Gender is a powerful structural determinant of mental health (WHO, 2000). In particular, women have been found to be at higher risk for poor mental health outcomes such as disproportionately higher prevalence rates of trauma and post traumatic stress disorder (Tolin & Foa, 2006), anxiety disorder (Merikangas, Burstein, & Schmitz, 2013), mood disorder like unipolar depression (Kessler & Gadermann, 2013), and suicide ideation, suicide planning, and suicide attempt (Klonsky, May, & Saffer, 2016). This may be because of a number of systemic factors that disadvantage girls and women across societies. These factors include higher experience of sexual violence (including sexual assault and childhood sexual abuse), poverty and income inequality, subordinate social status, and even gender bias in medical diagnosis and health care (Lee, 2014).

One gender-related observation in the mental health literature is the so-called “gender paradox of suicide” (Canetto & Sakinofsky, 1998). Women have higher rates of suicide ideation and attempt, while men have higher rates of suicide deaths (Hawton, 2000). Various hypotheses have been proposed for this finding, for example, differential access to methods of suicide (men are more likely to have access to and utilize means such as firearms that are more likely to be fatal). However, feminist critics have pointed out that the strong focus on suicide mortality—deaths by suicide—has the effect of drawing attention away from the importance of morbidity reduction (Reyes & Reyes, 2004). That is, individuals, especially women, who do not end their life are likely to be still dealing with the stressors that burdened them to contemplate suicide in the first place, such as depression, violence, and poverty.

Beyond examining gender disparities in suicide deaths and women’s greater risk for suicide ideation and attempted suicide, researchers are now challenged to consider the intersectional nature of social categories (Cole, 2009). One basic way to do this is by reflecting on who is included, and not included, within an analytic category like “women.”

One historically overlooked variable in gender research on mental health and suicide is sexual orientation, a person’s enduring pattern of romantic and sexual attractions for and desire to be in intimate relationships with those of the same sex, the other sex, or both (IOM, 2011). Traditional thinking in the mental health professions, driven by historical anti-LGBT stigma, did include sexual minority status in mental health practice—but only by equating same-gender sexual orientations as problematic and indicative of psychopathology (a pathologization model). This pre-1973 view has since been replaced by more empirically based theories that attempt to explain findings showing associations between sexual minority status and negative mental health outcomes such as disproportionately higher rates in depression, anxiety disorders, substance abuse, and suicide ideation and attempted suicide (IOM, 2011). Contemporary frameworks such as minority stress (Meyer & Frost, 2013) and structural stigma (Hatzenbuehler, Phelan, & Link, 2013) argue that these findings do not indicate that sexual minority status is pathological per se, the same way that gender is not a causal factor in depression or suicide ideation and attempted suicide. Rather, like gender, sexual minority status structures exposure to particular additional stressors such as those related to experiences of discrimination, as well as the quantity and quality of resources (e.g., competent, LGBT-friendly health care by medical professionals), all of which are important for coping with negative mental health outcomes.

With regard to suicide ideation and attempted suicide, there is now considerable evidence that indicates sexual minorities, especially sexual minority youth, are at disproportionately higher risk (Haas et al., 2011; Liu & Mustanski, 2012; Poteat & Rivers, 2014; Pflum et al., 2015). In particular, lesbian and bisexual women have been shown to have an increased risk of suicide attempts and suicide ideation, as well as depression, alcohol dependence, and substance abuse, compared to heterosexual women (King et al., 2008). This association has been found in populations such as Norwegian youth (Hegna & Wichstrøm, 2007), young adults in New Zealand (Skegg, Nada-Raja, Dickson, Paul, & Williams, 2003), adolescents in Guam (Pinhey & Millman, 2004), sexual minorities in Japan (DiStefano, 2008), and even in large Internet-based samples (Mathy, 2002) and in studies using prospective cohort designs (Mustanski & Liu, 2013). For instance, in one early paper using nationally representative data from the USA, Russell and Joyner (2001) showed that the prevalence of suicide ideation was much higher among lesbian and bisexual adolescent girls (28%) compared to heterosexual peers (15%). Suicide attempt prevalence was likewise elevated (12% versus 5%). A more recent analysis suggests that bisexual women may have the highest levels of suicide ideation and attempted suicide, compared to lesbian and heterosexual women, at least in the United States (Bolton & Sareen, 2011). The overall finding from the literature is that significant disparities exist for sexual minority populations—lesbian and bisexual girls and women are at disproportionately elevated risk for suicidal behaviors than heterosexual peers (Hass et al., 2011).

