According to the World Health Organization (2017), suicide mortality has increased among young people ages 15–29 years and is the second leading cause of death globally. Suicide is identified as a “priority condition,” targeted for at least a 30% reduction between 2013 and 2020 (Saxena & Setoya, 2014). Depression, a major contributing factor to suicide, has a global prevalence over 5% among females and 3% among males (World Health Organization, 2017). A comparison of 14-year-old adolescents in the United Kingdom (2005 vs. 2015) showed that depressive symptoms had become more common in 2015 compared to 2005 (Patalay & Gage, 2019).
Currently, cross-national estimates for these factors are incomplete, vary from country to country, and are difficult to compare, but overall trends in suicidal ideation and suicide attempts increased from 2003 to 2014 globally—especially among females (McLoughlin, Gould, & Malone, 2015). The Global School-Based Student Health Survey of 13- to 17-year-olds in 32 low- and middle-income countries estimates that suicidal ideation prevalence is 16.2% among females and 12.2% among males (McKinnon, Gariépy, Sentenac, & Elgar, 2016). Although surveillance standards vary across countries, these results of low- and middle-income countries appear similar to high-income European countries—although low- and middle-income countries bear the largest global burden of suicide (Kokkevi, Rotsika, Arapaki, & Richardson, 2012).
Within the United States, suicide rates parallel global trends: Suicide mortality increased substantially among adolescents (15–19 years) from 2007 to 2015. Female suicide mortality per 100,000 in this age group more than doubled from 2.4 in 2007 to 5.1 in 2015 (Centers for Disease Control and Prevention (CDC), 2017). Adolescent and young adult depression rates also increased in the United States from 8.75% in 2005 to 11.3% in 2014 (N = 172,495; Mojtabai, Olfson, & Han, 2016).
Suicide among adolescents and young adults is largely preventable (Labouliere et al., 2018), and a reduction in suicide attempts is an important population health goal incorporated into the U.S. Office of Disease Prevention and Health Promotions’ Healthy People 2010 and 2020 objectives. The Youth Risk Behavior Survey (YRBS)—a nationally representative biennial health survey of U.S. high school students—is designated as the data source for tracking progress toward 24 of these goals, including suicide attempts (U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, 2010). During the period covered by Healthy People 2010, the rate of suicide attempts that required treatment decreased 26.9% from 26 suicide attempts per 1,000 students (1999) to 19 suicide attempts per 1,000 students (2009). Since then, the rate of suicide attempts that required treatment increased 47% from 19 attempts per 1,000 students (2009) to 28 suicide attempts per 1,000 students (2015; CDC, 2017). Interactions between gender and trends for suicidality have not been reported for 2009–2017.
Earlier trends (1991–2011) from YRBS showed linear decreases in the prevalence of suicidal thoughts and attempts among female students (Lowry, Crosby, Brener, & Kann, 2014). A quadratic trend for female suicidal ideation was identified: The prevalence of having seriously considered suicide decreased from 37.2% (1991) to 17.4% (2009) and then increased to 19.3% (2011). In contrast, male students had negative linear trends for the prevalence of suicidal ideation, but not suicide attempts (Lowry et al., 2014). National YRBS results from 1991 to 2015 showed no linear trend (increase or decrease) overall in suicide attempts requiring treatment (Kann et al., 2016). Trends for rates of suicide attempts that required treatment have not been examined by gender from 2009 to 2017.
Overall, the rate of suicide attempts decreased 26.9% between 1999 and 2009 (National Institutes of Health, Substance Abuse and Mental Health Services Administration, 2012). Since then, YRBS results indicate that suicide attempts requiring treatment increased from 2009 to 2015; trends by gender were not estimated (CDC, 2017) and are an important variable to explore.
Interaction Between Gender and Linear Time Trend (Year) for Depressive Symptoms and Suicidality
Previous research with YRBS data showed that, among female students, a positive linear trend for depressive symptoms and suicidality occurred during 2007–2017 (CDC, 2018) and 2009–2017 (Ford et al., 2018). Among male students, there was no positive linear trend for depressive symptoms or suicidality during 2007–2017 (CDC, 2018). The CONSORT 2010 (Item 12b, p. 14) recommends that the correct analysis to test for differences in effects across subgroups is to perform statistical tests for interactions (Moher et al., 2010). It is incorrect to infer a significant interaction effect from one significant and one nonsignificant p value for separate analyses within subgroups (Altman & Bland, 2003; Moher et al., 2010).
