Perceived health status of Francophones and Anglophones in an officially bilingual Canadian province.
Belanger, Mathieu ; Bouchard, Louise ; Gaboury, Isabelle 等
Inequalities in health represent a major public health issue for
many countries and their governing bodies. (1) Many factors have been
suggested to contribute to health inequalities, including socio-economic
disparities (2,3) and social capital. (4) Actual and perceived health
inequalities are of particular concern among minority groups. In the US,
inequalities in health and mortality have been documented between
Caucasians and the African-American minority. (5,6) In comparison to the
majority, lower self-rated health has also been demonstrated in minority
populations of Australia (Aboriginal and Torres Straight Islander
communities). (7) Despite boasting a public health care system,
discrepancies in health exist between the Caucasian majority and racial
and ethnic minorities in Canada. (8,9) Differences in health also appear
to exist between the Anglophone majority and the Francophone minority in
Canada. (10)
Canadian Francophones outside the province of Quebec are described
as living in a minority setting. They are dispersed throughout the
country in communities that tend to be rural and present a relatively
old population with low levels of education and socio-economic status.
(11) A recent Canadian study suggested that, compared to Anglophone men
outside of Quebec, more Francophone men living in minority settings rate
their health as being poor. (10) No such difference was noted among
women. In Ontario, Francophone men and women were also more likely to
report being in poorer health than Anglophones and Allophones. (12)
New Brunswick represents Canada's only officially bilingual
province. Similar to Canada, about one third of its population is
Francophone which, on a relative scale, represents the largest
concentration of Francophones outside of Quebec. Two ecological studies
have explored the health of Francophones in this province. They
indicated that whereas the health of populations in mainly Francophone
regions of New Brunswick has improved between 1985 and 2003, disparities
still exist between French and English regions. (13,14) An analysis of
self-rated health using individual-level data and accounting for known
determinants of health has not yet been conducted in New Brunswick. It
therefore remains unclear whether language is an important correlate of
perceived health in the only province with an officially bilingual
public health care system. This study aimed at determining whether there
are differences in perceived health between Francophones (linguistic
minority) and Anglophones in New Brunswick when accounting for known
determinants of health.
METHODS
Population
Data were obtained from the Canadian Community Health Survey. To
meet analytical needs, data from cycles 1.1, 2.1, 3.1, and 4.1
(conducted in 2001, 2003, 2005 and 2007, respectively) were combined
as suggested by Thomas and Wannell. (15) This analysis was
restricted to New Brunswick household residents over 25 years old. The
response proportions for cycle 1.1, 2.1, 3.1 and 4.1 were 84.7%, 80.6%,
79.0% and 77.6%, respectively and there were 4996, 4929, 5100 and 2704
New Brunswickers sampled in these respective cycles.
We assigned a Francophone or Anglophone linguistic identity to
participants according to the algorithm presented in Figure 1. (16) More
specifically, the following steps were followed until one mutually
exclusive category was identified: 1) participants were sorted based on
the language(s) they can use to converse; as necessary, additional
sorting was performed based on 2) mother tongue, 3) language of
interview, and 4) preferred language.
Variables
Self-perceived health of participants was measured with the
question: "In general, would you say your health is?" followed
by five response options which were dichotomized into good (excellent,
very good, and good) or poor (fair or poor). Such a measure of
self-rated health has been shown to be a valuable predictor of
mortality. (17) Socio-demographic variables included quintiles of
household income (adjusted for Canadian region), level of education
(university, postsecondary, high school, or high school not completed),
and employment status (full-time, part-time, unemployed). Contextual
information on the respondents' lives was complemented with a
variable representing living in a rural or urban setting and type of
household (living alone, with a partner, with a partner and children,
single parent, or other). Health-related behaviours included physical
activity (active or inactive), tobacco use (regular, occasional or
former smoker, or never-smoker), diet ([greater than or equal to] 5 or
<5 fruits and vegetables per day), body mass index (<25, 25-30,
>30 kg/[m.sup.2] calculated from self-reported height and weight),
and alcohol use (regular, occasional or former drinker, or
never-drinker). Medical status was based on the presence or absence of
11 chronic conditions, physical incapacity (needing assistance for daily
activities), and self-perceived stress experienced on typical days.
