Mental health problems of children and adolescents, with and without migration background, living in Vienna, Austria

For the present study, the statistical analysis was conducted with IBM SPSS Statistics 21.0. We calculated descriptive parameters, and used ANOVA and MANOVA for analysing differences in parametric data. The “Chi-Quadrat-Tests” and “Fisher´s exact tests” were used for non-parametric data, assuming significant differences at an alpha level ≤ 0.05.

302 children participated in the study: 152 were native Austrians, and 150 were children and adolescents with a Turkish migration background. The mean age of the whole study sample was 11.7 ± 3.1 (SD) years. The clinical group consisted of 200 patients, the control group of 102 subjects. These groups were divided according to their age in two groups, 7–11 years and 12–18 years (Table 1).

Table 2 lists the clinical diagnoses of the clinical groups and their sex distribution.

Analysing differences in the distribution of diagnoses among children and adolescents with and without migration background, we observed significant differences for gender (Fisher’s Exact Test = 17.20, p = 0.01) and migration background (Fisher’s Exact Test = 18.38, p < 0.01). The clinical and control groups differed particularly in their distribution of the F9 diagnoses. F9 diagnoses were more frequently observed in males (44.4%) than in females (25.8%), and more frequently in Austrian children (42.1%) than in the Turkish-speaking sample (28%). F8 diagnoses were observed only in the Turkish-speaking sample. Age dependency was observed only as a tendency (Fisher’s Exact Test = 11.84, p = 0.06; Table 2).

The mean externalizing problems scale score was 13.95 (SD = 12.47).

Significantly more (F (1, 286) = 11.14, p < 0.01) native patients scored higher on the externalizing problems scale (M = 20.92) than the Turkish-speaking patients (M = 15.53).

In the clinical group, the Austrian patients (M = 20.92) scored higher on the externalizing problems than the Turkish patients (M = 7.11) and the control groups did. Turkish-speaking patients (M = 15.53) had significantly more externalizing problems than native (M = 4.01) and Turkish speaking (M = 7.11) children and adolescents of the control groups (Table 4).

We found significant gender differences for externalizing problems (F (1, 146) = 4.84, p = 0.03). Significantly more males (M = 14.88) than females (M = 12.00) had externalizing problems.

The mean score of internalizing problems was 15.34 (SD = 9.45) and the mean score of externalizing problems was 14.33 (SD = 8.05). The multivariate analysis showed significant differences for gender (F (2, 145) = 3.22, p = 0.04) and study group (F (2, 145) = 9.44, p < 0.01).

The influence of the migration background was not significant (F (2, 145) = 2.62, p = 0.08). The univariate analysis showed that males reported significantly more externalizing problems than females (M = 14.88 vs. M = 12.00, F (1, 146) = 4.84, p = 0.03).

The univariate analysis showed significant effects for both scales of internalizing (F (1, 146) = 7.18, p = 0.01) and externalizing problems (F (1, 146) = 16.89, p < 0.01). The clinical sample reported significantly more internalizing (M = 16.67) and externalizing (M = 16.13) problems than the controls (internalizing problems: M = 12.41, F = 7.18, p = 0.01; externalizing problems: M = 10.74, F = 16.89. p < 0.01).

The clinical group presented significantly (F (1, 286) = 157.77, p < 0.01) more behavioural problems (M = 55.33) than the control group (M = 18.37).

The multivariate analysis showed significant differences for behavioural problems between nationality and study group (F (1, 286) = 11.33, p < 0.01; Turkey’s HSD = 10.11).

The univariate analysis showed that, native children and adolescents (M = 11.70) had significantly lower scores in behavioural problems than their Turkish-speaking (M = 25.03) peers as well as Austrian (M = 58.57) and Turkish-speaking (M = 52.09) patients (Table 5).

Furthermore, Austrian and Turkish-speaking minors in the clinical groups showed significantly higher scores for behavioural problems compared to the control groups.

We found no significant gender differences for behavioural problems between the two study groups.

We found significant differences for depression between the two study groups (F (1, 286) = 30.61, p < 0.01) and between the two age groups (F (1, 286) = 5.24, p = 0.02).

The clinical sample (M = 14.36) had higher depression scores than the control sample (M = 9.38). Furthermore, depression was significantly more frequent among older children and adolescents (M = 12.89) in comparison to the younger (M = 10.84) sample.

We observed significant interactions between depression score and migration background (F (1, 286) = 4.01, p = 0.05; Turkey’s HSD = 3.08).

Austrian (M = 14.61) and Turkish-speaking (M = 14.10) clinical samples had higher mean depression scores than both control groups. Whereas children of the control group with a Turkish migration background (M = 10.93) compared to Austrian control children (M = 7.83) exhibited significantly higher mean depression scores (Table 6).

We found no significant gender differences for depression between the two study groups.

