Neighbourhood income and neonatal, postneonatal and sudden infant death syndrome (SIDS) mortality in Canada, 1991-2005.
Gilbert, Nicolas L. ; Auger, Nathalie ; Wilkins, Russell 等
Infant mortality declined in Canada from 6.5 per 1,000 live births
in 1991 to 5.1 in 20001 and 5.0 in 2007.2 However, the extent to which
all socio-economic levels benefitted from this decline is unknown.
Low socio-economic status is associated with a higher risk of
adverse perinatal health outcomes, particularly infant mortality. (3,4)
In Canada, findings vary by province and depend on the particular
indicator used. In Nova Scotia, low family income is associated with a
higher risk of postneonatal, but not neonatal mortality. (5) In Quebec,
low maternal education and low neighbourhood income are associated with
higher risks of both neonatal and postneonatal mortality, and the
association between maternal education and infant mortality remains
statistically significant after adjustment for covariates (including
neighbourhood income quintile), while the association between
neighbourhood income and infant mortality is no longer significant after
adjustment for covariates (including maternal education). (6) In British
Columbia, low neighbourhood income is associated with a higher risk of
postneonatal mortality in urban but not rural areas, although the
association is no longer statistically significant after adjustment for
covariates. (7) Thus the nature and extent of the association between
indicators such as neighbourhood income and infant mortality in Canada
remain unclear, and time trends in infant mortality within socioeconomic
strata have not been investigated.
Among causes of death, country-level data are of particular
interest for sudden infant death syndrome (SIDS) because of widespread
education efforts undertaken through the "back to sleep"
campaign in Canada. (8) Although SIDS rates declined in Canada between
1985 and 2005, (9,10) trends over time by socio-economic strata have not
been examined. In Quebec urban areas between 1991 and 2000, poorer
neighbourhoods had a higher risk of SIDS relative to richer
neighbourhoods, but the low number of SIDS deaths led to statistically
non-significant associations, and time trends were not examined. (4)
The objective of this study was to investigate differences and time
trends in neonatal, postneonatal and SIDS mortality across neighbourhood
income quintiles in Canada.
METHOD
We used data from the Canadian linked live birth and infant death
file created by Statistics Canada through probabilistic linkage of birth
and death registrations. (11) Births from 1991 through 2005 were
included, except those from Ontario because of previously documented
concerns related to the completeness of birth registrations, (12) and
those from the three territories (Yukon, Northwest Territories and
Nunavut) because of a high proportion (27%) of missing postal codes.
Births with non-residential postal codes were also excluded.
Maternal residential address at birth was linked to
neighbourhood-level income quintile (based on family income adjusted for
family size and area of residence, as in previous studies (6,7) for the
Canadian census closest to the birth year (1991, 1996, 2001 or 2006)
using Statistics Canada's Postal Code Conversion File Plus. (13)
Adjustment of income for family size and area of residence gives a
better indicator of income adequacy in relation to need, since a family
of four requires more resources than a family of two, and housing costs
vary widely across Canada.
Neighbourhoods were defined at the lowest level for which census
population characteristics were released: enumeration areas for the 1991
and 1996 censuses, and dissemination areas for the 2001 and 2006
censuses. Because neighbourhood income is a less valid marker of
socio-economic status in small towns and rural areas (since rich and
poor people are more likely to share postal codes), analyses were
conducted separately for urban areas and for small towns and rural
areas. Urban areas were defined as census metropolitan areas or census
agglomerations (population 10,000 or more), while small towns and rural
areas included all other areas. (14)
Study outcomes included neonatal death (0-27 days), post neonatal
death (28-364 days), and SIDS (0-364 days), defined as codes 798.0 and
R95 of the ninth and tenth revisions, respectively, of the International
Classification of Diseases (ICD-9 and ICD-10). Starting in 2000, ICD-10
was used for coding the underlying cause of death in Canada.
Neonatal and SIDS mortality rates were computed using live births
in the denominator, while the denominator for postneonatal mortality
rates was the number of infants at risk (infants who had survived 28
days or more). Rates were calculated for the entire population and study
period, by 5-year periods (1991-1995, 1996-2000, 2001-2005) and by
quintiles of neighbourhood income (based on Census data). Changes in
rates between 1991-1995 and 2001-2005 were expressed as differences in
deaths per 1,000 live births or infants at risk and as percentages of
the 1991-1995 rate.
