Ethnocultural Differences in Prevalence of Adolescent Depression

E thnocultural Differences in Prevalence of Adolescent Suicidal Behaviors

Robert E. Roberts, PhD, Y. Richard Chen, PhD, and Catherine R. Roberts, MPH, PhD

Data from an ethnically diverse sample of middle school students (grades 6-8; n = 5,423) are analyzed for ethnic differences in suicidal ideation, thoughts about suicide in the past 2 weeks, suicide plans, and suicide attempts. Ideation was examined using a four-item scale and a single item on suicidal thoughts. Ideation was higher among females, older youths, and lower status youths. The same general pattern held for recent suicidal plans and attempts, with the exception of gender, where the trend was for males to report more attempts. Lifetime plans and attempts were higher for females, older youths, and lower status youths. Data were sufficient to compare nine ethnic groups. Multivariate logistic regression analyses, adjusting for the effects of age, gender, and socioeconomic status, yielded significant odds ratios using the Anglo group as the reference, for suicidal ideation for the Mexican (OR = 1.76, p < .001), Pakistani (OR = 2.0, p < .Ol), and Vietnamese (OR = 1.48, p < .05) American groups. For thoughts about suicide in the past 2 weeks, only Pakistani and Mixed Ancestry youths had elevated risk. For suicidal plans in the past 2 weeks, Mixed Ancestry youths (OR = 2.02, p < .05) and Pakistani youths (OR = 3.20, p < .01) had elevated risk. For recent attempts, only the Pakistani American youths had elevated risk (OR = 3 . 1 9 , ~ < .01). Future research needs to address whether these results hold in other ethnically diverse communities and, if so, what factors contribute to increased risk among some minority youth and not others.

Ethnicity is a central theme of the Ameri- can experience. In fact, Alba (1990) ar- gues that ethnic differences form a possi- bly permanent substructure, if not the ultimate bedrock, of American society. Yet in spite of the acknowledged role of race and ethnicity in the United States, and the increasing diversity of our population, there have been remarkably few epidemiological studies of adolescent mental health in the diverse groups that make up American society.

A case in point is suicidal behaviors. Suicide is a major public health problem in the United States. For example, suicide is the third leading cause of death for per- sons 15-24 years old (the rate was 13.1 per

100,000 in 1991; National Center for Health Statistics, 1994). Suicidal behavior has been reported as a leading cause of psychiatric emergencies among children and adolescents (Robinson, 1986) and one of the strongest predictors of psychiatric hospital admissions for adolescents (Hil- lard, Slomowitz, & Deddens, 1988).

In the United States, as well as world- wide, suicide rates differ by age, gender, race or ethnicity, marital status, and socioeconomic status (MoScicki, 1995). Most of these studies have had adults as their focus. While there have been studies of suicidal behaviors of adolescents (see reviews by Reynolds & Mazza, 1994; Shaf- fer & Hicks 1994), there are virtually no

Robert E. Roberts is with the Department of Behavioral Sciences, School of Public Health, and Depart- ment of Psychiatry and Behavioral Sciences, Medical School. The University of Texas-Houston Health Sci- ence Center. Y. Richard Chen and Catherine R. Roberts are with the Department of Psychiatry and Behav- ioral Sciences, Medical School, The University of Texas-Houston Health Science Center. Address correspondence to Robert E. Roberts, The University of Texas School of Public Health, P.O. Box 20186, Houston, Texas 77225 USA.

r was presented at the 123rd Annual Meeting of the American Public Health Association. October 29-Gvember 2, 1995. San Diego, California. This research was supported in part by Reseaerch Grant No. MH51687 from the National Institute of Mental Health to the first author.

Suicide and Life-Threatening Behavior, Vol. 27(2), Summer 1997 208 0 1997 The American Association of Suicidology

