Insights on Canadian Society
Association between breastfeeding and select chronic conditions among off-reserve First Nations, Métis and Inuit children in Canada

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

by Nadine Badets, Tamara Hudon and Michael Wendt

Release date: March 20, 2017

Start of text box

Overview of the study

This paper examines associations between breastfeeding and select chronic conditions—asthma/chronic bronchitis and chronic ear infections—among off-reserve First Nations, Métis and Inuit children in Canada aged 1 to 5 years. Data are from the 2006 Aboriginal Children’s Survey, and each Aboriginal group was studied separately. Two aspects of breastfeeding are examined: feeding history (e.g. bottle-fed, breastfed, or both) and duration of breastfeeding.

  • About 60% of off-reserve First Nations, Métis and Inuit children aged 1 to 5 were both bottle-fed and breastfed. The proportion of those who were only breastfed varied between 14% and 17%, while the proportion of those who were only bottle-fed varied between 23% and 25%.
  • Breastfed children were found to have a lower prevalence of asthma/chronic bronchitis. Among off-reserve First Nations children, for instance, 14% of those who were only bottle-fed had asthma/chronic bronchitis, compared with 10% of those who were only breastfed.
  • Among off-reserve First Nations children, those who were breastfed had a lower prevalence of chronic ear infections. Of those who were only bottle-fed, 5%E reported chronic ear infections compared with 2%E of those who were only breastfed.
  • Off-reserve First Nations children who were breastfed for more than six months had a lower prevalence of asthma/chronic bronchitis and chronic ear infections than their counterparts who were never breastfed.
  • The findings described above remained significant even when clinical, demographic, and socio-economic factors were taken into account in a model. Results indicate that factors other than breastfeeding are also associated with health outcomes.

End of text box

Introduction

Past research has shown breastfeeding to be protective for infant and child health.Note 1 Human milk contains biologically active substances that stimulate and strengthen the infant immune system with antimicrobial and anti-inflammatory agents, unlike infant feeding formula.Note 2 In addition, the functional physiology of infants who are breastfed is different from infants who are bottle-fed due to pressure and suck, which prevents the pooling of milk in the middle ear and the consequent proliferation of bacteria.Note 3 Health Canada recommends that infants should be exclusively breastfed for the first six months of life to ensure the best nutrition for the optimal health, growth and development of infants and toddlers.Note 4

Previous literature on infant and child health has suggested that breastfeeding is protective against several negative health outcomes.Note 5 In particular, research specific to Aboriginal infants and children in Canada has found that breastfeeding is related to better general health and a lower prevalence of specific health problems such as asthma, respiratory tract infections, ear infections (otitis) and gastrointestinal infections.Note 6

That said, previous research has often focused on a single Aboriginal identity group (sometimes within small levels of geography) or has combined data for multiple Aboriginal groups. This study uses data from the 2006 Aboriginal Children’s Survey to examine associations between breastfeeding and select chronic conditions—asthma/chronic bronchitis and chronic ear infections—for each of three Aboriginal identity groups. The sample was restricted to off-reserve First Nations, Métis and Inuit children, aged 1 to 5 years,Note 7 whose birth mother responded to the survey (see Data sources, methods and definitions).

Exclusive breastfeeding, as defined by Health Canada, is when an infant is only fed breastmilk—no other liquids (not even water) or food.Note 8 However, it was not possible to create a variable with the data based on this definition as it was not possible to determine whether children in the study received solid foods as infants or not. Thus, two distinct breastfeeding variables were created to test for differences between breastfeeding and bottle-feeding, and to test for differences by duration of breastfeeding. Neither should be confused nor conflated with exclusive breastfeeding (i.e., without solid food or any other liquid), and it should be noted that children who were only bottle-fed/never breastfed may have been fed breastmilk in their bottles.

The first section of this article uses descriptive statistics to examine the distributions of select demographic, breastfeeding and health characteristics of off-reserve First Nations, Métis and Inuit children aged 1 to 5. The following sections present results of the descriptive and logistic regression analyses for two outcome variables (asthma/chronic bronchitis and chronic ear infections) by each breastfeeding variable (feeding history and breastfeeding duration). The purpose of the logistic regression analysis is to determine whether associations between breastfeeding and health outcomes remain after the influence of several demographic, social, economic and health characteristics are controlled for.

