Somali Study- Kroll

Soc Psychiat Epidemiol DOI 10.1007/s00127-010-0216-0

ORIGINAL PAPER

Psychoses, PTSD, and depression in Somali refugees in Minnesota Jerome Kroll • Ahmed Ismail Yusuf Koji Fujiwara

•

Received: 7 July 2009 / Accepted: 11 March 2010 Ó Springer-Verlag 2010

Abstract Introduction Initial clinical observation of Somali patients seen at a busy inner-city community clinic (CUHCC) suggested that, in addition to the expected pictures of Posttraumatic Stress Disorder (PTSD) and depression previously seen in the clinic’s Southeast Asian refugee population from 1980 to 2000, there was an unusually high number of young Somali men presenting with acute psychotic disturbances. Objectives The aim of this study of health care utilization of Somali refugees (N = 600) seen in the mental health unit of the clinic from 2001 to 2009 was to investigate the major patterns of psychiatric disorders in this outpatient population and compare these findings with a cohort of non-Somali patients (N = 3,009) seen at the same outpatient clinic during the years 2007–2009. If the results supported the initial clinical observations that the rate of psychoses was higher among young Somali men than nonSomali men attending CUHCC clinic, then several areas of further research would recommend itself. First, since this study was not a study of prevalence of mental illness in the Somali community, the next step would be to undertake a study of community prevalence of mental illness among Paper read at the 24th ISTSS Meeting, Chicago, IL, 13–15 November 2008. J. Kroll (&)
A. I. Yusuf Community-University Health Care Clinic, University of Minnesota Medical School, 2001 Bloomington Avenue South, Minneapolis, MN 55404, USA e-mail: [email protected] K. Fujiwara Duluth Medical Research Institute, University of Minnesota Medical School, Duluth, MN, USA

different age and gender cohorts. Second, further research should look into likely causative and contributory risk factors to explain the development of psychoses among Somali young men. Methods Somali and non-Somali patients were diagnosed according to DSM-IV-R criteria. Main outcome measures (diagnoses, age cohort, sex) were analyzed by Chi-square tests. Patterns of illness and adjustment varied significantly by age and gender cohorts, reflecting the relevance of age and gender at time of trauma on different trauma and loss experiences and cultural and religious shaping of subsequent adjustment and symptoms. Results The study confirmed that almost half of the Somali male patients are under age 30, 80% of whom presented with psychoses, compared with the rate of psychosis (13.7%) in the non-Somali control group of sameaged males at the clinic. The older male, and the majority of Somali female patients, show predominantly depressive and PTSD symptomatology. Conclusions War trauma experienced in childhood, early malnutrition from famines, head trauma, and excess Khat use in male adolescents provide partial explanations for the large number of young psychotic Somali men seen in the clinic from 2001 to 2009. Keywords Somali refugees
Psychoses
PTSD
Khat
Refugee mental health

Introduction There are approximately 50,000 Somali individuals living in Minnesota, the majority of whom reside in Minneapolis [1]. This represents the largest Somali community in the world outside of the Horn of Africa. Other large Somali

123

Soc Psychiat Epidemiol

communities in the US are located in Atlanta, Georgia, and Columbus, Ohio. Approximately 5% of school-age children in Minneapolis speak Somali at home. Despite traditional cultural resistance to acknowledgement of mental illness, the violence and dislocation of the Somali Civil War have brought about psychiatric syndromes that cannot be contained, especially in an urban environment and within a family unit, itself also fractured by the Somali diaspora. The psychiatric service of the Community-University Health Care Clinic, a general medical clinic, has assessed and treated more than 600 Somali persons with mental disorders.

Background Somalia occupies the entire coastline of the Horn of Africa, stretching 1,800 miles from Djibouti eastward along the Gulf of Aden and then southward to the Kenyan border. It has an estimated population (2005) of eight and a half million people, 85% of whom are ethnic Somalis and almost exclusively of the Sunni Muslim religion. Its recent history is, in broad outline, similar to other African territories that were colonized from the sixteenth century onward by European powers. Modern Somalia was partitioned in the nineteenth and early twentieth centuries into French, British, and Italian Somaliland Protectorates. It was pulled into World War II when Italian troops attacked British Somaliland in 1940 [2, 3]. After the war, Somalia was divided into British (northern) and Italian (southern) protectorates under a UN pact, and given its independence in 1960. The northern and southern protectorates voted to merge into a single nation at that time. From the start, the country was destabilized by clan-based rivalries. In 1969, General Mohamed Siad Barre took control of the government and established a repressive Soviet-supported socialist dictatorship. The Barre government was overthrown in January 1991 in a civil war, a war that has continued with varying degrees of mayhem and violence to this day. The efforts of UN and US to negotiate and ensure a peace have been unsuccessful and the country has been in near-total anarchy since the ouster of Barre. Rival clans and their militia have seized power and operate in different regions of the country, with the northern more stable than the southern regions of the country. Since June 2006, Somali Islamic militias have entered the Civil War, fighting against the clan-based militias. Fighting continues unabated, as coalitions between various clans and Islamic groups form, shift, and regroup. The relevance of this history for the Somali diaspora groups is that it contributes a burden of despair to their troubles regarding the possibility of return, as well as re-igniting their own Posttraumatic Stress Disorder (PTSD) symptoms and ongoing worries about kin left behind in Somalia and Kenya.

123

The intensity and extent of the violence and lawlessness occurring in the wake of the 1991 Civil War caught most Somalis unprepared both materially and psychologically. Youths of rival clans, armed with machine guns and grenades, often intoxicated on the amphetamine-like stimulant Khat [4– 6], and ostensibly representing their clan lines, attacked the homes of other-clan families, robbing, raping, and killing. Families fled, leaving behind their worldly possessions, but the flight was itself perilous. The same militia gangs set up roadside checkpoints and further robbed, raped, and killed those families trying to reach safe ground either in the countryside or in the neighboring countries, primarily Kenya and Ethiopia. Families could not bury their dead, and corpses littered the streets of Mogadishu and other areas of violence. Entrance as a refugee into a first host country does not guarantee safety or an end to exploitation and violence. Somali refugees were stopped at the Kenyan border by police and forced to pay bribes in lieu of proper identification papers to enter the country. Many refugees were thrown into jails and otherwise handled poorly. Once in the refugee camps, the population had to contend with ongoing intimidation and violence by its own youth gangs and by incursions of Somali clan militia into the Kenyan refugee camps. Travel out of the camp and into Nairobi was also dangerous because of hostile Kenyan police and gangs of youths. The humanitarian efforts of this host country and various UN and private organizations were at times compromised by Kenya’s concerns about violence spreading from Somalia into their own country, especially with its own large ethnic Somali population along the historically disputed northern border with Somalia [7]. Somali refugees also met with rape, robbery, and arrests in Yemen and the United Arab Emirates. From the camps and capital cities, there were various entrances, both legal and illegal, into a distribution process that relocated refugees into various countries in the Middle East, Asia (primarily Pakistan), Europe, and the Western hemisphere.

