Introduction

Canada is the fastest growing G8 nation in the world (Statistics Canada 2012b). This population growth is attributed almost exclusively to immigration (Statistics Canada 2012b). In 2006, 1 in 5 Canadians was foreign-born (19.8 % of the population) (Human Resources and Skills Development Canada 2007), and Canada is home to over 200 different ethnic origins (HRSDC 2007). This diverse immigrant population is projected to continue to increase; by 2031 it is expected that immigrants will account for 25–28 % of the population (Statistics Canada 2012a).

Immigration impacts upon mental health. However, there are mixed reports on how migration affects mental health. Some reports have found evidence for a “healthy immigrant effect” for mental health where the foreign-born population enjoys better levels of mental health than the Canadian-born population only to deteriorate and converge to national-born levels shortly after arrival (Ali 2002; Bergeron et al. 2009; Lou and Beaujot 2005; Ng and Omariba 2010). Others studies have found that immigrants maintain their mental health advantage (Aglipay et al. 2013), while others report that immigrants have comparable prevalence rates of anxiety and depression symptoms to the Canadian-born populations (Centre for Research on Inner City Health 2012). Taking these mixed results into consideration, Canada’s Mental Health Strategy emphasizes the importance of further research into the mental health needs of diverse populations to aid the development and delivery of mental health services that are inclusive of Canada’s growing immigrant populations (Mental Health Commission of Canada 2012).

Immigrants face many stressors settling in a new land. Difficulties with economic integration (unemployment, low income, underemployment, and loss of status) all contribute to post-migration stress (Aycan and Berry 1996; McKenzie 2009). Immigrants also face diminished social networks, low education and literacy, and language barriers (McKenzie 2009). Discrimination, threats to ethnic identity, and homesickness (Jibeen and Khalid 2010) all pose challenges to immigrant mental health. Immigrant youth also face added difficulties such as combatting intergenerational conflict (Islam 2012; Samuel 2009), renegotiating identity in a new land, and poverty (Beiser et al. 2002, 2012; Khanlou and Crawford 2006; Lim et al. 2008; Mahoney 2002; Yeh 2003). Parental social status is a critical determinant of distress in adulthood (Schoon et al. 2003). Whether employed or not, racialized and immigrant populations are at almost a three times greater risk of living in poverty with one in four immigrant families in Toronto living in poverty (United Way of Greater Toronto 2004).

Age at time of immigration has been found to be an important factor of migration related to mental health (Patterson et al. 2013; Rumbaut 2004; Wu and Schimmele 2005). However, the literature reports mixed findings on the impact of age at time of immigration on mental health. Rumbaut (2004) examined age at time of immigration in the US and categorized them into the following migration generations: first generation immigrant (immigration during adulthood 18+ years old), 1.25 generation (immigration during adolescence 13–17 years old), 1.5 generation (immigration during middle childhood 6–12 years old) and 1.75 generation (immigration during early childhood 0–5 years). The 1.25 generation was found to have a difficult time adapting to the new country, experiencing lower college graduate rates, while first generation immigrants were the least likely to be English-language proficient. The 1.5 generation had the fewest criminal justice experiences, lowest level of linguistic assimilation, and was most likely to identify with their national origin.

Canadian literature has also examined the impact of age at time of immigration on social, economic, and mental health outcomes. Statistics Canada carried out a study examining childhood immigrants who arrived in the 1960’s, 1970’s, and 1980’s and found that across these three groups those in the 1.5 generation (defined as those who had immigrated at the age of 12 or younger) were more likely than their Canadian-born counterparts to have received a university degree; despite this, men (not women) earned less than their Canadian-born counterparts (Bonikowska and Hou 2011). Wu and Schimmele (2005) examined immigrants in Canada and categorized age at time of immigration into those who migrated when they were 17 or younger and those 18 and older (which correlates to Rumbaut’s “first generation” group). It was found that those who immigrated as adults experienced fewer depressive symptoms and had a lower risk of major depressive episode. Xu and McDonald (2010) analyzed Canadian Community Health Survey (CCHS) 2001–2005 data and found that men who immigrated to Canada before the age of 12 were at a mental health disadvantage. Similarly, Patterson et al. (2013) found the highest prevalence rates and risk of mood disorders, anxiety disorders and substance abuse amongst those who had immigrated to Canada before the age of 6 even after adjusting for age, sex, region of origin, marital status, urbanicity, household income, and household size. Patterson et al. (2013) grouped those who migrated during middle childhood (6–12 years old) and adolescence (13–17 years old) together and used data from the Canadian Community Health Survey (CCHS) 2002.

