Chapter 2: The Health of Canada's Young People: a mental health focus: Mental Health – Mental health

Mental Health

by John Freeman and King Luu

There is growing recognition within Canada of the critical nature of improving the mental health of young Canadians with a particular emphasis on positive mental health1 in contrast to what has historically been a focus on mental illness. As Buote (2009) states, in her literature review commissioned by the Public Health Agency of Canada (PHAC), "we know that children who have greater social-emotional health experience greater overall success in life and have better abilities that can be used to overcome life's challenges"; we also know that "children who suffer from more mental health problems also have more adjustment problems in the area of health, school, self and home" (pp. 9-10).

This trend to move away from a deficit model of mental health toward a more balanced view (Keyes, 2006) measuring the dual (positive and negative) nature of mental health is further reinforced by a growing acknowledgment of the importance of developmental assets in young people's lives (Scales, 1999). Assets are characteristics that help youth develop to their fullest potential, shaped by external influences (e.g., parents, teachers) or personal influences (e.g., self-esteem, determination; Benson & Scales, 2009). Assets tend to promote greater personal resilience, namely, the ability to achieve success despite obstacles (Fergus & Zimmerman, 2005).

The commitment to better understanding and improving mental health with a focus on asset-based approaches has resulted in major Canadian initiatives exploring this domain. For example, in September 2009, PHAC invited over 80 cross-sectoral participants to a consultation on social-emotional protective factors for young people and their families (PHAC, 2009). One of the conclusions from this two-day meeting was that "population and administrative data from multiple jurisdictions is needed to compare [provincial/ territorial] child/youth/family health outcomes with national trends" (p. 24). Similarly, in May 2010, the Canadian Population Health Initiative (CPHI) – in cooperation with ministries within the New Brunswick provincial government – invited approximately 50 individuals from across the country to a one-day workshop (Canadian Institute for Health Information [CIHI], 2010). Two of the four key strategies suggested by participants at this workshop were (i) focusing on determinants, and (ii) developing, strengthening and using evidence, both of which depend on well-validated data, such as those provided by the HBSC. Indeed, Schonert-Reichl, Stewart Lawlor, Oberle, and Thomson (2009), having interviewed six experts in mental health about emotional-social healthy living for children aged 5-12, highlight the HBSC as the tool within Canada currently providing national information to understand mental health for that age group, even though the HBSC monitors health only for 11-15 year olds.

While these reports provide strong evidence for focusing the HBSC national report on mental health, our final decision to examine this topic so extensively arose through our ongoing discussions with policymakers and practitioners in the field. While consulting with the Joint Consortium for School Health (JCSH), attending provincial and national meetings of health and educational professionals, and conversing informally with colleagues, we repeatedly heard the request for more information about mental health, particularly positive mental health/emotional well‑being.

Development of mental health indicators

For the current HBSC national report, we created four mental health scales: emotional problems, behavioural problems, emotional well‑being, and prosocial behaviours. In this chapter, the four scales are presented through discussion of the items comprising the scale, with an examination of one item from each scale, an investigation of the scale with respect to grade and gender, and an exploration of the scale in terms of family arrangement and perceived wealth/socio-economic status. In subsequent chapters, these scales are related to each of the topics under consideration.

Construction of the scales was goal-informed, research-informed, and data-informed. On a goal level, our initial consultations with PHAC, Health Canada, and JCSH revealed a desire for more survey items to measure mental health. Thus our goal became to find and include a wider variety of well-validated mental health questions from the international HBSC data packages and appropriate other surveys within Canada.

The research literature revealed two dimensions of mental health. Psychologists have long recognized the presence of both externalizing and internalizing signals for mental health (Hopwood & Grilo, 2010; Krueger & Markon, 2011). Thus, some students reveal their mental health in a more behavioural manner, while other students display their mental health on an emotional level. There are also positive and negative aspects of mental health: mental health is no longer thought of as solely the absence of negative aspects but the presence of positive aspects as well, such that a person can lack negative aspects but still not have positive aspects and thereby not possess optimum mental health (Keyes, Dhingra, & Simoes, 2010; Strauss, 2007). Therefore, we needed to construct four scales: internalizing-negative (emotional problems), externalizing-negative (behavioural problems), internalizing-positive (emotional well‑being), and externalizing-positive (prosocial behaviours).

