Individual- and community-level determinants of Inuit youth mental wellness.

Gray, Andrew Paul ; Richer, Faisca ; Harper, Sam 等

Like many Indigenous groups, Inuit youth have experienced high rates of suicide following the historical trauma and ongoing marginalization associated with Euro-Canadian colonialism. (1) Suicide rates in Nunavik are among the highest in Canada. (2) Yet suicide rates vary substantially among communities, (3) and many Inuit youth are thriving. Moreover, some Inuit "feel that it is detrimental to the health of individuals to continually tell them that they are at the highest risk for disease" and "are interested in research on ... concepts of wellness, and wellness indicators". (4)

Mental wellness, defined by Inuit Tapiriit Kanatami (ITK) as "self-esteem and personal dignity flowing from harmonious physical, emotional, mental and spiritual wellness and cultural identity", (5) is a priority for Inuit organizations. Inuit cite a wide range of social and environmental factors that affect mental wellness, (6-9) including a clear cultural identity and social role;supportive relationships with friends, family and community, particularly Elders; access to and knowledge of the land and the Inuktitut language; and regular consumption of locally harvested food ("country food"). Salient stressors include emotional stress in relationships, violence, poverty, overcrowded housing and environmental change.

However, epidemiological research involving Inuit has focused primarily on individual-level risk factors for psychobehavioural pathology. (10-13) Little attention has been paid to protective factors or to community contexts as determinants of mental health, or to positive dimensions of mental wellness. This phenomenon is not specific to research involving Inuit: a recent review (14) of the scope of research on Indigenous mental health and wellness emphasized the need to include community-level perspectives in epidemiological research, as well as a focus on resilience and well-being. Such studies can be useful in guiding and supporting mental wellness promotion efforts for Inuit youth, and in directing research attention towards key dimensions of the complex community dynamics that give rise to variations in mental wellness among communities.

The current study aimed to quantitatively explore both the individual- and community-level determinants of Inuit youth mental wellness, emphasizing factors that have been identified as important to Inuit but that have rarely been considered in prior epidemiological research.

METHODS

Setting and sample

Nunavik is an Arctic region consisting of 14 fly-in communities ranging from approximately 200 to 2,500 people, of whom 90% identify as Inuit. The communities share a considerable degree of cultural and political history.

We obtained data from the 2004 Nunavik Inuit Health Survey, which has been thoroughly described elsewhere. (15) Briefly, households in all communities were sampled by random systematic draw. If at least one household resident self-identified as Inuk, all household residents aged 15 or older were invited to participate. Questionnaires covered various health and social themes selected in consultation with Inuit representatives. The number of Inuit participating was 1,006, for an overall response rate of 51%. The current study focused on the 452 Inuit participants aged 15-30. This is the upper age limit for "youth" used by the National Inuit Youth Council.

Measures of mental wellness

Mental wellness was measured by means of one positive outcome (self-esteem) and one negative outcome (suicidal ideation) in order to provide a balanced portrait. Self-esteem, part of ITK's definition of mental wellness, (5) was measured by a seven-item version of the extensively validated Rosenberg Self-Esteem Scale. The original 21-point score was converted to a percentage value for ease of interpretation. Suicidal ideation was measured by the dichotomous question: "In the past 12 months, have you thought seriously about committing suicide?", a minor modification of a question from the 2000-2001 Canadian Community Health Survey.

Individual- and community-level explanatory factors

Factors judged by Inuit to influence mental wellness were selected from the measures available in the survey. Table 1 provides definitions for each variable.

Individual-level factors were social support, education, income, pride in Inuit identity, activities relating to the land, difficulty finding animals to hunt, emotionally demanding relationships, histories of physical or sexual violence, and household crowding. Ordinal variables, treated as continuous in order to preserve information content and therefore maximize study power, were scaled to percentages for ease of comparability.

Community-level factors covered the same domains and were calculated as the survey-weighted means of individual-level measures among Inuit respondents of all ages, in order to reflect the characteristics of entire communities. The one exception was average educational attainment, which was calculated only among Inuit under 40. This reflects the cohort that attended Nunavik-run schools, rather than federal day schools or residential schools.

Statistical analysis

Missing data were handled by multiple imputation (Supplementary Appendix 1--see ARTICLE TOOLS section on journal site).

Survey-weighted means of each explanatory factor and outcome were calculated overall, by sex and by community. Community-level variation was assessed by calculating intra-class correlation coefficients (ICCs) for each community-level explanatory factor and each outcome, using random-effect linear regression models.

