TO: pjtremblay@hotmail.comTO:
TO: The Introduction / Study Results - Suicidality Problems of Gay and Bisexual Male Youth
To: Index Page for Bisexual, Gay, Queer Male Suicidality
TO: The Data Tables / Notes - Suicidality Problems of Gay and Bisexual Male Youth
At A Glance! The result of more than 30 studies in 4 categories.
The replication results of this study have again been replicated.
A GLBT Internet-Based Education with InfoSearch Pages
Part 2 of 4: Discussion

Final edited copy, Dec. 1996. Published as "Chapter 12" in C. Bagley & R. Ramsay, Eds. (1997) Suicidal Behaviours in adolescent and adults: taxonomy, understanding and prevention. Brookfield, Vermont: Avebury. A shorter version of this paper was published as Suicidal Behaviors in Homosexual and Bisexual Males in Crisis, The International Journal of Suicide and Crisis Studies, Vol. 18(1), 1997, pp. 24-34 (PubMed "abstract" link with a document delivery service. The abstract is also available on a CRISIS webpage.). The results of this study, combined with the another major demographic study also published in 1998, and additional information have been used to produce a monumentally important  CAVEAT ALERT! In 2000, a CDC-Based study - from an NHANES sample taken between 1988 and 1992 - produced a subsample of young homosexually oriented young adult males (age = 17 to 29 years) comparable in age range of the sample taken for this study (age = 18 to 27 years, sample taken in 1991) with "suicide attempt result - compared to same-age 'heterosexual' males - which replicated the results of this study. See Addendum.


Discussion Introduction / Study ResultsEnd Notes, TablesBibliography

Bell and Weinberg (1978) were meticulous in obtaining a sample of white predominantly homosexual males; their total sample consisted of 3,538 white males [2], and random sampling was used within the many contact cells to produce a final stratified sample of 575 homosexual males highly representative of all such males living in the San Francisco Bay Area. The data were then compared with the data produced from a stratified random sample of predominantly heterosexual males living in the same area. In contrast, our samples of homosexually and heterosexually oriented males are subsets of a randomly generated sample. The Bell and Weinberg analysis produced a '13.6 times' higher suicide attempt risk factor (to the age of 20 years) for predominantly homosexual compared to predominantly heterosexual males, and our replication of this factor (13.9 times, to the average age of 22.7 years) is an important cross-validation.

The 6.1 percent attempted suicide rate for our sample, compared to the Bell and Weinberg's 9.6 percent estimate to the age of 20 years, and the 31.3 percent estimate from contemporary community-based samples [1] probably result from sampling differences. The Bell & Weinberg homosexual male sample represents a largely gay community urban population, while our suburban sample excluded gay and bisexual males living in the gay community area. Community-based samples of gay and bisexual male youth are (unlike our own sample) generally overrepresented by males with attributes associated with high attempted suicide rates. In the Remafedi et al. (1991) sample of 137 males, 40 percent had been runaways, 39 percent had been sexually abused, 35 percent had been arrested, 23 percent were classified gender nonconformable (feminine), and 17 percent had engaged in prostitution. The associated attempted suicide rates for males having these (often overlapping) attributes are 36, 46, 44, 47, and 52 percent; the average for the sample is 30 percent.

Gender nonconformable gay and bisexual males were estimated to be 3-times more at risk for a suicide attempt than other homosexual males (Remafedi et al. 1991), a risk factor also suggested by Harry (1983). Most effeminate males are homosexually oriented (Bell, Weinberg & Hammersmith, 1981; Green, 1987) and, as youth, these 'visibly' gay males tend to live in gay community areas, often for safety reasons. The same applies for male youth engaged in prostitution. Their stroll is usually in the gay community area where they most often live. These at risk males, as well as other at risk homosexually oriented males living within the gay community area, are not represented (or would be greatly underepresented) in our subsample of homosexually oriented young men living generally stable lives outside the gay community area.