Our final consideration was the student data from the survey. We conducted factor analysis on the data from across Canada and were able to extract five scales, each with good to very good reliability. The fifth scale, psychosomatic symptoms, was not used in this report because of its item overlap and conceptual similarity with emotional problems. However, three of its individual items are highlighted later in the chapter as is the item on life satisfaction, which was not included in the factor analysis.

Emotional problems

The emotional problems scale consists of nine items (see Table 2.1) with a reliability of 0.84. Four of the items are measured on frequency: feeling low (depressed), feeling nervous, having difficulties getting to sleep, and feeling sad. The first three items are measured for a six-month time period, while the last item is measured over a week. The remaining five questions ask for strength of agreement. They assess students' trouble making decisions, wishing they were someone else, feeling helpless, feeling left out of things, and feeling lonely. When students are divided into three approximately equally sized groups with respect to emotional problems, 33.8% of students are in the group with the greatest number of emotional problems.

Figure 2.1 shows the percentages by grade and gender for one item in the emotional problems scale, students wishing they were someone else. At each grade level, girls were more likely than boys to wish they were someone else. The percentages for both boys and girls wishing they were someone else held relatively steady across grade levels, ranging from 22-24% for boys and 29-32% for girls.

The emotional problems scale shows a similar gender pattern to the "wishing they were someone else" item with girls consistently reporting a higher level of emotional problems than boys (Figure 2.2). However, while the numbers stay fairly stable for boys, they increase across grades for girls, so that, while there is an 8% difference between genders in Grade 6 (27% for boys and 35% for girls), the difference increases to 16% by Grade 10 (28% as opposed to 44%).

2.1 Students who agree or strongly agree they often wish they were someone else, by grade and gender (%)

[Click to enlarge Figure 2.1]

2.2 Students reporting high levels of emotional problems, by grade and gender (%)

[Click to enlarge Figure 2.2]

2.3 Students reporting high levels of emotional problems by perceived wealth and living with both parents (%)

[Click to enlarge Figure 2.3]

There is some statistical difference in the emotional problems reported by students in different family living arrangements with a lower percentage of such problems reported by adolescents living with both parents. However, the bigger difference in reported emotional problems comes from differences in perceived wealth. Those young people who reported lower wealth were much more likely than those reporting higher wealth to indicate having emotional problems (41% to 24% for living with both parents; 47% to 34% for other family arrangements) (Figure 2.3).

Behavioural problems

There are six items on the behavioural problems scale with a reliability of 0.84 (see Table 2.2). Students were asked to indicate the extent to which each of these statements described them. Items encompassed cutting classes/skipping school, making other people do what one wants, talking back to teachers, getting into fights, regularly saying mean things to get what one wants, and taking other person's things. When students were divided into three approximately equally sized groups with respect to behavioural problems, 35.8% of students were in the group reporting the highest number of behavioural problems.

2.4 It is somewhat or definitely like me to get into fights, by grade and gender (%)

[Click to enlarge Figure 2.4]

The illustrative item for behavioural problems is getting into fights (Figure 2.4), chosen for its higher frequency level than the other items. In considering those students who indicated that getting into fights was somewhat or definitely like them² , there was a very low reported frequency. While the percentage of girls getting into fights ranged from 6% in Grade 6-8 to 4% in Grade 9-10 , the pattern for boys was inconsistent. At each grade level, more boys than girls indicated they were likely to get into fights.

2.5 Students reporting high levels of behavioural problems, by grade and gender (%)

[Click to enlarge Figure 2.5]

2.6 Students reporting high levels of behavioural problems by perceived wealth and living with both parents (%)

[Click to enlarge Figure 2.6]

Figure 2.5 shows the small but consistent pattern for boys to be in the highest group for behavioural problems (differences between boys and girls ranging from 3-6%, by grade). For both girls and boys, there is a consistent increase in behavioural problems across grades with the lowest level for both being reported in Grade 6 (27%, girls; 30% boys) and the highest level in Grade 10 (45% girls; 48% boys).

