Introduction

Suicide-related behaviors among adolescents are a serious health concern in the United States. Suicide becomes increasingly prevalent with adolescent development: with rates increasing from .61 per 1000,000 for 12 year olds to 9.24 per 100,000 for 18 year old adolescents (CDC-Centers for Disease Control and Prevention 2011). The Centers for Disease Control and Prevention (CDC) report suicide is the third leading cause of death among adolescents aged 15–24 in the United States, accounting for 12.2 % of all deaths annually (CDC-Centers for Disease Control and prevention 2010). However, the true number of deaths from suicide may be higher owing to suicide being misclassified or masked by other mortality diagnoses (Committee on Adolescents 2000; Wasserman et al. 2005). Concomitant to an attempted or completed suicide is suicide ideation, which includes any thoughts of harming or killing oneself (Bridge et al. 2006; Hill et al. 2011). In 2009, 13.8 % of U.S. high school students in grades 9–12 admitted to having serious thoughts of suicide and 6.3 % reported making at least one suicide attempt in the previous 12 months (CDC-Centers for Disease Control and Prevention 2010).

During adolescence, emotions are heightened and marked mood swings are common (Harrington 2001). Immaturity of affective, cognitive and behavior processes may lead an adolescent to act impulsively, be emotionally reactive and ignore the negative consequences of their behavior (Dawes et al. 2008). The ability to cope with these negative emotions is an important determinant in the contemplation and completion of a suicide attempt (Dawes et al. 2008; Dougherty et al. 2004; Shain 2007).

Adolescent suicide is often the endpoint of chronic and multifaceted problems (Harrington 2001). In addition to a previous suicide attempt being one of the most effective predictors for a subsequent suicide attempt and suicide completion (Bridge et al. 2006; Committee on Adolescents 2000; Shaffer and Craft 1999), three risk factors, often in combination, have been theorized to trigger a suicide attempt in young people (Harrington 2001). The first is drug or alcohol use. Those who have attempted suicide are more likely to have substance abuse disorders than those who have experienced suicide ideation, suggesting that substance use may facilitate the transition from ideation to behavior (Gould et al. 1998). A better understanding of emotional regulation during this transition may help reduce the prevalence of suicide in adolescence. For example, there is evidence that misuse of substances impairs the part of the brain associated with judgment and impulse control (Dawes et al. 2008) and underlying psychopathological conditions may magnify this effect. Thus, it is not surprising that those suffering from mental illness or psychological distress often self-medicate with alcohol or other drugs as an inappropriate coping mechanism (Anhalt and Klein 1976; Ford and Schroeder 2009; Weiss et al. 1992).

The second risk factor is an opportunity for suicide, and more specifically access to a firearm or another means for suicide. Harrington states that the method young people use to kill themselves varies according to where they live, suggesting that it is in part determined by availability (Harrington 2001). Firearms are the most common suicide method in the United States (Bridge et al. 2010; Centers for Disease Control and Prevention 2010) and empirical studies have shown that having a gun in the home is a risk factor for completed suicide in adolescents (Grossman et al. 2005; Bridge et al. 2010). Suicide by way of firearm has a high likelihood of fatality; therefore, an emotionally unstable adolescent with access to a lethal agent such as a firearm is more likely to initiate and complete a suicide attempt (Brent et al. 1993).

The third risk factor is an acute event such as a break up with a romantic partner, disciplinary crisis, or death of a family member which results in stress or depression. Stressful life events are a risk factor for the onset of substance use (van den Bree and Pickworth 2005) and hence an indirect pathway for impulsive behavior. Some adolescents have reported that they engage in self-harm and substance use to “stop bad feelings” or “to get relief” and it has been postulated that suicide attempts triggered by stressful life events are an attempt to relieve or reduce stress (Light et al. 2003).

