Variations in psychiatric morbidity between traumatized Norwegian, refugees and other immigrant patients in Oslo

Abstract

Background: There is a lack of clinical studies that focus on different psychiatric disorders after trauma and the relationship with migration status.

Purpose: To examine differences in psychiatric morbidity in traumatized patients referred to psychiatric treatment in Southern Oslo.

Materials and methods: Hundred and ten patients with trauma background attending an outpatient clinic in Southern Oslo were studied. Forty-four of the participants (40%) were ethnic Norwegians, 25 (22.7%) had refugee background and 41 (37.3%) were first- or second-generation immigrants without refugee background. Thorough diagnostic assessment was done by experienced psychiatrists through several structured clinical interviews and self-report questionnaires.

Results: Ninety-eight patients (89%) were diagnosed with at least one Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR) disorder. There was a clear difference in the presentation of certain psychiatric disorders between the groups. Ethnic Norwegian patients were more likely to have anxiety disorders: agoraphobia, social phobia and panic disorder than non-refugee immigrant patients. They also had higher rates of alcohol abuse/dependence. Somatoform pain disorder was more common in both the refugee and other-immigrant groups than among the ethnic Norwegian patients. The refugee patients had significantly more major depressive disorder, post-traumatic stress disorder (PTSD) and both co-occurring.

Conclusion: Trauma is frequently associated with depression, anxiety disorders, somatoform pain disorder and PTSD in a clinical population. The clinical presentation and comorbidity of these disorders seem to vary significantly between traumatized patients with Norwegian, refugee and non-refugee immigrant backgrounds. After a major trauma, refugees may be at greater risk for both PTSD and depression than other immigrants and the native population.

Keywords:

• Immigrants
• refugees
• trauma
• diagnostics
• comorbidity

Introduction

Background

Migration between countries is common and steadily increasing. Over 3% of the world’s population were migrants in 2017. The major direction of inter-continental migration has been from low- to high-income countries. Nordic countries are popular destinations for many of these migrants. The Nordics are considered safe, highly developed countries and have well established social security systems. In Norway 17% of the country’s total population are immigrants. In the capital Oslo immigrants and second-generation immigrants account for about 36% of the total population [1,2].

The mental health of immigrants has been studied for decades with a pioneer study conducted on Norwegian immigrants in Minnesota in the 1920s and 1930s. It reported higher rates of psychosis among the Norwegian immigrants than Norwegians living in Norway [3]. In general, immigrants have been found to have a higher prevalence of many different mental health problems both in Norway, the Nordics and elsewhere [4–6]. Refugees and asylum seekers in high income countries have been shown to have particularly higher prevalence of trauma and trauma related disorders [7,8].

There have been some population-based studies comparing immigrant groups with the native populations on different mental health problems [9]. Posttraumatic stress disorder (PTSD) has been the most studied disorder in refugee populations [10,11]. Other psychiatric disorders like depression and anxiety are equally common in traumatized populations, but have been less studied [7,12,13]. Furthermore, there is a lack of studies in clinical populations that compare migration status and geo-cultural background to diagnosis. Particularly there are few studies that have compared psychiatric disorders between traumatized immigrant and traumatized native patients [14].

Norway has been ranked as the most prosperous and one of the most peaceful countries in the world [15,16]. In later years, however, there has been more focus on psychological trauma and its consequences in Norway as well. After the major terror attack in Oslo and Utøya in 2011, there has been an increased awareness on trauma and trauma related disorders [17]. Previous studies have shown low prevalence of trauma and PTSD in the general population in Norway [18]. A recent large population study found, however, much higher rates. They found that trauma affects about 85% of the population of which 8.5% of women and 3.8% of men have PTSD [19]. This may be due to increased awareness on trauma related disorders in the Norwegian population and a growing immigrant population in Norway. There is a need to know more about trauma in both the populations.

The health care systems in larger cities in the Nordic countries, including Oslo, are facing new challenges due to growing immigrant populations. This challenge seems to be particularly evident in mental health care [20–23]. There is a clear need for more information on how refugees and other immigrants differ from the native population so that we can provide culturally relevant mental health services.

The aim of our study was to compare psychiatric disorders among treatment-seeking traumatized patients with immigrant, refugee and Norwegian background living in an area with high proportion of immigrants. We wished to explore if there was a difference between the groups and what difference this could be. Our hypothesis was that traumatized patients with immigrant or refugee background could have more and different psychiatric disorders than traumatized native Norwegians living in the same area. We tried to examine possible explanations such as the effect of immigration and psychosocial situation, different trauma types, acculturation stress and ethnic and cultural differences.

Methods

Study subjects and procedures

The Study of Health Outcome after Trauma (SHOT-study), is a cross-sectional study of health outcome after major psychological trauma. This part of the study recruited patients referred to Southern Oslo District Psychiatric Centre (Søndre Oslo DPS or SOD). The second line mental health center is a part of Oslo University Hospital which is one of the largest hospitals in Northern Europe. This area in Oslo is known to have a culturally and ethnically diverse population. Non-ethnic Norwegians account for more than 50% of the population in many sectors, with a large proportion of refugee and non-Western immigrant population.