Associations between each explanatory factor and each outcome were estimated by unweighted marginal multilevel regression models. Models were estimated by generalized estimating equations with an exchangeable correlation structure within households. Standard errors were estimated by the clustered sandwich estimator, adjusted for clustering by household. The link function was linear for self-esteem and logistic for suicidal ideation. No clear non-linearity was detected for any bivariate association, so all exposures were treated as linear.

[FIGURE 1 OMITTED]

We first calculated unadjusted bivariate associations between each explanatory factor and each outcome. Each association was then adjusted for age, sex and a distinct set of potential confounders (Supplementary Table 1). Minimally sufficient sets of potential confounders were selected separately for each association on the basis of a formal causal diagram, depicted in Figure 1. This diagram was developed through a review of qualitative literature on Inuit mental health.

For the sake of comparability across explanatory factors, we present predicted differences in average self-esteem and predicted differences in the prevalence of suicidal ideation as the measures of association, with contrasts corresponding approximately to the range of values in the population. For individual-level factors, the 5th and 95th percentiles in the population were contrasted. For community-level factors, two standard deviations below and above the mean were contrasted. Statistical significance was set at the 95% confidence level.

All calculations were carried out in Stata (Version 12.1, StataCorp, College Station, TX).

Community involvement

The original survey was designed in consultation with, and with oversight from, Nunavik community members. (15) The current secondary analysis was requested as part of regular public health assessment and planning in the context of long-term collaboration between Dr. Richer's institution and the locally governed Nunavik Regional Board of Health and Social Services (NRBHSS). The results of the study were used by multi-sector working groups convened by the NRBHSS (including youth and Elders) to plan the next health survey and to develop mental health promotion recommendations for the region. As recommended by participatory research principles, (16) the research question, study design, choice of variables, interpretation of results and recommendations were generated or validated through discussions with the mental health promotion working group and the NRBHSS's Nutrition and Health Committee, which oversees health research in Nunavik. Both entities granted permission to publish the findings.

Ethics approval

Survey participants gave consent for their anonymized information to be used in future research. Data from the original survey are owned by the NRBHSS. Approval from the Nutrition and Health Committee is required to access the data and was granted for this study. Additional ethical approval was granted by the Research Ethics Board of the McGill University Faculty of Medicine (A11-B54-13B).

RESULTS

Sample characteristics

Table 2 summarizes individual and community characteristics. Inuit youth reported an average self-esteem score of 61%, and 21% reported suicidal ideation in the preceding 12 months. Young women reported higher levels of suicidal ideation and slightly lower self-esteem than young men. Young women also reported higher rates of violence and higher emotional demands, but experienced greater emotional support. Young men reported harvesting and processing animals more frequently, and reported more difficulty finding animals when they went hunting.

Across communities, the average self-esteem score among youth ranged from 56% to 69% (ICC 7.0%), and the prevalence of past-year suicidal ideation among youth ranged from 7% to 34% (ICC 0.8%). Communities varied the most in household crowding (ICC 10.8%), income (ICC 5.6%) and harvesting frequency (ICC 5.1%). Communities were similar in terms of pride in Inuit identity, emotional support and animal processing (ICC < 0.1%).

Figures 2a and 2b show associations with 95% confidence intervals between explanatory factors and each mental wellness outcome.

Land and culture

Greater mental wellness was associated with greater pride in Inuit identity and with more frequent harvesting or processing of animals. There was some evidence of higher self-esteem among youth living in communities with greater collective pride in their Inuit identity. There was some evidence of a lower prevalence of suicidal ideation among youth living in communities that processed or shared more country food, and some evidence of more suicidal ideation among youth experiencing greater difficulty finding animals to hunt.

Social factors

Youth experienced greater mental wellness in communities with more positive interaction. Similar trends were seen in communities with more emotional support. Youth mental wellness was poorer in communities with a higher prevalence of sexual abuse. There was some evidence of more suicidal ideation among youth in communities with more threats of physical violence.

Education, income and housing

Youth mental wellness was higher in communities with greater average educational attainment and incomes. Mental wellness was poorer for youth living in an overcrowded household or community.

Adjusted associations

Adjusting associations for potential confounders did not contradict any of these results, although the magnitude of some associations was attenuated.

In adjusted models, greater cultural pride and living in a community with more positive social interaction, greater educational attainment and higher income were associated with greater mental wellness. Increasing difficulty finding animals to hunt was associated with reduced mental wellness.

In contrast, there was no clear evidence of a harmful effect of living in a community with scarce housing, after adjustment for potential confounders. The magnitudes of the associations with harvesting and processing animals were also reduced considerably by adjustment.