Stressful and taxing situations generally elicit emotional arousal, a constituent source of information that can affect perceived self-efficacy in coping with threatening situations (Bandura 1977). Self-efficacy is defined as the belief about personal ability to perform behaviors that bring desired outcomes and cross sectional studies have reported a negative relationship between self-efficacy and depression (Muris 2001). Consequently, stress, when combined with depressive symptoms may reach a level where a suicide attempt is viewed as a means to cope with perceived insurmountable difficulties and emotional arousal (Spirito and Esposito-Smythers 2006).

Depression has been linked to a low sense of emotional self-efficacy, an aspect of self-efficacy that pertains to the perceived capability of coping with negative emotions (Muris 2002). Specifically, emotional self-efficacy beliefs are operationalized as an individual’s ability to avoid negative emotional states (e.g., preventing nervousness, suppressing negative thoughts) or restore a normal emotional state when experiencing a negative emotional state (e.g., self-talk to regain a positive attitude, calming yourself once scared or anxious). Those who cease their coping efforts prematurely will retain self-debilitating expectations and fears for longer durations (Bandura 1977) which may place these individuals at an increased risk for suicide ideation or worse, especially when suicidal behaviors may be considered as a means of escape when other mechanisms fail.

Understanding emotional self-efficacy, a construct of emotional regulation—the ability to regulate affective responses in regard to specific environmental demands (Suveg and Zeman 2004) may bring the public and mental health field one step closer to understanding and treating the precursors to depression and suicidality. Depression and negative emotions may, for adolescents, lead to substance abuse, emotional crisis and ultimately suicide. In fact, approximately 90 % those who die from suicide suffer from depression or other mental disorders, a substance-abuse disorder, or combination thereof (NIMH 2010; Moscicki 2001). Moreover, research examining illicit drug use suggests that individuals use the physiological and psychological effects of alcohol or other drugs to regulate distressful feelings and/or to achieve emotional stability (NIMH 2010; Khantzian 1997). In turn, if sustained, emotional self-efficacy can help adolescents overcome risk factors that may ignite the transition from emotional crisis to suicidality. However, models must account for the clustering of substance use that often accompanies suicide ideation, as well as attempts to ascertain the potential protective effects emotional self-efficacy may offer. In addition to established risk factors such as depression, substance use, firearm opportunity, hopelessness, lower life satisfaction, and past suicide attempts (Brown et al. 2000; Harrington 2001; McKeown et al. 1998; Valois et al. 2004; Garrison et al. 1993) problems with controlling emotions, emotional confidence or emotional self-efficacy may also increase the risk for suicidality (Wagner and Zimmermann 2006).

Associations found between low emotional self-efficacy and adolescent suicide ideation and attempts may be explained in part by the small body of previous research involving suicide ideation/attempts and other forms of positive emotion. In a study of 5,557 secondary school adolescents (Level 1–Level 4) from 42 secondary schools in Hong Kong, Kwok and Shek (2010), using Goleman’s (1996) theory of emotional competence (emotional self-efficacy) found a correlation between suicide ideation and low emotional competence. This study also found that adolescent suicide ideation increased when there were lower scores in self-management of emotions, social skills and utilization of emotions. Kowk and Shek (2010) also found no significant differences between male and female study participants for the relationship between suicide ideation and low emotional competence.

In 2011, Dour and colleagues conducted a laboratory-based study on self-injurious behavior with a sample of 87 adolescents in Massachusetts. These researchers determined that emotion reactivity was strongly associated with the probability of a suicide attempt among adolescents with poor problem solving skills and moderately associated among those with moderate problem-solving skills (Dour et al. 2011). These researchers concluded that interventions aimed at enhancing problem-solving and emotion regulation skills can decrease the risk of suicidal behavior.

Tahmassian and Moghadam (2011) conducted a study with 549 high school adolescents looking at the relationship between self-efficacy and symptoms of anxiety, depression and worry in a sample of normal students in Iran. This study involved adolescents from high school and pre-university centers. These researchers found significant and negative relationships between emotional self-efficacy and anxiety. In particular, low emotional self-efficacy was significantly associated with high levels of self-reported anxiety.