In line with current understandings of sexuality in psychology, contemporary research has shown that sexual orientation per se does not determine poor mental health outcomes (Meyer & Frost, 2013). Disparities in mental health outcomes among lesbian and bisexual women are in part due to minority stressors─the set of stressors experienced by sexual minorities who live in a culture of anti-LGBT social stigma. Across different cultures, sexual minorities like lesbian and bisexual women often grow up, work, and live in contexts of heteronormativity (Jackson, 2006) where cultural norms operate around and privilege heterosexuality, while stigmatizing lesbian and bisexual women, relationships, and populations, positioning same-sex love and relationships as nonexistent, deviant, or “immoral” and at the same time defining relationships between a man and a woman as good, “normal,” and “natural.” These conditions lead to stressors that include experiences of homonegative discrimination, pressure to conceal one’s sexuality, social rejection for being lesbian or bisexual, and self-stigma (“internalized homophobia”). According to the minority stress model, these minority stressors join normative stressors—the demands present in modern life for the general population, such as major life events, conflict, violence, poverty, etc.—in impacting on lesbian and bisexual mental health and well-being (Meyer & Frost, 2013).

Understanding lesbian and bisexual suicide risk requires examination of the social ecology of sexual minority populations wherever they may be located (Hong, Espelage, & Kral, 2011). In the Philippines, the minority status of lesbian and bisexual Filipinas is evident at structural and individual levels (Lim & Jordan, 2013). Sexual orientation here is not a fully protected category, though it was mentioned in the Magna Carta of Women (Ruizo, 2013), a national law that broadly seeks to eliminate discrimination against women in the Philippines. A handful of cities have legislated local anti-discrimination policies but only beginning the year 2003 (Manalastas, Ortega, Encinas-Franco, & Arugay, 2015), and even now there is no specific national law prohibiting anti-LGBT discrimination in employment, education, and health care, despite what one observer has described as “one of the longest running campaigns in the Philippines” (Thoreson, 2012, p. 552) by local LGBT rights activists and advocates in civil society, a campaign dating back to the 1990s. Though being lesbian or bisexual has never been criminalized as in neighboring Southeast Asian countries like Malaysia and Singapore (UNDP, USAID, 2014), no legal recognition of same-sex partnerships exists, restricting access of lesbian and bisexual women, especially those living in poverty, to important social protections such as social security, socialized housing, and government- administered loans which are enjoyed by married heterosexual couples (Lim & Jordan, 2013). At the same time, powerful anti-LGBT religious institutions like the Roman Catholic Church impose conservative moral ideologies, influencing the public as well as state policy on matters like education, LGBT representation in government, and sexual health services (Ruiz-Austria, 2004). As a country in Asia where majority of the population is Roman Catholic, traditional religion is often used as a means to suppress equality for sexual minorities, as when clergy publicly oppose proposed anti-discrimination laws, claiming that they will pave the way for “immoral” practices like same-sex marriage (UNDP, USAID, 2014). Finally, public opinion research also indicates that anti-lesbian prejudice is alive and well among Filipinos. Though Filipinos are comparatively less homonegative than survey respondents in Indonesia and Malaysia but have similar levels of homonegativity as Singaporean and Vietnamese respondents (Manalastas et al., 2015), nationally representative surveys from the mid-1990s to the present indicate that about 1 out of 4 Filipinos would reject lesbian and gay people as neighbors and that 28% of the Filipino adult population consider being lesbian as morally “unacceptable” (Manalastas & del Pilar, 2005; Manalastas, 2015).

The literature on suicide is scarce in the Philippines, where suicide mortality rates are officially low but increasing in incidence over the past three decades, possibly due to better reporting (Redaniel, Lebanan-Dalida, & Gunnell, 2011). The most recent available official estimates by the World Health Organization (2014) peg the suicide mortality rate for women ages 15 to 29 in the Philippines at 2.2 per 100,000 which is higher than in Malaysia (1.0 per 100,000) and in Vietnam (1.8 per 100,000) but relatively lower than in Indonesia (3.6 per 100,000), the United States (4.7 per 100,000), Singapore (6.0 per 100,000), Japan (10.8 per 100,000), and South Korea (14.9 per 100,000). Similar to the global mental health findings, suicide mortality among Filipinos is highly gendered, with a male-to-female suicide fatality ratio of 3.3:1 but with higher incidence of non-fatal suicide attempts among women, especially in the 15 to 24 age group (Redaniel, Lebanan-Dalida, & Gunnell, 2011).