Among off-reserve First Nations, Métis and Inuit children aged 1 to 5, 6 in 10 were both breastfed and bottle-fed

With the 2006 Aboriginal Children’s Survey, it is possible to derive proportions of Aboriginal children aged 1 to 5 who were only breastfed, only bottle-fed, or who were both breastfed and bottle-fed.Note 9 The majority of off-reserve First Nations (61%), Métis (60%), and Inuit (60%) children, aged 1 to 5, were both breastfed and bottle-fed (Table 1). Inuit children were most likely to have been only breastfed (not bottle-fed), which was the case for 17% of Inuit children, versus 14% for off-reserve First Nations children and 15% for Métis children. Conversely, almost one‑quarter of off-reserve First Nations, Métis and Inuit children were only bottle-fed.

Table 1
Distribution of children aged 1 to 5 by selected demographic, breastfeeding and health characteristics, by Aboriginal identity, 2006
Table summary
This table displays the results of Distribution of children aged 1 to 5 by selected demographic Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using percent units of measure (appearing as column headers).
Aboriginal identity
Off-reserve First Nations Métis Inuit
percent
Sex
Females 50.1 47.8 49.3
Males 49.8 52.1 50.6
Age group
1 to 3 years 59.6 58.1 58.6
4 to 5 years 40.3 41.8 41.3
Feeding history
Only breastfed 14.4 14.9 17.2
Both breastfed and bottle-fed 61.4 60.2 59.8
Only bottle-fedTable 1 Note 1 24.0 24.8 22.8
Breastfeeding duration
Never breastfedTable 1 Note 2 24.8 25.2 25.0
Breastfed 0 to 6 months 41.1 42.4 31.6
Breastfed more than 6 months 33.9 32.3 43.3
Asthma/chronic bronchitis
Yes 11.2 11.1 8.1
No 88.7 88.8 91.8
Ear infections
At least one since birth 54.3 55.0 50.7
ChronicTable 1 Note 3 4.0 3.7 8.7
None 45.6 44.9 49.2

More Inuit children were breastfed for more than six months than off-reserve First Nations or Métis children. A larger proportion of off-reserve First Nations (41%) and Métis (42%) children were breastfed from zero to six months, whereas a larger proportion of Inuit children (43%) were breastfed for more than six months. About one‑quarter of children in each Aboriginal identity group were not breastfed. It is important to note that cultural practices among Inuit, such as traditional adoption, play a role in breastfeeding initiation and practices,Note 10 and that although Inuit generally have lower breastfeeding rates than other Aboriginal groups and the non-Aboriginal population, Inuit children tend to be breastfed for longer periods of time.

In 2006, about 11% of off-reserve First Nations, 11% of Métis and 8% of Inuit children aged 1 to 5 had diagnosed asthma or chronic bronchitis. According to another study based on data from the National Longitudinal Survey of Children and Youth (NLSCY), asthma prevalence in remote and northern communities in Canada is lower for Aboriginal children than it is for non-Aboriginal children.Note 11

Off-reserve First Nations and Métis children who were breastfed had lower rates of asthma/chronic bronchitis than bottle-fed children

Asthma has been named as one of the most common chronic conditions among off-reserve First Nations, Métis and Inuit children in Canada.Note 12 In addition, Aboriginal children suffer from more severe respiratory infections and are more frequently admitted for hospitalization than non-Aboriginal children.Note 13 Previous research on the general population has found that breastfeeding is associated with a lower prevalence of respiratory illnesses, such as asthma and lower respiratory tract infections.Note 14 Research specific to Aboriginal children in Canada has produced similar results.Note 15

Over the next sections, descriptive results will be discussed, as well as the logistic regression analysis, which examined the relationship between breastfeeding and asthma/chronic bronchitis while taking numerous important factors into account. The analysis was conducted separately for each Aboriginal identity group.

Among off-reserve First Nations and Métis children, children who were only breastfed had lower rates of asthma/chronic bronchitis (10% of off-reserve First Nations and 8%E of Métis) than their counterparts who were only bottle-fed (14% and 13%, respectively) (Chart 1). In addition, Métis children who were only breastfed had a lower prevalence of asthma/chronic bronchitis (8%E) than Métis children who were both breastfed and bottle-fed (11%).

