An investigation of the healthy migrant hypothesis: pre-emigration characteristics of those in the British 1946 birth cohort study.
Fuller-Thomson, Esme ; Brennenstuhl, Sarah ; Cooper, Rachel 等
Increasingly, studies on large immigrant-recipient countries, such
as Canada, (1,2) the US (3-5) and Australia, (6) reveal that immigrants
have better morbidity and mortality outcomes than non-immigrants, once
adjustments have been made for socio-economic position (SEP).
Researchers propose that this may be due to the "healthy migrant
effect": individuals who are in better health are more likely to
choose to migrate (i.e., self-selection) and/or are given preference by
receiving countries (i.e., external selection through health
examinations of prospective immigrants). That is, factors contributing
to the healthy immigrant effect may begin prior to migration itself. (7)
This explanation is supported by research showing that immigrants have
better health in their host countries than non-migrants in their
countries of origin. (5) It is also consistent with evidence that
positive health selection in migrant populations begins as far back as
childhood. For example, several studies have shown that Mexican
immigrants to the US are taller than their American-born counterparts
(8) and their same-age counterparts who did not migrate or who migrated
and then returned home. (9)
Often studied in the context of individuals migrating from lower-
to higher-resource countries, the healthy migrant effect in those
relocating from one high-resource country to another has been
investigated in fewer studies (although there are exceptions (4,6,10)).
Moreover, substantiating evidence of positive health selection, such as
comparisons of pre-emigration health profiles with those of individuals
remaining in the sending country, is lacking, especially evidence based
on prospective data from a country of origin. Studying the movement of
individuals between high-resource countries is important for a better
understanding of how health is distributed within and across nations. In
the current study, we use the Medical Research Council National Survey
of Health and Development (NSHD), the oldest British birth cohort study,
started in 1946, to investigate the healthy migrant effect among those
leaving Britain by comparing their childhood health with that of their
contemporaries who stayed. The UK's long history of emigration
makes it a good country to study health and migratory selection
processes. Moreover, because many of its post-war emigrants would have
migrated to other Commonwealth countries, (11) this cohort provides a
good case study of the relationship between health and migration among
higher-income countries. Canada has historically received large numbers
of British emigrants. According to 2011 census data, approximately 5% of
adults aged 65 and older living in Canada were born in the UK. (12)
A number of indicators of childhood health are associated with
adult health outcomes, especially cardiovascular and metabolic health
and, thus, may be associated with successful migration. For example, low
SEP and poor material living conditions in childhood have been
associated with increased risk of morbidity (13) and earlier mortality.
(14) Nutritional factors, such as having been breastfed, are also
associated with adult health status. (15) Greater height (an indicator
of better nutrition, less illness and higher SEP in childhood) is
related to lower risk of coronary heart disease, although it may be
linked to higher risk of some cancers. (16) Low birth weight is
associated with increased cardiovascular disease and type 2 diabetes.
(17) Not surprisingly, poorer adult health status is predicted by having
a serious illness in childhood. (18) In addition to predicting adult
occupational prestige, (19) higher childhood cognitive ability is
associated with improved longevity in adulthood. (20) Whether this is
through socio-economic or biological pathways is a matter of debate.
(21)
To further test the healthy migrant hypothesis, we also
investigated age at immigration, which may play an important role in the
healthy migrant effect. If self-selection is integral for determining
the "hardiness" of those who migrate, adults who choose to
emigrate under their own volition should have better health profiles
than those who emigrate in their childhood. This speculation is
supported for the outcomes of functional limitations (3) and cognitive
decline (22) by studies of Mexicans migrating to the US; however, as far
as we are aware, it has yet to be shown for those relocating between
high-resource countries.
Finally, we explore another explanation for the unusually good
health status of immigrants in host countries, called the
"selective reverse migration hypothesis". This hypothesis
suggests that immigrants who develop a serious illness are more likely
than healthy immigrants to return to their native country, thereby
lowering the observed morbidity and mortality rates of immigrants in the
host country. Again, this explanation has been investigated primarily in
Hispanic populations in the US. (3,8) Moreover, evidence in favour of it
has been mixed, only some studies finding support among Mexican
Americans. (3,8) If the childhood health status of emigrants who remain
abroad is better than that of emigrants who subsequently return to their
native country, this would provide evidence of reverse migration bias.
This may be particularly true for source countries, such as the UK,
which provide universal health care.
Using the NSHD, the goal of the current study was to compare the
childhood, pre-emigration characteristics of the following:
non-emigrants with that of emigrants; those who emigrated before the age
of 20 with those who emigrated after age 20; and emigrants who remained
abroad with those who returned to live in their native country. To our
knowledge, this is the first study of its kind.
METHODS
Study sample
The NSHD is based on a representative sample of 5,362 single,
legitimate births in England, Scotland or Wales in March 1946.