An earlier version of this pa

ROBERTS ET AL. 209

data on suicidal behaviors among minority adolescents. This is surprising, since epidemiological studies suggest such behaviors are not uncommon. Recent community-based, epidemiological studies report prevalence of suicidal ideation ranging from about 2% to as high as 60% (Garrison, Addy, Jackson, McKeown, & Waller, 1991b). The prevalence of suicide attempts tends to be much lower, with lifetime prevalence for adolescents ranging from 3.5% to 11% (Andrews & Lewin- sohn, 1992). In general, ethnicity has been little studied in relation to suicidal behaviors. Results from the few studies that have examined ethnic differences have been equivocal, with several studies reporting no difference between White and Black adolescent suicidal behaviors (Dubow, Kausch, Blum, Reed, & Bush, 1989; Garri- son et al., 1991b) and several reporting higher suicidal behavior among Black adolescents (Garrison, Addy, Jackson, Mc- Keown, & Waller, 1991a; Harkavy Fried- man, Asnis, Boeck, & DiFiore, 1987). Lester and Anderson (1992), using data from a very small school-based sample in New Jersey, report that Hispanic students had higher scores on both depression and suicidal ideation than did African American students. Most of the Hispanic students were of Puerto Rican origin.

Vega, Gil, Zimmerman, and Warheit (1993), using data from a large school- based survey of Cuban, Nicaraguan, other Hispanic, African American, and Anglo American males in grades 6 and 7, report that African Americans had the highest prevalence of suicidal ideation in the previous 6 months. Nicaraguans and other Hispanics had the highest levels of lifetime suicide attempts.

Roberts and Chen (1995) examined suicidal ideation in a large sample of Anglo and Mexican American middle school students. Mexican Americans had rates of ideation almost twice as high as those of their Anglo counterparts. Adjusting for the effects of covariates such as age, gender, language use, and household struc- ture, Mexican American adolescents still had 1.7 times the risk of suicidal ideation.

The percentage of Mexican American youths reporting they had thought about killing themselves on 1 or more days during the past week was 25.2%. This is simi- lar to a rate of 23.4% reported by Swan- son, Linskey, Quintero-Salinas, Pumariega, and Holzer (1992) for Mexican American students.

Reynolds and Mazza (1992) assessed suicidal behaviors in a sample of junior and senior high schools in eight states. The highest rate of a history of suicide attempts was found for Native American adolescents (25.5%) and the lowest rate was among African American youths. Hispanic youths also reported a high rate (16.3%) of lifetime suicide attempts.

Garrison, Addy, Jackson, McKeown, and Waller (1991b), reporting data from a statewide survey in South Carolina, found that while suicide attempts were higher for Whites than for Blacks, Black females reported the highest frequency of attempts requiring medical treatment. Blacks were less likely to report plans, but significant ethnic differences were not found for suicidal thoughts.

Given the limited and sometimes equivocal results thus far, our purpose here is to further explore ethnic differentials in adolescent suicidal behaviors. To this end, we present prevalence rates for suicidal behaviors for nine diverse ethnic groups. For most of these groups, these are the first data ever reported on risk of suicidal behaviors. The data are from a larger study exploring the origins and conse- quences of ethnic status and other indicators of ethnicity, such as ethnic identity (Alba, 1990; Phinney, 1990), among An- glo, Asian, African, and Latino American adolescents, and the relation of ethnicity to indicators of adolescent mental health.

METHOD

Data presented are from a school-based survey conducted in a school district in the Houston metropolitan area in March 1994. The survey included five middle schools enrolling about 6,400 students in grades 6-8. Questionnaires were obtained

210 SUICIDE AND LIFE-THREATENING BEHAVIOR

from 5,496 subjects, ranging from 10-17 years of age.

The response rate was 85.3%. Nonparti- cipants (14.7%) included students absent during the first class period of the day (9.3%), when the survey was conducted, students whose parents declined for them to participate (4.0%), and students who themselves declined to participate (1.4%). Another 73 questionnaires were elimi- nated due to large numbers of missing data, resulting in 5,423 usable questionnaires. We were not able to follow up those absent on the day of the survey, either in school or at home. Since absentees tend to have higher rates of problematic behaviors, the absence of these youths undoubt- edly underestimates suicidal ideation and behaviors.

The questionnaires were administered in classroom settings, monitored by proj- ect field staff. Questionnaires were in En- glish. Passive parental and active student consent procedures were used.

Items on suicidal behaviors were part of the DSM Scale for Depression (DSD), a self-administered checklist developed by the first author from the stem questions on major depression in the Diagnostic In- terview Schedule for Children (National Institute of Mental Health, 1992).