Hazards ratios (HR) and 95% confidence intervals (CI) were
determined by Cox proportional hazards models using days survived as the
time axis. Infants were followed for 364 days from the time of birth, at
which point follow-up ended. Unadjusted HRs and HRs adjusted (aHR) for
province of residence, maternal age (<20, 20-29, 30-34, and [greater
than or equal to] 35 years), parity (1, 2, 3, or [greater than or equal
to] 4 live births, including the current birth), infant sex, and
multiple birth (yes, no) were estimated by neighbourhood income quintile
and period of birth. No adjustment was undertaken for gestational age, a
potential intermediate on the pathway between socio-economic status and
infant death. (15) Robust sandwich estimators were used to adjust
standard errors for clustering of births in dissemination areas. (16) We
used Cox models as they allow us to easily test the possibility that HRs
vary with age of infants at time of death, which would be the case if
hazards were non-proportional (i.e., if the assumption of proportional
hazards was rejected). The proportionality of hazards was verified by
log (-log survival) plots. HRs that are proportional are interpreted in
a similar manner as relative risks averaged over the follow-up period.
(17)
Analyses were carried out using SAS 9.1.
RESULTS
Of 3,258,394 live births in the nine study provinces from 1991 to
2005, 3,173,694 (97.4%) were retained in our analyses. The remaining
2.6% were excluded because of missing, invalid or nonresidential postal
codes. Among the included births, there were 11,765 neonatal, 5,379
postneonatal, and 1,668 SIDS deaths, corresponding to mortality rates of
3.7 per 1,000 live births, 1.7 per 1,000 neonatal survivors, and 0.52
per 1,000 live births, respectively, over the entire study period. Among
SIDS deaths, 93.0% occurred in the postneonatal period.
Higher rates of neonatal, postneonatal and SIDS deaths were
observed in the two poorest neighbourhood income quintiles compared with
the two richest quintiles (Table 1). Infants whose mothers resided in
small towns and rural areas were also at higher risk of death compared
to infants whose mothers resided in urban areas.
In urban areas, differences in mortality rates across neighbourhood
income quintiles were more pronounced. Relative to the richest quintile,
the poorest had aHRs of 1.24, 1.58, and 1.83 for neonatal mortality,
postneonatal mortality and SIDS, respectively (Table 2). * Except for
neonatal mortality, similar but weaker associations were found in small
towns and rural areas, where the association between neighbourhood
income and SIDS was statistically non-significant after adjustment for
covariates (Table 3). * There was little evidence that hazards were
non-proportional, suggesting that the magnitude of the association
between neighbourhood income and SIDS varied little with age at time of
death.
Between 1991-1995 and 2001-2005, rates of neonatal mortality,
postneonatal mortality and SIDS in urban areas (all neighbourhood income
quintiles combined) declined by 3%, 38% and 57%, respectively.
Changes over time in neonatal, postneonatal and SIDS mortality by
neighbourhood income quintiles are shown in Figures 1-3. The decrease in
rates of postneonatal mortality and SIDS occurred across all income
quintiles. Most of the decrease for SIDS took place between 1991-1995
and 1996-2000, while the decrease for postneonatal mortality continued
from 1996-2000 to 2001-2005. Over the entire study period, postneonatal
mortality rates declined by 35% in the poorest quintile and 39% in the
richest quintile. For SIDS, the poorest quintile experienced a 56%
decline and the richest quintile a 67% decline, but the absolute
decrease in rates was greatest in the poorest quintile.
DISCUSSION
We found that socio-economic inequalities in infant mortality,
measured using neighbourhood income, have narrowed in Canada over a
15-year period but have not disappeared. However, all income quintiles
experienced a decline in infant mortality. Unlike previous studies,
(6,7) the association between neighbourhood income and infant mortality
in our study remained statistically significant after adjustment for
covariates, most likely due to the greater statistical power of our
larger dataset.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Reasons why low socio-economic status is associated with infant
mortality need further investigation, but higher prevalence of risk
factors associated with infant death is likely involved. In Canada,
mothers who do not complete high school have significantly higher rates
of smoking and alcohol use during pregnancy and lower rates of
breastfeeding than women with higher levels of education. (9) In
addition, lower maternal education is associated with higher risks of
preterm (18) and small-for-gestational-age birth. (19) These
inequalities in outcomes were stable or increased over time. (18,19)
In the poorest neighbourhood income quintile, aHRs were higher for
SIDS than for neonatal and overall postneonatal death, indicating that
socio-economic gradients in relative risks are more important for SIDS
than for other types of infant death. The high hazard ratio of SIDS for
lower socio-economic strata is consistent with observations from the
United Kingdom, where both unemployment and area deprivation were
associated with a higher risk of SIDS. (20,21) In Canada, the proportion
of mothers placing their infants on their back to sleep (the safest
position for SIDS prevention) (22) increases with higher maternal
education. (23) Maternal smoking during pregnancy and absence of
breastfeeding vary by socio-economic status and are also known to be
risk factors for SIDS. (24,25)
As shown in a previous study, (10) changes in the coding of cause
of death are unlikely to have substantially affected SIDS counts.
