Depression among adolescents in Northern Ireland.

Donnelly, Michael

The present research attempts to describe the nature, frequency, and pattern of self-reported depression among a sample of secondary school students living in Northern Ireland. The study focuses on adolescents in the community rather than on those who have been referred for professional help since community surveys provide important data on the distribution of psychiatric problems in the population. Many psychiatrically disturbed youngsters are never referred to professional services and, therefore, studies of unreferred populations are required in order to achieve a complete understanding of the nature and frequency of depression in adolescence (Angold, 1988).

Compared with research into adult psychiatric and psychological conditions there have been few community studies of the prevalence of adolescent psychiatric disorder and, more specifically, fewer again of depression. In a series of studies, Rutter, Tizard and Whitmore (1970/1981) made multiple assessments of children and adolescents at two ages, 10 and 14, using instruments administered to teachers, parents, psychiatrists, and the young people who participated in the research. With respect to 10-11-year old children Rutter (1986) wrote:

". . . 13% showed a depressed mood at interview, 9% appeared preoccupied with depressive topics, 17% failed to smile, and 15% showed poor emotional responsiveness . . . at age 14-15 years depressive feelings were considerably more prevalent. Over 40% of the adolescents reported substantial feelings of misery and depression during a psychiatric interview, 20% expressed feelings of self-depreciation, 7-8% reported that they had suicidal feelings, and 25% described ideas of reference."

It would seem that feelings of misery and depression including quite severe depression (even suicidal) feelings causing personal suffering are not often detected by parents, teachers, and other adults. A self-report measure such as the Child Depression Inventory (CDI; Kovacs, 1983) affords a ready opportunity for a young person to express his or her personal feelings. In addition, a number of studies have shown that youngsters' reports of their own affective states are reliable (e.g., Angold et al., 1987).

Several studies have investigated the psychological well-being of Northern Irish children (Cairns, 1987). For example, Fee (1980; 1983), using the Rutter Teacher Rating Questionnaire, reported a 15% incidence of psychiatric disturbance in 10-year-old Belfast children. This proportion was higher than the rate for children living in rural Isle of Wight (11%) but lower for children living in inner London (19%). Fee concluded that the incidence of psychiatric disturbance "is probably no worse than might have been expected in similar urban areas in other parts of the United Kingdom." There have been fewer studies of the psychological well-being of adolescents living in Northern Ireland. In one study, McWhirter (1989) found that a sample of 18-year-old adolescents participating in a longitudinal study on the transition from school to work had a similar mean score on the General Health Questionnaire as did two English cohorts of school leavers. In other words, the psychological well-being of older Northern Irish adolescents was comparable to their English counterparts.

A review of the relatively few studies of adolescent psychiatric disorder in the general population illustrate: (1) an increase in both depressive feelings and depressive disorder during adolescence (which is mirrored by a dramatic increase in both attempted and completed suicide (Shaffer, 1986)); and (2) A greater increase in the rate of depression for females than males. The adult female preponderance in depressive disorders is a well-established finding both in community and clinical studies. Although the data from hospital statistics and clinical research with young people is limited, it tends to demonstrate a similar pattern to those found in community surveys. For example, in a cross-sectional study of 547 clinic-referred children, Pearce (1978) found that depressive symptomatology increased with age, from 11% among children to 25% in adolescents, with depressive symptoms more than twice as common in adolescent girls than adolescent boys.

However, as Angold (1988) noted, further research is required in order to determine the relationships and interactions among depression, sex, and age. Studies of depression in children and adolescents have not paid detailed attention to the variables of age and sex and their influence on the developmental course of depression (Rutter, 1986). The present study seeks to extend knowledge of those issues by examining the pattern of depression over the adolescent years 11 to 15 in the general population in Northern Ireland.

METHOD

Sample Structure

Students (N = 887) in forms 1 to 4 from 6 secondary schools: 3 grammar schools, 3 intermediate schools, situated in the North West of Northern Ireland participated in the research. While the initial aim was to select a stratified random sample, in practice, the choice of subjects was determined by the availability and cooperation of pupils, principals, teachers, and parents.

The mean age of the entire sample was 13.2 years while for each form from i to 4, the average age was 11.7 (97 males; 119 females); 12.6 (123 males; 94 females); 13.6 (118 males; 102 females); and 14.7 (119 males; 115 females), respectively. Since school form would appear to be a more dominant visible marker of development in the adolescent years than chronological age (Healy & Stewart, 1984), subjects were grouped according to school form for analyses. In terms of religious affiliation, 55% of the sample were Protestants and 42% were Catholics. The composition of the sample according to the current occupation of each respondent's father was as follows: Professional - 4%; Intermediate - 25%; Skilled - 39%; Semi-skilled/Unskilled - 8%; and Unemployed - 23%;.

Instrument

Depression was measured by the Child Depression Inventory (CDI), a 27-item, self-report scale suitable for school-aged children and adolescents (8 to 17 years old). Kovacs (1983) studied the validity of the CDI and found that it discriminated between young people with a major depressive disorder and "normal" adolescents, and that it was significantly correlated (r = 0.60) with clinicians' ratings of depression. The CDI is a downward version of the Beck Depression Inventory. Each of the 27 items consists of three statements graded in severity from 0 to 2. A sample item is as follows: "I am sad once in a while" (0); "I am sad many times" (1); and "I am sad all the time" (2). Scores range from 0 to 54, with a recommended cut-off point of 13 for designating a depressed case (Kovacs, 1983). The CDI is the most thoroughly researched self-report measure of childhood and adolescent depression, and its reliability and validity has been documented in several studies (e.g., Finch, Saylor, & Edwards, 1985; Smucker Craighead, & Green, 1986). The CDI was group-administered in classroom settings. Students were assured of the anonymity and confidentiality of their responses.

RESULTS

The distribution of CDI scores for the entire sample was positively skewed. This pattern of scores is in agreement with the findings of other large-scale random samples of school children (Kovacs, 1983).

In the cross-sectional analysis a cut-off point of 13 on the CDI was used to designate depression as recommended by Kovacs (1983). A higher discriminating cut-off point of 17 also was employed. This was equivalent to the mean CDI score plus one standard deviation for the entire sample. Due to the skewedness of the distribution of CDI scores, the results of the following parametric tests should be viewed with caution. Previous research with the CDI has tended to report the results of parametric analysis. The present study has done likewise so that the results can be compared with the findings of other research.

Table I gives descriptive statistics for the various school forms by sex subgroups as well as for the entire sample. The presentation of normative data should prove useful to other researchers and clinicians who are planning to use the CDI to assess the mental health of young people in Northern Ireland. The mean CDI score was 10.26, with a standard deviation of 6.35 and a range of 41. (The median value was 9.) For males (N = 457), the mean CDI score was 9.25, the standard deviation was 6.56, and the range was 41. For the female sample (N = 430), the mean CDI score was 10.30, the standard deviation was 6.11, and the range was 32. These results were above the normative mean score of 9.00 for the CDI (Smucker et al., 1986).

Analysis of variance on mean CDI scores indicated no main effect for school form, F (3, 879) = 1.23, p [greater than] 0.05) or sex, F (1, 879) = 0.08, p [greater than] 0.05). An examination of the mean values suggested that form and sex interact with each other. However, the resulting effect was not statistically significant, F (3, 879) - 1.63, p [greater than] 0.05).

[TABULAR DATA FOR TABLE 1 OMITTED]

The proportion of the sample judged to be "cases" according to a cutoff point of 13, did not increase with school form. On the whole, higher frequencies of depressed cases were found in forms 3 and 4 than in forms 1 or 2, but these were not statistically significant.