Four symptom queries related to suicidal ideation and four related to suicidal behavior (see Table 1). The ideation items were “felt life was hopeless,“ “thought more than usual about death or dying,” “wished you were dead,” and “thought about suicide or killing yourself .” The time frame was the past 2 weeks and the response categories were (1) hardly ever or never, (2) sometimes, (3) often, and (4) almost every day. Score range is 0-12. Cron- bachs alpha was .84 overall, .80-.85 across age groups, and .78-.91 across 10 ethnocultural subgroups. There were two items that inquire whether the participants ever made a plan to kill themselves or made such a plan in the past 2 weeks (yesho), and two items inquiring whether participants have ever tried to kill themselves or have tried to kill themselves in the past 2 weeks (yeslno).

TABLE 1 Suicidal Behavior Questions

1. Have you felt that life was hopeless and that there was nothing good for you in the future? ...............................................................

2. Have you thought more than usual about death or dying? .............................................

4. Have you thought about suicide or killing yourself? ........................................................ Response format: (1) Hardly ever or never, (2) sometimes, (3) often, (4) almost every day (in the past 2 weeks)

5. Have you ever made a plan to kill yourself? 6. Have you made a plan to kill yourself in the

past 2 weeks? ................................................ 7. Have you ever tried to kill yourself? ........... 8. Have you tried to kill yourself in the past 2

weeks? ........................................................... Response format: (1) Yes, (2) No

3. Did you wish you were dead? .......................

Ethnic status is self-designated, using these categories: (1) Anglo or White; (2) African or Black; (3) Native American; (4) specific Hispanic groups, including Cu- ban, Puerto Rican, Mexican, or Central American; (5) Specific Asian groups, including Cambodian, Vietnamese, Indian, Pakistani, Chinese, Korean, and Pacific Islander; (6) Other, and (7) Mixed An- cestry.

Characteristics of the sample are presented in Table 2. The sample is limited to the nine ethnic groups with sufficient sample sizes for statistical comparisons and with complete data on age, gender, socioeconomic status, and ethnicity (n = 4,186). There was almost equal represen- tation by gender and grade level. About half of the students reported that their family’s socioeconomic status was about the same as others a t their school. There was a clear positive skew, with dispropor- tionate numbers of subjects rating their socioeconomic status as “better or much better off,” compared with “worse or much worse off.” The ethnic composition was diverse, with over 20 distinct groups identi- fied. Since some of the groups have small numbers, subsequent analyses focus on the nine largest groups.

ROBERTS ET AL. 211

TABLE 2 Characteristic of the Study Participants

Characteristics N 70

Age 11 or younger 12 13 14 15 or older

Gender Male Female

Grade 6th 7th 8th

Much worse off Worse off About the same Better off Much better off NIA

African American Anglo American Central American Chinese American Indian American Mexican American Mixed Ancestry Pakistani American Vietnamese American

Socioeconomic Status

Ethnic Status

417 1138 1329 982 297

2071 2032

1355 1407 1389

76 257 1958 950 560 23

1237 775 253 177 188 755 342 155 304

10.0 27.3 31.9 23.5 7.1

50.5 49.5

32.6 33.9 33.5

2.0 6.7 46.8 24.8 14.6 0.6

29.6 18.5 6.0 4.2 4.5 18.0 8.2 3.7 7.3

Total 4186 100.0

RESULTS

Suicidal ideation was prevalent (see Table 3), with 7.3% feeling hopeless, 9.1% having thoughts of death, 5.4% wishing they were dead, and 4.3% having thoughts of suicide almost every day for the previous 2 weeks. The prevalence of suicide plans in the past 2 weeks and “ever” (lifetime) were 5.7% and 15.1%. respectively. The prevalence of suicide attempts in the past 2 weeks was 3.5% and of lifetime attempts was 10.5%. Females were more likely to report lifetime plans and attempts ( p < .001), but males were more likely to report recent attempts ( p < .05). On the other hand, 83% of the participants reported no suicidal ideation in the past 2 weeks.

In a previous study using these same four ideation items, a score of 5 or greater was equivalent to a score 1 standard devi- ation or more above the mean (Roberts & Chen, 1995). A score of 5 or more also yielded the same distribution in this sample, defined as indicating high, pervasive ideation. The mean score was 7.69 for those with suicide plans in the past 2 weeks versus 3.46 for those with no plans, and 7.57 for those with suicide attempts in the past 2 weeks versus 3.64 for those with no attempts. We also use the single item on thoughts of suicide almost every day in the past 2 weeks as an indicator by itself.