Our underestimate of the suicide attempt rate, however, is not confined to homosexual males given that our sample's 1.07 percent suicide attempt rate (8/750) is much lower that the 3.2 percent estimated lifetime rate in adolescent males not differentiated on the basis of sexual orientation [3]. Street youth, often with an adolescent runaway history, are not represented in our sample. Runaway youth have attempted suicide rates ranging from 15 to 29 percent (Stiffman, 1989; Rotheram-Borus, 1993), and gay and bisexual male youth form about 20 to 40 percent of homeless and street youth male populations (Savin-Williams, 1994). In Calgary, these youth and related services are concentrated in the city centre where one safe house reports that about 40 percent of their clients are homosexually oriented (Calgary Board of Education, 1996). In addition to not sampling youth leading relatively unstable lives, our sampling excluded institutionalized youth at high risk for life-threatening suicide attempts (Memory, 1989). On the basis of our results, it is therefore proposed that sampling youth in suburban areas of North American cities will produce relatively low suicide attempt rates, given that male youth suicide attempters tend to be greatly overrepresented in populations concentrated in city centres, often enough within gay community areas.

The lower attempted suicide rate for homosexually oriented males in our sample may also be related to the age of self-identification as homosexual or bisexual. On the basis of a multivariate analysis, Remafedi et al. (1991) reported that the age of self-identification was closely related to suicide attempts. About one-third of suicide attempts occurred within the year following self-identification, most others occurring soon thereafter. The risk also decreases significantly with delay in the age of self-identification. The average ages for self-identification, first homosexual experience, and a suicide attempt in the Remafedi et al. (1991) study is 14.9, 15.6, and 15.5 years, respectively. Herdt (1989a) cites similar self-identification ages (12.5 to 16.3 years) for community-based samples.

In our sample, the average age of the first homosexual experience for self-identified homosexual males is likely higher, probably 18 or 19. By their eighteenth birthday, only 43 percent (38/83) of the combined numbers of self-identified homosexual and bisexual males had become homosexually active. The close association between the age of first homosexual experience (15.6 years) and age of self-identification (14.9 years) reported by Remafedi et al. (1991) therefore suggests that our sample of suburban homosexually oriented males probably self-identified around 17 or 18 years, suggesting the likelihood of a lower suicide attempt rate. Their 6.1 percent attempted suicide rate is nonetheless high compared to the 0.45 percent rate for heterosexual males living in the same area, but low compared to the 31.3 percent average for twelve community samples [1]. Our results also imply that suicidality, as was the case in the Remafedi et al. (1991) study, is related to homosexuality issues; but these factors may be different from those identified by Remafedi. In our study, 80 percent (4/5) of homosexually oriented suicide attempters, and 77 percent (10/13) of homosexually oriented males in the "self-harm" category were either in the celibate homosexual or in the homosexually active bisexual categories.

Some celibate homosexuals males in our sample may be experiencing prolonged psychological problems with regards to the coming out process. Coming out events occur in stages usually tackled in linear fashion (Troiden, 1988; Martin, 1992) but variations occur (such as the first homosexual experience occurring before or after self-identification), and regressions to a celibate status may happen. These regressions may be related to self-acceptance problems, a factor also existing for some self-labelling bisexual males. When bisexuality is a transitional status (Gochros, 1989; Klein & Wolf, 1985), it is a part of the coming out process. Binson et al. (1995) report that about 50 percent of bisexually active male aged 18 to 29 years will become exclusively homosexually active after the age of 30 years.

As a group, homosexually oriented young adult males in our study were more likely to have experienced suicide-related life crises at some point in their lives, compared to their heterosexual counterparts. Their three-times greater likelihood of having engaged in some form of self-harm activities, and especially their 14-times greater risk for having engaged in the most serious form of self-harm (a deliberate attempt at self-killing), suggest their possible overrepresentation in hospitalizations (and possibly deaths) resulting from their suicide attempt(s). Their determination to die is reflected in the Remafedi et al. (1991) results: for 45 out of a total 68 suicide attempts studied (performed by 41 gay and bisexual male suicide attempters) for which the rescue was initiated by another person, "fifty-eight percent (26/45) of these cases received scores in the moderate to least' rescuable range. In other words, the predictable likelihood of rescue was moderate to low, despite the actual occurrence of an intervention" (P. 871). Twenty-one percent of the suicide attempts studied resulted in medical or psychiatric hospitalization, which contrasts sharply with the one to two percent rate in studies of youth (Meehan et al., 1992; CDC, 1991), the implications being that homosexually oriented males form the majority of male youth hospitalizations (and probably deaths) resulting from suicide attempts. This conclusion is also supported by the information related to the suicide reattempter rates for gay and bisexual male youth: 20 to 52 percent of the suicide attempters, mean: 39.6 percent [1]. Given that the probability of male suicide reattempters eventually committing suicide ranges from 10 to 14 percent (Diekstra, 1989; Spirito et al, 1989; Sakinofsky et al., 1990), and that gay and bisexual male youth have a very high reattempter rate, homosexually oriented males may therefore feature largely in populations of young males who have killed themselves. Preliminary work with high school population by Tremblay (1994 and 1996) indicates that more than half of male youth suicide victims were homosexually oriented.

