The changing suicide pattern in Canadian adolescents and youth, compared to their American counterparts.

Leenaars, Antoon A. ; Lester, David

Many nations experienced a rising youth suicide rate during the 1970s and 1980s (Lester, 1988). This was especially true for Canada where the suicide rate for youth rose to become almost as high as that for the elderly, traditionally the age group with the highest suicide rate (Leenaars & Lester, 1990). Clark (1962), Cunliffe (1974), Lipset (1990) and others have recommended that an understanding of such social patterns in nations can often be gleaned from cross-cultural comparisons, especially if the nations are similar. Canada and the United States are obvious units of comparison; they are in close proximity and have similar language and cultural backgrounds. They have also been open to continued comparisons, e.g., art, literature, politics, and religion (Lipset, 1990). A neglected area of comparison, even in encyclopedic volumes, is suicide (Lipset, 1990).

A recent series of studies have examined the differences in suicide rates between Canada and the United States (Domino & Leenaars, 1989; Leenaars, 1989; 1992ab; Leenaars & Domino, 1993; Leenaars & Lester, 1990, 1992ab, 1994ab; Leenaars, Yang, & Lester, 1993). It has been found that Canada's rate is higher, and the relation between suicide and homicide in these two countries is different in some ways. Although level of knowledge are the same in both countries, as is the content of suicide notes, attitudes toward suicide are strikingly different. Canadians see suicide as being related more to mental illness and as a cry for help, and they more strongly endorse the right to die. Americans see suicide as being related more to religion and moral evil. In particular, Canadian youth see suicide as a more normal way to cope with problems.

However, little research has addressed the fact that the suicide rate among Canada's youth, especially for males, is higher than that of the United States. For example, in 1988 the suicide rate for men aged 74 or older in Canada was 30.6 per 100,000 followed by 29.2 for those aged 25-34, 28.0 for those aged 55-64, and 26.9 for those aged 15-24 (Lester, 1994). In contrast, the suicide rate for men in the United States aged 75 or older in 1988 was 57.8, and for youths aged 15-24 was 21.9. The present study is an attempt to understand these patterns from a societal perspective.

Study of the variation in societal suicide rates is still guided primarily by the theory proposed by Durkheim (1897) - that the rate of suicide is affected by two social characteristics. The rate is high when the degree of social integration, i.e., the extent to which members of the society are bound together in social relationships is very low (leading to egoistic suicide) or very high (leading to altruistic suicide). Suicide rates are also high when the degree of social regulation, i.e., the degree to which the desires and behaviors of the members of the society are controlled by societal norms and customs is very low (leading to anomic suicide) or very high (leading to fatalistic suicide).

Modern sociologists have noted that altruistic and fatalistic suicides are rare in modern industrialized societies and that it is often difficult to measure empirically social regulation apart from social integration. Thus, one modern version of Durkheim's theory (Johnson, 1965) states that societal suicide rates are higher when the degree of social integration/regulation is lower.

Marriage, births, and divorces have been the most common indices of social integration used in sociological research into suicide. Marriage and children have been thought to increase the level of social integration, while divorce decreases the level (Durkheim, 1897; Stack, 1992). Thus, the present study examined the relationship to suicide of these three social indicators in Canada and the United States.

However, economic conditions have also been thought to play a role (Henry & Short, 1954). A review by Platt (1984) indicated that unemployment was strongly associated with suicide rates, both in aggregate and individual studies and in cross-sectional and time-series research designs. Therefore, unemployment was also examined for its impact on the suicide rates.

Leenaars, Yang, and Lester (1993) analyzed the rates of birth, divorce, marriage, and unemployment in comparison to rates of suicide in the general population from 1950 to 1985 in Canada and the United States. Results generally supported Durkheim's theory in the general population except for the variable of marriage; there was no association between marriage and suicide in Canada, whereas in the United States, marriage has a protective effect once other variables are taken into account. Divorce rates were positively associated and birth rates were negatively associated with suicide in both countries. Unemployment rates added little predictive power to the rates. However, that study raised the question of whether the relationship of these domestic and economic factors to the total suicide rate would also apply to the rising rate of suicide among youth in Canada. In fact, there is evidence that the social and economic predictors of suicide vary depending upon the particular subgroup of the society studied - men or women and young or old (Yang, 1992). The present study was designed, therefore, to explore the social and economic correlates of the suicide rate in Canada over time, and whether these correlates are different for the young than for the elderly, and whether Canada's pattern differs from that of the United States.

METHOD

Data on the suicide rates for those aged 15-24, 25-34, 35-44, 45-54, 55-64, 65-74, and 75 and older for Canada were obtained for the period 1965 to 1985 from the World Health Organization (annual). Data on the marriage, divorce and birth rates were obtained from the United Nations (annual), and on the unemployment rate from the International Labor Office (annual). A comparable set of data was also obtained from these sources for the United States.

The data were subjected to a simple Pearson correlational analysis and a time-series regression analysis using the Cochrane-Orcutt technique to correct for the serial autocorrelation in the data (Doan, 1990).

RESULTS

The results of the simple correlational analysis for Canada are shown in Table 1. It can be seen that unemployment, divorce, and birth rates were associated with male suicides rates more strongly than was the marriage rate. These associations were also stronger for adolescents and youth and for the elderly. For these groups, suicide rates were higher in years when unemployment was higher, divorce was more common, and births less common. These results are consistent with those reported by Stack (1981) for the total United States population. For women, again, youth suicide rates resembled elderly suicide rates in being associated with divorce and birth rates, while the suicide rates of middle-aged adults were associated with marriage rates. The results for the United States, as shown in Table 1, were similar except that the correlates of suicide rates were present for almost all age groups.

Table 1
Suicide Rates and Social Indicators over Time 1965-1985:
Correlational Analysis

 unempl divorce marriage birth
Canada
males:
15-24 0.82(*) 0.96(*) -0.23 -0.91(*)
25-34 0.88(*) 0.93(*) -0.37 -0.85(*)
35-44 0.59(*) 0.74(*) 0.18 -0.77(*)
45-54 0.38 0.57(*) 0.26 -0.59(*)
55-64 -0.13 -0.31 0.03 0.30
65-74 0.33 0.35 -0.24 -0.22
75+ 0.57(*) 0.48(*) -0.48(*) -0.33

females:
15-24 0.41 0.74(*) 0.27 -0.75(*)
25-34 0.16 0.42 0.61(*) -0.57(*)
35-44 0.06 0.21 0.62(*) -0.41
45-54 -0.17 0.15 0.71(*) -0.25
55-64 -0.17 0.09 0.47(*) -0.12
65-74 0.28 0.52(*) 0.32 -0.57(*)
75+ 0.44(*) 0.57(*) -0.09 -0.55(*)

USA
males:
15-24 0.78(*) 0.98(*) 0.35 -0.85(*)
25-34 0.82(*) 0.97(*) 0.27 -0.83(*)
35-44 0.52(*) 0.32 -0.28 -0.34
45-54 -0.59(*) -0.81(*) -0.37 0.51(*)
55-64 -0.74(*) -0.94(*) -0.47(*) 0.71(*)
65-74 -0.56(*) -0.70(*) -0.29 0.52(*)
75+ 0.24 0.13 -0.15 -0.03

females:
15-24 0.57(*) 0.81(*) 0.41 -0.82(*)
25-34 -0.25 -0.16 0.01 -0.09
35-44 -0.60(*) -0.58(*) -0.04 0.32
45-54 -0.57(*) -0.54(*) -0.21 0.23
55-64 -0.61(*) -0.66(*) -0.15 0.34
65-74 -0.58(*) -0.72(*) -0.36 0.51(*)
75+ -0.44(*) -0.48(**) -0.36 0.22

* significant at the two-tailed 5% level or better

The regression analysis is shown in Table 2. For Canadian men, the power of the social and economic variables in predicting the suicide rate (indicated by the [r.sup.2]) was greatest for youth suicide rates and decreased for the older age groups. The same trend was found for Canadian women, except that the power of the variables rose for elderly women. For the United States, the power of the variables decreased with age for men, as in Canada, but the decrease for women with age was much less pronounced.