Chi-square analyses did not reveal significant associations between the four school forms controlling for the variable of sex (Males: [[Chi].sup.2] = 2.96, df= 3, p [greater than] 0.05; Females: [[Chi].sup.2] = 5.10, df = 3, p [greater than] 0.05). However, it would seem that the first year of second level education was more difficult for boys than girls. Over 30% reported being depressed compared with 20% of first-year girls. After form I the number of depressed males decreased to 23% at form 4. The opposite was the case for females. Here, there was a steady increase to over 32% at form 4. Only in form 1 did boys outnumber girls in terms of the percentage of depressed cases.

When a more conservative score of 17 was used to designate a case, a similar pattern was observed. In forms 1 and 2, more boys than girls scored above the cut-off point, but by form 4, a significantly higher proportion of girls scored above the cut-off point than did boys ([[Chi].sup.2] = 3.86, df = 1, p [less than] 0.05). Approximately 35% of young people whose fathers were unemployed and a similar proportion whose fathers had few formal skills, scored above 13 on the CDI scale. This is more than double the rate reported for young people in the professional group (14%) and significantly higher than the rate found in either the intermediate (23%) or skilled (24%) groups ([[Chi].sup.2] = 14.61, df = 5, p [less than] 0.05). Comparisons of the proportions of depressed adolescents (using either cut-off point 13 or 17) in the social groups by school form and sex did not reveal significant associations.

Analysis of CDI Items

Table 2 presents the data for the CDI items as the percentage of subjects who endorsed item values 0, 1, or 2. It will be recalled that the CDI items consisted of three sentences. The young person was asked to pick the one sentence which best described how they were feeling. Each sentence was scored 0, 1, 2 in the direction of increasing severity. For example, the first item on the CDI scale found in Table 2 revealed that 86% of the sample were sad ". . . once in a while"; 13% were sad ". . . many times"; and 1% were sad ". . . all the time." As might be expected, most score distributions per item were rather positively skewed, with the majority of subjects endorsing 0 and few subjects scoring 2.

The items with the greatest percentage of adolescents endorsing the most severe version/sentence, that is, those who scored 2, were items 13 and 15. These concerned "indecisiveness" and "motivation due to school work," respectively. Around 20% of young people reported that they "could not make up their mind about things" and that they "had to push themselves all the time to do their school work." Approximately 11% reported that "most days, [they] do not feel like eating" and a similar proportion also believe that they "can never be as good as other youngsters." A lower percentage of young people (8% or less) endorsed the most severe version of the remaining items.

[TABULAR DATA FOR TABLE 2 OMITTED]

A number of items merit attention. While around 4% reported that they wanted to kill themselves, 33% stated: "I think about killing myself but I would not do it." As might be expected with this age range, concerns about the future occupied the minds of 56% of the sample. Around 45% worried about "bad things happening to them." Also characteristic of adolescent development, 56% were unhappy about the way they looked, while 30% did not like themselves. Between 20 and 30% had thoughts concerning loneliness, not having fun at school, not feeling loved, trouble sleeping, worry about aches and pains, tiredness, doing badly in school subjects, and general worry or anxiety. Approximately 14% endorsed items 1 and 10 concerning the degree of sadness experienced or frequency of reported crying.

Associations between CDI Items and School Form and Sex

The degree of association between each of the 27 CDI items and (a) school form and (b) sex was examined by means of chi-square analyses.

A higher proportion of females than males endorsed the item "I do not like myself." This was a consistent result across the school years with statistically significant associations found in form 2 ([[Chi].sup.2] = 18.0, df = 2, p [less than] 0.001); form 3 ([[Chi].sup.2] = 5.99, df = 2, p [less than] 0.05) and form 4 ([[Chi].sup.2]= 8.03, df = 2, p [less than] 0.01).

More females also reported that they ". . . feel like crying . . .," an association which was statistically significant in form 3 ([[Chi].sup.2] = 7.85, df = 2, p [less than] 0.01).

