Health Reports
Family networks and health among Métis aged 45 or older

by Pamela L. Ramage-Morin and Evelyne Bougie

Release date: December 20, 2017

The importance of social networks for promoting health and well-being is well-established.Note 1Note 2Note 3Note 4Note 5Note 6 These networks, which include family and friendship ties, provide emotional, practical and other support; enhance social participation and engagement; and help combat isolation and loneliness.Note 7Note 8

Interpersonal relationships are a component of successful aging.Note 9 Transitions such as retirement, relocation, diminishing health, and the death of family members and friends can change social networks.Note 9Note 10Note 11Note 12 The National Senior’s Council identified Aboriginal seniors as one of the groups most at risk of social isolation.Note 11Note 13

Métis comprise about a third of the Aboriginal population in Canada. Although their numbers are growing rapidly,Note 13 Métis-specific research is relatively scarce.Note 14Note 15 While the health benefits of social networks have been investigated for decades,Note 4Note 6Note 16 little information is available about networks among this Aboriginal group.Note 15Note 17Note 18Note 19Note 20 Métis history has been shaped by losses of language, culture, and access to the land,Note 15Note 18Note 19Note 20Note 21 yet at the same time, Métis history is one of resilience, with evidence of the durability of family bonds and social networks.Note 15Note 22

Using a population-based sample from the 2012 Aboriginal Peoples Survey, this analysis examines associations between family networks and self-perceived health among Métis aged 45 or older.


Data source

The 2012 Aboriginal Peoples Survey (APS) is a national survey of First Nations people living off reserve, Métis, and Inuit who, as of February 1, 2012, were aged 6 or older and lived in private dwellings. Residents of Indian reserves and settlements and certain First Nations communities in Yukon and the Northwest Territories were excluded.

The 2012 APS sample was drawn from respondents to the 2011 National Household Survey who reported either Aboriginal identity or Aboriginal ancestry. "Aboriginal identity" refers to people who reported: that they identified with at least one Aboriginal group, namely, First Nations (North American Indian), Métis or Inuit; being a Status Indian (Registered Indian or Treaty Indian, as defined by the Indian Act of Canada); or being a member of a First Nation or Indian band. The response rate to the 2012 APS was 76%, with a final sample of 28,410, of whom 11,270 (40%) reported Métis identity.

Study sample

In this study, Métis were defined as APS respondents who self-identified as “Métis” with a single identity. The study sample consisted of 2,142 respondents aged 45 or older, representing a population of 143,000 Métis, with 47% men, and an average age of 57. The analyses excluded 34 respondents who self-identified as Métis among other Aboriginal identities, and 110 proxy respondents, because they did not complete the APS questions on self-perceived mental health.


Two APS questions were used to create an indicator of strong family networks:

  1. “On a scale of 1 to 5, with 1 being very weak and 5 being very strong, how strong are the ties among members of your family living in your city, town or community, but in another household?” Respondents without family members living in another household in their city, town or community scored “0.”
  2. “Who would you turn to for support in times of need?”

Respondents were classified as having a strong family network if they reported very strong ties among family members identified in question 1, and in question 2, at least one relative (spouse, child aged 15 or older, parent, sibling, other relative) to whom they could turn in times of need. That relative may or may not have resided in the same household or community (Appendix Table A). The variables derived from these questions were moderately correlated (Cramer’s V 0.30).

It was hypothesized that knowledge of an Aboriginal language would be associated with stronger family networks. Even though relatively few Métis (2.5%) can converse in an Aboriginal language,Note 23 this indicator is important. Knowledge of traditional languages is regarded as a measure of cultural continuity.Note 24Note 25Note 26 Loss of traditional languages has contributed to the isolation of individuals from their families and communities,Note 20 the impact of which has reverberated through successive generations.Note 27Note 28 APS respondents were asked if they spoke and/or understood (even if only a few words) any Aboriginal language(s), such as Cree, Ojibway, Inuktitut, Mi’kmaq, Michif, or Dene. Those who replied affirmatively could specify up to three languages; identified the language they spoke/understood best; and rated their ability (very well, relatively well, with effort, or only a few words). Respondents who could speak and/or understand one or more Aboriginal languages very well or relatively well were classified as having current knowledge of Aboriginal language versus those with minimal or no knowledge.

Self-perceived general and mental health were based on answers to: “In general, would you say your health/mental health is: ....” Those who responded excellent, very good, or good (versus fair or poor) were considered to have positive self-perceived general health/mental health.

Analytical techniques

Men and women were analyzed separately to account for potential gender differences in the relationship between family networks and health. Multivariate analyses controlled for age, residence (rural versus population centre), socioeconomic characteristics (education and labour force status), and Aboriginal language knowledge.