Rationale for Estimation of Additive and Multiplicative Interactions
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines recommend the estimation of additive interactions as the additive scale is more appropriate for public health decision-making (Vandenbroucke et al., 2007, p. 1639). It should be noted that rate differences are measures on an absolute scale, and an interaction estimated with a difference measure of association (rate difference) is an additive interaction (Rothman, 2014). The absence of an additive interaction between female gender and linear time trend on suicide attempts that required treatment implies that the linear trend for the rate of suicide attempts in females (RDF) is not significantly different from the linear trend for the rate of suicide attempts in males (RDM). The absence of an additive interaction between female gender and year on suicide attempts that required treatment implies that the rate difference for suicide attempts among females between 2009 and 2017 (RDF) is not significantly different from the rate difference for suicide attempts among males between (RDM). The additive interaction is significant if the difference (RDF – RDM) > 0 and the 95% confidence interval (CI) for the difference does not include 0. Because STROBE guidelines recommend additive interactions, we estimate the additive interactions between female gender and linear time trend.
Although risk differences are the recommended effect size when testing for interactions (Vandenbroucke et al., 2007), odds ratios are commonly used as the effect size in suicide research (Holt et al., 2015; van Geel, Vedder, & Tanilon, 2014). Interactions estimated with a ratio measure of association (odds ratio) are multiplicative interactions (Rothman, 2014). The absence of a multiplicative interaction between female gender and linear time trend on suicide attempts that required treatment implies that the odds ratio for the linear trend for suicide attempts in females (ORF) is not different from the odds ratio for the linear trend for suicide attempts in males (ORM). The absence of a multiplicative interaction between female gender and year on suicide attempts that required treatment implies that the odds ratio for suicide attempts among females between 2009 and 2017 (ORF) is not different from the odds ratio for suicide attempts among males between 2009 and 2017 (ORM). The multiplicative interaction is significant if the ratio ORF/ORM > 1, and the 95% CI for the ratio does not include 1.
Among 15- to 19-year-olds, research showed that the rate of suicide mortality increased substantially from 2007 to 2015: females from 2.4 to 5.1 and males from 10.8 to 14.2. Thus, calculations show that, in this age group, during 2007–2015, suicide mortality increased among males by 31% and females by 113% (CDC, 2017). Therefore, on a relative risk (ratio) scale, suicide mortality increased substantially more among females (113%) than males (31%). Yet on an additive (rate difference) scale, suicide mortality increased by 3.4/100,000 among males and 2.7/100,000 among females; on a rate difference scale, the increase in suicide mortality was greater among males. These results highlight the importance of following STROBE guidelines and estimating and reporting rate differences and additive interactions.
This study uses U.S. high school YRBS data (2009–2017) to investigate additive interactions between female gender and (a) linear time trend and (b) survey year for depressive symptoms, suicidal ideation (seriously considered or planned suicide), suicide attempts, and suicide attempts that required treatment by a doctor or nurse. A major contribution of the present research is the estimation of adjusted additive interactions between female gender and (a) linear time trends and (b) survey year as recommended by STROBE guidelines (Vandenbroucke et al., 2007). Results show divergent trends for suicide attempts by gender and reveal that suicide attempts have increased substantially for female students but not male students.
The YRBS is a three-stage, nationally representative survey of U.S. high school students conducted biennially by the CDC since 1991. Student participation in the research was anonymous and voluntary. The paper-and-pencil survey was self-administered on computer-scannable paper during regular class time with the assistance of trained data collectors and included more than 80 questions on school bullying, electronic bullying, fighting, suicidality, substance use, sexual behavior, body image, nutrition, physical activity, and other health behaviors (CDC, 2018). The CDC’s internal review board approved the protocol for the YRBS and gave exempt status for this project because we used de-identified data that are publicly available with unrestricted access for researchers (CDC, 2013). Details of the complex sampling design, data collection methodology, and survey response rates for the national YRBS are available from the CDC (http://www.cdc.gov/healthyYouth/data/yrbs/index.htm). For these data sets, sampling design variables and sample weights are provided, which need to be incorporated for the estimation of nationally representative population statistics (CDC, 2018). In order to perform trend analyses (2009–2017), data were combined from each of the cross-sectional surveys conducted in years 2009, 2011, 2013, 2015, and 2017 (CDC, 2018). The five-wave YRBS sample contained complete data on 75,807 students. In order to compare the changes between 2009 and 2017, data were combined from the cross-sectional surveys conducted in years 2009 and 2017. The two-wave YRBS sample contained complete data on 31,175 students.