DISCUSSION

The present study has shown, therefore, that the correlates of Canadian youth suicide rates (unemployment, divorce, and birth rates) are similar to those for Canadian elderly and different from those for middle-aged Canadians. In addition, the social and economic indicators were more powerful in predicting youth suicide rates than in predicting adult and elderly suicide rates. For the United States, similar social variables predicted the suicide rates of men and women of all ages.

Table 2
Suicide Rates and Social Indicators over Time 1965-1985:
Regression Analysis (b coefficients shown)

 unempl divorce marriage birth [R.sup.2]

Canada
males:
15-24 -0.28 4.45(*) -0.95 -2.29(*) 0.94
25-34 0.65 2.03 -0.83 -1.33 0.89
35-44 0.49 2.09 2.76(*) 0.06 0.71
45-54 0.24 1.80 2.31 0.21 0.49
55-64 0.93 -1.51 1.55 0.86 0.18
65-74 -0.03 0.73 0.33 -0.43 0.20
75+ 1.24 2.98 -1.16 2.10 0.49

females:
15-24 0.01 0.04 0.58 0.94 0.84
25-34 0.08 0.46 1.94(*) -0.37 0.78
35-44 0.24 0.09 2.14(*) -0.11 0.54
45-54 -0.03 2.70(*) 3.00(*) 1.01 0.69
55-64 -0.09 2.13 1.89 0.79 0.41
65-74 -0.13 0.43 0.96 -0.72 0.73
75+ -0.17 -0.14 -0.41 -0.77(*) 0.57

USA
males:
15-24 -0.01 4.12(*) -0.50 0.01 0.95
25-34 0.13 3.39(*) -0.91(*) 0.06 0.96
35-44 0.20(*) -0.31 -0.43 -0.23 0.55
45-54 0.52(*) -0.87 0.22 -0.23 0.87
55-64 0.18 -1.86 -0.13 -0.24 0.95
65-74 0.18 -1.59 0.41 0.05 0.61
75+ -0.07 0.83 -0.41 -0.83 0.34

females:
15-24 -0.01 0.37 -0.27 0.07 0.75
25-34 0.12 1.07 -0.04 0.10 0.78
35-44 0.21 1.34 0.54 0.23 0.89
45-54 0.05 1.87 0.04 0.03 0.87
55-64 0.23 0.66 0.56 -0.31 0.84
65-74 0.08 -0.94 -0.69 0.06 0.63
75+ -0.03 -0.60 -0.25 -0.42(*) 0.52

* significant at the 5% level or better

The particular predictors identified here for youth suicide rates are in agreement with previous research and theory. The association between male Canadian youth suicide rates and unemployment rates is consistent with previous research findings (Platt, 1984) and theory (Henry & Short, 1954). The associations between youth suicide rates and measures of domestic integration (divorce and birth rates) are consistent with Durkheim's (1897) theory that the level of social integration affects the suicide rate, since divorce weakens social integration while the presence of children strengthens it.

Overall, social-economic factors seem important to the understanding of suicide in both countries although, as suggested by Leenaars (1989), predictors of the suicide rate depend upon age. Other variables e.g., cultural group, may be equally important. Therefore it is important to check the validity of the associations in different nations and, at least, for age and sex. However, cause-and-effect conclusions cannot be drawn from correlational studies, and caution is warranted in attributing the causes of suicide in Canadian adolescents and youth to domestic integration or economic stress.

