Results
All the indices of socio-demographic and SE characteristics differed significantly (p < 0.001) between the participants from these two groups. The ethnic Pakistanis were younger, more often married, had lower education, were less often in paid jobs (Table
1) compared to Norwegians.
Table 1
Socio-demographic characteristics of the sample
Age cohort
| | |
30 years | 28.3 | 23.0 |
40/45 years | 54.0 | 33.8 |
59/60 years | 15.6 | 23.9 |
75/76 years | 2.1 | 19.3 |
Gender
| | |
Male | 51.3 | 44.8 |
Female | 48.7 | 55.2 |
Marital status*
| | |
Married | 90.6 | 47.3 |
Others | 9.4 | 52.7 |
Education
| | |
1–9 years | 26.8 | 14.0 |
10–12 years | 33.9 | 24.7 |
13+ years | 39.2 | 61.3 |
Employment
| | |
status
| | |
With jobs | 44.6 | 69.3 |
Without jobs | 38.9 | 10.7 |
Missing | 16.5 | 20.0 |
Pakistani respondents also reported lower number of friends, less interest from others in their activities, less participation in the social gatherings and higher level of powerlessness compared to the Norwegians. These differences were statistically significant both unadjusted and when adjusted for age and sex. In contrast, ethnic Pakistanis reported significantly higher influence on the local community compared to Norwegians (Table
2).
Table 2
Psychosocial variables by ethnicity. Means with 95% Confidence Interval adjusted for age and sex by ANOVA.
Social support
| | | | | |
Good friends | 258 | 4.20 (3.85–4.58) | 12323 | 6.95 (6.89–7.0) | <0.001 |
Other's interest | 273 | 2.66 (2.56–2.76) | 10612 | 3.03 (3.02–3.05) | <0.001 |
Social participation
| | | | | |
Participation in groups | 243 | 0.88 (0.69–1.06) | 12935 | 1.16 (1.13–1.18) | 0.003 |
Influence on local community | 263 | 1.76 (1.66–1.86) | 10760 | 1.62(1.60–1.64) | 0.007 |
powerlessness
| 120 | 2.53 (2.45–2.62) | 11129 | 2.15 (2.14–2.16) | <0.001 |
While looking into the association between SE factors and the psychosocial variables in these two ethnic communities, significant interactions emerged. Interaction analysis indicated differences between Pakistanis and Norwegians for the effects of education (p < 0.000) and employment status (p = 0.001) on distress. In Norwegians, psychological distress improved with the increase in education level but no such association was observed for ethnic Pakistanis. Where as employment status was related to reduce psychological distress in ethnic Norwegians, but being employed was not related to any significant reduction in distress among ethnic Pakistanis.
The overall mean distress (adjusted for age and sex) for ethnic Pakistanis was 1.53 (95% CI: 1.48–1.59) and 1.30 (95% CI: 1.29–1.30) for ethnic Norwegians. All the psychosocial variables showed significant association with psychological distress for ethnic Norwegians. In the case of ethnic Pakistanis, significant associations were observed with number of good friends and powerlessness (Table
3). Interaction analyses further indicated that association with number of good friends was significantly stronger (p = 0.002) for ethnic Pakistanis than in Norwegians. With respect to others' interest and influence on the local community, these associations were on the same level in the two groups. Opposite to the findings in the Norwegians, participation in groups for ethnic Pakistanis was associated with increased level of psychological distress (p < 0.000).
Table 3
Association between psychological distress (dependent variable) and psychosocial factors, adjusted for age and sex, by ethnicity: findings from linear regression
Social support
| | | | | | |
Good friends | -0.06 | -0.09 to -0.03 | <0.001 | -0.03 | -0.03 to -0.03 | <0.001 |
Other's interest | -0.09 | -0.16 to -0.01 | 0.021 | -0.11 | -0.12 to -0.10 | <0.001 |
Social participation
| | | | | | |
Participation in groups | 0.06 | -0.01 to 0.12 | 0.061 | -0.03 | -0.04 to -0.03 | <0.000 |
Influence on local community | -0.02 | -0.11 to 0.06 | 0.601 | -0.04 | -0.05 to -0.03 | <0.001 |
Powerlessness
| 0.37 | 0.19 to 0.56 | <0.001 | 0.25 | 0.23 to 0.26 | <0.001 |
Adjusting for age and sex, the expected difference in distress between ethnic Pakistanis and Norwegians was 0.23 (0.19–0.29), with Pakistanis having most distress. This difference was reduced to 0.07 (0.01 to 0.12) and 0.12 (0.07 to 0.18) when adjusted for SE indicators (education and employment status) and for social support (number of good friends and people's interest in your activity), respectively. The differences, however, were still statistically significant. Adjusting for all of these variables simultaneously eliminated the difference in distress between ethnic Pakistanis and Norwegians (Table
4).
Table 4
The association between distress (dependent variable) and ethnic group (Norwegian as reference) in multiple regression analysis.