Subsequent waves of data collection have been conducted at least every
two years during childhood and at regular intervals during adulthood,
the most recent being an ongoing series of data collections completed in
2006-2011, when study participants were aged 60-64 years. (23,24) For
more information about study design, please see Wadsworth et al. and Kuh
et al. (23,24)
Measures
In each wave of the NSHD, if an individual was no longer available
to be interviewed, information was recorded on whether the respondent
had died, emigrated and/or was living abroad, refused to participate or
could not be contacted. We classified all those who emigrated and/or
were living abroad as "emigrants". Those respondents who were
interviewed in at least one wave after the first wave in which it was
recorded that they had emigrated or were living abroad were classified
as "return emigrants".
The age at emigration was defined as the age of the birth cohort at
the first wave of data collection in which the respondent was living
abroad. For example, if the respondent had been interviewed in the UK in
1953 (at age 7) but had emigrated by the 1954 wave of data collection
(at age 8), the respondent would have been classified as emigrating by
age 8. Emigrants were divided into two groups: those who had emigrated
before age 20 and those who had emigrated between ages 20 and 53. Age 53
was chosen as the upper limit in order to avoid post-retirement
emigrants.
In total, there were 984 emigrants and 4,378 non-emigrants. Of the
emigrants, 427 migrated before and 557 after age 20, 602 emigrants
remained abroad, and 382 returned to Britain. The sample sizes varied
slightly within waves of data collection as a result of non-response or
missing data, and from wave to wave because of attrition.
We investigated a variety of variables known to influence adult
health. Childhood SEP and material living conditions were assessed
through six indicators ascertained in the early waves of data
collection: father's occupation, mother's education, home
ownership, housing quality, care of house and child, and parental
involvement in child's education. Father's occupation,
assessed when the respondent was aged 4, was categorized according to
the UK Registrar General's social classification into five levels
(I or II professional and managerial, IIINM skilled nonmanual, IIIM
skilled manual, IV manual partly skilled or V nonskilled; and father
dead). Mother's highest level of education assessed when study
members were aged 6 was dichotomized into primary only versus higher
than primary. Whether the respondent's childhood home was owned by
the parents was assessed at age 6. A score of housing quality (range
0-3) was ascertained when the respondent was aged 4; one point was
allocated for each of the following: dwelling in very good repair,
dwelling built since 1919 and no overcrowding ([less than or equal to]
1.5 persons per room). Scores were divided into best, intermediate and
worst. Also at age 4, care of the house and child was assessed. One
point was allocated for each of the following: very clean house, very
clean child, at least adequate shoes, at least adequate clothes and
mother coped well. Scores were divided into best, intermediate and
worst. Parental interest in the study member's education, based on
comments made by class teachers when the child was aged 7 and on the
records of the number of times parents visited the primary school to
discuss their child's progress, distinguished parents with high,
fair or low levels of interest.
Measures of childhood health included low birth weight, body mass
index (BMI) at age 4, stunting at age 6 and having a serious childhood
illness before the age of 5. Low birth weight was defined as < 2.5
kilograms. Body mass index at age 4 was divided into three categories:
low (<-2 standard deviations [SD] of the mean), normal (between -2
and +2 SD) and high (> 2 SD). Individuals were defined as stunted in
height at age 6 if their height was less than two SD of the average
height of 6-year-olds based on 2007 World Health Organization measures.
(25,26) Finally, if respondents had been hospitalized for a minimum of
28 days before the age of 5, they were classified as having had a
serious childhood illness. We also measured whether the respondent had
been breastfed according to mother's reports of this at age 2.
Finally, childhood cognitive ability was measured at age 8 using tests
developed by the National Foundation for Educational Research, which
measured both verbal and nonverbal ability, including reading
comprehension, word reading ability, vocabulary and picture
intelligence. (27)
Data Analysis
Chi-squared tests were used to analyze categorical data and
independent t-tests were used for continuous data. To adjust the
analyses for other covariates, logistic regression was conducted.
Statistical significance was set at p < 0.05. All analyses were
undertaken using SPSS 21.
RESULTS
The results of the analysis testing the healthy migrant hypothesis
are presented in Table 1. Note that the sample sizes varied slightly
depending on the number of missing values in the measures used.
Comparing the childhood characteristics of Britons who had ever
emigrated with those who had never emigrated revealed that a larger
proportion of emigrants-to-be had a higher SEP (measured by
father's occupation and mother's education), better material
living conditions (measured by housing quality, home ownership and care
of house and child) and greater parental interest in their education
(each p < 0.01). Also, a greater share of emigrants compared with
non-emigrants were breastfed, not stunted and did not have a low birth
weight or a serious illness in childhood (each p < 0.05). Finally,
emigrants were taller and had higher scores on cognition in childhood
than non-emigrants (each p < 0.001). The only childhood indicators
that were not associated with emigration status were sex, BMI at age 4
and birth weight.
To further investigate the healthy migrant effect, we undertook a
multivariable analysis with emigration status as the dependent variable.
To avoid problems with multi-collinearity and to achieve parsimony in
the final model we ran a preliminary model, which included the outcome
and all six measures of childhood SEP and material living conditions, to
determine which of the variables should be included in the final model.
