Homosexual related legislation does not reduce suicidal intent in sexual minority groups

Editor—Bagley and D'Augelli contend that suicidal behaviour in bisexual, gay, and lesbian youth is an international problem associated with homophobic legislation.¹ Heretofore, no data existed to examine this. I recently found considerable variation in suicidal behaviour by sexual orientation cross culturally.² I re-examined the data to determine whether variations in suicidal behaviour are associated with national legislation on homosexual (a) adoption, (b) military service, (c) employment, and (d) marriage or domestic partnership.

A more detailed description of participants and methods can be found elsewhere.³,⁴ Four English speaking countries were selected from the intercontinental study²: Australia (n=185, 17.3% sexual minority groups), Canada (n=1178, 10.95% sexual minority groups), United Kingdom (n=116, 12.9% sexual minority groups), and United States (n=34 843, 13.41% sexual minority groups). Only countries with 15 or more participants in sexual minority groups were included to detect a proportional difference of 0.10, assuming a population proportion of M=0.50 (no difference), alpha=0.05, and 70% power. National governments' positions on issues (a)-(d) above were scored 0 (no position), 1 (prohibited), and −1 (protected), with a range of 4 to −4.

The scoring was as follows:

• Australia −1: adoption 0, marriage or domestic partnerships 0, employment discrimination 0, military service −1

• Canada −4 (see Canadian Charter of Rights and Freedoms section 15(1), 1982; Human Rights Act 1996; judicial decisions): adoption −1, marriage −1, employment −1, and military service −1

• United Kingdom −2: adoption 0, marriage 0, employment −1, military service −1

• United States 2: adoption 0, marriage 1 (Public Law 104-199), employment 0, military service 1 (Department of Defense directive 1332.14, 28 January 1982).

The table shows suicide ideation and suicide attempts by each sexual orientation in country, with the odds ratio and 95% confidence interval for risk. The final two columns reflect the Pearson's correlation and its P value for the relation between the cumulative score for homosexual related legislation (index) and suicidal behaviour cross culturally. The percentage of homosexual suicide attempts were strong and significant but inversely related to the index (r=−0.952, P<0.05). The index was not significantly associated with other variables of suicidal behaviour.

In contrast to Bagley and D'Augelli's public policy thesis,¹ I tested a social constructivist model that suggests cultural attitudes towards human sexuality mediate the relation between suicidal behaviour and sexual orientation.² The present study provides empirical evidence to refute the public policy model. Thus, changing cultural attitudes towards human sexuality seems to be a more effective target of intervention for the suicidal behaviour of bisexual, gay, and lesbian youth than direct challenges to public policy related to homosexuality.

Table.

Percentage of people with suicidal ideation and who attempted suicide, by sexual orientation and country, with national index scores for legislative policies on homosexuality, Pearson product moment correlation across countries, and associated probabilities

	Australia	Canada	United Kingdom	United States	r	P value
Legislation scores	−1	−4	−2	2	—	—
Suicidal ideation:
Homosexual	28.1	28.5	18.8	25.3	−0.102	0.898
Heterosexual	24.2	14.9	18.0	13.8	−0.096	0.904
OR (95% CI)	1.23 (0.52 to 2.88)	1.05 (0.27 to 4.08)	2.28 (1.50 to 3.45)	2.13 (1.98 to 2.29)	0.564	0.436
Suicide attempts:
Homosexual	12.5	13.1	12.5	10.3	−0.952	0.048*
Heterosexual	7.2	4.8	7.0	4.8	−0.105	0.895
OR (95% CI)	1.84 (0.55 to 6.21)	1.90 (0.36 to 0.07)	3.00 (1.68 to 5.38)	2.28 (2.05 to 2.54)	0.099	0.901

^*P<0.05.