The “How” Question of the Healthy Immigrant Paradox: Understanding Psychosocial Resources and Demands as Pathways Linking Migration to Mental Health Risks

Abstract

The current migrant health literature tends to focus on what determines immigrants’ mental health rather than how pathways such as psychosocial resources mediate the relationship between years since migration and mental health. Based on 4,282 foreign-born Canadian immigrant samples, this study includes both psychological distress and positive mental health as mental health measures because immigrants do not necessarily respond to stress by exhibiting distress. The correlation between psychological distress and positive mental health shows that these two measures are interrelated but distinctive concepts. Using piecewise regression models, this study finds that midterm immigrants have the highest levels of psychological distress and interpersonal strain. Guided by the stress process model, this study indicates that interpersonal strain acts as a major pathway to immigrants’ psychological distress but not positive mental health.

Keywords

immigrants positive mental health psychological distress psychosocial resources stress process model

Introduction

Empirical studies indicate that immigrants are positively selected for educational attainment, job skills, and health profiles (Belot and Hatton 2012). As a result, new immigrants tend to be in better health than nonimmigrants and long-term immigrants in major immigrant-receiving countries, such as the United States, the United Kingdom, Canada, and Australia—a phenomenon known as the healthy immigrant paradox (Farre 2016; Kennedy et al. 2015; Kwak 2016; Wu and Schimmele 2005).

Several migrant health researchers have demonstrated that the healthy immigrant paradox eventually fades (Newbold 2009; Rivera, Casal, and Currais 2016; Wu and Schimmele 2005); however, there is little consensus as to when and in what contexts this deterioration is most likely to begin. While some studies highlight the initial settlement stages as the most critical period for immigrants’ health (Christodoulou 2009; Newbold 2009), a number of others indicate that long-term immigrants tend to be of poorer physical and mental health than more recent immigrants (Rivera et al. 2016; Wu and Schimmele 2005). To be sure, the relationships between years since migration and immigrants’ health profiles have always been complex. A 2004 study by McDonald and Kennedy, for instance, demonstrated a quadratic relationship between years since migration and chronic morbidity in women who immigrate to Canada. Such rates of health deterioration can also differ based on the timing of one’s arrival, not just the cumulative number of years since migration. Using a longitudinal survey, Newbold (2005) showed that the health profiles of immigrants who landed in Canada between 1990 and 1994 deteriorated at faster rates than immigrants who had arrived earlier.

To begin, this study does not assume a linear relationship between years since migration and mental health. Rather, I take a piece-wise regression approach that allows for the nonlinearity of such a focal association. More specifically, the relationship between years since migration and mental health follows an inverted J, where the curve is smooth and shallow and not steep enough for a quadratic slope (Marsh and Cormier 2002).

Guided by the stress process model (SPM), this study provides psychosocial explanations as to why midterm immigrants tend to be of poorer mental health than both recent and long-term immigrants. The SPM is well suited for examining psychosocial resources—what Portes and Rivas (2011) called “soft variables”—that are highly responsive to the structural forces of the receiving context and often function as inequality-generating pathways, with greater effects on mental health than demographic factors such as marital status and occupation (Lin et al. 1979).

Canada and its Immigrant-Receiving Context

Various studies have pointed to the positive selection of healthy immigrants in Canada (Lu and Ng 2019; Newbold 2005; Wu and Schimmele 2005); however, the specificities of Canada’s receiving context—especially given its effects on immigrants’ psychosocial resources—have yet to be sufficiently addressed in the migrant health literature.

Prior to the 1970s, the Canadian government favored immigrants from the United Kingdom, the United States, and northern European countries, while those from southern or eastern European countries were only allowed restrictive entry during periods of economic growth (Arat-Koc 1999). With the 1967 implementation of the points-based system that remains today, immigrants seeking entry to Canada were required to have enough “points” based on a set of universal standards for selection, such as age, education, knowledge of the Canadian official languages (English or/and French), and labor market skills. This soon prompted a shift in Canada’s sourcing of immigrants from European countries to Asian, African, and South and Central American countries (Antecol, Cobb-Clark, and Trejo 2003).

For immigrants, settlement tends to be a lifelong journey involving three distinct phases: (1) initial reception (language training and accessing information), (2) secondary stages of education and employment, and (3) developing a sense of belonging (Richmond and Shields 2005; Schmdit 2007). Unfortunately, despite Canada’s consistent use of robust immigrant selection policies since the implementation of the points system in 1967, the country’s settlement policies have actually been in decline since the 1990s, now focusing mostly on initial reception (Arat-Koc 1999; Richmond and Shields 2005; Schmidt 2007). In terms of health promotion, the Migrant Integration Policy Index (2014) classifies Canada’s provincial-level migrant health policies as less cohesive than those coordinated by the U.S. Office of Minority Health. The American health care system seems more amenable to immigrant integration than universal health care in Canada because the latter does not provide immigrants with services comparable to the interpreters and other language supports funded by Medicaid.

The Healthy Immigrant Paradox in Mental Health

The healthy immigrant paradox was first noted by Markides and Coreil (1986), who found that while Mexican immigrants share many socioeconomic disadvantages with African Americans, they exhibit health advantages similar or close to those of non-Hispanic whites. Studies based on diagnoses of chronic conditions or biomarkers such as allostatic load have provided the strongest and most consistent evidence of this paradox (Chen, Ng, and Wilkins 1996; Kaestner et al. 2009; Lu and Ng 2019), whereas those relying on immigrants’ self-rated health measures tend to generate mixed evidence (Lu and Ng 2019; Newbold 2005).

Extant research on migrant mental health is even less consistent. In general, there is a foreign-born advantage when it comes to pathological psychological measures such as anxiety, mood disorders, and substance use disorders (Algeria et al. 2008; Dolly et al. 2012; Gee et al. 2007; Takeuchi et al. 2007). For nonpathological psychological measures such as self-rated mental health and psychological distress, however, the empirical testing of hypotheses on the healthy immigrant paradox rarely yields consistent results in this area (Dolly et al. 2012; Hurh and Kim 1990; Zhang et al. 2012). Some studies have found that immigrants’ mental health tends to decline after spending some time in the receiving country (evidenced through increased rates of distress and depression; Montazer and Wheaton 2017; Newbold 2009), while others point to improvements in immigrants’ mental health postmigration—be it over time or on nonlinear trajectories (Beiser 1988; Hurh and Kim 1990; Tran, Manalo, and Nguyen 2007). Further studies have pointed to discrepancies in the very measures of immigrants’ mental health. For instance, Dolly et al. (2012) found that immigrants have fewer mental disorders but worse self-rated mental health than nonimmigrants. In light of such inconsistencies, Harker (2001) argued that individual studies should draw on multiple mental health measures, including both positive and negative psychological outcomes.

Certain researchers have framed behavioral assimilation as a major pathway to health deterioration among immigrants (Chen et al. 1996); however, Newbold (2005) refuted this understanding by showing that such deterioration tends to occur at a much faster rate than behavioral changes postmigration. Other researchers point to the importance of psychosocial resources as mechanisms for maintaining immigrants’ physical and mental health (Angel, Buckley, and Sakamoto 2001; Kao and Tsai, 1986; Vega et al., 1991).

The Stress Process in the Context of Migration

The SPM is an emerging analytical framework that centers psychosocial resources and allows researchers to interpret health outcomes based on the structural locations of social agents (Brown and Turner 2010; Noh and Avison 1996; Young and Montazer 2018). This model treats status inequality as the starting point from which marginalized groups are exposed to greater psychosocial stressors that erode their psychosocial resources, eventually manifesting in greater mental health risks than those faced by more privileged groups (Brown and Turner 2010; Noh and Avison 1996; Pearlin et al. 1981). Mirowsky and Ross (2003) argued that life as a displaced person often creates feelings of alienation from the receiving society and, in turn, drives psychological distress. Indeed, migration is a major life event that disrupts the routines of premigration life and causes strain to the point of affecting psychosocial resources (Noh and Avison 1996). These include—but are not exclusive to—access to health care, linguistic familiarity, job stability and prestige, and social support (Christodoulou 2009; Finch, Bohdan, and Vega 2000; Finch, Reanne, and Vega 2004; Noh and Avison 1996).

Psychosocial Resources and Demands

There also seems to be a lack of consensus regarding immigrants’ abilities to forge stable social support networks. While some researchers have found immigrants to be highly resilient and capable of building new social support networks (Kao and Tsai 1986; Vega et al. 1991), others point to their fragmented social ties on arrival as a cause of significant—and often persistent—vulnerability (Menijivar 1997). A 1997 study by Menijivar showed how Salvadorian refugees seeking asylum in the United States face restrictive immigration policies, dwindling local economies, and scarce community resources—all of which make for a “handicapped context of reception” not conducive to building or maintaining ties with kinship or friends (p. 119). Nevertheless, another study indicated that at least in Canada, established Salvadorian refugees were more satisfied with the level of social support they received than newer cohorts (Young 2001).

The benefits of social support can be offset by interpersonal conflicts, and it is important that these be studied together where possible. According to Menijivar (1997) and Portes (1998), accessing social supports can result in negative social interactions because the recipients of said supports then feel obligated to reciprocate and stay permanently in the network. One form of such interpersonal conflict can arise from immigrant women’s increasing economic responsibilities combined with their husbands’ downward economic mobility (Min 2001). Also common are interpersonal conflicts between immigrant parents and their children, wherein the latter shoulder more communication responsibilities and the former feel a loss of parental control (Villanueva and Raymond 2001).

Psychological Distress and Positive Mental Health

Following the World Health Organization’s (2018) assertion that “Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity,” Statistics Canada has included the concept of complete mental health or positive mental health in several waves of its Canadian Community Health Survey. The concept and the scale of positive mental health were originally developed by Keyes (2005), who argued that mental disorders such as major depressive episodes, generalized anxiety disorder, and panic disorders are moderately and negatively correlated with positive mental health, representing two distinct psychological constructs.

On the other hand, Mirowsky and Ross (2003:28) suggested that good mental health can be defined by absence of distress because their estimation showed a strong negative correlation (–.7) between distress and positive psychological outcomes. Further to the work of Keyes (2002) and Mirowsky and Ross (2003), Payton (2009) used structural equation modeling to find that psychological distress and psychological well-being were moderately and negatively correlated, suggesting that these two constructs should be studied in their own right.

A 2009 study by Payton drew heavily on Ryff and Keyes’s (1995) six-dimension scale for measuring psychological well-being; however, researchers have yet to investigate a correlation between Keyes et al.’s (2008) positive mental health scale and Kessler et al.’s (2002) psychological distress scale. It is important to confirm that positive mental health and psychological distress are related but distinct psychological measures. Simply put, the justification for running separate regression models is premised on an understanding of psychological distress and positive mental health as distinct constructs (the strength of the coefficient no larger than ±.6). In the aforementioned studies by Payton (2009) and—less explicitly—Mirowsky and Ross (2003), a strong correlation is defined by a coefficient larger than ±.6. A moderate correlation, on the other hand, is defined by a coefficient falling between ±.4 and ±.59 (Evans 1996).

Linking Psychosocial Resources and Demands to Positive and Negative Psychological Outcomes

In previous studies, the SPM has mainly been used to find significant associations between psychosocial resources (demands) and negative psychological outcomes such as distress and depression (Noh and Avison 1996; Young and Montazer 2018). Nevertheless, Ingersoll-Dayton, Morgan, and Antonucci (1997) found that interpersonal strain has a weak and negative association with positive psychological outcomes. In this case, interpersonal strain might not mediate the relationship between migration and positive mental health given that a strong association is required for mediation to occur.

Additionally, Newsom et al. (2005) found that interpersonal strain is associated with both negative and positive psychological outcomes for older adults, while social support is only associated with positive psychological outcomes. This suggests that interpersonal strain can perhaps mediate the relationship between migration and positive psychological outcomes, but it also casts some doubts on the idea that social support has a suppressing effect on immigrants’ psychological distress (with which it is weakly associated).

These inconsistencies are mostly attributable to different measures of positive psychological outcomes (e.g., life satisfaction, happiness) and diverse negative psychological outcomes (e.g., depression, distress; Ingersoll-Dayton et al. 1997; Newsom et al. 2005; Okun and Keith 1998). The measures of social support and interpersonal strain also vary by study (Ingersoll-Dayton et al. 1997; Newsom et al. 2005; Okun and Keith 1998). Furthermore, studies comparing the effect of social support and interpersonal strain tend to focus largely on seniors because many of their relationships involve the dimension of caregiving and care-receiving that can be especially complicated when a person lacks the necessary resources to terminate any conflict-ridden relationships (Krause and Rook 2003). Immigrants’ network characteristics resemble those of senior citizens in that they are more likely to be comprised of family members engaged in resource-pooling and sharing (Hagan, Leal, and Rodriguez, 2015; Menijivar 1997). However, few studies compare the effect of social support and interpersonal strain on psychological distress and positive mental health. To do so, these two constructs must be standardized to compare their effects and relative contributions to psychological distress and positive mental health.

The Present Study

The first part of this study considers whether the relationship between years since migration and psychological distress/positive mental health follows a linear regression. The second part examines the correlation matrix between social support, interpersonal strain, positive mental health, and psychological distress. Finally, the third part investigates social support and interpersonal strain as mediating and moderating influences of both psychological distress and positive mental health.

Data and Methods

Sample

This study examines a total of 4,282 foreign-born immigrant samples extracted from the Canadian Community Health Survey, Mental Health 2012 (CCHS-MS), which originally had a national sample of 25,113 and a response rate of 86.3 percent. The CCHS-MH 2012 is a cross-sectional survey based on a stratified cluster design and covers respondents 15 years of age and older in all Canadian provinces and territories, excluding institutionalized persons, individuals living on reserves and other Indigenous settlements, and full-time members of the Canadian Forces. Missing values were handled using multiple imputation methods.

Piecewise Regression Modeling Approach

Past studies tend to treat years since migration as either a continuous measure with a linear relationship to health (Tran et al. 2007; Wu and Schimmele 2005) or as a categorical variable (broken into 10-year chunks) with a curvilinear relationship to health (Angel et al. 2001; Kaestner et al. 2009). However, the current literature has yet to fully articulate why the 10-year mark is automatically seen as a marker of long-term immigrant status. The decision to group years since migration into equal 10-year bins ultimately rests on considerable assumptions about the relationship between migration and health outcomes, as does fitting a linear regression line to address the relationship between the two. The current study thus employs a piecewise regression analysis as an alternative to such assumptions.

Focal Dependent Variables: Psychological Distress and Positive Mental Health

The association between years since migration and positive mental health is relatively linear; however, the relationship between years since migration and psychological distress shows a shallow curve with a peak at around 20 years since migration (see Figure 1). Figure 1 shows that initially, years since migration has a positive relationship with psychological distress. Then, approaching 20 years since migration, the relationship becomes negative. Given this curvilinearity, fitting one regression line masks the reality that immigrants’ psychological distress will increase or decrease at rates depending on the stage of settlement. Previously, the piecewise regression approach has been used in sociological studies of outcomes not proportional with time. For instance, Lavelle and Smock (2012) used this approach by placing the knot at the time of divorce and showing women’s health insurance coverage as being on a discrete downward slope at the time of divorce and for at least two years after. In a somewhat similar fashion, Aittomaki et al. (2010) demonstrated that decrease in self-rated health is more drastic for individuals whose household wealth falls below the second tercile point.

Figure 1.

Duration of migration and psychological distress.

Because mental health outcomes are not necessarily proportional to years since migration, this article uses a piecewise regression modeling approach to reflect a sudden increase in psychological distress. Visually, it is difficult to pin down a particular year since migration in which immigrants experience more drastic changes to their mental health, although Figure 1 suggests this to be anywhere around 20 years since migration. To estimate the peak, I used Stata 13.0 to place a knot that produced the smallest errors based on the ordinary least squares. When it comes to psychological distress, the estimation placed the knot at 17 years. A Wald test was applied to ensure there were significant differences in psychological distress scores before and after 17 years since migration.

Although the relationship between years since migration and positive mental health appears flat, I repeated the procedure for positive mental health to avoid inaccurate visual diagnosis. Unsurprisingly, Stata 13.0 did not detect any knot. Thus, fitting only one slope is enough to describe the relatively flat relationship between years since migration and positive mental health.

Mediators and Suppressors: Interpersonal Strain and Social Support

The relationship between years since migration and social support appears to be linear with a relatively flat slope. Additional estimations using Stata 13.0 suggest that fitting one slope is sufficient for describing the relationship between years since migration and social support.

However, the relationship between years since migration and interpersonal strain is curvilinear, just like the relationship between years since migration and psychological distress. Again, I used Stata 13.0 to estimate the knot between the two, and again the results suggested placing the knot at 20. For the sake of simplicity, I chose to place the knot at 17 instead. The Wald test still showed a significant difference in interpersonal strain scores before and after 17 years since migration.

Because two slopes were fitted to describe the curvilinear relationship between years since migration and interpersonal strain, one may observe these slopes as occurring in opposite directions: one positive and the other negative. The opposite direction of slope will influence whether interpersonal strain mediates or suppresses the focal association between years since migration and psychological distress. The following paragraph briefly explains the reasoning behind determining whether a psychosocial variable functions as a mediator or as a suppressor and provides a guideline for statistical interpretation of the relevant findings.

According to Aneshensel (2013), whether a psychosocial variable acts as a suppressor or a mediator will depend on the directions of their relationships with the focal independent variable and the dependent variables. The initial focal association (i.e., years since migration and psychological distress) might be suppressed or concealed by failing to consider the influence of psychosocial factors. By considering the influence of a psychosocial variable as a suppressor, the initial null focal association should become significant. On the other hand, if the focal association is significant, its magnitude should be weakened when a psychosocial variable (e.g., social support or interpersonal strain) operates as a mediator.

Measures

Focal Dependent Variables

The first focal dependent of this study is psychological distress. This is measured by the Kessler Distress Scale (K10), a 10-item index that asks respondents how often in the past four weeks they felt “tired out for no good reason,” “nervous,” “hopeless,” “restless and fidgety,” “depressed,” “everything was an effort,” “so sad that nothing would cheer them up,” or “worthless.” Scores on the K10 range from 0 to 40 (α = .85).

Another focal dependent variable is positive mental health, measured by a 14-item scale constructed by Keyes (2002). The factor structure of the positive mental health scale includes three dimensions of well-being: emotional, psychological, and social (Keyes 2005). The positive mental health scale has a total score of 70 (α = .86). Each item involves a 6-point response scale, which asked respondents to report how often that they felt “happy,” “interested in life,” “satisfied with [their] life,” “that [they] had something important to contribute to society,” “that [they] belong to a community,” “that people are basically good,” “that our society makes sense to [them],” “that [they] like most parts of [their own] personality,” “that [they] are good at managing the responsibilities of [their own] life,” “that [they] have warm and trusting relationships with others,” and “that [their] life has a sense of direction and meaning.” All items were rescaled and reverse-coded from 1 to 6 to 5 to 0 (1 = every day, 2 = almost every day, 3 = about 2 to 3 times per week, 4 = about once a week, 5 = once or twice, 6 = never) so that higher scores reflected better positive mental health.

Focal Independent Variable: Years since Migration

Piecewise regression was used for all models regressing years since migration on interpersonal strain and psychological distress. For these models, years since migration is broken into two continuous measures: the first representing the years 0 to 17 since migration and the second representing 18 or more years since migration. In models regressing years since migration on social support and positive mental health, only one slope is fitted. Both social support and positive mental health are modeled as one continuous measure.

Mediator and Suppressor

The social support scale involves 10 five-point Likert-scale questions (α = .93). Respondents were asked to report to what extent (from strongly agree to strongly disagree) they felt that “there are people to depend on to help [them] if [they] really need it,” “there are people who enjoy the same social activities as [they] do,” “there is someone [they] can talk to about important life decisions,” “[they] have relationships where [their] competence and skills are recognized,” “[they] feel part of a group of people who share [their] beliefs and attitudes,” and “[they] feel a strong emotional bond with a least one other person.”

Interpersonal strain is determined by the summate score of four 5-point Likert-scale questions (α = .81). Respondents were asked to report how often (from never to very often) in the past month they felt “others made too many demands on [them],” “others were critical of [them] and the things [they] did,” “others did things that were thoughtless and inconsiderate,” and “others acted angry or upset with [them].”

Control Variables

The demographic and socioeconomic factors considered in this study include gender, race, household composition, marital status, household income, education, current employment status, full-time/part-time employment status, language spoken at home, chronic conditions, and caregiving responsibilities. Age was not included in the model because it is strongly correlated with years since migration. I also controlled for health behaviors, including smoking and drinking.

Gender was dummy-coded as males (reference) and females. Race or ethnic background based on country of origin was dummy-coded as British (reference), American, German, Dutch, Italian, whites from other European countries, whites from non-European countries, blacks from Africa, blacks from other countries, South/Central Americans/Caribbeans, Filipino, Chinese, South Asians, South East Asians, Arabs, Latino, Korean/Japanese/West Asian, other race, and multiple origins.

Family size was coded as one-person household (reference), two persons, three to five persons, and more than six persons. Marital status was dummy-coded into three categories: (1) single (reference), (2) married or common-law, and (3) divorced, separated, or widowed.

To create the income adequacy variable, I divided total household income by the number of household members. The upper 25% income adequacy level was coded as the highest, the lower 25% (reference) was coded the lowest, and those falling in between were coded as upper-middle and lower-middle. Education was coded into three dummy variables: less than high school (reference), high school, and postsecondary degree. Employment status was coded as: currently working (reference), not working, retired, and permanently unable to work. Full-time versus part-time employment status was coded as full-time, part-time, and school/household responsibilities/retirement.

Language spoken at home was categorized as not speaking either English or French, speaking either English or French and another foreign language, and only speaking either English or French (reference).

I opted to include chronic conditions in the model due to the association between mental health and physical health—especially among older immigrants. This variable was dummy-coded as yes versus no. Additionally, because immigrants are more likely than nonimmigrants to delay the institutionalization of mentally ill family members (Mausbach et al. 2004), informal caregiving responsibilities for family members with mental health issues was dummy-coded into three variables: not a caregiver (reference), currently a caregiver, and no family members.

The health behaviors I opted to include in the model were smoking and drinking. Smoking behaviors were dummy-coded as nonsmokers, former smokers, occasional smokers, and daily smokers. Drinking behaviors were dummy coded as nondrinkers, occasional drinkers, and daily drinkers.

Plan of Analysis

Table 1 provides descriptive statistics on Canada’s immigrant population. Table 2 shows the correlation matrix of mental health measures and psychosocial resources. Table 3 presents the association between years since migration and social support fitting only one slope (Model 1), and a piecewise regression model reveals differential exposure to interpersonal strain before and after the knot (Model 2). In Table 4, Models 1 through 5 were produced using piecewise regression modeling. These show the focal association between years since migration and psychological distress after controls, how it changes with the additions of social support and interpersonal strain (separately and together), and finally, how social support and interpersonal strain interact with one another. In Table 5, Models 1 through 5 show the focal association between years since migration and positive mental health after controls (fitting only one regression line), changes to said association upon entering social support and interpersonal strain (separately and together), as well as the interaction between social support and interpersonal strain.

Table 1.

Sample Descriptions (Canadian Community Health Survey, Mental Health 2012; N = 4,282.)

Variable	Percentage, Mean, and SD (weighted)
Years since migration	0–maximum
Mean	23
Standard deviation	.01
Race and ethnicity	Percentage (weighted), total = 100
English (reference group)	6.96
American	2.52
German	2.23
Dutch	1.58
Italian	3.60
White (other European countries)	15.69
White (non-European countries)	4.70
Black (South and Central America/Caribbean)	3.88
Black (Africa)	2.22
Black (other countries)	.18
Filipino	7.23
Chinese	13.79
South Asian	13.57
South East Asian	3.14
Arab	5.39
Japanese/Korean/West Asian	3.87
Latino	5.59
Other races/ethnicities	3.14
Multiple race/ethnicities	2.62
Gender	Percentage (weighted), total = 100
Male (reference group)	49.08
Female	50.92
Language spoken at home	Percentage (weighted), total = 100
English or French (reference group)	34.25
English, French, and other	34.00
Other only	31.75
Household size	Percentage (weighted), total = 100
1 person (reference group)	12.38
2 persons	29.95
3–5 persons	50.65
>6 persons	7.02
Marital status	Percentage (weighted), total = 100
Single (reference)	19.26
Married or common-law	67.64
Divorced, separated, and widowed	13.10
Education	Percentage (weighted), total = 100
< Secondary (reference group)	16.07
Secondary graduate	13.35
Postsecondary degree	70.58
Income level	Percentage (weighted), total = 100
Lowest	30.30
Lower-middle	26.06
Upper-middle	24.15
Highest	19.49
Work status-1	Percentage (weighted), total = 100
Currently working (reference)	58.36
Currently not working	32.59
Retired	7.20
Permanently unable to work	1.84
Work status-2
Full-time (reference)	52.59
Part-time	8.82
School/domestic responsibilities/retired	35.59
Chronic conditions	Percentage (weighted), total = 100
Yes	51.46
No (reference)	48.54
Caregiving responsibilities	Percentage (weighted), total = 100
Yes	20.12
No(reference)	78.23
No family members	1.65
Smoking behavior	Percentage (weighted), total = 100
Nonsmoker (reference)	50.19
Former smoker	30.13
Occasional smoker	5.86
Daily smoker	8.82
Drinking behavior	Percentage (weighted), total = 100
Nondrinker (reference)	34.65
Occasional drinker	19.96
Regular drinker	45.39
Social support	Range,10–40
Mean	35.11
Standard deviation	.11
Interpersonal strain	Range, 0–12
Mean	2.34
Standard deviation	.05
Psychological distress	Range, 0–40
Mean	4.44
Standard deviation	.11
Positive mental health	Range, 0–70
Mean	54.24
Standard deviation	.01

Table 2.

Correlation Matrix of Psychosocial Resources and Mental Health Measures (N = 4,282).

	Psychological Distress	Positive Mental Health	Interpersonal Strain	Social Support
Psychological distress	1
Positive mental health	−.502	1
Interpersonal strain	.370	−.253	1
Social support	−.289	.405	−.089	1

Note. All significant at .001 level.

Table 3.

Effect of Years since Migration on Social Support and Interpersonal Strain (Canadian Community Health Survey, Mental Health 2012; N = 4,282).

Variable	Model 1 (OLS: Social Support) Unstandardized Coefficients	Model 2 (OLS: Interpersonal Strain) Unstandardized Coefficients
Migration
Years since migration	.000
Migration
0–17 years		.010*
18+ years		−.010**
Female (male = 0)	.060***	.148***
Number of persons in household
1	—	—
2	.110***	.078
3–5	.159***	.219***
6+	.167***	.007
Income adequacy
Lowest	—	—
Lower-middle	.026	−.039
Upper-middle	.105***	−.085
Highest	.179***	−.001
Work status
Employed	—	—
Not employed	.080	−.181
Permanently unable to work	−.214*	.030
Retired	.112*	−.376**
Have caregiving responsibilities
No	—	—
Yes	−.042*	.439***
No family members	.010	−.077
Race and ethnicity
English	—	—
American	.080*	.005
German	.009	−.005
Dutch	−.111	.014
Italian	−.001	−.292*
White (other European countries)	−.026	−.174*
White (non-European countries)	−.021	−.010
Black (South and Central America/Caribbean)	−.082	−.215*
Black (Africa)	−.055	−.067
Black (other countries)	.006	.398
Filipino	−.082	−.108
Chinese	−.238***	−.178
South Asian	−.079*	−.204*
South East Asian	−.094	−.017
Arab	−.057	−.007
Latino	−.030	−.173
Other	−.069	−.090
Multiple origins	−.067	−.056
Japanese, Korean, West Asian	−.144**	.066
Chronic morbidity (yes = 1)	−.062***	.173***
Smoking behavior
Nonsmoker	—	—
Former smoker	.005	.060
Occasional smoker	−.016	.157
Daily smoker	−.020	.161*
Drinking behavior
Nondrinker	—	—
Occasional drinker	.012	.025
Daily drinker	−.005	.135**
Constant	5.822***	.897***
Adjusted R²	11.822%	18.555%

Note. Other control variables include: marital status, education, language spoken at home, full-time/part-time employment status. OLS = ordinary least squares.

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

Table 4.

Effect of Years since Migration on Psychological Distress (Canadian Community Health Survey, Mental Health 2012; N = 4,282).

Variable	Model 1 (Controls) Unstandardized Coefficients	Model 2 (Social Support) Unstandardized Coefficients	Model 3 (Interpersonal Strain) Unstandardized Coefficients	Model 4 (Social Support and Interpersonal Strain) Unstandardized Coefficients, Standardized Coefficients in Parentheses	Model 5 (Interaction between Social Support and Interpersonal Strain) Unstandardized Coefficients
Migration
0–17years	−.007	−.010	−.012*	−.011*	−.011*
18+ years	−.011***	−.010***	−.008**	−.008**	−.008***
Female (male = 0)	.211***	.262***	.141**	.190***	.195***
Income adequacy
Lowest	—	—	—	—	—
Lower-middle	.074	.096	.092	.109	.107
Upper-middle	−.012	.076	.028	.100	.103
Highest	−.089	.061	−.089	.040	.042
Work status
Employed	—	—	—	—	—
Not employed	.113	.180	.199	.249*	.236*
Permanently unable to work	1.376***	1.196***	1.362***	1.209***	1.198***
Retired	.025	1.232	.203	.268	.246
Have caregiving responsibilities
No	—	—	—	—	—
Yes	.462***	.427***	.255	.243***	.244***
No family members	.072	.035	.109	.073	.076
Race and ethnicity
English	—	—	—	—	—
American	.155	.222*	.153	.210*	.205
German	.394**	.402*	.397*	.403	.399*
Dutch	.353**	.260	.347*	.267*	.256
Italian	.273*	.272*	.411***	.398***	.380**
White (other European countries)	.113	.092	.196*	.170*	.163
White (non-European countries)	.172	.154	.168	.152	.147
Black (South and Central America/Caribbean)	.219	.150	.321	.253	.240
Black (Africa)	−.145	−.191	−.113	−.155	−.170
Black (other countries)	.174	.179	−.014	.007	.010
Filipino	−.030	−.100	.021	−.043	−.038
Chinese	.051	−.150	.135	−.044	−.032
South Asian	.170	.103	.267**	.201	.186
South East Asian	.450**	.370**	.457**	.389**	.377**
Arab	.367**	.318*	.370**	.328*	.300*
Latino	−.045	−.070	.037	.008	−.008
Other	.240	.181	.282	.228	.221
Multiple origins	−.096	−.153	−.069	−.120	−.133
Japanese, Korean, West Asian	.332*	.210	−.301	.199	.192
Chronic morbidity (yes = 1)	.405**	.353***	.323***	.286***	.287***
Smoking behavior
Nonsmoker	—	—	—	—	—
Former smoker	−.036	−.032	−.065	−.058	−.054
Occasional smoker	.194	.180	.119	.114	.132
Daily smoker	.183*	.167*	.107	.100	.093
Drinking behavior
Nondrinker	—	—	—	—	—
Occasional drinker	.126	.136	.114	.124	.117
Regular drinker	.227***	.223***	.163**	.165**	.168**
Social support		−.841***		−.719*** (–.233)	−.489***
Interpersonal strain			.474***	.431*** (.318)	1.474***
Social support × interpersonal strain					−.177**
Constant	1.036***	5.930***	.611***	4.835***	3.475***
R²	13.906%	20.465%	23.842%	28.557%	28.855%

Note. Other control variables include: number of persons in household, marital status, education, language spoken at home, full-time/part-time employment status.

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

Table 5.

Effect of Years since Migration on Positive Mental Health (Canadian Community Health Survey, Mental Health 2012; N = 4,282).

Variable	Model 1 (Controls) Unstandardized Coefficients	Model 2 (Social Support) Unstandardized Coefficients	Model 3 (Interpersonal Strain) Unstandardized Coefficients	Model 4 (Social Support and Interpersonal Strain) Unstandardized Coefficients, Standardized Coefficients in Parentheses	Model 5 (Interaction between Social Support and Interpersonal Strain) Unstandardized Coefficients
Migration
Years since migration	.066***	.061***	.061***	.057***	.058***
Female (male = 0)	−.467	−1.106**	−.022	−.710	−.755
Number of persons in household
1	—	—	—	—	—
2	.925	−.347	1.181	−.062	−.123
3-5	1.540**	−.158	2.252**	.519	.559
6+	2.152	.369	2.278	.581	.561
Income adequacy
Lowest	—	—	—	—	—
Lower-middle	.275	−.000	.164	−.072	−.041
Upper-middle	−.746	−1.867***	−.978	−1.983***	−2.023***
Highest	.787	−1.122	.808	−.986	−1.009
Work status
Employed	—	—	—	—	—
Not employed	−1.436	−2.288	−2.005	−2.691*	−2.540*
Permanently unable to work	−11.192***	−8.909***	−11.088***	−8.970**	−8.836**
Retired	−1.312	−2.505*	−2.516	−3.395*	−3.144*
Have caregiving responsibilities
No	—	—	—	—	—
Yes	−2.412**	−1.967***	−1.112	−.953	−.965*
No family members	−.224	.253	−.452	.041	.013
Race and ethnicity
English	—	—	—	—	—
American	−.126	−.977	−.068	−.877	−.810
German	−3.605**	3.706*	−3.625**	−3.715*	−3.669*
Dutch	−2.919	−1.728	−2.946	−1.825	−1.695
Italian	−2.913*	−2.905*	−3.828*	−3.639**	−3.432**
White (other European countries)	−2.324**	−2.049*	−2.770*	−2.423**	−2.340**
White (non-European countries)	−1.112	−.884	−.965	−.781	−.714
Black (South and Central America/Caribbean)	−1.390	−.516	−1.852	−.942	−.782
Black (Africa)	1.531	2.118	1.436	2.004	2.178
Black (other countries)	−8.171*	−8.232**	−6.866**	−7.182*	−7.193**
Filipino	1.526	2.402	1.252	2.131	2.077
Chinese	−4.402***	−1.854	−4.777***	−2.315*	−2.456*
South Asian	−1.064	−.216	−1.536	−.648	−.471
South East Asian	−2.214	−1.210	−2.097	−1.179	−1.042
Arab	−2.743*	−2.134	−2.687*	−2.128	−1.791
Latino	.514	−.198	−.940	−.560	−.363
Other	−.752	−.566	−1.461	−.738	−.652
Multiple origins	−1.304	.235	−.509	.168	.326
Japanese, Korean, West Asian	−6.131***	−.589***	−5.814***	−4.432***	−4.347***
Chronic morbidity	−1.725***	−1.065*	−1.188**	−.676	−.695
Smoking behavior
Nonsmoker	—	—	—	—	—
Former smoker	−.745	−.800	−.556	−.645	−.694
Occasional smoker	−1.056	−.882	−.550	−.487	−.696
Daily smoker	−2.063*	−1.852*	−1.563*	−1.645*	−1.385
Drinking behavior
Nondrinker	—	—	—	—	—
Occasional drinker	−1.231*	−1.365**	−1.168*	−1.305*	−1.228*
Regular drinker	−2.064***	−2.008***	−1.668**	−1.694**	−1.734**
Social support		10.682***		10.011*** (0.379)	7.332***
Interpersonal strain			−2.968***	−2.379*** (–0.205)	−14.538***
Social support × Interpersonal strain					2.066**
Constant	57.600***	−4.588	60.412***	1.574	17.482**
Adjusted R²	9.857%	24.303%	15.209%	27.685%	28.240%

Note. Other control variables include: marital status, education, language spoken at home, full-time/part-time employment status.

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

Results

Descriptive Statistics

Table 1 shows that among immigrants, the average number of years spent in Canada is 23. Immigrants from the UK, Germany, the Netherlands, and other European countries comprise over one-third of the immigrant population. The other major groups of Canadian immigrants are Chinese and South Asian, each constituting roughly 15 percent of the entire immigrant population. Filipino immigrants and black immigrants from Africa, the Caribbean, South America, and Central America account for approximately 7 percent and 6 percent, respectively, while Arabs and Latinos each represent around 5 percent of the overall immigrant population.

Today’s Canadian immigrants are well educated, with approximately 70 percent of those surveyed holding postsecondary degrees. Over half are employed, and one-third are not in the labor force due to domestic responsibilities, difficulties findings jobs, or being in school. The average income per immigrant household is $68,581, with half of said incomes falling in the range of $20,000 to $69,999 per year (not shown in the table). Two-thirds of those surveyed were married or in common-law relationships, one-fifth were single, and the rest were divorced, separated, or widowed. Around half of respondents report living in a household with three to five family members, around one-third are in two-person households, slightly over 10 percent live alone, and roughly 7 percent live with more than six people. The sample consisted of slightly more female than male respondents (50.92 percent and 49.08 percent, respectively). As for languages spoken at home, approximately one-third (31.8 percent) of respondents speak their native language only, another third (34 percent) speak both their native language and one official language, and the remaining third speak only an official language.

Half of the immigrants surveyed self-identified as nonsmokers, and approximately one-third self-identified as former smokers. Less than 10 percent are daily smokers, and around 6 percent are occasional smokers. In terms of drinking behaviors, over one-third are nondrinkers, but around 45 percent are regular drinkers. Almost one in five are occasional drinkers.

The mean psychological distress of Canadian immigrants is 4.44 on the K10 scale, which is significantly lower than the overall Canadian sample (mean = 5.27; not shown in the table). The mean positive mental health score of Canadian immigrants is 55.17, which is significantly higher than the overall Canadian sample (mean = 54.24; not shown in the table). Immigrants on average scored 35.11 out of 40 on the social support scale and 2.34 out of 12 on the interpersonal strain scale.

Table 2 presents the correlation between social support, interpersonal strain, psychological distress, and positive mental health, all of which are significant at the .001 level. The correlation between distress and positive mental health is negative and moderate (–.5), thus supporting the interpretation that these are interrelated but distinctive psychological constructs. Social support and interpersonal strain, however, are weakly and negatively associated.

Years since Migration and Psychosocial Resources: Mediation and Suppression

Model 1 in Table 3 shows a nonsignificant relationship between years since migration and social support after controlling for sociodemographic background and socioeconomic conditions. The coefficient of 0 indicates that staying longer in Canada does not lead to increases in immigrants’ social support.

Without the use of any controls, years since migration and interpersonal strain appear to have a curvilinear relationship (results not shown). However, after controlling for sociodemographic background, socioeconomic conditions, and health behaviors, years since migration before and after the knot are both significantly associated with interpersonal strain (Table 3, Model 2). Specifically, the association between years since migration and interpersonal strain is positive up to year 17 postmigration, at which point the relationship is reversed and continues downward for years to come.

Pathways to Psychological Distress: The Effect of Psychological Demands

Without any controls, up to 17 years since migration and psychological distress are shown to have a nonsignificant positive association. After 17 years, the association becomes significantly negative, as Figure 1 suggests (coefficients not shown in models because the results reflect Figure 1).

Model 1 in Table 4 shows that after controlling for demographic background, socioeconomic factors, and health behaviors, the association between years since migration before the knot becomes negative, although it is still insignificant. Demographic variables such as gender, race and ethnicity, and health behaviors share significant associations with psychological distress; however, socioeconomic factors such as income and education do not predict immigrants’ psychological distress. Caregiving responsibilities are shown to increase psychological distress. Taken together, these variables explain 14 percent of the variance in psychological distress.

Model 2 in Table 4 shows that social support explains around 6.6 percent of the variance in psychological distress (R² in Model 2 – R² in Model 1 = 20.47% − 13.91% = 6.56%). Recalling how Model 1 in Table 3 establishes that years since migration have no effect on social support, this model might be taken as an indication that years since migration also has no indirect effect on psychological distress through the added dimension of social support.

Model 3 in Table 4 was designed to determine whether interpersonal strain mediates the relationship between years since migration and psychological distress. Interpersonal strain has a significant effect on and contributes to 10 percent of the explained variance in psychological distress. In Table 3, Model 2 shows years since migration to have a curvilinear relationship with interpersonal strain, where the relationship is positive before the knot and negative after the knot. After introducing interpersonal strain to Model 3 for Table 4, the initial nonsignificant negative relationship between years since migration and psychological distress in Model 1 (Table 4) at the aforementioned knot (17 years postmigration) becomes significantly negative. The Sobel test confirms that years since migration (before and after the knot) has a significant indirect effect on psychological distress—specifically, via interpersonal strain. Interpersonal strain is regarded as a mediator between years since migration and psychological distress after the knot (placed at 17 years since migration) because the coefficient of the focal association becomes smaller (from –.11 to –.08). Before the knot, however, it functions as a suppressor between years since migration and psychological distress because the coefficient of the focal association becomes larger (from –.007 to –.012). The interpretation is that when a direct effect (years since migration on psychological distress) is negative in direction while a concurrent, indirect effect (via interpersonal strain) is positive, it causes an initial null association. The significant focal association becomes significant only after controlling for interpersonal strain. This is a case of suppression in the SPM.

Moving forward with this interpretation, it appears that immigrants living in Canada less than 18 years (who landed after 1995) would feel less distressed were it not for their high exposure to interpersonal strain. Adding interpersonal strain to the model (holding it constant) reflects how immigrants in the mid-1990s would feel if their psychosocial demands were equal to those of earlier immigrant cohorts.

Model 4 in Table 4 presents the standardized effects of social support and interpersonal strain. Put simply, the results indicate that interpersonal strain has a greater effect on immigrants’ psychological distress than social support (and the Wald test shows that the effects of interpersonal strain and social support are significantly different). One unit of increase in interpersonal strain increases .32 in psychological distress, while one unit of increase in social support only decreases .23.

Model 5 in Table 4 shows the interaction effect of social support and interpersonal strain on psychological distress. For immigrants experiencing the same level of interpersonal strain, those with an additional social support score will experience lower psychological distress. Although the interaction is significant, the additional explained variance is very small (less than .5%).

Pathways to Positive Mental Health: the Effect of Psychological Demands

Model 1 in Table 5 controls for sociodemographic background, socioeconomic conditions, and health behaviors, ultimately showing a significant positive focal association between years since migration and positive mental health. Together, these variables explain around 10 percent of the variance in positive mental health.

Model 2 in Table 5 shows that both years since migration and social support are significantly and positively associated with positive mental health. Social support explains around 14 percent of the variance in positive mental health. However, while years since migration has a direct effect on positive mental health, it has no indirect effect on positive mental health routed through social support because longer residence in Canada does not help increase social support (see Model 1, Table 3 for the insignificant association between years since migration and social supports).

Interpersonal strain has a negative and significant association with positive mental health, but its contribution is much smaller, explaining around 5.5 percent of the variance. Unlike with psychological distress, the findings of this study do not favor interpretations of interpersonal strain as a mediator for the association between years since migration and positive mental health. Visually speaking, the relationship between years since migration and positive mental health appears flat, while the relationship between years since migration and interpersonal strain is curvilinear. Fitting one slope still reveals a significant focal association between years since migration and positive mental health; however, years since migration and interpersonal strain have no significant association fitting only one slope.

Model 4 in Table 5 compares the standardized effect of social support and interpersonal strain on positive mental health. Results suggest that compared to interpersonal strain, social support has a greater effect on positive mental health. With each unit of increase in social support, there is a .38 unit increase in positive mental health. By comparison, each unit of increase in interpersonal strain brings only a .21 unit decrease in positive mental health. Wald testing shows that social support and interpersonal strain have different effects on positive mental health.

Finally, Model 5 in Table 5 shows a significant interaction between social support and interpersonal strain as affecting positive mental health. This suggests that social support can buffer the negative effects of interpersonal strain. For immigrants who experience interpersonal strain, those with more social support experience additional gains in positive mental health over those without social support. Nonetheless, the increased variance is smaller than .5%.

Discussion

Much of the existing research assumes (1) that the relationship between years since migration and health is linear and (2) that immigrants’ health advantages dissipate over time (Torres and Wallace 2013; Zhang et al. 2012). Yet, years since migration and psychological distress follows a shallow and inverted J curve, with immigrants arriving to Canada in the mid-1990s experiencing the highest level of psychological distress. When such relationships are curvilinear, fitting one slope will fail to detect the contextual factors that contribute to their curvilinearity.

Midterm immigrants were those who landed in Canada during the 1990s. A Canadian study that used the 1994–95 National Population Household Survey to focus specifically on these immigrants indicated they had reported better health than those who had landed one decade prior (Chen et al. 1996). By comparison, based on the findings using CCHS-MH 2012, this study reveals that 20 years later, these immigrants from the 1990s were experiencing higher levels of distress than earlier or later cohorts.

One key takeaway from this study is that the relationship between migration and mental health must be interpreted in ways that center historical context. It should be understood, for example, that Canada’s socioeconomic receiving context was particularly hostile to immigrants arriving in the mid-1990s and improved over the following decade (Arat-Koc 1999; Richmond and Shields 2005). What is more, the results of the current study further point to how such receiving contexts come to shape immigrants’ psychosocial and mental health experiences. In recent years, the American receiving context has been increasingly difficult for immigrants to integrate. The Illegal Immigration Reform and Immigrant Responsibility Act passed in 1996 expanded border control and limited noncitizens’ access to welfare and health care; 30 percent of the legal immigrants lost their health coverage as a result (Hagan et al. 2003). Furthermore, although the US Deferred Action for Childhood Arrivals (DACA) has been found to reduce distress for eligible recipients (Venkataramani et al. 2017), the future of this program is now highly uncertain. Here it should be noted that when immigrants are deported, part of the human and social capital embedded in the low-resource immigrant communities are deported, making it harder to thrive for immigrants who stay behind (Hagan et al. 2015).

Most quantitative studies treat individuals as units of analysis. This study is no exception; however, social support and interpersonal strain are also viewed as modifiable resources existing among network members. This study shows that the Canadian receiving context is not conducive to enhancing social support for immigrants and that interpersonal strain spikes during the economic recession and welfare restructuring. Using piecewise regression analysis, this study demonstrates that years since migration has an indirect effect on psychological distress through interpersonal strain. However, years since migration has no similar indirect effect on positive mental health through interpersonal strain.

Both social support and interpersonal strain affect psychological distress and positive mental health. Social support does have a buffering effect (albeit minimal) against interpersonal strain when it comes to both psychological distress and positive mental health. Immigrants who originally have higher levels of social support would experience some boost in positive mental health or a reduction in psychological distress; however, the additional explained variance is very small.

The moderate and negative association between psychological distress and positive mental health revealed through this study highlights the distinctiveness of the two mental health measures and, by extension, loans support to the argument that for research involving immigrants in particular, these two measures should be studied in their own right. Additionally, the study also indicates that psychosocial resources contribute more to positive psychological constructs than psychological demands. By contrast, psychosocial demands contribute more to negative mental health measures than psychosocial resources.

Of course, the cross-sectional design of the CCHS-MH 2012 limited my ability to consider how the period effect and the duration effect might separately affect immigrants’ mental health. It is most difficult to disentangle the lasting impacts of a harmful receiving context in the past from duration of migration and immigrants’ current health outcomes, although some researchers have found interpersonal strain to be a persistent challenge that, once begun, is often difficult to solve (Krause and Rook 2003). Moreover, it should be understood that causal interpretation is not possible in a study such as this. One cannot determine with certainty, for example, whether interpersonal strain increases mental health risks or the opposite.

Future researchers may also want to consider the policy implications of the weak but nonetheless unique association between immigrants’ social support and interpersonal strain. It raises the question as to whether immigrants experience support and demands simultaneously from the same network members. Failing to measure sources of social support and interpersonal strain (whether they come from friends, co-workers, or strangers) excludes the possibility of introducing clear policy intervention in workplaces, schools, and families.

Finally, although this study controls for race and ethnicity, it does not offer interpretation. A major reason for this is because the age structures of different ethnic groups often vary a great deal. German and Italian immigrants have a mean age around 60 years, while the average Chinese immigrant is approximately 40 years old. For the future, comparing the health outcomes of ethnic groups with such different age structures will require an age-standardized approach and considerably larger sample sizes.

Footnotes

Acknowledgements

I thank Margaret Denton, Vic Satzewich, Paul Glavin, James M. Stutely, Mollie McQuire, and the three anonymous reviewers for their valuable comments on this article.

ORCID iD

Fei-Ju Yang

References

Aittomaki

Akseli

Martikainen

Pekka

Laaksonen

Mikko

Lahelma

Eero

Rahkonen

Ossi

. 2010. “The Associations of Household Wealth and Income with Self-Rated Health – A Study on Economic Advantage in Middle-Aged Finnish Men and Women.” Social Science & Medicine 71(5):1018–26.

Algeria

Margarita

Canino

Glorisa

Shrout

Patrick E

Woo

Meghan

Duan

Naihua

Vila

Doryliz

Torres

Maria

Chen

Chih-Nan

Meng

Xiao-Li

. 2008. “Prevalence of Mental Illness in Immigrant and Non-immigrant U.S. Latino Groups” American Journal of Psychiatry 165(3):359–69.

Aneshensel

Carol

. 2013. Theory-Based Data Analysis for the Social Sciences. Thousand Oaks, CA: Sage Publications.

Angel

Jaqueline L.

Buckley

Cynthia J.

Sakamoto

Art

. 2001. “Duration or Disadvantage? Exploring Nativity, Ethnicity, and Health in Midlife.” The Journals of Gerontology: Series B 56(5):S275–84.

Antecol

Heather

Cobb-Clark

Deborah A.

Trejo

Stephen J

. 2003. “Immigration Policy and the Skills of Immigrants to Australia, Canada, and the U.S.” The Journal of Human Resources 38(1):192–218.

Arat-Koc

1999. “Neo-Liberalism, State Restructuring and Immigration: Changes in Canadian Policies in the 1990s.” Journal of Canadian Studies 34(2):31–56.

Beiser

Morton

. 1988. “Influence of Time, Ethnicity, and Attachment.” American Journal of Psychiatry 1:46–51.

Belot

Michèle V. K.

Hatton

Timothy J.

2012. “Immigrant Selections in the OECDs.” The Scandinavian Journal of Economics 114(4):1105–28.

Brown

Robyn L.

Turner

Jay R.

, 2010. “Physical Disability and Depression: Clarifying Racial/Ethnic Contrasts.” Journal of Aging and Health 22(7):977–1000.

10.

Chen

Jiajian

Edward

Wilkins

Russell

. 1996. “The Health of Canada’s Immigrants in 1994-95.” Health Reports 7(4):33–45.

11.

Christodoulou

Irene

. 2009. “Immigration-Induced Syndromes.” International Journal of Health Sciences 2(1):145–46.

12.

Dolly

Castro

A. B. de

Martin

Diana P.

Duran

Bonnie

Takeuchi

David T.

2012. “Does an Immigrant Health Paradox Exist among Asian Americans? Associations of Nativity and Occupational Class with Self-Rated Health and Mental Disorders.” Social Science & Medicine 75(12):2085–98.

13.

Evans

James D.

1996. Straightforward Statistics for the Behavioral Sciences. Los Angeles, Brooks/Cole Publishing Company.

14.

Farre

Lidia

. 2016. “New Evidence on the Healthy Immigrant Effect.” Journal of Population Economics 29(2):365–94.

15.

Finch

Brian K.

Bohdan

Kolody

Vega

William A.

2000. “Perceived Discrimination and Depression among Mexican-Origin Adults in California.” Journal of Health and Social Behavior 41(3):295–313.

16.

Finch

Brian K.

Reanne

Frank

Vega

William A.

2004. “Acculturation and Acculturation Stress: A Social-Epidemiological Approach to Mexican Migrant Farmworkers’ Health.” International Migration Review 38(1):236–62.

17.

Gee

Gilbert C.

Spencer

Michael

Chen

Juan

Yip

Tiffany

Takeuchi

David T.

2007. “The Association between Self-Reported Racial Discrimination and 12-Month DSM-IV Mental Disorders among Asian Americans Nationwide.” Social Science & Medicine 64(10):1984–96.

18.

Hagan

Jacqueline

Leal

David

Rodriguez

Nester

. 2015. “Deporting Social Capital: Implications for Immigrant Communities in the United States.” Migration Studies 3(3):370–92.

19.

Hagan

Jacqueline

Rodriguez

Nester

Capps

Randy

Kabiri

Nika

. 2003. “The Effects of Recent Welfare and Immigration Reform on Immigrants’ Access to Health Care.” International Migration Review 37(2):444–63.

20.

Harker

Kathryn

. 2001. “Immigrant Generation, Assimilation, and Adolescent Psychological Well-Being.” Social Forces 79(3):969–1004.

21.

Hurh

Won-Moo

Kim

Kwang-Chung

. 1990. “Adaptation Stages and Mental Health of Korean Male Immigrants in the United States.” The International Migration Review 24(3):456–79.

22.

Ingersoll-Dayton

Berit

Morgan

David

Antonucci

Toni

. 1997. “The Effects of Positive and Negative Social Exchanges on Aging Adults.” Journal of Gerontology, Series B: Psychological Sciences and Social Sciences 52B(4):S190–99.

23.

Kaestner

Robert

Pearson

Jay A.

Keene

Danya

Geronminus

Arline T.

2009. “Stress, Allostatic Load and Health of Mexican Immigrants.” Social Science & Medicine 90(5):1089–11.

24.

Kao

Wen H.

Tsai

Yung M.

1986. “Social Networking, Hardiness and Immigrant’s Mental Health.” Journal of Health and Social Behavior 27(2):133–49.

25.

Kennedy

Steven

Kidd

Michael P.

McDonald

James Ted

Biddle

Nicholas

. 2015. “The Healthy Immigrant Effect: Patterns and Evidence from Four Countries.” Journal of International Migration and Integration 16(2):317–32.

26.

Kessler

Ronald C.

Andrews

Gavin

Colpe

Lisa J.

Hiripi

Mroczek

Daniel K.

Normand

Sharon-Lise

Walters

Ellen E.

Zaslavsky

Alan M.

2002. “Short Screening Scales to Monitor Population Prevlances and Trends in Non-specific Psychological Distress.” Psychological Medicine 32:959–76.

27.

Keyes

Corey L.

2002. “The Mental Health Continuum: From Languishing to Flourishing in Life.” Journal of Health and Social Behavior 43(2):207–22.

28.

Keyes

Corey L.

2005. “Mental Illness and/or Mental Health? Investigating Axioms of the Complete State Model of Health.” Journal of Consulting and Clinical Psychology 73(3):539–48.

29.

Keyes

Corey L.

Wissing

Marie P.

Potieger

Johan P.

Temane

Miachel

Kruger

Annamarie

Rooy

S. van

. 2008. “Evaluation of the Mental Health Continuum-Short Form (MHC-SF) in Setswana-Speaking South Africans.” Clinical Psychology & Psychotherapy 15(3):181–92.

30.

Krause

Neal

Rook

Karen S.

2003. “Negative Interaction in Late Life: Issues in the Stability and Generalizability of Conflict across Relationships.” The Journals of Gerontology: Series B 58(5):88–99.