This study has limitations. We did not have individual-level data
on income, education and employment, or other potentially relevant
characteristics such as smoking. Neighbourhood income was assigned using
postal codes and non-differential misclassification errors in the
attribution of neighbourhoods may have attenuated the results,
especially in small town and rural areas. (26) Neighbourhoods were
assigned at time of birth, which may not reflect true residential
exposures if mothers had moved either during or after pregnancy. We
could not adjust for clustering of births within mothers. Finally,
neighbourhood income quintile is only one measure of socio-economic
status; we do not know if trends would have been different for other
markers of area socio-economic status.
In conclusion, our study shows that despite a decrease in rates of
infant mortality and SIDS across all socio-economic strata over time,
socio-economic inequalities in those rates have persisted in Canada.
This finding highlights the need for effective infant health promotion
strategies in vulnerable populations.
Acknowledgements: The authors are grateful to Statistics Canada and
to provincial and territorial vital statistics registrars for providing
access to the births and deaths files, and to Juan Andres Leon and
Michael Tjepkema for reviewing and commenting on an earlier version of
this article. Conflict of Interest: None to declare.
REFERENCES
(1.) Health Canada. Canadian Perinatal Health Report 2003. 2003.
(2.) Public Health Agency of Canada. Perinatal Health Indicators
for Canada 2011. 2012.
(3.) Arntzen A, Samuelsen SO, Bakketeig LS, Stoltenberg C.
Socioeconomic status and risk of infant death. A population-based study
of trends in Norway, 1967-1998. Int J Epidemiol 2004;33(2):279-88.
(4.) Oakley L, Maconochie N, Doyle P, Dattani N, Moser K.
Multivariate analysis of infant death in England and Wales in 2005-06,
with focus on socioeconomic status and deprivation. Health Stat Q
2009;42:22-39.
(5.) Joseph KS, Liston RM, Dodds L, Dahlgren L, Allen AC.
Socioeconomic status and perinatal outcomes in a setting with universal
access to essential health care services. CMAJ 2007;177:583-90.
(6.) Luo ZC, Wilkins R, Kramer MS. Effect of neighbourhood income
and maternal education on birth outcomes: A population-based study.
CMAJ2006;174:1415-20.
(7.) Luo ZC, Kierans WJ, Wilkins R, Liston RM, Mohamed J, Kramer
MS. Disparities in birth outcomes by neighborhood income: Temporal
trends in rural and urban areas, British Columbia. Epidemiol
2004;15:679-86.
(8.) Injury Prevention Committee, Canadian Paediatric Society.
Reducing the risk of sudden infant death. Paediatr Child Health
1996;1:63-67.
(9.) Rusen ID, Liu S, Sauve R, Joseph KS, Kramer MS. Sudden infant
death syndrome in Canada: Trends in rates and risk factors, 1985-1998.
Chronic Dis Can 2004;25:1-6.
(10.) Gilbert NL, Fell DB, Joseph KS, Liu S, Leon JA, Sauve R.
Temporal trends in sudden infant death syndrome in Canada from 1991 to
2005: Contribution of changes in cause of death assignment practices and
in maternal and infant characteristics. Paediatr Perinat Epidemiol
2012;26:124-30.
(11.) Fair M, Cyr M, Allen A, Wen SW, Guyon G, MacDonald RC. An
assessment of the validity of a computer system for probabilistic record
linkage of birth and infant death records in Canada. The Fetal and
Infant Health Study Group. ChronicDis Can 2000;21:8-13.
(12.) Public Health Agency of Canada. Canadian Perinatal Health
Report. 2008 Edition. 2008.