There are almost no differences based on perceived wealth for behavioural problems (Figure 2.6). Family structure has a greater impact on behavioural problems than perceived wealth with 8% fewer behavioural problems for those students with lower perceived wealth living with both parents compared to students in other living arrangements and a 7% difference across family arrangements for students with higher perceived wealth.

Emotional well‑being

Emotional well‑being is measured with five items: two (happy home life and self-confidence) using agreement with the statement, and three (feeling fit and well, feeling full of energy, and having fun with one's friends) using frequencies (Table 2.3). These indicators have a reliability of 0.73. When students were divided into three approximately equally sized groups with respect to emotional well‑being, 36.9% of students were in the group reporting the highest level of well‑being.

2.7 Students who strongly agree they have confidence in themselves, by grade and gender (%)

[Click to enlarge Figure 2.7]

As shown in Figure 2.7, students' self-confidence is influenced by both grade and gender. Fifty percent of boys in Grade 6 strongly agreed that they had confidence in themselves, but this level dropped fairly substantially in each advancing grade to just 29% in Grade 9 and 26% in Grade 10. The proportion of girls with self-confidence started off lower than that of boys in Grade 6 at 40% and declined in a similar fashion to just 17% in Grade 9 and 18% in Grade 10. The biggest drop in self-confidence for boys occurred between Grades 6 and 7 and again between Grades 7 and 8 (8%); for girls, the largest drop was between Grades 7 and 8 (11%).

2.8 Students reporting high levels of emotional well‑being, by grade and gender (%)

[Click to enlarge Figure 2.8]

The results for emotional well‑being in Figure 2.8 mirror those for self-confidence in the previous figure. Boys were consistently more likely than girls to be in the high group for emotional well‑being. As grades increased, emotional well‑being steadily decreased, with both boys and girls reporting the highest level in Grade 6 (56% and 46%, respectively) and the lowest level in Grade 10 (30% and 21%, respectively). For both genders, the smallest drop was between Grades 9 and 10 (3% for boys and 2% for girls) and the largest drop was between Grades 7 and 8 (9% for boys and 10% for girls).

2.9 Students reporting high levels of emotional well‑being by perceived wealth and living with both parents (%)

[Click to enlarge Figure 2.9]

Both perceived wealth and living arrangements were related to emotional well‑being (Figure 2.9). Perceived wealth had the bigger influence in that there was a 21% increase in emotional well‑being associated with wealth for students living with both parents (27% and 48%, respectively) and a 17% increase for students in other living arrangements (21% and 38%, respectively). While not as strong an influence, family structure contributed to better emotional well‑being with a 6% increase for less wealthy students and a 10% increase for more wealthy students associated with living with both parents.

Prosocial behaviours

The five-item prosocial behaviours scale has a reliability of 0.85 (Table 2.4). Similar to the behavioural problems scale, each item is rated in terms of likeness to oneself. Prosocial behaviours are seen in terms of performing positive actions for people without being asked including doing favours, lending things, helping, complimenting, and sharing things. When students were divided into three approximately equally sized groups with respect to prosocial behaviours, 31.5% of students were in the highest group.

2.10 It is definitely like me to often help people without being asked, by grade and gender (%)

[Click to enlarge Figure 2.10]

While girls were more likely than boys to report helping people without being asked when followed across all grades (Figure 2.10), the difference was quite small, ranging from 2% (Grade 9) to 6% (Grades 6 and 8). The decline over time was steady and pronounced with the boys dropping from 22% in Grade 6 to 11% in Grade 10, and the girls decreasing from 28% in Grade 6 to 16% in Grades 9 and 10.

2.11 Students reporting high levels of prosocial behaviour, by grade and gender (%)

[Click to enlarge Figure 2.11]

As seen in Figure 2.11, the results on the prosocial behaviours scale are similar to those on helping people in one respect. At each grade level, girls were more likely than boys to report prosocial behaviours, consistently at 12-15% higher. However, unlike helping others, the only marked decline in prosocial behaviours occurred between Grade 6 and Grade 8 when both boys' and girls' reports decreased by 8%. The numbers stabilized from Grades 8-10 for both genders.

2.12 Students reporting high levels of prosocial behaviour by perceived wealth and living with both parents (%)

[Click to enlarge Figure 2.12]

Figure 2.12 shows there is almost no difference between students in different family arrangements with respect to prosocial behaviours. In contrast, there is a difference related to perceived wealth (6% for both parents; 5% for other living arrangements), favouring those with greater perceived wealth.

Psychosomatic symptoms in Canadian students

Over the past several cycles of the HBSC national report, we have commented on a trio of psychosomatic symptoms. Two of these symptoms (backaches and being bad-tempered or irritable) are not part of any of the four scales we constructed for this report (scales based on factor analyses), while the third symptom (feeling depressed or low) is part of the emotional problems scale. To continue our trend analysis of these three symptoms, we include them here.

2.13 Having backaches at least once a month, by grade and gender (%)

[Click to enlarge Figure 2.13]

Figure 2.13 shows the percentage of students who reported having a backache at least once a month in the past six months. Proportions of young people who have backaches increased for both boys and girls over the grades, starting at 30% for boys and 31% for girls in Grade 6, growing to 49% and 62% by Grade 10, respectively. The increase was thus steeper for girls than it was for boys, such that the 1% difference in Grade 6 grew to a 13% difference in Grade 10. The largest increases for girls (at 10%) were between Grade 6 and 7 and between Grade 8 and 9. The largest increase for boys (8%) was between Grade 7 and 8.

2.14 Having backaches at least once a month, by grade, gender, and year of survey (%)

[Click to enlarge Figure 2.14]

Two trends can be noted in the 2010 backache data in comparison with previous surveys (Figure 2.14). The greater number of boys than girls reporting backaches in Grade 6 for the 2006 survey has not continued in 2010, where, as in 1994, 1998, and 2002, the numbers for both sexes were almost the same. Similarly, there is no noticeable trend for Grade 8 students. However, the increasing differential between Grade 10 girls and boys with respect to backache noted in 2006 (60% and 53% respectively) has been further augmented in 2010 (62% and 49%). Thus, while Grade 10 boys' reports of backaches have consistently decreased since 1998 (after a sharp increase from 1994 to 1998), the Grade 10 girls' reported incidence remained virtually unchanged.

2.15 Feeling depressed or low at least once a week, by grade and gender (%)

[Click to enlarge Figure 2.15]

Figure 2.15 gives a picture of the percentage of students who felt depressed or low at least once a week in the past six months. For boys, the number was relatively stable across grades, ranging from 20% to 24%. In contrast, for girls, the number was lowest in Grade 6 (28%) and increased steadily to 38% in Grade 10. The gender gap was smallest in the two lowest grades (6% Grade 6; 9% Grade 7) and largest in the three highest grades (14% Grades 8 and 10; 13% Grade 9).

2.16 Feeling depressed or low at least once a month, by grade, gender, and year of survey (%)

[Click to enlarge Figure 2.16]

The pattern for feeling low or depressed by grade and gender varies little by survey year (Figure 2.16). In all surveys, the percentages for girls exceeded those for boys in Grades 8 and 10, as the symptom remains relatively constant for boys across grades while increasing steadily for girls. While there was virtually no gender gap in feeling depressed or low in the past four surveys for Grade 6 students, there was a Grade 6 gender gap in 2010.

2.17 Feeling bad tempered or irritable more than once a week, by grade and gender (%)

[Click to enlarge Figure 2.17]

Similar to the other two psychosomatic symptoms reported, the pattern for being bad tempered or irritable more than once a week in the past six months showed gender differences that generally grew in older grades (Figure 2.17). The percentage of boys being in a bad mood/irritable remained constant across the grades, whereas the percentage of girls suffering from these conditions increased by grade, with the biggest jump between Grade 7 and Grade 8. Thus, while boys and girls had almost the same rate of being bad tempered/ irritable in Grade 6, at 14% and 15% respectively, there was a gender gap in all the other grades, with girls reporting higher incidences of bad mood/irritability in each grade. The gap was widest in Grade 10 at 10%.

2.18 Feeling bad tempered or irritable more than once a week, by grade, gender, and year of survey (%)

[Click to enlarge Figure 2.18]

The findings related to gender differences in bad temper and irritability were consistent with those from the surveys since 1998 (Figure 2.18). In each survey year since 1998, girls' reporting of bad temper/irritability spiked between Grades 6 and 8 but was quite steady between Grades 8 and 10, while the boys' reporting of the symptom showed little grade-related change. While there was a marked decrease in feeling bad tempered or irritable between the 1994 and 1998 surveys, the percentages within each grade-gender combination (Grade 6 boys and girls, Grade 8 boys and girls, Grade 10 boys and girls) have not changed substantially since 1998.