31.

Kwak

Kyunghwa

. 2016. “An Evaluation of the Healthy Immigrant Effect with Adolescents in Canada: Examinations of Gender and Length of Residence” Social Science & Medicine 157:87–95.

32.

Lavelle

Bridget

Smock

Pamela J.

2012. “Divorce and Women’s Risk of Health Insurance Loss.” Journal of Health and Social Behavior 53(4):413–31.

33.

Lin

Simeone

Ronald S.

Ensel

Walter M.

Kao

Wen

. 1979. “Social Support, Stressful Life Events, and Illness: A Model and an Empirical Test.” Journal of Health and Social Behavior 20(2):108–19.

34.

Chaohui

Edward

. 2019. “Healthy Immigrant Effect by Immigrant Category in Canada.” Health Reports 30(4):3–11.

35.

Markides

K. S.

Coreil

Jeannine

. 1986. “The Health of Hispanics in the Southwestern United States: An Epidemiologic Paradox.” Public Health Reports 101(3):253–65.

36.

Marsh

Lawrence C.

Cormier

David R

. 2002. Spline Regression Models. Thousand Oaks, CA: Sage Publications, Inc.

37.

Mausbach

Brent T.

Coon

David W.

Depp

Colin A.

Rabinowitz

Yaron G.

Wilson-Arias

Esther

Kraemer

Helena C.

Larry

Thompson

Lane

Geoffrey W.

Gallagher-Thompson

Dolores

. 2004. “Ethnicity and Time to Institutionalization of Dementia Patients: A Comparison of Latina and Caucasian Female Family Caregivers.” Journal of the American Geriatrics Society 52(7):1077–84.

38.

McDonald

James Ted

Kennedy

Steve

. 2004.“Insights into the ’Healthy Immigrant Effect’: Health Status and Health Service Use of Immigrants to Canada” Social Science & Medicine 59(8):1613–27.

39.

Menijivar

Cecilla

. 1997. “Immigrant Kinship Networks and the Impact of the Receiving Context: Salvadorans in San Francisco in the Early 1990s.” Social Problems 44(1):104–23.

40.

Migrant Integration Policy Index. 2014. “Migrant Integration Policy Index.” http://www.mipex.eu/

41.

Min

Pyong, G.

2001. “Changes in Korean Immigrants’ Gender Role and Social Status, and Their Marital Conflicts.” Sociological Forum 16(2):301–20.

42.

Mirowsky

John

Ross

Catherine E

. 2003. Social Causes of Psychological Distress. London and New York: Routledge.

43.

Montazer

Shirin

Wheaton

Blair

. 2017. “Economic Conditions in Countries of Origin and Trajectories in Distress after Migration to Canada.” Society and Mental Health 7(1):1–20.

44.

Newbold

Bruce K.

2005. “Self-Rated Health within the Canadian Immigrant Population: Risk and the Healthy Immigrant Effect.” Social Science & Medicine 60(6):1357–70.

45.

Newbold

Bruce K.

2009. “The Short-Term Health of Canada’s New Immigrant Arrivals: Evidence from LSIC.” Ethnicity & Health 14(3):315–36.

46.

Newsom

Jason T.

Rook

Karen S.

Nishishiba

Masami

Sorkin

Dara H.

Mahan

Tyrae L.

2005. “Understanding the Relative Importance of Positive and Negative Social Exchanges: Examining Specific Domains and Appraisals.” The Journals of Gerontology: Series B 60(6):304–12.

47.

Noh

Samuel

Avison

William R

. 1996. “Asian Immigrants and the Stress Process: A Study of Koreans in Canada.” Journal of Health and Social Behavior 37(2):192–206.

48.

Okun

Morris A.

Keith

Verna M.

1998. “Effects of Positive and Negative Social Exchanges with Various Sources on Depressive Symptoms in Younger and Older Adults.” The Journals of Gerontology: Series B 53(1):4–20.

49.

Payton

Andrew R.

2009. “Mental Health, Mental Illness, and Psychological Distress: Same Continuum or Distinct Phenomena?” Journal of Health and Social Behavior 50(2):213–27.

50.

Pearlin

Leonard, I.

Menaghan

Elizabeth G.

Lieberman

Morton A.

Mullan

Joseph T.

1981. “The Stress Process.” Journal of Health and Social Behavior 22(4):337–56.

51.

Portes

Alejandro

. 1998. “Social Capital: Its Origins and Applications in Modern Sociology.” Annual Review of Sociology 24:1–24.

52.

Portes

Alejandro

Rivas

Alejandro

. 2001. “The Adaptation of Migration Children.” The Future of Children 21(1):219–46.

53.

Richmond

Ted

Shields

John

. 2005. “NGO-Government Relations and Immigrant Services: Contradictions and Challenges.” Journal of International Migration and Integration 6(3–4):513–26.

54.

Rivera

Berta

Casal

Bruno

Currais

Luis

. 2016. “The Healthy Immigrant Effect on Mental Health: Determinants and Implications for Mental Health Policy in Spain.” Administration and Policy in Mental Health and Mental Health Services Research 43(4):616–27.

55.

Ryff

Carol D.

Keyes

Corey

. 1995. “The Structure of Psychological Well-Being Revisited.” Journal of Personality and Social Psychology 69:719–27.

56.

Schmidt

Ron

. 2007. “Comparing Federal Government Immigrant Settlement Policies in Canada and the United States.” American Review of Canadian Studies 37(1):103–22.

57.

Takeuchi

David T.

Zane

Nolan

Hong

Seunghye

Chae

David H.

Gong

Fang

Gee

Gilbert C.

Walton

Emily

Sue

Stanley

Alegria

Margarita

. 2007. “Immigration-Related Factors and Mental Disorders among Asian Americans.” American Journal of Public Health 97(1):84–90.

58.

Torres

Jaqueline M.

Wallace

Steven P.

2013. “Migration Circumstances, Psychological Distress, and Self-Rated Physical Health for Latino Immigrants in the United States.” American Journal of Public Health 103(9):1619–27.

59.

Tran

Thanh V.

Manalo

Victor

Nguyen

Victor T.

2007. “Nonlinear Relationship between Length of Residence and Depression in a Community-Based Sample of Vietnamese Americans.” International Journal of Social Psychiatry 53(1):85–94.

60.

Vankataramani

Atheendar S.

Shah

Sachin J.

O’Brien

Rourke

Kawachi

Ichiro

Tsai

Alexander C.

2017. “Health Consequences of the US Deferred Action for Childhood Arrivals (DACA) Immigration Programme: A Quasi-experimental Study.” The Lancet Public Health 2(4):175–81.

61.

Vega

William A.

Kolody

Bohdan

Valle

Ramon

Weir

Judy

. 1991. “Social Networks, Social Support, and Their Relationship to Depression among Immigrant Mexican Women.” Human Organization 50(2):154–62.

62.

Villanueva

Christina M.

Buriel

Raymond

. 2001. “Speaking on Behalf of Others: A Qualitative Study of the Perceptions and Feelings of Adolescent Latina Language Brokers.” Journal of Social Issues 66(1):197–210.

63.

World Health Organization. 2018. “Mental Health: Strengthening Our Response.” https://www.who.int/news-room/fact-sheets/detail/mental-health-strengthening-our-response

64.

Zheng

Schimmele

Christopher M.

2005. “The Healthy Migrant Effect on Depression: Variation over Time?” Canadian Studies in Population 32(2):271–95.

65.

Young

Marisa

Montazer

Shrin

. 2018. “Neighborhood Effects on Immigrants’ Experiences of Work-Family Conflict and Psychological Distress.” Society and Mental Health 8(1):25–49.

66.

Young

Marta Y.

2001. “Moderators of Stress in Salvadoran Refugees: The Role of Social and Personal Resources.” International Migration Review 35(3):840–69.

67.

Zhang

Wei

Hong

Seunghye

Takeuchi

David T.

Mossakowski

Kyrsia N.

2012. “Limited English Proficiency and Psychological Distress among Latinos and Asian Americans.” Social Science & Medicine 75(6):1006–14.