Dependent and Independent Variables
The dependent variable to measure depressive symptoms was the survey item that measured feelings of sadness or hopelessness that lasted at least 2 weeks during the past 12 months with a dichotomous yes/no response. The dependent variable to measure suicide ideation was constructed from two survey items: (a) “During the past 12 months, did you ever seriously consider attempting suicide?” and (b) “During the past 12 months, did you make a plan about how you would attempt suicide?” Respondents who indicated “yes” to either question were coded as “yes” for suicide ideation. The dependent variable to measure “suicide attempt” was constructed from two survey items: (a) “During the past 12 months, how many times did you actually attempt suicide?” and (b) “If you attempted suicide during the past 12 months, did any attempt result in an injury, poisoning, or overdose that had to be treated by a doctor or nurse?” Respondents who indicated that they attempted suicide one or more times or “yes” to the second question were coded as “yes” for suicide attempt. The dependent variable to measure attempted suicide that required treatment from a doctor or nurse was the corresponding item in the YRBS questionnaire (note that attempted suicides were measured by self-report; they exclude fatal suicide attempts).
The independent variables were gender (male and female), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, non-Hispanic, other), grade level (Grades 9, 10, 11, 12), and survey year (note that the race/ethnicity variable is precoded in YRBS). Gender and grade level were included in the model because previous research with YRBS data showed that female gender, race/ethnicity, and grade level were related to depressive symptoms and suicidal behaviors (Kann et al., 2016). To prevent confusion, we use the term “gender” instead of “sex” in our model, because several YRBS questions about sexual activity also use the term “sex.” For analytic purposes, gender and race ethnicity were coded as categorical variables, grade level, and survey year were coded as continuous linear covariates (note that consecutive integers were used for consecutive survey waves). For each of the dependent variables, the logistic regression model was estimated for all students and separately (without gender as independent variable) for male and female students. All results reported are rate difference (RD) per 1,000 male (female) students.
For this research, multivariate analyses were performed with R and the R survey package, which incorporate the sampling design variables and sampling weights and generate nationally representative weighted estimates (Lumley, 2017; R Core Team, 2018). Estimates of multivariate additive and multiplicative interactions were performed as described (Lumley, 2017). The R survey package function, “svypredmeans,” was used to estimate adjusted marginal percentages by gender, and the R survey package function, “svycontrast,” was used to estimate adjusted risk differences and their respective CIs by gender (Lumley, 2017; White = Non-Hispanic White, Black = Non-Hispanic Black, Other = Non-Hispanic other).
Additive Interactions: Gender and Linear Time Trend/Survey Year
Results showed that there were positive additive interactions between female gender and linear trend for depressive symptoms, RD = 15, t = 3.12, p = .002, suicide ideation, RD = 13, t = 3.49, p < .001, and suicide attempts that needed treatment, RD = 3, t = 2.47, p = .014 (Table 1 and Figure 1). The additive interaction between female gender and linear trend was nonsignificant for suicide attempts, RD = 4, t = 1.48, p = .140 (Table 1 and Figure 1). There were also positive additive interactions between female gender and year (2017 vs. 2009) for depressive symptoms, RD = 51(69 minus 19), t = 2.37, p = .020, suicide ideation, RD = 36 (45 minus 9), t = 2.25, p = .028, and suicide attempts that needed treatment, RD = 10 (8 minus -2), t = 2.04, p = .045 (Table 1 and Figure 2). The additive interaction between female gender and year was nonsignificant for suicide attempts, RD = 7 (9 minus 3), t = 0.65, p = .520 (Table 1 and Figure 2).