Chart 1 Proportion of children aged 1 to 5 with asthma/chronic bronchitis, by Aboriginal identity and feeding history, 2006

Description for Chart 1
Data table for Chart 1
Table summary
This table displays the results of Data table for Chart 1. The information is grouped by Feeding history (appearing as row headers), Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using percent units of measure (appearing as column headers).
Feeding history Aboriginal identity
Off-reserve First Nations Métis Inuit
percent
Only breastfed (ref.) 9.7 7.8Note E: Use with caution Note F: too unreliable to be published
Both breastfed and bottle-fed 10.3 11.4Note * 8.6
Only bottle-fedData table Note 1 13.9Note * 12.5Note * 7.6Note E: Use with caution

Similar results were obtained when a variable indicative of breastfeeding duration was used among off-reserve First Nations children. Specifically, off-reserve First Nations children who were ever breastfed—for up to six months (11%) or for more than six months (10%)—had lower rates of asthma/chronic bronchitis than children who were never breastfed (14%) (Chart 2).

Chart 2 Proportion of children aged 1 to 5 with asthma/chronic bronchitis, by Aboriginal identity and breastfeeding duration, 2006

Description for Chart 2
Data table for Chart 2
Table summary
This table displays the results of Data table for Chart 2. The information is grouped by Breastfeeding duration (appearing as row headers), Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using percent units of measure (appearing as column headers).
Breastfeeding duration Aboriginal identity
Off-reserve First Nations Métis Inuit
percent
Never breastfed (ref.) 14.1 12.5 8.4Note E: Use with caution
Breastfed 0 to 6 months 10.5Note * 11.4 10.1Note E: Use with caution
Breastfed more than 6 months 10.3Note * 10.1 7.5Note E: Use with caution

Off-reserve First Nations children who were breastfed had a lower likelihood of asthma/chronic bronchitis after several factors were controlled for

After several clinical, demographic and socioeconomic characteristics were controlled for in a logistic regression model, the results showed that breastfeeding was associated with a lower likelihood of asthma/chronic bronchitis among off-reserve First Nations children. First Nations children living off reserve who were only breastfed were less likely to have asthma/chronic bronchitis (9%) than their counterparts who were only bottle-fed (14%). There was no significant relationship between breastfeeding and the presence of asthma/chronic bronchitis among Métis or Inuit children (Table 2).

Table 2
Predicted probability of asthma/chronic bronchitis for two breastfeeding variables, by Aboriginal identity for children aged 1 to 5, 2006
Table summary
This table displays the results of Predicted probability of asthma/chronic bronchitis for two breastfeeding variables Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using predicted probability units of measure (appearing as column headers).
Aboriginal identity
Off-reserve First Nations Métis Inuit
predicted probability
Feeding history
Only breastfed (ref.) 0.09 0.08 0.06
Both breastfed and bottle-fed 0.11 0.12 0.09
Only bottle-fedTable 2 Note 1 0.14Note * 0.12 0.08
Breastfeeding duration
Never breastfed (ref.) 0.14 0.12 0.09
Breastfed 0 to 6 months 0.11 0.12 0.09
Breastfed more than 6 months 0.10Note * 0.10 0.09

The breastfeeding duration variable also indicated that children who had never been breastfed had a higher likelihood of asthma/chronic bronchitis among First Nations children living off reserve. Children who were breastfed up to six months (11%) and who were breastfed for more than six months (10%) were less likely to have asthma/chronic bronchitis than children who were never breastfed (14%).Note 16

Model results, however, also show that other factors were associated with a lower prevalence of asthma. Having respiratory allergies or being diagnosed with tuberculosis, for instance, was significantly associated with a higher probability of asthma/chronic bronchitis in all three groups of Aboriginal children. Similarly, a higher prevalence of asthma/chronic bronchitis was found among off-reserve First Nations and Inuit children whose mothers had difficulty accessing health care. Full model results are available in tables A1 (with the feeding history variable) and A2 (with the breastfeeding duration variable) in the "Supplementary information" section.

Off-reserve First Nations children who were only breastfed had a lower prevalence of chronic ear infections

Ear infections are a leading cause of hospitalization among Aboriginal children in Canada.Note 17 More specifically, middle ear infections are a common health problem for young children in Canada,Note 18 which may be associated with a number of health and developmental issues such as hearing impairment, delayed speech and language development, as well as delayed academic and educational development.Note 19 First Nations, Métis and Inuit children in northern Canada are disproportionately affected by middle ear infections, with prevalence rates in some communities around 40 times higher than in the urban South.Note 20

In this study, chronic ear infections were defined as four or more ear infections in the last 12 months. This section presents the results of the descriptive and logistic regression analyses, which examined the relationships between feeding history and breastfeeding duration for chronic ear infections for each of the three Aboriginal identity groups.