The participant subjects were recruited from the outpatient clinic of SOD. Both newly referred and patients already receiving treatment in this clinic were given an opportunity to participate. If their therapist found that their patients had experienced a major trauma, they were given an opportunity to participate in the study. The trauma or trauma-related symptoms needed not be the main reason for the referral as long as the therapist uncovered a serious trauma in the history.

The inclusion criteria were having experienced at least one serious trauma according to the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR)criteria, referred to or already receiving treatment at SOD, age between 18 and 60, proficiency in a Scandinavian language or English and able to give informed consent. The exclusion criteria were less than one month since trauma, serious brain- or other organic disease, serious head injury, chronic psychotic disorder, serious alcohol or substance abuse (Alcohol Use Disorders Identification Test (AUDIT) >20 or Drug Use Disorder Identification Test (DUDIT) >25) and serious dyslexia or other serious difficulties with oral or written language. The most common exclusion criterion was lacking satisfactory proficiency in Scandinavian or English.

The study design was approved by the Regional Ethical Committee in South-Eastern Norway (REK Sør-Øst) in January 2016. A total of 119 patients read and signed a written consent to participate between 9 February 2016 and 22 June 2018. Two patients withdrew from the study before starting the study examinations and seven during the interviews or before the physical examination and blood tests. No patients have withdrawn their consent after having completed their study participation.

Finally, 110 patients were included in the study. Forty two participants were men (38.2%) and 68 were women (61.8%). The patients were divided into the three groups for comparison and analysis. Ethnic Norwegians (abbreviated NOR) were defined as being born in Norway and both parents also being born there. Immigrants (IMI) were defined as being born in any other county or being born in Norway but both parents being born in another country. We also included one patient in this group who was born and raised in another country and moved to Norway because one parent was born here.

Refugees often have a more forced reason for migration. We therefore separated this group from the rest of the immigrants. We also included one patient who was granted residency on humanitarian grounds. We abbreviate this group as RFG. The IMI group included 15 participants who came to Norway for family reunion. Some of these could have reunited with a refugee and could possibly have ties to refugee status. However, they might not have as forced and difficult emigration situation as the refugees themselves and we wished to separate only the refugees themselves, so all the family reunified participants were categorized in the IMI group. The NOR group consisted of 44 patients (40%). The RFG group had 25 (22.7%) patients and IMI group 41 (37.3%).

Instruments for assessment

The psychiatric examination consisted of two parts. First, the patients received a set of self-reported questionnaires to be completed at home or in the waiting room. The handout included a general questionnaire on socio-demographic data and several verified self-report instruments. Alcohol and drug use were assessed by using World Health Organization (WHO)’s AUDIT and DUDIT. These tests can be used to determine if the patient suffers from alcohol use disorder (abuse or dependence) or substance use disorder (abuse or dependence) [24].

To assess current symptom level, we used the Symptom Checklist 90- Revised (SCL-90-R) [25]. A higher score indicates more psychiatric symptoms. To assess the degree of current somatic pain symptoms in more depth we used ‘Norsk Smerteforenings Minimumsskjema’ (NOSF-Miss). This is a validated comprehensive pain assessment tool developed by the Norwegian Society for Pain Medicine [26]. A higher score indicates more pain and dysfunction due to pain.

The patients then met one of the two experienced psychiatrists for the clinical (semi)-structured interview part of the study. Firstly, they were examined using MINI Plus International Neuropsychiatric Interview 5.0.0 [27]. The psychiatrist then made a clinical decision on diagnose(s) according to DSM-IV-TR as appropriate.

Then, the life events checklist (LEC) was used to thoroughly review all traumatic events throughout their life. This validated instrument also evaluates each traumatic event’s impact and helps the clinician select the index trauma which has the highest impact on the patients’ life [28]. The index trauma was used in the structured clinical interview for DSM-IV-TR PTSD module (SCID-I PTSD) to assess current and past PTSD diagnosis.

These instruments have been validated and used in Norway for many years [29,30]. We assessed current psychiatric comorbidity by combining the results of MINI Plus, SCID-I-PTSD, AUDIT and DUDIT. The psychiatrists also assessed the participants on the two global assessment of functioning scales (GAF function and GAF symptom) to assess overall symptom level and functioning [31]. A higher score on this scale indicates better functioning or lower symptom load.