Method Setting CUHCC is an inner-city university outpatient clinic with medical, psychiatric, and dental services. It serves primarily an underserved population, 67% of whom fall below the federal poverty guidelines [8]. Total patient population at CUHCC for 2008 was 7,746 individuals. Over forty percent of patients speak minimal or no English and require interpreter services. Twenty-eight percent of patients have no health insurance, while 56% are insured by various government health programs that cover most but not all costs of services. Only 16% of patients have private insurance. Forty-one percent of staff is bilingual,

Soc Psychiat Epidemiol

comprising Cambodian, Hmong, Laotian, Oromo, Spanish, Somali, and Vietnamese language services on site; agency interpreters are called in for other linguistic groups (West African, Tibetan, Arabic, and East European languages). The clinic is located adjacent to a high-rise public housing project that now has a very large Somali population. The mental health unit alone employs nine full-time Somali-speaking clinical staff. Pathways to the clinic for Somali patients are referrals from Somali agencies, other county and voluntary agencies, hospital inpatient units, and self- and family-generated initiatives. Although it is often presumed that cultural attitudes keep non-Western peoples from utilizing mental health services, in fact the broad range of medical, psychiatric, and social services needed by refugees with psychiatric disabilities usually trumps whatever cultural and religious reticence may exist in this regard. Families, and the patients themselves, understand that psychiatric care and the various social and economic benefits to which mentally ill persons are eligible are an absolute necessity in the care and support of disabled family members in the urban US. In general, older adults come to the clinic because of a combination of PTSD and depressive symptoms (insomnia, nightmares, somatic complaints) and need for important social services, such as disability benefits and assistance in citizenship applications and disability waivers, whereas the majority of younger adults are brought in by their families or referred following a psychiatric hospitalization because of the behavioral manifestations of their psychoses (pacing and laughing in their room, bizarre and erratic behavior, running out of the house at night). Whereas marijuana and, less often, khat and methamphetamine use have been a contributing cause of some psychoses, families were often unaware of the drug use and did not raise drug use as a reason for bringing their young family member to the clinic or emergency room. Overall, this patient population is quite disabled, with very few having even part-time temporary employment, and with no young adults able to enroll and complete a semester in a college or technical course. Subjects The study population consists of all Somali patients (N = 600: 321 women; 279 men) above age 18 assessed in the Mental Health section of the Community-University Health Care Clinic (CUHCC) between January 2001 and October 2009. An additional ten Somali patients (4 men, 6 women), diagnosed with dementia or pervasive developmental disabilities, were not included in this study. The control group consisted of all non-Somali patients age 18 and older (N = 3,009: males = 1,342, females = 1,667) seen by Mental Health from October 2007 to October 2009. This 2-year time frame reflects the introduction of

electronic medical records at the clinic, which allowed computer access to diagnostic, gender and age cohort data. Diagnostic and assessment procedures Diagnoses were made upon initial evaluation, based on a clinical interview by the senior author, in accordance with DSM-IV-R criteria. Most interviews were conducted with the use of one of five Somali interpreters employed fulltime at the clinic. Information was also obtained from relatives and hospital records when available. Diagnoses in the non-Somali control group were made by the senior author, one other clinic psychiatrist, and five clinical nurse specialists credentialed in psychiatry. In the Somali patient population, the differentiation between schizophrenia, schizophrenic-like psychoses, and acute mania was in many cases impossible to make. Histories of drug use and mental illness episodes in the initial host countries were very unreliable for a variety of reasons, most likely related to concerns about possible effects of admission of drug use on immigration status and citizenship adjudication as well as cultural reluctance to acknowledge mental illness in oneself or one’s family. Many young Somali men were hospitalized in the US with an acute psychotic episode with disorganized, aggressive, and excitatory symptoms. They displayed poorly formed and shifting delusional thinking, often with prominent religious or paranoid themes. For statistical analysis purposes in this study, all patients with schizophrenia, mania, drug-induced psychosis, and psychosis not otherwise specified (DSM-IV codes 295, 297, 298, 293, 296.4 and 296.6) were classified in a single category as Psychoses. The diagnosis of PTSD was also difficult to make at times. Somali patients, especially the young men, were particularly reluctant to acknowledge, let alone discuss, details and impact of their traumas. This pattern is manifested in the remarkably low rate of PTSD in young Somali men given the almost universal exposure to the violence of the Civil War (see Table 1). At times, trauma experiences were acknowledged (e.g., a young man who as a 10-yearold-boy saw his father shot and killed in front of him), but there was strong denial of any emotional impact or lasting effect of witnessing this event. Men initially denied that their women kin had been raped. Women patients were equally reluctant to acknowledge or discuss sexual assaults despite the presence of PTSD symptoms suggestive of such trauma experiences. Even inquiries about nightmares and intrusive imagery were at times resented, as if the interviewer was implying that such ‘symptoms’ were pathological rather than a normal response to abnormal experiences. As an opening probe, all patients were routinely asked whether their home was attacked when the Civil War began in 1991 or shortly thereafter, with relevant

123

Soc Psychiat Epidemiol Table 1 Somali patients distribution of diagnosis by age and gender cohort (N = 600) Age

Diagnosis

Gender

Frequency Yes

\30

Psychosis

31–50 50[ \30

Depression_PTSD

31–50 50[ \30

PTSD

31–50 50[ \30

Depression

Chi-square test (df = 1) No

% Yes

Total 46.52 (p \ 0.0001)

M

104

26

80.00

130

F

25

52

32.47

77

M

36

40

47.37

76

F

12

96

11.11

108

M

6

67

8.22

73

F

6

130

4.41

136

M F

5 21

125 56

3.85 27.27

130 77

24.16 (p \ 0.0001) 7.26 (p \ 0.01)

M

29

47

38.16

76

F

63

45

58.33

108

M

50

23

68.49

73

F

108

28

79.41

136

M

4

126

3.08

130

F

7

70

9.09

77

M

4

72

5.26

76

F

14

94

12.96

108

M

11

62

15.07

73

F

9

127

6.62

136

M

17

113

13.08

130 77

30.41 (p \ 0.0001) 1.27 (p = 0.26)a

3.07 (p = 0.08) 3.48 (p = 0.06)a 3.00 (p = 0.08) 3.92 (p \ 0.05) 9.97 (p \ 0.01)

F

24

53

31.17

31–50

M

7

69

9.21

76

2.58 (p = 0.11)

50[

F M

19 6

89 67

17.59 8.22

108 73

0.10 (p = 0.75)

F

13

123

9.56

136

a

Chi-square may not be a valid test because of small cell sizes

questions following up on their particular responses. The interviewer’s questions about trauma experiences were low-keyed but covered likely scenarios. Nevertheless, a quantitative measure of types and number of trauma experiences was not possible in this clinical population and would lead to unreliable reporting, as follow-up interviews with patients who initially denied personal experience of assaults demonstrated. An indication of the reluctance of patients to acknowledge trauma histories and symptoms is seen in the 3–4 year follow-up study by Al-Saffar et al. [9], of Iranians, Turks, and Arab-speaking patients in an outpatient clinic in Sweden. Higher rates of probable PTSD were found in the three immigrant and refugee groups than in the Swedish group on follow-up, but in the original clinical evaluation 3 or 4 years earlier, no cases of PTSD were reported, indicating that the clinicians failed to inquire and the patients did not volunteer any information about their PTSD symptoms. Two additional studies caution about the cross-cultural use of scales initially designed for a Western population. First, reports of PTSD symptoms