Given the inconsistency in the literature, lack of prevalence rates of mental health outcomes for migration generations, and the need for updated statistics, this study seeks to elucidate the impact of age at time of immigration upon mental health in Canada across different categories of age at time of immigration groups using data from the Canadian Community Health Survey (CCHS) 2011. This study asked the following research question: does migration to Canada during early childhood, middle childhood, adolescence or adulthood pose the greatest risk for mental health? Based on US research (Rumbaut, 2004), it could be hypothesized that those who immigrated between the ages of 6–12 years old (middle childhood; 1.5 generation) would be at a mental health advantage. On the other hand, an examination of Canadian research offers the hypothesis that immigrating during childhood may pose some vulnerability. Rumbaut (2004) provides the most comprehensive and widely used breakdown of age at time of immigration. To establish consistency and better pinpoint which group may be vulnerable in terms of mental health, Rumbaut’s (2004) four categories were selected for this analysis. The CCHS 2011 was used to calculate the prevalence rates of mood disorders (mental health outcome measure) across the four different age at time of immigration groups. Secondly, adjusted odds ratios (OR’s) of mood disorders were calculated for each migration generation group to determine which group experienced the highest risk of mood disorders even after adjusting for important contextual factors.

Methods

Data Source: Canadian Community Health Survey (CCHS)

Statistics Canada conducts the Canadian Community Health Survey (CCHS) every year to collect health-related data from January to December of each year, surveying individuals over the age of 12 in all provinces and territories, excluding those residing on Indian Reserves, institutions, remote regions, and full-time members of the Canadian Forces. In order to administer surveys to a variety of households, the CCHS samples households by area framing, telephone list framing, and random digit dialing. Response to the survey is voluntary with a response rate of about 72 % (Statistics Canada 2011a).

Sample Population

Estimated prevalence rates for this study were calculated using the CCHS 2011. Thomas and Wannell (2009) recommend that data from a single year be used for the analysis of descriptive statistics, while pooled data can be utilized for higher order modeling. The immigrant population was selected using the derived variable “Immigration Flag” which divided respondents into those who were immigrants and those who were not. The derived variable “Age at Time of Immigration” was then used to stratify the immigrant population by years since immigration into the following four age at time of immigration categories: 0–5 years old (early childhood, 1.75 generation), 6–12 years old (middle childhood, 1.5 generation), 13–17 years old (adolescence, 1.25 generation), and 18+ years old (adulthood, first generation immigrant) (Rumbaut 2004). In order to increase sample size and power, CCHS was merged across 5 yearly cycles (2007–2011) for the multivariable logistic regression analysis. The immigrant population was selected from the pooled data using the “Immigration Flag” variable.

Mental Health Outcome Variable

The presence of mood disorders (depression, bipolar disorder, mania or dysthymia) was assessed based on the survey question where participants were asked if they had such disorders diagnosed by a health professional (yes, no). The four additional mental health outcome measures using the anxiety disorders, self-perceived mental health, self-reported life stress, and positive mental health variables were also examined but could not be utilized for high-level statistical analysis due to low sample size of these rare outcomes in the CCHS 2011.

Covariate Variables

Sociodemographic, acculturative, socioeconomic status, and health and behavior variables were examined as key contextual factors related to mood disorder risk. Age (recategorized into ≤24 years old, 25–44 years old, 45–64 years old, and 65+ years old), gender (male, female), racialized status (racialized, non-racialized/white), sense of belonging to community (somewhat weak–very weak and somewhat strong–very strong), working status last week (recategorized into yes, no (did not have a job, permanently unable to work), and not applicable (those not of working age), self-rated health status (poor–fair, good–excellent), leisure time physical activity index (moderately active–active and inactive), and current smoking status (yes, daily/occasional smoker and no, former/never smoker).