Statistical analysis

The data were input in IBM SPSS statistical program version 25. We analyzed using one-way analysis of variance (ANOVA) and Pearson’s Chi-square tests for continuous and categorical variables, respectively. One-way between-groups ANOVA was conducted with post-hoc comparisons using the Turkey HSD test if we found a statistically significant difference among groups. If Levene’s test for homogeneity of variances were <0.05, we used Welche and Brown-Forsyhe tests to compute p values. The non-parametric Kruskal–Wallis test was used when normal distribution of continuous scores was not assumed. In Chi-square test we used adjusted residual, acting like a post-hoc test helping to identify between groups differences. Effect size (ES) was judged using Cramer’s V. The alpha level was set at p < 0.05. Beyond statistical significance we also commented on possible clinical relevance of trends that were not statistically significant.

Results

Country of origin of the immigrants

The participants originated from all continents. The most common origin in the RFG group was the Middle East and North-Africa (46%). A large portion also originated from Eastern Europe (25%), the remaining from Southern/Central Asia (13%), Southern/Eastern Africa (12%) and South-East Asia (4%). In the IMI group most also came from the Middle East and North-Africa (31%). The remaining were quite evenly distributed between origins from Europe, Africa and Asia, and only three patients came from Central/South America and Oceania combined.

Of the IMI group, more than half of them came to Norway for family reunion, 22% were born in Norway of immigrant parents and 17% were migrant workers.

Sociodemographic characteristics

The sociodemographic characteristics of the study subjects are summarized in Table 1 below.

Table 1. Socio-demographic characteristics and smoking habits in the clinical sample of 110 traumatized patients in Southern Oslo.

Socio-demographic parameters	NOR (n = 44)	RFG (n = 25)	IMI (n = 41)	Total (n = 110)
Age, years^a, mean (SD)	37.81 (12.4)	44.72 (11.0)	39.69 (8.7)	40.4 (11.1)
Age groups, n (%)
18–30 years	15 (34.1)	3 (12.5)	7 (16.7)3	25 (22.7)
31–45 years	14 (31.8)	9 (37.5)	23 (54.8)	46 (41.8)
46–61 years	15 (34.1)	12 (50)	12 (28.6)	39 (35.5)
Gender, n (%)
Female	31 (70.5)**	8 (32)^††,‡‡	29 (70.7)**	68 (61.8)
Male	13 (29.5)**	17 (68)^††,‡‡	12 (29.3)**	42 (38.2)
Marital status, n (%)
Married/cohabitant	21 (47.7)	18 (72)^†	16 (39)*	55 (50)
Any other	23 (52.3)	7 (28)^†	25 (61)*	55 (50)
Number of children, n (%)
0	25 (56.8)^**,††	2 (8)^††,‡‡	13 (31.7)^**,‡‡	40 (36.4)
1–2	14 (31.8)	12 (48)	18 (43.9)	44 (40)
≥3	5 (11.4)**	11 (44)^‡‡	10 (24.4)	26 (23.6)
Number of close friends, n (%)
0	10 (22.7)	3 (12)	12 (29.3)	25 (22.7)
1–3	23 (52.3)	19 (76)	20 (48.8)	62 (56.4)
≥4	11 (25)	3 (12)	9 (22)	23 (20.9)
Education years, n (%)
≤13	23 (52.3)	17 (68)	21 (51.2)	61 (55.5)
≥14	21 (47.7)	8 (32.0)	20 (48.8)	49 (44.5)
Job any, n (%)
Yes	21 (47.7)	8 (32.0)	10 (24.4)	39 (35.5)
No	23 (52.3)	17 (68.0)	31 (75.6)	71 (64.5)
Use of welfare system, n (%)
None	6 (13.6)	6 (24)	3 (7.3)	15 (13.6)
Fully paid sick leave	10 (22.7)	4 (16)	7 (17.1)	21 (19.1)
Temporary benefit, AAP	22 (50)	12 (48)	22 (53.7)	56 (50.9)
Any other	6 (13.6)	3 (12)	9 (22)	18 (16.4)
Current daily smoking, n (%)	12 (27.3)*	10 (40)^‡	14 (34.1)	36 (32.7)

NOR: ethnic Norwegian patients; RFG: refugee patients; IMI: immigrant patients without refugee background; AAP: ‘Arbeidsavklaringspenger’; Any other benefit: permanent benefit (‘uføretrygd’), unemployment benefit, social benefit and maternal/paternal benefit.

^a Scores was not normal distributed; the non-parametric Kruskal–Wallis test was used.

*p < 0.05; **p < 0.01 (compared with refugee patients).

^†p < 0.05; ^††p < 0.01 (compared with non-refugee immigrants).

^‡p < 0.05; ^‡‡p < 0.01 (compared with Norwegian).

In the NOR and the IMI groups, the number of women was higher than the number of men, the opposite was true in the RFG group (ES = 0.33). The participants in the RFG group were most married (72%) while the IMI group patients were more often divorced (ES = 0.25). The NOR group had fewest children. The RFG patients had the highest number of children (ES = 0.30) and IMI in between. The NOR group had fewer daily smokers than RFG.