Using the summary measure, ideation was higher among females, older youths (particularly 15 and older), and those whose socioeconomic circumstances were worse to much worse than their peers (see Table 4). In general, the same pattern held for suicide plans and attempts, with the exception of gender, where the trend was for males to report higher levels of attempts. Comparing ethnic groups, Mexi- can Americans had higher prevalences of recent suicide ideation, plans, and attempts. Youths of Mixed Ancestry had a higher rate suicide plans and Pakistani youths had a higher rate of suicide attempts.

Using the single item on suicidal thoughts, the patterns were different. There was no gender effect, and no age effect (although those 15 or older were higher). The socioeconomic effect was much less pronounced. In terms of ethnic contrasts, Mexican and Pakistani youths and those of Mixed Ancestry also had higher prevalences of suicidal thoughts in the past 2 weeks.

As can be seen in Table 5, lifetime prevalences of plans and attempts were higher for females ( p < .001), for older youths, and for those worse off economically. In- dian Americans have significantly lower rates of lifetime plans, as do Chinese Americans. Youths of Mixed Ancestry have significantly higher rates of lifetime plans.

Multivariate logistic regression analy-


TABLE 3 Distribution of Suicide Items in Sample

Gender

Male Female Overall

N % N 70 N %

Suicidal ideation (past 2 weeks) Hopeless Thought of death Dead wish Thought about suicide

Suicidal plans Lifetime Past 2 weeks

Suicidal attempts Lifetime Past 2 weeks

274 7.3 116 6.3 158 8.2* 341 9.1 167 9.1 174 9.0 203 5.4 86 4.7 117 6.1 161 4.3 79 4.3 82 4.2

575 15.1 224 12.0 351 18.2*** 215 5.7 107 5.7 108 5.6

395 10.5 157 8.5 238 12.4*** 133 3.5 78 4.2 55 2.9*

*p < .05; ***p < .001.

ses, adjusting for the effects of age, gender, and socioeconomic status, yielded significant odds ratios (see Table 6), using the Anglo group as the reference, for suicidal ideation for the Mexican American (OR = 1.76, p < .001), Pakistani (OR = 2.0, p < . O l ) , and Vietnamese (OR = 1.48, p < .05) groups. For suicide plans in the past 2 weeks, Mixed Ancestry youths (OR = 2.02, p < .05) and Pakistani youths (OR = 3.19, p < .01) had elevated risk. For recent attempts, only the Paki- stani youths had elevated risk (OR = 3 . 1 9 , ~ < .01).

Adjusted odds ratios also indicate youths 13 and older have increased risk of suicidal ideation, with those 14 and above at greater risk for recent plans and attempts. Females are at greater risk only for ideation, while those who report their economic status as much worse off are at greater risk for suicidal ideation, plans, and attempts.

Again, using only the single item on suicidal thoughts as our measure of ideation, the patterns of association were quite different. There were no effects for gender, age, or socioeconomic status. Only two ethnic groups had higher suicidal thoughts, Pakistani and Mixed Ancestry youths.

DISCUSSION

There have been few community-based epidemiological studies of suicidal behaviors among younger adolescents, and almost none comparing diverse ethnocultural groups. The prevalences of suicidal ideation, plans, and attempts in this sample of middle school students suggest that suicidal behaviors are a significant mental health problem in early to middle adolescence and, further, that this problem varies to some extent by ethnocultural background of the youths.

How do our rates compare with those reported in the literature? Perusal of the literature suggests that rates of suicidal ideation among adolescents fall generally in the range of 10-20% (Andrews & Lew- insohn, 1992; Garrison et al., 1991a. 1991b; Roberts & Chen, 1995; Swanson et al., 1992). Our overall rate of ideation in the past 2 weeks using the four-item scale was 16.6%, well within this range. Our range of pervasive ideation (almost every day for the past 2 weeks) was 4-7%, with 4.3% having thought about suicide or killing themselves almost every day in the past 2 weeks. Likewise, available research suggests that the lifetime prevalence of

ROBERTS ET AL. 213

TABLE 4 Prevalence of Suicidal Ideation, Suicide Plans, and Attempts for

Past 2 Weeks Overall, by Gender, Age, E thnicity, and Socioeconomic Status

Suicide Suicide Suicide Suicide ideation thought Plans Attempts

(past (past (past (past 2 weeks) 2 weeks) 2 weeks) 2 weeks)