Only two variables remained statistically significant: father's
occupation and housing quality. These variables were included in the
final model, along with the variables most closely related to childhood
health: low birth weight, having had a serious childhood illness by age
5, height at age 6 and cognitive ability at age 8. As shown in Table 2,
when all factors were considered simultaneously, height at age 6,
father's occupation, and cognitive ability at age 8 remained
statistically significant. Specifically, the odds of emigration
increased by 3% for every centimetre in height at age 6 (p < 0.01)
and 36% for every one SD increase in cognitive ability (p < 0.001).
Those with fathers who had an occupation in the professional or skilled
manual labour categories had 56% and 34% higher odds respectively of
emigrating than those whose fathers were partly or non-skilled manual
labourers (p < 0.01 and p < 0.05). In short, both childhood
socio-economic and health factors, when mutually adjusted, predicted
emigration from the UK.
Table 3 reports the results of the second descriptive analysis
examining the hypothesis about age of migration. Comparison of the
childhood characteristics of Britons who emigrated before and after age
20 revealed only one difference: Older emigrants had a marginally higher
cognitive ability in childhood (p = 0.05). All other measures of
childhood SEP and health were comparable across older and younger
emigrants and, thus, further multivariate analysis was not undertaken.
In sum, we found little evidence supporting the speculation that
early-life characteristics relate to age at emigration.
The results of the third descriptive analysis investigating the
selective reverse-migration hypothesis are presented in Table 4.
Comparing return emigrants with those staying abroad shows that a
greater share of returning Britons had a higher childhood SEP as
measured by their mother's level of education (p < 0.05). All
other factors were comparable. Again, further multivariate analysis was
not undertaken because we found little evidence suggesting that
childhood characteristics were related to return migration.
DISCUSSION
Consistent with the healthy migrant hypothesis, British emigrants
born in 1946 had higher childhood SEP, better childhood health
(indicated by height at age 6) and superior cognitive ability at age 8
than their counterparts who stayed in Britain. Moreover, each of these
factors predicted emigration status, when all others had been controlled
for. Our research adds to the literature on the healthy migrant effect
in several important ways. First, we introduce support for the idea that
the factors contributing to positive health selection in migrant
populations begin as far back as childhood. Previous research has shown
that Mexican immigrants are taller than their non-migrant counterparts
and, thus, has implied that they also had better nutrition in childhood.
(8,9) Our research is able to show this directly using prospective
measures of childhood health. Second, we reveal that the healthy migrant
effect can be found in a sample of those emigrating from a high-resource
sending country. This finding suggests that positive health selection
does not depend entirely on factors affecting individuals moving from a
low- to a higher-income country.
Counter to the selective-reverse migration hypothesis, we found no
evidence that emigrants with worse childhood health were more likely to
return home. The only childhood indicator that differed between
emigrants staying abroad and those returning home was mother's
level of education, which was higher among return emigrants. This
relationship was in the opposite direction than we had anticipated and
was not corroborated by father's occupation, suggesting that it may
have been a chance finding, especially given the multiple tests
undertaken. However, it is also possible that higher maternal education
is associated with greater perceived employment opportunities for the
mother in her home country, especially if she had not been employed to
her highest capabilities abroad. Further research is needed to explore
this. The lack of support for selective-reverse migration is somewhat
surprising given that returning emigrants would have regained access to
the National Health Service, which is free for all British citizens.
However, while this may have been an attractive option for some migrants
who fell ill, it would not necessarily have made a difference for those
ending up in Canada or another country with universal health care. That
said, it is also possible that the original survey respondents would not
have been quite old - or sick--enough to move back home for this reason.
The current study also reveals that age at emigration does not
appear to matter with respect to the selection process for British
emigrants, except when it comes to cognitive ability: Britons emigrating
after age 20 had higher childhood cognitive ability at age 8. This
finding deserves some attention. First, in contrast to what we had
originally hypothesized it suggests that self-selection at the level of
the parent may take into account the health of children. That is, with
the exception of some special cases, we suspect that parents from
high-income countries, particularly those from countries with universal
health care, are not likely to choose to emigrate if they have unhealthy
children. Second, if all else is equal with respect to childhood health,
it makes sense that childhood cognitive ability, which predicts adult
occupational prestige, (19) would emerge as an important selective
factor for successful immigration in adulthood. Cognitive ability is
related not only to enhanced ability to adapt to a new environment but
also to the type of opportunity for which an adult may move. (28) By the
mid-1970s, countries receiving large numbers of British emigrants, such
as Canada and Australia, were increasingly focused on attracting the
most skilled migrants (a policy that continues until today). (29)
Immigration policies favouring highly skilled workers, therefore, would
have further contributed to selection based on cognitive ability for at
least some of the NSHD emigrants.