There were no statistically significant differences between the groups on age, social network (reported number of close friends), education, employment status and use of welfare system. There was a tendency, however, that patients in the RFG group were older, had smaller social network, lower education level and receiving the (lowest) social benefit, than the two other groups.

RFG were, however, mostly married or living in a cohabitant relationship whereas IMI most often lived without a spouse. NOR patients tended to be more often employed and/or receiving the fully paid sick-leave.

Types of traumatic events

Of the index trauma type identified during the LEC interview, the NOR patients had mostly been exposed to physical-(39%) or sexual assault/rape (25%). The RFG group’s most common index trauma was combat or war-zone experience (33%) or physical assault (17%). The IMI group patients had mostly been exposed to sexual assault/rape (26%) or physical assault (21%). Only one person in the RFG group (4%) reported torture as index trauma, but 29% reported experiencing torture as non-index trauma. One person (2%) in the IMI group had experienced torture, none in the NOR group. Most of the participants reported their index trauma as having been repeated, continuous or longstanding. In the NOR group this was 75%, RFG 76% and IMI 63.4%. The differences were not statistically significant, and the average was 70.9%.

Symptom burden and functional impairment

We found some differences in the subjective symptoms obtained from the self-report on the SCL-90-R between the groups. NOR tended to score lower (better) on most scales and subscales of the instrument including somatization, compulsivity, depression, anxiety, hostility, paranoia and global severity index. None of these were statistically significant, though. RFG scored highest (worst) on all subscales but the difference was only significantly higher on the Psychotic delusions and other symptoms subscales. On the NOSF-Miss pain instrument the IMI group scored highest, however. The difference was large and significantly higher (worse) than the NOR group (ES = 0.33).

Some of these differences in self-reported symptom load were reaffirmed by the researcher’s assessment of Global symptom level and functioning. The NOR group scored highest (best) on both the GAF-F (function measure) and GAF-S (symptom measure). The RFG group scored significantly lower than the two other groups on GAF-S and this coincides with the findings from SCL-90-R. The results are summarized in Table 2.

Table 2. Self-report psychometric instruments: symptoms according to SCL-90-R, pain according to NOSF-MISS and GAF-score in a clinical sample of 110 traumatized patients in Southern Oslo.

Self-report psychometric instruments	NOR (n = 44)	RFG (n = 25)	IMI (n = 41)	Total (n = 110)
Average SCL-90-R scores (range 0–4)
Somatization^a	1.72	2.16	2.12	1.97
Compulsive tendency^a	2.02	2.43	2.22	2.19
Interpersonal hypersensitivity	1.89	2.19	1.85	1.94
Depression^a	2.16	2.34	2.24	2.23
Anxiety	1.76	2.19	1.92	1.91
Hostility^a	0.82	1.29	1.00	0.99
Phobia	1.46	1.77	1.43	1.52
Paranoid delusions	1.42	1.85	1.64	1.59
Psychotic delusions	0.82*	1.29^‡	1.03	1.01
Other symptoms	1.95*	2.42^‡	2.08	2.11
Global Severity Index	1.66	2.04	1.82	1.80
NOSF-MISS pain score (range 0–100)
Normalized total pain score ratio, mean^a	31.95^†††	42.51	50.41^‡‡‡	41.23
Global assessment of function score (range 0–100)
GAF-function^a, mean	59.18	54.84	54.85	56.58
GAF-symptoms^a, mean	59.43*	54.00^†,‡	58.00*	57.66

SCL-90-R: The Symptom Checklist-90-R; NOR: ethnic Norwegians; RFG: refugee patients; IMI: immigrants without refugee background; NOSF-MISS: ‘Norsk smerteforenings minimumsspørreskjema for smertepasienter’ - Norwegian pain association’s minimum pain questionnaire for pain patients; GAF: global assessment of functioning.

^aScores was not normal distributed; the non-parametric Kruskal–Wallis test was used.

*p < 0.05 .

^†p < 0.05; ^†††p < 0.001 (compared with IMI).

^‡p < 0.05; ^‡‡‡p < 0.001 (compared with NOR).

Psychiatric disorders

Rates of the different psychiatric disorders in the three groups are presented in Table 3.

Table 3. Rates of current and previous disorders according to MINI Plus, SCID-I-PTSD, AUDIT and DUDIT of the clinical sample of 110 traumatized patients in Southern Oslo.