A significantly higher number of females than males in form 1 ([[Chi].sup.2] = 12.92, df = 2, p [less than] 0.001); form 2 ([[Chi].sup.2] = 13.31, df = 2, p [less than] 0.001); form 3 ([[Chi].sup.2] = 3.27, df = 2, p [greater than] 0.19); and form 4 ([[Chi].sup.2] = 8.47, df = 2, p [less than] 0.01) reported that there were bad things about their looks or that they looked ugly.

Difficulties with appetite became appreciably greater with age. In form 3, 16% of females compared with 7% of males, and in Form 4, 18% of females and 1% of males reported that on most days they did not feel like eating ([[Chi].sup.2]= 6.17, df = 2, p [less than] 0.05; ([[Chi].sup.2] = 27.47, df = 2, p [less than] 0.001).

There was also a tendency for more females than males to report ". . . worry about aches and pains." This sex difference reached statistical significance in form 2 ([[Chi].sup.2] = 9.43, df = 2, p [less than] 0.01), and form 4 ([[Chi].sup.2] = 12.27, df = 2, p [less than] 0.01).

More males than females endorsed the most extreme version of item 9: "I want to kill myself." Statistically significant associations were found in form 2 ([[Chi].sup.2] = 6.13, df = 2, p [less than] 0.05) and form 3 ([[Chi].sup.2] = 5.69, df = 2, p [less than] 0.05).

In both forms 2 and 3 a significant association was found between being male and "not wanting to be with people at all," ([[Chi].sup.2] = 6.52, df = 2, p [less than] 0.05; [[Chi].sup.2] = 6.96, df = 2, p [less than] 0.05). Higher proportions of males in form 1 and form 4 also expressed this asocial attitude, but the association was not statistically significant.

Finding motivation to do school work appeared to be a problem for a higher proportion of males. In forms 1 to 3 significantly more males endorsed the item: "I have to push myself all the time to do my school work."

More males than females tended to have trouble with school work. This was particularly so in form 1 ([[Chi].sup.2] = 14.9, df = 2, p [less than] 0.05) where a significantly higher number of boys reported ". . . doing badly in subjects they used to be good in" and form 2 ([[Chi].sup.2] = 4.9, df = 2, p [greater than] 0.08) where a similar but nonsignificant result was found.

DISCUSSION

When discussing these results it is important to bear in mind the criterion used for designating a depressed case. Kovacs (1983) explored a number of cut-off points and concluded: ". . . the criterion score of 13 will miss about 50% of the clinically depressed cases, [but] it will yield the maximal true negative rate and the highest diagnostic confidence." (Kovacs explored a cut-off point of 13 because children clinically diagnosed as having a major depressive disorder had a mean score of 13.63.)

Other researchers have used different methods. For example, Helsel and Matson (1984) selected depressed subjects on the basis of one standard deviation greater than the mean CDI score. In the present sample this calculated to a cut-off point of 17. Using this cut-off point meant that around 12% of the sample were designated depressed cases. This proportion of cases increased to approximately 27% when the lower cut-off point of 13 was employed.

In an examination of the normative properties of the CDI, Smucker et al., (1986) selected the top 10% of the distribution as depressed. This, in effect, meant a cut-off score of 19. In the present study, less than 10% scored above 19 on the CDI scale.

This lack of agreement in the utilization of the CDI scores is problematic for investigations that seek to provide comparative data on the prevalence of depression among children and adolescents. In addition to variation in case rates due to differences in the way depression is classified, prevalence rates may vary according to the mode of assessment employed.