Life satisfaction among students

2.19 How students rate life out of 10, by grade and gender (%)

[Click to enlarge Figure 2.19]

In Grade 6, 62% of boys and 60% of girls gave themselves a high score (between 8 and 10) on the life satisfaction ladder (Figure 2.19). This proportion for boys remained steady in Grade 7 but went down to 58% in Grade 8 and 55% in Grades 9 and 10. In contrast, the percentage of girls reporting high life satisfaction decreased sharply from Grade 6 (60%) to Grade 7 (55%), to Grade 8 (50%). The ratings dropped to under 50% in Grades 9 and 10 (49% and 44%, respectively). As with many other measures of emotional health, there were limited gender differences in Grade 6 (2%), but marked gender differences in Grade 10 (11%).

What young people thought about these findings

The major goal of the youth engagement workshop in March 2011 was to document the insights of a sample of young people from across Canada with respect to mental health and its determinants. As such, mental health was more prominent in the young people's minds than other aspects of health. More specifically, these young people were: (1) presented with core definitions of the four indicators of mental health being considered in this report (emotional well‑being, prosocial behaviour, emotional problems, behavioural problems); (2) asked to provide insights into some of the basic patterns observed in these key indicators; (3) asked to provide additional interpretation of observed relationships among mental health, health behaviours, and environmental factors as outlined in subsequent chapters.

It is clear that young people can provide deep and meaningful insights into many of the patterns that emerged from the HBSC analyses. Many of these insights focused on the role of developmental and gender factors in shaping their emotions and feelings. Others focused on how various environments such as home, peer group, and school shaped their feelings and emotions, and how relationships in these environments were core determinants of mental health. For them, mental health transcended all aspects of their daily lives and affected them and their situation no matter where they were or what they were doing. The youth agreed with the current research that good mental health is not just about the absence of negative aspects but also about the presence of positive aspects, the assets in their lives. This balanced interpretation was appreciated.

It was clear from the workshop that there is value in asking young people directly about statistical patterns to get at deeper meanings and interpretations of research findings. This workshop was a trial of sorts for HBSC. The workshop format offered a novel approach to gaining richer perspective on statistical findings with the final report benefiting from the process.

Summary and implications

Commentary

This chapter focuses on four types of mental health indicators, contrasted with respect to internalizing/emotional versus externalizing/behavioural, and positive versus negative outcomes. These indicators were examined with reference to grade, gender, perceived wealth, and family structure.

Regardless of the way internalizing/emotional outcomes were examined, girls reported more negative outcomes. They had higher levels of emotional problems and lower levels of emotional well‑being and life satisfaction than boys. Girls, compared to boys, also indicated that they more often wished they were someone else. They also had lower self-confidence, a higher number of backaches, felt depressed or low more often, and more often felt bad-tempered or irritable. The trends on these last three variables indicate that the situation has not changed over time and, on the few occasions when it has, the gender gap seems to have worsened. Furthermore, while on many internalizing/ emotional variables, boys' scores remained fairly consistent across grades, scores for girls consistently worsened.

In contrast, externalizing/behavioural outcomes favour girls, who reported fewer behavioural problems and more prosocial behaviours than boys, although the size of this gap was smaller than that for internalizing/emotional outcomes. However, for both boys and girls, externalizing/behavioural outcomes worsened across grades such that older students demonstrated more behavioural problems and showed fewer prosocial behaviours.

Perceived wealth is most related to internalizing/emotional indicators of mental health, most particularly emotional well‑being, but minimally related to externalizing/behavioural indicators. Family structure has small but consistent relationships with all indicators of mental health, with students who live with both parents having the advantage.

One of the most intriguing findings, given the recent research literature, is the small difference in findings between positive and negative indicators of mental health. Specifically, one can arrive at the same determinants of mental health irrespective of whether one models positive or negative mental health outcomes. Thus the strong gender and perceived wealth effects were distinguished in terms of internalizing/emotional as opposed to externalizing/behavioural. They were unrelated to whether the measure was positive or negative mental health, although the two types of measures could be distinguishing different adolescents and different underlying causes.