Psychiatric disorders, n (%)	NOR (n = 44)	RFG (n = 25)	IMI (n = 41)	Total (n = 110)
Major depression current, n (%)	13 (29.5)***	17 (68)^†††,‡‡‡	11 (26.8)***	41 (37.3)
Major depression past, n (%)	37 (84.1)	22 (88)	38 (92.7)	97 (88.2)
PTSD current, n (%)	17 (38.6)*	17 (68)^†,‡	17 (41.5)*	51 (46.5)
PTSD past, n (%)	39 (88.6)	22 (88)	37 (90.2)	98 (89.1)
Any current anxiety disorder, n (%)	31 (70.5)^†	17 (68)	25 (61)^‡	73 (66.4)
Any past anxiety disorder, n (%)	39 (88.6)	20 (80)	31 (75.6)	90 (81.8)
Agoraphobia current, n (%)	16 (36.4)^†	13 (52)^†	7 (17.1)^*,‡	36 (32.7)
Agoraphobia past, n (%)	25 (56.8)	15 (60)	16 (39)	56 (50.9)
Social phobia current, n (%)	22 (50)	7 (28)	13 (31.73)	42 (38.2)
Social phobia past, n (%)	25 (56.8)^†	9 (36)	14 (34.1)^‡	48 (43.6)
Specific phobia current, n (%)	7 (15.9)	6 (24)	8 (19.5)	21 (19.1)
Panic disorder current, n (%)	13 (29.5)	7 (28)	6 (14.6)	26 (23.6)
Panic disorder past, n (%)	30 (68.2)^†	14 (56)	17 (41.5)^‡	61 (55.5)
General anxiety disorder current, n (%)	10 (22.7)	7 (28)	9 (22)	26 (23.6)
OCD current, n (%)	7 (15.9)	4 (16)	6 (14.6)	17 (15.5)
Somatization disorder current, n (%)	14 (31.8)	6 (24)	14 (34.1)	34 (30.9)
Somatization disorder past, n (%)	18 (40.9)	7 (28)	16 (39.0)	41 (37.3)
Hypochondriacal disorder current, n (%)	2 (4.5)*	5 (20)^‡	4 (9.8)	11 (10)
Dysmorphophobia current, n (%)	9 (20.5)	5 (20)	4 (9.8)	18 (16.4)
Somatoform pain disorder current, n (%)	9 (20.5)^*,†	11 (44)^‡	18 (43.9)^‡	38 (34.5)
Alcohol use disorder (AUDIT > 7), n (%)	16 (36.4)*	2 (8)^‡	9 (22)	27 (24.5)
Drug use disorder (DUDIT > 2F and >6M)	11 (25)*	1 (4)^‡	4 (9.8)	16 (14.5)

NOR: ethnic Norwegian patients; RFG: refugee patients; IMI: immigrant patients without refugee background; OCD: obsessive compulsive disorder; PTSD: posttraumatic stress disorder. GAD: general anxiety disorder; AUDIT: Alcohol Use Disorders Identification Test; DUDIT: Drug Use Disorders Identification Test; F: female; M: male.

*p < 0.05; ***p < 0.001 (compared with RFG).

^†p < 0.05; ^†††p < 0.001 (compared with IMI).

^‡p < 0.05; ^‡‡‡p < 0.001 (compared with NOR).

Anxiety disorders, PTSD, major depressive disorder and somatoform pain disorder were the most common current psychiatric disorders. Of the participants 66.4% had a current anxiety disorder, 46.5% had a current PTSD diagnosis, 37.3% had a major depressive disorder and 34.5% had somatoform pain disorder. When assessing past disorders, the rates were higher, 81.8% for anxiety disorders, 89.1% for PTSD and 88.2% had a past major depressive disorder.

A past disorder included both cases where there were still symptoms but currently below the threshold for diagnosis, and where the instruments in the study and clinical evaluation indicated a past/lifetime disorder from recollection and/or journal reports. Results based on past diagnoses should therefore be treated with more caution than the current diagnoses.

The types of disorders had significant differences between the groups. The NOR patients had significantly higher rates of current agoraphobia (ES = 0.29), past social phobia (ES = 0.22) and past panic disorder (ES = 0.24) than patients in the IMI group. When all anxiety disorders were pooled together, NOR had higher rates than IMI although the effect size was relatively small (70.5% and 61% respectively, ES = 0.18).

Patients in the RFG group clearly had higher rates of a current major depressive disorder than both the NOR and the IMI patients (68vs. 29.5 and 26.8%, respectively, ES = 0.35). They were also significantly more likely to suffer from current agoraphobia than the IMI group (ES = 0.29). Hypochondriacal disorder was most common in the RFG group and significantly higher than in the NOR group. Somatoform pain disorder was more common in both the immigrant groups (RFG and IMI) than in the NOR group (ES = 0.24).

The NOR group had the lowest rates of post-traumatic stress disorder (PTSD). The RFG group had highest rate and IMI’s rates were close to the NOR group. The difference was significantly higher in the RFG group than in the NOR and IMI groups, (68 vs. 29.5 and 26.8% respectively, ES = 0.24).

The NOR patients had significantly higher rates of alcohol use disorder and drug use disorder than the RFG group. The IMI group had a rate in between the two other groups.

Comorbidity between psychiatric disorders

Comorbidity between the most common disorders are listed in Table 4.

Table 4. Comorbidity of the different psychiatric disorders in the clinical population of 110 patients with trauma background in Southern Oslo.