(13.) Wilkins R, Khan S. PCCF+ Version 5J User's Guide.
Automated geographic coding based on the Statistics Canada Postal Code
Conversion files, including postal codes through May 2011. Report
82F0086-XDB. Ottawa: Statistics Canada, 2011.
(14.) du Plessis V, Beshiri R, Bollman RD, Clemenson H. Definition
of Rural. Rural and Small Town Canada Analysis Bulletin. Vol. 3, No. 3.
Report 21-006-XIE. Ottawa: Statistics Canada, 2001.
(15.) Wilcox AJ, Weinberg CR, Basso O. On the pitfalls of adjusting
for gestational age at birth. Am J Epidemiol 2011;174(9):1062-68.
(16.) Lin DY, Wei LJ. The robust inference for the Cox proportional
hazards model. J Am Stat Assoc 1989;84:1074-78.
(17.) Allison PD. Survival Analysis Using SAS[R]: A Practical
Guide, Second Edition. Cary, NC: SAS Institute Inc., 2010.
(18.) Auger N, Roncarolo F, Harper S. Increasing educational
inequality in preterm birth in Quebec, Canada, 1981-2006. J Epidemiol
Community Health 2011;65:1091-96.
(19.) Auger N, Park AL, Harper S, Daniel M, Roncarolo F, Platt RW.
Educational inequalities in preterm and term small-for-gestational-age
birth over time. Ann Epidemiol 2012;22:160-67.
(20.) Leach CE, Blair PS, Fleming PJ, Smith IJ, Platt MW, Berry PJ,
et al. Epidemiology of SIDS and explained sudden infant deaths.
Pediatrics 1999;104:e43.
(21.) Wood AM, Pasupathy D, Pell JP, Fleming M, Smith GC. Trends in
socioeconomic inequalities in risk of sudden infant death syndrome,
other causes of infant mortality, and stillbirth in Scotland: Population
based study. BMJ 2012;344:e1552.
(22.) Fleming PJ, Blair PS, Bacon C, Bensley D, Smith I, Taylor E,
et al. Environment of infants during sleep and risk of the sudden infant
death syndrome: Results of 1993-5 case-control study for confidential
inquiry into stillbirths and deaths in infancy. BMJ 1996;313:191-95.
(23.) Public Health Agency of Canada. What Mothers Say: The
Canadian Maternity Experiences Survey. 2009.
(24.) Mitchell EA, Milerad J. Smoking and the sudden infant death
syndrome. Rev Environ Health 2006;21(2):81-103.
(25.) Hauck FR, Thompson JM, Tanabe KO, Moon RY, Vennemann MM.
Breastfeeding and reduced risk of sudden infant death syndrome: A
meta-analysis. Pediatrics 2011;128(1):103-10.
(26.) Wilkins R. Neighbourhood income quintiles derived from
Canadian postal codes are apt to be misclassified in rural but not urban
areas. Ottawa: Statistics Canada, 2004.
Received: November 8, 2012 Accepted: February 28, 2013
* For more complete data for Tables 2 and 3, please see
Supplemental Appendix A and Supplemental Appendix B, respectively.