123

in populations not exposed to Western psychoeducational and stress research methodology tend to emphasize somatic and economic complaints rather than our more familiar categories of intrusive imagery and numbing [10]. Second, use of scale cut-off points determined in one cultural group for study of a different group may lead to misidentification (false positives and false negatives) of those with mental health problems [11]. The descriptions of trauma when given were tragically stereotyped in the sense that once a home was attacked, slaughter of the men, robbery of family possessions, and gang rape of the women inevitably ensued. The decision to classify a person as either depressed or having PTSD was problematic, since in almost every case of PTSD, depression was present, and most depressed patients, especially women, had strong components of anxiety and PTSD. Patients presenting with prominent depressive and PTSD symptoms were placed in a combined ‘‘depression/PTSD’’ category. Methodologically, this is similar to Bhui et al.’s [12] decision, in a community survey of 180 Somalis living

Soc Psychiat Epidemiol

in London, of using symptoms of depression and anxiety (as measured by the subscales of the Hopkins symptom checklist) to designate those individuals with PTSD, depression, and anxiety symptoms. This strategy served to limit cross-cultural confounding in applying the translation of instruments developed in one culture to another culture. Clinically, these two categories of PTSD and depression formed a unified picture of depression, demoralization, and PTSD symptoms in response to complex trauma, forced displacement, multiple losses, and difficulties in the resettlement and acculturation processes [13–15]. Patients who showed symptoms of one but not both disorders, were diagnosed accordingly. There were some, but not many, who were not in the country at the outbreak of the Civil War and had not been directly exposed to trauma. The study population was divided into broad age cohorts because the chronological age at which trauma occurred appeared to greatly affect the type of trauma experienced and the psychological response to it. Teenage boys and men were at greatest risk of being shot and killed; younger boys were often ignored, at times clubbed on the head, and almost always were helpless witnesses to the violence; women from their early teens to late maturity were at high risk of assault, rape, and kidnap and, if resistant, murder. Analysis Chi-square tests were used to test the null hypothesis (with the expectation of empirical refutations) that no gender or age differences would be found within specific diagnostic categories and that the rates of psychoses would not vary between Somali and non-Somali patients seen at the clinic.

Results Within the Somali cohorts Table 1 shows the demographics and associations between diagnostic category, age, and gender for the Somali group. There were 279 men and 321 women. Almost half (46.6%) of the male patients, compared with almost one-quarter (23.0%) of women patients, were 30 years of age or younger. Comparing men and women patients of age 30 years or younger, we find that 80% of the men had psychotic disorders compared with 32.5% of the women (v2 = 46.5, df = 1, p \ 0.0001). Similarly, men 31– 50 years of age have significantly higher rates of psychoses than women aged 31–50 years (47.4 vs. 11.1%, v2 = 30.4, df = 1, p \ 0.0001). There are no significant differences in rates of psychoses between men and women aged 51 years and above. With respect to the diagnostic category ‘Depression/PTSD,’ Somali women had higher rates than

men in the age groups 30 years and below (v2 = 24.16, df = 1, p \ 0.0001) and 31–50 years (v2 = 7.26, df = 1, p \ 0.01). For ages 50 years and above, there was no significant difference between men and women (v2 = 3.07, df = 1, p \ 0.08). Within the non-Somali cohorts Similar to the Somali male groups, non-Somali males in the age cohorts 18–30 and 31–50 had higher rates of psychoses than non-Somali females: age 18–30, v2 = 8.24, df = 1, p \ 0.01; age 31–50, v2 = 7.35, df = 1, p \ 0.01. The other findings of significance in the non-Somali patients fall into the expected range: more women than men of ages 31–50 years with combined PTSD/depression diagnoses (v2 = 6.6, df = 1, p \ 0.01); more women than men with diagnoses of PTSD alone in the older age ranges (ages 31–50 years: v2 = 17.7, df = 1, p \ 0.0001; ages [ 50, v2 = 3.72, df = 1, p = 0.05); and more women than men in all age cohorts with significantly higher rates of depression. Comparing Somali cohorts with non-Somali cohorts Table 2 compares rates of illness of Somali patients versus non-Somali patients by age and gender cohorts. Eighty percent of male Somali patients of ages 18–30 years had psychotic illnesses compared with 13.7% of non-Somali males in the same age cohort (v2 = 223.49, df = 1, p \ 0.0001). Somali males between ages 31–50 years had a 47.4% rate of psychoses compared with 23.3% of nonSomali males (v2 = 19.95, df = 1, p \ 0.0001). These rates are reversed in the males aged 51 years and older, where the rates are 8.2% for Somali and 17.6% for nonSomali men (v2 = 3.95, df = 1, p \ 0.05). Somali women of ages 18–30 years had higher rate (32.5%) of psychoses than non-Somali women (8.0%) in the same age range (v2 = 40.41, df = 1, p \ 0.0001), but the direction is again reversed in the age group[50 years, where the non-Somali women have higher rates (16.5%) than the Somali women (4.4%) (v2 = 12.82, df = 1, p \ 0.001). For the combined PTSD/depression category, rates are higher for all Somali versus non-Somali age and gender groups, except for males of ages 18–30 years, where the Somali/non-Somali rates are low and almost identical (3.85 vs. 3.78%). Khat and other drug use in young Somali males Table 3 compares drug use versus no-drug use across all four diagnostic categories for Somali males of age below 30 years. The main drugs were khat and marijuana, with a few young men using alcohol, cocaine, and methamphetamine in addition to khat and marijuana. Parenthetically,

123

Soc Psychiat Epidemiol Table 2 Somali versus non-Somali patients (N = 3,009) by diagnosis, age, gender Age

Gender

Diagnosis

Group

Frequency Yes

\30

M

Psychosis

Somali Non-Somali

31–50

Somali Non-Somali

50[

Somali Non-Somali

\30

F

Psychosis

31–50

50[ \30

F

Depr_PTSD

31–50 50[ PTSD

F

PTSD

31–50 50[

Somali

Depression

6

130

4.41

136

71

360

16.47

431

5

125

3.85

130

18

458

3.78

476

Somali

50

23

68.49

73

Non-Somali

16

313

4.86

329

Somali

21

56

27.27

77

Non-Somali

18

505

3.44

523

Somali

63

45

58.33

108

Non-Somali

56

657

7.85

713

108

28

79.41

136

27

404

6.26

431

4

126

3.08

130

14

462

2.94

476

4

72

5.26

76

17

520

3.17

537

Somali

Somali

11

62

15.07

73

Non-Somali

14

315

4.26

329

7

70

9.09

77

Non-Somali

Somali

20

503

3.82

523

Somali

14

94

12.96

108

Non-Somali

65

648

9.12

713

Somali

Somali Somali Somali

Chi-square may not be a valid test because of small cell sizes

123

108 713

76

Non-Somali a

11.11 17.11

537

Non-Somali 50[

96 591

4.28

Non-Somali 31–50

12 122

38.16

Somali Depression

73 329

47

Somali

F

8.22 17.63

514

Non-Somali \30

67 271

29

Non-Somali 50[

6 58

23

Non-Somali 31–50

76 537

Somali

Somali M

47.37 23.28

Non-Somali

Non-Somali \30

40 412

77

Non-Somali

\30

36 125

523

Somali

50[

476

8.03

Non-Somali 31–50

130

13.66

32.47

Somali M

80.00

411

52

Non-Somali \30

26

65

481

Non-Somali 31–50

104

25

Somali Depr_PTSD

Total

42

Somali

M

%Yes

Somali

Non-Somali \30

No

Non-Somali Non-Somali 50[

Chi-square test (df = 1)