Psychiatric disorders	NOR (n = 44)	RFG (n = 25)	IMI (n = 41)	Total (n = 110)
At least 1 current disorder, n (%)	40 (90.9)	22 (88)	32 (78)	94 (85.5)
At least 2 current disorders, n (%)	29 (65.9)	22 (88)	25 (61)	76 (69.1)
≥3 current disorders, n (%)	23 (52.3)*	18 (72)^‡,†	19 (46.3)*	18 (16.4)
PTSD + depression (current), n (%)	8 (18.2)**	14 (56)^††,‡‡	10 (24.4)**	32 (29.1)
Depression + any current anxiety disorder, n (%)	12 (27.3)	12 (48)^†	10 (24.4)*	34 (30.9)
PTSD + any anxiety disorder (current), n (%)	15 (34.1)	13 (52)	15 (36.6)	43 (39.1)
PTSD + panic disorder (current), n (%)	10 (22.7)^†	5 (20)	3 (7.3)^‡	18 (16.4)
PTSD + social phobia (current), n (%)	3 (6.8)*	6 (24)^‡	6 (14.6)	15 (13.6)
PTSD + somatoform pain disorder (current), n (%)	4 (9.1)*	8 (32)^‡	10 (24.4)	22 (20)
PTSD + alcohol use disorder	7 (15.9)	1 (4)	3 (7.3)	11 (10)
Depression + alcohol use disorder	4 (9.1)^†	2 (8)	0 (0)^‡	6 (5.5)

NOR: ethnic Norwegians; RFG: refugee patients; IMI: immigrants without refugee background; PTSD: posttraumatic stress disorder; GAD: general anxiety disorder.

*p < 0.05; **p < 0.01.

^†p < 0.05; ^††p < 0.01.

^‡p < 0.05; ^‡‡p < 0.01.

Of the participants 9.1% in the NOR group and 12% in the RFG group did not have any current psychiatric disorder. In the IMI group 22% were not diagnosed with any current disorder, but this difference was not statistically significantly higher than the other two groups.

All the participants in the RFG group who were diagnosed with one psychiatric disorder also had at least one other comorbid disorder. Seventy percent of the RFG group were diagnosed with three or more comorbid disorders. This rate was significantly higher than in the NOR (52.3%) and IMI groups (46.3%).

On a whole, PTSD and depression were the most common co-occurring pair of diagnoses in the clinical population. More than half of the RFG patients had this combination and they were significantly more often diagnosed with these two together than the other two groups. RFG also more commonly had depression comorbid with any anxiety disorder (48%) compared with IMI (24.4%). RFG patients were also more likely to have social phobia and somatoform pain disorder comorbid to current PTSD compared to the NOR group. The combination PTSD and current panic disorder, however, was more common in the NOR group than in the IMI group. NOR more commonly had depression and alcohol use disorder in combination than IMI. No patients in the IMI group had this comorbidity.

Discussion

Disorders after trauma

This clinical study included patients who were referred to our clinic and had a history of trauma. Trauma and trauma related symptoms were, however, not necessarily the main reason for referral. In many of the cases the major traumatic event was uncovered during the treatment. This finding in itself will indicate that traumatic events can be challenging for patients to talk about. Some reasons may be avoidance of painful memories, feelings of shame or fear of stigmatization.

Trauma may play an important role in many different types of disorder and it may be important to explore possible traumatic events in many different clinical settings. We found that a serious traumatic event may be related to a wide variety of symptoms and disorders, not only symptoms directly related to the trauma like in PTSD.

Other psychiatric disorders than trauma disorders may follow trauma exposure, including alcohol and substance abuse, somatoform and pain disorders, different anxiety disorders and major depressive disorder. The latter has in some studies been found to be almost as common as PTSD after major trauma [32]. Our overall findings were similar to the literature as major depressive disorder, anxiety disorders and somatoform pain disorder were common in addition to PTSD. A previous cross-sectional study from Norway has found particularly high rates of comorbid depression with PTSD in multi-traumatized refugees in outpatient clinics [30]. Similar results have been found internationally [33]. In our study, however, we were able to compare the types of disorders among the refugees with those of the native Norwegian population and other immigrants in the same clinical setting. We did find differences in comorbidity between the groups.

Refugees

Having to flee from your home country is known to be a large risk factor for many mental health problems and refugees have been shown to have more mental health problems than other immigrants [34]. Many have experienced war and torture, which are important risk factors. The traumatized refugee patients in our study (RFG) had the highest rates of depression and PTSD. More than half of them had PTSD comorbid with a depression diagnosis. In addition, RFG patients had significantly more often current agoraphobia, hypochondriacal disorder and somatoform pain disorders. They also more commonly had depression comorbid with any anxiety disorder compared with both the other groups (NOR and IMI) and were also more likely to have social phobia and somatoform pain disorder comorbid to current PTSD than the NOR group.