% N % N % N 70 N

Overall Gender

Male" Female

Anglo American' African American Mexican American Central American Indian American Pakistani American Vietnamese American Chinese American Mixed Ancestry

12" 11 or younger 13 14 15 or older

Socioeconomic status About the same" Much worse off Somewhat worse Better off Much better off

Ethnic Status

Age

16.7 715

14.4 304 19.0*** 393

14.3 111 14.1 174 23.6*** 178 17.3 59 13.3 25 20.0 31 19.1 58 9.0 16

18.4 63

13.6 159 13.7 58 16.4 222 19.8*** 199 23.9*** 73

16.4 328 30.3** 23 26.9*** 70 15.4 149 13.7 79

4.4

4.4 4.3

3.2 4.3 5.4* 4.7 2.8 7.1* 4.0 1.2 6.7**

4.2 3.8 4.2 4.8 6.1

4.0 9.0* 6.6 4.0 4.6

163 5.7 222 3.6

78 5.7 107 4.2* 81 5.6 108 2.9

23 5.1 37 3.0 45 3.9 42 2.3 36 8.0 54 5.5* 10 4.3 13 1.3 5 5.6 10 3.9

10 9.2 13 7.4* 11 6.1 17 5.0 2 2.4 4 2.9

21 10.3** 32 5.3

42 4.7 50 2.9 14 2.8 11 1.8 49 5.6 68 3.7 42 7.4* 67 4.3 16 9.6** 26 7.4**

70 5.1 93 3.1

15 8.5* 20 4.7 35 5.0 45 3.2 23 5.7 30 3.4

6 14.5** 10 16.2***

141

78 55

22 25 37

4 7

10 14 5

17

36 7

47 42 20

57 11 11 29 18

"Indicates reference group. * p c .05; **p < .01; ***p < .001.

suicide attempts among adolescents is in the 3.5-11% range (Andrews & Lewin- sohn, 1992; Vega et al., 1993). Our rate, at 10.670, is a t the upper limit of published studies. Interestingly enough, Meehan et al. (1992) report a lifetime prevalence of suicide attempts of 10.4% in a sample of 18-24-year-old college students using a self-administered questionnaire.

There are almost no data on prevalence of suicidal ideation, plans, or attempts across different ethnocultural groups. Our results suggest minority status may increase the risk for suicidal behaviors, but also that the effects vary depending on the outcome examined (ideation vs. at-

tempts vs. plans) and the group studied. Mexican American, Pakistani American, Vietnamese American, and Mixed Ances- try youths all had elevated prevalences on one or more indicators of suicidal behavior. Adjusting for the effects of age, gender, and socioeconomic status, Pakistani American students reported higher prevalences of recent ideation, plans, and attempts. An intriguing finding was that youth of Mixed Ancestry appear to be at elevated risk of several suicidal indicators. To our knowledge, this is the first report of such an effect.

Sorenson and Golding (1988) used the Epidemiologic Catchment Area (ECA)


TABLE 5 Prevalence of Suicide Plans and Attempts for Lifetime Overall,

by Gender, Age, Ethnicity, and Socioeconomic Status

Suicide plans Suicide attempts (Lifetime) (Lifetime)

% N T O N

Overall Gender

Male" Female

Anglo American' African American Mexican American Central American Indian American Pakistani American Vietnamese American Chinese American Mixed Ancestry

12" 11 or younger 13 14 15 or older

Socioeconomic status About the same' Much worse off Somewhat worse Better off Much better off

Ethnic status

Age

15.2

12.0 18.2***

15.5 13.5 18.3 12.6

13.2 15.8 9.4*

23.7**

13.8 8.7*

15.1 18.1** 21.7**

14.8 23.2* 23.8*** 14.3 12.6

7.7**

588

224 351

113 144 124 38 14 19 44 16 76

145 34

184 165 59

269 16 56

129 67

10.5 403

8.5 157 12.4*** 238

11.0 81 8.7 92

13.3 90 8.7 25 8.3 15 7.8 11

10.4 29 9.0 15

14.1 45

8.6 90 6.7 26

10.2 124 12.7** 115 16.9*** 46

10.2 186 19.4* 13 18.4*** 43 10.1 91 7.4 39

"Indicates reference group. *p < .05; **p < .01; ***p < .001.