In a study of 78 older adolescents (college students) in the UK who reported some degree of suicidality, Johnson et al. (2010) found that positive self-appraisals moderated the association between stressful life events and suicidality. For those participants reporting moderate or high levels of positive self-appraisals, raised incidence of stressful life events did not lead to increases in suicidality. These results suggest that positive self-appraisals may confer resilience to suicide ideation and suicide attempts (Johnson et al. 2010).

Tamas et al. (2007) conducted a study using a sample of 407 children/adolescents with depressive disorder recruited from clinical sites across Hungary. These researchers found that high maladaptive and low adaptive emotion regulation response tendencies increased the odds of suicidal behaviors, above and beyond the risk posed by depressive illness severity.

In a study of 596 normal adolescents from a secondary school in Belgium, Muris (2002) found that emotional self-efficacy was strongly linked to anxiety and depression, recognized precursors to suicide ideation and attempt. More specifically, Muris (2002) found that a low sense of emotional self-efficacy was accompanied by high levels of anxiety and depression in a sample of normal adolescents. This study also found that females exhibited lower levels of emotional self-efficacy, but higher levels of trait anxiety/neuroticism, anxiety disorders symptoms and depressive symptoms than did male adolescents.

It appears reasonable to conclude that research exploring specifically the relationship between suicide ideation and suicide attempts and emotional self-efficacy for adolescents is relatively sparse (Kwok and Shek 2010; Dour et al. 2011; Johnson et al. 2010; Tamas et al. 2007) and inconclussive. Emotional self-efficacy as defined by Muris (2001) is the perceived capability of coping with negative emotions for adolescents. Anxiety and negative emotions for adolescents may lead to decreased scholastic performance, social maladjustment, depressed mood and possibly an increased risk for depression, and other emotional health challenges such as suicide (Muris 2001) and health risk behaviors such as substance abuse and sexual risk behaviors (Cooper et al. 1998; Farrell and Danish 1993; Hessler and Fainsibler-Katz 2010).

Therefore, the purpose of this study was to investigate potential relationships between self-reported suicide ideation and suicide attempts and emotional self-efficacy among adolescents in a large cross-sectional, statewide sample. Specifically, we hypothesized that low levels of emotional self-efficacy would be significantly associated with higher levels of suicide ideation and suicide attempt for study participants and that these associations would vary by race and sex. Associations were investigated separately by four race/sex groups (black males, black females, white males, and white females), given the demographic differences in adolescent suicide behaviors. From a national USA perspective, adolescent female suicide ideation and suicide attempts are more prevalent than male and white suicide ideation and suicide attempts are more prevalent for blacks (CDC-MMWR 2011). Previous studies in the USA have determined significant race/sex differences in adolescent emotional self-efficacy (Valois et al. 2008, 2013) emotional self-efficacy associations with sedentary behavior (Valois et al. 2008) and emotional self-efficacy associations with sexual risk-taking behavior for high school adolescents (Valois et al. 2013).

Method

The Youth Risk Behavior Survey (YRBS) used a sampling procedure designed to obtain a representative sample (with the exception of special education students) of all public high school students’ grades 9 through 12 in South Carolina. The questionnaire was previously determined to have adequate reliability (Brener et al. 2002). YRBS methodology has been utilized extensively, and these methods have been explained in detail in previous publications (see Brenner et al. 2002, 1995). For this study, 215 schools were stratified by enrollment size into 3 categories: schools with 74–874 students were classified as small; enrollments from 875 to 1,278 were medium; and enrollments greater than 1,278 were large. Our data are reported on 39 of the 68 eligible schools (57 % school response rate). The student response rate was 89 %. The overall response rate (.57 × .89) was 51 %. This overall response rate did not meet CDC’s Division of Adolescent and School Health criteria for weighted data (60 %); therefore, these data were not weighted, but treated as a simple random sample. This research was approved by the referent university’s review board for the rights of human participants in research.