[FIGURE 1 OMITTED]
Analyses
The chi-square statistic was used to compare frequencies for
various variables between Francophones and Anglophones. Additive
multivariate logistic regressions modeled the association between
linguistic identity in New Brunswick and self-perceived health. In
sex-specific analyses, model 1 accounted for age, model 2 accounted for
age and health-related behaviours (physical activity, tobacco use, body
mass index, diet, and alcohol use), model 3 additionally accounted for
socio-demographic variables (income, education, employment status),
model 4 added variables related to living context (urban/rural, type of
household), and model 5 included measures of medical condition
(morbidity, and stress). Analyses were conducted using SAS (version 9.2)
and standard errors were estimated with the Bootstrap method to account
for the complex study design. (18)
RESULTS
In comparison to Anglophones, there were fewer Francophone men and
women reporting their health as being good (Table 1). More Francophone
men reported physical incapacities than Anglophones. However, a similar
number of French- and English-speaking women had physical incapacities.
In both men and women, reports of chronic conditions were more prevalent
among Anglophones than Francophones. Compared to Anglophones, more
Francophones were among the low-income quintiles, were less educated,
and lived in rural areas. Consumption of fruits and vegetables was more
frequent among Francophones and fewer Francophones than Anglophones
reported being obese (BMI [greater than or equal to] 30 kg/[m.sup.2]).
Age-adjusted odds ratios suggest no statistically-significant
difference between the two linguistic groups in likelihood of having a
good self-reported health (Model 1, Table 2). Among men, further
adjustments continued to suggest an absence of difference in the odds of
reporting good health. Most models also suggested no difference between
the odds of reporting good health between French and English women.
However, it must be noted that the confidence intervals for the
estimates were wide and only the end of the tail of the intervals
overlapped the null value of 1. In both men and women, additional
adjustments did not substantively change the estimated effect of
language on the likelihood of reporting good health.
DISCUSSION
Although Francophones, and in particular Francophone women, tended
to be more numerous to rate their health as poor, this study shows that
once we account for well-known determinants of health, similar
proportions of Francophones and Anglophones perceive themselves as being
in good health within the only officially bilingual province in Canada.
Unlike reports of the 1980s and 1990s, (13,14) which were based on
regional-level data and included no statistical adjustments, our results
suggest that there may be no disparity in perceived health between
Francophones and Anglophones in New Brunswick.
Alternatively, the lack of statistical difference between perceived
health of the linguistic groups may be explained by varying
interpretations of health as a concept. Although self-rated health was
shown to be a reliable predictor of morbidity and mortality, (17) the
measurement of self-rated health remains subjective and could be
affected by cultural background. For example, when Hispanics were more
likely to report fair to poor health than non-Hispanic Whites in a
previous study, (19) the authors, who could not identify differences in
morbidity or socio-economic factors, proposed that Hispanics who are
deeply rooted in tradition and culture might be disinclined to rate
their health in a positive manner. (19) In another study, marked
differences in self-reported health between older Black and White adults
with similar levels of physical and psychological function suggested
increased "pessimism" towards health among the Black elderly
population. (20) This relates to the "enduring self-concept"
which proposes that self-rated health be assessed with the recognition
that cultural differences exist in how health is interpreted and
influenced by external factors. (20,21) Challenges associated with the
potentially different meaning of health across different groups have
also been documented among chronically ill and non-ill individuals and
among individuals of varying socio-economic status. (22)
The absence of significant differences in perceived health in our
analysis may also be attributed to changes in the levels of access to
health services. The importance of having access to health services in
one's language was highlighted in a Health Canada report at the
turn of the century. (23) The presence of language barriers has also
been identified as an important determinant of health among Canadian
Francophone minorities. (24) In a recent report, Bouchard et al.
highlighted important developments in Francophone social capital in
Canada and growing vitality of Francophone minorities over the past
half-century, ranging from the Official Languages Act in 1969 to the
survival of the French Montfort Hospital in Ottawa, Ontario. (11)
Elements of growth have also taken place for the Francophone population
of New Brunswick. Among them, we note a substantial increase in the
number of French-speaking physicians and in the number of opportunities
for future physicians to obtain part of their medical training in French
in New Brunswick, (25) the creation of the "Societe sante et
mieux-etre en francais du Nouveau-Brunswick", and the appointment
of a deputy minister of health for Francophones.
The influence of social capital on perceived health has been
observed to promote better health in linguistic minority groups
elsewhere. For example, in Finland, the Swedish-speaking minority
presents more social capital (characterized by social participation and
contacts, trust, and sense of security) than the Finnish-speaking
majority, which is associated with better self-rated health among the
minority group. (26) Similarly, a study in Hungary reported better
self-rated health among German, Romanian and Serbian minorities than
among the Hungarian population, (27) suggesting that in some instances,
being part of a minority appears to be associated with favourable health
indicators. This also suggests a complex relationship between social
positioning and health. With a long history of struggle for recognition
of its identity, the Francophone population of New Brunswick
distinguishes itself from minority groups in these examples since it
does not have a history of elevated social position. The bilingual
status of New Brunswick and its institutions may therefore have
contributed to modifying the relationships of power in the province and
improved the well-being of Francophones. This would be in accordance
with the theory of health gradient, which proposes that social gains can
translate into health gains. (28,29)
The cross-sectional design of this study limits the assessment of
causality. However, it would be impracticable to control assignment of a
variable such as language in any type of study design. Statistics Canada
used both telephone and face-to-face interviews to collect data for this
series of surveys (approximately one third by telephone). It is possible
that people interviewed by telephone overestimated their health to a
greater extent than those interviewed in person. (30) There is
nevertheless no evidence to suggest that this could have introduced bias
in our results since the error of estimation was likely similar among
Francophones and Anglophones. There is no gold standard for the
self-rated assessment of health. The outcome used in this analysis was
based on a single self-reported item. The potential for attenuation of
the odds ratio as a result of non-differential misclassification has to
be taken into consideration.
In conclusion, this study suggests that there may be no difference
in the proportion of Francophones and Anglophones perceiving themselves
as being in good health in New Brunswick. It is nevertheless unclear if
the apparent absence of difference between perceived health status of
the two official linguistic groups in this province is attributable to
improvements in access to health care in one's language, varying
interpretation of the concept of health, random misclassification, or a
combination of these and other factors.
Acknowledgement of support: The analyses were performed in the
context of research programs funded by the Canadian Institutes of Health
Research (Bouchard et al. Les determinants de la sante des minorites
francophones, une analyse secondaire de l'ESCC) and the Consortium
national de formation en sante, volet Universite de Moncton (Bourque et
al.).
REFERENCES
(1.) Commission on Social Determinants of Health. Final Report:
Closing the gap in a generation: Health equity through action on the
social determinants of health. Geneva, Switzerland: World Health
Organization, 2008.
(2.) Marmot M. Social determinants of health inequalities. Lancet
2005;365(9464):1099-104.
(3.) Tarlov AR. Public policy frameworks for improving population
health. Ann N Y Acad Sci 1999;896:281-93.
(4.) Kawachi I. Social capital and community effects on population
and individual health. Ann N Y Acad Sci 1999;896:120-30.
(5.) Willams DR, Collins C. US socioeconomic and racial differences
in health: Patterns and explanations. Annu Rev Sociol 1995;21:349-86.
(6.) Davey Smith G, Neaton JD, Wentworth D, Stamler R, Stamler J.
Mortality differences between black and white men in the USA:
Contribution of income and other risk factors among men screened for the
MRFIT. MRFIT Research Group. Multiple Risk Factor Intervention Trial.
Lancet 1998;351(9107):934-39.
(7.) Australian Bureau of Statistics/Australian Institute of Health
and Welfare. The Health and Welfare of Australian's Aboriginal and
Torres Strait Islander Peoples. Commonwealth of Australia: The
Australian Government, 2008.
(8.) Kobayashi KM, Prus S, Lin Z. Ethnic differences in self-rated
and functional health: Does immigrant status matter? Ethn Health
2008;13(2):129-47.
(9.) Kopec JA, Williams JI, To T, Austin PC. Cross-cultural
comparisons of health status in Canada using the Health Utilities Index.
Ethn Health 2001;6(1):41-50.
(10.) Bouchard L, Gaboury I, Chomienne M-H, Gilbert A, Dubois L. La
sante en situation linguistique minoritaire. Healthcare Policy
2009;4(4):33-40.
(11.) Bouchard L, Gilbert A, Landry R, Deveau K. Social capital,
health, and Francophone minorities. Can J Public Health 2006;97(Suppl
2):S16-S20.
(12.) Institut Franco-Ontarien/Programme de recherche,
d'education et de developpement en sante publique. Deuxieme rapport
sur la sante des francophones de l'Ontario. Ontario: Office of
Francophone Affairs, 2005.
(13.) Desjardins L. La sante des francophones du Nouveau-Brunswick.
Les Editions de la Francophonie. 2003;258.
(14.) Robichaud J-B. La sante des francophones. Objectif 2000, Vol.
1. Moncton, Nouveau-Brunswick: Editions Acadie, 1985;189.
(15.) Thomas S, Wannell B. Combining cycles of the Canadian
Community Health Survey. Health Report (Statistics Canada, Catalogue
82-003-X) 2009;20(1):53-58.
(16.) Bouchard L, Gaboury I, Dubois L, Gilbert A, Chomienne MH,
Beauregard N, Berthelot JM. Disparites de sante et francophonie
minoritaire. 96e Conference annuelle de l'Association canadienne de
sante publique, Ottawa, 2005.