Three studies have sought to estimate the proportion of homosexually oriented young males who commit suicide (Rich et al., 1986; Bagley, 1992; Shaffer et al., 1995). All of these studies are beset by the problem that coroners and medical examiners may not be told about homosexuality; if parents have such information (or have reasons to suspect), they may suppress the information. Some young males, suffering from the psychological agony imposed by homophobia, may have told no one about their acute crises. Bagley (1992) who studied medical examiner records in Alberta found that for 11 of 130 males (8.5%) aged 10 to 30 who committed suicide, homosexuality was identified as a background factor. Bagley notes that one male suicide victim in his series of completed suicides was known from other information to have experienced identity crises about coming out; but this was not recorded in the medical examiner's records. Rich et al. (1986) did not identify one homosexually oriented victim below the age of 21, but produced a 10.9 percent estimate (13/119) for males aged 21 to 42. Shaffer et al. (1995) only identified three out of 95 male victims (3%), with homo-suspicious information existing for nine more. Generally, professionals have overlooked the importance and clinical implications of obtaining sexual orientation information from surviving homosexually oriented youth suicide attempters (Uribe & Harbeck, 1992; Martin & Hetrick, 1988), and the same has applied for runaway and street youth (Savin-Williams, 1994; Mohr, 1988). It is therefore not surprising that most researchers have not yet acquired the skills needed to discover the homosexual orientation of gay and bisexual male youth after their suicidal death.

A silence based on both denial and contempt has existed in our society (and in the youth problem research field) with respect to gay, lesbian, and bisexual youth issues, and the silence has also been a socially imposed attribute of homosexually oriented suicide attempters. At the New York Institute for the Protection of Gay and Lesbian Youth, the situation was described for nine youth receiving therapy elsewhere for their suicide attempt: "[They] had not yet told their therapist either that they were homosexual or that that was a factor in their suicide attempt" (Martin & Hetrick, 1988, p. 173). Uribe and Harbeck (1992) reported that most suicide attempters in their Los Angeles high school sample of 37 gay and bisexual males (indicating a near-50 percent attempted suicide rate in homosexual youth) had sought mental health services outside the school and that, 'without exception, those seeking help from private sources denied their sexual orientation to them' (P. 22).

These and other findings suggest that mental health professionals are poorly educated and trained about homosexual realities (Murphy's 1992, Kourany, 1987), and that these issues continue to be ignored in most universities. At Vancouver's University of British Columbia, for example, homosexual realities are not included in the training of counsellors (Abrams, 1996), in spite of Vancouver having the largest gay and lesbian community in western Canada. This education only began in 1995 in some University of Calgary faculties producing professionals who will be working with youth; requests have now been made for the inclusion of homosexuality issues in the training of Canadian family therapists (Steel & Gyldner, 1993).

Gay, lesbian, and bisexual youth are misunderstood and often harmed in the youth residence system (O'Brien et al., 1993; Gibson, 1989), and rapes of gay boys (and heterosexual boys assumed to be gay) in institutional settings is reported to be likely or common (Martin & Hetrick, 1988; Tremblay, 1996). They are also misunderstood and harmed in substance abuse treatment programs (Simpson, 1994); substance abuse is a common problem for gay and bisexual male youth having coming out problems, especially those related to stigmatization and socially induced low self-esteem (Shifrin & Solis, 1992). Remafedi et al. (1991) reported, on the basis of a multiple logistic regression analysis, that the use of illicit drugs was one of the three most important factors implicated in the suicide attempts of gay and bisexual male youth, the age of self-identification and gender nonconformity being the other two.

Gay and bisexual male youth have been at risk for life experiences often linked to suicidal crises. These include being physically, verbally, and emotionally abused in families, schools, and society; experiencing declining academic achievement and truancy; becoming throwaways, runaways, street youth and delinquents; engaging in prostitution for survival purposes and/or as a way to act out learned negative stereotypes; and not being able to access qualified services when needed; or worse, they may be abused by professionals entrusted to help youth. (Remafedi,1987; Boyer, 1989; Gibson, 1989; Hunter, 1990; Kruks, 1991; Remafedi et al., 1991; Galst, 1992; American Academy of Pediatrics, 1993; Massachusetts Governor's Commission on Gay and Lesbian Youth, 1993; Savin-Williams, 1994; Pilkington & D'Augelli, 1995).