Weighted cross-tabulations were used to examine characteristics of Métis men and women with strong family networks and associations with self-perceived general and mental health. Separate multivariate models for the general and mental health outcomes examined the odds of reporting positive general health/mental health by the strength of family networks when taking account of the covariates in the bivariate analyses. Age groups were used in the cross-tabulations; multiple logistic regression models controlled for age (continuous) and age-squared because of the non-linear relationship between self-perceived health outcomes and age. Variance inflation factors (≤ 2.9) and tolerance estimates (≥ 0.2) demonstrated that multicollinearity was not a problem for the models. Model-adjusted prevalence estimates for those aged 45 to 64 were calculated to further examine relationships between family networks and self-perceived health by labour force participation.

Data were weighted to represent census counts of Métis aged 45 or older.

The bootstrap technique was used to estimate coefficients of variation and p-values and perform significance tests. SAS-callable SUDAAN was used for the analyses with a Fay adjustment factor to account for the number of bootstrap samples used to produce each mean bootstrap variable.Note 29


Strong family networks

An estimated 85% of Métis aged 45 or older had a family member to turn to in times of need (Appendix Table A). Men were more likely than women to identify their spouse as a support person (61% versus 46%), while women were more likely than men to report a child or other relative (Figure 1). Two-thirds of men (66%) identified just one relative, compared with 58% of women. Women were almost twice as likely (15% versus 8%) as men to name three to six family members as sources of support.

More than half (54%) of Métis aged 45 or older had strong family networks; that is, they had at least one relative to turn to for support and reported very strong ties among family members in the same community (Table 1). Women (60%) were more likely than men (48%) to have strong family networks, especially at age 65 or older (78% versus 53%).

Just over a quarter (28%) of Métis did not have a secondary school diploma (Appendix Table B). Women without a diploma were more likely than those with postsecondary education to have strong family networks. Over half (57%) of Métis women participated in the labour force—they were less likely than non-participants to report strong networks.

For men, neither education nor labour force participation was associated with strong family networks.

General and mental health

Among Métis aged 45 or older, 75% of men and 71% of women had positive self-perceived general health, and 93% of men and 88% of women reported positive mental health (Table 2). People with lower levels of education were generally less likely than those with postsecondary education to have positive self-perceived health. Labour force participants were more likely than non-participants to report positive general and mental health.

Strong family networks and health

Among Métis men with strong family networks, 81% had positive self-perceived general health, compared with 69% of those with weaker networks (Table 2). As well, men who had strong family networks were more likely than those who did not to have positive self-perceived mental health: 96% versus 90%. These associations remained significant when potentially confounding characteristics were taken into account (Table 3). The odds of positive self-perceived general health and mental health were, respectively, 2.3 and 3.5 times higher for men with strong family networks.

Among Métis women, a strong family network was associated with self-perceived mental health: 91% with strong family networks reported positive mental health, compared with 85% who had weaker networks. This association remained significant when other factors were considered; the odds of positive self-perceived mental health were 1.8 times higher for women with strong family networks. By contrast, no association between the strength of family networks and self-perceived general health emerged for Métis women.

The multivariate models revealed strong positive associations between labour force participation and self-perceived health for Métis men and women. Further analyses of 45- to 64-year-olds highlight the potential health value of having strong family networks, particularly for Métis who were not in the labour force (Figures 2 and 3). For example, among men who were not in the labour force, 65% of those with strong family networks had positive self-perceived general health, compared with 28% of those with weaker family networks.


More than half (54%) of Métis aged 45 or older had strong family networks, defined as very strong ties among family members in other households in the community and at least one relative to turn to in times of need. Family members in close proximity can help with activities such as grocery shopping and appointments, provide opportunities for social participation, and enhance feelings of community attachment, which, in turn, influence individual health status.Note 4Note 30 The subjective assessment of “very strong ties” between family members is an indicator of network quality. Having at least one family member to turn to in times of need reflects an individual’s perceived access to social support.Note 16

Métis women were significantly more likely than Métis men to have a strong family network: 60% versus 48%. By contrast, earlier studies of older Canadians in general revealed that women were more likely than men to be lonely and socially isolated.Note 10Note 31 Having a strong family network and feeling lonely or isolated may not be mutually exclusive—both may be true for women simultaneously. However, it may also indicate that gender differences among Métis differ from those among the Canadian population overall.

Research on sources of social support indicates that spouses are most important for men, while women rely on a slightly broader network.Note 15Note 32 Results from this study show that among Métis, men were more likely than women to identify their spouse as someone to whom they could turn in times of need. Women were more likely than men to mention children or other relatives as sources of support. These results reinforce the need to examine gender-specific social and cultural factors that shape networks.Note 32

Older Métis women were more likely than those in the younger age groups to have strong family networks. Connections may consolidate around family members as they age if additional support is required. Thus, although it is hypothesized that networks influence health status, the opposite may also be true. Reciprocal relationships are a characteristic of social networks,Note 16 and exchanges are not necessarily concurrentNote 33; parents may provide support when their children are younger, and receive support in their later years. Alternatively, the likelihood of having stronger family networks at older ages may be a cohort effect, reflecting differences between earlier and subsequent generations.