There are several study limitations that should be considered when
interpreting our results. First, we lacked data on where Britons
emigrated to or why, and so we cannot fully specify the selection
process behind the healthy migrant effect observed in our data. Second,
it is possible we misclassified some of those who could not be contacted
or refused to participate as "nonemigrants" when they may have
subsequently emigrated. Such an error would bias the results towards the
null, rendering our results more conservative. Third, because factors
underlying migratory selection processes are likely to vary according to
the historical time period and regions of the world between which
migrants are moving, our findings should not be generalized to other
birth cohorts or sending populations. For example, while many Britons
born in 1946 would have moved to other wealthy, English-speaking
countries, such as Canada, more recent British emigrants are
increasingly moving within the European Union (EU). (30) This shift in
receiving countries may affect the strength of selection effects,
especially given the ease with which migration has been possible between
nations of the EU. Also, the selection processes underlying the health
profiles of migrants moving between wealthy countries likely differ from
those of migrants moving from poorer areas of the world to more
economically advantaged ones. The latter, who are the most common types
of migrant today, may be more strongly selected for their health because
of the need to find work. (31)
Our finding that the childhood health of return migrants was
similar to those who stayed abroad provides no support for the selective
reverse migration hypothesis. Future research could more definitively
test this hypothesis by also gathering data about the adult health of
migrants and non-migrants. Finally, the sample sizes used for tests of
the selective reverse-migration and age of migration hypotheses were
relatively small. Consequently, these migrant-based analyses would have
had less power to detect significant differences (especially for the
categorical variables) than the main analysis using the full sample,
although in many cases the direction of the observed association was
opposite to that hypothesized.
In conclusion, using pre-emigration prospective data, we found
evidence of a healthy migrant effect among Britons born in 1946.
Moreover, differences found in the childhood characteristics of
emigrants and non-emigrants suggest that the underlying selection
processes likely began in childhood. In particular, British emigrants
had higher childhood SEP, better childhood health (as indicated by
greater height at age 6) and better cognitive ability at age 8 than
their counterparts who stayed at home. We also found some evidence that
adult emigrants were more strongly selected for cognitive ability at age
8 than those who emigrated before age 20. More research using
prospective birth cohort data and a range of wealthy and less
economically advantaged sending nations is needed to more fully
understand the selection processes that underlie the healthy migrant
effect.
REFERENCES
(1.) Ng E. The healthy immigrant effect and mortality rates. Health
Rep 2011;22(4):C1-C9. PMID: 22352149.
(2.) McDonald J, Kennedy S. Insights into the 'healthy
immigrant effect': Health status and health service use of
immigrants to Canada. Soc Sci Med 2004;59(8):1613-27. PMID: 15279920.
doi: 10.1016/j.socscimed.2004.02.004.
(3.) Fuller-Thomson E, Nuru-Jeter A, Richardson D, Raza F, Minkler
M. The Hispanic paradox and older adults' disabilities: Is there a
healthy migrant effect? Int J Environ Res Public Health
2013;10(5):1786-814. PMID: 23644828. doi: 10.3390/ijerph10051786.
(4.) Huang C, Mehta NK, Elo IT, Cunningham SA, Stephenson R,
Williamson DF, et al. Region of birth and disability among recent US
immigrants: Evidence from the 2000 census. Popul Res Policy Rev
2011;30(3):399-418. PMID: 21666828. doi: 10.1007/s11113-010-9194-x.
(5.) Argeseanu Cunningham S, Ruben JD, Narayan KMV. Health of
foreign-born people in the United States: A review. Health Place
2008;14(4):623-35. PMID: 18242116. doi:
10.1016/j.healthplace.2007.12.002.
(6.) Guo S, Lucas RM, Joshy G, Banks E. Cardiovascular disease risk
factor profiles of 263,356 older Australians according to region of
birth and acculturation, with a focus on migrants born in Asia. Plos One
2015;10(2): e0115627. PMID: 25695771. doi: 10.1371/journal.pone.0115627.
(7.) Franzini L, Ribble JC, Keddie AM. Understanding the Hispanic
paradox. Ethn Dis 2001;11(3):496-518. PMID: 11572416.
(8.) Riosmena F, Wong R, Palloni A. Migration selection,
protection, and acculturation in health: A binational perspective on
older adults. Demography 2013;50(3):1039-64. PMID: 23192395. doi:
10.1007/s13524-0120178-9.
(9.) Crimmins EM, Soldo BJ, Kim JK, Alley DE. Using anthropometric
indicators for Mexicans in the United States and Mexico to understand
the selection of migrants and the "Hispanic paradox". Soc Biol
2005; 52(3/4):164-77. PMID: 17619609.
(10.) Kennedy S, McDonald JT, Biddle N. The Healthy Immigrant
Effect and Immigrant Selection: Evidence from Four Countries, 2006;
SEDAP Research Paper No. 164. Available at:
https://ideas.repec.org/p/mcm/sedapp/164.html (Accessed July 13, 2015).