Using Kovacs (1983) criterion-tested cut-off point of 13 on the CDI scale to identify depressed adolescents there was a higher rate of depression among Northern Irish adolescents (27%) than was found with other "normal" populations. For example, Kovacs (1983) found that approximately one-fifth of young people scored above 13 on the CDI, and a similar proportion has been reported by Worchel et al. (1988) and Reinherz et al. (1990). However, using a score of 17 as the cut-off point, approximately 12% of the sample of young people studied by Kovacs and an almost identical proportion in the present study were designated cases. Brown, Fitzgerald, & Kinsella (1990) found a similar prevalence rate (15%) for psychological distress (as assessed by the Youth Self-Report Questionnaire) in a group of 15-16 year-old girls living in inner city Dublin. In addition, Ehrenberg, Cox, & Koopman (1990), using the Beck Depression Inventory, reported 20% of 13- 15 year-old adolescents were mildly depressed (BDI score: 11 to 16) and 12% were clinically depressed (BDI score: 17 or higher).

It may be that more adolescents in Northern Ireland than those living elsewhere report low or mild levels of depression or distress (as measured by the CDI) while comparable rates exist for severe depression. The higher reported frequency of "mild" depression or distress among Northern Irish adolescents may be associated with living in a society that appears unable or unwilling to resolve long-standing differences which often lead to violence. In a review of the psychiatric aspects of violence in Northern Ireland, Cairns and Wilson (1985) suggested that "it is unlikely that the political violence has caused any marked increase in serious psychiatric illness but rather stimulated an increase in normal anxiety, particularly among the more vulnerable and especially those with a previous history." Given the significant association found between the type of social group (as indicated by father's occupation) and depression, further investigations should address the set of relationships encompassing socioeconomic disadvantage, political violence, and the health of young people in Northern Ireland.

The results also suggest that the pattern of depression in adolescence may be represented by the notion of "peak stressful periods" (Coddington, 1972; Newcomb, Huba, & Bentler, 1986) which occur at certain transitional times. It is possible that the change from primary to secondary level may be a more difficult time for boys than girls. Some researchers have suggested that more pressure is put on boys to succeed academically (e.g., Gove & Herb, 1974; Maccoby & Jacklin, 1974; and Healy & Stewart, 1984). This may be all the more pertinent in Northern Ireland where grammer school entry is via selection tests which admit only 25 to 30% of the population of 11-year-olds. Form 3 would appear to be the peak age for adolescent female depression. Although quite difficult to assess, it is impossible to ignore the potential psychological effects which appear to accompany the development of female secondary sexual characteristics. In form 4 a significantly higher proportion of females scored above the stronger discriminating score of 17 on the CDI scale. This is similar to research by Teri (1982) who found that significantly more females scored more than one standard deviation above the mean on the BDI, and Ehrenberg et al. (1990) who reported that 15% of females and 9% of males were clinically depressed.

The detailed analysis of CDI items indicated that depression appears to be expressed differently by each sex. There were a number of significant associations between sex and depressive symptomatology. Adolescent males tended to endorse what might be described as behavioral problems or symptomatology while females tended to report a constellation of items of an emotional type. Females, for example, reported relatively more loss of appetite, crying, worry about aches and pains, and low self-esteem (particularly concerning physical appearance). Adolescent males, on the other hand, tended to report not wanting to be with people, trouble with motivation and school work, and suicidal tendencies.

The findings confirm the need for further attention to the variables of age and sex when considering appropriate intervention strategies for adolescents experiencing depression. Additional research is required in order to assess the impact of political violence and socioeconomic deprivation on the mental health of young people living in Northern Ireland. Finally, it is encouraging to note that the large majority of adolescents were not depressed or distressed and appeared to be coping well.

REFERENCES

Angold, A. (1988). Childhood and adolescent depression: Epidemiological and etiological aspects. British Journal of Psychiatry, 152, 601-617.

Angold, A., Weisman, M.M., John, K., Merikangas, K.R., Prusoff, B.A., Wickramaratre, P., & Gammor, G.D. (1987). Parent and child reports of depressive symptoms. Journal of Child Psychology and Psychiatry, 28, 901-915.

Brown, K., Fitzgerald, M., & Kinsella, A. (1990). Prevalence of psychological distress in Irish female adolescents. Journal of Adolescence, 13, 341-350.

Cairns, E. (1987). Caught in the crossfire: Children and the Northern Ireland conflict. Belfast: Appletree Press.