The multivariate analysis showed significant differences for anxiety between nationality (F (2, 285) = 3.63, p = 0.03), the study groups (F (2, 285) = 5.64, p < 0.01) and age groups (F (2, 285) = 5.82, p < 0.01). The group with a Turkish migration background (M = 51.21) presented higher state-anxiety scores than the Austrian group (M = 48.13).

The univariate analysis showed that Austrian subjects (M = 44.40), females (M = 45.95) and the younger group (M = 44.02) of the control sample had lower state-anxiety scores in comparison to the other groups (Table 7).

Both, parental CBCL and self-assessed YSR scores were higher for internalizing and externalizing problems compared to the respective control groups.

Janssen et al. [6] reported higher behavioural problem scores for internalizing problems among youths with Turkish migration background when compared to Dutch peers. Our results support these findings. We speculate that Turkish-speaking minors are more exposed to external problems like migration-related stress, and maybe do not seek adequate professional help in time. In addition, migrant children tend to hide their problems in order not to appear feeble.

Austrian psychiatric patients exhibited higher scores for externalizing problems than their Turkish-speaking peers. Gao et al. [5] found in a Chinese sample, that children with migration background had more internalizing and externalizing mental health problems compared to their native peers. This difference may probably be explained by population-based differences in education standards and the related social behaviour.

Darwish Murad et al. [57] reported results that did not differ from our analyses. Turkish speaking adolescents reported more problem behaviour such as internalising problems than their Dutch peers. This ethnic difference between adolescents with and without migration background, may also be due to divergences in their socio-economic status [57].

The psychiatric diagnoses F8 (disorders of psychological development) were present only among Turkish-speaking patients. Whereas, the psychiatric disorder, F9 (behavioural and emotional disorders with onset usually occurring in childhood and adolescence) were diagnosed more frequently among native patients compared to the Turkish-speaking sample.

Akkaya-Kalayci et al., Ceri et al. and Webb et al. [24, 25, 31] also reported that specific psychiatric diagnoses were more commonly observed among migrants compared to the indigenous population. The difference may also be explained by population-based differences. On the other hand, ADHD may have a civilizing background.

In accord with previous studies [54, 58], children and adolescents in our study had higher depression scores compared with children and adolescents of the control sample. Similar to studies [7, 9, 16, 59], we found no gender or age-related differences in our study sample but an age-dependent increase of depression symptoms.

Children of the Austrian control sample exhibited lower depression scores when compared to the Turkish control subjects and the Austrian and Turkish patients.

In line with our results, higher prevalences of depression [6, 60] and anxiety [31, 61, 62] have been reported among migrant populations. The higher depression scores may be interpreted as migration-stress related even in the 2nd and 3rd generations (LIT).

Anxiety scores were elevated in both clinical groups, with no age or gender-related differences. Age and gender-related differences were only seen in the control groups. Moreover, Austrian and Turkish-speaking control groups differed in their anxiety levels. Turkish-speaking children and adolescents in our study and control groups had similar anxiety levels (Table 7). These results are partly in line with the study of Brunner [58].

Anxiety levels usually increase with age, particularly during adolescence [16, 17, 63, 64]. We only found such differences in the control groups, where females and younger children exhibited lower anxiety scores. Usually, anxiety scores are lower in boys [13, 14, 63, 65].

Migrant children and adolescents, similar to migrant adults, were described to have higher overall anxiety scores compared to autochthonous children [39, 66‐68]. The study of Borgna and Contini [69] report, that children and adolescents with migration background are disadvantaged in the school system, as most of the features of the educational system do not cover the needs of migrant children. The differences in our study may be related to the specific situation of the Turkish-speaking children in the school system.

Our findings show that non-clinical Turkish-speaking children and adolescents had higher levels of internalizing, depressive and anxiety symptoms, which probably are in relation to migration-induced acculturation stress that impaired their mental health. Since the migrant population does not adequately and timely use mental health services, they only ask for professional help when severe problems occur [70].

These results can also be explained with cross-cultural differences among native Austrians and Turkish speaking migrants, especially in the cultural dimensions of individualism and collectivism. Austrians are individualistically oriented, in contrary the Turkish community is collectively oriented which makes their lifestyles quite different in many aspects [71]. Among individualistically oriented societies, members see themselves as autonomous agents, therefore their own preferences and goals have priority. In contrary, in collectivist societies, a strong connection among members of the social group is present and loyalty is quite important, so that they choose goals, which do not threaten group harmony [72]. Turkish families living in Austria are mostly very traditionally and hierarchically oriented. As the majority of the Turkish population is collectively oriented and also belongs to Islamic faith, Turkish speaking children probably suffer from acculturation stress in Austria because of the larger cultural and religious differences between the culture of origin and the dominant host culture [25, 73‐75].

Furthermore, in the present study we analysed factors like socio-economic status or acculturation stress, which may be related to transcultural differences. Our results showed that children and adolescents with Turkish migration background had a lower socio-economic status compared to their indigenous peers; additionally they were impaired due to a high level of acculturative stress.