Data Analysis

The estimated prevalence rates (percentages) of mood disorders were calculated for immigrants in Canada using the CCHS 2011 across the four categories of age at time of immigration. To determine the risk of mood disorders, multivariable logistic regression modeling was carried out using data from CCHS 2007–2011 on the dichotomous outcome of presence of a mood disorder (yes, no) with age at time of immigration group as the main independent variable of interest. The forced entry: enter model was used to include all the covariate variables outlined above. Significance was set at p < 0.05. Missing cases were excluded from analysis. Prevalence rates, odds ratios (OR’s) and 95 % confidence intervals (CI’s) were reported and population sample weights provided by the CCHS were applied to all results. All data analysis was carried out using SPSS and Stata12 statistical software at the Statistics Canada Research Data Centre at York University.

Results

Sample Profile

In the CCHS 2011, 22.17 % (weighted percentage) of the Canadian population surveyed identified themselves as immigrants. The sample characteristics of immigrant populations (unweighted n = 8,839) in the CCHS 2011 are displayed in Table 1. Of the immigrant population surveyed in the CCHS 2011, 1,094 (12.38 %) immigrated during early childhood, 1,129 (12.77 %) during middle childhood, 759 (8.59 %) during adolescence, and 5,857 (66.29 %) during adulthood (unweighted sample sizes). The weighted samples sizes and frequencies are presented in Table 1. The four age at time of immigration groups were significantly different in distribution of age, gender, racialized status, working status, self-rated health, physical activity level, and current smoking status. It was only the factor of sense of belonging to community which the migration generations did not differ across.

Table 1 Weighted sample demographics for age at time of immigration cohorts in Canada 12+ years old, CCHS 2011 (unweighted n = 8,839)
Full size table

Prevalence Rates of Mood Disorders

The prevalence rates of mood disorders across the four migration generations are displayed in Table 2. Those who immigrated during early childhood (0–5 years old; 1.75 generation) had significantly higher prevalence rates of mood disorders (6.83 %; 95 % CI 6.77–6.89) than all other age at time of migration groups. Although significant differences were found between the other three groups, in terms of clinical significance, those who migrated during middle childhood (4.88 %; 95 % CI 4.83–4.93), adolescence (4.84 %; 4.79–4.90) and adulthood (4.58 %; 95 % CI 4.56–4.60) had comparable prevalence rates of mood disorders.

Table 2 Weighteda prevalence rates of mood disorders for age at time of immigration cohorts in Canada 12+ years old, CCHS 2011 (unweighted n = 8,839)
Full size table

Risk of Mood Disorders

Unadjusted and adjusted odds ratios were calculated to determine mood disorder risk across the four migration generations using CCHS 2007–2011 data (unweighted n = 42,228). Unadjusted odds ratios of mood disorders were calculated and it was found that those who immigrated during early childhood had a significantly higher risk of mood disorders compared to those who migrated during adulthood (OR 1.41, 95 % CI 1.11–1.78). Those who immigrated during middle childhood and adolescence had a comparable risk of mood disorders to those who immigrated during adulthood. After adjusting for age, gender, racialized status, sense of belonging to the community, working status last week, self-rated health, physical activity level, and smoking status, the same trend emerged with those who migrated during early childhood having a significantly higher risk of mood disorders (OR 1.40, 95 % CI 1.04–1.88) compared to those who migrated during adulthood. Immigrating during middle childhood and adolescence was not associated with elevated risk of mood disorders. The weighted unadjusted and adjusted odds ratios are presented in Table 3.