9

127

6.62

136

33

398

7.66

431

17

113

13.08

130

159

317

33.40

476

7

69

9.21

76

254

283

47.30

537

6

67

8.22

73

172

157

52.28

329

24

53

31.17

77

275

248

52.58

523

19

89

17.59

108

426

287

59.75

713

13

123

9.56

136

283

148

65.66

431

223.49 (p \ 0.0001) 19.95 (p \ 0.0001) 3.95 (p \ 0.05) 40.41 (p \ 0.0001) 2.47 (p = 0.12) 12.82 (p \ 0.001) 0.001 (p = 0.97) 98.41 (p \ 0.0001) 176.27 (p \ 0.0001) 62.72 (p \ 0.0001) 192.84 (p \ 0.0001) 304.91 (p \ 0.0001) 0.01 (p = 0.93)a 0.89 (p = 0.35)a 11.98 (p \ 0.001) 4.33 (p \ 0.05) 1.60 (p = 0.21) 0.16 (p = 0.69) 20.47 (p \ 0.0001) 39.51 (p \ 0.0001) 47.01 (p \ 0.0001) 12.31 (p \ 0.001) 67.14 (p \ 0.0001) 130.41 (p \ 0.0001)

Soc Psychiat Epidemiol Table 3 Somali men under age 30: drug versus no-drug use (N = 130) Diagnosis

Drug use (%)

No drug use (%)

Total

Psychoses

47 (45.2)

57 (54.8)

104

Depr/PTSD

2 (40)

Depression

6 (35.3)

3 (60)

5

11 (64.7)

17

PTSD

2 (50)

2 (50)

Total

57 (43.8)

73 (56.2)

4 130

rates of tobacco use were very high in this group of young men. We compared the rate of drug use between the psychosis and non-psychosis groups. There was no significant difference between numbers of individuals with drug use (N = 47) versus no-drug use (N = 57) in the cohort of 104 young men with psychosis (v2 = 0.3827, df = 1, p = 0.536). If anything, the percentage of psychotic men who did not use drugs was higher than psychotic men who used drugs. In the group of 26 Somali young men without psychoses, 61.54% reported no drug use versus 38.46% who reported drug use. We did not collect data on Khat use in older males because the majority of them did use Khat socially in Somalia and Kenya, and there was no way to determine accurately patterns of frequency and quantity of use. There are several caveats in collecting data about drug use in this patient cohort, which will be amplified in the discussion section.

Discussion Two major findings emerge from the study data: first the unexpected very high rates of psychoses in young Somali men seen in the clinic, which has not been noted to this extent in other studies of Somali refugees, and second an expected pattern of co-morbid depressive and PTSD symptoms in the women and older men patients that are in keeping with other studies of war refugee populations. High rate of psychotic disorders among Somali young men The unexpected and dramatic finding in our study of an outpatient clinical sample is that half of the Somali men are of less than 30 years of age and that fully 80% of these young men present with acute psychotic symptoms. In the non-Somali control group at our clinic, only 35.5% of men are aged less than 30 years, and only 13.7% in this group have a psychotic illness. The high rate of psychosis in young Somali men in our study is also a far different pattern from one seen at this same clinic among Southeast Asian refugee patients two decades ago [16]. There were

almost no cases of psychoses seen in young Hmong men, and the psychoses found in the Vietnamese, Cambodian, and Lao young men showed almost classical Dementia Praecox symptoms of either paranoid delusions or slow withdrawal into negative symptomatology, with little apparent influence of the trauma experiences of years of open and guerilla warfare in Indo-China. Specific explanations for the high rate of psychoses seen in young Somali men are not immediately apparent, but likely risk factors include history of trauma, age at which the trauma occurred, specific head trauma (smashed on head with rifle butts) meted out to young boys, starvation and malnutrition, excessive khat use in early adolescence followed by marijuana use in later adolescence, and cultural role expectations of young Somali men. The divergence of clinical symptoms along gender and age patterns suggests that the chronological age and developmental stage at which traumatic experiences occur and the gender of the person experiencing traumas are important considerations. Men and women, young and old, literally experience different forms of trauma and loss events, and process, respond to, and understand these traumas and losses differently. For a 5-year-old child, separation from mother is a greater trauma than observation of violence, whereas for a 15-year-old, the experience of physical and sexual assault, especially under conditions of enforced passivity, is the greater trauma. Underlying these different ways of experiencing and responding to trauma and loss are the cultural narratives that provide meaning, if such a term can be used, by which to make sense of such mayhem and rape, as well as a cultural pattern by which one is expected to respond to such disasters. A number of significant risk factors appear to converge in the life experiences of the young Somali men [16]. In a study of 135 Somali youth (ages 11–20 years old) in three New England communities, designed specifically to search for correlates of depression and PTSD in a middle-school sample, Ellis et al. [17] found that cumulative trauma was related to PTSD symptoms, and that post-resettlement stressors, acculturative stressors, and perceived discrimination were also associated with greater PTSD symptoms. Presence of psychotic symptoms was not measured. The few psychotic symptoms that were observed were thought to be related to PTSD (email communication from Dr. Ellis). This study utilized a community not a clinical sample, used research instruments designed to elicit PTSD and depression symptoms rather than diagnostic interviews, and excluded young Somali persons who could not be interviewed in English. Almost all of the young men evaluated in our study experienced major exposure to trauma, dislocation, and starvation in the Civil War of 1991 and the years beyond. In addition, many young men give a history of extensive

123

Soc Psychiat Epidemiol

Khat-chewing after forced displacement in their pre-teen and teen years to Kenya and Ethiopia [18] as well as when serving as combatants within conflict zones in Somalia [19–21]. Approximately 40% of the acute psychotic episodes in the US in Somali young men in our present study occurred in the context of marijuana abuse, which is becoming more prevalent among Somali youth. Khatchewing is present but less common in the US, due to the bulkiness in smuggling and degraded potency of the plant, although use of khat is making its appearance in the US and UK. Bhui et al. [11], in their London study, found that 62.6% of men and 16.9% of women endorsed Khat chewing at the present time. Marijuana use was at 1.1% for both genders. According to Bhui et al.’s later study [22], utilizing patients in a general medical practice registry and a non-patient community sample in London, increased risk for mental illness was associated with use of Khat, expressed intent of asylum seeking, and recruitment from the primary care registry (patient status) rather than the community setting. Employment and educational experiences emerged as protective factors. The authors note that the prevalence of substance use, including Khat, may have been underreported because of religious reasons for not using (alcohol, drugs) or for not acknowledging use when actually present. Prevalence of khat use in our Minneapolis cohort is unclear because most of our young patients initially deny khat use ever, and only later, if they are seen at follow-up appointments, do they acknowledge mild or even heavy khat use in Africa and the US. For example, one young man who, on initial interview, denied Khat use in Africa, recently told us that not only did he begin regular Khat use at age 17, but also that his father was a Khat grower. The young boys who experienced the violence, starvation, khat use, and civil breakdown in Somalia of the early 1990s have grown into the young men of the 2000s. As a general consideration, most studies of immigrants and refugees have found a higher incidence of mental illness, especially psychoses, in first- and-second generation migrant populations than in the host population [23–27]. This finding overlaps with growing recognition of the relationship between trauma and psychoses, especially in young people. A Dutch study of early adolescents (N = 1,290) found that nonclinical delusional ideation and hallucinatory experiences were strongly and independently associated with sexual trauma and a history of having been bullied [28]. In a study by Kilcommons and Morrison, 53% of 32 subjects with a diagnosis of psychosis met criteria for PTSD [29]. Severity of trauma was associated with severity of PTSD and psychotic symptoms, with physical abuse associated with positive psychotic symptoms and sexual abuse specifically related to hallucinations. In a report of findings from the National Comorbidity Survey