The obvious question arises, despite similarities, as to why there are differences between Canada and the United States. Why would measures of domestic integration (divorce and birth rates) and the economy (the unemployment rate) predict youth suicide rates more successfully than they did adult suicide rates in Canada? Why would there be less variation in the prediction of suicide by age in the United States? These are critical questions and, regrettably, there are few answers.

The suicide rate in youth has risen in many nations in the last two decades, but less so in Muslim and Roman Catholic nations (Barraclough, 1988). Leenaars, Yang, and Lester (1993) showed that marriage, a common variable used to assess social integration/regulation, had a protective effect on suicide rates in the United States, whereas there was no association between marriage and suicide in Canada. They speculated that marriage and family played a negative role in Canada's suicide rate. Trovato (1992) pointed out the role of the reduction in family ties in Canada's suicide rate among youth, but he also stressed the possible role of their declining religiosity. Public opinion polls have noted an increase in the proportion of Canadians who report having no religious affiliation, especially among youth. Domino and Leenaars (1989) have shown that Canadian youth indicated greater disagreement than did American youth with attitudinal statements linking lack of religious beliefs to suicide attempts and ideation. Leenaars and Domino (1993) have replicated that finding in the more general population. Weber (1949) had already shown that from the beginning of the twentieth century, religion has had much more influence on values and behavior in the United States. Declining religiosity would be expected to have an impact on both social integration (through a decrease in the strength of the cultural norms governing behavior). This hypothesis is difficult to study due to the absence of annual indices of religiosity in the general population, let alone by age. However, future research may be able to find ways of incorporating this variable into the analyses. Further, it would be of interest to document the relative changes in religiosity by age in both Canada and the United States during the period of changing youth suicide rates.

It has been suggested, as an alternative hypothesis, that a youth cohort effect was causing the rise in youth suicide in Canada and the United States (Holinger, 1987). It was argued that a large proportion of dependents might cause economic and social problems for other groups in that nation, problems that might be reflected in high rates of deviant behavior such as suicide. Easterlin (1980) had suggested that the size of the cohort might also have an impact on the behavior of its own cohort as well as the behavior of other cohorts. Holinger (1987), as an example, found that for youth (aged 15 to 24), the size of the cohort was positively related to the risk of dying from suicide, whereas for those aged 25 to 64, the size of the cohort was negatively related. However, Leenaars and Lester (1994) have shown that there is no youth cohort effect in Canada and the United States in the more recent data, suggesting that Holinger's findings may have been related to an historical event - WWII. Therefore, it is likely that we can dismiss the youth cohort effect as an explanation for Canada's rate of suicide in youth at this time.

We know that Canadian adolescents and youth, especially males, are at higher risk for suicide (Leenaars & Lester, 1990). For example, for boys (15-19) the suicide rate was 50 to 60% greater in the 1980s in Canada than in the United States. The only other study that addressed this question, but at a micro-level, was by Domino & Leenaars (1989). They reported that Canadian youth more frequently see suicide as more acceptable and more normal as a solution to problems than do their American counterparts. Canadian youth see suicide as a more private matter and perceive it to be "lethal" in the sense that they are more pessimistic than American youth about intervention. Domino and Leenaars (1989) concluded that Canadian youth perceive suicide as a part of everyday life. Obviously, cultures foster "life" differently, as Durkheim (1897) had already suggested. Further research is needed to understand the suicide crisis among Canadian youth.

REFERENCES

Barraclough, B. (1988). International variation in the suicide rate of 15- 24-year olds Social Psychiatry & Psychiatric Epidemiology, 23, 75-84.

Clark, S.D. (1962). The developing Canadian community. Toronto: University of Toronto Press.

Cunliffe, M. (1974). New world, old world: The historical antithesis. In R. Rose (Ed.), Lesson from America. London: MacMillan.