Table 3 Weighteda odds ratios of mood disorders for immigrants in Canada 12+ years old for age at time of immigration cohorts, CCHS 2007–2011
Full size table

In addition to migration during early childhood, multivariable regression analysis also revealed the following factors associated with mood disorders for immigrant populations in Canada: female gender (OR 1.94, 95 % CI 1.64–2.29) (reference group: male), age between 25 and 44 years old (OR 1.89, 95 % CI 1.45–2.46), age between 45 and 64 years old (OR 1.89, 95 % CI 1.49–2.38) (reference group: 65+ years), racialized status (OR 0.51, 95 % CI 0.42–0.61) (reference group: non-racialized/white), weak sense of belonging to the community (OR 1.42 (95 % CI 1.20–1.67) (reference group: strong sense of belonging), not being employed (OR 1.60, 95 % CI 1.32–1.95) (reference group: currently employed), inactive physical activity status (OR 5.69, 95 % CI 4.71–6.88), being a current smoker (OR 1.75, 95 % CI 1.42–2.16) (reference group: not a current smoker), and fair–poor self-rated health (OR 5.69, 95 % CI 4.71–6.88) (reference group: good–excellent self-rated health).

Discussion

Those who immigrated during early childhood (0–5 years old, 1.75 generation) were found to have the highest prevalence rates and risk of mood disorders compared to their other migration generation group counterparts. This study suggests that those who immigrate during early childhood have a similar prevalence rate of mood disorders to the general Canadian population (6.3 %) (Statistics Canada 2011b). This study corroborates Patterson et al.’s (2013) findings of an increased mood disorder risk for the 1.75 generation compared to first generation immigrants based on CCHS 2002 data. However, Patterson et al. (2013) found more than double the risk for mood disorders for the early childhood migration generation group (OR 3.41, 95 % CI 1.7–7.0) compared to the mood disorder risk found using CCHS 2007–2011 data (OR 1.40, 95 % CI 1.04–1.88). There may be a couple of explanations. This may be a true difference and mood disorder risk for the 1.75 generation may have decreased from 2002 to 2007–2011. The newer generation of immigrants may experience less difficulty during the migration and resettlement process as the Canadian mental health service system has attempted to adapt and change to meet immigrants’ needs. As more immigrants settle in Canada, the stressors of coming to a new land may be mitigated by the presence of immigrant-dense neighborhoods and cultural services. On the other hand, it may be due to other factors, for example sampling error in the CCHS survey (more immigrants with mood disorders may have been surveyed in 2002 compared to 2011 just due to chance). It may also be due to statistical analytical differences, perhaps related to pooling of data across cycles, the application of weights, method of analysis, or covariates included in the model. A time trends analysis of mood disorder risk by migration generation during this time period would help elucidate if there has indeed been a drop in risk.

Is it actually childhood migration that increases the risk of mood disorders or is this increase related to acculturating to Canadian conceptions of mental health and increased ease in navigating the health system that leads to a greater probability of being diagnosed? A parallel can be drawn to coronary heart disease. Immigrants arrive in Canada with lower levels of heart disease only to markedly increase in prevalence rates and risk within about 10 years of migration (Health Canada 2010; Hyman 2007). There is not much difference in the conceptions and diagnosis of this condition in the host country, rather, the change seems to be linked to lifestyle changes upon arrival in Canada, such as increase in sedentary behavior and unhealthy change in diet (Koya and Egede 2007; Yang et al. 2007). Similarly, the changes in mental health risk may be attributed to factors experienced post-migration, but trying to discern what these lifestyle changes immigrants undergo upon arrival that are leading to mental health issues is not as clear cut. McDonald and Kennedy (2004) found that the decline in physical health experienced by immigrants cannot be attributed to better screening and detection of pre-existing health conditions in the host country. Factors such as parents’ loss of social status and precarious employment could have long-standing ramifications for children and youth. Schoon et al. (2003) found that children of parents who experienced loss of social status experienced greater levels of distress in adulthood. Moreover, they found that social risk events cumulate throughout life, influencing behavioral and psychosocial adjustment in both childhood and adulthood. Migration to a new land may lead to changes in parenting style (Ochocka and Janzen 2008). Differences in parent and child attitudes toward issues such as dating, academic success, discipline, and religion and culture may become sources of tension in the household after migration (Ochocka and Janzen 2008). These post-migration factors could also be related to experiences directly felt by child and youth immigrants, for example, rapid adoption of Canadian cultural conceptions of mental health and illness or attempts at group inclusion with a new peer group. Further examination is needed to understand which factors impact those who migrate in childhood differently from those who migrate as adults. Can anything be done to mitigate this downward spiral or is this deterioration of mental health an inevitability for childhood immigrants? Moreover, what does this say about life in Canada, where people often migrate to in hopes for a better life and access to universal health care? Put another way, is living in Canada making people unhealthy?