(11.) Constantine S. British emigration to the empire-commonwealth
since 1880: From overseas settlement to diaspora? In: Fedorowich K (Ed),
The British World: Diaspora, Culture, and Identity. London: Frank Cass,
2003;16-35.
(12.) Statistics Canada. National Household Survey. Statistics
Canada Catalogue no.99-010-X2011026, 2011.
(13.) Strand BH, Murray ET, Guralnik J, Hardy R, Kuh D. Childhood
social class and adult adiposity and blood-pressure trajectories 36-53
years: Gender-specific results from a British birth cohort. J Epidemiol
Community Health 2012;66(6):512-18. doi: 10.1136/jech.2010.115220.
(14.) Kuh D, Hardy R, Langenberg C, Richards M, Wadsworth MEJ.
Mortality in adults aged 26-54 years related to socioeconomic conditions
in childhood and adulthood: Post war birth cohort study. BMJ
2002;325(7372):1076-80. PMID: 12424168. doi: 10.1136/bmj.325.7372.1076.
(15.) Parikh NI, Hwang S, Ingelsson E, Benjamin EJ, Fox CS, Vasan
RS, et al. Breastfeeding in infancy and adult cardiovascular disease
risk factors. Am J Med 2009;122(7):656-80. PMID: 19559168. doi:
10.1016/j.amjmed.2008.11.034.
(16.) Batty GD, Shipley MJ, Gunnell D, Huxley R, Kivimaki M,
Woodward M, et al. Height, wealth, and health: An overview with new data
from three longitudinal studies. Econ Hum Biol 2009;7(2):137-52. PMID:
19628438. doi: 10.1016/j.ehb.2009.06.004.
(17.) Barker DJ. The developmental origins of well-being. Philos
Trans R Soc Lond B Biol Sci 2004;359(1449):1359-66. PMID: 15347527. doi:
10.1098/rstb.2004.1518.
(18.) Kuh D, Wadsworth M, Yusuf E. Burden of disability in a post
war birth cohort in the UK. J Epidemiol Community Health
1994;48(3):262-69.
(19.) Cheng H, Furnham A. Childhood cognitive ability, education,
and personality traits predict attainment in adult occupational prestige
over 17 years. J Vocat Behav 2012;81(2):218-26. doi:
10.1016/j.jvb.2012.07005.
(20.) Batty GD, Deary IJ, Gottfredson LS. Premorbid (early life) IQ
and later mortality risk: Systematic review. Ann Epidemiol
2007;17(4):278-88. PMID: 17174570. doi: 10.1016/j.annepidem.2006.07.010.
(21.) Kuh D, Shah I, Richards M, Mishra G, Wadsworth M, Hardy R. Do
childhood cognitive ability or smoking behaviour explain the influence
of lifetime socio-economic conditions on premature adult mortality in a
British post war birth cohort? Soc Sci Med 2009;68(9):1565-73. PMID:
19269077. doi: 10.1016/j.socscimed.2009.02.006.
(22.) Hill T, Angel J, Balistreri K. Does the "healthy
immigrant effect" extend to cognitive aging? In: Angel JL,
Torres-Gil F, Markides K (Eds), Aging, Health, and Longevity in the
Mexican-Origin Population. New York: Springer US, 2012; 19-33.
(23.) Wadsworth M, Kuh D, Richards M, Hardy R. Cohort profile: The
1946 national birth cohort (MRC National Survey of Health and
Development). Int J Epidemiol 2006;35(1):49-54. doi: 10.1093/ije/dyi201.
(24.) Kuh D, Pierce M, Adams J, Deanfield J, Ekelund U, Friberg P,
et al. Cohort profile: Updating the cohort profile for the MRC National
Survey of Health and Development: A new clinic-based data collection for
ageing research. Int J Epidemiol 2011;40(1):e1-9. PMID: 21345808. doi:
10.1093/ije/dyq231.
(25.) Ong KK, Hardy R, Shah I, Kuh D. National Survey of Health and
Development Scientific and Data, Collection Teams. Childhood stunting
and mortality between 36 and 64 years: The British 1946 birth cohort
study. J Clin Endocrinol Metab 2013;98(5):2070-77. PMID: 23533234. doi:
10.1210/jc.2012-3595.
(26.) World Health Organization. Growth Reference 5-19 years, 2013.
Available at: http://www.who.int/growthref/who2007_height_for_age/en/index. html (Accessed July 13, 2015).
(27.) Pigeon DA. Tests used in the 1954 and 1957 surveys (appendix
1). In: Douglas JWB(Ed), The Home and the School. London; Macgibbon
& Kee, 1964.
(28.) Lu Y. Test of the 'healthy migrant hypothesis': A
longitudinal analysis of health selectivity of internal migration in
Indonesia. Soc Sci Med 2008;67(8):1331-39. PMID: 18639967. doi:
10.1016/j.socscimed.2008.06.017.
(29.) Khoo S, Hugo G, McDonald P. Skilled migration from Europe to
Australia. Population Space and Place 2011;17(5):550-66. doi:
10.1002/psp.651.
(30.) Hawkins O. Migration Statistics, 2013. Available at:
http://www.parliament.uk/briefing-papers/SN06077 (Accessed July 13,
2015).
(31.) Malmusi D, Borrell C, Benach J. Migration-related health
inequalities: Showing the complex interactions between gender, social
class and place of origin. Soc Sci Med 2010;71(9):1610-19. PMID:
20869798. doi: 10.1016/j. socscimed.2010.07.043.
Received: July 13, 2015
Accepted: September 27, 2015
Esme Fuller-Thomson, PhD, [1] Sarah Brennenstuhl, PhD, [2] Rachel
Cooper, PhD, [3] Diana Kuh, PhD, FMedSci [3]
[1.] Factor-Inwentash Faculty of Social Work, University of
Toronto, Toronto, ON
[2.] Lawrence Bloomberg Faculty of Nursing, University of Toronto,
Toronto, ON
[3.] MRC Unit for Lifelong Health and Ageing at UCL, London, UK
Correspondence: Esme Fuller-Thomson, PhD, Factor Inwentash Faculty
of Social Work & Institute for Life Course and Aging, University of
Toronto, 246 Bloor St. W., Toronto, ON M6S 3W6, Tel: ^416-978-3269,
E-mail: esme.fuller.thomson@utoronto.ca
Funding: E. Fuller-Thomson is supported by the Sandra Rotman Chair
in Social Work. R. Cooper and D. Kuh are supported by the UK Medical
Research Council (Programme code: MC_UU_12019/1). The MRC National
Survey of Health and Development is funded by the UK Medical Research
Council.
Conflict of Interest: None to declare.
Table 1. Comparison of the childhood characteristics of
Britons born in 1946 who never emigrated versus emigrants
Never emigrated
N % or
mean
(SD)
Sex
Male 2286 52.2%
Female 2092 47.8%
Socio-economic position & material living conditions
Father's occupation
Professional or zmanagerial 809 20.8%
Skilled non-manual 691 17.7%
Manual--skilled 1215 31.2%
Manual--partly or non-skilled 1126 28.9%
Father dead 56 1.4%
Mother's education
Primary only 2417 65.1%
More than primary 1296 34.9%
Parents own home, age 6
Not owner 2827 73.8%
Owner 1004 26.2%
Housing quality, age 4
Best 1982 51.6%
Intermediate 1025 27.7%
Worst 835 21.7%
Care of house and child
Best 1426 39.9%
Intermediate 939 26.2%
Worst 1213 33.9%
Parental interest in school progress
Very interested 1481 41.7%
Average interest 1715 48.3%
Low interest 356 10.0%
Health indicators
Serious illnesses
No serious illness 3892 88.9%
Serious illness 486 11.1%
Height (cm), age 6 3480 113.9 (5.4)
Stunted ([dagger])
Yes 204 5.9%
No 3276 94.1%
Birth weight (g) 4346 3372.8 (552.3)
Low birth weight ([double dagger])
Yes 188 4.3%
No 4158 95.7%
Body mass index, age 4 ([section])
Low 37 1.0%
Medium 3383 95.1%
High 136 3.8%
Cognitive ability, age 8 3665 99.3 (15.0)
Breastfed
Yes 3037 75.7%
No 974 24.3%
Emigrated
N % or
mean
(SD)
Sex
Male 529 53.8%
Female 455 46.2%
Socio-economic position & material living conditions
Father's occupation
Professional or zmanagerial 201 30.0%
Skilled non-manual 133 19.8%
Manual--skilled 206 30.7%
Manual--partly or non-skilled 119 17.7%
Father dead 12 1.8%
Mother's education
Primary only 342 54.4%
More than primary 287 45.6%
Parents own home, age 6
Not owner 429 66.7%
Owner 214 33.3%
Housing quality, age 4
Best 409 59.4%
Intermediate 162 23.5%
Worst 118 17.1%
Care of house and child
Best 311 48.7%
Intermediate 164 25.7%
Worst 163 24.4%
Parental interest in school progress
Very interested 277 47.9%
Average interest 264 45.7%
Low interest 37 6.4%
Health indicators
Serious illnesses
No serious illness 706 5.0%
Serious illness 37 5.0%
Height (cm), age 6 581 115.1 (5.1)
Stunted ([dagger])
Yes 19 3.3%
No 562 96.7%
Birth weight (g) 981 3393.2 (489.8)
Low birth weight ([double dagger])
Yes 28 2.9%
No 953 97.1%
Body mass index, age 4 ([section])
Low 8 1.3%
Medium 599 95.4%
High 21 3.3%
Cognitive ability, age 8 591 104.2 (14.5)
Breastfed
Yes 614 79.4%
No 159 20.6%
p-value *
Sex
Male 0.38
Female
Socio-economic position & material living conditions
Father's occupation
Professional or zmanagerial <0.001
Skilled non-manual
Manual--skilled
Manual--partly or non-skilled
Father dead
Mother's education
Primary only <0.001
More than primary
Parents own home, age 6
Not owner <0.001
Owner
Housing quality, age 4
Best <0.001
Intermediate
Worst
Care of house and child
Best <0.001
Intermediate
Worst
Parental interest in school progress
Very interested <0.01
Average interest
Low interest
Health indicators
Serious illnesses
No serious illness <0.001
Serious illness
Height (cm), age 6 <0.001
Stunted ([dagger])
Yes <0.05
No
Birth weight (g) 0.25
Low birth weight ([double dagger])
Yes <0.05
No
Body mass index, age 4 ([section])
Low =0.74
Medium
High
Cognitive ability, age 8 <0.001
Breastfed
Yes <0.05
No
SD, standard deviation.
