Predictive association between immigration status and chronic pain in the general population: results from the SwePain cohort

Immigration is considered to be a major determinant in health disparities [1, 2]. Several studies have demonstrated an association between immigration and mental well-being [3‐7]. For example, the risk of developing psychosis is about two times higher in immigrants and about three times higher in immigrants from developing Eastern European countries and developing countries with high and middle income [6]. The prevalence of posttraumatic stress disorders and depression is also high among immigrants [3, 7]. Previous research using longitudinal data has also shown that immigrants are generally at higher risk of poor health [8] (e.g., ischaemic heart disease, diabetes, and stroke) [9‐11]. Additionally, being an immigrant has been associated with a number of psychosocial issues such as economic stress, difficulties in adaptation, increased ambiguity for the future, changes in living conditions and in personal ties, and disruptions of usual social roles and networks [12‐15].

Previous research has also highlighted the association between immigration and chronic health conditions. Understanding how immigration status and chronic health conditions are related is important because both immigration and chronic pain constitute highly complex topics that have an enormous impact on both individuals and society [11]. Emerging evidence has revealed that the prevalence of chronic pain is high among immigrants [16‐20]. For example, Soares et al. [18] found that non-western born immigrants residing in Sweden experience a greater impact of chronic pain than their Swedish-born counterparts. A study from the UK found that south Asian ethnicity (i.e., people who define themselves as being of Indian, Pakistani, or Bangladeshi origin) was a significant predictor of spinal pain with disability [16], while other studies have shown that immigrants have higher odds of social difficulties, chronic widespread musculoskeletal pain [19, 21], higher pain-related psychological consequences, and higher rates of pain-related disability [16, 18, 19]. Kurita et al. [19] also found that immigrants in general report a higher pain prevalence and higher pain intensity than native-born individuals.

However, the causal pathway between immigration and chronic health outcomes remains not entirely clear, as the reasons that immigrant status seems to be associated with an increased risk for pain appear to be multifaceted and previous studies have shown some contradictory results. Choudhury et al. [20] found that Bangladeshi ethnic minority group in East London who have lower levels of acculturation (i.e. assimilation to a different culture, typically the dominant one) experience more pain, however other studies have shown that increased acculturation in immigrants is associated with higher reports of chronic back and neck problems [17]. In general, it has been suggested that the stress of the immigrant experience can lead to a higher report of chronic back or neck problems among immigrant respondents [17]. Given the lower use of anxiolytics and opioids in immigrants in Denmark who report increased pain, a previous study has questioned whether immigrants are undertreated, or whether healthcare professional attitudes and lack of resources are contributing to reduced access to care for immigrant populations [19]. Finally, the fact that immigrants are more likely to work in riskier jobs with poor working conditions which are more physically strenuous and demanding can be another potential explanation for the increased pain among immigrants [22].

Immigrants in Sweden have poor somatic health, including musculoskeletal disorders, compared to native-born Swedes, and are over-represented among those who get an early retirement due to musculoskeletal disorders [21]. A Swedish study investigating patient reported outcomes after hip arthroplasty highlighted that immigrant groups indicated more pain than those born in Sweden [23]. The differences between immigrants and native-born Swedes can be due to immigrant specific factors (e.g., discrimination, cultural adjustment, language), whilst a further, noticed association between poor health and being born in a country other than Sweden, was greatly reduced when the social network, social support, and economic factors were controlled for [21].

However, one line of research indicates that foreign-born status may represent a health advantage, a phenomenon known as the ‘healthy immigrant effect’ [24]. A systematic review of healthcare outcomes in Canada determined that on average the immigrant population is healthier than the Canadian-born population in terms of mental health, chronic conditions, disability/functional limitations, and risk behaviours [25]. Furthermore, as immigrants worldwide are increasing, studies are needed that examine health problems such as chronic pain among immigrants [12, 13].

To this end, this population-based study with a two-year follow-up investigates whether immigration status is associated with chronic pain. First, we tested the hypothesis that the odds ratio (OR) of having chronic pain at a two-year follow-up is higher in immigrants than native born Swedes. We then applied a path analysis approach to explore whether the relationship between immigrant status and chronic pain is mediated by mood status (i.e., anxiety and depression). The unique contribution of our study to the literature lies in the exploration of the causal pathway between immigration and chronic health condition outcomes, and in particular in investigating the role of mental health problems in the association between immigrant status and chronic pain.

All statistical analyses were performed using IBM SPSS Statistics (version 25.0; IBM Inc., New York, USA) and R statistical language and environment (version 3.6.1) using the lavaan package [38]. Two-sided statistical tests were used and a P < 0.05 was considered significant. We calculated means and standard deviations (SDs) for continuous variables and frequencies with percentages (n; %) for categorical variables.

To examine the predictive association between baseline immigration status (foreign-born vs Swedish-born) with the pain outcomes at follow-up (presence of CP, CWSP, and severe CP), we used logistic regression models under the Generalized Estimated Equations (GEE) with robust standard errors and a logit link function, while we employed an unstructured correlation matrix [39]. GEE is a flexible method for longitudinal analysis and can be used to analyse correlated data with binary, discrete, or continuous outcomes, also considering the dependency between repeated measures. This technique also allows all participants to be included in the analysis even when data are missing [39]. Particularly, it allows missing values within a subject without losing all the data from the subject, and time-varying predictors that can appear in the model [40]. The statistical significance of the models was determined using the Wald test [41]. For this analysis, we adjusted for unequal possibilities of sample selection by weighting cases regarding age, strata, gender, and city. These weights were calculated by SCB [30]. We produced two models per outcome of interest and per immigration status: one unadjusted in which crude ORs with corresponding 95% (CIs) were calculated; and one adjusted model including time independent variable of sex (women vs. men), and time dependent variables of age, marital status (married vs. other), financial hardship (yes vs. no), anxiety, depression, and changes in CP, CWSP, and severe CP. We also performed a sensitivity analysis excluding those with missing information on the baseline variable of the financial hardship and one including only those who had developed chronic pain at follow-up.

We then explored whether the relationship between immigrant status and chronic pain is mediated by mood status (i.e., anxiety and depression) via a path analysis approach. Path analysis can be used to describe the directed dependencies among a set of variables and can estimate both the magnitude and significance of causal links between variables [42]. For this analysis we used baseline data for immigration status and covariates while we used the follow-up data for the outcomes of interest i.e., all three chronic pain conditions. Participants with missing values were excluded from this analysis. The final sample size for the analysis after excluding missing values was n = 11,152 for CP and CWSP while the final sample size for severe CP was n = 6870. Path models identification (i.e., just-identified model, over-identified model, and under-identified model) were based on degrees of freedom (df) which are related to the number of parameter estimates. The models df must be equal or bigger than 0 [43]. We tested the path model using the maximum likelihood estimation using the fit indices proposed by Hu and Bentler [44] as well as Barrett [45]. Briefly, we used the Chi-Square (χ²) value, which is the traditional measure for evaluating overall model fit and ‘assesses the magnitude of discrepancy between the sample and fitted covariances matrices’ [44]. A good model fit should provide an insignificant result at a 0.05 threshold [45]. Other indicators were the Tucker Lewis Index (TLI), the normed fit index (NFI), the non-normed fit index (NNFI), the comparative fit index (CFI), and the goodness-of-fit index (GFI), which shows the model fit relative to the null model. Typically, all indices are considered acceptable when estimates ≥0.90 [44]. The root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMSR) were also included. For both latter indices, estimates ≤0.05 were considered a good fit. We presented three models: one for CP, one for CWSP, and one for severe CP. We tested the mediation effect of mood status (i.e., anxiety and depression) with bootstrapping procedures using the mediate function from the mediation package in R [46]. We also transformed the standardized regression coefficients (beta) into ORs.

Based on our large cohort of the general population, a high prevalence of chronic pain, chronic widespread pain, and severe chronic pain was observed among first generation immigrants 2 years after baseline data collection. Moreover, an increased risk of having any chronic pain outcome was found. Both unadjusted and adjusted models and sensitivity analysis showed similar results. The adjusted risk was almost one and a half times higher for chronic widespread pain and severe chronic pain for migrants compared to Swedes. Our exploratory analysis also found that baseline financial hardship, depression, and anxiety may play an important role in chronic pain among immigrants. Especially, baseline mood aspects seem to mediate the relationships between baseline immigration status and chronic pain outcomes at follow-up. Mood status fully mediates the relationship between immigration and chronic widespread pain, while in the case of chronic pain it may also have a suppressor effect [47].

To our knowledge, this large population-based study is the first study evaluating three common chronic pain outcomes to spotlight the role of immigration in chronic health conditions such as chronic pain by comparing the chronic pain ORs in immigrants with native populations. In addition, our study is unique in the sense that it sheds light on the pathway between immigration and chronic health outcomes, exploring the mediating role of mental health problems in the association between chronic pain and immigrant status. Our results now provide important evidence in an otherwise sparse area of study.

Overall, the present study confirmed the findings regarding inequalities in chronic pain prevalence among immigrant populations. Similar to our study, Kurita et al. [19] documented a higher prevalence of pain in individuals with a foreign background compared to native Danes. This was also the case for the studies conducted by Soares et al. [18] and of Choudhury et al. [20]. The latter study found that chronic widespread pain was more common and more severe in the Bangladeshi than in the white population in East London. Our study shows that being an immigrant is not only associated with increased chronic pain, but also other relevant factors (i.e., age and financial hardship) and more importantly mental health conditions such as anxiety and depression may play an important role in the experience of chronic pain among immigrants. Moreover, our results are in agreement with previous studies that suggest immigrants have higher odds of chronic musculoskeletal pain [19‐21]. However, the majority of the earlier studies used a cross-sectional study design, making a direct comparison with our findings not fully relevant. Our results are partly in agreement with a recent population-based study in Germany, showing that although ‘migration background’ as per official statistics definition is not related with increased mental health problems, identification as an immigrant (self and/or by others) was found as significant predictor for PTSD and depression [48]. A previous report investigating health-related quality of life outcomes have also documented important disparities between racial/ethnic groups related to the experience and management of pain [23]. Taking into consideration our findings, our study did not seem to follow the hypothesis of a ‘healthy immigrant effect’ [24]. Today, there is extensive ongoing research about the above mentioned phenomenon, yet the findings have been inconclusive [25, 49‐51].

Our path analysis showed that immigration status, along with age, financial hardship, and mood variables were associated with a higher risk for all chronic pain outcomes, whilst on the other hand, university education was associated with a lower risk for chronic pain. These findings seem to be in concordance with previous literature from Sweden which showed that experiencing pain was more severe in the older immigrants, suffering from depression with a background of limited education [52]. The path analysis in our study also highlighted the importance of co-existing mental health problems in the experience of chronic pain, since the associations between immigrant status and all three pain outcomes were mediated by mental health status, i.e., anxiety and depression. Anxiety and depression are known to be more prevalent in immigrant populations and have been associated with increased pain [52]. In general, our results agree with earlier findings showing strong associations between anxiety, depression, and socio-economic situation and future chronic pain [11, 26, 53, 54]. Other well-known sociodemographic factors related to future chronic pain outcomes such as age, sex, and education [26, 35, 53, 54] were also confirmed in our analysis. Future research should thoroughly investigate immigration-related factors including a wide-range of sociodemographic and health-related factors that may contribute to the health status among immigrants.

The results of this study should be interpreted taking into consideration some limitations. While our study used a longitudinal study design in conjunction with large and representative sample size, the response rate was low and the proportion of the immigrant population was relatively very low compared to the native population (10% vs 90%, respectively). This low proportion alongside the declined response rate at follow-up among immigrants in our data may underestimate the observed predictive associations between immigrant status and chronic pain outcomes. Likewise, there was great heterogeneity among immigrants. Thus, our findings should be interpreted with caution also considering that the short follow-up (i.e., 2 years) may not be long enough to properly explore the changes of chronic pain status. Moreover, as this study collected data using postal surveys rather than interviews, it was not possible to include refugees or to examine other immigration-related factors such as language skills, age at immigration, second immigrant generation, and acculturation status, factors that have been proven to affect the relationship between immigration and health status [4, 7, 17]. Furthermore, the postal design of the study means that some of the most mentally unwell immigrants (who in general have more frequent and severe mental health difficulties seemingly associated with higher levels of chronic pain) may have found it more difficult to complete the questionnaires and therefore immigrants with severe chronic widespread pain/chronic severe pain might have been under-represented in the survey response and outcomes. Finally, the hypothesized relationships between the variables may be in different directions. For example, we found that mood (anxiety and depression) mediates the relationship between immigrant status and pain; yet it could also be plausible that pain would partially explain the relationship between immigrant status and mood (anxiety and depression). Generally, mediation is ideal in the context of experimental designs (which have many controls); accordingly, it should be fully acknowledged that our study design/model was unable to account for many other potential explanations.

In conclusion, our study highlights the importance of evaluating the chronic pain prevalence and experience among immigrants and verifies a predictive association between immigrant status and increased risk of chronic pain, widespread pain, and severe chronic pain after adjustments for known risk factors. More importantly, our study provides health care practitioners with a deeper knowledge of the factors influencing the relationship between immigration status and chronic pain, which, in turn, could help enable targeted interventions better tailored to socio-economic and psychological status of immigrants with chronic pain. These findings are important because pain, anxiety, depression, and social factors like financial strain may lead to greater ill-health. Future research with larger samples should thoroughly investigate immigration-related factors including a wide-range of sociodemographic and health-related factors that may contribute to the health status among immigrants are needed to evaluate our findings, considering the difficulty of trans-cultural care.