While this study suggests that migration impacts child and youth immigrants negatively, it seems that migration in adulthood on the other hand, may confer some mental health advantage. There has been much research to support that migration fosters resilience. The migration trajectory can be viewed as having three phases: pre-migration, migration, and post-migration resettlement (Kirmayer et al. 2011). Pre-migration factors such as ties to the origin culture may play a role. The migration process itself may make a person stronger. In the case of voluntary immigration, migration can be a process of empowerment, where an individual takes their destiny into their own hands and moves in hopes for a better future (Hull 1979). Or it may be factors related to the post-migration resettlement process where connection to the origin culture may help immigrants resist fully adopting the host nation’s lifestyle and prevent convergence to Canadian-born levels of lower mental health. Stafford et al. (2011) found that visible minority immigrants who lived in immigrant-dense neighborhoods had a lower risk of depression. Living in an immigrant-dense neighborhood could potentially allow an immigrant to maintain cultural ties and create social networks with those who can empathize with the migration experience. Further studies are needed to examine if/what aspects of Canadian-born culture or lifestyle that are related to lower levels of mental health. Indeed rapid integration into the host nation is encouraged because it mitigates the immediate stressors or resettlement (Boyd 1989); however, there may be long term mental health consequences to such rapid adoption of the host nation’s lifestyle. “Integration” in Canada has been critiqued as being an expectation of conformity to the dominant culture’s norms rather than being respectful of maintaining cultural differences (Li 2003). Aspects of the origin culture may prove to be protective for mental health. What these aspects might be and if they can be retained in the host country need to be explored. Moreover, additional immigrant host countries such as the United States, Australia, and Europe need to be investigated for similar trends.

Comparable mood disorder prevalence rates were found for the 1.75 migration generation and the general Canadian population. Those who immigrate during childhood are often regarded as indistinguishable from the second generation (Canadian-born, in this case) population (Rumbaut 2004), and immigrating early in life (0–5 years old) can be akin to being born in the host nation. Young children integrate much more easily into the host nation’s culture and lifestyle and do not retain memory or ties to their origin culture as readily. This parallels findings of the “duration/years since immigration effect” for mental health, where immigrants initially arrive in Canada with better levels of mental health only to deteriorate to Canadian-born levels the longer they reside in Canada (Kirmayer et al. 2011). Immigrating early in life and residing in the host nation for many years can have the same effect of making the individual lose connection to their origin culture and way of life. Lifestyle and traits of the origin culture may afford some protective mental health benefit for newcomers, but for those who immigrate during early childhood, the absence of memory and connection to the origin culture may lead to rapid adoption of the host nation’s lifestyle and explain the similar prevalence rates of mood disorders found for those who immigrated during early childhood compared to the Canadian population. Further studies are needed to compare the mental health of second generation immigrants to those of the 1.75 generation. Unfortunately, the CCHS does not have information on second generation immigrant status (are/were your parents first generation immigrants to Canada?). The addition of this question to the survey would help facilitate further investigation. Longitudinal studies that can follow the mental health of newcomers who arrive in Canada in the same year from the first, 1.25, 1.5 and 1.75 migration generations and make comparisons to the second-generation immigrant population would be illuminating. Past data cycles of the now inactive Longitudinal Survey of Immigrants to Canada (LSIC) could be analyzed for further elucidation.

Strengths and Limitations

The CCHS offers a nationally representative database with a large sample size, allowing modeling such as multivariable logistic regression analysis. This study helps to clear the confusion in the Canadian literature over which age at time of immigration group is most at risk in terms of mental health in Canada and provides updated statistics. To our knowledge, this is the only study to separately examine the mental health of the four age at time of immigration groups in Canada. The analysis of both prevalence rates and odds ratios led to convergence of findings. Triangulation of results from both the prevalence rate and logistic regression analysis confirm the finding that immigrating to Canada during early childhood poses the greatest risk to mental health.