Most of the patients in the RFG group were diagnosed with three or more comorbid disorders. Self-reported symptoms (GAF-symptoms and SCL-R-90) confirmed that these patients experienced higher burden of psychiatric symptoms compared to the other groups. They also scored significantly worse on the Psychotic delusion subscales. This could possibly be related to more severe forms of PTSD with secondary psychotic symptoms and dissociative symptoms. Many of the refugees came from the Middle East and cultural factor may also play a role.

Ethnic Norwegians and non-refugee immigrants

The native Norwegian patients in the study (NOR) had a tendency towards higher rates of social phobia and panic disorder, and of having any anxiety disorder. They also had higher current comorbidity of PTSD with panic disorder and combination of depression and alcohol use disorder. Traditional Norwegian culture has been known for suspiciousness of strangers and anxiety and this may explain this finding [35]. The prevalence of anxiety disorders also seems to be higher than other countries, especially in the younger generation [36]. The NOR group also had lower rates of somatoform pain disorder than both the immigrant groups. Somatic pain has been shown to be a more common presentation of mental illness in non-Western societies [37]. This has repeatedly been shown to be particularly evident in refugees re-settled in Western countries [29,38,39]. In our study, however, there was not a statistically significant difference between pain in the RFG and IMI group, but interestingly the IMI group tended to have more pain and pain disorders.

Alcohol and drug use may also be related to culture and religion. Ethnic Norwegians in Norway have been shown to have higher consumption rates than immigrants but lower rates than the rest of Europe [40,41]. The difference between immigrants and native Norwegians seems to be even more pronounced after trauma, possibly indicating a coping mechanism related to ethnic background.

Overall rates of disorders

A significant portion of the patients did not have any of the diagnoses we assed. There are several possible explanations for this. Many of the patients were already receiving treatment and some could have remitted before entering the study or could have subthreshold symptoms. Furthermore, some could have other disorders that were not evaluated in the study, such as personality disorders. It could also be the case that some of the patients were referred to the clinic for other reasons than a manifest mental illness, such as a requirement for welfare benefits or because of a request from family or a primary care physician.

The non-refugee immigrants seemed to have the lowest rates of many disorders including of having any anxiety disorder. Fewer of them had repeated/longstanding trauma, but the difference was not statistically significant. Some other possible explanations for this including social network and religion are discussed below.

Possible explanations

Each individual has a unique history and background with risk and resilience factors and each experience of a traumatic event is unique to that individual. Immigration status seems to be one of several important factors in this respect. Family, social network, education, employment and gender are other important factors. Unemployment, weak social network and weak social integration have been shown to be correlated to psychopathologic load in refugees in Norway [30]. Our findings indicate that some of these factors may differ between ethnic Norwegians, refugees and other immigrant patients. These psychosocial factors could possibly explain some of the observed differences in our study.

Gender

Gender distribution in a clinical sample can create a sex bias. In our sample there were significantly more men in the RFG group than in the other two groups. It is well known that women of reproductive age are about twice as likely to develop anxiety, depression and PTSD than men of the corresponding age. This is probably in part related to biological factors [42–44]. In our study, however, the group with most men also had the highest burden of symptoms and mental illness.

There are several possible explanations for this disparity. Different trauma burden and severe psychosocial stressors under forced migration may overweigh the possible protective effects of being male. Another possibility may be that although the threshold for getting a disorder may be higher for men, once this threshold has been exceeded, the symptom load may be less different than that of women.

Effects of migration and integration

Moving to a new country and culture is known to be a psychosocial stressor and risk factor for developing mental illness, also after trauma. This has also been found in other Nordic countries [45,46].

In our sample, however, the non-refugee immigrants (IMI) did not have more mental illness than comparable ethnic Norwegians (NOR). One possible explanation is so-called relative socio-economic status. Norwegians living in this area are often in relatively lower socio-economic status compared with other Norwegians living in other parts of Oslo. Conversely, the immigrants living in this area often are of higher socio-economic status compared with a corresponding population in their native home country. It often takes a lot of resources to be able to move from a developing country to a developed country.

Furthermore, since inclusion in our study required proficiency in English or Norwegian, some of the less well integrated immigrants, possibly with more mental illness, might have been excluded. Finally, it could be that Norway’s introductory programs are more generous and efficient than comparable Western countries and a better integration of first generation immigrants may have led to fewer psychiatric disorders.

Socio-demographic differences

Some of the variations can possibly be explained by demographic characteristics. The NOR group drank more alcohol than the non-Norwegians, a clear risk factor. Lack of social network is a known risk-factor for developing mental illness after trauma [47]. The refugees (RFG) had more children, were more often married and had a tendency towards having more friends. These factors did not seem to be protective for this group in our sample, however. When comparing the non-refugee immigrants (IMI) to the ethnic Norwegians (NOR), living with more children could be a protective factor. Having children can possibly give purpose in life and they can be a positive distraction from mental and social problems.