data from Los Angeles to contrast Anglos and Hispanics. The latter reported less lifetime suicidal ideation and fewer attempts, almost half the rates for Anglos. Why the ethnic differential for adults in the ECA survey is the opposite of that observed for adolescents in this study or in our earlier study in New Mexico (Roberts & Chen, 1995), is unclear. Many factors could account for the observed differences, including differences in time, place, age of respondents, and measurement. In Los Angeles structured psychiatric inter- views were used to obtain data on lifetime episodes. As Simon and Vonkorff (1992) have demonstrated, lifetime reports in the ECA study appear to have substantial er-

ror. Whether this error differs by ethnicity is unclear. Clearly, this is an area in need of further study.

We are unaware of any published r e search that has reported elevated risk of suicidal behaviors among Pakistani American adolescents. At this point, it is not clear why this is so. Likewise, the Chi- nese youths appear to be at lower risk of all three forms of suicidal behaviors. Why this is so is not apparent, although some studies have reported lower risk of depression for Chinese as well (Chang, Morris- sey, & Koplewicz, 1995; Lin, 1953; Shen, 1981; Kleinman, 1982; Kuo, 1984; Xia & Zhang, 1984; Leung, Chan, & Cheng, 1992) Mexican American adolescents had

TA

BL

E 6

Odd

s R

atio

s (O

R) o

f Su

icid

al I

deat

ion,

Sui

cide

Pla

ns, a

nd A

ttem

pts

for

Past

2 W

eeks

, A

djus

ted

for A

ge, G

ende

r, an

d So

cioe

cono

mic

Sta

tus

Suic

idal

idea

tion

Suic

ide

thou

ght

Suic

ide

plan

s Su

icid

e at

tem

pts

Adj

uste

d A

djus

ted

Adj

uste

d A

djus

ted

OR

p-va

lue

OR

p-va

lue

OR

p-va

lue

OR

p-va

lue

Gen

der

Mal

e"

Fem

ale

Ang

lo A

mer

ican

" A

fric

an A

mer

ican

M

exic

an A

mer

ican

C

entr

al A

mer

ican

In

dian

Am

eric

an

Paki

stan

i Am

eric

an

Vie

tnam

ese

Am

eric

an

Chi

nese

Am

eric

an

Mix

ed A

nces

try

12"

11

13

14

15 o

r old

er

Soci

oeco

nom

ic st

atus

A

bout

the

sam

ea

Muc

h w

orse

off

Som

ewha

t wor

se o

ff B

ette

r of

f M

uch

bette

r of

f

Eth

nic

grou

p

Age

1.00

-

1.46

.ooo*

* *

1.00

- 1.07

.608

1.76

.ooo*

** 1.10

.612

1.17

.508

2.00

.003**

1.48

.036*

.56

.058

1.35

.098

1 .oo

-

1.07

.678

1.29

.034*

1.67

.ooo

***

1.98

.ooo

***

1.00

- 2.07

.006**

1.88

. ooo*

* *

.97

201

.78

.082

1.00

1.05

1 .oo

1.43

1.59

1.14

1.13

2.53

1.01

.19

2.17

1 .oo

.84

.95

1.16

1.24

1 .oo

1.42

.58

.59

.63

-

.774

- .184

.lo6

.772

.756

.024*

.982

.lo3

.015*

-

.607

306

.530

.534

- .494

.062

.loo

.194

1.00

.63

1.00

.72

1.18

.25

1.96

3.20

1.25

.80

2.27

1 .oo

.64

1.31

1.85

2.90

1 .oo

3.73

1.08

.97

1.01

- .029*

- .351

.624

.072

.145

.007**

.609

.738

.022*

-

.394

.344

.035*

.004**

- .004**

.852

.930

.969

1.00

.69

1 .oo

.84

1.27

.32

1.68

3.19

1.34

.92

1.89

1.00

.66

1.55

1.72

2.86

1.00

5.75

1.35

.96

.86

-

.061

-

.577

.433

.078

.249

.004**

.460

.a91

.065

-

.385

.097

.050

.002**

-

.ooo*

** .397

A92

.618


higher rates of suicidal behaviors, whereas those of Central American origin did not. What accounts for this differential risk also needs clarification.