123

(N = 5,877), childhood physical abuse was the only significant predictor of psychosis in the total sample after depression was controlled [30]. A second line of research has demonstrated an association between early cannabis use and an increased risk for development of psychotic disorders. Moore et al. [31], in a review of 35 studies, concluded that there is sufficient evidence to warn young people that cannabis use increases their risk of developing a psychotic illness later in life. Degenhardt and Hall [32], in a review of six studies from five countries that examined relationships between selfreported cannabis use and the risk of diagnosis with a psychosis or of reporting psychotic symptoms, showed that regular cannabis use predicted an increased risk of a schizophrenia diagnosis or reported symptoms of psychosis after controlling for confounding variables in adolescents and young adults. In a separate study of young adults with schizophrenia and related disorders, Degenhardt et al. [33] found that cannabis use predicted a small but statistically significant increase in psychotic symptoms. Increased psychotic symptoms or depression did not predict increased cannabis use, indicating that cannabis was not used as selftreatment in the face of worsening symptoms. Henquet et al. [34] in a prospective study that assessed cannabis use, predisposition for psychosis, and psychotic symptoms in 2,437 young people in Munich and environs, concluded that cannabis use moderately increases the risk of psychotic symptoms in young people, with a much stronger effect on those with evidence of predisposition for psychosis. Schweinsburg et al. [35] in a review of the influence of marijuana on neurocognitive functioning in adolescence, concluded that heavy marijuana use in adolescence is associated with persisting abnormalities in working memory, attention, and learning. A similar problem regarding the relationship of substance abuse to psychosis arises in regard to Khat, the plant indigenous to the Horn of Africa and parts of the Mideast. Cathinone, the pharmacologically active substance in the khat leaf, is a mild stimulant with amphetamine-like properties [4]. Khat leaves are chewed recreationally by many men in that region. This habit is culturally endorsed and socially regulated. Most khat-chewers obtain a mild high from the plant without developing psychoses or longlasting personality aberrations. A recent literature review of the evidence for a causal relationship between khat use and mental illness sifted through 450 papers published between 1945 and 2006 identified only 24 papers that met inclusion criteria for relevance and methodology and concluded that the evidence supports a weak association between khat use and mental illness, with no scientific evidence of causality [36]. However, the changing pattern of khat use recently, together with extensive social breakdown, is beginning to implicate excessive khat use as an important risk factor.

Soc Psychiat Epidemiol

Odenwald [37] has written an extensive review of the pharmacology of Khat and the changing patterns from socially regulated use to increasingly unregulated abuse of Khat in the youthful populations in the Horn of Africa as the social fabric of these countries have imploded. Soon after the outbreak of the Somali Civil War, Randall [38] and, more recently with greater urgency, Bhui and Warfa [19] have written about the economic effects and the exacerbation of poverty of Khat consumption in Somalia, where per capita income is about US $1 per day, yet 80% of the male population spends about US $4 per day on khat. Our findings of a predominance of acute psychoses in young male patients are similar to the findings of a 2002 community survey of a representative sample of 612 Somali households in Hargeisa, the capital of the northern provinces, by Odenwald et al. [18]. These investigators found that 21% of the surveyed households care for at least one member with severe mental health problems, and that 15% of the ex-combatants (primarily men) suffered from a severe mental disorder, mostly psychoses. The authors link the high prevalence of psychoses in young men to a combination of heavy Khat use and traumatic stress. Odenwald et al. repeated their study on a larger scale with 8,723 interviews of Somali combatants and found a dose effect, mediated by PTSD, between higher Khat consumption and paranoia. Respondents with PTSD used Khat more frequently, saying that it helped them forget their war experiences, but more frequent use of Khat was also correlated with greater odds of paranoid ideation [20, 21]. Last, there is extensive medical literature on the effects of starvation on fetuses and young children, including risk involving central nervous system development and higher than expected rates of schizophrenia and other serious mental illnesses [39–41]. Two years of drought and famine (1992–1993) immediately followed the Somali Civil War [42]. An estimated 300,000 persons died of starvation during these early years, most drawn from the approximately 2 million internally displaced persons [43, 44]. Child mortality rates were 40%. Relief agencies were unable to bring sufficient quantities of foodstuff and medicines to the starving population because of the anarchy brought on by fighting between rival clans [45]. The Somali population now reaching adolescence and young adulthood belongs to this cohort exposed to prenatal and infant malnutrition in the early 90s [46–48]. Low prenatal Vitamin D has been hypothesized as a risk factor for schizophrenia (possibly as an inducer of nerve growth factor synthesis) [49], as well as a risk factor in adult schizophrenia in first- and second-generation immigrants in high-latitude countries [50]. Similarly, maternal iron deficiency has been associated with increased risk of schizophrenia in offspring [51].

Clinical presentation of older men The clinical presentation of the older men, roughly of age 50 years and beyond, is vastly different. Psychosis is rare; a mixture of depression and PTSD predominate. The older men were extensively traumatized. They were shot and beaten, and then forced to look on while their male relatives were killed and their wives and daughters raped and sometimes murdered. Some will describe the events during an interview; most will bury their faces in their hands, weep, and say they cannot talk about it. Shame and demoralization rather than anger are expressed by the older men. The rates of psychoses in the older men are lower than that seen in the non-Somali control group, reflecting the clinic’s regular aging-in population of patients with chronic schizophrenia and bipolar disorders. PTSD/depressive patterns in Somali women and older men The high rates of co-morbid depressive and PTSD symptoms in women and older men readily fit with the trauma and loss histories most of these patients experienced. Somali women across all age cohorts, and Somali men of age over 31 years, have much higher rates of combined PTSD/depression diagnoses than in the nonSomali control group. This high rate is consistent with rates seen in other studies of refugees from war-devastated countries. In a study comparing Somali and Rwandese war refugees in an Ugandan refugee settlement, Onyut et al. [52] found that Somali persons had higher prevalence of PTSD than Rwandese persons (48.1 vs. 32.0%). In a community survey of mental disorders among Somali refugees in London, utilizing samples from general practice registries and community venues within two boroughs, Bhui et al. [22] utilizing the MINIneuropsychiatric interview [53] adapted for cross-cultural use to screen for common psychiatric disorders, found relatively low PTSD rates, but 90% of those with PTSD also had a common mental disorder. Almost 60% of all Somali women seen in our mental health clinic fall into the combined depression/PTSD category, compared with just 3.4% of the non-Somali women. Some of the women present primarily with nightmares, intrusive imagery, avoidance, and physiological arousal while most show a mixed clinical picture of PTSD symptoms along with fatigue, passivity, helplessness, sadness, and inability to function even in basic household chores, such as shopping, cooking, and responsible care for children. Most women have lost several of their children and remain in chronic bereavement about these losses, as well as the fracturing of their secure domestic existence prior to