DISCUSSION

This study aimed to explore, quantitatively, determinants of Inuit youth mental wellness with explicit guidance from Inuit perspectives. (6) To the best of our knowledge, this is the first quantitative study in an Inuit population that assesses the determinants of a positive mental wellness outcome, documents the potential protective role of cultural identity and traditional activities, and simultaneously assesses individual- and community-level factors in a multilevel analysis. Its results are likely generalizable to other Inuit communities and perhaps to other remote Indigenous communities.

In this study, pride in Inuit identity and participation in traditional activities relating to the land were associated with greater mental wellness. A clear and positive cultural identity is widely recognized as a crucial resource for mental wellness among Indigenous peoples (17) and features prominently in the ITK definition of mental wellness. This cultural identity includes a strong relationship to the land, a relationship that is enacted, in part, through the harvesting, processing, sharing and consuming of traditional food. (18) The food itself has both nutritional and symbolic value but is also embedded in social activities that serve to strengthen relationships, reinforce cultural values and transmit traditional knowledge. Barriers to hunting, associated with greater suicidality in this study, may deprive men, especially, of both a livelihood and a valued social role. Additionally, animal scarcity in an era of resource development and climate change may be understood by hunters as evidence of environmental degradation, which may be equivalent to a "direct assault on the person" (19) for peoples who identify directly with the land.

At the community level, there also appeared to be positive associations with lower suicidality among residents of communities in which more country food was processed and shared. This may reflect the presence of extensive community networks through which sharing occurs, the availability of positive role models in a community with more successful and generous hunters, or greater optimism for the future in areas where food species remain abundant. The stronger collective relationship with the land and the traditional knowledge underlying these activities may also reflect a higher degree of cultural continuity, famously linked to lower youth suicide rates among BC First Nations and seen as a key resource for youth to feel invested in their own future. (20)

Despite widespread recognition of these links in grey and qualitative literature, few epidemiological studies have investigated the link between culture, land and wellness among Inuit youth. One recent article, which used data from the same survey as the current study, also found that hunting was protective against suicide attempts among young Inuit women. (21)

The current study found that greater mental wellness was associated with living in a community with greater social cohesion and greater socio-economic success. These results agree with previous ecological studies of Inuit and Alaska Native communities, which found lower suicide rates in communities with greater socio-economic success and with stronger ties to culture and family. (22,23) Similar findings in other Indigenous communities in Canada include an association between individual perception of community caring and lower suicidality among adolescents in Manitoba First Nations communities. (24) Our study expands these results to a positive measure of mental wellness, and finds that these associations are robust to adjustment for individual characteristics.

Many mechanisms have been proposed to link Indigenous community social and economic capital to individual mental wellness: Inuit communities with greater social cohesion may provide individuals with greater social support, a sense of belonging, and positive role models. (25) Communities with higher educational achievement and income may also have more resources to invest socially, more successful role models for youth and more optimism based on visible opportunities. (26)

A number of stressors were also identified in this study, including living in an overcrowded household, living in a community with inadequate housing and living in a community with highly prevalent violence.

Many authors have discussed the negative collective impact of violence in Indigenous communities, which can produce feelings of fear and insecurity even for those who have not directly experienced violence, and may be a manifestation of widespread feelings of anger, shame and helplessness as a legacy of historical and collective forms of violence. (27)

Overcrowded housing at the individual level has been previously linked to physical markers of stress in Nunavik. (28) We are not aware of previous studies that have demonstrated an association directly with mental health or wellness or an association with community-level crowding. The latter association was not confirmed after adjustment, though this finding should be viewed with caution since household-level crowding is likely both a confounder and a mediator of the effect of community-level crowding. In overcrowded housing, the lack of private space may impair school performance and increase exposure to violence; at the community level, housing shortages may reduce the availability of safe havens for those fleeing violence.

Finally, prior analysis of the data set (29) has already demonstrated, and this analysis further confirms, that mental wellness is positively associated with individual-level social support, educational attainment and income; it is negatively associated with experiences of excessive emotional demands and violence.

Limitations

The study had several limitations. First, the scope of factors examined was limited by the availability of data. Macro-level factors such as collective political autonomy, climate change and systemic racism could not be assessed. Data on some important individual-level factors were also not available, such as culturally relevant knowledge, intergenerational relationships, childhood family disruption, and experiences of discrimination. The specific content of the measures used here may not have been ideal for the purposes of the current study, and none of the measures used here has been extensively validated in this population. Selection bias may also arise from the somewhat low response rate.

Our ability to adjust associations for confounding factors and to draw conclusions about causality was also limited by the complex multifactorial nature of mental health and wellness, the lack of temporality in cross-sectional data, and limited statistical power. Studies that pool data from multiple Inuit regions may partially overcome this last limitation.

RECOMMENDATIONS AND CONCLUSIONS

Extensive recommendations for improving the determinants of Inuit health have recently been published. (6) Many of these recommendations find support in the current study. While the root causes of high rates of mental health problems among Inuit remain historical, structural and political, the current results can inform support for community-led mental wellness promotion efforts:

* Promotion of cultural identity and traditional activities: Land and culture camps bring together Elders and youth, providing opportunities to build positive relationships and to transmit cultural values and knowledge. (30) Culture and identity may also be supported by greater opportunities to learn Inuit history and language. Many non-Indigenous people, including health and social service workers, require anti-racism education. Economic barriers to harvesting can be reduced through harvester support programs, and environmental protection strategies are needed to maintain wildlife populations in the face of climate change and local resource development.

* Promotion of social cohesion and community wellness: Ongoing support is needed for locally developed activities to promote healing from historical trauma and to prevent violence. Natural helpers and community wellness workers require additional outside support to prevent burnout. Activities to promote traditional culture, which emphasizes interconnection as a core value, may also promote youth resilience.

* Improvement of community living conditions: The stress of overcrowding is one more of many reasons to provide adequate housing in Nunavik. Culturally appropriate economic development and educational opportunities are also badly needed. Improved opportunities for collective discussion and planning, from greater access to community centres to greater political autonomy for Nunavik overall, are likely to facilitate such work.

The results of this study are broadly consistent with Inuit perspectives on the determinants of mental wellness. This echoes the prediction of Pauktuutit Inuit Women of Canada and collaborators: "as researchers become familiar with Indigenous-specific health indicator frameworks, we can expect more 'new' discoveries of 'old truths', as they incorporate the insights of Inuit and other Indigenous cultures into their assessments of Indigenous health". (8(p31)) These results remind us that in addition to their challenges, Inuit communities have important strengths that must be acknowledged and supported. Mental wellness promotion in Nunavik must include support for multifaceted, community-led initiatives that seek to create healthier environments for all.

REFERENCES

(1.) Kral MJ. Postcolonial suicide among Inuit in Arctic Canada. Cult Med Psychiatry 2012;36(2):306-25. PMID: 22392639. doi: 10.1007/s11013-012-9253-3.

(2.) Oliver LN, Peters PA, Kohen DE. Mortality rates among children and teenagers living in Inuit Nunangat, 1994 to 2008. In: Health Reports. Vol. 23. Statistics Canada Catalogue no. 82-003-X. Ottawa, ON: Statistics Canada, 2012.

(3.) Hicks J. Statistical Data on Death by Suicide by Nunavut Inuit, 1920 to 2014. Iqaluit, Nunavut: Nunavut Tunngavik Incorporated, 2015.

(4.) Canadian Institutes of Health Research, Institute of Aboriginal Peoples' Health, Institute of Nutrition, Metabolism and Diabetes. Health Research Needs North of 60: Northern Town Hall Meetings, September 2001. Ottawa, ON: Canadian Institutes of Health Research, 2002.

(5.) Inuit Tapirisat of Canada. Inuit-Specific Mental Wellness Framework. Ottawa, ON: Inuit Tapirisat of Canada, 2001.

(6.) Inuit Tapiriit Kanatami. Social Determinants ofInuit Health in Canada. Ottawa, ON: Inuit Tapiriit Kanatami, 2014.

(7.) Morris M, Crooks C. Structural and cultural factors in suicide prevention: The contrast between mainstream and Inuit approaches to understanding and preventing suicide. J Soc Work Pract 2015;29(3):321-38. doi:10.1080/ 02650533.2015.1050655.

(8.) Pauktuutit Inuit Women of Canada; Rasmussen D, Guillou J. Developing an Inuit-specific framework for culturally relevant health indicators incorporating gender-based analysis. J AborigHealth 2012;8(2):24-35.

(9.) Petrasek MacDonald J, Ford JD, Willox AC, Ross NA. A review of protective factors and causal mechanisms that enhance the mental health of Indigenous Circumpolar youth. Int JCircumpolarHealth 2013;72:21775. PMID: 24350066. doi: 10.3402/ijch.v72i0.21775.

(10.) Kirmayer LJ, Boothroyd LJ, Hodgins S. Attempted suicide among Inuit youth: Psychosocial correlates and implications for prevention. Can JPsychiatry 1998; 43:816-22.PMID: 9806088.