Sexual abuse may be another part of the problem. Our study has revealed that sexual abuse (experiencing unwanted sexual acts before the age of 17 years) is a factor in suicidality problems (Bagley et al. 1994). A further analysis of the data on our sample of 750 Calgary males indicates however that child sexual abuse is not significantly related to the self-harm behaviors reported by sexually active homosexual, bisexual, and heterosexual males, but is significant in the celibate groups who have the highest mean depression (CES-D) scores. Remafedi et al. (1991), on the basis of multivariate analysis, also did not find sexual abuse to be significantly implicated in the suicide attempts of sexually active gay and bisexual male youth when other factors were controlled. Our study also indicates that child sexual abuse may be a relatively insignificant factor in suicide attempts. The two suicide attempters in the celibate male group are homosexual males.


A substantial body of knowledge has become available about homosexually oriented males, and our results are an addition to the mostly German and British 100-year-old documentation of their suicidality problems (Ellis, 1948; Kennedy, 1984; Symonds, 1984; Rofes, 1983). The ongoing international scope of the problem is indicated by the 20 to 50 percent attempted suicide rates from community-based samples of British homosexually oriented youth (Plummer, 1989). The problem has also been officially recognized in western Australia (YouthLink, 1994).

Elsewhere, related information is only beginning to emerge, such as in Sri Lanka, a highly homophobic country (Selvadurai, 1996). The problem may also exist in China given the highly homophobic attitudes which have developed in the past 150 years (Hinsch, 1990). Historically, a homosexuality-related suicide problem in American aboriginal cultures is noted by Williams (1993). Relevant data suggest that aboriginal gay youth, often subjected to higher levels of homophobia than white youth, and also overrepresented in male prostitution (Boyer, 1989), have serious suicidality problems (Tremblay, 1994).

From the results of our study, the Bell & Weinberg study, other studies, and the issues addressed in this discussion, it is predicted that the level of homophobia manifested in a particular country, or culture, may be directly linked to the extent of gay and bisexual male youth suicidality problems. Ross (1989) for example, studied homosexually oriented males in four countries (Sweden, Finland, Ireland, and Australia); the analysis of the data 'suggests that homosexual adolescents are likely to have more problems in the more antihomosexual countries, and they are also likely to have less accepting ideas about homosexuality than older homosexual men' (p. 313). The need for international studies of suicidal behaviour in youth (and adults) is therefore indicated, and similar studies are also required within countries where different cultural groups co-exist and manifest varying levels of homophobia.

Data from studies of gay and bisexual male youth [1] suggest that the attempted suicide rates of North American homosexually oriented male youth of colour is about 30 percent higher that the rate for their white counterpart (Tremblay, 1995). White racism, both in society and in gay communities, and the often more intense homophobia which exists in minority groups, are added stressors for these youth (Gibson, 1989; Schneider et al., 1989; Tremble, Schneider & Appathurai, 1989; Erwin, 1994; Tremblay, 1994). Clearly, this is an area for future research.

The data in our study suggest that homosexually oriented male youth are about three-times more at risk for engaging in self-harm activities than heterosexual males. They may also accountfor more than half (62.5%) of male youth suicide attempters, reflecting a 14-times greater risk for a suicide attempt. These results are consonant with the Bell and Weinberg (1978) findings, indicating the probability that homosexually oriented males have been greatly over-represented in male youth who engage in deliberate self-harm for at least 50 years.

The relevant data also suggest that gay and bisexual males form an even greater proportion of male youth hospitalized and/or dying from their suicide attempts. The potential magnitude of the problem, extending into adulthood, is chillingly rendered by a Canadian gay community leader, Hellquist (1993): 'I certainly know considerably more people in our [gay] community who have taken their own lives than I know who have died from AIDS' (p. 8).

Discussion Introduction / Study ResultsEnd Notes, TablesBibliography

TO: The Introduction / Study Results - Suicidality Problems of Gay and Bisexual Male Youth
A GLBT Internet-Based Education with InfoSearch Pages
TO: The Data Tables / Notes - Suicidality Problems of Gay and Bisexual Male Youth

To The Top Of The Page!
To: Home Page - Gay and Bisexual Male Suicidality Studies