Cairns, E., & Wilson, R. (1985). Psychiatric aspects of violence in Northern Ireland. Stress Medicine, 1, 193-201.

Coddington, R. (1972). The significance of life events as etiological factors in the diseases of children: A study of a normal population. Journal of Psychosomatic Research, 16, 205-213.

Ehrenberg, M.F., Cox, D.N., & Koopman, R.F. (1990). The prevalence of depression in high school students. Adolescence, 25(100), 905-912.

Fee, F. (1980). Responses to a behavioral questionnaire of a group of Belfast children. In J. Harbison, & J. Harbison, (Eds.), A society under stress: Children and young people in Northern Ireland. Somerset: Open Books.

Fee, F. (1983). Educational change in Belfast school children 1975-1981. In J. Harbison (Ed.), Children of the troubles. Belfast: Stranmillis LRU.

Finch, A.J., Saylor, C.F., & Edwards, G.L. (1985). Children's Depression Inventory: Sex and grade norms for normal children. Journal of Consulting and Clinical Psychology, 53(3), 424-425.

Gove, W.R., & Herb, J.R. (1974). Stress and mental illness among the young. Social Forces, 53, 256-265.

Healy, J., & Stewart, A.J. (1984). Adaptation to life changes in adolescence. In P. Karoly, & J.J. Steffen, (Eds.), Adolescent behavior disorders: Foundations and contemporary concerns. Lexington, MA: D.C. Health and Co.

Helsel, W.J., & Matson, J.L. (1984). The assessment of depression in children: The internal structure of the Child Depression Inventory (CDI). Behavior Research and Therapy, 22(3), 289-298.

Kovacs, M. (1983). The Children's Depression Inventory: A self-rated depression scale for school-aged youngsters. University of Pittsburg: (unpublished manuscript).

Maccoby, E., & Jacklin, H.J. (1974). The psychology of sex differences. Palo Alto, CA: Stanford University Press.

McWhirter, L. (1989). Longitudinal evidence on the teenage years. In J. Harbison (Ed.), Growing up in Northern Ireland. Belfast: Stranmillis LRU.

Newcomb, M.D., Huba, G.J., & Bentler, P.M. (1986). Desirability of various life change events among adolescents: Effects of exposure, sex, age and ethnicity. Journal of Research in Personality, 20, 207-227.

Pearce, J. (1978). The recognition of depressive disorder in children. Journal of the Royal Society of Medicine, 71, 494-500.

Reinherz, H.Z., Frost, A.K., Stewart-Berghauer, G., Pakiz, B., Kennedy, K., & Schille, C. (1990). The many faces of correlates of depressive symptoms in adolescents. Journal of Early Adolescence, 10, 455-471.

Rutter, M., Tizard, J., & Whitmore, K. (Eds.). (1970). Education, health and behavior. London: Longman, (Reprinted 1981).

Rutter, M. (1986). The developmental psychopathology of depression. In M. Rutter, C.E. Izard, & P.B. Read (Eds.), Depression in young people: Developmental and clinical perspectives. New York: Guilford Press.

Shaffer, D. (1986). Developmental factors in child and adolescent suicide. In M. Rutter, C.E. Izard, & P.B. Read (Eds.), Depression in young people. New York: Guilford Press.

Smucker, M.R., Craighead, L.W., & Green, B.J. (1986). Normative and reliability data for the CDI. Journal of Abnormal Child Psychology, 14, 25-39.

Teri, L. (1982). Depression in adolescence: Its relationship to assertion and various aspects of self-image. Journal of Clinical Child Psychology, 11, 101-106.

Worchel, F.F., Nolan, B.F., Wilson, V.L., Purser, J.S., Copeland, D.R., & Pfefferbaum, B. (1988). Assessment of depression in children with cancer. Journal of Pediatric Psychology, 13, 101-112.

The author wishes to express his thanks to Dr. Ronnie Wilson for his help and advice.