(11.) Boothroyd LJ, Kirmayer LJ, Spreng S, Malus M, Hodgins S. Completed suicides among the Inuit of northern Quebec, 1982-1996: A case-control study. Can Med Assoc J 2001;165(6):749-55. PMID: 11584562.

(12.) Haggarty J, Cernovsky Z, Bedard M, Merskey H. Suicidality in a sample of Arctic households. Suicide Life ThreatBehav 2008;38(6):699-707. doi: 10.1521/ suli.2008.38.6.699. PMID: 19152300.

(13.) Chachamovich E, Kirmayer LJ, Haggarty JM, Cargo M, McCormick R, Turecki G. Suicide among Inuit: Results from a large, epidemiologically representative follow-back study in Nunavut. Can J Psychiatry 2015;60(6): 268-75. PMID: 26175324.

(14.) Wexler L, Chandler M, Gone JP, Cwik M, Kirmayer LJ, LaFromboise T, et al. Advancing suicide prevention research with rural American Indian and Alaska Native populations. Am JPublic Health 2015;105(5):891-99. PMID: 25790403. doi: 10.2105/AJPH.2014.302517.

(15.) Rochette L, Blanchet C. Qanuippitaa? Methodological Report. Quebec: Institut national de sante publique du Quebec, Nunavik Regional Board of Health and Social Services, 2007.

(16.) Inuit Tapiriit Kanatami, Nunavut Research Institute. Negotiating Research Relationships with Inuit Communities: A Guide for Researchers. Ottawa, ON and Iqaluit, NU: Inuit Tapiriit Kanatami, Nunavut Research Institute, 2007.

(17.) King M, Smith A, Gracey M. Indigenous health part 2: The underlying causes of the health gap. Lancet 2009;374(9683):76-85. PMID: 19577696. doi: 10.1016/S0140-6736(09)60827-8.

(18.) Kral MJ, Idlout L, Minore JB, Dyck RJ, Kirmayer LJ. Unikkaartuit: Meanings of well-being, unhappiness, health, and community change among Inuit in Nunavut, Canada. Am J Community Psychol 2011;48(3-4):426-38. PMID: 21387118. doi: 10.1007/s10464-011-9431-4.

(19.) Kirmayer LJ, Brass GM, Tait CL. The mental health of Aboriginal peoples: Transformations of identity and community. Can J Psychiatry 2000;45(7): 607-16. PMID: 21387118.

(20.) Chandler MJ, Lalonde CE. Cultural continuity as a moderator of suicide risk among Canada's First Nations. In: Kirmayer LJ, Valaskakis GG (Eds.), Healing Traditions: The Mental Health ofAboriginal Peoples in Canada. Vancouver, BC: UBC Press, 2009;221-48.

(21.) Fraser SL, Geoffroy D, Chachamovich E, Kirmayer LJ. Changing rates of suicide ideation and attempts among Inuit youth: A gender-based analysis of risk and protective factors. Suicide Life ThreatBehav 2015;45(2):141-56. PMID: 25255825. doi: 10.1111/sltb.12122.

(22.) Penney C, Senecal S, Guimond E, Bobet E, Uppal S. Suicide in Inuit Nunaat: An analysis ofsuicide rates and the effect ofcommunity-level factors. Paper presented at: 5th Northern Research Forum Open Assembly; June 27, 2008; Anchorage, AK.

(23.) Berman M. Suicide among young Alaska Native men: Community risk factors and alcohol control. Am J Public Health 2014;104(S3):S329-35. PMID: 24754505. doi: 10.2105/AJPH.2013.301503.

(24.) Mota N, Elias B, Tefft B, Medved M, Munro G, Sareen J. Correlates of suicidality: Investigation of a representative sample of Manitoba First Nations adolescents. Am J Public Health 2012;102(7):1353-61. PMID: 22676500. doi: 10.2105/AJPH.2011.300385.

(25.) Petrasek MacDonald J, Cunsolo Willox A, Ford JD, Shiwak I, Wood M. Protective factors for mental health and well-being in a changing climate: Perspectives from Inuit youth in Nunatsiavut, Labrador. Soc Sci Med 2015; 141:133-41. PMID: 26275362. doi: 10.1016/j.socscimed.2015.07.017.

(26.) Kirmayer LJ, Sehdev M, Whitley R, Dandeneau SF, Isaac C. Community resilience: Models, metaphors and measures. J Aborig Health 2009;5(1):62-117.