Nicolas L. Gilbert, MSc, [1,2] Nathalie Auger, MD, MSc, [2,3]
Russell Wilkins, MUrb, [4,5] Michael S. Kramer, MD [6]
Author Affiliations
1. Maternal and Infant Health Section, Public Health Agency of
Canada, Ottawa, ON
2. Department of Social and Preventive Medicine, Universite de
Montreal, Montreal, QC
3. Institut national de sante publique du Quebec, Montreal, QC
4. Health Analysis Division, Statistics Canada, Ottawa, ON
5. Department of Epidemiology and Community Medicine, University of
Ottawa, Ottawa, ON
6. Departments of Pediatrics and of Epidemiology, Biostatistics and
Occupational Health, McGill University, Montreal, QC
Correspondence: Nicolas Gilbert, Maternal and Infant Health
Section, Public Health Agency of Canada, 785 Carling Avenue 6804A,
Ottawa, ON K1A 0K9, Tel: 613-960-0434, E-mail:
nicolas.gilbert@phac-aspc.gc.ca
Table 1. Infant Mortality in Canada (Excluding Ontario
and Territories), 1991-2005
Live Births Neonatal Mortality
Rate per Unadjusted Hazard
1000 Ratio (95% CI)
Live
Births
Total 3,173,694 3.71
Neighbourhood
income
quintile
1 (poorest) 753,143 4.14 1.22 (1.16-1.30)
2 650,605 3.72 1.10 (1.03-1.17)
3 621,244 3.69 1.09 (1.02-1.16)
4 607,563 3.46 1.02 (0.96-1.09)
5 (richest) 541,139 3.38 Reference
Area of
residence
Urban 2,414,309 3.56 Reference
Rural and 759,385 4.17 1.17 (1.12-1.22)
small
town
Year of birth
1991-1995 1,144,957 3.84 Reference
1996-2000 1,031,217 3.59 0.93 (0.89-0.97)
2001-2005 997,520 3.68 0.96 (0.91-1.00)
Postneonatal Mortality
Rate per Unadjusted Hazard
1000 Ratio (95% CI)
Neonatal
Survivors
Total 1.70
Neighbourhood
income
quintile
1 (poorest) 2.54 1.95 (1.79-2.14)
2 1.64 1.26 (1.14-1.39)
3 1.46 1.13 (1.02-1.25)
4 1.33 1.02 (0.92-1.14)
5 (richest) 1.3 Reference
Area of
residence
Urban 1.53 Reference
Rural and 2.25 1.48 (1.39-1.56)
small
town
Year of birth
1991-1995 2.10 Reference
1996-2000 1.62 0.77 (0.72-0.82)
2001-2005 1.33 0.63 (0.59-0.68)
SIDS
Rate per Unadjusted Hazard
1000 Ratio (95% CI)
Live
Births
Total 0.53
Neighbourhood
income
quintile
1 (poorest) 0.86 2.53 (2.16-2.96)
2 0.53 1.56 (1.31-1.86)
3 0.44 1.29 (1.07-1.55)
4 0.35 1.02 (0.84-1.25)
5 (richest) 0.34 Reference
Area of
residence
Urban 0.47 Reference
Rural and 0.69 1.46 (1.32-1.62)
small
town
Year of birth
1991-1995 0.75 Reference
1996-2000 0.46 0.61 (0.55-0.68)
2001-2005 0.33 0.44 (0.38-0.49)
Table 2. Infant Mortality in Urban Areas of Canada (Excluding Ontario
and Territories), 1991-2005
Live Neonatal Mortality
Births
Rate Unadjusted Hazard
per Ratio (95% CI)
1000
Live
Births
Total 2,414,309 3.56
Neighbourhood
income quintile
1 (poorest) 569,279 4.04 1.25 (1.16-1.34)
2 501,546 3.65 1.13 (1.04-1.22)
3 476,071 3.5 1.08 (1.00-1.17)
4 465,752 3.22 0.99 (0.92-1.08)
5 (richest) 401,661 3.24 Reference
Year of birth
1991-1995 861,835 3.69 Reference
1996-2000 781,476 3.42 0.93 (0.88-0.98)
2001-2005 770,998 3.56 0.97 (0.92-1.02)
Neonatal Postneonatal Mortality
Mortality
Adjusted Hazard Rate Unadjusted Hazard
Ratio * (95% CI) per Ratio (95% CI)
1000
Neonatal
Survivors
Total 1.53
Neighbourhood
income quintile
1 (poorest) 1.24 (1.15-1.34) 2.24 1.96 (1.76-2.19)
2 1.14 (1.05-1.24) 1.55 1.36 (1.21-1.53)
3 1.09 (1.01-1.18) 1.34 1.18 (1.04-1.33)
4 1.01 (0.93-1.09) 1.15 1.01 (0.89-1.14)
5 (richest) Reference 1.14 Reference
Year of birth
1991-1995 Reference 1.92 Reference
1996-2000 0.90 (0.85-0.95) 1.41 0.73 (0.68-0.79)
2001-2005 0.92 (0.87-0.97) 1.20 0.62 (0.58-0.68)
Postneonatal SIDS
Mortality
Adjusted Hazard Rate
Ratio * (95% CI) per
1000
Live
Births
Total 0.47
Neighbourhood
income quintile
1 (poorest) 1.58 (1.41-1.76) 0.8
2 1.20 (1.07-1.35) 0.5
3 1.07 (0.94-1.20) 0.42
4 0.96 (0.84-1.08) 0.26
5 (richest) Reference 0.29
Year of birth
1991-1995 Reference 0.70
1996-2000 0.74 (0.69-0.80) 0.40
2001-2005 0.65 (0.60-0.70) 0.30
SIDS
Unadjusted Hazard Adjusted Hazard
Ratio (95% CI) Ratio* (95% CI)
Total
Neighbourhood
income quintile
1 (poorest) 2.76 (2.26-3.37) 1.83 (1.49-2.26)
2 1.71 (1.38-2.12) 1.32 (1.06-1.65)
3 1.43 (1.13-1.80) 1.17 (0.92-1.48)
4 0.89 (0.69-1.15) 0.80 (0.62-1.04)
5 (richest) Reference Reference
Year of birth
1991-1995 Reference Reference
1996-2000 0.57 (0.49-0.65) 0.58 (0.50-0.67)
2001-2005 0.43 (0.37-0.50) 0.46 (0.40-0.54)
* Adjusted for neighbourhood income quintile, province of
residence at birth, year of birth (5-year interval),
parity, sex, and multiple birth.