Given the relatively small sample size of children who reported chronic ear infections, the results must be interpreted with caution. However, there was a significant difference in the prevalence of chronic ear infections between off-reserve First Nations children who were only breastfed (2%E) and those who were only bottle-fed (5%E) (Chart 3). Off-reserve First Nations children who were breastfed for more than six months had a lower prevalence of chronic ear infections (3%E) than off-reserve First Nations children who were not breastfed (5%E) (data not shown).Note 21

Chart 3 Proportion of children aged 1 to 5 with chronic ear infections, by Aboriginal identity and feeding history, 2006

Description for Chart 3
Data table for Chart 3
Proportion of children aged 1 to 5 with chronic ear infections,Data table Note 1 by Aboriginal identity and feeding history, 2006
Table summary
This table displays the results of Data table for Chart 3 Proportion of children aged 1 to 5 with chronic ear infections. The information is grouped by Feeding history (appearing as row headers), Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using percent units of measure (appearing as column headers).
Feeding history Aboriginal identity
Off-reserve First Nations Métis Inuit
percent
Only breastfed (ref.) 2.3Note E: Use with caution Note F: too unreliable to be published 8.7Note E: Use with caution
Both breastfed and bottle-fed 3.9 3.9 8.9Note E: Use with caution
Only bottle-fedData table Note 2 5.2Note E: Use with cautionNote * 4.1Note E: Use with caution 8.2Note E: Use with caution

Off-reserve First Nations children who were breastfed had a lower prevalence chronic ear infections after several factors were controlled for

After several factors were controlled for in a logistic regression model, First Nations children living off reserve who were only breastfed were significantly less likely to have chronic ear infections (2%) than children who were only bottle-fed (5%) (Table 3). Likewise, off-reserve First Nations children who were breastfed for more than six months were significantly less likely to have chronic ear infections (3%) than their counterparts who were not breastfed (5%). In contrast, no association was found between chronic ear infections and breastfeeding among Métis and Inuit children.

Table 3
Predicted probability of chronic ear infections for two breastfeeding variables, by Aboriginal identity for children aged 1 to 5, 2006
Table summary
This table displays the results of Predicted probability of chronic ear infections for two breastfeeding variables Aboriginal identity, Off-reserve First Nations, Métis and Inuit, calculated using predicted probability units of measure (appearing as column headers).
  Aboriginal identity
Off-reserve First Nations Métis Inuit
predicted probability
Feeding history  
Only breastfed (ref.) 0.02 0.03 0.13
Both breastfed and bottle-fed 0.04 0.04 0.08
Only bottle-fedTable 3 Note 1 0.05Note * 0.04 0.06
Breastfeeding duration  
Never breastfed (ref.) 0.05 0.04 0.07
Breastfed 0 to 6 months 0.04 0.04 0.10
Breastfed more than 6 months 0.03Note * 0.03 0.07

As was the case with the results for asthma/chronic bronchitis, the models also showed that other factors were associated with chronic ear infections. Exposure to second hand smoke at home was associated with a higher prevalence of chronic ear infections among Métis and Inuit children. Among Inuit children, those living outside of Inuit Nunangat had a lower probability of having chronic ear infections than those who lived in Inuit Nunangat.Note 22 Readers interested in the full model results are invited to consult tables A3 and A4 in the "Supplementary information" section.

Conclusion

The findings for First Nations children living off reserve suggest that there is a relationship between breastfeeding and the health outcomes in this study. This relationship continued to be significant after several clinical, demographic and socioeconomic characteristics were controlled for. In addition, children who were breastfed for longer periods of time, such as for more than six months, had lower prevalence and lower likelihood of asthma/chronic bronchitis and chronic ear infections. These results are consistent with past research on breastfeeding and health among the general population, and with research specific to Aboriginal populations.Note 23

However, breastfeeding was not found to be associated with better health outcomes across all Aboriginal populations. Specifically, no association was found between breastfeeding and asthma/chronic bronchitis or between breastfeeding and chronic ear infections among Métis or Inuit children.

The logistic regression results also indicate that there may be other factors with more influence on health outcomes than breastfeeding, such as socioeconomic status and living conditions (full model results are available in the "Supplementary information" section). Past research has also shown that infants living in low income are less likely to have been breastfed than infants from higher socioeconomic backgrounds.Note 24

Overall, these findings have implications for research specific to Aboriginal peoples, and as such, care should be taken not to generalize or apply findings from general population studies or from studies that combine multiple Aboriginal groups to single Aboriginal populations.