Sample

The total number who participated in the YRBS for this southern state was 3,836 with usable data from 3,376 students. However, only 2,566 valid observations were available owing to: (a) non-response by subjects to variables of interest; (b) responses that could not be read; (c) out-of-range responses (n = 549, 16.3 %); and (d) exclusion of subjects reporting their race as “Other” than black or white (n = 261, 7.7 %). Sex was split at 41 % female and 59 % male. Ethnicity was split between African Americans at 54 % and Caucasians at 46 %. The final sample contained 1,037 (40.4 %) females and 1,529 males (59.6 %) of whom 579 (22.6 %) were black females, 809 (31.5 %) were black males, 458 (17.8 %) were white females, and 720 (28.1 %) were white males. Student participants were in grades 9–12 with ages ranging from 12 to 18 years. There were 833 (32.5 %) 9th grade students, 741 (28.9 %) 10th grade students, 518 (20.2 %) 11th grade students, and 474 (18.5 %) 12th grade students in the final sample for analysis.

Instrumentation

Emotional Self-Efficacy

The Emotional Self-Efficacy (ESE) Scale included in the YRBS was developed by Muris (2001) and validated with Dutch adolescents through exploratory and confirmatory factor analytic procedures. In a recent study (Valois and Zullig 2013) this scale was modified slightly and determined to be valid and reliable with high school adolescents from the USA. The ESE scale (Valois and Zullig 2013) included seven items. The 7 item scale included the following items: “How well do you succeed at cheering yourself up when an unpleasant event has happened”; “How well do you succeed in becoming calm again when you are very scared”; “How well can you prevent becoming nervous”; “How well can you control your feelings”; “How well can you give yourself a pep-talk when you feel low”; “How well do you succeed in suppressing unpleasant thoughts” and “How well do you succeed in not worrying about things that might happen?” Each item was scored on a 5-point Likert-type scale with the following options: (a) not at all, (b) a little bit, (c) pretty well, (d) well and (e) very well. Valois and Zullig (2013) report a Cronbach’s alpha of .85 for initial development of the ESE scale for use with adolescents in the USA providing an estimate of internal consistency.

Suicide Ideation and Attempts

Four standard YRBS suicide ideation and attempts questions were utilized in this study: “During the past 12 months, did you ever seriously consider attempting suicide?”; “During the past 12 months, did you make a plan about how you would attempt suicide?”; “During the past 12 months, how many times did you actually attempt suicide?”; and “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?” Response options for each item was (a) yes and (b) no with the exception of the fourth question which had response options of (a) I did not attempt suicide during the past 12 months, (b) yes, and (c) no. These response options (a) and (c) were collapsed into a single “no” option for this study to create a fourth dichotomous variable.

Data Analysis

All seven discrete variables from the ESE Scale were pooled to form a pseudo-continuous variable ranging in score from 7 (1 × 7) to 35 (5 × 7), expressed as a mean emotional self-efficacy (MESE) score with lower scores indicative of being less emotionally self-efficacious. A numeric collapse of the pooled dependent variable was performed because adolescents reporting their ESE as “pretty well” were measured with one response option, whereas the other two categories were composed of two response options. As a result, those receiving MESE scores of 15 or less were categorized as having lower emotional self-efficacy, MESE scores between 16 and 21 were deemed as having mid-range emotional self-efficacy, while those with MESE scores of 22 or greater were designated as having high emotional self-efficacy. To account for the three levels of the MESE outcome variable (low, medium [mid-range] and high emotional self-efficacy), unadjusted and adjusted logistic regression analyses were conducted with those reporting high emotional self-efficacy serving as the referent group. Logistic regression was chosen as the appropriate analytical technique in this study because: (1) the ESE scale items are ordinal variables; and (2) because we wanted to be conservative in our estimates.