Doan, T. (1990). Regression analysis of time series. Evanston, IL: Var Econometrics.

Domino, G., & Leenaars, A. (1989). Attitudes towards suicide: A comparison of Canadian and United States college students. Suicide & Life-Threatening Behavior, 19, 160-172.

Durkheim, E. (1897). Le suicide. Paris: Felix Alcan.

Easterlin, R. (1980). Birth and fortune. New York: Basic Books.

Henry, A., & Short, J. (1954). Suicide and homicide. Glencoe, IL: Free Press.

Holinger, P. Violent deaths in the United States. New York: Guilford Press.

International Labor Office. Year book of labor statistics. New York: International Labor Office, annual.

Johnson, D. (1965). Durkheim's one cause of suicide. American Sociological Review, 30, 875-886.

Leenaars, A. (1989). Rates and patterns of suicide for Canada and the United States: Their implications for socio-political action. Presidential address presented at Canadian Association for Suicide Prevention Conference, Toronto, Ontario.

Leenaars, A. (1992a). Suicide notes from Canada and the United States. Perceptual and Motor Skills, 74, 278.

Leenaars, A. (1992b). Suicide in Canada: Our responsibility. Plenary address presented at Canadian Association for Suicide Prevention Conference, Saskatoon, Sask.

Leenaars, A., & Domino, G. (1993). A comparison of community attitudes towards suicide in Windsor and Los Angeles. Canadian Journal of Behavioral Science, 25, 253-266.

Leenaars, A., & Lester, D. (1990). Suicide in adolescents: A comparison of Canada and the United States. Psychological Reports, 67, 867-873.

Leenaars, A., & Lester, D. (1992a). A comparison of rates and patterns of suicide in Canada and the United States, 1960-1988. Death Studies, 16, 433-446.

Leenaars, A., & Lester, D. (1992b). Facts and myths of suicide in Canada and the United States. Journal of Social Psychology, 132, 787-789.

Leenaars, A., & Lester, D. (1994a). Suicide and homicide in Canada and the United States. Suicide & Life-Threatening Behavior, 24, 184-191.

Leenaars, A., & Lester, D. (1994b). Domestic and economic correlates of personal violence in Canada and the United States. Gionale Italiano di Suicidologia, 4, 7-12.

Leenaars, A., Yang, B., & Lester, D. (1993). The effects of domestic and economic stress on suicide rates in Canada and the United States. Journal of Clinical Psychology, 49, 918-921.

Lester, D. (1988). Youth suicide. Adolescence, 23, 955-958.

Lester, D. (1994). Suicide in the elderly. In D. Lester, & M. Tallmer (Eds.), Now I lay me down. Philadelphia, PA: Charles Press, pp. 1-16.

Lipset, S. (1990). Continental divide. New York: Routledge.

Platt, S. (1984). Unemployment and suicidal behavior. Social Science and Medicine, 19, 93-115.

Stack, S. (1981). Divorce and suicide. Journal of Family Issues, 2, 77-90.

Stack, S. (1992). Marriage, family, religion and suicide. In R. Maris, A. Berman, J. Maltsberger, & R, Yufit (Eds.), Assessment and prediction of suicide. New York: Guilford.

Trovato, F. (1992). A Durkheimian analysis of youth suicide: Canada, 1971 and 1981. Suicide & Life-Threatening Behavior, 22, 413-427.

United Nations. Demographic yearbook. New York: United Nations, annual.

Weber, M. (1949). The methodology of social sciences. Glencoe, Ill: The Free Press. (Original published in various years, early 20th century.)

World Health Organization. Statistics annual. Geneva, Switzerland: World Health Organization.

Yang, B. (1992). The economy and suicide. American Journal of Economics and Sociology, 51, 87-99.

Antoon Leenaars, Ph.D., University of Leidon, The Netherlands.

David Lester, Ph.D., Center for the Study of Suicide, Blackwood New Jersey.