Association between unmet post-arrival expectations and psychological symptoms in recently arrived refugees

Abstract

Thwarted expectations regarding one's post-settlement life may challenge the mental health of refugees. The present study aimed to investigate the relationship between pre-arrival expectations and the course of psychological symptoms across time. A secondary analysis of 1,496 principal visa applicants across five waves of the Building a New Life in Australia (BNLA) study was conducted. The cross-sectional associations between expectations on the one hand, and post-traumatic stress (PTSD-8) symptoms and psychological distress (Kessler-6; K6) on the other, were assessed using multiple regression. Latent class growth analysis (LCGA) was used to identify discrete symptom trajectories of psychological symptoms across five years following settlement, and multinomial regressions were used to determine if violated expectations predicted membership of identified PTSD-8 and K6 class trajectories. LCGA supported a four-class solution for the PTSD-8 “Resilient Post Traumatic Stress (PTS)” (54.1%), “Improving PTS” (15.0%), “Deteriorating PTS” (17.3%), and “Persistently High PTS” (13.6%). For the K6, three classes were identified: “Persistently Mild K6” (60.4%), “Resilient K6” (9.4%), and “Persistently High K6” (30.2%). Thwarted expectations were found to significantly predict membership of less favourable symptom trajectories classes in the context of other established predictors. Post-settlement expectations may thus have weak but unique predictive value for the course of psychological symptoms alongside other factors such as older age and financial stress. Implications of these findings for service provision and policy are discussed.

Keywords

expectations post-traumatic stress pre-migration psychological distress PTSD refugee

The United Nations High Commissioner for Refugees reported that as of June 2019, there were 70.8 million forcibly displaced people around the world, with 25.9 million formally recognized refugees (UNHCR, 2019). Multiple pre-migration, migration, and post-migration stressors contribute to increased rates of depression, anxiety, and post-traumatic stress disorder (PTSD) among refugees when compared to non-migrant populations (Porter & Haslam, 2005). While there appears to be consensus that rates of mental health disorders in refugees are higher than in non-migratory populations, there is nonetheless large variance observed in these estimates. A meta-analysis conducted by Lindert et al. (2016) found the prevalence of depression, anxiety, and PTSD in refugees to be 44%, 40%, and 36% respectively.

The mental health outcomes observed in refugee populations have been attributed to a large range of contextual and socio-political factors (Porter & Haslam, 2005). Historically, research has focused on refugee mental health as a direct consequence of pre-migration trauma, such as exposure to torture, war, and natural disasters (Steel et al., 2002). However, this overlooked the ongoing factors during the migration experience occurring across time and contexts (Bracken et al., 1995; Porter & Haslam, 2005). Refugees commonly experience a wide range of stressors during the ‘pre-flight, flight, exile, resettlement/repatriation periods’ (Morina et al., 2018; Porter & Haslam, 2005, p. 603). These early stressors are believed to have lasting effects on mental health and are further compounded by environmental stressors which may include socioeconomic disadvantage (Murali & Oyebode, 2004), marginalization (Horyniak et al., 2016), loss of social support (Wachter & Gulbas, 2018), cultural bereavement (Smeekes et al., 2017), and acculturation difficulties (Porter & Haslam, 2005).

In addition, those who are forced to flee the familiarity of their home may be inclined to envisage their new life based on a set of preconceived ideas possibly with no first-hand account of what it is likely to be (Negy et al., 2009). These preconceived ideas of what life will be like in a new country may lead to high or even potentially unrealistic expectations for their future and their families (Negy et al., 2009). However, it is likely that once resettled, refugees will experience a range of positive and negative experiences which may or may not align with their expectations. If these expectations are violated it will likely result in ‘psychological burdens’ related to the process of adapting to a host country (Berry, 1997; Roccas et al., 2000). This process of acculturation can be experienced as a stressor, especially when the changes are perceived as aversive or coercive (Negy et al., 2009). Higher observed levels of acculturative stress have been consistently linked to less desirable mental health outcomes in both immigrant and refugee populations (D’Abreu et al., 2019; Sirin et al., 2019). The role of violated expectations in this process has not been widely researched (Negy et al., 2009), despite expectations being included as a moderating factor by Berry (1997) in his theory of acculturation.

Negy and colleagues (2009) suggested that unmet expectations can be further explained through the lens of expectancy violation theory (EVT) as posited by Burgoon (1978). EVT argues that when expectations are not met or are violated it results in a negative psychological reaction (Burgoon, 1978). In this respect, “violation” of one's expectations is considered synonymous with those expectations not being met. Burgoon’s (1978) theory also asserts that if expectations are exceeded above simply being met it can result in positive psychological responses. Furthermore, EVT emphasizes that poorer outcomes are more likely when the expectations were formed well before it was possible to know how realistic they may be (Negy et al., 2009).

Simich and colleagues (2006, p. 436) asserted that “no matter whether the decision to migrate is forced or voluntary, migration is undertaken in the expectation of a better life” and a similar sentiment is present in much of the published literature (Magro, 2009; Negy et al., 2009). Magro (2009) noted that there is often a clash between the high expectations of one's intended destination and the ensuing experience once there. For instance, interviews with young adults from the Sudanese refugee community in Winnipeg, Canada, indicated that expectations of opportunities such as education and support had not materialized (Magro, 2009). Furthermore, disappointment was reported with regard to the barriers that had prevented them from fulfilling their expectations and a collective feeling that these barriers had never been discussed prior to their arrival (Magro, 2009). Further studies have consistently identified that thwarted pre-flight expectations have a negative impact on mental health and general well-being (Covington-Ward, 2017; Mähönen et al., 2013; Simich et al., 2006).

McKelvey and colleagues (1993) attempted to complete a prospective longitudinal study exploring the link between pre-migratory expectations and mental health symptomology post settlement in Vietnamese American youth of American fathers who were initially assessed at the Amerasian Transit Center in Ho Chi Minh City. Expectations and mental health symptomology were recorded at baseline prior to migration. However, when attempting to collect the follow-up data, only 25 of the 161 original participants were able to be contacted. McKelvey and Webb (1996) asserted that a discriminant function analysis suggested that the group was representative of the original sample despite the 85% attrition rate. Higher expectations around support from their new communities were found to be associated with higher levels of depressive symptomology (McKelvey & Webb, 1996). However, the large loss of participants between studies must be viewed as a major limitation of this study and the study did not assess whether expectations were met or unmet, but rather only the pre-migratory expectations themselves.

Established predictors of poor post-migration mental health include discrimination (Beiser & Hou, 2017; Sangalang et al., 2019), family conflict (Beiser & Hou, 2017), and high levels of daily stressors (Beiser & Hou, 2017; LeMaster et al., 2018; Sangalang et al., 2019). Violated expectations may serve as an independent predictor of poor mental health alongside these variables, as well as in an interacting fashion, if these other stressors serve to kindle thwarted expectations which may in turn precipitate unfavourable post-migration mental health trajectories.

The present study aimed to investigate the effect of violated pre-migratory expectations on post-settlement mental health and mental health trajectories in a refugee population. To date, research in this area has been almost exclusively cross-sectional and retrospective, without controlling for any recall bias, and has failed to control for variables such as past trauma. To address this question, the present study reports data from the Building a New Life in Australia (BNLA) study, a five-year representative cohort study of recently arrived refugees in Australia (Fogden et al., 2020).

The aforementioned literature suggests that the mental health outcomes of refugees may be affected when their expectations are violated and may be protected when their expectations are met (Mähönen et al., 2013). Given that refugees report high rates of PTSD on account of fleeing from their home countries as well as the often-perilous migration journey (Slewa-Younan et al., 2015), specific mental health consequences may be evident in the trajectories of their post-settlement PTSD symptoms. Given that a sense of safety rather than enduring threat is thought to contribute to favourable recovery from PTSD (Kelly et al., 2014), the meeting of refugees’ expectations regarding safety, predictability, and security may be important for the course of any PTSD symptoms. However, there may also be implications for the course of their mental health more generally, as reflected by the non-specific notion of “psychological distress” (Kessler et al., 2002) which encompasses anxiety and depression symptoms among others. This informed the following hypotheses. First, it was hypothesized that unmet expectations would be associated with unfavourable post-traumatic stress symptomatology at the baseline assessment of BNLA. Second, it was predicted that unmet expectations would be associated with unfavourable psychological distress at the baseline assessment of BNLA. Third, it was hypothesized that unmet expectations at the baseline assessment would be predictive of unfavourable post-traumatic stress symptomatology over time; and fourth, that unmet expectations at baseline will be predictive of unfavourable psychological distress over time.

Method

The BNLA study was a population-based longitudinal study with data collected from refugees granted humanitarian visas in Australia in 2013. Data was collected annually, with the final wave collected in 2018. These five waves were used as a basis for the current analyses. Data was initially collected in a face-to-face setting (either administered by an interview or self-interview on a computer) and alternated with telephone interviews at each collection point. Participants provided information pertaining to a broad range of domains, including demographic information, language, housing, education, pre-migration experiences, health, self-sufficiency, employment and income, community support, life satisfaction, and life in Australia (Chen et al., 2017). Interviews were conducted in multiple languages, with either questions translated or with an interviewer who was matched to the participants’ native language. Ethics approval was obtained from the University of Technology Sydney Human Research Ethics Committee (reference number: 2018002456-12).