* p value derived from chi-squared tests
for categorical variables and
independent t-tests for continuous
variables.
([dagger]) Individuals were defined as
stunted if their height was less than 2
SD of the average height of 6-year-olds
based on 2007 World Health Organization
measures.
([double dagger]) Low birth weight was
defined as <2.5 kilograms.
([section]) Low body mass index was
defined as <-2 SD of the mean, normal
was between -2 and +2 SD, and high was
>2 SD.
Table 2. Logistic regression of childhood characteristics
of British emigrants born in 1946 versus their
non-emigrating counterparts (sample with
complete data on all covariates n = 3529)
OR (95% CI) p-value
Socio-economic position & material living conditions
Father's occupation
Professional or managerial 1.56 (1.15, 2.12) 0.03
Skilled non-manual 1.19 (0.85, 1.65)
Manual--skilled 1.34 (1.01, 1.78)
Manual--partly or non-skilled 1.00 (Ref)
Father dead 1.56 (0.63, 3.84)
Housing quality, age 4
Best 1.00 (Ref) 0.12
Intermediate 0.78 (0.61, 1.00)
Worst 0.95 (0.72, 1.25)
Health indicators
Low birth weight * 0.68 (0.32, 1.42) 0.3
Serious illnesses before age 6 0.88 (0.56, 1.39) 0.59
Cognitive ability, age 8 ([dagger]) 1.36 (1.20, 1.55) <0.001
Height (cm), age 6 1.03 (1.01, 1.05) <0.01
Nagelkerke R square 0.038
-2 Log likelihood 2696.4
OR, odds ratio; CI, confidence interval.
* Low birth weight was defined as <2.5 kilograms.
([dagger]) Per 1 SD change.
Table 3. Comparison of the childhood characteristics of
Britons born in 1946 who emigrated before
age 20 versus age 20 or older
Emigrant Emigrant
<20 years [greater
than or
equal to]
20 years
N % or N
mean (SD)
Sex
Male 219 51.3% 310
Female 208 48.7% 247
Socio-economic position & material living conditions
Father's occupation
Professional or 39 25.8% 162
managerial
Skilled non-manual 34 22.5% 99
Manual--skilled 51 33.8% 155
Manual--partly 24 15.9% 95
or non-skilled
Father dead 3 2.0% 9
Mother's education
Primary only 61 50.8% 281
More than primary 59 49.2% 228
Parents own home, age 6
Not owner 85 67.5% 344
Owner 41 32.5% 173
Housing quality, age 4
Best 100 58.8% 309
Intermediate 43 25.3% 119
Worst 27 15.9% 91
Care of house and child
Best 76 50.0% 235
Intermediate 42 27.6% 122
Worst 34 22.4% 129
Parental interest in school progress
Very interested 38 42.2% 239
Average interest 43 47.8% 221
Low interest 9 10.0% 28
Health indicators
Serious illnesses
No serious illness 178 95.7% 528
Serious illnesses 8 4.8% 29
Height (cm), age 6 117 114.1 (4.9) 464
Stunted ([dagger])
Yes 2 1.7% 17
No 115 98.3% 447
Birth weight (g) 426 3372.3 (497.8) 555
Low birth
weight ([double
dagger])
Yes 16 3.8% 12
No 410 96.2% 543
Weight, age
4 ([section])
Low 2 1.3% 6
Medium 148 97.4% 451
High 2 1.3% 19
Cognitive ability, 90 101.4 (14.1) 501
age 8
Breastfed
Yes 198 79.5% 416
No 51 20.5% 108
Emigrant P-
[greater value *
than or
equal to]
20 years
% or
mean (SD)
Sex
Male 55.7% 0.17
Female 44.3%
Socio-economic position & material living conditions
Father's occupation
Professional or 31.2% 0.59
managerial
Skilled non-manual 19.0%
Manual--skilled 29.8%
Manual--partly 18.3%
or non-skilled
Father dead 1.7%
Mother's education
Primary only 55.2% 0.39
More than primary 44.8%
Parents own home, age 6
Not owner 66.5% 0.84
Owner 33.5%
Housing quality, age 4
Best 59.5% 0.77
Intermediate 22.9%
Worst 17.5%
Care of house and child
Best 48.4% 0.57
Intermediate 25.1%
Worst 26.5%
Parental interest in school progress
Very interested 49.0% 0.23
Average interest 45.3%
Low interest 5.7%
Health indicators
Serious illnesses
No serious illness 94.8% 0.62
Serious illnesses 5.2%
Height (cm), age 6 115.1 (5.2) 0.77
Stunted ([dagger])
Yes 3.7% 0.29
No 96.3%
Birth weight (g) 3409.3 (483.4) 0.24
Low birth
weight ([double
dagger])
Yes 2.2% 0.14
No 97.8%
Weight, age
4 ([section])
Low 1.3% 0.28
Medium 94.7%
High 4.0%
Cognitive ability, 104.6 (14.5) 0.05
age 8
Breastfed
Yes 79.4% 0.97
No 20.6%
SD, standard deviation.
* p-value derived from chi-squared tests for categorical
variables and independent t-tests for continuous variables.
([dagger]) Individuals were defined as stunted if their
height was less than 2 SD of the average height of
6-year-olds based on 2007 World Health Organization
measures.
([double dagger]) Low birth weight was defined as <2.5
kilograms.
([section]) Low body mass index was defined as < -2 SD of
the mean, normal was between -2 and +2 SD, and high was >2
SD.
Table 4. Comparison of the childhood characteristics of British
emigrants born in 1946 who stayed abroad versus
emigrants who returned home
Emigrants Emigrants--
--stayed returned home
abroad
N % or N
Mean (SD)
Sex
Male 334 55.5% 195
Female 268 44.5% 187
Socio-economic position & material living conditions
Father's
occupation
Professional or 115 29.4% 86
managerial
Skilled non- 77 19.7% 56
manual
Manual-skilled 130 33.2% 76
Manual-partly 66 16.9% 53
or non-
skilled
Father dead 3 0.8% 9
Mother's education
Primary only 208 58.8% 134
More than 146 41.2% 141
primary
Parents own home, age 6
Not owner 242 67.2% 187
Owner 118 32.8% 96
Housing quality, age 4
Best 231 59.4% 178
Intermediate 86 22.1% 76
Worst 72 18.5% 46
Care of house and child
Best 160 45.5% 151
Intermediate 100 28.4% 64
Worst 92 26.1% 71
Parental interest in
school progress
Very interested 154 47.4% 123
Average interest 153 47.1% 111
Low interest 18 5.5% 19
Health indicators
Serious illnesses
No serious 397 94.7% 309
illness
Serious illness 22 5.3% 15
Height (cm), 326 115.1 (5.2) 255
age 6
Stunted ([dagger])
Yes 9 2.8% 10
No 317 97.2% 245
Birth weight (g) 599 3384.2 (479.2) 382
Low birth
weight ([double
dagger])
Yes 15 2.5% 13
No 584 97.5% 369
BMI, age
4 ([section])
Low 4 1.1% 4
Medium 332 95.1% 267
High 13 3.7% 8
Cognitive 332 104.7 (14.1) 259
ability,
age 8
Breastfed
Yes 355 79.2% 259
No 93 20.8% 66
Emigrants-- P-
returned home value *
% or
Mean (SD)
Sex
Male 51.0% 0.17
Female 49.0%
Socio-economic position & material living conditions
Father's
occupation
Professional or 30.7% 0.09
managerial
Skilled non- 20.0%
manual
Manual-skilled 27.1%
Manual-partly 18.9%
or non-
skilled
Father dead 3.2%
Mother's education
Primary only 48.7% <0.05
More than 51.3%
primary
Parents own home, age 6
Not owner 66.1% 0.76
Owner 33.9%
Housing quality, age 4
Best 59.3% 0.42
Intermediate 25.3%
Worst 15.3%
Care of house and child
Best 52.8% 0.13
Intermediate 22.4%
Worst 24.8%
Parental interest in
school progress
Very interested 48.6% 0.54
Average interest 43.9%
Low interest 7.5%
Health indicators
Serious illnesses
No serious 95.4% 0.7
illness
Serious illness 4.6%
Height (cm), 115.2 (5.0) 0.83
age 6
Stunted ([dagger])
Yes 3.9% 0.44
No 96.1%
Birth weight (g) 3407.3 (506.3) 0.48
Low birth
weight ([double
dagger])
Yes 3.4% 0.41
No 96.6%
BMI, age
4 ([section])
Low 1.4% 0.8
Medium 95.7%
High 2.9%
Cognitive 103.5 (14.9) 0.34
ability,
age 8
Breastfed
Yes 79.7% 0.88
No 20.3%
SD, standard deviation.
* p value derived from chi-squared tests for categorical
variables and independent t-tests for continuous variables.
([dagger]) Individuals were defined as stunted if their
height was less than 2 SD of the average height of
6-year-olds based on 2007 World Health Organization
measures.
([double dagger]) Low birth weight was defined as <2.5
kilograms.
([section]) Low body mass index was defined as <-2 SD of the
mean, normal was between -2 and +2 SD, and high was >2 SD.