While the gender and perceived wealth differences are a cause of concern, so too is the number of students whose mental health is not as strong as one would hope. Regardless of grade, gender, or year of survey, between 14% and 27% of students reported feeling irritable or bad-tempered more than once a week, and between 19% and 38% reported feeling depressed or low more than once a week. Regardless of survey year, almost half of Grade 8 students and more than half of Grade 10 students reported monthly backaches. By Grades 9 and 10, only about 1 in 4 boys and 1 in 5 girls indicated they had high self-confidence, while just over 20% of boys and about 30% of girls across all grades often wished they were someone else. Furthermore, between 14% and 20% of students graded their life satisfaction as 5 or less on a 10-point scale. It is necessary to try to find ways to ameliorate gaps in mental health across gender, grade, perceived wealth, and family structure. It is even more necessary to address mental health comprehensively to help the significant minority of students who are struggling with mental health issues.

How might we reach this goal? We suggest a three-pronged approach. First, Canadians need to recognize the significance of mental health issues in our current society. They must understand that young people experience poor mental health and/or lack positive mental health at relatively young ages. It is not sufficient or prudent to wait until individuals reach adulthood or they reach a crisis situation before we take their mental health seriously. Second, Canadians need to understand that mental health is gender and age-related. It has positive and negative, internal and external aspects, with these distinctions especially important for better understanding how different people experience their lives. Finally and most critically, Canadians all need to take responsibility for the mental health of adolescents. Responsibility lies not just with schools, parents, or even just adults, but with all of us to actively listen to the voices and input of youth themselves.

References

• Benson, P. L., & Scales, P. C. (2009). The definition and preliminary measurement of thriving in adolescence. Journal of Positive Psychology, 4:85-104.
• Buote, D. (2009). Social-emotional health in middle childhood: What we know. Ottawa, ON: Public Health Agency of Canada.
• Canadian Institute for Health Information [CIHI]. (2010). Recognizing and exploring positive mental health. Ottawa, ON: Author.
• Fergus, S. & Zimmerman, M.A. (2005). Adolescent resilience: A framework for understanding healthy development in the face of risk. Annual Review of Public Health, 26:399-419.
• Friedli, L. (2009). Mental health, resilience and inequalities. Copenhagen, Denmark: World Health Organization—Europe.
• Hopwood, C.J. & Grilo, C.M. (2010). Internalizing and externalizing personality dimensions and clinical problems in adolescents. Child Psychiatry and Human Development, 41:398-408.
• Keyes, C.L.M. (2006). Mental health in adolescence: Is America's youth flourishing? American Journal of Orthopsychiatry, 76:395-402.
• Keyes, C.L.M., Dhingra, S.S. & Simoes, E.J. (2010). Change in level of positive mental health as a predictor of future risk of mental illness. American Journal of Public Health, 100:2366-2371.
• Krueger, R.F. & Markon, K.E. (2011). A dimensional-spectrum model of psychopathology: Progress and opportunities. Archives of General Psychiatry, 68:10-11.
• Public Health Agency of Canada [PHAC]. (2009). Consultation on evidence for developing social-emotional protective factors for children, youth, and families. Ottawa, ON: Author.
• Scales, P.C. (1999). Reducing risks and building developmental assets: Essential steps for promoting adolescent health. Journal of School Health, 69:113-119.
• Schonert-Reichl, K.A., Stewart Lawlor, M., Oberle, E. & Thomson, K. (2009). Identifying indicators and tools for measuring social-emotional healthy living: Children ages 5-12 years [Spring 2009 Final Report]. Vancouver: University of British Columbia.
• Strauss, R. (2007). A positive perspective on adolescent mental health. Continuing Medical Education, 25:232-235.

1. Although definitions vary, positive mental health is generally seen as including emotion (affect/feeling), cognition (perception, thinking, reasoning), social functioning (relations with others and society), and coherence (sense of meaning and purpose in life) (Friedli, 2009, p. 2).
2. The actual question has six points, anchored at one end by "definitely not like me" (1) and anchored at the other end by "definitely like me" (6). We gave the term "somewhat like me" to the responses of 5.