This study has a number of limitations as well. As this was secondary data analysis of cross-sectional data, causal inferences cannot be made about the relationship between age at time of immigration and mood disorder risk. Due to low sample sizes and the relative rarity of negative mental health outcomes, other mental health measures beyond diagnosed mood disorders could not be assessed in this study. Moreover, the presence of mood disorders was a self-reported variable. Self-reported clinical diagnoses of mood disorders presents the problem where it is difficult to determine if those who migrated in early childhood actually have a higher prevalence rate of mood disorders or if it is the case that this group is actually utilizing mental health services more than other groups and as a result more likely to be diagnosed with mood disorders. Psychological scales would be a better measure of mental health outcomes in the CCHS. These scales are available in the CCHS; however, they are optional content in the survey and not carried out across all provinces and territories in Canada. Future CCHS cycles could potentially include psychological scales in the core content to allow for Canada-wide analysis of mental health outcomes. However, studies that have used psychological scales (e.g. measured psychological distress using the General Health Questionnaire) have reported higher levels of negative mental health outcomes for immigrants compared to non-immigrants (Lai and Surood 2008; Mirsky et al. 2008; Miszkurka et al. 2010). In addition, the CCHS does not distinguish between immigrant class/type or circumstance of migration. Refugees, non-status individuals, economic immigrants, and family class immigrants are all grouped together. The addition of a CCHS question that distinguishes between immigrant types would help further research in this field.

Implications

Canada’s Mental Health Strategy recommends that research on immigrant populations is crucial in order to develop a mental health system that is responsive to the needs of an increasingly diverse population (Mental Health Commission of Canada 2012). This study found that those who migrate to Canada during early childhood may be at a greater vulnerability compared to those who migrate at other ages. The age of onset of many mental health issues present within childhood and adolescence (Kessler et al. 2005). Early intervention programs need to address the unique experiences of childhood immigrants (e.g. culture clash with parents, getting bullied at school). Programs to mitigate the stress upon arrival and family conflict associated with migration can help bring families together and reinforce the familial social support network. Anti-stigma and educational campaigns that can help immigrant families access pathways to care are vital. Moreover, community programming could offer support for immigrant parents through social networking events and parenting and family education programs. The Triple P-Positive Parenting Program has been shown to be efficacious across the globe and is recommended by the Public Health Agency of Canada as an evidence-based best practice intervention (Public Health Agency of Canada 2013). Early childhood intervention in the form of parent education and family support programs coupled with early childhood education lead to better mental health outcomes (Karoly 2010; Mental Health Commission of Canada 2013). Moreover, cost-benefit and return on investment analysis make the case the cost-effectiveness of such programs (Mental Health Commission of Canada 2013). Dedication to the health and well-being of Canada’s immigrants needs to be demonstrated at the government level. Recent immigrants and racialized populations face higher levels of poverty and precarious employment (Access Alliance 2013; Citizens for Public Justice 2012). In addition to increasing funding for mental health programming, the government needs to develop effective anti-poverty and economic security strategies in consultation with newcomer and immigration populations and support better job-matching and skills bridging programs, raise minimum wage, and decrease precarious employment to mitigate the loss of social status experienced by new immigrants. Ontario, British Columbia, and Quebec are home to the largest immigrant populations in Canada (Statistics Canada 2011b). However, the provincial governments of Ontario, British Columbia, Quebec, and Manitoba impose a three-month wait period on permanent residents before they are eligible for provincial health insurance coverage. This provincial government policy needs to be reconsidered in order to move forward and ensure the health of Canada’s immigrants.

Conclusions

Age at time of immigration is an important social determinant of mental health that warrants further investigation. This study finds that those who immigrate to Canada before the age of six are at a significantly greater risk of mood disorders compared to those who migrate as adults. The differences in post-migration stress, integration, and acculturation patterns of those who migrate in childhood compared to those who migrate later in life need to be explored to further understand the impact of early childhood migration on mental health. Early intervention mental health programming needs to be developed that is responsive to the needs of those who have undergone childhood migration.