Article Text
PDF
Abstract
Background In Western settings, migration is associated with psychological well-being, but studies inevitably focus on culturally distinct ethnic minorities, making it difficult to distinguish migration from cultural assimilation. Many children in Hong Kong, a developed non-Western setting, have migrant parents with the same Chinese ethnicity. This study examined the association of migration with the child's psychological well-being in Hong Kong.
Methods Multivariable linear regression was used in Hong Kong's ‘Children of 1997’ Chinese birth cohort to examine the adjusted associations of migration (both parents Hong Kong born n=4285, both parents migrant n=1921, mother-only migrant n=462, father-only migrant n=1110) with a parent-reported Rutter score for child behaviour at ∼7 (n=6294, 80% follow-up) and ∼11 years (n=5598, 71% follow-up), self-reported Culture-Free Self-Esteem Inventory score at ∼11 years (n=6937, 88% follow-up) and self-reported Patient Health Questionnaire-9 (PHQ-9) depressive symptom score at ∼13 years (n=5797, 73% follow-up), adjusted for sex, highest parental education and occupation, household income, maternal and paternal age at birth, age of assessment and survey mode (PHQ-9 only).
Results Migration was unrelated to the overall self-esteem or depressive symptoms, but both parents migrant was associated with better behaviour (lower Rutter scores) at ∼7 years (β-coefficient (β) −1.07, 95% CI –1.48 to −0.66) and ∼11 years (−0.89, 95% CI −1.33 to −0.45).
Conclusions In a non-Western context, migration appeared to be protective for childhood behaviour.
- PUBLIC HEALTH
- MIGRATION
- PAEDIATRIC
Statistics from Altmetric.com
Introduction
Poor mental health is a major contributor to the global burden of disease. Mental health disorders are expected to account for 15% of the total disease burden by 2020.1 Behavioural and mental health problems can be identified at a young age, as early as 3–4 years for behavioural disorders2 ,3 and 7 years for mood and anxiety.3 Child and adolescent psychological well-being has been linked to ongoing or recurrent adult mental health disorders.4 ,5
With globalisation, migration is increasingly common and may affect the psychological well-being of adults and their children.6–8 Many studies have examined the psychological well-being of migrant children from minority ethnic groups in Europe, the USA and Canada.8 Extensive research has investigated the role of psychological well-being of first-generation child migrants (ie, non-native born) compared with second-generation migrant children (native born children born to at least one migrant parent), but few have compared second-generation migrant children with third-generation migrant children (native born and both parents native born). Most found little or no differences for behavioural problems,9–12 externalising13 ,14 and internalising,15 in second-generation compared with third-generation migrant children, whereas some found that second-generation migrant children had fewer behavioural problems,16 and others reported that second-generation migrant adolescents had more internalising problems than third-generation migrant adolescents.13 ,14 Most of these studies pertain to Western settings where migrants come from other countries. To the best of our knowledge, no previous study has examined the association of migration, focusing on second-generation and third-generation migrant children, within a homogeneous ethnic group, with child and adolescent psychological well-being. Uniquely, Hong Kong is a developed non-Western setting with a population largely formed by migration at different historic times from Southern China,17 enabling the assessment of the effects of migration within a culturally, linguistically and ethnically homogeneous population. We took advantage of this setting to examine the associations of migration status (second-generation and third-generation migrant children) with emotional and behavioural problems at ∼7 and ∼11 years, self-esteem at ∼11 years and depressive symptoms at ∼13 years using Hong Kong's Chinese population-representative birth cohort, ‘Children of 1997.’
Methods
Participants
The Hong Kong ‘Children of 1997’ birth cohort is a population-representative Chinese birth cohort (n=8327) that covered 88% of all births in Hong Kong from 1 April 1997 to 31 May 1997 and has been described in detail elsewhere.18 The study was initially established to investigate the effects of secondhand smoke exposure on infant health. Families were recruited at the first postnatal visit to 1 of the 49 Maternal and Child Health Centres (MCHCs) in Hong Kong, which parents of all newborns were encouraged to attend. There were no specific exclusion criteria. However, certain groups were not included in our ethnic Chinese Hong Kong study, such as ethnic minorities unable to answer a questionnaire in Chinese, non-Hong Kong residents who left Hong Kong after birth and possibly a few very sick infants unable to attend well-baby services at the MCHCs. Characteristics obtained using a self-administered questionnaire in Chinese at recruitment and subsequent routine visits included maternal and birth characteristics, parental education, parental migration and early life exposures, such as breastfeeding.