Our results are generally consistent with other studies in terms of higher risk for females, older youths, and those reporting lower socioeconomic status. Our findings also corroborate studies that report increased risk of suicidal behaviors among Hispanic youths, particularly Mexican Americans (see Roberts & Chen, 1995). Our results suggest that risks of suicidal behaviors vary somewhat, depending on the ethnic group and the indicator of suicidal behaviors. That is, there were differential effects across groups observed for suicidal ideation, plans, and attempts. Our data support the contention by some that suicidal behaviors are com- plex, ranging from ideation to attempts to completed suicide (see Reynolds & Mazza, 1994), with presumed severity in terms of risk of death in that order. Our prevalence rates reflect that. Prevalence in the past 2 weeks was 16.7% for ideation, 5.7% for plans, and 3.6% for attempts.

Our findings are limited in some re- spects by our measure of suicidal behaviors. Although our measure included a range of suicidal behaviors and was em- bedded in a set of items inquiring about episodes of major depression, the data are self-report. This is particularly problematic in the case of reported suicide attempts. As Reynolds and Mazza (1994) point out, suicide attempters represent a heterogeneous population that includes youths who have made serious and life- threatening attempts as well as young- sters whose attempts are very mild or of limited lethality potential. These two ex- tremes probably represent very different risk groups for completed suicide. For example, Meehan et al. (1992) report that of those young adults who reported ever at- tempting suicide, 44% actually suffered injury, 25% sought medical care, and 10% were admitted to the hospital. Garrison et al. (1991b) report 5.9% of high school students reported a suicide attempt in the past year not requiring medical care in the past year, and 1.6% reported attempts requiring medical treatment.

In our school-based survey it was not possible to assess methods used or to judge severity of the attempts due to re- straints imposed by the school district. Thus, in the case of the observed ethnic differences in reported suicide attempts in our study, it is not possible to determine whether the increased risk observed for Pakistani youths, for example, would re- main if method and severity were con- trolled.

This issue relates to our estimate of 3.6% for suicide attempts in the past 2 weeks. We could find no epidemiological studies that examined prevalence of attempts in the past 2 weeks. Our rate of 3.6% is comparable to the 12-month prevalences reported by Garrison et al. (1991b), which ranged from 2.3% to 8.5%, but is higher than that reported (1.9%) by Mee- han et al. (1992).

Given the results reported here, and the limited data on suicidal behaviors among adolescents, more comparative research is needed to ascertain whether and how ethnocultural background constitutes a risk factor for suicide among youths.

Analyses are underway to further explore ethnocultural differences in suicidal behaviors, including the role of language, generational status, and ethnic identity in addition to ethnic status and perceived discrimination.

REFERENCES

Alba, R. D. (1990). Ethnic identity: The tmnsforma- tion of White America. New Haven, CT Yale Uni- versity Press.

Andrews, J. A., & Lewinsohn, P. M. (1992). Suicidal attempts among older adolescents: Prevalence and co-occurrence with psychiatric disorders. Journal of American Academy of Child and Ado- lescent Psychiatry, 32.655-662.

Chang, L., Morrissey, R. F., & Koplewicz, H. S. (1995). Prevalence of psychiatric symptoms and their relation to adjustment among Chinese- American youth. Journal of the American Acad- emy of Child and Adolescent Psychiatry, 34(1), 91- 99.

Dubow, E. F., Kausch D. F.. Blum. M. C., Reed, J., & Bush, E. (1989). Correlates of suicidal ideation and attempts in a community sample of junior high and high school students. Journul of Clinical Child

Garrison, C. Z. , Addy, C. L., Jackson, K. L.. Mck- eown, R. E., & Waller, J. L. (1991a). A Longitudi- nal study of suicidal ideation in young adoles-

Psychology, 18,158-166.

ROBERTS ET AL. 217

cents. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 397-603.

Garrison, C. Z., Addy, C. L., Jackson, K. L., Mc- Keown, R. E., & Waller, J. L. (1991b). Suicidal behaviors in young adolescents. American Journal of Epidemiology, 133.1005- 1014.

Harkavy Friedman, J. M. N., Asnis, G. M., Boeck M., & DiFiore, J. (1987). Prevalence of specific suicidal behaviors in a high school sample. American Journal of Psychiatry, 144, 1203-1206.