In terms of disparities along sexual orientation, findings from a national sample of young Filipino men ages 15 to 24 indicate that gay and bisexual youth in the Philippines have disproportionately higher levels of non-fatal suicide behaviors compared to heterosexual peers (Manalastas, 2013). Sexual minority male youth had more than two times greater odds than heterosexual peers to report suicide ideation (odds ratio=2.09), and among those reporting suicide ideation, 39% of sexual minority male youth had attempted suicide, compared to 26% of heterosexual counterparts. This study was the first published report on mental health and suicidality disparities for sexual minorities, albeit only among men, based on a national sample in Southeast Asia.

For this paper, I present a secondary analysis of data from sexual minority and heterosexual female youth in the Philippines to provide evidence of disparities in suicide ideation and attempt among women. In addition, I explore other risk factors aside from sexual minority status available in the dataset, as suggested by minority stress theory (Meyer & Frost, 2013).

What is the prevalence of suicide ideation and attempted suicide for sexual minority female youth in the Philippines? And following the global literature on LGBT mental health, are young Filipina lesbian and bisexual women at disproportionately higher risk for suicide ideation and suicide attempt, compared to the general population? Specifically the goals of this analysis were twofold: (a) to determine the prevalence of suicide ideation and attempted suicide among young sexual minority Filipina women using existing nationally representative data; and (b) to examine any disparities in suicide ideation and attempted suicide among sexual minority Filipina youth, versus the general population. Additionally, I explore possible factors available in the data that could contribute to suicide ideation and attempt among lesbian and bisexual youth.

I reanalyzed archival survey data from the Third Young Adult Fertility & Sexuality Study (YAFS3), a large questionnaire-based interview study conducted by the University of the Philippines Population Institute. The data are based on a nationally representative sample of young Filipino adults from 16 administrative regions of the Philippines including the National Capital Region and the Autonomous Region of Muslim Mindanao, and including 894 barangays (administrative neighborhoods), using two-stage cluster sampling. Respondents in the female subsample were 8,891 young Filipina women ages 15 to 24 (M=18.9, SD=2.9). Though collected a decade ago, the nationally representative YAFS3 dataset remains the largest and most comprehensive publicly available source of information for Filipino young adult sexuality and health outcomes (Raymundo & Cruz, 2004). YAFS3 is also the first nationally representative dataset in the Philippines to include items that allow analysis along women’s sexual orientation.

Suicide ideation and attempt. Respondents were asked if they had ever thought about suicide. Second, for those reporting suicide ideation, a follow-up probe asked if the respondents had ever attempted suicide. Both items were dichotomously scored (yes/no). In the overall dataset, 18.7% of women reported suicide ideation. Among those with ideation, 24.1% indicated having attempted suicide (4.5% of the overall women’s sample).

Sexual minority status. Sexual minority status was operationalized in the dataset in three ways: self-reported same-sex attraction, same-sex sexual behavior, and same-sex romantic relationships. For the attraction item, respondents were asked if they were attracted to men, to women, or to both. Those reporting attractions to other woman were classified as sexual minority. Missing responses were excluded from the analysis (n=254, 2.9%). In the analytic sample, 10.3% of women (n=885) reported attractions to both men and women, while 1.31% (n=113) reported attractions to other women exclusively. The rest (88.45%, n=7,639) were attracted only to men. Because of the centrality of attraction in the conceptual definition of sexual orientation and sexual minority status (IOM, 2011) as well as the low proportions on the sexual behavior and same-sex relationships items, this paper reports on analyses based on the attraction measure. To achieve sufficient statistical power, data from respondents with attraction to women exclusively and to both women and men were pooled during analysis to form the sexual-minority classification similar to precedents in the LGBT health literature (e.g., Russell & Joyner, 2001; Russell, 2003).

Normative risk factors. A number of variables in the dataset were available for inclusion in the analysis as general risk factors for suicide ideation and attempt. These normative factors are not specific to lesbian and bisexual youth and may apply to the entire female youth population. The general risk factors included were: (a) depressive symptomatology in the past three months, a three-item continuous scale ranging from 3 to 12, higher scores indicating higher levels of depressive symptoms (overall M=5.14, SD=1.81, Cronbach’s alpha=0.75); (b) recent peer suicide attempt, a dichotomous item (yes=23.9%); (c) experience of being physically threatened in the past three months, a dichotomous item (yes=16.7%); and (d) experience of being physically attacked by someone in the past three months, a dichotomous item (yes=10.7%). Global self-esteem was measured using a six-item five-point Likert-type scale (overall M=23.5, SD=3.1, Cronbach’s alpha= 0.67). Age, measured as a continuous variable (in years), was also included as a control variable, as in other suicide ideation and attempt research (Russell & Joyner, 2001; Russell, 2003).

Separate cross-tabulation analysis of the two nonfatal suicide behavior indicators and sexual minority status were conducted in order to determine the prevalence of suicide ideation and suicide attempt among lesbian and bisexual Filipina youth, relative to heterosexual peers, and to test the hypothesis of disparities across sexual orientation. In addition, I ran logistic regressions with suicide ideation and attempt as the outcome variables. Sexual minority status was entered in a model along with age as a control to test the main hypothesis. Normative risk factors (depression, peer suicide attempt, threat, and victimization) were entered in a second regression model to determine if these general stressors could partly account for any effects of sexual minority status, as posited by the minority stress model (Meyer & Frost, 2013).

Suicide ideation, as hypothesized, was disproportionately higher among sexual-minority Filipina youth, compared to heterosexuals (see Figure 1). Young lesbian and bisexual Filipina women had a higher rate of suicide ideation compared to other female youth. About one out of four women who were attracted to other women reported having thought of suicide (27%), disproportionately higher than other youth (18%),  (1)=43.84, Cramer’s V=.07, p<.001.

Figure 1. 
Disproportionately Higher Rates of Suicide Ideation and Attempt among Sexual-minority Filipina Youth versus Heterosexual Peers

Sexual minority youth also had disproportionately higher rates of suicide attempt than heterosexual peers. More young bisexual and lesbian Filipina women reported attempting suicide (6.6%), than other young women (3.9%),  (1)=16.71, Cramer’s V=.04, p<0.001.

Results from the logistic regression analyses for suicide ideation and attempt are presented here. In the first analysis, an analytic sample of N=8,487 cases was used. Prediction success was 81.04%. Odd ratios for the different predictors in the model are presented in Table 1. The initial model indicated that age and sexual minority status were both significantly associated with suicide ideation among young Filipina women. Specifically, sexual minority women had 67% higher odds of having thought of suicide, compared to other women,  (1)=44.42, p<.001, odds ratio=1.67 (95% CI=[1.44, 1.95]). This key finding supports the hypothesis of elevated suicide ideation for sexual minority women in the Philippines.

In addition to sexual minority status as a predictor for suicide risk, normative stressors were significantly associated with thinking about suicide, namely: peer suicide attempt,  (1)=122.34, p<.001, odds ratio=2.98; experience of being threatened,  (1)=49.04, p<.001, odds ratio=1.65; victimization,  (1)=17.20, p<.001, odds ratio=1.42; and depression,  (1)=224.55, p<.001, odds ratio=1.26. Increasing global self-esteem was associated with lower odds of suicide ideation,  (1)=4.82, p<.05, odds ratio= 0.98. Filipina female youth regardless of sexual orientation who knew of a friend attempting suicide, who had experienced being threatened, who had been the victim of interpersonal violence, or who felt depressed were more likely to consider taking their own life. The results indicate that while experiences of threat, victimization, and depression do predict suicide ideation among Filipina youth as expected, sexual minority status (i.e., being lesbian or bisexual) still remained a significant predictor over and above those normative factors as suggested by the minority stress model (Meyer & Frost, 2013) and as found in previous research (Russell & Joyner, 2001; Russell, 2003).

Similar results for suicide attempt were obtained. Odd ratios for this set of analyses are shown in Table 2. The initial model, based on an analytic sample of N=8,631 cases with prediction success at 95.44%, indicated sexual minority status was significantly associated with suicide attempt among young Filipina women. Sexual minority women had twice higher odds of having attempted suicide, compared to other women,  (1)=32.9, p<.001, odds ratio=2.13 (95% CI=[1.65, 2.76]), supporting the hypothesis of elevated suicide risk, in both suicide ideation and attempt, for sexual minority women in the Philippines.

Apart from sexual minority status as a correlate for suicide attempt, normative stressors were again significantly associated with suicide risk, according to a regression model with all predictors (analytic sample n=8,484; prediction success 95.43%). These factors were peer suicide attempt,  (1)=34.91, p<.001, odds ratio=1.92; experience of being threatened,  (1)=31.07, p<.001, odds ratio=1.95; victimization,  (1)=28.89, p<.001, odds ratio=2.02; and depression,  (1)=115.70, p<.001, odds ratio=1.30. Increasing global self-esteem was likewise associated with lower odds of suicide risk,  (1)=6.41, p<.01, odds ratio=0.96.

Examination of the odds ratios across the two regression models similarly indicated that while factors like experiencing threat, victimization, and depression may drive suicide attempt among Filipina youth, sexual minority status (i.e., being bisexual or lesbian) again remained a significant predictor for suicide attempt over and above those normative stressors as found in previous research (Russell & Joyner, 2001; Russell, 2003) and predicted by the minority stress model (Meyer & Frost, 2013).