(27.) Fast E, Collin-Vezina D. Historical trauma, race-based trauma and resilience of indigenous peoples: A literature review. First Peoples Child Fam Rev 2010; 5(1):126-36.

(28.) Riva M, Plusquellec P, Juster R-P, Laouan-Sidi EA, Abdous B, Lucas M, et al. Household crowding is associated with higher allostatic load among the Inuit. J Epidemiol Community Health 2014;68(4):363-69. PMID: 24385548. doi: 10.1136/jech-2013-203270.

(29.) Kirmayer LJ, Paul KW. Mental Health, Social Support and Community Wellness. Quebec: Institut national de sante publique du Quebec, Nunavik Regional Board of Health and Social Services, 2007.

(30.) Takano T. Connections with the land: Land-skills courses in Igloolik, Nunavut. Ethnography 2005;6(4):463-86. doi: 10.1177/1466138105062472.

Received: October 21, 2015

Accepted: February 14, 2016

Andrew Paul Gray, MD, MSc, [1] Faisca Richer, MD, MSc, [1,2] Sam Harper, PhD [1]

Author Affiliations

[1.] Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC

[2.] Secteur sante des Autochtones, Institut national de sante publique du Quebec, Montreal, QC Correspondence: Dr. Andrew Paul Gray, Department of Epidemiology, Biostatistics and Occupational Health, McGill University, 1020 Pine Avenue West, Montreal, QC H3A 1A2, Tel: 514-398-6258, E-mail: andrew.p.gray@mail.mcgill.ca

Acknowledgements: We thank the people of Nunavik, particularly the Mental Health Promotion and Prevention Working Group of the Nunavik Regional Board of Health and Social Services, for invaluable insights regarding interpretation and recommendations; all those who planned, carried out or participated in the 2004 Nunavik Inuit Health Survey; and all those who work hard every day to support the wellness of the communities of Nunavik. We also thank Veronique Paradis and Isabelle Duguay, who provided helpful feedback on the article. AG and FR were supported by the Nunavik Regional Board of Health and Social Services to travel to Nunavik for meetings with community representatives. SH was supported by a Chercheur-boursier Junior 2 award from the Fonds de recherche du Quebec-Sante.