Analyses were conduced separately for four sex and race groups. Use of dummy variables to represent the four groups would have assumed that the coefficients of all risk (suicide) and confounding variables were constant across all four sex and race groups. Because that assumption was false, the four sex/race groups were analyzed separately controlling for tobacco, alcohol, marijuana, and cocaine use (past 30 days); relationship violence (past 12 months); feelings of depression for 2 or more weeks; socioeconomic status (as measured by eligibility for free or reduced priced lunch); and weapon or gun access (past 30 days) as covariates in the adjusted models. Unadjusted and adjusted odds ratios (OR) and 95 % confidence intervals (CI) were calculated to determine which suicide ideation and behavior variables were associated with reduced ESE for each sex and race group for both the midrange and high ESE levels. Participants who did not report suicide ideation, planning a suicide attempt, a suicide attempt, or being injured by a suicide attempt (past 12 months) and had high ESE served as referent categories. Level of significance was determined a priori at alpha α ≤ .05. Frequency and percent prevalence for the suicide ideation and suicide behavior variables and ESE variables are presented in Table 1.

Table 1 Prevalence of depressed mood, suicide ideation and suicide attempts
Full size table

Results

Prevalence Results

The data in Table 1 on prevalence of the variables of interest suggest some trends that are in line with, and contrary to, national/international trends in adolescent emotional self-efficacy, depressed mood and suicide. For emotional self-efficacy, this study found females (BF 31.6 %; WF 44.1 %) to have lower ESE compared to black males (27.6 %), however white males had a significantly lower (52.9 %) ESE than all other race/sex groups. With the exception of white males, these prevalence rates are similar to previous studies (Kwok and Shek 2010; Muris 2002). In regard to depressed mood (sad every day for 2 weeks in the past 12 months) our results are similar to national trends with females (BF 33.0 %; WF 22.1 %) at a higher percentage than males (BM 31.3 %; WM 21.0 %) (CDC-MMWR 2012). Our results for serious consideration of suicide with black females at 14.9 %, black males at 21.4 %, white females at 10.0 % and white males at 11.5 % are in contrast to consistent national trends (CDC-MMWR 2012) where our male participants report a greater prevalence of suicide ideation than females and our black study participants report a higher prevalence than our white participants and this pattern continues for planned suicide attempts (BF 13.1 %; BM 16.9 %; WF 9.9 %; WM 11.0 %). For suicide attempts, our results for whites are consistent with national trends (CDC-MMWR 2012) with females (8.3 %) at a greater prevalence than white males (5.8 %). However, our results are contrary to national trends for blacks adolescents with black males (11.7 %) at a greater prevalence than black females (9.8 %). This pattern continues for suicide attempts resulting in injury where our results for whites are consistent with national trends (CDC-MMWR 2012) with females (4.2 %) at a greater prevalence than white males (2.1 %). However, again, our results for blacks are contrary to national trends with black males (3.5 %) at a greater prevalence than black females (2.2 %).

Relationship Between Suicide Ideation and Attempts and Midrange ESE

No significant associations were detected between suicide ideation and behavior and midrange ESE for any of the race and sex groups. Owing to the lack of association, these variables are neither tabled nor discussed.

Relationship Between Suicide Ideation and Attempts and Low ESE

Black Females

Significant relationships were established between low ESE and all four suicide ideation and suicide attempts variables for black females in both the unadjusted and adjusted analyses (Table 2). In the adjusted analyses, black females who reported considering suicide, planned a suicide attempt, attempted suicide, or had an injury from a suicide attempt significantly increased the odds of reporting low ESE ranging from 2.01 (considered suicide) to 2.44 times (attempted suicide) in comparison to black female adolescents who reported high ESE and no suicide ideation or attempts (p < .05, Table 2).

Table 2 Unadjusted and adjusted odds ratios and 95 % confidence intervals for association between low emotional self-efficacy and suicidal risk behaviors by race and gender
Full size table

Black Males

A significant association was established between low ESE for all four suicide ideation and attempts variables for black males in the unadjusted analyses (Table 2). However, only the variable of “seriously considered suicide” (past 2 weeks) remained significant in the adjusted analyses (OR 1.56; CI 1.01–2.39, p < .05) for black males. Thus, among black males who reported having seriously considered suicide (past 2 weeks), the odds of reporting low ESE increased by approximately 1.6 times in comparison to black males reporting high ESE and not seriously considering suicide (past 2 weeks).

White Females

Significant associations were established between low ESE for three of four suicide ideation and suicide attempt variables for white females in the unadjusted analyses (Table 2). However, after adjustment for the covariates, no significant relationships remained.

White Males

Significant relationships were established between low ESE and all four suicide ideation and suicide attempts variables for white males in the unadjusted analyses and for three of four variables in the adjusted analyses (Table 2). In the adjusted analyses, white males who reported considering suicide, attempted suicide, or had an injury from a suicide attempt significantly increased the odds of reporting low ESE ranging from 3.15 (injured from a suicide attempt) to 3.95 times (attempted suicide) in comparison to white males who reported high ESE and no suicide ideation or attempts (p < .05, Table 2).

Discussion

Results suggest that a substantial number of public high school adolescents are engaging in suicide ideation and attempts (Table 1). This study also suggests that a substantial number of public high school students are reporting low or only modest ESE (Table 1). Most importantly, this study demonstrated a meaningful association between two distinct adolescent health research literatures: health risk behavior (suicide ideation and suicide attempts) literature and emotional self-efficacy literature.

Our findings suggest that low emotional self-efficacy is associated with serious consideration of suicide for black (females and males) and especially for white male adolescents in this study. Results also suggest that planning a suicide attempt was associated with low emotional self-efficacy for black females. Attempting suicide and suicide attempts resulting in injury were associated with low ESE for black females and strongly associated with white males in this large state wide sample of adolescents.

Results of this study are also in line with the Broaden-and-Build Theory of Positive Emotions (Fredrickson 2001). The Broaden-and-Build Theory posits that experiences of positive emotions broaden people’s momentary thought-action repertories, which in turn serves to build their enduring personal resources, ranging from physical and intellectual resources to social and psychological resources (Fredrickson 2001; Fredrickson and Joiner 2002; Tugade et al. 2004). From this study, high emotional self-efficacy would be considered a social and psychological resource/asset for adolescents as they cope with the adolescent developmental process in an ever-evolving intellectual, psychological and social environment.

The present study examined associations between emotional self-efficacy (the capacity of adolescents to cope with and regulate their own emotional reactions) and adolescent suicide ideation (past 2 weeks) and suicide attempts (past 12 months). Our hypotheses (that lower emotional self-efficacy would be related to increased suicide ideation and attempts) were largely supported for all race/sex groups with each of the four suicide variables with the exception of white females in the adjusted analyses. Findings suggest that low emotional self-efficacy formed during adolescence may be associated with suicide ideation and suicide attempts, especially for black females, white males and to a lesser extent, black male adolescents.

Our findings that compare suicide and ESE results by sex could be the result of prevalence differences in our sample (females = 1,039 vs. males 1,529) helping to explain why white males (n = 720) had significant associations with ESE and suicide ideation, suicide attempts and suicide attempts that resulted in injury compared to females in the sample and to a lesser extent, why black males (n = 809) had a significant association for suicide. White females are lowest in prevalence in our sample (n = 458) however they could be the subset that is more comfortable with recognizing and processing their emotions related to suicide compared to the other race/sex groups in this study. White females in this study could be from more socio-economic advantaged families and communities established factors that help combat emotional distress, depression and anxiety (Yoshikawa et al. 2012). These differences in suicide prevalence could also be the result of other risk behavior prevalence and the clustering of health risk behaviors (Hessler and Fainsilber-Katz 2010) that influence emotional self-efficacy and mental health.

Adolescent suicide has typically been perceived as an event that affects females more than males in the U.S. (Joe et al. 2006; CDC-MMWR 2012). Our results for ESE associated with suicide consideration, attempts and attempts resulting in injury for white males (Odds Ratios 3.35; 3.95; 3.15 respectively) is somewhat perplexing vis-à-vis national trends. Our findings (Table 1) for adolescents reporting that they felt sad almost every day for 2 or more weeks in a row so that they stopped doing some usual activities (past 12 months) with males reporting lower prevalence rates than females is consistent with national trends (CDC-MMWR 2012). However, the prevalence data from this study shows that white males report both the highest frequency of low ESE (n = 381) and the highest percent (52.9) of low ESE compared to all other race/sex groups and levels of reported ESE. South Carolina, similar to other states, has a significant rural population. Rural areas are often resource challenged in regard to economic opportunity, education, recreation and mental health services. Socio-economic disadvantage is an important dynamic associated with emotional distress, anxiety, and depression (Yoshikawa et al. 2012). Suicide intervention should consider these findings for tailoring and cultural competency for white male adolescents.

Although typically perceived as affecting whites more than any other group in the USA, recently the rate of suicide among black adolescents has been increasing (Joe et al. 2006; Balis and Postolache 2008; CDC 2010, 2011, 2012). Contextual hardships experienced by many black youth can precipitate suicidality (Osiezagha et al. 2009) or interact with psychological challenges historically associated with adolescent suicidality (Sani et al. 2011). In addition to differences in etiology, suicidality may manifest differently in black youth (Goldston et al. 2008). Related to cultural mores condemning suicide within the African American community (Bergmooser 1994), black youth may engage in activities and behaviors that can be self-harming, as opposed to direct suicide. Gibbs (1988) suggests that some blacks may put themselves in harm’s way as a noble escape from intolerable environments. Putting oneself in harm’s way may be considered as indirect suicide (Langhinrichsen-Rohling et al. 2009), and it has been posited that suicidality and interpersonal violence among black youth represent a continuum of related self-destructive behaviors (Chu et al. 2010; Gibbs 1988; Hillbrand 2001).

In 1999, the National Comorbidity Survey found that black adolescents compared with whites, were somewhat less at risk for lifetime suicide attempts (Kessler et al. 1999). Data from the CDC Youth Risk Behavior Survey (1987) and an ethnocultural differences study of adolescent suicide by Roberts et al. (1997) found similar results. Research suggests that black females have specific risk factors for suicide to include: “psychological distress,” substance abuse, post-traumatic stress disorder, relationship difficulties, poor social support, childhood abuse and abuse by her (intimate/sexual) partner (Kaslow et al. 2000). A study by Groves and colleagues (Groves et al. 2007) found depression to be a major risk factor for black females; and, depression is associated with an even greater risk of suicide compared to males in this study. Preliminary findings from the Pittsburg Girls’ Study indicate that black adolescent females are suicidal at lower thresholds of depression, relative to their white counterparts (Hipwell and Stepp 2011; Keenan et al. 2008), yet our study controlled for depressive mood and findings continued to suggest elevated risk for African American females.

However, black males appear to commit suicide more frequently than females (Willis et al. 2002; Thompson et al. 2002). Willis et al. (2002) have theorized that postmodernity loosens the bonds between the individual and society for black males, thereby increasing the vulnerability to depression, related pathologies (such as substance abuse) and suicide. Researchers suggest that young black males are more exposed to such stresses and the usual social institutions that provide support in the black culture have been unable to be as effective in maintaining that support currently (Willis et al. 2002; Thompson et al. 2002).

Three potentially important pathways should also be noted in understanding why low emotional self-efficacy appeared to be associated with suicide ideation and suicide attempts in this study. First, when individuals are placed in situations where they are required to meet highly valued standards (e.g., ability to manage emotional challenges with peers, parents and teachers), a low sense of affect regulation may lead to anticipatory apprehension, thereby lowering emotional self-efficacy. This is especially important if an individual’s personal performance standards are set much higher than their perceived efficacy to attain them. Second, a low sense of affect regulation may obstruct the benefits that emotional confidence and controlling emotions offers adolescents in terms of forming positive relationships with others that enable individuals to manage stressful situations. Third, low emotional self-efficacy may inhibit youth in a variety of ways and may lead to increased anxiety and depression (Muris 2002) and increased health risk behaviors (Hessler and Fainsilber-Katz 2010).

This study had limitations. These limitations include: use of a cross-sectional study where no temporal sequence of reduced emotional self-efficacy as a precursor for suicide ideation or suicide attempts can be determined; use of a geographical sample that may represent youth only in this southern state; and elimination of participants with missing data on variables of interest for logistic regression analysis. Such elimination can introduce biases; however, this bias is not seriously threatening. This study did not examine change processes for emotional self-efficacy. It would be useful to test the effectiveness of interventions in producing changes in emotional self-efficacy. In turn, it would be imperative to examine any positive changes in emotional self-efficacy by decreased rates of suicide ideation and suicide attempts for adolescents. Studies of this nature would provide additional perspective on emotional self-efficacy as a contribution to interventions to decrease adolescent suicide.

Study strengths include the large final sample (n = 2,566 valid observations) the random sampling technique, stratification of schools by size, student response rate of 89 %, reliability and construct validity of the emotional self-efficacy measure, the examination of both suicide ideation and suicide attempts in the data analysis, and analysis of results by race and sex. In addition, dichotomized variables and associations established between suicide ideation, suicide attempts and emotional self-efficacy could be even stronger and more conservative (i.e., an underestimation of the magnitude of risk and more toward the null hypothesis (of no association) than reported.

Conclusion

Results from this study have implications for suicide prevention and risk reduction interventions. In regard to interventions for increasing emotional self-efficacy for reductions in adolescent suicide ideation and attempts, experts suggest skills that are necessary for navigating the challenges of the immediate social environment (Saarni et al. 1998; Saarni 1999; Halberstadt et al. 2001; Buckley et al. 2003). A basic set of emotional self-efficacy skills for adolescents could include: Developing an awareness of one’s emotional state; Skills for discerning emotions in others; Skills in using the language and vehicular of emotion and expression in a culturally competent manner; Developing the capacity for empathy and sympathy involving the emotions of others; Skills in recognizing that inner emotional state does not need to translate to outer emotional expression; Developing a capacity for managing stress; and Developing the ability to recognize that the formation of relationships, to some degree, is determined by the level of emotional genuineness of expression and the degree of balance within a relationship (Saarni et al. 1998; Saarni 1999; Halberstadt et al. 2001; Buckley et al. 2003). These skills and strategies are adaptive and assist the adolescent to: reach goals; cope with life challenges; manage emotional arousal for effective problem solving; discern what others feel and to respond sympathetically as the situation is presented; and recognize how emotion communication and self-presentation affect relationships (Buckley et al. 2003). More importantly, in regard to effective learning for improved emotional self-efficacy for suicide risk reduction, is skill development to the degree where an adolescent can begin to trust their ability to reach their goals when faced with emotion-laden interactions with others. In turn, increasing self-efficacy, in particular emotional self-efficacy, appears to be an important component for interventions designed to prevent adolescent engagement in risky emotion driven behaviors (DiClemente et al. 2008; Hennessy et al. 2013; Romer et al. 2009; Hessler and Fainsilber-Katz 2010) often associated with the risk factors for adolescent suicide (Balis and Postolache 2008; van den Bree and Pickworth 2005; Harrington 2001; Borowsky et al. 2001).

Poverty and socio-economic disadvantage are among the most salient factors found to precipitate general emotional distress, anxiety, and depression (Jensen 2009; Yoshikawa et al. 2012), each of which may be linked to suicidality (Murry et al. 2011). These findings suggest that risk factors other than psychiatric disorder are operative for black (male and female) and white male adolescents and, as such, preventive interventions must consider youth’s social ecology and their cognitive and behavioral management of their environments in order to be optimally effective and culturally and contextually relevant for adolescents. Future research and practice needs to address mechanisms that might elucidate sex and ethnic/racial differences in relationships between emotional self-efficacy and suicide.

Adolescent suicide risk reduction intervention research should also consider enhancing emotional self-efficacy as a program component for adolescents (Buckley et al. 2003; Hessler and Fainsibler-Katz 2010; Valois and Zullig 2013). Research and program evaluation should include sound emotional self-efficacy measurement instrumentation for documenting program effectiveness (Valois and Zullig 2013; Muris 2001).