(17.) Idler EL, Benyamini Y. Self-rated health and mortality: A
review of twenty-seven community studies. J Health Soc Behav
1997;38(1):21-37.
(18.) Rao JNK, Wu CFJ, Yue K. Some recent work on resampling
methods for complex surveys. Survey Methodology (Statistics Canada,
Catalogue 12-001) 1992;18(2):209-17.
(19.) Shetterly SM, Baxter J, Mason LD, Hamman RF. Self-rated
health among Hispanic vs. non-Hispanic white adults: The San Luis Valley
Health and Aging Study. Am J Public Health 1996;86(12):1798-801.
(20.) Spencer SM, Schulz R, Rooks RN, Albert SM, Thorpe RJ Jr,
Brenes GA, et al. Racial differences in self-rated health at similar
levels of physical functioning: An examination of health pessimism in
the health, aging, and body composition study. J Gerontol B Psychol Sci
Soc Sci 2009;64(1):87-94.
(21.) Bailis DS, Segall A, Chipperfield JG. Two views of self-rated
general health status. Soc Sci Med 2003;56(2):203-17.
(22.) Quesnel-Vallee A. Self-rated health: Caught in the crossfire
of the quest for "true" health. Int J Epidemiol
2007;36(6):1161-64.
(23.) Commission on the Future of Healthcare in Canada. Building on
values: The Future of Health Care in Canada, Final Report. Ottawa, ON:
Health Canada, 2002.
(24.) Bowen S. Language Barriers in Access to Health Care/Barrieres
linguistiques dans l'acces aux soins de sante. Minister of Public
Works and Government Services, Health Canada, 2001.
(25.) Schofield A, Bourgeois D. Socially responsible medical
education: Innovations and challenges in a minority setting. Med Educ
2010;44(3):263-71.
(26.) Nyqvist F, Finnas F, Jakobsson G, Koskinen S. The effect of
social capital on health: The case of two language groups in Finland.
Health Place 2008;14(2):347-60.
(27.) Komar M, Nagymajtenyi L, Nyari T, Paulik E. The determinants
of self-rated health among ethnic minorities in Hungary. Ethn Health
2006;11(2):121-32.
(28.) Marmot M. Status Syndrome. London, UK: Bloomsbury Publishing,
2004;288.
(29.) Wilkinson RG. Health, hierarchy and social anxiety. Ann N Y
Acad Sci 1999;896:48-63.
(30.) St-Pierre M, Beland Y. Mode Effects in the Canadian Community
Health Survey: A Comparison of CAPI and CATI. Proceedings of the Annual
Meeting of the American Statistical Association, Survey Research Methods
Section, American Statistical Association, 2004.
Received: June 3, 2010
Accepted: October 28, 2010
Author Affiliations
Mathieu Belanger, PhD, [1-4] Louise Bouchard, PhD, [5] Isabelle
Gaboury, PhD, [6] Brigitte Sonier, MSc, [3,7] Isabelle Gagnon-Arpin,
MSc, [5] Aurel Schofield, MD, [1,2,4] Paul-Emile Bourque, PhD [4]
[1] Centre de formation medicale du Nouveau-Brunswick, Moncton, NB
[2] Department of Family Medicine, Universite de Sherbrooke,
Sherbrooke, QC
[3] Vitalite Health Network Research Centre, Moncton, NB
[4] Faculte des sciences de la sante et des services
communautaires, Universite de Moncton, Moncton, NB
[5] Institute of Population Health, University of Ottawa, Ottawa,
ON
[6] Department of Community Health Sciences, University of Calgary,
Calgary, AB
[7] Atlantic Cancer Research Institute, Moncton, NB
Correspondence: Mathieu Belanger, Centre de formation medicale du
Nouveau-Brunswick, Pavillon J.-Raymond-Frenette, 15, rue des Aboiteaux,
Moncton, NB E1A 3E9, Tel: 506-863-2221, Fax: 506-863-2284, E-mail:
mathieu.f.belanger@usherbrooke.ca
Conflict of Interest: None to declare.
Table 1. Socio-demographic and Health-related Characteristics of
Francophones and Anglophones in New Brunswick
Men, %
Francophone Anglophone
(n=2336) (n=4743) p-value
Perceived health
Good 83.3 85.0
Poor 16.7 15.0 0.07
Age (years)
25-44 41.7 43.4
45-64 42.2 39.4
[greater than or equal to] 65 16.0 17.2 0.07
Physical activity
Very active 20.1 19.0
Moderately active 22.5 22.6
Inactive 57.5 58.4 0.6
Tobacco use
Regular 20.8 22.5
Occasional or former 54.9 52.9
Never 24.3 24.6 0.2
Diet (fruits and vegetables)
[greater than or equal to] 5
per day 27.2 24.9
<5 per day 72.8 75.1 0.08
Body mass index (kg/[m.sup.2])
<25 35.5 33.2
25-30 45.0 44.2
>30 19.5 22.6 0.01
Alcohol use
Regular 17.8 16.3
Occasional 65.0 61.7
Never or former 17.3 22.0 <0.001
Household income
Quintile 1 (lower) 18.4 13.9
Quintile 2 20.0 15.6
Quintile 3 18.8 16.9
Quintile 4 17.0 20.7
Quintile 5 (higher) 15.8 20.4
Missing 10.4 12.6 <0.001
Education
University 16.8 19.9
Postsecondary 44.2 46.1
High school 13.1 18.3
Less than high school 25.9 15.7 <0.001
Employment
Active 67.7 70.0
Inactive 32.3 30.0 0.05
Place of residence
Urban 44.6 53.0
Rural 55.4 47.0 <0.001
Type of household
Living alone 14.7 13.6
Single parent 46.9 51.2
Living with partner and
children 38.4 35.2 0.003
Chronic condition
None 57.1 54.4
One 26.5 26.2
Two or more 16.5 19.4 0.01
Has a physical incapacity 13.2 11.4 0.02
Perceives a lot of stress 61.1 61.0 0.8
Women, %
Francophone Anglophone
(n=2524) (n=5051) p-value
Perceived health
Good 83.6 86.3
Poor 16.4 13.8 0.002
Age (years)
25-44 40.4 41.4
45-64 39.7 37.8
[greater than or equal to] 65 19.9 20.9 0.3
Physical activity
Very active 15.3 14.2
Moderately active 23.7 23.0
Inactive 61.0 62.9 0.3
Tobacco use
Regular 18.2 20.4
Occasional or former 46.9 44.3
Never 18.2 20.4 0.03
Diet (fruits and vegetables)
[greater than or equal to] 5
per day 45.7 38.8
<5 per day 54.3 61.2 <0.001
Body mass index (kg/[m.sup.2])
<25 46.0 49.6
25-30 34.3 27.4
>30 19.8 23.0 <0.001
Alcohol use
Regular 30.8 28.1
Occasional 43.6 38.5
Never or former 25.7 33.5 <0.001
Household income
Quintile 1 (lower) 24.1 16.7
Quintile 2 19.0 17.0
Quintile 3 16.6 17.4
Quintile 4 13.2 16.3
Quintile 5 (higher) 13.6 17.0
Missing 13.5 15.7 <0.001
Education
University 16.3 19.1
Postsecondary 42.2 46.1
High school 13.1 18.0
Less than high school 28.4 16.9 <0.001
Employment
Active 55.4 58.0
Inactive 44.6 42.1 0.04
Place of residence
Urban 48.0 55.5
Rural 52.0 44.5 <0.001
Type of household
Living alone 18.2 19.4
Single parent 49.2 47.1
Living with partner and
children 32.6 33.5 0.2
Chronic condition
None 55.5 49.6
One 24.7 27.0
Two or more 19.8 23.5 <0.001
Has a physical incapacity 21.0 20.7 0.7
Perceives a lot of stress 63.3 62.9 0.8
Table 2. Adjusted Odds Ratios (OR) and 95% Confidence Intervals (CI)
for Reporting Good Self-perceived Health in Men and Women
Francophone Model 1 * Model 2 * Model 3
(Reference = Adjusted OR (95% ([dagger]) *([dagger])
Anglophone) CI) Adjusted OR (95% ([double dagger])
CI) Adjusted OR (95%
CI)
Men 0.86 (0.63-1.16) 0.80 (0.58-1.10) 0.93 (0.66-1.30)
Women 0.81 (0.62-1.06) 0.72 (0.54-0.97) 0.84 (0.62-1.13)
Model 4 Model 5
*([dagger])([double*([dagger])([double
dagger])([section])dagger])([section])([parallel])
Adjusted OR (95% Adjusted OR (95%
CI) CI)
Men 0.93 (0.66-1.29) 0.88 (0.61-1.26)
Women 0.82 (0.60-1.11) 0.71 (0.49-1.04)
* Odds ratio adjusted for age; ([dagger]) Adjusted for physical
activity, tobacco use, body mass index, diet, and alcohol use;
([double dagger]) Adjusted for income, education, employment status,
([section]) Adjusted for urban/rural, type of household, ([parallel])
Adjusted for morbidity, and stress.