123

Soc Psychiat Epidemiol

the Civil War. In many cases, marriages have fallen apart under the weight of deaths of children, sickness in one or both of the spousal partners, disintegration of the social structures that maintain marital cohesiveness and resolve disputes, and loss particularly of the husband’s role as protector and breadwinner of the nuclear or extended family. Abandonment of murdered relatives, because of the imperative to leave the scene of devastation immediately, with the consequent failure to properly handle and respectfully bury the dead, also leave a powerful residue of humiliation and resentment between the marital couple. Although the data on our Somali patients in the clinic represents, for the most part, a cross-sectional assessment at the time of their clinical evaluation, in fact our longitudinal experience with many of the older Somali men and women patients in this study is that they are well on their way toward chronicity of symptoms and disability. In a study of Afghan, Iranian, and Somali refugees and asylum seekers in the Netherlands, Gerritsen et al. [54] studied prevalence rates of physical and mental health problems and identified risk factors for illness in these groups. Asylum seekers, whose future was quite uncertain because they had not been formally accepted for refugee status by the host country, scored higher on scales of poor general health, PTSD, and depression/anxiety symptoms. Eighty-seven of the 410 subjects in the study were Somali. The Somali and Afghan individuals had lower rates of PTSD, depression/anxiety symptoms and self-reported poor health than did the Iranian group. The authors’ comment that the changed political climate (less tolerance toward asylum seekers; more people threatened to be sent back to their country of origin) may contribute to the greater symptomatology of the more recent arrivals (Iranian asylum seekers) compared with the earlier arrivals (refugees). This impression of chronicity and poor outcome is supported by a 10-year-follow-up study of Vietnamese refugees in Australia, in which the risk of mental illness, and particularly of PTSD, increased with age, a pattern opposite that seen in Australians [55]. Similarly, in a follow-up study by Al-Saffar et al. [56] of Arab-speaking, Iranians, and Turks in Sweden, refugee and guest worker patients with PTSD symptoms had a considerably poorer outcome in terms of general health and disability compared with those without PTSD symptoms. The rate of depressive disorders (without co-morbidity with PTSD) in our study is higher in the non-Somali than the Somali women across all ages, reflecting the fact that depression is the most common mental disorder seen in women in outpatient clinics. Somali women have very high rates of depression too, but their depressions are intermingled with PTSD symptoms.

123

Limitations of the study The data presented in this paper are derived from a consecutive clinical population seen in a community medical/ psychiatric clinic. It is not a study of community prevalence of mental illness in general, psychoses in young men, Khat and marijuana use, or PTSD/depressive symptoms in women and older men, nor does it suggest that the rates of these illnesses seen in clinic patients are carried over to community rates. Methodologically, we have chosen to err on the side of broad diagnostic categories based upon clinical interviews rather than upon the use of clinical research instruments. We have provided our justification for this, certainly disputed by many, that use of questionnaire-type instruments designed to elicit very personal and potentially high-risk disclosures that might have an impact on immigration status and citizenship eligibility (from a refugee’s point of view) runs a great risk of under-reporting and denial of the very information the study is seeking. We can easily report much anecdotal evidence to the effect that initial denial of Khat use and trauma experiences, to give just two categories of information, change to acknowledgment of drug use and descriptions of trauma experiences after the patient comes to know us. The opposite pattern is also a possibility, as it is in many studies, that symptoms and trauma experiences may be embellished when a patient is seeking disability benefits, asylum protection, and citizenship disability exemption from learning English. The risk of our not using validated instruments in this study is that the diagnoses may reflect the biases of the author and the study group. Further, whereas the DSM-IV-R diagnoses of all the Somali patients were made following interview by the senior author, some of the DSM-IV-R diagnoses of the non-Somali control group were made by clinical psychiatric nurse practitioners rather than by the senior author. The senior author and the nurse practitioners have worked closely for a number of years, and cases are reviewed regularly, but no inter-rater reliability studies were done. There is some safety afforded by condensing various categories of psychoses, such as schizophrenia, mania, and schizoaffective disorder, into higher-order categories such as Psychoses, whereby the risks of diagnostic disagreement among specific diagnoses would be minimized. We have not reported on marital, employment, and living situations because these have been extremely fluid in this refugee population, as well as the observation that definitions of marriage, separation, divorce, kinship relationships (e.g., nieces and nephews are often referred to as biological children), changing housing patterns, and parttime employment are not what they appear to be. Even asking the age of some patients is problematic and raises suspicions, because many patients had incorrect birth dates

Soc Psychiat Epidemiol

placed on their immigration documents, often not understanding why this was done. It has been our sense that to provide ‘hard data’, i.e., data in print of responses to these remarkably tangled questions, gives a sense of accuracy and permanence that we could not stand behind. In a similar way, ascertaining how many Somali persons live in the catchment area, in order to compare percentage of Somali persons in the community seen in the clinic with percentage of non-Somali persons in the community seen in the clinic, is problematic, First, the actual catchment area of the clinic includes not only Hennepin County (Minneapolis), but also Ramsey County (St. Paul), as well as the other five counties constituting the Greater Twin Cities Metropolitan area, plus other areas of Minnesota, Wisconsin, and Iowa. Second, it is very hard to ascertain the size of the Somali immigrant population. The last actual census was 2000. There will be a census this year. What we have offered in this paper is the best estimate of government officials. One problem is that the Somali immigrant population, especially the young men, is a very fluid and mobile one, moving back and forth within the Twin Cities and greater metropolitan area, but also back and forth to Atlanta, Columbus (Ohio), Alaska, and to relatives at the meat-packing plants of Iowa, South Dakota, and Nebraska. One last problem involves the question of whether the rate of psychotic young Somali men seen in the clinic is artificially elevated compared with non-psychotic young Somali men because young Somali men with psychoses are seen at the clinic only when dragged in by their relatives. This implies that non-psychotic but troubled Somali young men are permitted by their families to avoid coming to clinic, which at face value seems true, but conversely that non-Somali young psychotic men are more willing to come to the clinic than psychotic Somali young men. This does not ring true to us. It suggests that non-Somali psychotics willingly bring themselves to psychiatrists. It has been our observation in many years of psychiatric work that only rarely do any schizophrenics or manics or depressives with psychosis (or others with acute psychotic experiences) bring themselves into the clinic. The Somali young men seen in the clinic do not appear to have any less insight or be less open to recognizing and talking about their illnesses than are the other young psychotic men (and women) whom we have long evaluated and treated. This is why ACT teams, case managers, and outpatient commitment are so essential in the care and treatment of such patients, and why there are such high rates of readmission of psychotic individuals to hospital, because these individuals resist psychiatric diagnoses and treatment. In general, most young persons with psychosis avoid and resist psychiatric evaluation and treatment, and stop taking their medicine or participating

in their rehabilitation programs unless supported and pressured by family and community.