The healthy immigrant effect

It could also be possible that non-refugee immigrants in Western countries have less mental illness than their corresponding native population. The ‘healthy immigrant effect’ has been established in the literature in later years. Larger population-based studies from Canada have found that the mental health of non-Western immigrants tends to be better than that of the general population in both the sending and receiving countries [48,49]. This is contrary to many studies in the USA, UK and EU where immigrants have generally been found to have more mental illness. Immigrants to Canada have shown slightly lower rates of mental disorders than the general population. They had particularly lower rates of depression and alcohol dependence compared with the ethnic Canadian population [50,51]. Even in the Canadian sample, however, the health of the immigrants tended to worsen over time to match that of the general population in the third generation [52]. Some of the variations in this study are attributed to positive selection in work-immigrant populations. This was not the case in our study, however, since only 17% came as migrant workers. Our findings can shed more light on these disparities since refugee status clearly separated them from the other immigrants as being a major risk factor for mental illness after trauma.

The non-refugee immigrants did not have overall higher rates of mental illness than the ethnic population in our sample. This seems contrary to many studies in comparable European countries. Psychosocial factors, cause of move to Norway and alcohol/substance use or referral practices may explain some of the variations but cultural and religious factors probably also play an important role. These findings suggest that politicians, health care providers and clinicians in the mental health field should be aware of ethno-racial differences in the presentation of mental illness after trauma.

Limitations

This study has several limitations. The study has a cross-sectional design, without a control group and excluding patients without a major trauma experience and persons not referred to specialized treatment may limit the generalizations of the findings. Moreover, for practical reasons and because we lack validated instruments in many languages patients who could not speak Norwegian or English were excluded from our study. This may cause an obvious selection bias. The actual number of participants who were excluded due to language was low, however. In general refugees and other immigrants tend to use less health care services than natives. This is particularly evident in mental health care and among those who do not speak the local language [53]. We thus believe that this selection bias is relatively minor and that the results largely are representative of the clinical sample.

Furthermore, even though the participants all spoke English or a Scandinavian language, the definition of some terms and concepts may vary between cultures and subcultures, for instance what can be defined as ‘a close friend’. Time since arrival in the host country is also known to influence psychological distress [54], but this information is unfortunately not known in the sample. The IMI group consisted of participants with many reasons for moving to Norway and is therefore heterogeneous.

All the patients came from the same outpatient clinic. This can be both strength and a weakness. It may cause stronger validity of the local population, but lower with other, less similar, clinical populations.

Moreover, we have carried out multiple tests with many variables in our analyses on a relatively limited sample. This causes some statistical limitations. Generalization to other populations may be limited. However, most of the similar instruments showed consistent inter-test results indicating that the multiple tests problem may be relatively lower in this study. For instance, those who had somatic pain syndrome also scored high on all the other pain measurements.

Sample selection

The participants in the study had gone through several selection processes before being included in the results. There is a possibility of selection bias in several of the steps. The patients had to seek primary care and the physician there had to find that they required referral to specialized treatment. It could for instance be that refugees are less aware of health care services or may be reluctant to seek treatment, or primary care physicians might less familiar with their way of suffering. This could possibly explain why we found refugees to have a higher psychopathological load.

Conclusion

The 110 traumatized patients in this study had a wide variety of different disorders. There was a difference in the rates of some psychiatric disorders between the ethnic Norwegian, refugee and other immigrant patients. Norwegians had more anxiety disorders while the immigrants had more pain related disorders. The refugees had higher rates of PTSD and depression. It is important for clinicians to recognize traumatic events in their patients history and that there may be differences in presentation of disorders after trauma.

Acknowledgments

The authors thank The University of Oslo, Oslo University Hospital and SOD for all the practical support for the study. The authors thank all the participants and therapists for their interest in the study. The authors also thank the research group on traumatic stress, forced migration and global mental health in Oslo for all the discussions before, during and after the study.

Disclosure statement

The authors reported no conflicts of interest. The authors alone are responsible for the content and writing.

Additional information

Funding

The financial support of the project was provided by the Regional Health Authority of South-Eastern Norway and Oslo University Hospital.

Notes on contributors

Erik Ganesh Iyer Søegaard

E.G.I.S. is a psychiatrist at Oslo University Hospital in Norway. He is a PhD candidate at the University of Oslo and the head of department at Southern Oslo District Psychiatric Centre.

Zhanna Kan

Z.K. is a medical doctor in specialization in psychiatry at Oslo University Hospital in Norway. She has a PhD in cardiology.

Rishav Koirala

R.K. is a psychiatrist at Brain and Neuroscience Center, Nepal, Kathmandu, Nepal. He is a PhD candidate at the University of Oslo.

Edvard Hauff

E.H. is a psychiatrist in private practice. He is professor emeritus at the University of Oslo.

Suraj Bahadur Thapa

S.B.T. is a psychiatrist at Oslo University Hospital in Norway. He is an assistant professor at the University of Oslo.