Passive follow-up via record linkage was instituted in 2005 to obtain routinely collected information including biannual assessments of emotional and behavioural problems using the Revised Parent's Rutter Scales in Chinese (grades 2, 4 and 6 (ie, ages 7–8, 9–10 and 11–12 years, respectively)) and self-esteem using Form A of the Culture-Free Self-Esteem Inventories for Children by Battle in Chinese (grade 4 (age 9–10 years) onwards) from the Student Health Service, Department of Health, which provides free annual check-ups for all school students. Active follow-up via direct contact was instituted in 2007, including postal surveys. Survey I, sent in July 2008, then re-sent a second and third time as necessary to non-respondents over the following 9 months, included questions on the mother's and father's birthplace. Survey II, including the Patient Health Questionnaire-9 (PHQ-9), was first sent in February 2010 and then re-sent a second and third time as necessary over the following 5 months. Non-responses were followed up via two waves of telephone interviews from November 2010 to April 2011 and from July 2011 to June 2012, and during pilot studies for in-person follow-up (June–August 2011 and June–August 2012). Survey III, sent as a postal survey from July 2011 to June 2011 and then followed up via telephone interviews from July 2011 to June 2012, included questions about the mother's and father's birthplace. Informed consent was obtained at recruitment from the next of kin, caretakers or guardians by completing the original questionnaire. Written consent was obtained from a parent or guardian and child for these postal surveys and the in-person follow-up, with verbal consent for the telephone survey. The study obtained ethical approval from the University of Hong Kong—Hospital Authority Hong Kong West Cluster Joint Institutional Review Board.
Exposures
Migration status was based on information about the mother's and father's birthplace from the baseline questionnaire (parental age, year of parental migration to Hong Kong and residency eligibility) supplemented by information from surveys I and III in answer to the question “Where was each of your child's parents born?” Migration status (Hong Kong born or not) for 7844 of the mothers and 7912 of the fathers was obtained. To provide granularity, migration status was classified as both parents Hong Kong born (corresponding to third-generation migrant or higher) and both parents migrant, mother-only migrant and father-only migrant (corresponding to second-generation migrant).
Outcomes
Emotional and behavioural problems
Emotional and behavioural problems at ∼7 years (6 to <9 years) and ∼11 years (9 to <13 years) were assessed from the Revised Parent's Rutter Scales in Chinese for parents.19 ,20 The scales consist of a set of 31 items describing emotional and behavioural difficulties, with each item scored 0 for does not apply, 1 for applies somewhat or 2 for certainly applies. A total score and subscores for conduct problems (5 items), emotional problems (5 items) and inattention/hyperactivity (3 items) were calculated, where a higher score indicated more emotional and behavioural problems. A total score of 13 or more was considered as overall emotional and behavioural problems.21 The presence of conduct problems, emotional problems or hyperactivity was defined as having the corresponding subscore above the sex-specific 97th centile (≥5 for boys and ≥4 for girls for conduct, ≥4 for boys and ≥4 for girls for emotion, 6 for boys and ≥5 for girls for hyperactivity) for our population.