Conclusions The psychiatric symptoms and adjustment problems of a cohort of Somali individuals seen in a Minneapolis community clinic, while sharing many common factors with refugees from other continents and other wars, are powerfully shaped by Somali culture and the specific savagery of the Civil War that began in 1991. Against the Null hypothesis that diagnostic categories would not correlate with age and gender considerations, half of the Somali male patients in the study are under age 30, of whom the overwhelming majority present with acute psychoses that have mixed paranoid and affective components. The older men show predominantly depressive and PTSD symptomatology. The majority of Somali female patients aged over 30 years have predominantly depressive and PTSD symptoms. It is our impression that the actual rate of PTSD, especially in our young male patients, is higher than reflected in the tables because of many subjects’ strong reluctance to acknowledge their experiences of trauma and its effects. The symptom picture of Somali women and older Somali men conform to expectations in terms of the experiences of violence in the Civil War and the losses all have suffered. Likely explanations for the predominantly psychotic picture seen in the younger men include a combination of war trauma experienced in childhood, early malnutrition, heavy Khat and marijuana use in adolescence, and the demands of the male role in Somali culture, a model of environmentally additive risk factors that has gained much salience lately [57]. Research into these complex factors, along with biological and genetic studies, should provide some insights into these dramatic clinical patterns. Methodological problems relating to the willingness of vulnerable populations, such as war refugees with uncertain status in a host country, to provide trauma narratives, answer trauma questionnaires accurately, give descriptions of how other kin are related to them, and report accurate drug use histories represent another area that calls for painstaking and repeated research protocols. Acknowledgments This study was approved by the IRB Human Subjects Committee of the University of Minnesota Medical School. Dr. Fujiwara was supported in part by shared resources through the NIDA grant R21DA024626 (Khat Research Program: Neurobehavioral Impact of Long-Term Use; PI: Mustafa al’Absi). The authors thank Auke Tellegen, Ph.D. and Mustafa al’Absi, Ph.D. for discussion and analysis of the manuscript, Essa Hassan, Awo Qasim, Aisha Mohamed, Tahir Hassan, and Abdulahi Mohamed for dedicated involvement in this project and in the care of their patients at all times, and Amy Shellabarger for organizing the data retrieval of the control group.

123

Soc Psychiat Epidemiol Conflict of interest statement interest in this research project.

The authors report no conflict of

References 1. Population Notes, Minnesota State Demographics Center, June 2004. http://www.demography.state.mn.us/PopNotes/Evaluating Estimates.pdf. Accessed 23 December 2009 2. Lewis IM (2002) A modern history of the Somali, 4th edn. Ohio University Press, Athens 3. Besteman C (1999) Unraveling Somalia: race, violence and the legacy of slavery. University of Pennsylvania Press, Philadelphia 4. Feyissa AM, Kelly JP (2008) A review of the neuropharmacological properties of khat. Prog Neuropsychopharmacol Biol Psychiatry 32:1147–1166 5. Al-Motarreb A, Baker K, Broadley KJ (2002) Khat: pharmacological and medical aspects and its social use in Yemen. Phytother Res 16:403–413 6. Cox G, Rampes H (2003) Adverse effects of khat: a review. Adv Psychiatr Treat 9:456–463 7. Anthony C (2009) Africa’s displaced peoples: State-building realpolitik and stunted globalization. Harvard Int Rev 31(3):50–54 8. The 2009 HHS Poverty Guidelines. Federal Register Notice, 23 Jan 2009. http://aspe.hhs.gov/poverty/09poverty.shtml. Accessed 11 November 2009 9. Al-Saffar S, Borga P, Edman G, Hallstrom T (2003) The aetiology of posttraumatic stress disorder in four ethnic groups in outpatient psychiatry. Soc Psychiatry Psychiatr Epidemiol 38:456–462 10. Yeomans PD, Herbert JD, Forman EM (2008) Symptom comparison across multiple solicitation methods among Burundians with traumatic event histories. J Trauma Stress 21:231–234 11. Ichikawa M, Nakahara S, Wakai S (2006) Cross-cultural use of the predetermined scale cutoff points in refugee mental health research. Soc Psychiatry Psychiatr Epidemiol 41:248–250 12. Bhui K, Abdi A, Abdi M, Pereira S, Dualeh M, Robertson D, Sathyamoorthy G, Ismail H (2003) Traumatic events, migration characteristics and psychiatric symptoms among Somali refugees. Soc Psychiatry Psychiatr Epidemiol 38:35–43 13. Kroll J (2003) Posttraumatic symptoms and the complexity of responses to trauma. JAMA 290:667–670 14. Briggs L, Macleod AD (2006) Demoralisation—a useful conceptualization of non-specific psychological distress among refugees attending mental health services. Int J Soc Psychiatry 52:512–524 15. Scuglik DL, Alarcon RD, Lapeyre AC, Williams MD, Logan KM (2007) When the poetry no longer rhymes: mental health issues among Somali immigrants in the USA. Transcult Psychiatry 44:581–595 16. Kroll J, Habenicht M, Mackenzie T, Yang M, Chan S, Vang T, Nguyen T, Ly M, Phommasouvanh B, Nguyen H, Vang Y, Cabugao R (1989) Depression and post-traumatic stress disorder in Southeast Asian refugees. Am J Psychiatry 146:1592–1597 17. Ellis BH, MacDonald HZ, Lincoln AK, Cabral HJ (2008) Mental health of Somali adolescent refugees: the role of trauma, stress, and perceived discrimination. J Consult Clin Psychol 76:184–193 18. Odenwald M, Schauer M, Neuner F, Lingenfelder B, Horn R, Catani C, Klaschik C, Elbert T (2002) War trauma, Khat abuse and psychosis: mental health in the demobilization and reintegration program Somaliland. Final Report of the vivo mission 2 within the EC/GTZ, Sept–Nov. http://www.vivo.org. Accessed 27 August 2006