Model 0 | 0.23 | 0.19 to 0.29 | <0.001 | Age and sex |
Model 1 | 0.07 | 0.01 to 0.12 | 0.016 | Age, sex, education and employment status |
Model 2 | 0.12 | 0.07 to 0.18 | <0.001 | Age, sex, social support* |
Model 3 | 0.00 | -0.06 to 0.06 | 0.922 | Age, sex, education, employment status, social support |
Discussion
The differences observed between these two ethnic communities with regard to the socio-demographic characteristics were in accordance with the already existing statistics in Norway. For example, we know that immigrant population in general is younger than the total population of Norway. 50% of the immigrant population is in the age group 20–44 years compared to 35% of the total population [
30]. Likewise, existing information regarding employment rate indicates that, Pakistani and Turkish immigrants have lowest employment rate (44%) compared to 69.3% in the entire population [
31]. It has been further reported that the overall gap in income between non-Western immigrants and native born Norwegians is 20–30% for men and somewhat lower for women [
32]. Moreover, marriage is the only socially acceptable pattern of cohabiting among ethnic Pakistanis, where as this is not the only option in case of ethnic Norwegians. The difference in educational level among ethnic Pakistanis and Norwegians in this study might be explained by the sample consisting of first generation immigrants from Pakistan, who came to Norway as young guest worker with limited or no education.
Most of the psychosocial factors in the study showed less favorable outcome for the ethnic Pakistanis than for the Norwegians. With respect to influence on the local community, however, the opposite was true. Also as expected, these factors were associated to psychological distress, but somewhat different in the two ethnic groups.
The total number of close friends was more strongly associated to psychological distress in the ethnic Pakistanis than in the ethnic Norwegians. A possible explanation for this could be that the Pakistanis in Oslo are more exposed to economic and social stressors than the Norwegian born are, and for that reason are in stronger need of friends to cope. This is in accordance with the "buffer hypothesis" of social support, and also in accordance with the earlier findings that the disruption of traditional support systems has a negative impact on the psychological well-being of the immigrants [
10,
33‐
35].
For Norwegians participation in groups was significantly and negatively associated with distress, where as for Pakistanis this association was in opposite direction. This difference in trend and association was also confirmed by the interaction analysis. One possible explanation of this difference in trend observed in this study might represent the feelings of Pakistanis while participating in social gatherings organized by the host members of the society due to some language, cultural or due to their hierarchal perception about their social position. It is hard to believe that this community feels the same when they participate in their own social gatherings. This response needs further research in future.
To feel that one has some influence on the local community was negatively associated to psychological distress in both ethnic groups, but somewhat unexpectedly, Pakistanis reported more influence than the Norwegians. This response might indicate the diversity of these two communities in terms of their social structure and meaning to this variable. It seems that while reporting their influence on the local community ethnic Pakistanis might have referred to their influence in their own community based on the personal relations and affiliations. This is in agreement with the observations reported by other studies where it has been discussed that members of Western and non-Western society's represent two different approaches of social structure, i.e. individualistic and collectivistic. In contrast to more or less equal and horizontal relations and social ties in the individualistic societies, the relations and social ties in collective societies are vertical in nature [
36]. Consequently, individuals in collective societies are bound by relationships which emphasize common fate and interest. It is therefore possible that higher influence on the local community reported by Pakistanis here represent the naturally existing characteristics of their social structure based on common fate/interest or interpersonal affiliations. It might not necessarily correspond with the Western concept of community involvement/influence.
Like many other epidemiological studies, social support and health are also investigated from an egocentric approach in this study. This approach mainly deals and focuses on the structure and function of networks immediately surrounding the people ignoring the fact that social support and networks are embedded within a broader set of macro-social exchanges [
37]. To account for the macro-social concept of exchanges in terms of broader social networks, we included the sense of powerlessness as a psychosocial factor in this study. Powerlessness is a feeling that arises when an individual considers him/her worthless in terms of societal norms, attitudes and human models. The literature in this field suggests that living in poverty, with low self-esteem, high demands, low level of control, chronic stress, lack of social support and lack of resources are some of the key physical and social risk factors related to the feeling of powerlessness [
38,
39]. Most of these risk factors are an essential part of the migration process and resettlement experience. Therefore, by reporting higher feeling of powerlessness, Pakistanis might have pointed indirectly toward difficulties in their life situation due to their immigrant status.
Moreover, results of the study show that the expected difference in the level of distress between these two communities becomes insignificant after adjusting simultaneously for psychosocial and SE factors. When all these factors are taken into consideration, it seems that there is no difference in mental health between Norwegians and Pakistanis.
As already reported by the author in another paper from the same sample [
40], SE factors alone cannot explain the difference in psychological distress between the two ethnic groups. When stratifying for level of socioeconomic status, the rates of distress were still higher in the Pakistanis, and this difference was especially strong among those with high socioeconomic status, in particular among those with higher education. Whereas Norwegians with higher education displayed relatively low rates of distress, the opposite was true for the Pakistanis.