Self-esteem
Self-esteem at ∼11 years (9 to <13 years) was assessed from Form A of the Culture-Free Self Esteem Inventories (SEI) in Chinese for children.22 Assessment inventories developed in Western countries may be less valid in non-Western settings. However, items of the SEI were chosen with concern for content that is least sensitive to change across cultures.23 The SEI has good validation in assessing self-esteem in adolescents and has been used in Hong Kong Chinese.24 Responses are of the forced-choice variety where the respondent checks either ‘yes’ or ‘no’ for each item. A total score and subscores for general self-esteem (perception of self-worth in general; 20 items), social self-esteem (perception of quality of relationships with peer; 10 items), academic self-esteem (perception of ability to academic success; 10 items), and parent-related self-esteem (perception of status at home; 10 items) were calculated, where a lower score indicated lower self-esteem. The presence of low self-esteem was defined as having a total score of 19 or less.22 Each subscale score was classified into one of the five categories (‘very low,’ ‘low,’ ‘intermediate,’ ‘high’ or ‘very high’). A ‘very low’ subscale score indicated low self-esteem in the respective subscale category (≤7 for general self-esteem and ≤2 for social, academic and parent-related self-esteem).22 We excluded unreliable scores with a lie scale of 2 or less (n=158), although a lie scale of 5 or more had been used to indicate a lack of defensiveness and reliable self-esteem inventories.22 ,25
Depressive symptoms
Depressive symptoms at ∼13 years (12 to ≤15 years) were assessed from the PHQ-9 for children in Chinese.26 The PHQ-9 scale has good sensitivity and specificity in detecting depression in youth27 and is an effective screening tool for the risk of depression among adolescents.28 The scales consist of a set of nine items describing symptoms and functional impairment, with each item scored 0 for not at all, 1 for several days, 2 for more than half the days or 3 for nearly everyday. A total PHQ-9 score was calculated, where a higher score indicated more depressive symptoms. The presence of PHQ-9 depression was defined as having a total score of 11 or more mapping on the Diagnostic Statistical Manual of Mental Disorders, fourth edition criteria.27
Statistical analysis
We used multivariable linear regression to estimate the adjusted association of migration status with Rutter, SEI and depressive symptom scores from which estimated β-coefficients with 95% CIs are presented. We used multivariable logistic regression to examine the adjusted associations of migration status with the presence of emotional and behavioural problems, low self-esteem and PHQ-9 depression from which ORs with 95% CIs are presented. The potential confounders included were sex, highest parental education at birth, highest parental occupation at birth, monthly household income per head at birth, maternal and paternal age at birth and age of assessment. Additionally, the survey mode for depression was included since PHQ-9 was obtained by postal questionnaire, telephone interview and in-person interview. We further assessed whether associations varied by sex from the heterogeneity across strata and the significance of interaction terms. As a sensitivity analysis, we also adjusted for maternal depression when the participant was ∼13 years old.
We used multiple imputation for missing exposures (among 7681 cohort members studied, migration status was imputed for 4.2%) and potential confounders (highest parental education for 2.1%, highest parental occupation for 13.9%, household income for 11.8%, maternal age at birth for 2.2%, paternal age of birth for 2.3%) as multiple imputation is the gold standard in this situation.29 We predicted missing values of exposures and confounders incorporating data on the primary outcomes (Rutter, SEI and PHQ-9 scores), exposures (migration status), confounders and other factors potentially associated with parental migration status. We summarised the results from 20 imputed data sets into single estimated β-coefficients with CIs adjusted for missing data uncertainty using the ‘Hmisc’ package in R 3.0.1 (R Development Care Team, Vienna, Austria). We also performed a complete case analysis for comparison using Stata V.10 (Stata Corp, College station, Texas, USA).
Results
Of the original 8327 cohort members, as of December 2013, 16 (0.2%) were known to have died, 26 (0.3%) had permanently withdrawn, 96 (1.2%) were known to have migrated and 275 (3.3%) were untraceable, probably migrated, whereas 7914 (95%) continued to live in Hong Kong. Of these, 6294 (80% follow-up) had parent-reported Rutter scores at ∼7 years and 5598 (71% follow-up) at ∼11 years, while 6937 (88%) had self-reported self-esteem scores at ∼11 years and 5797 (73%) had depressive symptom scores at ∼13 years.
The majority, at 55% (n=4285), had both parents Hong Kong born, while 25% (n=1921) had both parents migrant, 6% (n=462) mother-only migrant and 14% (n=1110) father-only migrant. Table 1 shows that both parents Hong Kong born had higher education, occupation and monthly household income than migrant parents.
Baseline characteristics by migration status from H K's ‘Children of 1997 Birth Cohort’ (available case analysis)
Table 2 shows that both parents migrant was associated with lower Rutter scores at ∼7 and ∼11 years than both parents Hong Kong born, adjusted for sex, highest parental education, highest parental occupation, monthly household income, maternal and paternal age at birth and age at assessment. Children with one migrant parent (mother only or father only) had similar scores to children with both parents Hong Kong born. Both parents migrant was associated with lower Rutter scores consistently across the Rutter subscales. Parental migrant status was not associated with overall self-esteem. Both parents migrant was associated with higher general and academic self-esteem scores compared with both parents Hong Kong born. Again, children with one migrant parent had generally similar self-esteem scores to children with both parents Hong Kong born, although children with mother-only migrant had lower parent-related self-esteem scores. Parental migrant status was not associated with PHQ-9, and this association was unchanged in a sensitivity analysis adjusting for maternal depression at ∼13 years. The associations of migration status with the mental health outcomes did not vary by sex.
Adjusted* association of migration status with Mental Health Scores in H K's ‘Children of 1997’ birth cohort (after multiple imputation)
Table 3 shows that both parents migrant, as compared with both parents Hong Kong born, was associated with the absence of behavioural problems, including conduct and emotional problems, at ∼7 and ∼11 years. Both parents migrant was also associated with the absence of hyperactivity at ∼7 years. Father-only migrant was also associated with the absence of overall and emotional problems at ∼7 years. Both parents migrant was associated with the presence of higher social and academic self-esteem at ∼11 years and a lower risk of depression at ∼13 years (unadjusted and adjusted for maternal depression). Associations of migration status with presence of mental health problems did not vary by sex. Overall, the available case analysis produced similar results (online supplementary tables S1 and S2).
Adjusted*association of migration status with the Presence of Behavioural Problems, Low Self-Esteem and Depression in H K's ‘Children of 1997’ birth cohort (after multiple imputation)
Discussion
Our study from an understudied non-Western developed population focuses on migration within a homogeneous ethnic group and considers migration status in relation to several dimensions of the key public health issue of child and adolescent psychological well-being. Consistent with some previous studies,16 second-generation migration status compared with third-generation migration status was associated with better behaviour. We additionally found that second-generation migration status was associated with some aspects of higher self-esteem and with lower risk of depression.
Although this birth cohort study has strengths, including a large population-representative sample, high follow-up and a range of measures of psychological well-being, it has limitations. First, information on migration status and some potential confounders was not complete. We used multiple imputation to use all available data while preserving ‘uncertainty from missing data, minimising inclusion bias and increasing statistical power’.29 An available case analysis produced similar results. Second, we have no information on parent dialect and language skills. However, most migrants were from Guangdong province where Cantonese, the dialect spoken in Hong Kong, is common. Additionally, a US study found that heritage (foreign) language proficiency, as opposed to English proficiency, of Chinese American youth was not associated with depressive symptoms in second-generation children.30 Third, childhood characteristics, such as temperament31 and academic achievement,32 as well as parental factors including divorce,33 caretaking style and rearing practices,34 ,35 will affect child psychological well-being. However, these are unlikely to cause parental migration and hence are not potential confounders but possibly mediators. Fourth, our study utilises screening tools, validated in Chinese, for measurement of psychological well-being instead of clinical diagnosis.19 ,36 ,37 While clinical diagnoses are useful for identifying individuals requiring treatment, screening tools identify differences across a range, which is more relevant to population health. Fifth, although the proportion of adolescents with migrant parents is quite low in some groups, our large sample size means that even for the smallest group (mother-only migrant), we have 80% power with 5% α to detect a fairly small effect size of 0.14 standard deviations. Sixth, as current socioeconomic position (SEP) may be an important indicator of child psychological well-being, we only adjusted for SEP at birth. A key attribute of SEP, such as parental education, is unlikely to change, but parental occupation and income could have changed substantially. However, education is a strong predictor of later income and occupation.38 Seventh, we did not distinguish between third-generation and higher generation migrant children because the Hong Kong population was formed by mass migration in the mid-20th century,17 so most adolescents in Hong Kong have some grandparents born in the neighbouring province of Guangdong. Lastly, we lacked information on maternal postpartum depression, which is associated with parental migrant status39 and adolescent depression40 ,41 and other mental health diagnoses of parents and children. However, only diagnoses affecting decisions about parental migration would be confounders. Nevertheless, we cannot rule out the possibility that these parental migrants were self-selected mentally healthy migrants, although some of them may have migrated as children before any such problems would have emerged.