123

19. Bhui K, Warfa N (2007) Drug consumption in conflict zones in Somalia. PLoS Med 4(12):e354. doi:10.1371/journal.pmed. 0040354 20. Odenwald M, Hinkle H, Schauer E, Neuner F, Schauer M, Elbert TR, Rockstroh B (2007) The consumption of Khat and other drugs in Somali combatants: a cross-sectional study. PLoS Med 4(12):e341. doi:10.1371/journal.pmed.0040341 21. Odenwald M, Hinkle H, Schauer E, Schauer M, Elbert TR, Neuner F, Rockstroh B (2009) Use of khat and posttraumatic stress disorder as risk factors for psychotic symptoms: a study of Somali combatants. Soc Sci Med 69:1040–1048 22. Bhui K, Craig T, Mohamud S, Warfa N, Standfeld SA, Thornicroft G, Curtis S, McCrone P (2006) Mental disorders among Somali refugees: developing culturally appropriate measures and assessing socio-cultural risk factors. Soc Psychiatry Psychiatr Epidemiol 41:400–408 23. Mavreas V, Bebbington P (1989) Does the act of immigration provoke psychiatric breakdown? A study of Greek Cypriote immigrants. Acta Psychiatrica Scand 80:469–473 24. Selten JP, Sijben N (1994) First admission rates for schizophrenia in immigrants to the Netherlands: the Dutch National Registrar. Soc Psychiatry Psychiatr Epidemiol 29:71–77 25. Cantor-Graae E, Pedersen CB, McNeil TF, Mortensen PB (2003) Migration as a risk factor for schizophrenia: a Danish population based cohort study. Br J Psychiatry 182:117–122 26. Leao TS, Sundquist J, Frank G, Johansson L-M, Johansson E-S, Sundquist K (2006) Incidence of schizophrenia or other psychoses in first-and second-generation immigrants, a national cohort study. J Nerv Ment Dis 194:27–33 27. Coid JW, Kirkbride JB, Barker D, Cowden F, Stamps R, Yang M, Jones PB (2008) Raised incidence rates of all psychoses among migrant groups. Arch Gen Psychiatry 65:1250–1258 28. Lataster T, Van Os J, Drukker M, Henquet C, Feron F, Gunther N, Myin-Germeys I (2006) Childhood victimization and developmental expression of non-clinical delusional ideation and hallucinatory experiences: victimization and non-clinical psychotic experiences. Social Psychiatry Psychiatr Epidemiol 41:423–428 29. Kilcommons AM, Morrison AP (2005) Relationships between trauma and psychosis: an exploration of cognitive and dissociative factors. Acta Psychiatr Scand 112:351–359 30. Shevlin M, Dorahy MJ, Adamson G (2007) Trauma and psychosis: an analysis of the National Comorbidity Study. Am J Psychiatry 164:166–169 31. Moore THM, Zammit S, Lingford-Hughes A, Barnes TRE, Jones PB, Burke M, Lewis G (2007) Cannabis use and risk of psychotic or affective mental health outcomes: a systematic review. Lancet 370:319–328 32. Degenhardt L, Hall W (2006) Is cannabis use a contributory cause of psychosis? Can J Psychiatry 51:556–565 33. Degenhardt L, Tennant C, Gilmour S, Schofield D, Nash L, Hall W, McKay D (2007) The temporal dynamics of relationships between cannabis, psychosis and depression among young adults with psychotic disorders: findings from a 10-month prospective study. Psychol Med 37:927–934 34. Henquet C, Krabbendam L, Spauwen J, Kaplan C, Lieb R, Wittchen H-U, van Os J (2005) Prospective cohort study of cannabis use, predisposition for psychosis, and psychotic symptoms in young people. Br Med J 330:11–14 35. Schweinsburg AD, Brown SA, Tapert SF (2008) The influence of marijuana use on neurocognitive functioning in adolescents. Curr Drug Abuse Rev 1:99–111 36. Warfa N, Klein A, Bhui K, Leavey G, Craig T, Stansfeld A (2007) Khat use and mental illness: a critical review. Soc Sci Med 65:309–318 37. Odenwald M (2007) Chronic khat use and psychotic disorder: a review of the literature and future prospects. Sucht 53:9–22

Soc Psychiat Epidemiol 38. Randall T (1993) Khat abuse fuels Somali conflict, drains economy. JAMA 269:12–13 39. Susser E, Neugebauer R, Hoek HW, Brown AS, Lin S, Labovitz D, Gorman JM (1996) Schizophrenia after prenatal famine. Arch Gen Psychiatry 53:25–31 40. St. Clair D, Xu M, Wang P, Yu Y, Fang Y, Zhang F, Zheng X, Gu N, Feng G, Sham P, He L (2005) Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–61. JAMA 294:557–562 41. McClennan JM, Susser E, King M-C (2006) Maternal famine, de novo mutations, and schizophrenia. JAMA 296:582–584 42. Internal Displacement Monitoring Centre (iDMC): Somalia: Window of opportunity for addressing one of the world’s worst internal displacement crises. 10 January 2006, http://www. internal-displacement.org. Accessed 23 August 2006 43. Sargent J, Michael K (2005) The need for a digital aid framework in humanitarian relief. The 9th World Multi-Conference on Systematics, Cybernetics and Informatics, Orlando, FL, 10–13 July 2005. http://ro.uow.edu/infopapers/377. Accessed 26 August 2006 44. IRINnews: Somalia. All-out war catastrophic for southern Somalia––famine watchdog. 12 January 2007. http://www. IRINnews.org. Accessed 12 January 2007 45. Patman RG (1997) Disarming Somalia: the contrasting fortunes of US and Australian peacekeepers during UN intervention 1992– 93. Afr Aff 96:509–533 46. Moore PS, Marfin AA, Quenemoen LE, Gessner BD, Ayub YS, Miller DS, Sullivan KM, Toole MJ (1993) Mortality rates in displaced and resident populations of central Somalia during 1992 famine. Lancet 341:935–938 47. Seal AJ, Creeke PI, Mirghani Z, Abdalla F, McBurney RP, Pratt LS, Brookes D, Ruth LJ, Marchand E (2005) Iron and vitamin A deficiency in long-term African refugees. J Nutr 135:808–813 48. Collins S, Myatt M (2000) Short-term prognosis in severe adult and adolescent malnutrition during famine. JAMA 284:621–626

49. McGrath J (1999) Hypothesis: is low prenatal vitamin D a riskmodifying factor for schizophrenia? Schizophrenia Res 40:173–177 50. Dealburto MJ (2007) Why are immigrants at increased risk for psychosis? Vitamin D insufficiency, epigenetic mechanisms, or both? Med Hypotheses 68:259–267 51. Insel BJ, Schaefer CA, McKeague IW, Susser ES, Brown AS (2008) Maternal iron deficiency and the risk of schizophrenia in offspring. Arch Gen Psychiatry 65:1136–1144 52. Onyut LP, Neuner F, Ertl V, Schauer E, Odenwald M, Elbert T (2009) Trauma, poverty and mental health among Somali and Rwandese refugees living in an African refugee settlement––an epidemiological study. Confl Health 3. doi:10.1186/1752-15053-6 (electronic version) 53. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, Hergueta T, Baker R, Dunbar GC (1998) The miniinternational neuropsychiatric interview (MINI): the development and validation of a structured diagnostic interview for DSM-IV and ICD-10. J Clin Psychiatry 59(20):22–33 54. Gerritsen AAM, Bramsen I, Deville W, van Willigen LHM, Hovens JE, vander Ploeg HM (2006) Physical and mental health of Afghan, Iranian, and Somali asylum seekers and refugees living in the Netherlands. Soc Psychiatry Psychiatr Epidemiol 41:18–26 55. Silove D, Steel Z, Bauman A, Chey T, McFarlane A (2007) Trauma, PTSD and the longer-term mental health burden amongst Vietnamese refugees. Soc Psychiatry Psychiatr Epidemiol 42:467–476 56. Al-Saffar S, Borga P, Hallstrom T (2002) Long-term consequences of unrecognized PTSD in general outpatient psychiatry. Soc Psychiatry Psychiatr Epidemiol 37:580–585 57. Cougnard A, Marcelis M, Myin-Germeys I, De Graaf R, Vollebergh W, Krabbendam L, Lieb R, Wittchen HU, Henquet C, Spauwen J, Van Os J (2007) Does normal developmental expression of psychosis combine with environmental risk to cause persistence of psychosis? A psychosis proneness-persistence model. Psychol Med 37:513–527

123