Table 3. Infant Mortality in Rural Areas of Canada
(Excluding Ontario and Territories), 1991-2005
Live Neonatal
Births Mortality
Rate Unadjusted Hazard
per Ratio (95% CI)
1000
Live
Births
Total 759,385 4.17
Neighbourhood
income quintile
1 (poorest) 183,864 4.45 1.17 (1.05-1.31)
2 149,059 3.93 1.03 (0.92-1.16)
3 145,173 4.31 1.14 (1.01-1.27)
4 141,811 4.27 1.12 (1.00-1.26)
5 (richest) 139,478 3.80 Reference
Year of birth
1991-1995 283,122 4.32 Reference
1996-2000 249,741 4.10 0.95 (0.87-1.03)
2001-2005 226,522 4.06 0.94 (0.86-1.02)
Neonatal Mortality Postneonatal Mortality
Adjusted Hazard Rate Unadjusted Hazard
Ratio * (95% CI) per Ratio (95% CI)
1000
Neonatal
Survivors
Total 2.25
Neighbourhood
income quintile
1 (poorest) 1.14 (1.02-1.28) 3.47 1.98 (1.68-2.33)
2 1.03 (0.92-1.16) 1.92 1.09 (0.91-1.31)
3 1.15 (1.02-1.29) 1.87 1.06 (0.89-1.28)
4 1.13 (1.01-1.27) 1.92 1.09 (0.91-1.31)
5 (richest) Reference 1.76 Reference
Year of birth
1991-1995 Reference 2.63 Reference
1996-2000 0.92 (0.85-1.00) 2.26 0.86 (0.77-0.96)
2001-2005 0.90 (0.82-0.98) 1.78 0.68 (0.60-0.77)
Postneonatal SIDS
Mortality
Adjusted Hazard Rate
Ratio * (95% CI) per
1000
Live
Births
Total 0.69
Neighbourhood
income quintile
1 (poorest) 1.54 (1.31-1.82) 1.05
2 1.02 (0.85-1.23) 0.65
3 1.02 (0.85-1.23) 0.52
4 1.08 (0.90-1.29) 0.64
5 (richest) Reference 0.49
Year of birth
1991-1995 Reference 0.93
1996-2000 0.87 (0.77-0.98) 0.67
2001-2005 0.70 (0.61-0.79) 0.42
SIDS
Unadjusted Hazard Adjusted Hazard
Ratio (95% CI) Ratio ([dagger])
(95% CI)
Total
Neighbourhood
income quintile
1 (poorest) 2.16 (1.66-2.79) 1.54 (1.19-2.00)
2 1.34 (0.98-1.82) 1.19 (0.87-1.64)
3 1.07 (0.78-1.48) 1.00 (0.72-1.37)
4 1.32 (0.96-1.80) 1.28 (0.93-1.75)
5 (richest) Reference Reference
Year of birth
1991-1995 Reference Reference
1996-2000 0.72 (0.59-0.88) 0.74 (0.61-0.90)
2001-2005 0.46 (0.36-0.58) 0.46 (0.36-0.59)
* Adjusted for neighbourhood income quintile, province of residence
at birth, year of birth (5-year interval), parity, sex, and multiple
birth.
([dagger]) Adjusted for neighbourhood income quintile, province of
residence at birth, year of birth (5-year interval), parity, and sex.