Multiplicative Interactions: Gender and Linear Time Trend/Survey Year
Results showed that there were positive multiplicative interactions between female gender and linear trend for depressive symptoms, OR = 1.06, t = 2.53, p = .012, suicide ideation, OR = 1.07, t = 2.85, p = .005, and suicide attempts that needed treatment, OR = 1.11, t = 2.03, p = .044 (Table 1). The multiplicative interaction between female gender and linear trend was nonsignificant for suicide attempts, OR = 1.03, t = 1.01, p = .314. There were no significant positive multiplicative interactions between female gender and year (2017 vs. 2009) for depressive symptoms, OR = 1.19, t = 1.71, p = .092, suicide ideation, OR = 1.19, t = 1.70, p = .094, suicide attempts, OR = 1.05, t = 0.30, p = .764, and suicide attempts that needed treatment, OR = 1.11, t = 1.78, p = .079 (Table 1; none of the multiplicative interactions between female gender and year [2017 vs. 2009] were significant).
Trends for Rates of Depressive Symptoms and Suicidality by Gender (2000–2017)
Results showed that, among female students, there were positive linear time trends in the likelihood of depressive symptoms, RD = 18, t = 4.00, p < .001, and suicidal ideation, RD = 14, t = 4.50, p < .001 (Table 2). In contrast, among male students, there was no trend in the likelihood of depressive symptoms, RD = 3, t = 1.03, p = .305, or in the likelihood of suicidal ideation, RD = 2, t = .82, p = .413 (Table 2). Results showed that, among female students, there was a positive linear time trend in the likelihood of suicide attempts that required treatment by a doctor or nurse, RD = 2, t = 2.26, p = .025, but no positive linear time trend in the likelihood of suicide attempts, RD = 4, t = 1.66, p = .094 (Table 3). In contrast, among male students, there were no trend in the likelihood of suicide attempts, RD = 0, t = 0.31, p = .756, or of suicide attempts that required treatment RD = −0, t = −0.56, p = .578 (Table 3).
Changes in Rates of Depressive Symptoms and Suicidality
A comparison of rates, 2017 versus 2009, showed that, for female students, the rate of depressive symptoms was greater in 2017 (408) versus 2009 (340), RD = 69, t = 3.52, p < .01, and the rate of suicidal ideation was greater in 2017 (252) versus 2009 (207), RD = 45, t = 3.29, p < .001 (Table 2). For male students, neither of the comparisons were significant (Table 2). A comparison of rates, 2017 versus 2009, showed that neither of the comparisons for either suicide attempts or suicide attempts that required treatment were significant, either among male students or among female students (Table 3).
Comparison by Gender (2009–2017) for Suicide Attempts That Required Treatment
For every YRBS wave, there was a difference between female students and male students on the rate of suicide attempts that required treatment (except for 2009, where the gender rate difference was not different, RD = 6, t = 1.92, p = .056; Table 3). The rate difference (rate difference female students − rate difference male students) for suicide attempts that required treatment increased from 6 (2009) to 10 (2011) to 18 (2013). From 2013 onward, the rate difference was essentially constant, 17 (2015) and 16 (2017; Table 3).
Progress Toward Healthy People 2020 Goals by Gender
Healthy People 2020 sets a population health goal among high school students for a 10% decrease (2009–2019) in suicide attempts that required treatment and designated the YRBS to measure progress toward this goal (U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, 2010). Among female students, the rate of suicide attempts per 1,000 students increased from 22 to 30—an increase of 36%. These results show that female students in the United States are further from the Health People 2020 goal in 2017 than in 2009. In contrast, among male students, the rate of suicide attempts per 1,000 students decreased from 16 to 14, a decrease of 13%, which shows that male high school students in the United States attained the Health People 2020 goal by 2017.
These results are the second reported example with YRBS data demonstrating the need for surveillance by gender to determine progress toward Healthy People 2020 goals. Previous research with YRBS data showed that, between 2009 and 2015, the rate of school bulling victimization per 1,000 U.S. high school students increased by 17% among females (from 212 to 248), decreased 16% among males (from 187 to 158), and remained essentially unchanged among all students (from 199 to 202; Pontes, Ayres, Lewandowski, & Pontes, 2018). Given the divergent trends between male and female students, these results identify the need for surveillance by gender when measuring progress toward Healthy People goals.
Comparison Between Additive and Multiplicative Interactions
Results showed positive additive interactions as well as positive multiplicative interactions between female gender and linear time trend for depressive symptoms, suicide ideation, and suicide attempts that required treatment. However, results showed positive additive interactions but no positive multiplicative interactions between female gender and survey year (2009 vs. 2017) for depressive symptoms, suicide ideation, and suicide attempts that required treatment. Thus, for interactions between gender and survey year (2009 vs. 2017), the estimation of additive or multiplicative interactions results in different inferences about whether the increase in either (a) depressive symptoms, (b) suicide ideation, or (c) suicide attempts (2009 vs. 2017) was significantly greater in female rather than male students.
The results of this research add to the growing literature that additive and multiplicative interactions may lead to different statistical inferences when comparisons of effect sizes across strata are performed (Rothman, 2014; Vandenbroucke et al., 2007). Estimates of additive interaction with YRBS data showed that the relationships between bullying and depressive symptoms or suicidality were greater among females (Pontes, Ayres, & Pontes, 2018). In contrast, estimates of multiplicative interaction with the same YRBS data showed that the relationships between bullying and depressive symptoms or suicidality were not different among gender (Pontes, Ayres, & Pontes, 2018). Because STROBE guidelines recommend the estimation of additive interactions for public health decision-making (Vandenbroucke et al., 2007, p. 1639), these results highlight the importance for researchers to follow the STROBE guidelines and report additive interactions (and risk or rate differences) because multiplicative interactions can sometimes lead to incorrect inferences about the variation in effect sizes across strata.
Effect Sizes of Additive Interactions Are Relevant to Public Health
The rate difference (2017–2009) for suicide attempts that required treatment is 8/1,000 female students and −2/1,000 male students. This result indicates that if we compared 1,000 female (male) students in 2017 versus 1,000 female (male) students in 2009, we would expect to see 8 (−2) more females (males) make suicide attempts that required treatment in 2017 than in 2009. The additive interaction coefficient is the difference, 8 − (−2) = 10. Thus, estimation of the additive interaction coefficient provides a direct measure of how the increase over time (2009–2017) in the rate of suicide attempts that required treatment is greater among females than males by 10 and provides a 95% CI for this difference in rate difference: 10, 95% CI [0, 20], t = 2.04, p = .045.
This research highlights the importance of exploring gender and gender identity differences related to suicidal ideation and suicide attempts among adolescents (King, Horwitz, Czyz, & Lindsay, 2017). These results add to the growing literature about the need for more global surveillance related to depressive symptoms and suicidal behavior trends by gender. Current transnational surveillance comparisons lack uniformity, and comparing data interchangeably is challenging because of the question variability: Some surveys ask about suicidal behavior in the previous 12 months, whereas others ask about lifetime prevalence, and the age range of participants varies. These differences limit comparisons of current global surveillance.
Social Determinants That Correlate With Suicidality
Specific social determinants that correlate with suicidality in both male and female high school students are important to address. For example, bullying victimization, social isolation, depressive symptoms, previous suicide attempts by the student or family, community violence, school absence, lack of sleep, lack of social support, gender identity issues, substance use, sexual activity, perceived over- or underweight, and lack of physical activity have all shown an association with suicidality among this age group (Hill, Oosterhoff, & Kaplow, 2017; King et al., 2017; Lowry et al., 2014; Pontes, Ayres, & Pontes, 2018). Gender-specific associations with suicidality among high school females are injection drug use, weapon carrying, and methamphetamine use; suicidality in males is associated with a history of forced sex, disordered eating, alcohol use, and singleness (King et al., 2017; Lowry et al., 2014).
Felitti et al. (1998) reported the association between adverse childhood experiences and health risks among adults later in life. More recently, adolescents who reported adverse childhood experiences also reported more suicide attempts compared to those without adverse childhood experiences (Fuller-Thomson, Baird, Dhrodia, & Brennenstuhl, 2016). One common form of adverse childhood experiences is bullying victimization, which is an important risk factor for depressive symptoms and suicidality among youth; additive interactions demonstrate that the effects of bullying victimization on depressive symptoms and suicidality are greater among female students compared to male students (Pontes, Ayres, & Pontes, 2018). Previous research with YRBS data showed that during 2009 to 2015 the rate of school bullying victimization increased among females and decreased significantly among males (Pontes, Ayres, Lewandowski, & Pontes, 2018). One partial explanation of the present research finding—that the rate of depressive symptoms and suicidal behaviors increased more among females than among males during 2009 to 2017—is that the rate of school bullying victimization increased more among females than males during this same period and thus could be a contributing factor.
Implications for Schools
School suicide prevention programs are important because they are associated with a reduction in suicide among youth in the United States (Fond et al., 2016) and most countries studied globally (Kölves & De Leo, 2016). A study of European countries estimated a median decrease by 25% between 1990 and 2009 after suicide prevention programs were initiated (Fond et al., 2016). However, data are limited following 2009; the same time period that suicidal behaviors increased among U.S. high school students, according to our findings.
School personnel have an important role in ensuring a safe environment for students and fostering their sense of belonging and connectedness—free from school-related victimization. Researchers have shown that victimization at school and perceived lack of safety on school grounds are significant risk factors for suicidal behaviors, but further exploration is needed—especially among female students (Jiang, Perry, & Hesser, 2010; Messias, Kindrick, & Castro, 2014; Pontes, Ayres, & Pontes, 2018). School personnel, including school nurses and other mental health providers, have a unique opportunity—and a significant responsibility—to prevent suicidal behaviors and the school-related social determinants associated with them.
Implications for Nursing and Other Health Professionals
Unfortunately, prevalence rates of suicidal behaviors continue to increase into adulthood—which highlights the importance of primary prevention efforts from early adolescence—and throughout the transition into early adulthood and beyond (Han et al., 2018; Wang et al., 2016). Previous research has shown that adolescents are likely to seek healthcare services—mental health for females and primary care for males—within the month prior to their suicide attempt (King et al., 2017), which gives health professionals the opportunity to intervene. Widespread training of nurses and health professionals using evidence-based “gatekeeper” screening such as “Question Persuade Risk” or “Applied Suicide Intervention Skills” is recommended (Bolster, Holliday, Oneal, & Shaw, 2015). Even mental health nurses and other mental health professionals can benefit from this training, as they may also omit direct questions about suicidal thoughts and behaviors; these trainings encourage a direct approach with specific guidelines (Bolster et al., 2015; Litteken & Sale, 2018; O’Reilly, Kiyimba, & Karim, 2016).
Despite the importance of suicide screening, antecedent factors, such as adverse childhood experiences, should also be identified and addressed through a trauma-informed approach (Erlich & GAP Committee on Psychopathology, 2016); however, this approach is not a standard part of most nursing curricula (Li et al., 2019). Furthermore, gender-specific interventions must address other specific social determinants related to the increasing prevalence of suicide attempts among females. Most importantly, nurses have the expertise to provide leadership in setting national and global standards for prevention and early identification, treatment, and referral for adolescents with depressive symptoms and suicidal ideation or suicide attempts.
As secondary analysis, this research is limited by previously determined questions and data collection methods. The reliance upon self-report is a limitation, but responses to self-administered questionnaires have a higher validity than responses to face-to-face or telephone interviews. This study also relies upon recall, so it is possible that answers will be inaccurate due to telescoping errors or social desirability bias. Previous reliability and validity testing have shown this questionnaire to be suitable for high school students (estimates are accurate within ±5% at 95% confidence level), although over- and underreporting is possible (CDC, 2013). The lack of self-identification as transgendered is a limitation of the survey.
Despite the Healthy People 2020 goal to reduce adolescent suicide attempts by 10% that required treatment, we are further from this goal now than we were in 2009. This sobering reality is a call to action for nurses, other health professionals, educators, and researchers. Specific efforts should address divergent trends of depressive symptoms and suicidality by gender—especially among females. Nurses should be at the forefront of these efforts: leading research that can inform better practice models for suicide prevention among adolescents, training professionals, and communities in prevention and in implementing best practices for suicide prevention and trauma-informed care.
Altman D. G., & Bland J. M. (2003). Interaction revisited: The difference between two estimates. BMJ
, 326, 219. doi:
Bolster C., Holliday C., Oneal G., & Shaw M. (2015). Suicide assessment and nurses: What does the evidence show? OJIN: The Online Journal of Issues in Nursing
, 20, 81–87. doi:
Centers for Disease Control and Prevention (2013). Methodology of the youth risk behavior surveillance system—2013. Morbidity and Mortality Weekly Report: Recommendations and Reports
, 62(RR-1), 1–20. Retrieved from https://www.cdc.gov/mmwr/pdf/rr/rr6201.pdf
Centers for Disease Control and Prevention (2017). QuickStats: Suicide rates*,†
for teens aged 15–19 years, by sex—United States, 1975–2015. Morbidity and Mortality Weekly Report
, 66, 816. doi:
Centers for Disease Control and Prevention. (2018). YRBSS data and documentation
. Retrieved from https://www.cdc.gov/healthyyouth/data/yrbs/pdf/2017/2017_yrbs_sadc_documentation.pdf
Erlich M. D.GAP Committee on Psychopathology (2016). Envisioning zero suicide. Psychiatric Services
, 67, 255. doi:
Felitti V. J., Anda R. F., Nordenberg D., Williamson D. F., Spitz A. M., Edwards V., … Marks J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine
, 14, 245–258. doi:
Fond G., Llorca P. M., Boucekine M., Zendjidjian X., Brunel L., Lancon C., … Boyer L. (2016). Disparities in suicide mortality trends between United States of America and 25 European countries: Retrospective analysis of WHO mortality database. Scientific Reports
, 6, 20256. doi:
Ford J. H., Zollinger T. W., Zhang J., O’Neil J., Nelson D. R., & Steele G. K. (2018). Trends in depressed mood and suicidal behaviors among female high school students who engaged in physical fighting. Journal of Interpersonal Violence
, 886260518790599. doi:
Fuller-Thomson E., Baird S. L., Dhrodia R., & Brennenstuhl S. (2016). The association between adverse childhood experiences (ACEs) and suicide attempts in a population-based study. Child: Care, Health and Development
, 42, 725–734. doi:
Han B., Compton W. M., Blanco C., Colpe L., Huang L., & McKeon R. (2018). National trends in the prevalence of suicidal ideation and behavior among young adults and receipt of mental health care among suicidal young adults. Journal of the American Academy of Child & Adolescent Psychiatry
, 57, 20–27. doi:
Hill R. M., Oosterhoff B., & Kaplow J. B. (2017). Prospective identification of adolescent
suicide ideation using classification tree analysis: Models for community-based screening. Journal of Consulting and Clinical Psychology
, 85, 702–711. doi:
Holt M. K., Vivolo-Kantor A. M., Polanin J. R., Holland K. M., Degue S., Matjasko J. L., … Reid G. (2015). Bullying and suicidal ideation and behaviors: A meta-analysis. Pediatrics
, 135, e496–e509. doi:
Jiang Y., Perry D. K., & Hesser J. E. (2010). Suicide patterns and association with predictors among Rhode Island public high school students: A latent class analysis. American Journal of Public Health
, 100, 1701–1707. doi:
Kann L., McManus T., Harris W. A., Shanklin S. L., Flint K. H., Hawkins J., … Zaza S. (2016). Youth risk behavior surveillance–United States, 2015. MMWR Surveillance Summaries
, 65, 1–174. Retrieved from https://www.cdc.gov/mmwr/volumes/65/ss/ss6506a1.htm
King C. A., Horwitz A., Czyz E., & Lindsay R. (2017). Suicide risk screening in healthcare settings: Identifying males and females at risk. Journal of Clinical Psychology in Medical Settings
, 24, 8–20. doi:
Kokkevi A., Rotsika V., Arapaki A., & Richardson C. (2012). Adolescents’ self-reported suicide attempts, self-harm thoughts and their correlates across 17 European countries. Journal of Child Psychology and Psychiatry
, 53, 381–389. doi:
Kölves K., & De Leo D. (2016). Adolescent
suicide rates between 1990 and 2009: Analysis of age group 15–19 years worldwide. Journal of Adolescent Health
, 58, 69–77. doi:
Labouliere C. D., Vasan P., Kramer A., Brown G., Green K., Rahman M., … Stanley B. (2018). “Zero suicide”—A model for reducing suicide in United States behavioral healthcare. Suicidologi
, 23, 22–30. doi:
Li Y., Cannon L., Coolidge E., Darling-Fisher C., Pardee M., & Kuzma E. (2019). Current state of trauma-informed education in the health sciences: Lessons for nursing. Journal of Nursing Education
, 58, 93–101. doi:
Litteken C., & Sale E. (2018). Long-term effectiveness of the question, persuade, refer (QPR) suicide prevention gatekeeper training program: Lessons from Missouri. Community Mental Health Journal
, 54, 282–292. doi:
Lowry R., Crosby A. E., Brener N. D., & Kann L. (2014). Suicidal thoughts and attempts among U.S. high school students: Trends and associated health-risk behaviors, 1991–2011. Journal of Adolescent Health
, 54, 100–108. doi:
Lumley T. (2017). Survey: Analysis of complex survey samples. R package version 3.32( Version 3.32).
McKinnon B., Gariépy G., Sentenac M., & Elgar F. J. (2016). Adolescent
suicidal behaviours in 32 low- and middle-income countries. Bulletin of the World Health Organization
, 94, 340–350. doi:
McLoughlin A. B., Gould M. S., & Malone K. M. (2015). Global trends in teenage suicide: 2003–2014. QJM: An International Journal of Medicine
, 108, 765–780. doi:
Messias E., Kindrick K., & Castro J. (2014). School bullying, cyberbullying, or both: Correlates of teen suicidality in the 2011 CDC Youth Risk Behavior Survey. Comprehensive Psychiatry
, 55, 1063–1068. doi:
Moher D., Hopewell S., Schulz K. F., Montori V., Gøtzsche P. C., Devereaux P. J., … Altman D. G. (2010). CONSORT 2010 explanation and elaboration: Updated guidelines for reporting parallel group randomised trials. BMJ
, 340, c869. doi:
Mojtabai R., Olfson M., & Han B. (2016). National trends in the prevalence and treatment of depression in adolescents and young adults. Pediatrics
, 138, e20161878. doi:
National Institutes of Health, Substance Abuse and Mental Health Services Administration, National Center for Health Statistics (2012). Mental health and mental disorders. In Healthy people 2010 final review
(pp. 3–18). Retrieved from https://www.cdc.gov/nchs/data/hpdata2010/hp2010_final_review.pdf
O’Reilly M., Kiyimba N., & Karim K. (2016). “This is a question we have to ask everyone”: Asking young people about self-harm and suicide. Journal of Psychiatric and Mental Health Nursing
, 23, 479–488. doi:
Patalay P., & Gage S. H. (2019). Changes in millennial adolescent
mental health and health-related behaviours over 10 years: A population cohort comparison study. International Journal of Epidemiology
, 48, 1650–1664. doi:
Pontes N. M. H., Ayres C. G., Lewandowski C., & Pontes M. C. F. (2018a). Trends in bullying victimization by gender among U.S. high school students. Journal of Research in Nursing
, 41, 243–251. doi:
Pontes N. M. H., Ayres C. G., & Pontes M. C. F. (2018b). Additive interactions between gender and bullying victimization on depressive symptoms and suicidality: Youth Risk Behavior Survey 2011–2015. Nursing Research
, 67, 430–438. doi:
R Core Team (2018). R: A language and environment for statistical computing
(Version 3.5.0). Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.R-project.org
Rothman K. J. (2014). Six persistent research misconceptions. Journal of General Internal Medicine
, 29, 1060–1064. doi:
Saxena S., & Setoya Y. (2014). World Health Organization’s comprehensive mental health action plan 2013–2020. Psychiatry and Clinical Neurosciences
, 68, 585–586. doi:
U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. (2010). Healthy people 2020
. Retrieved from https://www.healthypeople.gov/
van Geel M., Vedder P., & Tanilon J. (2014). Relationship between peer victimization, cyberbullying, and suicide in children and adolescents: A meta-analysis. JAMA Pediatrics
, 168, 435–442. doi:
Vandenbroucke J. P., von Elm E., Altman D. G., Gøtzsche P. C., Mulrow C. D., & Pocock S. J.STROBE Initiative (2014). Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): Explanation and elaboration. PLoS Medicine
, 4, e297. doi:
Wang Z., Yu C., Wang J., Bao J., Gao X., & Xiang H. (2016). Age-period-cohort analysis of suicide mortality by gender among White and Black Americans, 1983–2012. International Journal for Equity in Health
, 15, 107. doi:
World Health Organization (2017). Depression and other common mental disorders: Global health estimates
. Retrieved from http://apps.who.int/iris/bitstream/10665/254610/1/WHO-MSD-MER-2017.2-eng.pdf