This study found that Métis men and women with strong family networks were more likely than those with weaker networks to report positive mental health, even when potential confounders were taken into account. Among Métis men, an association between strong family networks and positive general health was also observed.

According to Berkman et al. ,Note 4Note 30 social and cultural factors influence the quality of social networks, which, in turn, influence health through social support, access to resources, and other psychosocial mechanisms. Education and labour force participation are indicators of social advantage. Lower educational attainment has been associated with loneliness, social isolation, and perceptions of social support.Note 10Note 31 However, in the present study, Métis women with lower education and those who were not in the labour force were more likely to have strong family networks. Relocation for education or employment could explain the weaker family networks among these women, especially because having relatives in the community is a study criterion for a strong family network. The Métis population is highly mobile—22% migrated within Canada between 1996 and 2001.Note 34 The association could also reflect a greater need for support among those with lower levels of education or not in the labour force.

At first glance, this study reflects a paradox among the Aboriginal population,Note 35 where high levels of social support exist but do not always result in better health. Métis women with lower levels of education and who did not participate in the labour force were more likely to have strong family networks, but less likely to report positive general health. However, a sub-analysis of 45- to 64-year-olds suggested that strong social networks are associated with positive health among those not in the labour force. It is possible that strong family networks are protective against the impact of lower socioeconomic status, but cannot compensate completely for any disadvantage.

It was hypothesized that knowledge of an Aboriginal language would be an indicator of cultural affiliation and closer network ties. Only a small percentage of Métis aged 45 or older could speak or understand an Aboriginal language relatively well—this was not associated with having a strong family network and generally not associated with health.

Strengths and limitations

This study makes a contribution by demonstrating the importance of family networks to the health of Métis in mid- to later life. The results are consistent with numerous studies, which show that people who are socially integrated, engaged, and supported are less likely to experience morbidity and mortality.Note 4Note 16

The analysis pertains only to individuals who self-identified as Métis in the 2012 APS, recognizing that this is just one approach to capturing the Métis population in Canada.Note 36Note 37 In addition, the population who self-identified as Métis in the 2012 APS might differ from the population who self-identified in another data collection period and/or data source because of changes in how they report their Aboriginal identity over time, or the possibility that successive generations may identify as Métis for the first time.Note 37Note 38Note 39Note 40

Previous studies have evaluated social and family networks with objective (for instance, number of contacts, frequency of participation) and subjective (for example, satisfaction with social participation) measures.Note 4Note 8Note 16Note 41 Apart from the stipulation that other family members reside in the same community, this study relies on subjective criteria. The advantage is that this allows respondents to evaluate their family ties within the personal and cultural context of their lives. However, the APS questions were not designed to assess family networks. The definition of strong family networks in this analysis is relatively simple, compared with instruments specifically formulated to assess networks.Note 4 This limits comparisons with other studies, as does the absence of a non-Aboriginal reference group. Moreover, the APS questions do not consider the strength of dispersed networks, such as family members living in different communities or broader social networks that include friends, co-workers, and others. People with weak family ties may compensate with broader social networks, which could help explain why high percentages of those with weaker family networks still reported positive general and mental health. Furthermore, based on the 2012 APS questions, it was not possible to disentangle close proximity of kin and family ties.

The temporal order of family network characteristics and health cannot be established with the cross-sectional APS data. While it is hypothesized that a strong family network has a positive impact on health, the opposite may be true.Note 16Note 42

About 5% of Métis respondents aged 45 or older (110) relied on proxies to complete the APS and were excluded from this analysis. This group was older, on average, than the non-proxy population (61 versus 57) and more likely to be male. No difference was evident in the likelihood of reporting positive health based on proxy responses. The impact of excluding these respondents cannot be assessed, but is expected to be negligible, as they comprised a small percentage of APS respondents.


More than half of Métis aged 45 or older reported strong family networks, women more so than men. Those with strong family networks were generally more likely to report positive health. As well, Métis aged 45 to 64 who were not in the labour force were significantly more likely to report better health if they also had strong family networks. This study highlights the value of investigating family networks as a resource that contributes positively to Métis health, as well as the importance of family well-being in addition to individual behaviours for general health promotion. Future research could explore Métis social networks that extend beyond family members to include friends, co-workers, and others.


This study was funded by Indigenous and Northern Affairs Canada (INAC).

Date modified: