceding sections examined personal and social correlates of depression, lifetime and post-migration incidents. In both the prevalence and incidence of depression, the most significant risk factors were pre-migration factors – experiences of refugee camps and trauma. Significant post-migration correlates were stressful life events and the lack of emotional social support. The onset of major depression disorder was also related to instrumental support and ethnic identity. In addition, the prevalence rates of depression were elevated, although statistically not significant, among young adults (younger than 24), currently not working and not married, and non-believers. Further analyses on stressors revealed the nature of life strains.

These findings suggest urgent needs for developing mental health interventions for those experienced refugee camps and trauma. Settlement programs focusing on employment to ease financial strain is likely to help the mental health of Ethiopian immigrants and refugees. While unemployment was directly related to an increase in depression, presumably, it was also related to the increase of family and marital conflicts, lower income, and life events related to inadequate income.

Living alone or without a partner also constituted a risk for depression. Non-believers were also at risk. But, having social support for emotional needs seemed to protect mental health. Church based interventions for single persons may promote informal networks to meet the needs for emotional support.

11.1 Mental health care services.

It should be recalled that respondents were asked whether, and how many times, they sought medical or non-medical professional help for their emotional problems during the preceding year. Eighteen (5.3%) made at least one visit to medical services for mental health or emotional problems (Table 8.2). Since the number seeking help for emotional problems was small, detailed statistical analyses were not performed, rather patterns were observed. As can be seen in Table 11.1, demographic characteristics were not statistically related to the help-seeking behavior as measured. However, there were trends suggesting that younger and unmarried respondents may use services more frequently than their counter parts. The members of the Ethiopian Orthodox Church and those who reported that their religion is important also use mental health services more frequently than their counterparts.

Enabling factors associated with health service utilization for emotional problems are presented in Table 11.2a and 11.2b. It is apparent that individuals who were unemployed or working part-time were more likely to use health services than employed and full-time working counterparts; but these observations may be confounded with need i.e., these groups experienced more mental health problems.

Individuals who attended school in Canada and who were still in school used health services for emotional problems more frequently than those who did not attend school in Canada. The high service use among individuals attending school in Canada and still in school may be partially explained by the fact that this group was more likely to have knowledge about the mental health service system. Respondents from rural areas and individuals categorized as 'marginal' used mental health services more frequently compared to their urban and non-marginalized counterparts. This difference might be explained by the high prevalence of somatic symptoms reported among the rural and "marginalized" groups.

It was also apparent that respondents who had family or close friends in Canada, especially who could help, were more likely to seek help. This may be because they were more knowledgeable about resources to turn to for help.

Information on factors that describe ‘need’, which are associated with mental health service utilization are presented in Table 11.3. Individuals seeking help were more likely to have several somatic symptoms, to have experienced several stressful life events and to have a diagnosis of depression or anxiety.

Table 11.4 through 11.6 describes factors associated with the use of any health services. Generally speaking, service use was higher for females and individuals who were unemployed. Again there was a strong relationship between need and service use. Individuals who experienced somatic symptoms, depression, anxiety and pre-migration trauma were more likely to use health services compared to their counterparts.

One primary objective of this study was to determine the extent of mental health problems as experienced by Ethiopian immigrants in Toronto. The results suggested an estimated lifetime prevalence of depression of 9.8% (95% CI: 6.6%- 13.0%). This figure was not significantly different from the provincial rate. But, it was about three times higher than the rate estimated for southern Ethiopia (Awas, Kebede and Alem, 1999). It is difficult to determine the factors that explain such increase in depression among immigrants. However, given that the depression was significantly inflated among those Ethiopians who were previously exposed to refugee camps and traumatic stressful events, we may speculate that the increased depression in immigrants is unlikely to be independent of migratory traumas and settlement stressors.

Furthermore, depression was more prevalent among women than men (Awas, Kebede and Alem, 1999). The opposite pattern was observed in Toronto, a pattern that was also reported in other immigrant/refugee populations. As suggested in qualitative research, the re-establishment of family stability following migration is of prime importance to Ethiopian women. Compared to men, Ethiopian women may be more willing to accept a drop in professional or social status, set aside educational goals, and assume dual responsibilities at work and at home. On the other hand, men were less willing to accept changes in occupational roles and felt more threatened by changing gender roles (McSpadden & Moussa, 1993). Moreover, males in our sample were more likely to have experienced conditions associated with depression (i.e., migration traumatic experiences and refugee camp internment) than females. Thus, Ethiopian men compared to women are more likely to be exposed to adverse mental health consequences of migration and settlement stresses.

Lifetime prevalence and post-migration incidence of depression persistently demonstrated the importance of pre-migration trauma and refugee camp experiences. These traumatic experiences appeared to have long-lasting mental health effects. There is an urgent need for intervention programs that focus on the effect of torture, camps and other trauma. Post migration settlement stresses were also exerting severe adverse effects on the mental health of Ethiopian immigrants and refugees.

Aside from treating the post trauma wounds, social and health care programs need to focus on current financial strains, family conflict and systemic discriminations. Experienced stressful life events, including financial problems, family problems, difficulty at school, illness or death in the family, exerted a strong effect on depressive disorder. Not having any clear motive for migration was also significant risk factor. Again resorting in our qualitative data, many Ethiopians came to Canada simply just to follow other peoples, or because they heard that Canada is a good place to live, they had a chance to come, or they did not have much to do at home. This group of people lacks the resilience that is often demonstrated by the self-selected immigrants who met the admission criteria of Canadian immigration. Further analyses will explore the data to determine the characteristics of the immigrants of non-specific motives.

Having a strong Ethiopian identity was a protective factor. It is possible that individuals who had high identity scores had access to more personal and social resources than individuals with lower scores. The majority of individuals in the study population belonged to the Ethiopian Orthodox Church, a well-established religious institution that serves both spiritual and social needs for many Ethiopians in Toronto. The prevalence of depression was significantly associated with advice and satisfaction and emotional social support satisfactions. The data also showed protective effect of instrumental social support satisfaction, although statistically not significant.

The findings on the incidence of post-migration depression closely replicated those on lifetime prevalence of depression. Recent stressful life events, primary motives for migration, ethnic identity and the interaction of pre-migration trauma and racial discrimination have a statistically significant effect on the incidence of depression. The incidence of depression was also related to experiences of refugee camp and emotional support satisfaction.

Study findings have implications for the delivery of mental health services to the Ethiopian community in Toronto. It is important for mainstream health professionals to recognize that the effects of pre-migration experiences on mental health may persist in Canada. Settlement services must develop programs to address stressful life events experienced by many Ethiopian newcomers. There may be a need to further develop community-based programs that re-enforce Ethiopian identity and a sense of belonging.

It is alarming that a large majority of Ethiopians did not seek professional help for emotional problems. Some of the reasons why help was not sought were because help was thought to be unnecessary, or there was no solution for the problem and privacy. However, the utilization patterns in this sample were similar to those of non-immigrant residents of Montreal, and the rate of service use was higher than those found among immigrant groups in Montreal. This may be explained by the facts that Ethiopians were better educated, have had previous contact with Western health care, and have had access to Ethiopian physicians in Toronto. Nonetheless, the rate of service use suggests a need for more effective health education programs to be developed in the community.

Aside from scientific findings, the present research project has made substantial contributions toward mobilization of the community. To present a few specific examples, we list below the activities that had been initiated within the Ethiopian community and mainstream institutions as a result of our project:

• A quarterly project newsletter published in English and Amharic (Aemero) is being used to promote awareness and resolution of mental health issues within the Ethiopian community. It includes project updates, articles on mental health and local community events and items of interest to the community (e.g., summer activities). Six issues have been published to date and approximately 800 copies are distributed at Ethiopian churches, restaurants and doctors offices.
• The project has been featured at several Ethiopian community events e.g. Ethiopian-Canadian Day, International Soccer competition.
• The Ethiopian Association in Toronto established a Task Force to assess mental health problems and care needs within the community and to develop provisional strategies to respond to the identified needs.
• The Ethiopian community was featured in an educational series at the CAMH initiated "Meet Our Community Partners" to acquaint CAMH staff with perceptions of mental health and addictions, beliefs about treatment and access to care for Ethiopian immigrants in Toronto.
• The Ethiopian community is participating in a project of the Ethnoracial Initiative of CAMH and Mount Sinai Hospital to improve the accessibility of mental health and addiction service to ethnoracial communities in Toronto.

This study is the first community survey of the mental health needs of Ethiopian immigrants in North America. The particular merit of this study is that it has addressed the methodological issues of obtaining a representative sample of a minority population, (new immigrants and refugees), which is key in making valid inference about the target population.

In this study, it was observed that many Ethiopian immigrants in Toronto experienced pre-migration trauma, refugee camps and post-migration stressful events, which are known risk factors for the development of depressive disorder. However, the prevalence of depression among Ethiopian immigrants in Toronto was only slightly higher compared to the average rate of the Ontario population (9.8% versus 8.3%). This may suggest that many of the Ethiopian immigrants have been able to cope with the stresses or they might have had a good social support network in the Ethiopian community.

The lifetime prevalence of depression among Ethiopian immigrants in Toronto was found to be 3 times higher than the lifetime prevalence of depression in Ethiopia. This difference could be due to the fact that many Ethiopians in Toronto had been exposed to numerous risk factors for depression during their migration and resettlement process.

The results of this study also suggested a higher prevalence rate of depression among younger adults, males, currently unmarried individuals, the unemployed, and those with a low level of education. Although statistically significant evidence was lacking, due to a small study size, the findings confirm well-known epidemiological risk factors for depression. The higher depression rate among males compared to females, which was the opposite pattern in Ethiopia, may suggest that males were less likely to accept changes in occupational roles and felt more threatened by changing gender roles. Furthermore, in our sample males were more likely to have experienced factors associated with depression than females.

The results of this study also showed a statistically significant relationship between the prevalence of depression and the well-known risk factors, including experiences of living in refugee camps, motive of migration, and post migration stressful events. These finding suggest urgent needs for developing mental health intervention programs, particularly for those who experienced pre-migration trauma, refugee camps, and stressful life events.

Regarding healthcare utilization, the findings indicate that Ethiopian immigrants in Toronto tend to use more medical care services for emotional and non-emotional problems compared to the immigrant and refugee communities in Montreal, including Caribbean, Vietnamese and Filipinos. However, among our study respondents, who met CIDI depression criteria, only 19.4% sought medical services for emotional problems. Therefore, there is a need to develop programs that may help these individuals in accessing mental health care services.

Finally, we hope that this study will provide valuable information that helps to understand the extent and nature of the mental health care needs of Ethiopians in Toronto. It may also help in planning social and health care services for the individuals in the Ethiopian community, who were exposed to traumatic life experiences, as well as those immigrants and refugees in other communities with similar experiences.