In the past, in Western populations, it was theorised that first-generation immigrants struggle, the second-generation did better, and the third-generation even better in terms of personal health and achievement.41 The overall pattern has shifted to an ‘immigrant paradox’, where first-generation migrants outperform their second and third-generation children.41 ,42 Others contend that biculturalism is the most adaptive approach to acculturation,43 although this is less relevant in our culturally homogeneous setting.
There are several explanations for our findings. First, parents of second-generation migrant children may have different perspectives on psychological well-being compared with Hong Kong born parents of third-generation migrant children, who might have more westernised views on health and well-being. Problematic behaviour and mental health symptoms depend on the informant8 when culture and society influence mental health and its reporting.44 Second, parental self-selection, where migrants choose to leave a community to have better lives,41 and the healthy migrant effect, where migrants are those with the best physical and mental health,8 ,10 ,45 could play a role in their offspring's psychological well-being. Third, cultures that are coherent and supportive, such as traditional Chinese culture, may protect children with migrant parents from developing mental health problems.8 Fourth, children with both parents migrant may have a different frame of reference from other groups since they are most likely brought up by parents with the most similar values and customs. Finally, it is possible that our findings are temporary differences that will level off later in life10 due to assimilation and adaptation to the environment over time by the family.9 ,10 ,46
Conclusion
Our study contributes evidence suggesting that parental migration within an ethnically homogeneous population has no negative impact on childhood behaviour, self-esteem or depressive symptoms for second-generation migrants compared with third-generation migrants, but might even be protective. Future research is necessary to understand how these differences develop during assimilation for migrant families so that targeted interventions can be created to prevent future mental health issues.
What is already known on this subject
-
Migration studies involving child psychological well-being in Western settings focus on culturally distinct ethnic minorities, making it difficult to distinguish migration from cultural assimilation.
What this study adds
-
In a developed non-Western setting where many children have migrant parents with the same Chinese ethnicity, our study provides new evidence that migration is associated with better child behaviour.
Acknowledgments
The authors thank their colleagues at the Student Health Service and Family Health Service of the Department of Health for their assistance and collaboration. They also thank the late Dr Connie O for coordinating the project and all the fieldwork for the initial study in 1997–1998. CYL also thanks the University Grants Committee for funding her research through the Hong Kong PhD Fellowship Scheme.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Footnotes
-
Contributors CYL performed the literature review, conducted the data analysis, interpreted the findings and drafted the manuscript. CMS conceptualised the ideas, designed and directed the analytic strategy, interpreted the findings, revised drafts of the manuscript and supervised the study from conception to completion. GML and CMS resurrected the birth cohort. All authors have read and approved the submission of the manuscript.
-
Funding This work is a substudy of the ‘Children of 1997’ birth cohort funded by the Health and Health Services Research Fund (HHSRF Grants # 07080751), Government of the Hong Kong SAR. The study was initially supported by the Health Care and Promotion Fund, Health and Welfare Bureau, Government of the Hong Kong SAR (HCPF Grant #216106). Since 2005, the ‘Children of 1997’ birth cohort has been funded by the Health and Health Services Research Fund (HHSRF Grant #03040771), Government of the Hong Kong SAR, the Research Fund for the Control of Infectious Diseases (RFCID Grants #04050172), Government of the Hong Kong SAR, and the University Research Committee Strategic Research Theme (SRT) of Public Health, The University of Hong Kong.
-
Competing interests None.
-
Ethics approval Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster.
-
Provenance and peer review Not commissioned; externally peer reviewed.