Conflict of Interest: None to declare. Table 1. Definitions of study variables Possible individual- Variable Definition level values * Land and culture Pride in Agreement with the 0% = Never; 33% = Inuit identity statement "I am Rarely; 67% = proud to be an Inuk" Sometimes; 100% = Usually Harvesting animals Number of episodes Range: 0-320 of hunting or fishing in the last 12 months Processing animals Number of animal Range: 0-120 carcasses processed (skinning, washing, cutting) in the last 12 months Sharing country "Do you receive 0% = Never; 50% = food ([dagger]) country food from Sometimes; 100% = your friends or Often relatives outside your household?" (community-level only: question not asked of individuals) Difficulty hunting "Since 2000, have 0 = No; 1 = Yes any species you hunt for food been harder to catch/hunt/find?" Social factors Emotional support Has someone to talk 0% = Never; 25% = to if feeling Rarely; 50% = troubled or for Sometimes; 75% = emotional support Most of the time; 100% = All of the time Positive Has someone to have 0% = Never; 25% = interaction a good time with Rarely; 50% = Sometimes; 75% = Most of the time; 100% = All of the time High emotional Someone causes worry 0% = Never; 25% = demands ([dagger]) or demands too much Rarely; 50% = (individual-level Sometimes; 75% = only: the community- Most of the time; level mean was 100% = All of the considered a time consequence rather than a cause of prevalent mental health problems) History of Lifetime history of 0 = No; 1 = Yes sexual abuse forced or attempted forced sexual activity History of threats Adult history of 0 = No; 1 = Yes of physical threats of physical violence violence that were so serious that the person became afraid Education, income and housing Educational Highest level of 0% = No school; 25% attainment school reached = Primary only; 50% = Some secondary; 75% = Finished secondary; 100% = Some post-secondary Personal Gross income in the 0% = Less than income past 12 months $20,000; 33% = $20,000 to $40,000; 67% = $40,000 to $60,000; 100% = Over $60,000 Overcrowding People per room, Continuous calculated as number of people divided by (number of bedrooms + 2) * Community-level variables were calculated as the community mean of the corresponding individual-level variable. ([dagger]) All exposures were considered at both the individual and community levels, except where otherwise noted. Table 2. Survey-weighted means of individual and community factors experienced by Inuit youth in Nunavik, 2004 Individual-level factors (n = 452) Overall Women Men mean (n = 244) (n = 208) Self-esteem 60.9% 58.5% 63.6% Suicidal ideation in 21.8% 26.5% 16.2% past 12 months Pride in Inuit identity 90.1% 89.6% 90.6% Harvesting animals 73.5 55.6 94.4 (episodes/year) Processing animals 10.7 6.7 15.5 (animals/year) Sharing country food 57.6% 59.3% 55.6% Difficulty hunting 48.0% 43.5% 53.3% Emotional support 54.4% 59.3% 48.6% Positive interaction 71.7% 72.6% 70.6% High emotional demands 39.0% 42.5% 34.8% History of sexual abuse * 44.0% 61.9% 23.1% History of threats of 24.6% 33.1% 14.6% physical violence * High school completion1 27.6% 28.2% 26.8% Personal income > 27.3% 27.6% 26.8% $20,000/year Overcrowding (people/room) 1.12 1.13 1.11 Community-level factors (n = 14) Minimum Maximum Self-esteem 56.3% 68.5% Suicidal ideation in 6.7% 33.6% past 12 months Pride in Inuit identity 89.4% 97.4% Harvesting animals 71.7 157.8 (episodes/year) Processing animals 10.1 29.0 (animals/year) Sharing country food 38.2% 64.2% Difficulty hunting 37.2% 69.8% Emotional support 45.9% 60.3% Positive interaction 57.5% 72.7% High emotional demands 27.5% 53.3% History of sexual abuse * 25.5% 68.5% History of threats of 13.2% 31.1% physical violence * High school completion1 14.0% 41.0% Personal income > 24.5% 53.0% $20,000/year Overcrowding (people/room) 0.76 1.26 * Mean calculated for Inuit aged 18+ only, because these questions were not asked of those under 18. ([dagger]) Mean calculated for Inuit aged 18-39 only, because older cohorts did not attend Nunavik-run schools. Figure 2. (a) Associations between individual-and community- level factors and average self-esteem among Inuit youth in Nunavik, 2004. (b) Associations between individual-and community-level factors and prevalence of suicidal ideation among Inuit youth in Nunavik, 2004. Note: Percentages indicate ordinal variables treated as continuous, so that 0% corresponds to all community members giving the lowest possible answer to a multiple choice question, and 100% corresponds to all community members giving the highest possible answer. (a) Individual factors Unadjusted difference Factor Contrast (95% CI) Personal income $40-60k vs <$20k/y +10.8 (+6.0, +15.6) Positive interaction Always vs. never +9.1 (+2.1, +16.1) Emotional support Always vs. never +8.5 (+3.1, +13.9) Educational attainment Some post-secondary +8.3 (+3.5, +13.1) vs. some secondary Pride in Inuit identity Usually vs. rarely +7.5 (+3.1, +11.9) Processing animals 47 vs 0 animals/year +4.9 (+1.2, +8.7) Harvesting animals 200 vs 0 episodes/year +4.7 (+0.3, +9.2) Difficulty hunting Yes vs. no +0.2 (-3.1, +3.4) Threats of violence Yes vs. no -2.7 (-6.9, +1.5) Overcrowding 1.9 vs 0.5 people/room -5.6 (-11.4, +0.3) High emotional demands Always vs. never -6.6 (-11.9, -1.4) Sexual abuse Yes vs. no -6.9 (-10.3, -3.4) Community factors Positive interaction 75% vs 59% +12.3 (+2.1, +22.5) Educational attainment 65% vs 52% +10.7 (+3.5, +17.9) Average income 28% vs 10% +9.9 (+3.3, +16.5) Pride in Inuit identity 98% vs 88% +5.6 (-2.4, +13.7) Emotional support 61% vs 45% +5.5 (-1.2, +12.2) Sharing country food 70% vs 40% +3.6 (-3.4, +10.5) Processing animals 25 vs 7 animals/year +1.9 (-8.4, +12.1) Difficulty hunting 69% vs 31% +1.4 (-4.5, +7.2) Threats of violence 34% vs 14% -1.5 (-9.1, +6.1) Harvesting animals 145 vs 54 episodes/year -2.1 (-10.8, +6.6) Overcrowding 1.3 vs 0.8 people/room -8.0 (-15.7, -0.3) Sexual abuse 66% vs 21% -10.7 (-18.2, -3.3) Adjusted difference Factor (95% CI) Personal income +5.5 (+0.1, +10.8) Positive interaction +7.3 (+0.3, +14.2) Emotional support +7.1 (+1.7, +12.5) Educational attainment +8.2 (+3.5, +12.9) Pride in Inuit identity +4.8 (+0.5, +9.2) Processing animals +1.5 (-2.5, +5.4) Harvesting animals +1.2 (-3.2, +5.7) Difficulty hunting -0.9 (-4.1, +2.3) Threats of violence -0.7 (-4.6, +3.2) Overcrowding -3.3 (-8.0, +1.5) High emotional demands -8.6 (-13.6, -3.5) Sexual abuse -3.7 (-7.3, -0.2) Community factors Positive interaction +22.6 (+0.9, +44.3) Educational attainment +10.4 (+2.9, +17.8) Average income +12.5 (+1.5, +23.5) Pride in Inuit identity +5.7 (-10.1, +21.4) Emotional support +6.8 (-16.0, +29.5) Sharing country food +0.2 (-7.4, +7.9) Processing animals -0.8 (-30.2, +28.6) Difficulty hunting +1.1 (-4.7, +6.8) Threats of violence -5.0 (-16.9, +6.9) Harvesting animals +6.0 (-10.1, +22.2) Overcrowding +6.4 (-8.2, +21.1) Sexual abuse -6.3 (-14.3, +1.7) (b) Individual factors Unadjusted difference Factor Contrast (95% CI) Personal income $40-60k vs <$20k/y -17.0 (-27.7, -6.3) Educational attainment Some post-secondary -15.1 (-23.6, -6.6) vs. some secondary Harvesting animals 200 vs 0 episodes/year -12.1 (-23.6, -0.6) Processing animals 47 vs 0 animals/year -11.3 (-21.5, -1.1) Pride in Inuit identity Usually vs. rarely -9.8 (-21.6, +2.1) Positive interaction Always vs. never -1.0 (-17.3, +15.3) Emotional support Always vs. never +2.3 (-10.2, +14.7) Difficulty hunting Yes vs. no +8.3 (-0.2, +16.7) Overcrowding 1.9 vs 0.5 people/room +12.7 (+1.0, +24.4) Sexual abuse Yes vs. no +15.6 (+6.8, +24.5) Threats of violence Yes vs. no +20.4 (+8.5, +32.2) High emotional demands Always vs. never +25.6 (+11.5, +39.7) Community factors Positive interaction 75% vs 59% -15.3 (-35.3, +4.6) Processing animals 25 vs 7 animals/year -11.9 (-32.5, +8.6) Sharing country food 70% vs 40% -10.9 (-26.9, +5.0) Average income 28% vs 10% -10.5 (-25.9, +4.9) Educational attainment 65% vs 52% -9.7 (-25.8, +6.4) Emotional support 61% vs 45% -7.4 (-21.7, +6.9) Pride in Inuit identity 98% vs 88% -4.4 (-21.2, +12.4) Harvesting animals 145 vs 54 episodes/year -1.4 (-22.7, +19.9) Difficulty hunting 69% vs 31% -1.1 (-15.4, +13.2) Overcrowding 1.3 vs 0.8 people/room +6.0 (-12.2, +24.2) Sexual abuse 66% vs 21% +6.5 (-11.5, +24.5) Threats of violence 34% vs 14% +8.3 (-9.9, +26.6) Adjusted difference Factor (95% CI) Personal income -10.9 (-23.8, +2.0) Educational attainment -14.7 (-21.5, -7.8) Harvesting animals -4.0 (-15.9, +7.9) Processing animals -4.7 (-16.0, +6.5) Pride in Inuit identity -9.0 (-21.0, +2.9) Positive interaction +4.7 (-10.0, +19.4) Emotional support +5.7 (-5.6, +16.9) Difficulty hunting +10.0 (+1.4, +18.6) Overcrowding +8.8 (-2.1, +19.6) Sexual abuse +10.6 (+2.4, +18.7) Threats of violence +15.8 (+5.1, +26.6) High emotional demands +28.1 (+14.2, +42.1) Community factors Positive interaction -22.1 (-69.9, +25.8) Processing animals -16.3 (-80.7, +48.0) Sharing country food -9.1 (-26.8, +8.6) Average income -18.6 (-44.4, +7.2) Educational attainment -9.2 (-26.5, +8.1) Emotional support +5.2 (-35.7, +46.0) Pride in Inuit identity -11.5 (-54.6, +31.7) Harvesting animals +1.1 (-42.6, +44.8) Difficulty hunting +0.9 (-12.7, +14.5) Overcrowding -10.1 (-40.2, +20.0) Sexual abuse +1.0 (-19.0, +20.9) Threats of violence +17.5 (-11.7, +46.8)