Participants and procedure

The BNLA collected data from 2,399 humanitarian migrants, aged between 15 and 83 years. Humanitarian visas were granted to individuals and migrating units (MU) 3–6 months prior to their participation in the survey. Migrating units consisted of principal applicants (lead applicant for the MU) and secondary applicants (other names on the MU application over the age of 15 years). Within the dataset, 35 different countries were represented, with varying cultural backgrounds. Participants were sourced from 11 sites from around Australia, with participants from both regional and metropolitan areas. Of the 2,399 participants, 1,509 were principal applicants, with the remaining 890 classed as secondary applicants. The present study utilized the data of the primary applicants only, to avoid statistical dependencies arising from the clustering of multiple members of the same family.

Measures

The Post-Traumatic Stress Disorder Scale – 8 items (PTSD-8; Hansen et al., 2010) was used to assess post-traumatic stress (hypervigilance, avoidance, and intrusion). The PTSD-8 was derived from the Harvard Trauma Questionnaire (HTQ; Mollica et al., 1992), and has demonstrated strong psychometric properties. A 4-point Likert scale is used to assess symptoms over the week prior (Hansen et al., 2010). In the current study, the PTSD-8 was found to have adequate internal consistency (Cronbach's α) ranging from .92 to .95 across the five waves.

The Kessler Psychological Distress Scale (K6; Kessler et al., 2003) was used to assess psychological distress over the previous month (e.g., nervousness, restlessness, worthlessness; Mewton et al., 2016). The Australian version of the K6 was used and consists of six questions rated on a 5-point scale, with total scores reflecting the sum of all individual item scores. The K6 has been used extensively across multiple languages and cultures to measure non-specific psychological distress (Kessler et al., 2010). In the present study, the K6 was found to have internal consistency (Cronbach's α) ranging from .87 to. 92 across all five waves.

A range of other demographic, migration, and occupational-rated items were administered at each wave of assessment. However, the present study focused on the variables relating to pre-flight expectations. Participants were asked if quality of life; access to education; what they would be given to start their new life; adequacy of services; acceptance of cultural background; quality of accommodation; amount of government payments; time taken to learn English; opportunities to reunite with family; and job opportunities were better, worse, or as expected when thinking back to what they expected before coming to Australia.

Data analysis

The study was a within-groups design, with each participant being assessed on five occasions. A population sampling weighting was not applied, as the purpose of this study was to explore the relationships within and between the participants rather than serving estimates of the population as a whole. Descriptive statistics including the mean, standard deviation, and frequencies were used to summarize the demographic information of the participants at Wave 1. The K6 and PTSD-8 were modelled as numeric interval measures.

A confirmatory factor analysis (CFA) was conducted in order to determine the latent structure of the 10 expectations-related questions and that a single total score would reflect a unitary expectations construct. A single factor model was investigated using the “Categorical” specifier in Mplus version 8.3 such that the analysis was based on polychoric correlations. Alternative factor structures were not investigated given the absence of conceptually coherent subscales. A CFA approach was used in lieu of exploratory factor analysis, given that previous studies had implied a single unitary, ‘expectations’ construct which was thought to be sufficiently covered by the 10 items of the current study. Model fit was determined to be adequate if the Root Mean Square Error of Approximation (RMSEA) was ≤ 0.08 (as 0.08 reflects “mediocre” fit, MacCallum et al., 1996), Comparative Fit Index (CFI; Hu & Bentler, 1999) > 0.95, and Tucker Lewis Index (TLI; Tucker & Lewis, 1973) > 0.95.

Multiple regression analyses were conducted to determine whether expectations were associated with PTSD-8 and K6 respectively after controlling for sex, exposure to immigration detention, financial stress, age, and country of birth (Afghanistan, Iran, or Iraq vs. other). Sex was included as previous findings indicate sex differences in PTSD symptoms among refugees (Nickerson et al., 2010, Perera et al., 2013). Age was included as the risk of lifetime trauma exposure increases with age and therefore PTSD symptoms may be associated with age. Furthermore, distress levels can vary across the life span (Jorm et al., 2005). Financial stress was included as it was suggested by Simich et al. (2006) that both financial hardship and unmet expectations resulted in the increase of psychological distress.

Latent growth modelling analysis

A latent growth modelling approach was used to determine whether latent classes were identifiable in the data, allowing us to find groups of participants with similar trajectories of the PTSD-8 and the K6 symptoms across each of the annual waves of assessment. This approach separates the trajectories of groups into mutually exhaustive and exclusive classes (Collins & Lanza, 2010), which are ‘latent’ in that membership of any class is not directly observed or measured (O’Donnell et al., 2017). The present analyses extend the latent class growth analyses of this dataset by Fogden et al. (2020), who reported on symptom trajectories across the first four waves of the data that were available at the time. In contrast to the present study, Fogden et al. aimed to determine whether separation from family members and worry about family members were significant predictors of symptom trajectories.

The number of classes was determined partly on the basis of a number of fit indices and partly with respect to model parsimony (simplicity being favoured). The Akaike Information Criterion (AIC; Akaike, 1974) and Bayesian Information Criterion (BIC; Schwarz, 1978), Consistent Akaike Information Criterion (Bozdogan, 1987), and sample size-adjusted BIC (Sclove, 1987) goodness of fit indices were considered, where lower values correspond to improved model fit (Fogden et al., 2020). Two likelihood ratio tests were also considered: the Bootstrap Likelihood Ratio Test (BLRT; McLachlin & Peel, 2000) and the Vuong-Lo-Mendell Likelihood Ratio Test (VLMLRT; Lo et al., 2001), which provide a p-value indicating if a model demonstrates a significantly superior fit than the model with one fewer class (Nylund et al., 2007). The entropy values, each of which indicates the classification accuracy of a solution, are also reported. Entropy values closer to 1 indicate a more accurate classification (Geiser, 2010). In line with the approach of other researchers (e.g., Fogden et al., 2020; Infurna & Grimm, 2018), classes comprising less than 5% of the overall sample were not considered favourable on the basis that such solutions tend to be unstable (Fogden et al., 2020). Analyses were conducted in Mplus version 7.31 (Muthén & Muthén, 1998–2012). All models included an intercept and slope term, and the number of random starts was set to 1,000, and with 200 optimizations for the final stage of analysis. Models were re-run with twice the number of random starts to ensure that the best loglikelihood values were replicated. Mplus uses a Full Information Maximum Likelihood (FIML) approach to dealing with missing data as a default (Muthén & Muthén, 1998–2012). Principal applicants who were missing data at all five time points from either the PTSD-8 or K6 were excluded from the analyses, as Mplus analyses cannot be conducted on participants who are missing data on the dependent variable at all time points. Thus, of the 1,509 principal applicants, 1,496 were then included for the data analysis in Mplus.

Multinomial regression analyses were conducted in IBM SPSS 26.0 on the sample of 1,496 to determine whether expectations significantly predicted trajectory class membership in the context of other relevant variables. For predictor variables, the “reference” category corresponded to the absence of the noted variable (e.g. “0” corresponded to an absence of immigration detention). The same predictor variables were included, as stated previously, for the linear regression analyses.

For all regression analyses in SPSS, a multiple imputation approach with 25 imputations was used to estimate missing values.

Results

Participant profile

The demographic and social characteristics of all participants are displayed in Table 1. There were 1,496 participants included in the sample, of whom 1,053 were male (70.4%), and the mean age was 38.55 years (SD = 12.97, Mdn = 37, range 18–75 yrs).

Table 1.
Demographic characteristics of principal applicants at Wave 1 (N = 1496).

N (%)

Female 443 (29.6%)

Country of birth

Iraq 525 (35.1%)

Afghanistan 446 (29.8%)

Iran 158 (10.6%)

Myanmar 98 (6.6%)

Other 269 (18%)

Completed post-secondary school education 283 (19.1%)

Trauma pre-arrival

Harsh living conditions 510 (34.1%)

War/conflict 861 (57.6%)

Violence 322 (21.5%)

Imprisonment/kidnapping 264 (17.6%)

Political/religious persecution 831 (55.5%)

Natural disasters 95 (6.4%)

Other 263 (17.6%)

Experiencing financial stress 627 (41.9%)

Note: Some variables had incomplete data, and percentages reflect the available data.

The analysis used each point of assessment (Waves 1, 2, 3, 4, and 5) rather than the number of months between assessment, as the time-related variable as precise inter-wave intervals could not be ascertained from the data.

The numbers of principal applicants found to have complete data for the PTSD-8 were 1,443 (96.5%), 1,246 (83.3%), 1,124 (77.1%), 1,154 (79.1%), and 1,123 (75.1%) at Waves 1–5 respectively; and for the K6, 1,458 (97.5%), 1,264 (84.5%), 1,138 (76.1%), 1,169 (78.1%), and 1,126 (75.3%) for Waves 1–5.

Dimensionality of the expectations items

An initial single factor CFA model did not fit the data (χ² = 577.66, p < .01; RMSEA = 0.10, CFI = 0.97, TFI = 0.97). Inspection of modification indices indicated that the error terms of the following pairs of variables should be correlated: (i) expectations about learning English and expectations about education and (ii) expectations about what one would be given and expectations about quality of life. This model then provided an adequate fit to the data (RMSEA = 0.08, CFI = 0.99, TLI = 0.98), so total scores of all 10 expectations items were used in all further analyses.

Wave 1 multiple regression: Hypotheses 1 & 2

Multiple regression analyses were conducted to assess the association between demographic factors and post-migration stressors and PTSD-8 and K6 scores at Wave 1 respectively. Results are presented in Tables 2 and 3. Sex, no experience of immigration detention, post-school level of education, financial stress, country of origin, and age were controlled for. The analysis shows that the total sum score of baseline expectations was a significant independent predictor of (greater) levels of PTS symptomology (B = .22, t(.04) = 5.69, p < .001). Female sex, post-school education, financial stress, and age were also found to be significant predictors of greater levels of PTS symptomology (all ps < .05). For K6, the expectations total score similarly significantly predicted greater K6 scores (B = .24, t(.03) = 5.69, p < .001). Female sex, no experience of immigration detention, financial stress, and older age were also found to be significant predictors of higher K6 total scores (all ps < .05).

Table 2.
Multiple regression predicting PTSD-8 total scores at Wave 1 (N = 1496).

B (SE) 95% CI

Constant 6.16 (.95)* 4.3–8.02

Expectations .22 (.04)* .14–.29

Female sex 1.56 (.39)* .8–2.33

Immigration detention −.64 (.52) −1.66–.38

Post-secondary school education 1.53 (.44) .66–2.41

Financial stress 2.99 (.37)* 2.27–3.70

Afghanistan, Iran, or Iraq country of birth .49 (.41) −.31–1.29

Age .10 (.04)* .07–.29

Note: CI = confidence interval; OR = odds ratio. **p < .05, p < .01, p < .001.

Table 3.
Multiple regression predicting K6 total scores at Wave 1 (N = 1496).

B (SE) 95% CI

Constant 4.69 (.77)* 3.17–6.21

Expectations .24 (.03)* .18–.30

Female sex 1.72 (.32)* 1.08–2.36

Immigration detention −1.23 (.43) −2.07–−.38

Post-secondary school education 2.51 (.31)* 1.9–3.11

Financial stress .16 (.34) −.50–.83

Afghanistan, Iran, or Iraq country of birth .24 (.03)* .18.30

Note: CI = confidence interval; OR = odds ratio. **p < .05, p < .01, p < .001.

Latent growth modelling

A Latent Growth Mixture Model (LGMM) was run for both the PTSD-8 and K6 respectively, however the resulting models did not converge for either variable. Convergence problems persisted when the number of iterations and random starts was varied and when a Bayes estimator was used, which is more likely to achieve convergence than maximum likelihood approaches (Asparouhov & Muthén, 2010). Therefore, a specific type of LGMM, latent class growth analysis (LCGA), was used to identify these groups by setting the within-class variance for intercept and slope to zero (Frankfurt et al., 2016).

Post-traumatic stress symptoms (PTS; LCGA model)

Table 4 reports the classification accuracy, model fit, and relative model fit statistics for the PTSD-8 trajectory class memberships. The absolute fit indices continued to decrease with each class, although this is not uncommon for latent class analyses (Nylund-Gibson & Choi, 2018). The VLMRLRT p-value indicated that the 4-class solution was significantly better than the 3-class solution, and the 4-class solution had the greatest entropy value of all solutions (0.72). Additionally, previous studies have identified similar trajectory classes (e.g., Bonanno et al., 2012) indicating that the 4-class model was the best fit for the data and in the present analysis the smaller class for the 5-class solutions comprised of only 7% of participants, which, although greater than the 5% threshold applied by some researchers, nonetheless reflected a disproportionately small group. Thus, even though the 5-class solution also had a significant VLMRLRT value, for parsimony and the above discussed reasons, the 4-class solution was favoured. The classes of participants are described as: Class 1 “Resilient PTS” (n = 842, 56.3%), Class 2 “Improving PTS” (n = 224, 15.0%), Class 3 “Deteriorating PTS” (n = 228, 15.2%), and Class 4 “Persistently High PTS” (n = 202, 13.5%). Thus, the Improving and Deteriorating trajectories crossed over, while the Persistently High and Resilient trajectories remained consistency high and low respectively. Figure 1 provides a visual representation of the four classes of the PTSD-8 symptom trajectories.

Figure 1.
PTSD-8 symptom trajectories across the five waves of assessment (five years).

Table 4.
Incremental fit statistics and classification accuracy for latent class growth model for PTSD-8 total scores (N = 1496).

No. of classes Loglikelihood (LL) AIC BIC Sample size-adjusted BIC Entropy BLRT p-value VLMRLRT p-value Sample size per class based on most likely class membership

1 − 20,556.11 41,126.22 41,126.22 41,163.40 — — — 1,496

2 − 19,741.46 39,502.92 39,502.92 39,524.25 0.797 <.0001 <.0001 1,044/452

3 − 19,632.92 39,291.85 39,291.85 39,319.59 0.715 <.0001 <.0001 847/447/202

4 − 19,563.65 39,159.29 39,159.29 39,193.43 0.720 <.0001 .0004 842/225/228/202

5 − 19,543.62 39,125.23 39,125.23 39,165.77 0.697 <.0001 .0146 178/80/216/782/240

6 − 19,528.60 39,101.20 39,101.20 39,148.14 0.696 <.0001 .0756 157/263/69/755/66/186

7 − 19,518.58 39,087.16 39,087.16 39,140.50 0.726 <.0001 .2737 49/649/250/158/183/141/66

Note: AIC = Akaike information criterion; BIC = Bayesian information criterion; BLRT = bootstrap likelihood ratio test; K6 = Kessler-6 questionnaire; PTSD-8 = post-traumatic stress disorder scale – 8 items; VLMRLRT = Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test.

Psychological distress LCGA model

Table 5 reports the classification accuracy, model fit, and relative model fit statistics for the K6 trajectory class memberships. A decrease was observed in the fit indices across the successive classes. Ultimately, a 3-class solution was selected due to the significant VLMRLRT p-value. The classes of participants are described as: Class 1 “Persistently Mild K6” (n = 930, 62.2%), Class 2 “Resilient K6” (n = 425, 28.4%), and Class 3 “Persistently High K6” (n = 141, 9.4%), and their parallel form reflected a linear pattern of increasing severity across the three classes from Resilient to Persistently High. A visual representation of the three K6 symptom trajectories is shown in Figure 2.

Figure 2.
K6 symptom trajectories across the five waves of assessment (five years).

Table 5.
Incremental fit statistics and classification accuracy for latent class growth model for K6 total scores (N = 1496).

No. of classes Loglikelihood (LL) AIC BIC Sample size-adjusted BIC Entropy BLRT p-value VLMBLRT p-value Sample size per call based on most likely class membership

1 − 19,774.999 39,563.997 39,601.17 39,578.934 — — — 1,496

2 − 19,013.05 38,046.101 38,099.21 38,067.439 0.805 <.0001 <.0001 384/1112

3 − 18,874.588 37,775.177 37,844.21 37,802.917 0.754 <.0001 <.0001 930/141/425

4 − 18,808.82 37,649.64 37,734.61 37,683.781 0.752 <.0001 .1189 145/941/118/292

5 − 18,780.955 37,599.911 37,700.81 37,640.453 0.733 <.0001 .0956 127/301/141/53/874

6 − 18,760.535 37,565.069 37,681.90 37,612.014 0.722 <.0001 .0236 319/138/71/804/45/119

7 − 18,754.18 37,558.361 37,691.12 37,611.706 0.734 .0128 .3888 63/28/342/40/126/779/118

Note: AIC = Akaike information criterion; BIC = Bayesian information criterion; BLRT = Bootstrap likelihood ratio test; K6 = Kessler-6 questionnaire; PTSD-8 = post-traumatic stress disorder scale – 8 items; VLMRLRT = Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test.

Supplementary Table 1 summarizes the number and percent of participants in each respective K6 and PTSD-8 class and the extent of overlap between K6 and PTSD-8 trajectory classes. A chi-square test indicated a significant degree of association between K6 and PTSD-8 symptom trajectory classes (χ² = 873.07, df = 6, p < .0001).

Multinomial regressions predicting of trajectory class membership: Hypotheses 3 & 4

Multinomial logistic regression analyses were completed to assess the input of demographic factors and post-migration stressors on K6 and PTSD-8 class memberships respectively. Due to the possibility of multicollinearity between certain variables in the regression analysis, variance inflation factor (VIF) values were reviewed using SPSS. No VIF value was observed greater than 1.50 for any variables of any of the regression analyses, allowing confidence that this was not a meaningful concern.

Each trajectory grouping is presented in Tables 6 and 7, with the three and two most meaningful comparisons (for PTSD-8 and K6 respectively) shaded in each instance. For the PTSD-8, the three key comparisons included unfavourable trajectories (Persistently High and Deteriorating PTS) versus the Resilient trajectory class as well as the comparison between the improving PTSD-8 trajectory versus the Persistently High trajectory class. For K6, the most informative comparisons were the Persistently High and the Persistent Mild classes compared to the Resilient class.

Table 6.
Multinomial regression analyses predicting the PTSD-8 trajectory class memberships. Key comparisons are shaded.

Comparison group

Improving PTS trajectory (C2) Persistently High PTS trajectory (C4) Deteriorating PTS trajectory (C3) Low/resilient PTS trajectory (C1)

Reference group B (SE) OR [95% CI] B (SE) OR [95% CI] B (SE) OR [95% CI] B (SE) OR [95% CI]

Low/resilient PTS trajectory (C1)

Age .03 (.01) 1.03 [1.02, 1.04] .05 (.01)* 1.05 [1.04, 1.06] .03 (.01)* 1.03 [.101, 1.04] — —

Expectations .05 (.02)* 1.05 [1.02,1.09] .07 (.02) 1.07 [1.03, 1.11] −.01 (0.2) .99 [.96, 1.02] — —

Female sex .36 (.17) 1.44 [1.03, 2.02] .87 (.18)* 2.38 [1.67, 3.38] .24 (.17) 1.27 [.91, 1.78] — —

Immigration detention −.51 (.26)* .60 [.36,.99] −.80 (.33)* .45 [.23,.87] −.38 (.24) .69 [.43, 1.09] — —

Post-secondary school education .50 (.19)* 1.64 [1.14, 2.37] .56 (.21)** 1.75 [1.17, 2.62] .11 (.21) 1.11 [.74, 1.66] — —

Financial stress .63 (.16)* 1.88 [1.37, 2.59] 1.16 (.18)* 3.18 [2.23, 4.55] .49 (.16) 1.62 [1.19, 2.22] — —

Afghanistan, Iran, or Iraq country of origin .16 (.18) 1.18 [.83, 1.68] 1.20 (.26)* 3.30 [1.98, 5.52] .78 (.20)*** 2.19 [1.47, 3.25] — —

Persistently High PTS trajectory (C4)

Age −.02 (.01)* .98 [.94, 1.03] — — −.02 (.01) .98 [96,.99] −.05 (.01)* .95 [.94,.97]

Expectations −.02 (.02) .98 [.94, 1.03] — — −.08 (.02) .93 [.89,.97] −.07 (.02) .95 [.94,.97]

Female sex −.50 (.21)* .61 [.40,.92] — — −.63 (.21) .53 [.35,.81] −.87 (.18)* .42 [.30,.60]

Immigration detention .29 (.39) 1.33 [.62, 2.84] — — .42 (.38) 1.52 [.35,.81] .80 (.33)* 2.21 [1.15, 4.26]

Post-secondary school education −.06 (.23) .94 [.60, 1.47] — — −.45 (.25) .63 [.39, 1.03] −.56 (.21)** .57 [.38,.85]

Financial stress −.52 (.21)* .59 [.39,.89] — — −.67 (.21) .51 [.34,.77] −1.16 (.18)* .31 [.22,.45]

Afghanistan, Iran, or Iraq country of origin −1.03 (.29)* .36 [20,.63] — — −.41 (.31) .66 [.36, 1.21] −1.20 (.26)* .30 [.18,.51]

Improving PTS trajectory (C2)

Age — — .02 (.01)* 1.02 [1.00, 1.03] .00 (.01) 1.00 [.98, 1.01] −.03 (.01)* .97 [.96,.98]

Expectations — — .02 (.02) 1.02 [.97, 1.06] −.06 (.02) .94 [.90,.98] −.05 (.02)** .95 [.92,.99]

Female sex — — .50 (.21)* 1.65 [1.10, 2.48] −.12 (.21) .88 [.59, 1.33] −.36 (.17)* .70 [.50,.97]

Immigration detention — — −.29 (.39) .75 [.35, 1.61] .13 (.24) .68 [.42, 1.08) .51 (.26)* 1.66 [1.66, 1.01]

Post-secondary school education — — .06 (.23) 1.07 [.68, 1.67] −.39 (.24) .68 [.42, 1.08] −.50 (.19)** .61 [.42,.88]

Financial stress — — .52 (.21)* 1.69 [1.12, 2.55] −.15 (.20) .86 [.59, 1.27] −.63 (.16)* .53 [.39,.73]

Afghanistan, Iran, or Iraq country of origin — — 1.03 (.29)* 2.80 (1.6, 4.92) .62 (.24)* 1.86 [1.15, 2.99] −.16 (.18) .85 [.59, 1.21]

Note: CI = confidence interval; OR = odds ratio.

**p < .05, p < .01, p < .001.

Table 7.
Multinomial regression analyses predicting the K6 trajectory class memberships. Key comparisons are shaded.

Comparison group

Resilient K6 (C1) Persistently High K6 (C2) Persistently Mild (C3)

Reference group B (SE) OR [95% CI] B (SE) OR [95% CI] B (SE) OR [95% CI]

Resilient K6 (C1)

Age — — .04 (.01)* 1.04 [1.02, 1.05] .03 (.00)* 1.03 [1.02, 1.04]

Expectations — — .09 (.02)* 1.09 [1.04, 1.14] .05 (.01)* 1.06 [1.03, 1.09]

Female sex — — 1.03 (.20)* 2.80 [1.89, 4.15] .73 (.14)* 2.07 [1.58, 2.71]

Immigration detention — — −.63 (.36) .53 [.26, 1.08] −.57 (.21) .57 [.38, .85]

Post-secondary school education — — .04 (.24) [.65, 1.66] −.02 (.16) .98 [.72, 1.34]

Financial stress — — .89 (.20)* 2.43 [1.64, 3.60] .76 (.13)* 2.14 [1.66, 2.76]

Country of origin^# — — 1.14 (.30)* 3.14 [1.74, 5.66] .54 (.16) 1.71 [1.26, 2.33]

Persistently High K6 (C2) — —

Age −.04 (.01)* .96 [.95, .98] — — −.01 (.01) .99 [.98, 1.01]

Total sum score of baseline EXP −.09 (.02)* .36 [.24, .53] — — −.03 (.02) .97 [.93, 1.01]

Female sex −1.03 (.20)* .36 [.24, .53] — — −.30 (.20) .74 [.50, 1.10]

Immigration detention .63 (.36) 1.88 [.93, 3.81] — — .06 (.38) 1.06 [.50, 2.26]

Post-secondary school education −.04 (.24) .96 [.60, 1.55] — — −.06 (.25) .95 [.58, 1.53]

Financial stress −.89 (.20)* .41 [.28, .61] — — −.12 (.21) .88 [.59, 1.33]

Country of origin^# −1.14 (.30)* .32 [.18, .57] — — −.60 (.31) .55 [.30, 1.01]

Note. CI = confidence interval; OR = odds ratio^{. #} Where 1 = Afghanistan, Iran, or Iraq as country of origin (75.5%), 0 = any other country (24.5%). **p < .05, p < .01, p < .001.

Hypothesis 3

For PTS symptoms, the Persistently High trajectory was compared with the Resilient trajectory to determine if individuals whose expectations were unmet had persistently high PTS symptoms across the fives waves, after controlling for sex, experience of immigration detention, post-secondary school level of education, financial stress, country of origin, and age. When compared to the Resilient class, unmet expectations were a significant, but weak, independent predictor of membership of the Persistently High trajectory (OR = 1.07, p < .01). Female sex, not having experienced immigration detention, lack of post-secondary school level of education, financial stress, born in Afghanistan, Iran, or Iraq, and older age were also found to be significant predictors (all ps < .05).

When the Deteriorating trajectory class was compared with the Resilient trajectory class, membership of the Deteriorating class was not predicted by level of unmet expectations. However, older age, financial stress, and being from Afghanistan, Iran, or Iraq were found to be significant predictors of membership of the Deteriorating class for PTS symptoms (all ps < .05).

Finally, the improving trajectory was compared with the Persistently High trajectory class also for PTSD-8. When compared, membership of the Persistently High class was not predicted by unmet expectations when compared to the Improving class. Younger age, male sex, lack of financial stress, and not being from Afghanistan, Iran, or Iraq, however, were significant predictors (of membership of the improving trajectory class all ps < .05).

Hypothesis 4

Regarding K6, the Persistently High trajectory class was compared with the Resilient trajectory class to determine if those with higher levels of unmet expectations had comparatively high levels of psychological distress across the five waves, after controlling for sex, experience of immigration detention, post-secondary school level of education, financial stress, country of origin, and age. When compared with the Resilient class, unmet expectations were a significant, but weak, predictor of membership of the Persistently High trajectory class (OR = 1.09, p < .05). Furthermore, older age, female sex, financial stress, and being from Afghanistan, Iran, or Iraq were also significant predictors (all ps < .05).

In addition, the Persistent Mild class was then compared with the Resilient trajectory class. Upon comparison, membership of the Persistent Mild class was significantly predicted by unmet expectations when compared with the Resilient class (OR = 1.06, p < .5). However, older age, female sex, experiencing immigration detention, financial stress, and being from Afghanistan, Iran, or Iraq were also significant predictors (all ps < .05).

Discussion

Prior research has found that thwarted expectations have an adverse effect on mental health in refugee and migrant populations. Furthermore, expectations have been presented as a moderating factor in Berry’s (1997) theory of acculturation, positioning them as a potential key variable in the post-arrival adjustment of refugees. This theory and research have positioned violated expectations as a variable worthy of more rigorous investigation to determine their influence on post-arrival mental health trajectories of refugees specifically. This study is the first large-scale prospective cohort study investigating these variables for five years post settlement. Overall, our trajectory analyses suggest that most refugees demonstrated remarkable resilience in the immediate post-settlement years, with most participants reporting consistently low levels of psychological symptoms.

Preliminary analyses focused solely on Wave 1 data, to explore if unmet expectations had an adverse effect on mental health at approximately six months post arrival. The first hypothesis was supported, with analysis indicating that violated expectations predicted higher levels of PTS symptomology in the initial six months of settlement in Australia. Furthermore, the second hypothesis was also supported, as violated expectations were found to be predictive of higher levels of psychological distress within the first six months of settlement in Australia. These results align with the findings of Negy and colleagues (2009) who found discrepancies between pre-migration expectations and post-migration experience to be a predictor of acculturative stress in an immigrant population, which has been shown to be predictive of mental health symptomology (D’Abreu et al., 2019; Sirin et al., 2019).

As unmet expectations demonstrated significant predictive value for PTS symptoms and psychological distress at the initial time point, it was deemed appropriate to further investigate these effects over time. Hypothesis 3 was partially supported. In this respect, violated expectations were only found to be predictive of experiencing consistently high levels of PTS symptomology when compared to those who had experienced consistently low levels of symptomology. However, they were not found to be predictive when: a) a resilient symptom trajectory was compared to a deteriorating one, or b) when an improving trajectory was compared to a persistently high PTS symptom trajectory. Hypothesis 4 was supported, with violated expectations being predictive of both consistently high and moderate levels psychological distress when both were compared with those with consistently mild symptoms. These findings suggest that unmet expectations may be a more sensitive predictor of the course of general psychological distress than PTS symptomology specifically.

These findings align with EVT as posited by Burgoon (1978). EVT (Burgoon, 1978) posited that when expectations are violated this has a direct and adverse effect on mental health. This study's findings align with EVT not only at the initial time point but also across time, with thwarted expectations playing a significant role in the five-year trajectories of mental health symptomology.

Although expectations were found to play a significant role in the mental health trajectories of this sample of refugees, they were relatively weak predictors of trajectory class membership. Nonetheless, it is noteworthy that violated expectations were assessed only at a single time point, and were, with few exceptions, a significant predictor of the course of psychological symptoms among refugees across five years following settlement. While by no means the sole predictor of mental health trajectories, it is clear that violated expectations are associated with distinctly less favourable trajectories of symptoms across five years.

Certain demographic variables were also predictive of unfavourable trajectories of psychological distress and PTS symptomology. For instance, financial stress was found to be a significant predictor across all analyses, which is consistent with the findings of Simich and colleagues (2006). Thus, alongside more practical concerns related to financial stress such as cost of living including food and accommodation, it is likely refugees expect their adopted homeland to be a place of opportunity and security and may not have been prepared for this to not be the case (Simich et al., 2006).

The small effect sizes suggest that interventions to address thwarted expectations at the individual level may have limited value, but there remains scope for community-level programmes which may still result in overall improvements in mental health at a population level. Encouraging peer support for newly arrived refugees with members of their community may be helpful in communicating realistic lived experience early in their resettlement. Furthermore, making information about provided government and financial support and access to education more readily available to refugee populations prior to arrival, when expectations are most likely being formed, may be helpful. It may also be appropriate to use violated expectations as a possible screening measure in an individual clinical setting, identifying those who may be more likely to continue on higher symptom trajectories. However, more research would be required to explore this as a potential therapeutic screener. The present study is not without its limitations. There was unfortunately a sizable amount of missing data for the expectations items. Multiple imputation was used to mitigate the effect of this missing data but it is conceivable that biases remained. Despite the PTSD-8 and K6 being well-validated and widely used screening tools in large-scale populations, they do not allow for diagnosis. Furthermore, the K6 screens for psychological distress, which is comprised of both anxious and depressive symptomology. For any future research, it would be advisable to use a screening tool which is able to discern between the two diagnoses which have been shown to be sensitive to different aspects of expectations (McKelvey & Webb, 1996). Finally, we note that the LCGA results for the K6 may simply reflect an underlying continuous distribution of psychological severity, rather than distinct and discrete classes of individuals. Nonetheless, the LCGA analyses indicate that participants tended to stay in the same severity-defined classes across time.

There remains a high degree of variability in the measures of expectations used by previous researchers, which differ both in the definition of the construct as well as the range of domains of expectations that are assessed. The creation of a reliable and valid expectations measure would eliminate this uncertainty in future research and improve methodological rigour. However, despite the lack of a comprehensive expectations scale, the expectations items in the present study appeared to reflect a unidimensional construct which is brief to administer and appears to cover most of the key domains in which expectations might be held, providing opportunities for more extensive validation in future.

As noted, the majority of participants demonstrated trajectories of resilience across the five years of this BNLA study. Collectively, they appeared to be a broadly resilient population considering the hardship and challenges they have encountered. While our findings highlighted this resilience, it is also clear that pre-migratory expectations may play a small but important role in the symptom trajectories of those who do experience psychological distress and PTS symptoms.

Supplemental Material

sj-docx-1-tps-10.1177_13634615221111022 - Supplemental material for Association between unmet post-arrival expectations and psychological symptoms in recently arrived refugees

Supplemental material, sj-docx-1-tps-10.1177_13634615221111022 for Association between unmet post-arrival expectations and psychological symptoms in recently arrived refugees by Claire H. Allinson and David Berle in Transcultural Psychiatry

	N (%)
Female	443 (29.6%)
Country of birth
Iraq	525 (35.1%)
Afghanistan	446 (29.8%)
Iran	158 (10.6%)
Myanmar	98 (6.6%)
Other	269 (18%)
Completed post-secondary school education	283 (19.1%)
Trauma pre-arrival
Harsh living conditions	510 (34.1%)
War/conflict	861 (57.6%)
Violence	322 (21.5%)
Imprisonment/kidnapping	264 (17.6%)
Political/religious persecution	831 (55.5%)
Natural disasters	95 (6.4%)
Other	263 (17.6%)
Experiencing financial stress	627 (41.9%)

	B (SE)	95% CI
Constant	6.16 (.95)***	4.3–8.02
Expectations	.22 (.04)***	.14–.29
Female sex	1.56 (.39)***	.8–2.33
Immigration detention	−.64 (.52)	−1.66–.38
Post-secondary school education	1.53 (.44)**	.66–2.41
Financial stress	2.99 (.37)***	2.27–3.70
Afghanistan, Iran, or Iraq country of birth	.49 (.41)	−.31–1.29
Age	.10 (.04)***	.07–.29

	B (SE)	95% CI
Constant	4.69 (.77)***	3.17–6.21
Expectations	.24 (.03)***	.18–.30
Female sex	1.72 (.32)***	1.08–2.36
Immigration detention	−1.23 (.43)**	−2.07–−.38
Post-secondary school education	2.51 (.31)***	1.9–3.11
Financial stress	.16 (.34)	−.50–.83
Afghanistan, Iran, or Iraq country of birth	.24 (.03)***	.18.30

No. of classes	Loglikelihood (LL)	AIC	BIC	Sample size-adjusted BIC	Entropy	BLRT p-value	VLMRLRT p-value	Sample size per class based on most likely class membership
1	− 20,556.11	41,126.22	41,126.22	41,163.40	—	—	—	1,496
2	− 19,741.46	39,502.92	39,502.92	39,524.25	0.797	<.0001	<.0001	1,044/452
3	− 19,632.92	39,291.85	39,291.85	39,319.59	0.715	<.0001	<.0001	847/447/202
4	− 19,563.65	39,159.29	39,159.29	39,193.43	0.720	<.0001	.0004	842/225/228/202
5	− 19,543.62	39,125.23	39,125.23	39,165.77	0.697	<.0001	.0146	178/80/216/782/240
6	− 19,528.60	39,101.20	39,101.20	39,148.14	0.696	<.0001	.0756	157/263/69/755/66/186
7	− 19,518.58	39,087.16	39,087.16	39,140.50	0.726	<.0001	.2737	49/649/250/158/183/141/66

No. of classes	Loglikelihood (LL)	AIC	BIC	Sample size-adjusted BIC	Entropy	BLRT p-value	VLMBLRT p-value	Sample size per call based on most likely class membership
1	− 19,774.999	39,563.997	39,601.17	39,578.934	—	—	—	1,496
2	− 19,013.05	38,046.101	38,099.21	38,067.439	0.805	<.0001	<.0001	384/1112
3	− 18,874.588	37,775.177	37,844.21	37,802.917	0.754	<.0001	<.0001	930/141/425
4	− 18,808.82	37,649.64	37,734.61	37,683.781	0.752	<.0001	.1189	145/941/118/292
5	− 18,780.955	37,599.911	37,700.81	37,640.453	0.733	<.0001	.0956	127/301/141/53/874
6	− 18,760.535	37,565.069	37,681.90	37,612.014	0.722	<.0001	.0236	319/138/71/804/45/119
7	− 18,754.18	37,558.361	37,691.12	37,611.706	0.734	.0128	.3888	63/28/342/40/126/779/118

	Comparison group
Low/resilient PTS trajectory (C1)
Age	.03 (.01)	1.03 [1.02, 1.04]	.05 (.01)***	1.05 [1.04, 1.06]	.03 (.01)***	1.03 [.101, 1.04]	—	—
Expectations	.05 (.02)***	1.05 [1.02,1.09]	.07 (.02)**	1.07 [1.03, 1.11]	−.01 (0.2)	.99 [.96, 1.02]	—	—
Female sex	.36 (.17)**	1.44 [1.03, 2.02]	.87 (.18)***	2.38 [1.67, 3.38]	.24 (.17)	1.27 [.91, 1.78]	—	—
Immigration detention	−.51 (.26)*	.60 [.36,.99]	−.80 (.33)*	.45 [.23,.87]	−.38 (.24)	.69 [.43, 1.09]	—	—
Post-secondary school education	.50 (.19)*	1.64 [1.14, 2.37]	.56 (.21)**	1.75 [1.17, 2.62]	.11 (.21)	1.11 [.74, 1.66]	—	—
Financial stress	.63 (.16)*	1.88 [1.37, 2.59]	1.16 (.18)***	3.18 [2.23, 4.55]	.49 (.16)**	1.62 [1.19, 2.22]	—	—
Afghanistan, Iran, or Iraq country of origin	.16 (.18)**	1.18 [.83, 1.68]	1.20 (.26)***	3.30 [1.98, 5.52]	.78 (.20)***	2.19 [1.47, 3.25]	—	—
Persistently High PTS trajectory (C4)
Age	−.02 (.01)*	.98 [.94, 1.03]	—	—	−.02 (.01)**	.98 [96,.99]	−.05 (.01)***	.95 [.94,.97]
Expectations	−.02 (.02)	.98 [.94, 1.03]	—	—	−.08 (.02)**	.93 [.89,.97]	−.07 (.02)**	.95 [.94,.97]
Female sex	−.50 (.21)*	.61 [.40,.92]	—	—	−.63 (.21)**	.53 [.35,.81]	−.87 (.18)***	.42 [.30,.60]
Immigration detention	.29 (.39)	1.33 [.62, 2.84]	—	—	.42 (.38)	1.52 [.35,.81]	.80 (.33)*	2.21 [1.15, 4.26]
Post-secondary school education	−.06 (.23)	.94 [.60, 1.47]	—	—	−.45 (.25)	.63 [.39, 1.03]	−.56 (.21)**	.57 [.38,.85]
Financial stress	−.52 (.21)*	.59 [.39,.89]	—	—	−.67 (.21)**	.51 [.34,.77]	−1.16 (.18)***	.31 [.22,.45]
Afghanistan, Iran, or Iraq country of origin	−1.03 (.29)***	.36 [20,.63]	—	—	−.41 (.31)	.66 [.36, 1.21]	−1.20 (.26)***	.30 [.18,.51]
Improving PTS trajectory (C2)
Age	—	—	.02 (.01)*	1.02 [1.00, 1.03]	.00 (.01)	1.00 [.98, 1.01]	−.03 (.01)***	.97 [.96,.98]
Expectations	—	—	.02 (.02)	1.02 [.97, 1.06]	−.06 (.02)**	.94 [.90,.98]	−.05 (.02)**	.95 [.92,.99]
Female sex	—	—	.50 (.21)*	1.65 [1.10, 2.48]	−.12 (.21)	.88 [.59, 1.33]	−.36 (.17)*	.70 [.50,.97]
Immigration detention	—	—	−.29 (.39)	.75 [.35, 1.61]	.13 (.24)	.68 [.42, 1.08)	.51 (.26)*	1.66 [1.66, 1.01]
Post-secondary school education	—	—	.06 (.23)	1.07 [.68, 1.67]	−.39 (.24)	.68 [.42, 1.08]	−.50 (.19)**	.61 [.42,.88]
Financial stress	—	—	.52 (.21)*	1.69 [1.12, 2.55]	−.15 (.20)	.86 [.59, 1.27]	−.63 (.16)***	.53 [.39,.73]
Afghanistan, Iran, or Iraq country of origin	—	—	1.03 (.29)***	2.80 (1.6, 4.92)	.62 (.24)*	1.86 [1.15, 2.99]	−.16 (.18)	.85 [.59, 1.21]

Comparison group
Resilient K6 (C1)
Age	—	—	.04 (.01)***	1.04 [1.02, 1.05]	.03 (.00)***	1.03 [1.02, 1.04]
Expectations	—	—	.09 (.02)***	1.09 [1.04, 1.14]	.05 (.01)***	1.06 [1.03, 1.09]
Female sex	—	—	1.03 (.20)***	2.80 [1.89, 4.15]	.73 (.14)***	2.07 [1.58, 2.71]
Immigration detention	—	—	−.63 (.36)	.53 [.26, 1.08]	−.57 (.21)**	.57 [.38, .85]
Post-secondary school education	—	—	.04 (.24)	[.65, 1.66]	−.02 (.16)	.98 [.72, 1.34]
Financial stress	—	—	.89 (.20)***	2.43 [1.64, 3.60]	.76 (.13)***	2.14 [1.66, 2.76]
Country of origin^#	—	—	1.14 (.30)***	3.14 [1.74, 5.66]	.54 (.16)**	1.71 [1.26, 2.33]
Persistently High K6 (C2)	—	—
Age	−.04 (.01)***	.96 [.95, .98]	—	—	−.01 (.01)	.99 [.98, 1.01]
Total sum score of baseline EXP	−.09 (.02)***	.36 [.24, .53]	—	—	−.03 (.02)	.97 [.93, 1.01]
Female sex	−1.03 (.20)***	.36 [.24, .53]	—	—	−.30 (.20)	.74 [.50, 1.10]
Immigration detention	.63 (.36)	1.88 [.93, 3.81]	—	—	.06 (.38)	1.06 [.50, 2.26]
Post-secondary school education	−.04 (.24)	.96 [.60, 1.55]	—	—	−.06 (.25)	.95 [.58, 1.53]
Financial stress	−.89 (.20)***	.41 [.28, .61]	—	—	−.12 (.21)	.88 [.59, 1.33]
Country of origin^#	−1.14 (.30)***	.32 [.18, .57]	—	—	−.60 (.31)	.55 [.30, 1.01]

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

David Berle

Supplemental material

Supplemental material for this article is available online.

Claire H. Allinson achieved her Masters of Clinical Psychology from the University of Technology Sydney in 2020. She is currently working as a Clinical Psychology Registrar in a private practice setting.

David Berle, MPsychol, PhD, is an associate professor in the Discipline of Clinical Psychology, Graduate School of Health, University of Technology Sydney. His research interests include post-traumatic stress disorder (PTSD), obsessive-compulsive disorder (OCD), and depression.

References

Akaike

. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19(6), 716–723. https://doi.org/10.1109/tac.1974.1100705

Asparouhov

Muthén

(2010). Bayesian analysis of latent variable models using Mplus. Technical report. Muthén & Muthén.

Beiser

Hou

. (2017). Predictors of positive mental health among refugees: Results from Canada’s General Social Survey. Transcultural Psychiatry, 54(5-6), 675–695. https://doi.org/10.1177%2F1363461517724985

Berry

J. W

. (1997). Immigration, acculturation, and adaptation. Applied Psychology, 46(1), 5–34. https://doi.org/10.1111/j.1464-0597.1997.tb01087.x

Bonanno

G. A.

Mancini

A. D.

Horton

J. L.

Powell

T. M.

LeardMann

C. A.

Boyko E. J. & Millennium Cohort Study Team. (2012). Trajectories of trauma symptoms and resilience in deployed US military service members: prospective cohort study. The British Journal of Psychiatry, 200(4), 317–323. https://doi.org/10.1192/bjp.bp.111.096552

Bozdogan

. (1987). Model selection and Akaike’s Information Criterion (AIC): The general theory and its analytical extensions. Psychometrika, 52, 345–370. https://doi.org/10.1007/BF02294361

Bracken

P. J.

Giller

J. E.

Summerfield

. (1995). Psychological responses to war and atrocity: The limitations of current concepts. Social Science & Medicine, 40(8), 1073–1082. https://doi.org/10.1016/0277-9536(94)00181-R

Burgoon

J. K

. (1978). A communication model of personal space violations: Explication and an initial test. Human Communication Research, 4(2), 129–142. https://doi.org/10.1111/j.1468-2958.1978.tb00603.x

Chen

Hall

B. J.

Ling

Renzaho

A. M

. (2017). Pre-migration and post-migration factors associated with mental health in humanitarian migrants in Australia and the moderation effect of post-migration stressors: Findings from the first wave data of the BNLA cohort study. The Lancet Psychiatry, 4(3), 218–229. https://doi.org/10.1016/S2215-0366(17)30032-9

10.

Collins

L. M.

Lanza

S. T.

(2010). Latent class and latent transition analysis: With applications in the social, behavioral, and health sciences. Wiley Blackwell.

11.

Covington-Ward

. (2017). “Back home, people say America is heaven: Pre-migration experiences and post-migration adjustment for Liberians in Pittsburgh. Journal of International Migration and Integration, 18, 1013–1032. https://doi.org/10.1007/s12134-017-0511-7

12.

D’Abreu

Castro-Olivo

Ura

S. K

. (2019). Understanding the role of acculturative stress on refugee youth mental health: A systematic review and ecological approach to assessment and intervention. School Psychology International, 40(2), 107–127. https://doi.org/10.1177/0143034318822688

13.

Fogden

Berle

Steel

. (2020). The impact of family separation and worry about family on psychological adjustment in refugees resettled in Australia. Journal of Traumatic Stress, 33(6), 894–897. https://doi.org/10.1002/jts.22568

14.

Frankfurt

Frazier

Syed

Jung

K. R

. (2016). Using group-based trajectory and growth mixture modelling to identify classes of change trajectories. The Counselling Psychologist, 44(5), 622–660. https://doi.org/10.1177/0011000016658097

15.

Geiser

(2010). Data analysis with Mplus. Guilford Press.

16.

Hansen

Anderson

Armour

Elklit

Palic

Mackrill

. (2010). PTSD-8: A short PTSD inventory. Clinical Practice and Epidemiology in Mental Health, 6, 101–108. https://doi.org/10.2174/1745017901006010101

17.

Horyniak

Higgs

Cogger

Dietze

Bofu

. (2016). Heavy alcohol consumption among marginalised African refugee young people in Melbourne, Australia: Motivations for drinking, experiences of alcohol-related problems and strategies for managing drinking. Ethnicity & Health, 21(3), 284–299. https://doi.org/10.1080/13557858.2015.1061105

18.

Bentler

P. M

. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. https://doi.org/10.1080/10705519909540118

19.

Infurna

F. J.

Grimm

K. J

. (2018). The use of growth mixture modelling for studying resilience to major life stressors in adulthood and old age: Lessons for class size and identification and model selection. The Journals of Gerontology: Series B, 73(1), 148–159. https://doi.org/10.1093/geronb/gbx019

20.

Jorm

A. F.

Windsor

T. D.

Dear

K. B. G.

Anstey

K. J.

Christensen

Rodgers

. (2005). Age group differences in psychological distress: The role of psychosocial risk factors that vary with age. Psychological Medicine, 35(9), 1253–1263. https://doi.org/10.1017/S0033291705004976

21.

Kelly

Boyd

M A.

Valente

S. M.

Czekanski

. (2014). Trauma-informed care: Keeping mental health settings safe for veterans. Issues in Mental Health Nursing, 35(6), 413–419. https://doi.org/10.3109/01612840.2014.881941

22.

Kessler

R. C.

Andrews

Colpe

L. J.

Hiripi

Mroczek

D. K.

Normand

S. -L. T.

Walters

E. E.

Zaslavsky

A. M

. (2002). Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine, 32(6), 959–976. https://doi.org/10.1017/S0033291702006074

23.

Kessler

R. C.

Barker

P. R.

Colpe

L. J.

Epstein

J. F.

Gfroerer

J. C.

Hiripi

,… Zaslavsky

A. M.

(2003). Screening for serious mental illness in the general population. Archives of General Psychiatry, 60(2), 184–189. https://doi.org/10.1001/archpsyc.60.2.184

24.

Kessler

R. C.

Green

J. C.

Gruber

M. J.

Sampson

N. A.

Bromet

Cuitan

Furukawa

T. A.

Gureje

Hinkov

C-Y.

Lara

Lee

Mneimneh

Myer

Oakley-Browne

Posada-Villa

Sagar

Viana

M. C.

Zaslavsky

A. M

. (2010). Screening for serious mental illness in the general population with the K6 screening scale: Results from the WHO World Mental Health (WMH) Survey initiative. International Journal of Methods in Psychiatric Research, 19(Suppl 1), 4–22. https://doi.org/10.1002/mpr.310

25.

LeMaster

J. W.

Broadbridge

C. L.

Lumley

M. A.

Arnetz

J. E.

Arfken

Fetters

M. D.

Jamil

Pole

Arnetz

B. B

. (2018). Acculturation and post-migration psychosocial symptoms among Iraqi refugees: A path analysis. American Journal of Orthopsychiatry, 88(1), 38–47. https://doi.org/10.1037/ort0000240

26.

Lindert

Carta

M. G.

Schäfer

Mollica

R. F

. (2016). Refugees mental health—A public mental health challenge. The European Journal of Public Health, 26(3), 374–375. https://doi.org/10.1093/eurpub/ckw010

27.

Mendell

N. R.

Rubin

D. B

. (2001). Testing the number of components in a normal mixture. Biometrika, 88(3), 767–778. https://doi.org/10.1093/biomet/88.3.767

28.

MacCallum

R. C.

Browne

M. W.

Sugawara

H. M

. (1996). Power analysis and determination of sample size for covariance structure modelling. Psychological Methods, 1(2), 130–149. https://psycnet.apa.org/doi/10.1037/1082-989X.1.2.130

29.

Magro

(2009). Building bridges to social inclusion: Researching the experiences and challenges of the lost boys and girls of Sudan community in Winnipeg. Prairie Metropolis Centre.

30.

Mähönen

Leinonen

Jasinskaja-Lahti

(2013). Met expectations and the wellbeing of diaspora immigrants: A longitudinal study. International Journal of Psychology, 48(3), 324–333. https://doi.org/10.1080/00207594.2012.662278

31.

McKelvey

R. S.

Mao

A. R.

Webb

J. A

. (1993). Premigratory expectations and mental health symptomatology in a group of Vietnamese Amerasian youth. Journal of the American Academy of Child & Adolescent Psychiatry, 32(2), 414–418. https://doi.org/10.1097/00004583-199303000-00024

32.

McKelvey

R. S.

Webb

J. A

. (1996). Premigratory expectations and postmigratory mental health symptoms in Vietnamese Amerasians. Journal of the American Academy of Child & Adolescent Psychiatry, 35(2), 240–245. https://doi.org/10.1097/00004583-199602000-00018

33.

McLachlin

Peel

(2000). Finite mixture models. Wiley.

34.

Mewton, L., Kessler, R. C., Slade, T., Hobbs, M. J., Brownhill, L., Birrell, L., …Andrews, G. (2016). The psychometric properties of the Kessler psychological distress scale (K6) in a general population sample of adolescents. Psychological Assessment, 28(10), 1232–1242. https://doi.org/10.1037/pas0000239

35.

Mollica

R. F.

Caspi-Yavin

Bollini

Truong

Tor

Lavelle

. (1992). The Harvard Trauma Questionnaire: Validating a cross-cultural instrument for measuring torture, trauma and posttraumatic stress disorder in Indochinese refugees. Journal of Nervous and Mental Disease, 180(2), 111–116. https://psycnet.apa.org/doi/10.1097/00005053-199202000-00008

36.

Morina

Akhtar

Barth

Schnyder

. (2018). Psychiatric disorders in refugees and internally displaced persons after forced displacement: A systematic review. Frontiers in Psychiatry, 9, 433. https://doi.org/10.3389/fpsyt.2018.00433

37.

Murali

Oyebode

(2004). Poverty, social inequality and mental health. Advances in Psychiatric Treatment, 10(3), 216–224. https://doi.org/10.1192/apt.10.3.216

38.

Muthén

L. K.

Muthén

B. O.

(1998–2012). Mplus user’s guide: Seventh edition. Muthén & Muthén.

39.

Negy

Schwartz

Reig-Ferrer

. (2009). Violated expectations and acculturative stress among US Hispanic immigrants. Cultural Diversity and Ethnic Minority Psychology, 15(3), 255–264. https://doi.org/10.1037/a0015109

40.

Nickerson

Bryant

Steel

Silove

Brooks

. (2010). The impact of fear for family on mental health in a resettled Iraqi refugee community. Journal of Psychiatric Research, 44(4), 229–235. https://doi.org/10.1016/j.jpsychires.2009.08.006

41.

Nylund

K. L.

Asparouhov

Muthén

B. O.

(2007). Deciding on the number of classes in latent class growth analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14(4), 535–569. https://doi.org/10.1080/10705510701575396

42.

Nylund-Gibson

Choi

A. Y

. (2018). Ten frequently asked questions about latent class analysis. Translational Issues in Psychological Science, 4(4), 440–461. https://doi.org/10.1037/tps0000176

43.

O'Donnell, M., Schaefer, I., Varker, T., Kartal, D., Forbes, D., Bryant, R., A. … Steel, Z., (2017). A systematic review of person-centred approaches to investigating patterns of trauma exposure. Clinical Psychology Review, 57, 208–225. https://doi.org/10.1016/j.cpr.2017.08.009

44.

Perera

Gavian

Frazier

Johnson

Spring

Westermeyer

Butcher

Halcon

Robertson

Savik

Jaranson

. (2013). A longitudinal study of demographic factors associated with stressors and symptoms in African refugees. American Journal of Orthopsychiatry, 83(4), 472–482. https://doi.org/10.1111/ajop.12047

45.

Porter

Haslam

. (2005). Predisplacement and postdisplacement factors associated with mental health of refugees and internally displaced persons: A meta-analysis. Jama, 294(5), 602–612. https://doi.org/10.1001/jama.294.5.602

46.

Roccas

Horenczyk

Schwartz

S. H

. (2000). Acculturation discrepancies and well-being: The moderating role of conformity. European Journal of Social Psychology, 30(3), 323–334. https://doi.org/10.1002/(SICI)1099-0992(200005/06)30:3<323::AID-EJSP992>3.0.CO;2-5

47.

Sangalang

C. C.

Becerra

Mitchel

F. M.

Lechuga-Pena

Lopez

Kim

. (2019). Trauma, post-migration stress, and mental health: A comparative analysis of refugees and immigrants in the United States. Journal of Immigrant and Minority Health, 21, 909–919. https://doi.org/10.1007/s10903-018-0826-2

48.

Schwarz

. (1978). Estimating the dimension of a model. Annals of Statistics, 6(2), 461–464. https://doi.org/10.1214/aos/1176344136

49.

Sclove

. (1987). Application of model-selection criteria to some problems in multivariate analysis. Psychometrika, 52, 333–343. https://doi.org/10.1007/BF02294360

50.

Simich

Hamilton

Baya

B. K

. (2006). Mental distress, economic hardship and expectations of life in Canada among Sudanese newcomers. Transcultural Psychiatry, 43(3), 418–444. https://doi.org/10.1177/1363461506066985

51.

Sirin

S. R.

Sin

Clingain

Rogers-Sirin

. (2019). Acculturative stress and mental health: Implications for immigrant-origin youth. Pediatric Clinics, 66(3), 641–653. https://doi.org/10.1016/j.pcl.2019.02.010

52.

Slewa-Younan

Guajardo

M. G. U.

Heriseanu

Hasan

. (2015). A systematic review of post-traumatic stress disorder and depression amongst Iraqi refugees located in western countries. Journal of Immigrant and Minority Health, 17, 1231–1239. https://doi.org/10.1007/s10903-014-0046-3

53.

Smeekes

Verkuyten

Çelebi

Acartürk

Onkun

. (2017). Social identity continuity and mental health among Syrian refugees in Turkey. Social Psychiatry and Psychiatric Epidemiology, 52(10), 1317–1324. https://doi.org/10.1007/s00127-017-1424-7

54.

Steel

Silcove

Phan

Bauman (2002). Long-term effect of psychological trauma on the mental health of Vietnamese refugees resettled in Australia: A population-based study. The Lancet, 360(9339), 1056–1062. https://doi.org/10.1016/S0140-6736(02)11142-1

55.

Tucker

L. R.

Lewis

. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10.

56.

United Nations High Commissioner for Refugees. (2019). UNHCR global trend report: Forced displacement in 2018 . https://www.unhcr.org/5d08d7ee7.pdf

57.

Wachter

Gulbas

L. E

. (2018). Social support under siege: An analysis of forced migration among women from the Democratic Republic of Congo. Social Science & Medicine, 208, 107–116. https://doi.org/10.1016/j.socscimed.2018.04.056

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

	Comparison group
	Improving PTS trajectory (C2)		Persistently High PTS trajectory (C4)		Deteriorating PTS trajectory (C3)		Low/resilient PTS trajectory (C1)
Reference group	B (SE)	OR [95% CI]	B (SE)	OR [95% CI]	B (SE)	OR [95% CI]	B (SE)	OR [95% CI]
Low/resilient PTS trajectory (C1)
Age	.03 (.01)	1.03 [1.02, 1.04]	.05 (.01)***	1.05 [1.04, 1.06]	.03 (.01)***	1.03 [.101, 1.04]	—	—
Expectations	.05 (.02)***	1.05 [1.02,1.09]	.07 (.02)**	1.07 [1.03, 1.11]	−.01 (0.2)	.99 [.96, 1.02]	—	—
Female sex	.36 (.17)**	1.44 [1.03, 2.02]	.87 (.18)***	2.38 [1.67, 3.38]	.24 (.17)	1.27 [.91, 1.78]	—	—
Immigration detention	−.51 (.26)*	.60 [.36,.99]	−.80 (.33)*	.45 [.23,.87]	−.38 (.24)	.69 [.43, 1.09]	—	—
Post-secondary school education	.50 (.19)*	1.64 [1.14, 2.37]	.56 (.21)**	1.75 [1.17, 2.62]	.11 (.21)	1.11 [.74, 1.66]	—	—
Financial stress	.63 (.16)*	1.88 [1.37, 2.59]	1.16 (.18)***	3.18 [2.23, 4.55]	.49 (.16)**	1.62 [1.19, 2.22]	—	—
Afghanistan, Iran, or Iraq country of origin	.16 (.18)**	1.18 [.83, 1.68]	1.20 (.26)***	3.30 [1.98, 5.52]	.78 (.20)***	2.19 [1.47, 3.25]	—	—
Persistently High PTS trajectory (C4)
Age	−.02 (.01)*	.98 [.94, 1.03]	—	—	−.02 (.01)**	.98 [96,.99]	−.05 (.01)***	.95 [.94,.97]
Expectations	−.02 (.02)	.98 [.94, 1.03]	—	—	−.08 (.02)**	.93 [.89,.97]	−.07 (.02)**	.95 [.94,.97]
Female sex	−.50 (.21)*	.61 [.40,.92]	—	—	−.63 (.21)**	.53 [.35,.81]	−.87 (.18)***	.42 [.30,.60]
Immigration detention	.29 (.39)	1.33 [.62, 2.84]	—	—	.42 (.38)	1.52 [.35,.81]	.80 (.33)*	2.21 [1.15, 4.26]
Post-secondary school education	−.06 (.23)	.94 [.60, 1.47]	—	—	−.45 (.25)	.63 [.39, 1.03]	−.56 (.21)**	.57 [.38,.85]
Financial stress	−.52 (.21)*	.59 [.39,.89]	—	—	−.67 (.21)**	.51 [.34,.77]	−1.16 (.18)***	.31 [.22,.45]
Afghanistan, Iran, or Iraq country of origin	−1.03 (.29)***	.36 [20,.63]	—	—	−.41 (.31)	.66 [.36, 1.21]	−1.20 (.26)***	.30 [.18,.51]
Improving PTS trajectory (C2)
Age	—	—	.02 (.01)*	1.02 [1.00, 1.03]	.00 (.01)	1.00 [.98, 1.01]	−.03 (.01)***	.97 [.96,.98]
Expectations	—	—	.02 (.02)	1.02 [.97, 1.06]	−.06 (.02)**	.94 [.90,.98]	−.05 (.02)**	.95 [.92,.99]
Female sex	—	—	.50 (.21)*	1.65 [1.10, 2.48]	−.12 (.21)	.88 [.59, 1.33]	−.36 (.17)*	.70 [.50,.97]
Immigration detention	—	—	−.29 (.39)	.75 [.35, 1.61]	.13 (.24)	.68 [.42, 1.08)	.51 (.26)*	1.66 [1.66, 1.01]
Post-secondary school education	—	—	.06 (.23)	1.07 [.68, 1.67]	−.39 (.24)	.68 [.42, 1.08]	−.50 (.19)**	.61 [.42,.88]
Financial stress	—	—	.52 (.21)*	1.69 [1.12, 2.55]	−.15 (.20)	.86 [.59, 1.27]	−.63 (.16)***	.53 [.39,.73]
Afghanistan, Iran, or Iraq country of origin	—	—	1.03 (.29)***	2.80 (1.6, 4.92)	.62 (.24)*	1.86 [1.15, 2.99]	−.16 (.18)	.85 [.59, 1.21]

Comparison group
	Resilient K6 (C1)		Persistently High K6 (C2)		Persistently Mild (C3)
Reference group	B (SE)	OR [95% CI]	B (SE)	OR [95% CI]	B (SE)	OR [95% CI]
Resilient K6 (C1)
Age	—	—	.04 (.01)***	1.04 [1.02, 1.05]	.03 (.00)***	1.03 [1.02, 1.04]
Expectations	—	—	.09 (.02)***	1.09 [1.04, 1.14]	.05 (.01)***	1.06 [1.03, 1.09]
Female sex	—	—	1.03 (.20)***	2.80 [1.89, 4.15]	.73 (.14)***	2.07 [1.58, 2.71]
Immigration detention	—	—	−.63 (.36)	.53 [.26, 1.08]	−.57 (.21)**	.57 [.38, .85]
Post-secondary school education	—	—	.04 (.24)	[.65, 1.66]	−.02 (.16)	.98 [.72, 1.34]
Financial stress	—	—	.89 (.20)***	2.43 [1.64, 3.60]	.76 (.13)***	2.14 [1.66, 2.76]
Country of origin^#	—	—	1.14 (.30)***	3.14 [1.74, 5.66]	.54 (.16)**	1.71 [1.26, 2.33]
Persistently High K6 (C2)	—	—
Age	−.04 (.01)***	.96 [.95, .98]	—	—	−.01 (.01)	.99 [.98, 1.01]
Total sum score of baseline EXP	−.09 (.02)***	.36 [.24, .53]	—	—	−.03 (.02)	.97 [.93, 1.01]
Female sex	−1.03 (.20)***	.36 [.24, .53]	—	—	−.30 (.20)	.74 [.50, 1.10]
Immigration detention	.63 (.36)	1.88 [.93, 3.81]	—	—	.06 (.38)	1.06 [.50, 2.26]
Post-secondary school education	−.04 (.24)	.96 [.60, 1.55]	—	—	−.06 (.25)	.95 [.58, 1.53]
Financial stress	−.89 (.20)***	.41 [.28, .61]	—	—	−.12 (.21)	.88 [.59, 1.33]
Country of origin^#	−1.14 (.30)***	.32 [.18, .57]	—	—	−.60 (.31)	.55 [.30, 1.01]