The finding that the difference in psychological distress between Norwegians and Pakistanis disappears when adjusting for SE factors as well as social support, however not when adjusting for each variable separately, indicates that both these factors are important mediators between immigration and mental health. Taken into consideration that especially Pakistanis with higher education display a relatively high level of distress, it seems that the lack of support is especially important for mental health in this group. This is in accordance with the finding that social support did not increase with increasing education among Pakistanis, contrary to the situation for Norwegians.
As a conclusion, the study shows that access to work and income are important for health, for Norwegians as well as for immigrants from Pakistan, however in addition it shows that the situation is more complex for the immigrants than for the Norwegians. It seems that immigrants with higher education, and hence with higher aspirations with respect to a social career, run into special frustrations. This may be related to a certain degree of discrimination on the labor market against the immigrants, which may be especially strong against those with higher education. There are many stories about Pakistanis with higher education in Norway, or immigrants from other culturally different countries, who apply for an endless number of white collar jobs without any positive response, as if their name in itself was a reason for discrimination. Then it becomes easier for them to get a job as a taxi-driver or housecleaner. The existing statistics from Norway confirms that 25% of the non-Western immigrants are enrolled within elementary occupations compared to 6.7% of the entire working population [
31]. Moreover, it has been reported that immigrants do not perform as well on the labor market as natives with similar characteristics and a large proportion of immigrants from non-Western countries is characterized as self-employed marginalized, even when controlling for observed and unobserved individual characteristics [
41].
From a methodological point of view, the weakness of the study is that it is based on a cross-sectional design. The inherent problem of a cross-sectional design is that the outcome (in this case distress) and the exposure (in this case psychosocial and socio-economic condition) are collected simultaneously and thereby preventing conclusions regarding causality. Moreover, little attention has been paid to the information bias emerging from the dependent error in the cross-sectional studies, which means a possible correlation between the degree of error in measured exposure and measured outcome. Thus, it is possible that estimated associations between distress and psychosocial factors are falsely inflated in our study [
42].
The data collected by self-reporting has often raised the concerns about its validity. However, self-reported health and related psychosocial variables are widely used in European [
43‐
45] and American studies [
46,
47].
Besides the validity of self-reported data, we were also concerned about the apprehension of psychological variables by the ethnic Pakistanis in Oslo. A word to word translation was made from 'Norwegian' to 'Urdu' language for all the psychosocial instruments used in this study. However the instrument for powerlessness was not even translated. It might be possible that metaphors and phrases used in those instruments are not culturally relevant to Pakistani respondents. Hence, results from these instruments must be interpreted with caution. Keeping this issue in mind, we have planned to conduct another study in Pakistan and Oslo after securing the culturally sensitive translation of instruments used in the current study. Another methodological challenge was related to the low participation rate for both the ethnic Pakistanis and Norwegians in our study. Low participation in epidemiological studies may threaten the validity and generalizability of the results due to the possibility of selective participation [
48]. However, in case of the current study such an impact is negligible [
26].
In addition to low participation rate, it was a problem that not all the participants answered to every question. The lowest response rate among Pakistanis was related to the total household income. It was further noticed that 86.4% of the respondents with lower household income and 80% of the respondents in the younger age group (30–45 years) had not responded to HSCL among ethnic Pakistanis, where as no such association was seen for the Norwegians.
Conclusion
This study has revealed the importance of psychosocial factors in addition to SE factors on the psychological distress among Pakistani immigrants in Oslo. It seems that to improve the mental health of this immigrant community, we need to address both SE and psychosocial issues. This could be achieved through adopting a strategy that not only deals with better SE opportunities equivalent to their education and other socio-demographic characteristics, but also provide opportunities to bring them in contact with the main stream Norwegian society. In this way it might be possible that while gaining SE stability, this community will also find the way to reduce the burden of dispersed/reduced social support on their distress level by interacting and affiliating to the Norwegian society. It has been reported in studies that frequent interactions between immigrant communities and the individuals of the society of resettlement have a positive impact on the mental health of immigrants and vice versa [
49,
50].
By adopting such a strategy, it would be possible that this community would start taking an active part in the context of the broader Norwegian society. This active involvement would then reduce the prevailing feeling of powerlessness in this community. Gaining sense of empowerment in this way would enhance their abilities to meet their own needs, solve their own problems and would give them confidence to interact with the Norwegian society. In return, we can expect that prevailing prejudice in Norwegian society would reduce. This process in return will alleviate the acculturation process of this community and will give strength to the multicultural concept of Norwegian society by reducing the inequalities in health.
The second conclusion of this study is related to the importance of the more planned and solid methodological grounds to conduct research with immigrants. In comparative studies it is important that we focus on the substance of the information being compared rather than assuming that we are dealing with similar information.
Hence, we need to develop culturally sensitive and validated instruments in the field of immigrant's health in Norway.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
HR conducted the basic study. OS contributed with the intellectual discussions and inputs in draft. ID and BC have discussed the results and statistical methods. AH, RS and NA read the script and have given their expert comments and suggestions to improve the study.