Hillard, J. R., Slomowitz. M., & Deddens, J. (1988). Determinants of emergency psychiatric admis- sion for adolescents and adults. American Journal o f Psychiatry, 145,1416-1419.

Kleinman, A. (1982). Neurasthenia and depression: A study of somatization and culture in China. Cul- ture, Medicine and Psychiatry, 6, 117-190.

Kuo, W. H. (1984). Prevalence of depression among Asian-Americans. Journal of Nervous and Mental Disease, 172, 449-457.

Lester, D., & Anderson, D. (1992). Depression and suicidal ideation in African American and His- panic American high school students. Psychologi- cal Reports, 71, 618.

Leung, C. M., Chan, K. K., & Cheng, K. K. (1992). Psychiatric morbidity in a general medical ward Hong Kong’s experience. General Hospital Psy- chiatry, 14, 196-200.

Lin, T. (1953). A study of the incidence of mental disorders in Chinese and other cultures. Psychiatry,

Meehan, P. J., Lamb, J. A., Saltzman, L. E.. & O’Car- roll, P. W. (1992). Attempted suicide among young adults: Progress toward a meaningful estimate of prevalence. American Journal of Psychiatry, 149,

Moscicki, E. K. (1995). Epidemiology of suicidal behavior. Suicide and Life-Threatening Behavior, 25,

National Center for Health Statistics. (1994). Ad- vance report of final mortality of statistics, 1991. Monthly Vital Statistics Report, 42(2, Suppl.). Hy- attsville, MD: U.S. Public Health Service.

National Institute of Mental Health. (1992). Diag- nostic Inventory Schedule for Children: Child In- formant Interview about Self (NIMH Contract No.

Phinney. J. S. (1990). Ethnic identity in adolescents and adults: Review of research. Psychological Bul- letin, 108(3), 499-514.

16,313-336.

41-44.

22-35.

278-89-0001 ).

Reynolds, W. M., & Mazza, J. J. (1992). Suicidal behavior in nonreferred adolescents. Paper presented at the International Conference for Suicidal B e havior, Western Psychiatric Institute and Clinic, Pittsburgh, PA.

Reynolds, W. M., & Mazza, J. J. (1994). Suicide and suicidal behaviors in children and adolescents. In W. M. Reynolds & H. J. Johnson (Eds.), Handbook of depression in children and adolescents (pp. 525- 580). New York Plenum.

Roberts, R. E.. & Chen, Y. R. (1995). Depressive symptoms and suicidal ideation among Mexican Origin and Anglo adolescents. Journal of the American Academy of Child and Adolescent Psy- chiatry, 34(1), 81-90.

Robinson, J. (1986). Emergencies I. In K. S. Robson (Ed.), Manual of clinical childpsychiatry (pp. 185- 21 1). Washington, DC: American Psychiatric Press.

Shaffer, D., &Hicks, R. (1994). Suicide. In I. B. Pless (Ed.), Epidemiology of childhood disorders (pp. 339-365). New York: Oxford University Press.

Shen, Y. (1981). The psychiatric services in the urban and rural areas of People’s Republic of China. Bul- letin of the Neuroinformation Laboratory of Naga- saki University, 8, 131-137.

Simon, G. B., & Vonkorff, M. (1992). Reevaluation of secular trends in depression rates. American Jour- nal ofEpidemiology, 135,1411-1422.

Swanson, J. W., Linskey, A. O., Quintero-Salinas, R., Pumariega, A. J., & Holzer, C. E. (1992). A Bina- tional school survey of depressive symptoms, drug use, and suicidal ideation. Journal of the American Academy Child and Adolescent Psychiatry, 31,

Vega, W. A., Gil, A. G., Zimmerman. R. S.. & War- heit, W. J. (1993). Risk factors for suicidal behavior among Hispanic, African American, and non- Hispanic white boys in .early adolescence. Ethnicity and Disease, 3, 229-241.

Xia, Z. Y., & Zhang M. (1984). Affective psychoses. In Xia, Z. Y., e t al., eds. Clinical Psychiatry, Vol. 2 (in Chinese; pp.) Shanghai, Jingshenyixuecong- shu.

669-678.

Received: April 2, 1996 Revision Accepted: October 1, 1996

Ethnocultural Differences in Prevalence of Adolescent Depression

Documents