Nadine Badets and Tamara Hudon are research analysts in the Social and Aboriginal Statistics Division at Statistics Canada, and Michael Wendt is chief of the Social and Aboriginal Analysis section in the same division.

Start of text box

Data sources, methods and definitions

Data source

The Aboriginal Children’s Survey (ACS) is a postcensal survey that was conducted once—by Statistics Canada in 2006. The target population was all children in Canada under the age of 6 (as of October 31, 2006),Note 25 with First Nations (North American Indian), Métis or Inuit identity or ancestry. However, a small number of children were 6 years of age by the time of the survey. Children living in institutions and children living on reserve in the provinces were excluded. The final sample size was 12,845.

Data from this survey are weighted with respect to the child, unlike other data sources such as the Aboriginal Peoples Survey (APS) and the Canadian Community Health Survey (CCHS), which are weighted with respect to the mother. As such, breastfeeding prevalence is not measured in this study. For information related to breastfeeding prevalence among Aboriginal and non-Aboriginal people, consult CANSIM tables 105‑0512 and 105‑0515 on Statistics Canada’s website.

Methods

This study’s results were produced using descriptive and logistic regression analyses. Variances were estimated using bootstrap weightsNote 26 available from the Aboriginal Children’s Survey data files.Note 27 Four breastfeeding variables were tested separately in the models: ever breastfed versus never breastfed; feeding history; and two versions of breastfeeding duration (breastfed up to six months or more, and breastfed up to 12 months or more). Although feeding history and breastfeeding duration at six months were selected as the main breastfeeding variables, results are similar for all four breastfeeding variables.

Sample

The sample consists of off-reserve First Nations, Métis and Inuit children in Canada aged 1 to 5 years. Children with Aboriginal identity were selected, and only children with single responses for First Nations, Métis or Inuit categories. The sample was also restricted to children whose birth mother responded to the survey. Birth mothers were selected as respondents not only for consistency with previous research, but also because they are the most likely to have the best knowledge and recall of the child’s health and feeding history.Note 28 Given the small sample sizes in some cases, results that must be interpreted with caution have been marked with an ‘E’ and results that are too unreliable to be published have been marked with an ‘F’.

Definitions

The variables used in this study are based on the literature and previous studies, but are limited by data available in the 2006 Aboriginal Children’s Survey (ACS). Four types of independent variables were used: clinical, demographic, socioeconomic, and other variables. Clinical variables were included to control for the risk of illness due to co-morbid conditions. For example, ear infections have been found to be significantly associated with the risk of asthma and wheezing (for children who have not received an asthma diagnosis).Note 29

Multiple definitions of breastfeeding and breastfeeding duration were tested in this study. The variable that is closest to Health Canada’s definition of exclusive breastfeeding is feeding history, which examines children who were only breastfed, only bottle-fed, or who were both breastfed and bottle-fed. Please note that children who were only bottle-fed/never breastfed may have been fed breastmilk in their bottles.

Asthma/chronic bronchitis was diagnosed by a health professional, as reported by the child’s mother.

Chronic ear infections denotes four or more ear infections in the past 12 months, as reported by the mother.

Limitations

The analyses in this study are limited by the variables collected in the survey, as the 2006 Aboriginal Children’s Survey (ACS) was not conducted on reserve. Although the literature on asthma has shown that the presence of dampness and mould have an effect on the prevalence of asthma in children,Note 30 data on indoor air quality was not available through the survey. Furthermore, the ACS does not include information on other predisposing health conditions that could be related to the health outcomes examined in this study, such as respiratory tract infections.

Sample size was also an issue in this study. Power analysis indicated that the models for Inuit generally had small sample sizes. However, when the models were run for all types of respondents, the results did not change even though the sample size for Inuit had increased and was much closer to the desired size. Bivariate statistical tests between each breastfeeding variable and each outcome produced similar results to those from the multivariate logistic regression.

Cultural variables such as the birth mother’s primary language (Aboriginal language versus non-Aboriginal); the importance of First Nations, Métis and Inuit history and culture to the mother; and the frequency at which the child participated in traditional activities were tested throughout the model-building process. However, these variables were not significantly associated with the health outcomes in the study. It was also unclear whether these variables serve as adequate proxies for cultural involvement, thus they were excluded from analysis.

End of text box

Related information

Supplementary information

The tables in this section show the full logistic regression model results used for tables 2 and 3 in this article.

Data sources

Bibliographic references

Date modified: