! Alcohol-Related Problems as Specific Correlates of Interpersonal-Psychological Theory of Suicide Constructs by Jami Michele Gauthier A thesis submited to the Graduate Faculty of Auburn University in partial fulfilment of the requirements for the Degree of Master of Science Auburn, Alabama August 2, 2014 Approved by Tracy K. Wite, Chair, Asistant Profesor of Psychology Christopher J. Correia, Profesor of Psychology Frank W. Weathers, Profesor of Psychology ! ii! Abstract The relationship betwen alcohol misuse and suicide ideation among undergraduate students has recently been investigated through the potential mechanisms of coping motives for drinking (Gonzalez, Bradiza, & Collins, 2009), and alcohol-related problems (Lamis & Malone, 2011). In the later, the interpersonal constructs (i.e., perceived burdensomenes and thwarted belongingnes) of the Interpersonal-Psychological Theory of Suicide (IPTS) were found to partialy mediate the relationship betwen a uni-dimensional measure of alcohol-related problems and suicide pronenes. In the current study we atempted to replicate and extend upon this previous work, by utilizing both uni and multi-dimensional measures of alcohol-related problems, including the constructs of the IPTS asociated with acquired capablity for suicide, using a more precise measure of suicide ideation as our outcome variable, and controlling for the efects of coping motives for drinking and sadnes. In a sample of regularly drinking college students (N=295), our results suggested that the relationships betwen alcohol-related problems and the IPTS constructs are not specific to the subtype of alcohol-related problems experienced, and that these relationships are beter acounted for by recent experiences of sadnes. Implications of these findings for the IPTS, as wel as suggestions for future research on this topic are discussed. ! iii! Table of Contents Abstract...?.?????????????????????????????????ii List of Tables??.???????...????????????????..??????iv List of Figures???????????????????...??????????...??v Introduction...?..???????????????????????????????.1 Method..?..??...???????????????????..?????????......14 Participants and Procedure...?.?..???????????????..????.....14 Measures..???????????????????..?????????.?...16 Results..??????..?????????..??????????????????..23 Preliminary Analyses...???????????????????????..??23 Hypothesis 1: Replication and Extension of Lamis and Malone (2011) Model???....24 Hypothesis 2: Determination of Diferential Asociations...??????????.....26 Hypothesis 3: A more Stringent Test of the Lamis and Malone (2011) findings...?.......27 Additional Analyses: A more Parsimonious Path Model...?????????.?..?29 Discussion...??????????????????????????..???.???30 References..?????????????????????????????.....??...37 Appendix A: Tables..?????????????????????????..???...49 Appendix B: Figures..???????????????????????????..?..58 ! iv! List of Tables Table A1: Descriptive statistics for al variables for regular drinking sample?.............??..50 Table A2: Zero-order correlations betwen study variables in regular drinking sample??...51 Table A3: Standardized and unstandardized parameter estimates for path model 1????..52 Table A4: Indirect efects in path model 1???????????????????.....53 Table A5: Standardized and unstandardized parameter estimates for path model 2????..54 Table A6: Indirect efects in path model 2????????????????????.56 Table A7: Standardized and unstandardized parameter estimates for path model 3??..??57 ! v! List of Figures Figure B1: Path analytic model 1. Relationships among alcohol-related problems, interpersonal-psychological theory of suicide constructs, and pasive suicide ideation with gender as a covariate???.?????????????...???????..59 Figure B2: Path analytic model 2. Relationship\s among alcohol-related problems, interpersonal-psychological theory of suicide constructs, and pasive suicide ideation with gender, coping motives for drinking, and sadnes as covariates??.?????..60 Figure B3: Path analytic model 3. Relationships among sadnes, interpersonal-psychological theory of suicide constructs, and pasive suicide ideation with gender as a covariate....61 ! 1! As research on suicidal behavior in al forms (i.e., ideation, atempts, and death) has advanced, a number of distal and proximal risk factors have been identified as key antecedents to the behavior. Whereas distal factors, such as psychiatric ilness, self-harming behaviors, family history of suicide, hopelesnes, impulsivenes, and alcohol and substance abuse (Sher, 2006; Yoshimasu, Kiyohara, & Mityashita, 2008) increase an individual?s lifetime risk for suicidal behavior, proximal risk factors, including interpersonal problems, employment isues, increased psychiatric disturbance, and increased alcohol consumption (Sher, 2006), increase an individual?s imediate risk. Because alcohol misuse functions as both a distal and proximal risk factor, understanding the interplay betwen alcohol misuse and suicidal behavior is an important task for those interested in understanding and preventing suicide. Existing evidence for alcohol misuse as a distal risk factor for suicidal behavior largely comes from research on individuals diagnosed with alcohol use disorders (AUDs). AUDs have a demonstrated relationship with suicide ideation (Borges, Walters, & Kesler, 2000; Kesler, Borges, & Walters, 1999), non-fatal suicide atempts (Kely, Lynch, Donovan, & Clark 2001), and death by suicide (Cavanagh, Carson, Sharpe, & Lawrie, 2003; Wilcox, Conner, & Caine, 2004; Yoshimasu et al., 2008). Taken together, these findings suggest that those with an AUD are at an elevated risk of experiencing al types of suicidal behavior. Nevertheles, the increased risk for suicidal behavior conferred by alcohol misuse is not limited to those with a diagnosable AUD. Although operationalized in a number of ways and often used interchangeably with the term problematic drinking, heavy drinking can be described as a patern of higher alcohol consumption (encompasing a greater frequency, larger quantity, and greater number of binge drinking days) compared to one?s peers (Berkowitz & Perkins, 1986; Ham & Hope, 2003). This patern of alcohol consumption is asociated with suicide ! 2! ideation (Gonzalez, Bradiza, & Collins, 2009; Gonzalez, Collins, & Bradiza, 2009; Reifman & Windle, 1995), non-fatal suicide atempts (Pfaf, Almeida, Wite, Waesche, & Joiner, 2007; Reifman & Windle, 1995), and death by suicide (Klatsky & Armstrong, 1993; Nakaya et al., 2007; Schneider et al., 2011). Although some have suggested that the relationship betwen alcohol misuse and suicidal behavior may be explained through co-occurring depresive symptoms, several studies have demonstrated that such symptoms do not fully acount for the relationship (Borges et al., 2000; Gonzalez, Bradiza, et al., 2009; Gonzalez, Collins, et al., 2009; Reifman & Windle, 1995; Stephenson, Pena-Shaf, & Quirk, 2006). These findings suggest that the link betwen heavy drinking and suicidal behavior is non-spurious and extends beyond diagnosable AUD. Drinking Motives, Alcohol-Related Problems, and Suicidal Behavior Although the majority of the research on the link betwen alcohol misuse and suicidal behavior has been descriptive (Hufford, 2001), those who have looked for possible causal explanations have tended to focus on the relationship betwen suicide ideation and either coping motives for drinking (Gonzalez, Bradiza, et al., 2009; Gonzalez, Collins, et al., 2009) or the negative consequences of alcohol misuse (Duberstein, Conwel, & Caine, 1993; Lamis & Malone, 2011; Light, Grube, Madden, & Gover, 2003; Pirkola, Isometsa, Heikkinen, & Lonnqvist, 2000; Preuss et al., 2002). Drinking motives have been conceptualized as the reasons why people choose to drink in a particular situation and have typicaly been divided into four distinct categories based on both the source of the motivation and the valence of the emotion behind it (Cooper, 1994; Cox & Klinger, 1988). Although al motives have been found to be asociated with alcohol use, coping motives appear to be particularly pernicious. Coping motives have been found to be asociated ! 3! with heavy drinking (e.g., Labouvie & Bates, 2002), alcohol-related problems even when controlling for alcohol consumption (Kasel, Jackson, & Unrod, 2000), and alcohol dependence (Carpenter & Hasin, 1998). More recently, coping motives have been found to be asociated with solitary heavy drinking (Gonzalez, Collins, & Bradiza, 2009), suicide ideation among binge drinkers (Gonzalez & Hewel, 2012), and to statisticaly mediate the relationships betwen suicide ideation and both alcohol consumption and alcohol-related problems (Gonzalez, Bradiza, & Collins, 2009) among participants with a history of pasive suicide ideation (SI). Although such findings have been interpreted as SI leading to drinking for coping motives (i.e., Gonzalez, Bradiza, & Collins, 2009), alternative explanations for these relationships cannot be ruled out. For instance, coping motives may actualy increase negative afect over the long term, since avoidant coping strategies do not actualy resolve problems, which could lead to increased suicide ideation. Additionaly, the asociation may not persist once other related variables (e.g., depresive symptoms, alcohol-related problems) are taken into acount. Furthermore, despite the asociation betwen coping motives for drinking and suicide ideation, we do not yet know if coping motives are also asociated with more severe forms of suicidal behavior. In studies investigating alcohol-related problems as a potential cause of the relationship betwen alcohol use and suicidal behavior, certain alcohol-related problems are positively asociated with suicide ideation and atempts (Duberstein et al., 1993; Light et al., 2002; Preuss et al., 2002), as wel as death by suicide (Pirkola et al., 2000). However, alcohol-related problems are also positively asociated with coping motives for drinking (Cooper, Frone, Russel, & Mudar, 1995; Carey & Correia, 1997; Gonzalez, Bradiza, et al., 2009; Merril & Read, 2010). Although it is clear that there are asociations betwen alcohol-related problems and al forms of suicidal behavior, again the significance of these relationships when ! 4! conceptualy-related variables, such as drinking for coping motives and sadnes, are taken into acount cannot be determined from the acumulated research. Given that drinking motives are typicaly conceptualized as antecedents to drinking that may decrease in salience once drinking habits are established (Kuntsche, Knibbe, Gmel, & Engels, 2005), whereas alcohol-related problems are viewed as negative outcomes of alcohol misuse over both the short and long term (Perkins, 2002), in the current paper we focus more heavily on the ways in which alcohol-related problems may be conceptualy linked to suicidal behavior. At the same time we consider the possibility that coping motives for drinking and/or depresive symptoms may beter explain the relationship betwen suicidal behavior and alcohol misuse. Theoretical Explanations of Suicidal Behavior The Interpersonal-Psychological Theory of Suicide (IPTS; Joiner, 2005; Van Orden et al., 2010) has been proposed as a comprehensive, falsifiable theory purported to explain al forms of suicidal behavior. Acording to this theory, when experienced in isolation thwarted belongingnes (i.e., social disconnection, lack of close others) and perceived burdensomenes (i.e., negative self-view, felings of liability) cause pasive suicide ideation (Joiner, 2005; Van Orden et al., 2010). When experienced in combination with the idea that these states are stable and unchanging (i.e., hopelesnes), pasive suicide ideation wil transform into an active desire for death (Joiner, 2005; Van Orden et al., 2010). These tenets of the IPTS have been supported by preliminary empirical studies demonstrating a statistical interaction betwen perceived burdensomenes and thwarted belongingnes in the prediction of suicide ideation (Van Orden, Wite, Gordon, Bender, & Joiner, 2008; Joiner et al., 2009). In that the IPTS cites the experiences of thwarted belongingnes and perceived burdensomenes as causes of suicide ideation, it is in line with previous explanations of suicidal ! 5! behavior that emphasize social disconnection (Durkheim, 1897) and negative self-view (Baumeister, 1990; Joiner, 2005; Van Orden et al., 2010; Shneidman, 1998). However, the IPTS departs from previous explanations of suicidal behavior by stating that individuals wil not act on their suicidal desire, no mater its intensity, without the presence of a third construct (Joiner, 2005; Van Orden et al., 2010). This third construct is the acquired capability for suicide, and can be broken down into the components of fearlesnes about death and increased pain tolerance (Joiner, 2005; Smith & Cukrowicz, 2010; Van Orden et al., 2010). The rationale for this construct arose in part from the consistent finding that suicide ideation is far more common than non-fatal and fatal suicide atempts (e.g., McIntosh, 2012), indicating that suicide ideation in isolation is not sufficient to produce lethal suicidal behavior. Joiner (2005) states that although no one is born with the ability to take his or her own life, repeated exposure to fear-inducing and painful events can habituate an individual to physical pain and/or fear of death. Joiner (2005) proposes that although one?s initial response to any life- threatening stimulus is intense fear, repeated exposure causes habituation, resulting in the reduction of that fear (Van Orden et al., 2010), with the same proces holding true for painful stimuli. Evidence of how habituation to both the fear and pain asociated with suicidal behavior occurs is most clearly reflected in studies showing that those who have previously atempted suicide report les fear of suicide (Linehan, Goodstein, Neilsen, & Chiles, 1983; Malone et al., 2000) and higher pain tolerance (Orbach, Palgi, et al., 1996; Orbach, Stein, et al., 1996) than individuals who have not previously atempted. Although a history of non-fatal suicide atempts is proposed as the most potent way to acquire the capability for suicide, Joiner (2005) explains that exposure to other painful and life- threatening experiences can have a similar efect (Van Orden et al., 2010). In support of this ! 6! notion, researchers have found that exposure to various painful and provocative events (e.g., shooting a gun, physical fighting; Van Orden et al., 2008), combat exposure (e.g., Bryan, Cukrowicz, West, & Morrow, 2010), and engaging in impulsive behaviors (e.g., Bender, Gordon, Bresin, & Joiner, 2011), are asociated with heightened acquired capability for suicide. These findings suggest that habituation to the fear and pain asociated with suicide can arise from exposure to a variety of fear-inducing and painful stimuli and is not limited to previous suicidal behavior. Such an understanding of suicidal behavior explains why, although suicide attempts tend to increase in lethality as multiple atempters habituate to the pain and fear involved, many individuals (who have presumably acquired the capability for suicide through other experiences) die by suicide on their first atempt (Rudd, Joiner, & Rajab, 1996; Van Orden et al., 2010). In summary, the IPTS states that only under the simultaneous conditions of thwarted belongingnes, perceived burdensomenes, fearlesnes about death, and increased pain tolerance wil an individual make a lethal or near lethal suicide atempt (Joiner, 2005; Van Orden et al., 2010). Specifying Relations among Alcohol-Related Problems and IPTS Constructs Distinguishing the ways in which alcohol-related problems link to the constructs of the IPTS could be important because the kinds of alcohol-related problems individuals experience may put them at greater risk for suicidal behavior. Although the relationships among specific types of alcohol-related problems and the constructs of the IPTS have yet to be empiricaly investigated, others have theorized that the consequences of alcohol misuse afect diferent realms of a heavy drinker?s life (Perkins, 2002; Read, Kahler, Strong, & Colder, 2006). As discussed in more detail in the following section, depending upon the realm a problem afects, it ought to be diferentialy asociated with IPTS constructs. ! 7! Alcohol-related problems and thwarted belongingnes. Joiner (2005) argues that prolonged substance use has powerful negative efects on one?s social support system, and therefore, individuals with prolonged substance use may have higher levels of thwarted belongingnes compared to other populations. In support of this argument, Joiner (2005) cites a literature review examining the social ties of individuals with AUDs (Akerlind & Hornquist, 1992). A main conclusion of this review was that those with AUDs report more lonelines than others. This finding has been supported by more recent studies demonstrating that heavy drinking in the general population is asociated with marital problems ranging from disatisfaction to violence (Marshal, 2003), and that those with AUDs (and other drug dependencies) experience a wide range of interpersonal problems (Doumas, Blasey, & Mitchel, 2006). Additionaly, social disconnection and interpersonal conflict are asociated with death by suicide among heavy drinkers (Pirkola et al., 2000) and with suicide ideation and atempts among those with AUDs (Preuss et al., 2002; You, Van Orden, & Conner, 2011). Thus, one specific problem asociated with heavier alcohol use paterns is decreased positive social relationships, and as the IPTS would predict, this consequence is asociated with suicidal behavior. In the current study, we anticipated that alcohol-related problems of an interpersonal nature (i.e., social problems) would be uniquely and positively asociated with thwarted belongingness. Alcohol-related problems and perceived burdensomenes. There are also alcohol- related problems that sem relevant to perceived burdensomenes. McNaly, Palfai, Levine, and Moore (2003) demonstrated that one reason why individuals engage in heavy drinking is to cope with low self-esteem. Brady (2006) explains that although it remains unclear in the literature if alcohol misuse furthers negative perceptions of the self, some have argued that alcohol misuse ! 8! may lower self-esteem and through this, increase risk for suicide (Kendal, 1983; Lester, 1992). Some evidence for this view has been established in cross-sectional studies that found both an inverse correlation betwen self-estem and alcohol consumption (Glindermann, Geler, & Fortney, 1999; Lewis & O?Neil, 2000) and that heavy drinkers have a greater tendency to make shameful atributions for their mistakes in a variety of hypothetical situations (Dearing, Steuwig, & Tangney, 2005). Additionaly, a longitudinal study in a sample of high school students found that those with consistently low, and importantly those with decreasing levels of, self-esteem reported the highest rates of drinking (Zimerman, Copeland, Shope, & Dielman, 1997). This finding indicates a possible reciprocal relationship betwen alcohol misuse and self-esteem in which students with initialy low levels of self-estem drink heavily, and those with somewhat low levels drink heavily and experience additional drops in self-estem after the onset of heavy drinking. Though such findings have not definitively answered the question regarding the temporal ordering of the alcohol misuse-self-estem relationship, they clearly demonstrate its existence. Furthermore, low self-esteem has been specified as a key component of perceived burdensomenes (Joiner, 2005; Van Orden, 2010). Though lowered self-estem has not traditionaly been measured on alcohol-related problem scales, a recently developed scale includes such problems, and they have been found to correlate wel with other, more frequently measured problems (Read et al., 2006). Thus, one way that alcohol misuse may link to perceived burdensomenes is through its asociation with self-estem. Additionaly, Perkins (2002) lists academic impairment and legal repercussions among other consequences asociated with heavy drinking in college student populations. In that these particular types of problems could result in an individual feling like a liability to others, they too could be asociated with perceived burdensomenes. Similarly, others have found that job ! 9! loss and financial hardships are asociated with death by suicide among heavy drinkers (Pirkola et al., 2000) and suicide ideation and atempts among those with an AUD (Preuss et al., 2002). However, the relationship betwen a broadened sense of role failure, perceived burdensomenes, and suicidal behavior among alcohol misusers has not yet been empiricaly investigated. In the current study we investigated this link, expecting to find unique and positive asociations betwen perceived burdensomenes and the alcohol-related problems of negative self-esteem and occupational and academic impairment. Alcohol-related problems and acquired capability. Although experiences of the aforementioned types of alcohol-related problems would be expected to increase suicide ideation through their asociations with thwarted belongingnes and perceived burdensomenes, those who experience only these types of problems would not be expected to atempt or die by suicide without also acquiring the capability to do so. Joiner (2005) makes the case that substance users experience a variety of frightening and painful events as a direct result of their use. In support of this notion, heavy drinkers are at an increased risk of experiencing the les common alcohol- related problems asociated with physical risk (Kahler, Strong, Read, Palfai, & Wood, 2004; Vik, Carrelo, Tate, & Field, 2000) including risky sexual practices, driving while intoxicated, physical fights, near alcohol-poisoning, and acidental injury (Borges et al., 2006; Kely,!Lynch, Donovan, & Clark, 2001; McLeod, Stockwel, Stevens, & Philips, 1999; Perkins, 2002). Given that such experiences, especialy when they occur repeatedly, may habituate heavy drinkers to the pain and fear asociated with death, we expected that these types of alcohol-related problems would be uniquely asociated with the two components of acquired capability. ! 10! Previous Research on the IPTS and Alcohol-Related Problems To our knowledge, there has only been one study investigating the relationship betwen alcohol-related problems and the main constructs introduced by the IPTS. Specificaly, Lamis and Malone (2011) demonstrated that thwarted belongingnes and perceived burdensomenes mediated the relationship betwen alcohol-related problems (as measured by the Rutgers Alcohol Problem Index [RAPI]; White & Labouvie, 1989) and suicide-pronenes (as measured by the Life Atitudes Schedule-Short Form [LAS-SF], Rohde, Lewinsohn, Langhinrichsen-Rohling, & Langford, 2004). Although this study is a first step in discovering why those who engage in heavy drinking are at an increased risk for al types of suicidal behavior, there are several limitations that hinder our ability to draw firm conclusions from this study. The first limitation is the choice of measure for suicidal behavior used in this study. The outcome measure of suicide-pronenes (LAS-SF; Rohde et al., 2004) correlates only modestly with single-item measures of lifetime suicide ideation (r=.35) and past atempts (r=.25; Rohde, Seley, Langhinrichsen-Rohling, & Rohling, 2003). Relatedly, the lack of clarity in scale items makes it impossible to know which items demonstrate ideation and which are related to atempts. One major advantage of the IPTS is that it makes diferent predictions about the combination of constructs that lead to diferent suicidal behaviors (i.e., pasive ideation, suicidal desire, suicidal intent, and suicide atempt; Joiner, 2005; Van Orden et al., 2010); unfortunately, the LAS-SF does not alow for the parsing out of such behaviors. Similarly, the failure of Lamis and Malone (2011) to measure acquired capability limits the understanding of how alcohol-related problems might prepare individuals to be able to act on suicide ideation. Lastly, in that the LAS-SF includes at least one item pertaining to binge drinking, its use in a study examining the ! 11! relationship betwen suicide pronenes and alcohol-related problems is an example of criterion contamination (Miler, McIntire, & Lovler, 2011). In addition to the limitations noted above, the uni-dimensional nature of the measure used for alcohol-related problems (i.e., the RAPI; White and Labouvie, 1989) in the Lamis and Malone (2011) study does not provide the most informative understanding of which types of alcohol-related problems are asociated with thwarted belongingnes and which are asociated with perceived burdensomenes. As explained earlier, thwarted belongingnes and perceived burdensomenes are diferent constructs, and while they do have some commonalities, a recent factor analytic study has demonstrated that these constructs are distinguishable from one another (Van Orden, Cukrowicz, Wite, & Joiner, 2012). As explained above, there is good reason to believe that some problems asociated with alcohol misuse (e.g., lonelines, loss of friends, marital and family isues) are more indicative of thwarted belongingnes, whereas others (e.g., academic impairment, shame and low self-estem, legal isues) are more indicative of perceived burdensomenes. If such diferences do exist, this could have important consequences for tailoring treatment for individuals who present with diferent compilations of alcohol-related problems. The Curent Study In the current study we sought to both replicate and extend the findings of Lamis and Malone?s (2011) study while using more optimal measures of both pasive suicide ideation and alcohol-related problems. To this end, and in consideration of the findings presented earlier, we tested the following three hypotheses regarding the relationships among alcohol-related problems, IPTS variables, and pasive suicide ideation: ! 12! 1. In a model utilizing a uni-dimensional measure of alcohol-related problems, we predicted that alcohol-related problems would be positively asociated with al constructs of the IPTS and pasive suicide ideation. Furthermore, we predicted that thwarted belongingnes and perceived burdensomenes would partialy mediate the relationships betwen alcohol-related problems and pasive suicide ideation. Such findings would provide support for the model presented by Lamis and Malone, 2011, and would establish the existence of a relationship betwen alcohol-related problems in general and al constructs of the IPTS. However, such a model would not clarify which types of alcohol-related problems are most strongly related to each of the separate IPTS constructs. 2. We expected that diferential asociations would emerge among subtypes of alcohol- related problems and constructs of the IPTS when a multi-dimensional measure of alcohol-related problems (i.e., YACQ; Read et al., 2006) was utilized. Specificaly, we expected that interpersonal alcohol-related problems (i.e., social/ interpersonal YACQ subscale) would be most strongly asociated with thwarted belongingnes, that alcohol-related problems asociated with self-perception and care (i.e., self- perception and self-care YACQ subscales) as wel as academics and work (i.e., academic/occupational YACQ subscale), would be most strongly asociated with perceived burdensomenes, and that alcohol-related problems asociated with risk- taking (i.e., risky behavior YACQ subscale) and physical injury (i.e., blackout drinking YACQ subscale) would be most strongly asociated with both facets of acquired capability. Considering these anticipated diferential asociations among subtypes of alcohol-related problems, we further expected that significant ! 13! relationships would exist betwen pasive suicide ideation and the subtypes of alcohol-related problems asociated with thwarted belongingnes and perceived burdensomenes, and that these relationships would be partialy mediated by thwarted belongingnes and perceived burdensomenes. 3. We finaly proposed that the relationships betwen alcohol-related problems (conceptualized as a uni- or multi-dimensional construct dependent upon results of analyses testing hypothesis 2), IPTS variables, and pasive suicide ideation would remain significant when a symptom of depresion (i.e., sadnes) and coping motives for drinking are taken into acount. In sum, we expected that alcohol-related problems would be asociated with al constructs of the IPTS and pasive suicide ideation, that these asociations would be specific to the types of alcohol-related problems experienced, and that these asociations would be non-spurious when conceptualy related variables were taken into acount. ! 14! Method Participants and Procedure Participants in this study were undergraduate students, age 19 or older, who were enrolled in psychology courses at a large southeastern university. In order to achieve an approximately equivalent gender distribution, we intentionaly oversampled for males (i.e., added an inclusion criterion of male gender once a sufficient number of females participated during each semester of data collection). Following a protocol approved by the university?s internal review board (IRB), the study was advertised on the psychology department?s web- based research system. To protect participants from identifying themselves as regular drinkers, the study was described as a brief, online survey examining the thoughts, felings, and behaviors of young adults. Students who met the age criterion and were interested in participating were able to enroll through the department?s web-based system. Upon signing up, participants were provided with a one-time use web link which redirected them to the survey hosted on a secure website. After reading through the information leter and agreeing to the procedures, participants were presented with the measures, which were available for completion at the participant?s leisure. The majority of participants spent betwen 22 (25 th percentile) and 42 (75 th percentile) minutes completing the measures; however, the amount of time spent with the browser window open ranged from seven seconds to 12 days. Data for this study were collected anonymously and upon completion of the survey (or active opting to discontinue by clicking a box at the bottom of the screen), participants were automaticaly awarded research credit to count toward extra credit points for one of their psychology courses. Participants were included in the analyses reported ! 15! here if they 1) reported consuming at least four alcoholic beverages in the past month (excluded n=257; Gonzalez, Bradiza, & Collins, 2009), and 2) were not flagged as ?inatentive responders? based on the embedded screening measure (excluded n=92; Meade and Craig, 2011; described in more detail below). Our final sample consisted of 295 college students who engage in regular drinking (43% male; n= 127) with an average age of 20.6 years (S.D.= 1.7; range= 19-31). A problem with the display setings of survey questions prevented data on race from being collected from al participants; this problem was addresed after the majority of the data had already been colected. Of the participants that met inclusion criteria and from whom data on race was collected (n=84), 88% selected Caucasian as their primary race, 6% selected Asian American and the remaining 6% selected African American. These figures are comparable to those of both the liberal arts department (82% Caucasian) and the total student population (81% Caucasian) of the university (Auburn University Ofice of Institutional Research and Asesment, 2013), and therefore are likely representative of the sample as a whole. The vast majority of participants (97%, n= 287) reported their ethnicity as non-Hispanic/Latino. Al years in undergraduate studies were represented wel in our sample, with 13% (n= 38) of participants in their first year, 25% (n= 75) in their second year, 28% (n=81) in their third year, 26% (n= 78) in their fourth year, and 8% (n= 23) in their fifth year or beyond. The modal grade point average (GPA) range reported was betwen a 3.00-3.49 (37%; n=111). Five percent of participants (n=15) reported their GPA as a 4.00, 23% (n=68) reported it as faling betwen a 3.50-3.99, 23% (n=69) reported it as faling betwen a 2.50-2.99, and 12% reported it as faling at or below a 2.49. The majority (56%, n=166) of participants reported their marital status as single, 39% (n=115), reported being in a ! 16! commited relationship, and the remaining 5% reported being married (n=1), engaged (n=5), divorced (n=1), or living with a partner (n=7). Measures Alcohol consumption. For descriptive purposes, we asesed alcohol consumption in this study using the timeline follow back method (TLFB; Sobel & Sobel, 1992). Participants were provided with an online calendar marked with federal and university holidays and a standard drink conversion chart, and were instructed to fil in each calendar date with the total number of standard alcoholic beverages consumed on that date. This method has been used widely in both clinical and nonclinical populations, has demonstrated aceptable psychometric properties, and is often preferred over typical quantity-frequency measures when used with individuals who do not have regular daily drinking paterns (Sobel & Sobel, 1995). The calendar included the month in which the measures were completed as wel as the three months prior, and participants were asked to report for each of the 90 days prior to the date of their participation in the study. Although we had intended to use the full 90-day reporting period, many participants included only partial months for either the most and/or least recent months within the reporting period. An examination of the data showed that participants were most likely to provide complete data for the full month prior to the month in which they completed the survey. For this reason, we opted to use the total number of drinks reported in the 30-day month (or most recent thirty days in months with more or les than thirty days) prior to the month in which the participant completed the survey measures as the alcohol consumption variable (e.g., ?alcohol consumption? for a participant completing the survey on September 8 th would be the total number of drinks reported on August 2nd-31 st ). The mean number of total drinks reported was ! 17! 29.4 (SD= 26.2, range 4.0-115.2), and approximately 50% of the sample reported consuming 21 or more drinks in the specified time period. Alcohol related problems. The Brief Young Adult Alcohol Consequences Questionnaire (B-YACQ; Kahler, Strong, & Read, 2005) was used as a uni-dimensional measure of alcohol- related problems. The B-YACQ is a 24-item measure of alcohol-related problems that utilizes a dichotomous (present or absent) scoring format. Unlike the full YACQ (described below), the B-YAACQ was designed using Rasch modeling (Rasch, 1960). This modeling procedure creates a measure that aseses an underlying construct along a continuum; thus, higher scores on the B- YACQ not only indicate a wider variety of alcohol related problems but also reflect a more severe patern. In our sample, the mean score on this measure was 11.25 (SD= 6.35, range 0-24). Acording to Kahler and colleagues (2005), a score of about 10 indicates the likely presence of some important consequences, whereas a score of about 15 may indicate the presence of symptoms consistent with AUD; that said, it appears our sample adequately reflects a wide range of participants in terms of alcohol-related problems. In our sample, the internal consistency of the B-YACQ (Cronbach?s ?=.91) was excelent. In analyses testing for diferential asociations among subtypes of alcohol-related problems and the IPTS constructs (i.e., hypothesis 2), each of the eight subscales of the 48-item Young Adult Alcohol Consequences Questionnaire (YACQ; Read et al., 2006) were used as separate, domain-specific, measures of alcohol-related problems. The developers of this scale have shown it to have strong psychometric properties when used as either a uni-dimensional or multi-dimensional measure (Read, Merril, Kahler, and Strong, 2007), with a confirmatory factor analysis providing evidence of diferential item loading onto the eight separate but related factors that comprise the eight subscales (i.e., social/interpersonal, self-care, physiological dependence, ! 18! self-perception, impaired control, academic/ occupational, risky behavior, and blackout drinking; Read et al., 2006). The YACQ was created with a dichotomous scoring format (i.e., each problem is recorded as present or absent within the reporting period), which precludes the asesment of the frequency of occurrence of each of the alcohol-related problems. This inability to ases frequency has been cited as one of the major criticisms of this measure (Devos-Comby & Lange, 2008). However, Kahler et al. (2004) have argued that frequency of occurrence is relatively unimportant when asesing the degree of alcohol-related problems experienced. Furthermore, they found strong correlations among dichotomous and frequency based scoring methods for measures of alcohol related problems similar to the YACQ (Kahler et al., 2004). Given the importance that repeated experience plays in the habituation proces that must occur to acquire the capability for suicide (Joiner, 2005; Van Orden et al., 2010), we chose to modify the YACQ to include a frequency scale. On the modified version of this scale, participants were asked to rate how often they had experienced each alcohol-related problem on a 1(never) to 5( more than 10 times) scale (similar to the RAPI; White & Labouvie 1989). In our sample this version of the full scale was highly correlated with the dichotomously-scored format (r = .92). The mean score on the frequency version of the full scale was 33.50 (SD= 27.31). The internal consistency of both the full scale measure (Cronbach?s ?= .95) and each of the eight subscales (social/interpersonal subscale Cronbach?s ?=.81, self-perception subscale Cronbach?s ?=.78, self-care subscale Cronbach?s ?=.84, academic/ occupational subscale Cronbach?s ?=.79, risky behavior subscale Cronbach?s ?=.81, blackout drinking subscale Cronbach?s ?=.84, physical dependence subscale Cronbach?s ?=.66, impaired control subscale Cronbach?s ?=.74),was adequate. ! 19! Coping motives for drinking. The coping motives subscale (5 items) of the 20-item, self-report Drinking Motives Questionnaire (DMQ; Cooper, 1994) was used to measure a subset of the reasons for which participants reported they would be likely to drink. Items on this questionnaire are rated on a 5-point Likert scale from 1-5 to indicate how frequently the participant drinks for each specified reason. The DMQ has demonstrated strong psychometric properties (Cooper, 1994). The coping motives subscale of the DMQ was included as a separate covariate to test hypothesis 3. In our sample the internal consistency of this subscale (Cronbach?s ?=.85), was good. Thwarted belongingnes and perceived burdensomenes. The 15-item self-report Interpersonal Neds Questionnaire (INQ; Van Orden et al., 2012) was used to assess thwarted belongingnes (nine items) and perceived burdensomenes (six items). Participants used a 7- point Likert scale to indicate how wel each item described them. Scores for each subscale were computed by averaging the items thus total scores on each of the subscales range from 0-6. Higher scores on the subscales indicate higher levels of each of the constructs. A recent study of the psychometric properties of the INQ indicated that this scale measures perceived burdensomenes and thwarted belongingnes, can be used with diverse populations, and correlates wel and in the expected direction with similar constructs (Van Orden et al., 2012). In our sample, the mean was 0.98 (SD= 1.02) on the thwarted belongingnes subscale, and 0.20 (SD=0.51) on the perceived burdensomenes subscale. The internal consistencies for both the thwarted belongingnes (Cronbach?s ?=.89) and perceived burdensomenes (Cronbach?s ?=.92) subscales were excelent. Fearlesnes about death. The seven-item, self-report Acquired Capability for Suicide, fearlesnes about death subscale (ACSS-FAD; Ribeiro et al., 2014) was used to ases ! 20! fearlesnes about death. Participants rated items on a 5-point Likert scale, in order to indicate the degree to which each item describes them. Total scores can range from 0 to 28, with higher scores indicative of greater fearlesnes about death. In a recent study of the psychometric properties of this subscale, Ribeiro and colleagues (2014) found evidence of good factor structure and discriminant and convergent validity. In the current sample the mean score on fearlesnes about death was 21.48 (SD= 6.44). The internal consistency of this scale (Cronbach?s ?=.86) was good. Pain tolerance. The seven-item, self-report Discomfort Intolerance Scale (DIS; Schmidt, Richey, Fitzpatrick, 2006) was used to ases self-reported pain tolerance. Items are rated on a 7- point Likert scale to indicate the degree to which they are descriptive of the participant. Total scores can range from 0 to 42, with higher scores indicative of higher pain tolerance. In a preliminary investigation of the psychometric properties of the DIS, it was found to correlate wel and in the expected directions with relevant constructs, evidencing good convergent and discriminant validity (Schmidt, Richey, & Fitzpatrick, 2006). The mean score in our sample was 29.34 (SD= 6.87). The internal consistency for the DIS was aceptable (Cronbach?s ?=.71). Passive suicide ideation. To measure pasive suicide ideation in our sample we used the first five items of the Beck Suicide Scale (BSS; Beck & Ster, 1991). On this measure, participants are asked to rate 21 items on a scale ranging from 0-2, regarding their suicidal thoughts and behaviors. We opted to use only the first five items of this scale for two reasons. First, in this study we were interested only in pasive suicide ideation. Factor analytic research has indicated that the first 19 items of the BSS (items 20 and 21 ases frequency and severity of suicide atempts and are typicaly not included when the measure is intended to ases various degrees of suicidal ideation), measure two separate factors: suicide ideation and desire, and ! 21! resolved plans and preparation (RPP; Ster et al., 1993: Holden & DeLisle, 2005; Joiner, Rudd, & Rajab, 1997; Wite et al., 2006), with the first five items generaly loading more strongly onto the suicide ideation and desire factor. Second, the first five items were designed as a screener (Beck, Brown, and Ster, 1997). In this study, as is typical with this measure, only those participants who reported a desire for suicide (score of greater than 0 on items 4 and/or 5) were instructed to complete items 6-19. In our sample, 5% (n=14) of al participants were required to complete items 6-19. Thus because we only sought to analyze pasive suicide ideation, and because the majority of our sample did not complete items beyond the first five screener items, we felt justified in using the total score on the BSS items 1-5 as our measure of pasive suicide ideation. In our sample the mean score on this measure was 0.19 (SD=0.54, range 0.00-2.63). The internal consistency of the abbreviated version of this scale was aceptable (Cronbach?s alpha=.75). Sadnes. We used item 18 (I felt sad.) from the Center for Epidemiologic Studies of Depresion Scale (CES-D; Radloff, 1977) as a covariate in the second path analytic model in order to control for recent experiences of sadnes. This item was selected from the full 20-item scale as a covariate because it is a common symptom of depresion and because there was strong conceptual overlap betwen the other CES-D items and the INQ. Participants were asked to rate the frequency of times during the past wek when they felt sad on a four-point scale ranging from rarely, les than 1 day to most or all of the time, 5-7 days. The mean score on this item was 0.69 (SD=0.77, range 0-3). Bogus Items (Meade & Craig, 2011). In order to check for the presence of and screen out inatentive responders, we included eight bogus items embedded within our other study measures. Such items can only logicaly be endorsed in one direction (e.g., I slep les than one ! 22! hour per night) and have been found to be among the most sensitive identifiers of potential inatentive responders (Mead & Craig, 2011). In order to identify inatentive responders within our data set, individuals who completed the survey in five minutes or les (n=44) were first screened out, as such a completion time was dramaticaly lower than the average, and likely was an indicator that participants were clicking through the survey without reading the items. Once this was done, the screening criterion of reporting four or more drinks in the previous month was applied, and only those who met this criterion were included in the final screening procedure (n= 343). In the sample that met the inclusion criterion, we finaly flagged participants who answered two or more of the bogus items incorrectly (10% of the remaining sample; n=48). ! 23! Results Descriptive statistics for al study variables, computed using IBM SPS version 21.0, are provided in Table 1. Al univariate outliers were fenced in at 3 standard deviations above or below the mean. Zero-order correlations among variables are reported in Table 2. These correlations were inspected to gain a preliminary understanding of the relationships among study variables. As can be sen in Table 2, alcohol-related problems, regardles of the how this construct was measured (i.e., either the B-YACQ or the eight YACQ subscales) tended to have modest, positive correlations with both thwarted belongingnes and perceived burdensomenes, but not with pasive suicide ideation, pain tolerance, or fearlesnes about death. A similar patern of results emerged among coping motives for drinking and the IPTS and pasive suicide ideation variables, with the exception of a modest, negative correlation with fearlesnes about death. Notably, gender (coded 0 = male; 1 = female), thwarted belongingnes, perceived burdensomenes, and sadnes had positive correlations with pasive suicide ideation. Zero-order correlations were also utilized to test for diferential asociations betwen subtypes of alcohol-related problems and the constructs of the IPTS (described in detail below). To compute zero-order correlations and to test al path models, we used Mplus Version 7.1 (Muthen & Muthen, 1998-2012) and the robust maximum likelihood (MLR) estimator. This estimator was chosen because it is robust against non-normality (Brown, 2006), and some of the variables included within the models did not follow a normal distribution. When investigating the indirect efects in each model, we used a bias-corrected bootstrapping (BC) procedure with ! 24! 5,000 resamplings (Preacher & Hayes, 2008). Individuals who met criteria based on age, alcohol consumption, and atentive responding, and who had data for at least one of the included variables were included in al analyses. The degree of mising data across variables included in each of the models was low; the proportion of complete data for each pair of variables ranged from 83 to 95% in the first model, and 83 to 96% in the second. Full information maximum likelihood (FIML) was used to handle mising data. Al of the path analytic models that we tested were saturated models with zero degrees of freedom; therefore, we do not report fit statistics for these models. Hypothesis 1: Replication and Extension of Lamis and Malone (2011) Model First, we wanted to ensure that we could replicate the findings from Lamis and Malone (2011) in our sample, especialy given that we utilized diferent measures of alcohol-related problems and pasive suicide ideation. Additionaly, we sought to extend their findings by determining whether alcohol-related problems were asociated with the acquired capability constructs of the IPTS. To test the hypotheses that alcohol-related problems would be positively asociated with al four IPTS constructs (i.e., thwarted belongingnes, perceived burdensomenes, fearlesnes about death, and pain tolerance), as wel as that the interpersonal constructs (i.e., thwarted belongingnes and perceived burdensomenes) would partialy mediate the relationships betwen alcohol-related problems and pasive suicide ideation we conducted a path analysis (se Figure 1 for full model). Although in their analyses Lamis and Malone controlled for social desirability and a variety of demographic variables including gender, we did not measure social desirability. Additionaly, because none of the other demographic variables that we collected were asociated with alcohol-consumption, we only controlled for the efect of gender in this model. ! 25! In order to evaluate the above-mentioned hypotheses, we investigated the parameter estimates for al specified direct (Table 3) and indirect (Table 4) paths in the model; these estimates provided partial support for our first set of hypotheses. As expected, there was a significant direct path from alcohol-related problems to thwarted belongingnes (?=0.15, p<.02); the direct path from alcohol-related problems to perceived burdensomenes approached statistical significance (?=0.12, p=.07). Additionaly, there were significant direct paths from both thwarted belongingnes (?= 0.24, p<.01) and perceived burdensomenes (?= 0.28, p<.01) to pasive suicide ideation. Whereas the indirect path from alcohol-related problems to pasive suicide ideation through thwarted belongingnes (?=0.04, [95% CI=.00-.07], p=.05), approached statistical significance, the indirect path from alcohol-related problems to pasive suicide ideation through perceived burdensomenes (?=0.03, [95% CI= -.02-.08], p=.15), did not. Although in contrast to those found in Lamis and Malone (2011), the indirect paths in our sample were not statisticaly significant, the effects were similarly modest in both studies (i.e., in their sample the standardized values for the indirect efects were .08 and .01 for perceived burdensomenes and thwarted belongingnes, respectively). Thus, rather than indicating truly discrepant results, the diferences in statistical significance of these paths are likely due to our smaler sample size. In sum, the asociations among alcohol-related problems, thwarted belongingnes, perceived burdensomenes, and pasive suicide ideation reported here are similar to those reported by Lamis and Malone (2011). In the current study we atempted to extend the findings reported by Lamis and Malone (2011) by investigating the relationships betwen alcohol-related problems, and the acquired capability variables (i.e., fearlesnes about death and pain tolerance) of the IPTS. Contrary to our expectations, the direct path from alcohol-related problems to fearlesnes about death was ! 26! negative and significant (?= -0.14, p=.01). Additionaly, the direct path from alcohol-related problems to pain tolerance was not significant (?= -0.09, p=.12). Overal in this model, we acounted for 2% of the variance in perceived burdensomenes, 4% of the variance in thwarted belongingnes, 11% of the variance in fearlesnes about death, 11% of the variance in pain tolerance, and 22% of the variance in pasive suicide ideation. In order to ensure that the inclusion of gender as a covariate was not unduly influencing our results, we reran the model without it. In this version of the model, the patern and significance of the results remained in line with those reported above with some minor exceptions. Specificaly, when gender was not included as a covariate in the model, the direct path from alcohol-related problems to fearlesnes about death was reduced to non-significance (?= -0.09, p=.14), and the percentage of the variances explained in both fearlesnes about death and pain tolerance were each reduced to les than one percent. Hypothesis 2: Determination of differential associations betwen YACQ subscales and IPTS constructs In order to test hypothesis 2 regarding the diferential asociations among domain- specific types of alcohol-related problems (i.e., the eight YACQ subscales) and the IPTS constructs, we used Meng, Rosenthal, and Rubins?s (1992) test for diferences betwen correlated correlations. This test was used to determine whether alcohol-related problems should be divided into domain specific groupings when investigating their relationships with the IPTS constructs and pasive suicide ideation when testing our final path model. Despite the fact that the results of these analyses indicated some significant diferences among subscales, in general, the correlations with the IPTS variables were modest (ranging from r=-.12 to r=.22). Additionaly, rather than highlighting strong relationships betwen some ! 27! specific subtypes of alcohol-related problems and the IPTS constructs, these results indicated that there are some subtypes of problems that have particularly weak relationships with the IPTS constructs. This same patern of results was found regardles of whether the subscales were scored using the Likert or dichotomous scoring formats. Acordingly, we reasoned that there was litle utility in separating alcohol-related problems into separate subscales when investigating their relationships with the IPTS and suicide ideation variables. Hypothesis 3: A more stringent test of the Lamis and Malone (2011) findings As noted previously, although alcohol-related problems have been identified as one potential explanatory factor in the relationship betwen alcohol consumption and suicidal behavior, depresive symptoms and coping motives for drinking have as well. The preliminary results reported above indicate that both sadnes and coping motives for drinking had positive zero-order correlations with thwarted belongingnes and perceived burdensomenes (se Table 2). In order to ensure that the asociations reported in model 1 are not beter acounted for by sadnes or coping motives for drinking, we ran a second path analytic model, in which these two variables were included as covariates along with gender. In addition to the hypotheses tested in model 1, we expected that in this model there would be a positive and significant correlation betwen coping motives for drinking and alcohol-related problems. We also expected that there would be positive and significant asociations betwen sadnes and thwarted belongingnes, perceived burdensomenes, and pasive suicide ideation. We did not make a priori predictions regarding the other asociations betwen the covariates and variables included in the model. Similar to the results of model 1, in this model the direct paths from both thwarted belongingnes (?= 0.21, p<.01) and perceived burdensomenes (?=0.23, p=.03) to pasive suicide ideation were significant. As expected, there was also a positive correlation betwen ! 28! coping motives for drinking and alcohol-related problems (?= 0.55, p<.01). Similarly, the direct paths from sadnes to thwarted belongingnes (?=0.38, p<.01), perceived burdensomenes (?=0.49, p<.01), and pasive suicide ideation (?=0.14, p=.04), were all significant. Contrary to our expectations, when coping motives for drinking and sadnes were added to the model as covariates, the direct paths betwen alcohol-related problems and the IPTS variables found in model 1 were reduced to non-significance. Additionaly in this model there were no significant direct paths betwen coping motives for drinking and any of the IPTS variables nor betwen coping motives for drinking and pasive suicide ideation. Neither the indirect paths from alcohol-related problems nor from coping motives to pasive suicide ideation through either thwarted belongingnes or perceived burdensomenes were significant (se Table 6). Despite the fact that only gender and sadnes had significant asociations with any of the IPTS variables, this model acounted for 21% of the variance in thwarted belongingnes, 24% of the variance in perceived burdensomenes, 11% of the variance in fearlesnes about death, 12% of the variance in pain tolerance, and 23% of the variance in pasive suicide ideation. As with model 1, we reran this model without gender included as a covariate. Once again, in this model the general patern and significance of the results was similar to those reported in detail above with just a few exceptions. Specificaly, when gender was not taken into acount in this model the direct path from sadnes to pasive suicide ideation was reduced from significant to approaching significance (?=0.13, p=.07), and the percentage of variance explained in both fearlesnes about death and pain tolerance was reduced to one percent for each. ! 29! Aditional Analysis: A more parsimonious path model Although we had not planned to test a third path model, the surprising results of model two led us to question the utility of including any alcohol-related constructs in the model. In model two it appeared that sadnes was likely explaining the majority of the variance in thwarted belongingnes and perceived burdensomenes, which appeared to be explaining the majority of the variance in pasive suicide ideation in both models one and two. We decided to test these asumptions by running a path analytic model that included only sadnes, the IPTS variables, and pasive suicide ideation, as wel as gender as a covariate. In this model we found similar significant direct paths as those reported in model 2. Specificaly, we found that the direct paths from sadnes to thwarted belongingnes (?=0.42, p<.01), and from sadnes to perceived burdensomenes (?=0.18, p<.01) were significant. We also found that the direct paths from thwarted belongingnes (?=0.21, p<.01), and from perceived burdensomenes (?=0.24, p=.03) to pasive suicide ideation were significant. Additionaly in this model the direct path from sadnes to pasive suicide ideation approached statistical significance (?=0.12, p=.08). Exactly as was the case in model 2, this model acounted for 21% of the variance in thwarted belongingnes, 24% of the variance in perceived burdensomenes, and 23% of the variance in pasive suicide ideation. In addition this model acounted for 12% and 9% of the variance in fearlesnes about death and pain tolerance, respectively. As with the other models reported above, we reran this model without gender included as a covariate; in this case the patern and significance of al results remained the same, with only the percentage of variance in both fearlesnes about death and pain tolerance changing (each again being reduced to approximately one percent). ! 30! Discusion In this study we examined the relationships betwen alcohol-related problems, the IPTS variables, and pasive suicide ideation. We expanded upon the existing literature in a number of ways. First, whereas in the only other study of the relationship betwen alcohol-related problems and the IPTS variables in a college student population (.i.e., Lamis & Malone, 2011) the authors utilized the construct of suicide pronenes as their outcome measure, we used pasive suicide ideation as the outcome measure in this study. Second, we included al four of the IPTS constructs in our study in order to investigate whether the experience of alcohol-related problems was asociated with not only the desire but also the capability for suicidal behavior. Given that within the past decade researchers interested in alcohol-related problems have posited that there are diferent kinds of alcohol-related problems, each of which may have unique impacts on a drinkers life (i.e., Perkins, 2002; Read et al., 2006), we also investigated whether these subtypes of alcohol-related problems had diferential relationships with the IPTS constructs. Additionaly, we tested whether the asociations betwen alcohol-related problems and the interpersonal constructs of the IPTS remained significant when the two conceptualy related constructs of coping motives for drinking and sadnes were taken into acount. With regards to our first hypothesis, we were largely able to replicate the findings of Lamis and Malone (2011), finding that the direct path from alcohol-related problems to thwarted belongingnes was significant and that the direct path from alcohol-related problems to perceived burdensomenes approached significance. Although our results difered in that the indirect paths of alcohol-related problems on pasive suicide ideation through perceived burdensomenes and ! 31! thwarted belongingnes were not significant, the size of these efects were similar across both samples, and thus the diferences in significance were likely more asociated with sample size than true diferences in efects. However, one important diference betwen the results presented in our model and those presented in Lamis and Malone (201l) is that in their model the direct path from alcohol-related problems to suicide pronenes was significant, and in ours the path from alcohol-related problems to pasive suicide ideation was not significant. This diference, along with the fact that neither the uni-dimensional nor the multi-dimensional measures of alcohol-related problems in our study had significant zero-order correlations with pasive suicide ideation, appear to substantiate our concerns about the suicide pronenes construct. Specificaly, these discrepant results in the context of otherwise comparable findings sem to suggest that the relationship betwen suicide pronenes and alcohol-related problems found in Lamis and Malone (2011) was likely a reflection of the conceptual similarity betwen their measures rather than a true relationship. Based on the results of our study it sems unlikely that there is a relationship betwen alcohol-related problems and suicide ideation in undergraduate students, despite the recent line of research utilizing suicide pronenes as an outcome measure (i.e., Lamis & Malone, 2012; Lamis, Malone, & Jahn, in pres), that might lead one to asume otherwise. In our first model we also examined asociations betwen alcohol-related problems and the two components of the acquired capability for suicide: fearlesnes about death and pain tolerance. Although in the initial description of the IPTS, Joiner (2005) hypothesized that individuals with alcohol and other substance use problems are at an increased risk of experiencing painful and frightening circumstances and through these circumstances acquiring the capability for suicide, we found no such evidence of this in model one. Although our findings in this model did not support our hypotheses regarding the asociation betwen alcohol-related ! 32! problems and acquired capability, we recognized that in this particular formulation of the relationship the efects of repeated experience were not captured. In order to addres the above-mentioned isue, as wel as to determine whether subtypes of alcohol-related problems had diferential asociations with each of the IPTS constructs, we tested our second hypothesis. In general, we had anticipated that when the eight subscales of the YACQ, each of which measures alcohol-related problems asociated with a particular life domain, were measured on a frequency scale, distinct relationships betwen some of these subscales and the IPTS constructs would emerge. Contrary to our expectations, the results of this analysis indicated that in general alcohol-related problems of al types were modestly positively correlated with both thwarted belongingnes and perceived burdensomenes, and were not correlated with fearlesnes about death or physical pain tolerance. This was the case regardles of whether the subtypes of alcohol-related problems were scored dichotomously or on a frequency scale. We draw two conclusions from these findings. First, these results indicate that the kinds of alcohol-related problems measured on the YACQ are not related to the components of the acquired capability for suicide. Secondly, they suggest that, at least when it comes to predicting the variables included in our model, there was litle utility in diferentiating among subtypes of alcohol-related problems. Finaly, in order to examine diferential asociations betwen alcohol-related problems and the IPTS constructs when coping motives for drinking and sadnes are taken into acount, we ran a second path model in which we included both of these constructs as covariates. Contrary to our hypotheses, in this model the direct paths from alcohol-related problems were reduced to non-significance, and only the paths from sadnes to thwarted belongingnes, perceived burdensomenes, and pasive suicide ideation, as wel as those from gender to ! 33! thwarted belongingnes and the acquired capability constructs, remained significant. As a further test of the impact of any of the alcohol-related variables on the IPTS constructs and pasive suicide ideation, we conducted a third path analysis in which only sadnes and gender were included as exogenous variables. In this model we found that nearly the exact same amount of variance was explained in each of the endogenous variables, a finding that we interpret as additional evidence that the alcohol-related constructs did not have a significant impact beyond that acounted for by sadnes. In sum our findings suggest that there are modest relationships betwen alcohol-related problems and the interpersonal constructs of the IPTS, but these relationships appear spurious and beter acounted for by recent experiences of sadnes. Additionaly, whereas other authors (i.e., Lamis & Malone, 2011) have found a relationship betwen alcohol-related problems and the construct of suicide pronenes, the lack of an asociation found in our study betwen alcohol- related problems and pasive suicide ideation suggests that suicide pronenes is capturing a diferent construct. Although both the IPTS and recent research on alcohol-related problems would predict diferential asociations betwen IPTS constructs and subtypes of alcohol-related problems, our findings suggest that for at least those alcohol-related problems captured on the YAAC-Q (Read et al., 2006), there are modest asociations betwen al subtypes of alcohol- related problems and thwarted belongingnes and perceived burdensomenes but not the acquired capability constructs. Lastly, our results support the main tenets of the IPTS: we found that perceived burdensomenes and thwarted belongingnes have the strongest relationships with pasive suicide ideation, beyond that explained by sadnes alone. Despite what this study adds to the literature, it has several limitations. Most importantly, the fact that our sample consisted of generaly wel-adjusted college students may have limited ! 34! our ability to find efects. Specificaly, the severe restriction of range in the thwarted belongingnes, perceived burdensomenes, and pasive suicide ideation variables in our sample does not alow us to speak to the relationships that could exist betwen these and the alcohol- related variables at higher levels of the thwarted belongingnes, perceived burdensomenes, and pasive suicide ideation constructs. In contrasting our study with those that have found significant relationships betwen alcohol-related constructs and suicidal behavior in college student populations (e.g., Gonzalez, Bradiza, & Collins, 2009; Gonzalez, Collins, & Bradiza, 2009), our choice to not recruit specificaly for lifetime experiences of pasive suicide ideation clearly resulted in a les pathological, but likely more generalizable, sample of undergraduate students. At the same time, we were able to capture a wide range within the alcohol consumption, alcohol-related problems, and coping motives for drinking variables in our sample, which suggests that there is not a strong asociation betwen variation in these constructs and the IPTS constructs nor pasive suicide ideation. Again while this suggests the lack of a relationship betwen these constructs in a generaly wel-functioning sample of college students, the findings from our study cannot be generalized to more severe samples, such as those with an AUD who may experience severe physical and interpersonal consequences (e.g., hospitalization, homelesnes, job loss, divorce, jail time, etc.) as a result of their drinking (e.g., You, Van Orden, and Conner, 2011). Another notable limitation is that our data were collected cross-sectionaly. As others have expresed, although such design can be helpful in determining the existence of relationships among variables, causal inferences cannot be drawn from a study carried out through such methods. Given that we viewed our study as a first step in determining the existence and relative strength of relationships among al IPTS variables and alcohol related variables, such a design ! 35! was warranted. Considering the fact that we found recent experiences of sadnes to have the most robust relationships with the IPTS and pasive suicide ideation variables, future research should atempt to distinguish the causal order of these variables. Our results cannot speak to whether alcohol-related problems had a causal role in the sadnes that students reported, or whether coping motives for drinking were endorsed in response to current felings of sadnes. To determine the temporal order in which these constructs may be emerging, and particularly whether they relate to each other in a causal manner, more sophisticated methods are needed. A final limitation is that we did not measure the construct of suicide pronenes in our sample. With this in mind, we cannot say for certain that the significant asociations betwen alcohol-related problems and suicide pronenes found by Lamis and Malone (2011) was due to the overlap betwen the two constructs. These diferences also could have reflected general diferences betwen our samples; while the similar findings regarding al other relationships betwen their model and ours, as wel as similar demographics of the two samples makes this explanation unlikely, at this point it remains an empirical question. In the future researchers ought to measure both suicide pronenes and pasive suicide ideation, so that it wil be possible to determine how each relates to other relevant constructs. Despite the above-mentioned limitations, this study contributes to the literature in three distinct ways. As the first study to investigate the relationships among al four IPTS constructs and specific subtypes of alcohol-related problems, we provide a rigorous test of the hypotheses initialy proposed in the first articulation of the IPTS (Joiner, 2005), regarding the impacts of the negative experiences triggered by substance use on an individual?s desire and capability for suicide. Our findings do not support the specificity of the efects of alcohol-related problems on each of the IPTS components, but rather suggest that there are modest relationships among al ! 36! types of alcohol-related problems and the IPTS constructs asociated with suicidal desire. However, these relationships can best be captured by recent experiences of sadnes. Another important contribution is that this is the first study to investigate the utility of distinguishing among subtypes of alcohol-related problems when relating them to suicidal behavior. The fact that al subtypes were highly correlated with one another, and had similar relationships with al other variables in our model suggests that litle relevant information is gained by dividing problems into these subscales. Rather, our findings support the use of a brief, dichotomously scored alcohol-related problems measure (i.e., the B-YACQ, Khaler et al., 2005) when investigating these kinds of relationships among undergraduate students. This measure captured the same kinds of relationships as did the full scale, and frequency scored versions of the measure, while proving les demanding for participants, and utilizing a user-friendly scoring format. Finaly, this study provided additional support for the IPTS, given that we consistently found strong, theoreticaly predicted relationships among thwarted belongingnes, perceived burdensomenes, and suicide ideation. These findings suggest that, within a group of regular drinking undergraduate students, thwarted belongingnes and perceived burdensomenes have the strongest asociations with pasive suicide ideation. ! 37! References Akerlind, I., & Hornquist, J.O. (1992). Lonelines and alcohol abuse: A review of evidences of an interplay. Social Science & Medicine, 34, 405-414. doi: 10.1016/0277-9536(92)90300-F Auburn University Ofice of Institutional Research and Asesment (2013). Total enrollment by gender and ethnicity [Data file]. Retrieved from htps:/oira.auburn.edu/factbook/enrollment/enrtrends/ebg.aspx Baumister, R.F. (1990). Suicide as escape from self. Psychological Review, 97, 90-113. doi: 10.1037/0033-295X.97.1.90 Beck, A.T., Brown, G.K., & Ster, R.A. (1997). Psychometric characteristics of the scale for suicide ideation with psychiatric outpatients. Behaviour Research and Therapy, 35, 1039- 1046. doi:10.1016/S0005-7967(97)00073-9 Bender, T. W., Gordon, K. H., Bresin, K., & Joiner, T. E. (2011). Impulsivity and suicidality: A test of the mediating role of painful experiences. Journal of Afective Disorders, 129, 301?307. doi:10.1016/j.jad.2010.07.023 Berkowitz, A. D., & Perkins, H. W. (1986). Problem Drinking among College Students: A Review of Recent Research. Journal of American College Health, 35, 1, 21-28. Borges, G., Cherpitel, C. J., Orozco, R., Bond, J., Ye, Y., Macdonald, S., Giesbrecht, N., ... Poznyak, V. (2006). Acute alcohol use and the risk of non-fatal injury in sixten countries. Addiction, 101, 7, 993-1002. doi:10.1111/j.1360-0443.2006.01462.x Borges, G., Walters, E. E. and Kesler, R. C. (2000) Asociations of substance use, abuse, and ! 38! dependence with subsequent suicidal behaviour. American Journal of Epidemiology 151, 781?789. Brady, J. (2006). The asociation betwen alcohol misuse and suicidal behaviour. Alcohol & Alcoholism, 41, 473-478. doi:10.1093/alcalc/agl060 Brown, T.A. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford. Bryan, C.J., Cukrowicz, K.C., West, C.L., & Morrow, C.E. (2010). Combat experience and the acquired capability for suicide. Journal of Clinical Psychology. 66, 1044-1056. doi:10.1002/jclp.20703 Carey, K. B., & Correia, C. J. (1997). Drinking motives predict alcohol- related problems in college students. Journal of Studies on Alcohol, 58, 100 ?105. Carpenter, K.M., & Hasin, D.S., (1998). A prospective evaluation of the relationship betwen reasons for drinking and DSM-IV alcohol-use disorders. Addictive Behaviors, 23 (1), 41- 46. Cavanagh, J. T. O., Carson, A. J., Sharpe, M., & Lawrie, S. M. (2003). Psychological autopsy studies of suicide: a systematic review. Psychological Medicine, 33, 3, 395-405. doi: 10.1017/S0033291702006943 Cooper, M. L. (1994). Motivations for alcohol use among adolescents: Development of a four- factor model. Psychological Asesment, 6, 117-128. Cooper, M. L., Frone, M. R., Russel, M., & Mudar, P. (1995). Drinking to regulate positive and negative emotions: A motivational model of alcohol use. Journal of Personality and Social Psychology, 69, 990?1005. Cox, M., & Klinger, E. (1988). A motivational model of alcohol use. Journal of Abnormal ! 39! Psychology, 97, 168-180. Dearing, R. L., Stuewig, J., & Tangney, J. P. (2005). On the importance of distinguishing shame from guilt: Relations to problematic alcohol and drug use. Addictive Behaviors, 30, 7, 1392-1404. doi: 10.1016/j.addbeh.2005.02.002 Devos-Comby, L., & Lange, J. E. (2008). Standardized measures of alcohol-related problems: A review of their use among college students. Psychology of Addictive Behaviors, 22, 349-361. doi:10.1037/0893-164X.22.3.349 Doumas, D., Blasey, C., & Mitchel, S. (2006). Adult atachment, emotional distres, and interpersonal problems in alcohol and drug dependency treatment. Alcoholism Treatment Quarterly, 24, 4, 41-54. doi:10.1300/J020v24n04-04 Duberstein, P.R., Conwel, Y. & Caine, E.D., (1993). Interpersonal stresors, substance abuse, and suicide. Journal of Nervous and Mental Disease, 181, 80-85. doi: 10.1097/00005053-199302000-00002 Durkheim, E. (1897). Le Suicide: Etude de socologie. Paris, France: F. Alcan. Glindemann, K. E., Geler, E. S., & Fortney, J. N. (1999). Self-Estem and Alcohol Consumption: A Study of College Drinking Behavior in a Naturalistic Seting. Journal of Alcohol and Drug Education, 45, 1, 60-71. Gonzalez, V. M., Bradiza, C. M., & Collins, R. L. (2009). Drinking to Cope as a Statistical Mediator in the Relationship Betwen Suicidal Ideation and Alcohol Outcomes Among Underage College Drinkers. Psychology of Addictive Behaviors, 23, 3, 443-451. doi:10.1037/a0015543 Gonzalez, V. M., Collins, R. L., & Bradiza, C. M. (2009). Solitary and social heavy drinking, suicidal ideation, and drinking motives in underage college drinkers. Addictive ! 40! Behaviors, 34, 12, 993-999. doi:10.1016/j.addbeh.2009.06.001 Gonzalez, V.M., & Hewel, V.M. (2012). Suicidal ideation and drinking to cope among college binge drinkers. Addictive Behaviors, 37, 994-997. doi: 10.1016/j.addbeh.2012.03.027 Ham, L. S., & Hope, D. A. (2003). College students and problematic drinking: A review of the literature. Clinical Psychology Review, 23, 719-759. doi:10.1016/S0272-7358(03)00071-0 Holden, R.R., & DeLisle, M.M. (2005). Factor analysis of the beck suicide scale for suicidal ideation with female suicide atempters. Asesment, 12, 232-238. Hufford, M. R. (2001). Alcohol and suicidal behavior. Clinical Psychology Review, 21, 797?811. doi: 10.1016/S0272-7358(00)00070-2 Joiner, T.E., Jr., (2005). Why people die by suicide. Cambridge, MA: Harvard University Pres. Joiner, T.E., Jr., Rudd, M.D., & Rajab, M.H., (1997). The modified scale for suicidal ideation: Factors of suicidality and their relation to clinical and diagnostic variables. Journal of Abnormal Psychology, 106, 260-265. Joiner, T. E., Jr., Van Orden, K. A., Wite, T. K., Selby, E. A., Ribeiro, J., Lewis, R., & Rudd, M. D. (2009). Main predictions of the interpersonal- psychological theory of suicidal behavior: Empirical tests in two samples of young adults. Journal of Abnormal Psychology, 118, 634-646. doi:10.1037/a0016500 Kahler, C.W., Strong, D.R., & Read, J.P. (2005). Toward eficient and comprehensive measurement of the alcohol problems continuum in college students: The brief young adult alcohol consequences questionnaire. Alcoholism: Clinical and Experimental Research, 29, (7), 1180-1189. Kahler, C. W., Strong, D. R., Read, J. P., Palfai, T. P., & Wood, M. D. (2004). Mapping the ! 41! continuum of alcohol problems in college students: a Rasch model analysis. Psychology of Addictive Behaviors: Journal of the Society of Psychologists in Addictive Behaviors, 18, 4, 322-33. doi: 10.1037/0893-164X.18.4.322 Kasel, J.D., Jackson, S.I., & Unrod, M. (2000). Generalized expectancies for negative mood regulation and problem drinking among college students. Journal of Studies on Alcohol, 61 (2), 332-340. Kely, T. M., Lynch, K. G., Donovan, J. E., & Clark, D. B. (2001). Alcohol Use Disorders and Risk Factor Interactions for Adolescent Suicidal Ideation and Atempts. Suicide and Life- Threatening Behavior, 31, 2, 181. doi:10.1521/suli.31.2.181.21512 Kendal, R.E., (1983). Alcohol and suicide. Substance & Alcohol Actions/ Misuse, 4, 121-127. Kesler, R. C., Borges, G., & Walters, E. E. (1999). Prevalence of and risk factors for lifetime suicide atempts in the National Comorbidity Survey. Archives of General Psychiatry, 56, 7, 617-26. doi:10.1001/archpsyc.56.7.617 Klatsky, A. L., & Armstrong, M. A. (1993). Alcohol use, other traits, and risk of unnatural death: a prospective study. Alcoholism, Clinical and Experimental Research, 17, 6, 1156-62. doi: 10.1111/j.1530-0277.1993.tb05221.x Kuntsche, E., Knibbe, R., Gmel, G. & Engels, R. (2005). Why do young people drink? A review of drinking motives. Clinical Psychology Review, 25, 841-861. doi:10.1016/j.cpr.2005.06.002 Labouvie, E., & Bates, M.E., (2002). Reasons for alcohol use in young adulthood: Validation of a three dimensional measure. Journal of Studies on Alcohol, 63 (2), 145-155. Lamis, D. A., & Malone, P. S. (2011). Alcohol-related problems and risk of suicide among college students: the mediating roles of belongingnes and burdensomenes. Suicide & ! 42! Life-Threatening Behavior, 41, 5, 543-53. doi: 10.1111/j.1943-278X.2011.00052.x Lamis, D.A., & Malone, P.S., (2012). Alcohol use and suicidal behaviors among adults: A synthesis and theoretical model. Suicidology Online, 3, 4-23. Lamis, D.A., Malone, P.S., & Jahn, D.R., (in pres). Alcohol use and suicide pronenes in college students: a proposed model. Mental Health and Substance Use. Lester, D. (1992). Why people kil themselves: A 1990s summary of research findings on suicidal behavior (3 rd . ed.) Springfield, IL: Charles C. Thomas. Lewis, B. A., & O'Neil, H. K. (2000). Alcohol expectancies and social deficits relating to problem drinking among college students. Addictive Behaviors, 25, 2, 295-299.  doi:10.1016/S0306-4603(99)00063-5 Light, J. M., Grube, J. W., Madden, P. A., & Gover, J. (2003). Adolescent alcohol use and suicidal ideation: a nonrecursive model. Addictive Behaviors, 28, 4, 705-24. doi:10.1016/S0306-4603(01)00270-2 Linehan, M. M., Goodstein, J. L., Nielsen, S. L., & Chiles, J. A. (1983). Reasons for staying alive when you're thinking of kiling yourself: The reasons for living inventory. Journal of Consulting and Clinical Psychology, 51, 276-286. Malone, K.M., Oquendo, M.A., Haas, G.L., Elis, S.P., Li, S. & Mann, J.J. (2000). Protective factors against suicidal acts in major depresion: reasons for living. American Journal of Psychiatry, 157, 1084-1088. doi: 10.1176/appi.ajp.157.7.1084 Marshal, M. P. (2003). For beter or for worse? The efects of alcohol use on marital functioning. Clinical Psychology Review, 23, 7, 959-97. doi:10.1016/j.cpr.2003.09.002 McIntosh, J. L. (for the American Asociation of Suicidology). (2012). U.S.A. suicide: 2009 ! 43! official final data. Washington, DC: American Asociation of Suicidology, dated January 12, 2012, downloaded from http:/ww.suicidology.org McLeod, R., Stockwel, T., Stevens, M., & Philips, M. (1999). The relationship betwen alcohol consumption paterns and injury. Addiction, 94, 11, 1719-1734. doi:10.1080/09652149932497 McNaly, A.M., Palfai,T.P., Levine, R.V., & Moore, B.M., (2003). Atachment dimensions and drinking-related problems among young adults: The meditational role of coping motives. Addictive Behaviors, 28, 1115-1127. doi: 10.1016/S0306-4603(02)00224-1 Meade, A. W., & Craig, S. B. (2011, April). Identifying careles responses in survey data. Paper presented at the 26th annual conference of the Society for Industrial and Organizational Psychology, Chicago, IL. Meng, X.L., Rosenthal, R., & Rubin, D.B. (1992). Comparing correlated correlation coeficients. Psychological Bulletin, 111, 1, 172-175. doi: 10.1037/0033-2909.111.1.172 Merril, J. E., & Read, J. P. (2010). Motivational Pathways to Unique Types of Alcohol Consequences. Psychology of Addictive Behaviors, 24, 4, 705-711. doi: 10.1037/a0020135 Miler, L., McIntire, S., & Lovler, R. (2011). Foundations of psychological testing: A practical approach. Thousand Oaks, CA: Sage Publications Inc. Muthen, L. K., & Muthen, B. O. (1998-2012). Mplus User?s Guide: Seventh Edition. Los Angeles, CA: Muthen & Muthen. Nakaya, N., Kikuchi, N., Shimazu, T., Ohmori, K., Kakizaki, M., Sone, T., Awata, S., & Tsuji, I. (2007). Alcohol consumption and suicide mortality among Japanese men: the Ohsaki Study. Alcohol, 41, 7, 503-510. doi: 10.1016/j.alcohol.2007.08.001 ! 44! Orbach,I., Palgi, Y., Stein, D., HarEven, D., Lotem-Peleg, M., Asherov, J., & Elizur, A. (1996). Tolerance for physical pain in suicidal subjects. Death Studies, 20, 327-341. Orbach, I., Stein, D., Palgi, Y., Asherov, J., HarEven, D. & Elizur, A. (1996). Perception of physical pain in acident and suicide atempt patients: Self-preservation versus self- destruction. Journal of Psychiatric Research, 30, 307-320. doi:10.1016/0022-3956(96)00008-8 Perkins, H. W. (2002). Surveying the damage: A review of research on consequences of alcohol misuse in college populations. Journal of Studies on Alcohol. Supplement, 14, 91-100. Pfaf, J. J., Almeida, O. P., Wite, T. K., Waesche, M. C., & Joiner, T. E. (2007). Relationship betwen Quantity and Frequency of Alcohol Use and Indices of Suicidal Behavior in an Elderly Australian Sample. Suicide and Life-Threatening Behavior, 37, 6, 616-626. doi:10.1521/suli.2007.37.6.616 Pirkola, S.P., Isometsa, E.T., Heikkinen, M.E., & Lonnqvist, J.K. (2000). Suicides of alcohol misusers and non-misusers in a nationwide population. Alcohol & Alcoholism, 35, 70-75. doi: 10.1093/alcalc/35.1.70 Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for asesing and comparing indirect efects in multiple mediator models. Behavior Research Methods, 40, 879-891. doi:10.3758/BRM.40.3.879 Preuss, U.W., Schuckit, M.A., Smith, T.L., Danko, G.P., Buckman, K., Bierut, L., Bucholz, K.K., Heselbrock, M.N., Heselbrock, V.M., & Reich, T. (2002). Comparison of 3190 alcohol dependent individuals with and without suicide atempts. Alcoholism: Clinical and Experimental Research, 26, 471-477. doi: 0145-6008/02/2604-0471 Radloff, L. S. (1977). The CES-D Scale: A Self-Report Depresion Scale for Research in the ! 45! General Population. Applied Psychological Measurement, 1, 3, 385-401. doi:10.1177/014662167700100306 Rasch, G. (1960). Probablistic Models for Some Inteligence and Atainment Tests. Denmark?s Paedagogiske Institute, Copenhagen. Read, J. P., Kahler, C. W., Strong, D. R., & Colder, C. R. (2006). Development and preliminary validation of the young adult alcohol consequences questionnaire. Journal of Studies on Alcohol, 67, 1, 169-77. Read, J. P., Merril, J. E., Kahler, C. W., & Strong, D. R., (2007). Predicting functional outcomes among college drinkers: Reliability and predictive validity of the Young Adult Alcohol Consequences Questionnaire. Addictive Behaviors, 32, 11, 2597-2610. doi:10.1016/j.addbeh.2007.06.021 Reifman, A., & Windle, M. (1995). Adolescent suicidal behaviors as a function of depresion, hopelesnes, alcohol use, and social support: a longitudinal investigation. American Journal of Community Psychology, 23, 3, 329-54. doi: 10.1007/BF02506948 Ribeiro, J.D., Wite, T.K., Van Orden, K. A., Selby, E. A., Gordon, K. H., Bender, T.W., & Joiner, T. E. (2014). Fearlesnes about death: The psychometric properties and construct validity of the revision to the Acquired Capability for Suicide Scale. Psychological Asesment, 26, 115-126. doi: 10.1037/a0034858 Rohde, P., Lewinsohn, P. M., Langhinrichsen-Rohling, J., & Langford, R. (2004). Life atitudes schedule: short form (LAS-SF): A risk asesment for suicidal and life-threatening behaviors [Technical Manual]. Toronto, ON: Multi-Health Systems. Rohde, P., Seley, J. R., Langhinrichsen-Rohling, J., & Rohling, M. L. (2003). The life atitudes schedule-short form: psychometric properties and correlates of adolescent suicide ! 46! pronenes. Suicide & Life-Threatening Behavior, 33, 3, 249-60. doi: 10.1521/suli.33.3.249.23216 Rudd, M.D., Joiner, T., & Rajab, M.H. (1996). Relationships among suicide ideators, atempters, and multiple atempters in young-adult sample. Journal of Abnormal Psychology, 105, 541-550. doi:10.1037/0021-843X.105.4.541 Schmidt, N.B., Richey, J.A., & Fitzpatrick, K.K., (2006). Discomfort intolerance: Development of a construct and measure relevant to panic disorder. Journal of Anxiety Disorders, 20, 263-280. doi:10.1016/j.janxdis.2005.02.002 Schneider, B., Baumert, J., Schneider, A., Meisinger, C., Ladwig, K.-H., Marten-Mitag, B., Erazo, N., & Hamer, G. P. (2011). The efect of risky alcohol use and smoking on suicide risk: Findings from the German MONICA/KORA-Augsburg Cohort Study. Social Psychiatry and Psychiatric Epidemiology, 46, 11, 1127-1132. doi: 10.1007/s00127-010-0287-y Sher, L. (2006). Alcohol consumption and suicide. Qjm: an International Journal of Medicine, 99, 1, 57-61. doi:10.1093/qjmed/hci146 Shneidman, E. S. (1998). Perspectives on suicidology: Further reflections on suicide and psychache. Suicide and Life-Threatening Behavior, 28, 245?250. Smith, P. N., & Cukrowicz, K. C. (2010). Capable of suicide: a functional model of the acquired capability component of the Interpersonal-Psychological Theory of Suicide. Suicide & Life-Threatening Behavior, 40, 3, 266-75. doi:10.1521/suli.2010.40.3.266 Sobel, L.C., & Sobel, M.B. (1992) Timeline Followback: A technique for assesing self- ! 47! reported alcohol consumption. In: Liten, R.Z. and Alen J., eds. Measuring Alcohol Consumption: Psychosocial and Biological Methods. New Jersey: Humana Pres, 1992. Pp. 41-72. Sobel, L. C. & Sobel, M. B. (1995). Alcohol consumption measures.(In J. P. Alen & M. Columbus (Eds.), Assesing alcohol problems (pp. 55-74). Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism Steer, R.A., Rismiler, D.J., Ranieri, W.F., & Beck, A.T. (1993). Dimensions of suicidal ideation in psychiatric inpatients. Behavior Research and Therapy, 229-239. Stephenson, H., Pena-Shaf, J., & Quirk, P. (2006). Predictors of College Student Suicidal Ideation: Gender Diferences. College Student Journal, 40, 1, 109-117. Van Orden, K. A., Cukrowicz, K. C., Wite, T. K., & Joiner, T. E. (2012). Thwarted belongingnes and perceived burdensomenes: Construct validity and psychometric properties of the Interpersonal Neds Questionnaire. Psychological Asesment, 24(1), 197-215. doi:10.1037/a0025358 Van Orden, K. A., Wite, T. K., Cukrowicz, K. C., Braithwaite, S. R., Selby, E. A., & Joiner Jr., T. E. (2010). The interpersonal theory of suicide. Psychological Review, 117, 575?600. doi: 10.1037/a0018697 Van Orden, K. A., Wite, T. K., Gordon, K. H., Bender, T. W., & Joiner, T. E., Jr. (2008). Suicidal desire and the capability for suicide: Tests of the interpersonal-psychological theory of suicidal behavior among adults. Journal of Consulting and Clinical Psychology, 76, 72?83. doi: 10.1037/0022-006X.76.1.72 Vik, P. W., Carrelo, P., Tate, S. R., & Field, C. (2000). Progresion of consequences among ! 48! heavy-drinking college students. Psychology of Addictive Behaviors: Journal of the Society of Psychologists in Addictive Behaviors, 14, 2, 91-101. doi: 10.1037//0893164X.14.2.91 White, H. R., & Labouvie, E. W. (1989). Towards the asesment of adolescent problem drinking. Journal of Studies on Alcohol, 50, 1, 30-7. Wilcox, H. C., Conner, K. R., & Caine, E. D. (2004). Asociation of alcohol and drug use disorders and completed suicide: an empirical review of cohort studies. Drug and Alcohol Dependence, 76. doi:10.1016/j.drugalcdep.2004.08.003 Wite, T.K., Joiner, T.E., Jr., Brown, G.K., Beck, A.T., Beckman, A., Duberstein, P., & Yeates, C. (2006). Factors of suicidal ideation and their relation to clinical and other indicators in older adults. Journal of Afective Disorders, 94, 165-172. Yoshimasu, K., Kiyohara, C., Miyashita, K., & Stres Research Group of the Japanese Society for Hygiene. (2008). Suicidal risk factors and completed suicide: meta-analyses based on psychological autopsy studies. Environmental Health and Preventive Medicine, 13, 5, 243-56. doi:10.1007/s12199-008-0037-x You, S., Van Orden, K. A., & Conner, K. R. (2011). Social Connections and Suicidal Thoughts and Behavior. Psychology of Addictive Behaviors, 25, 1, 180-184. doi: 10.1037/a0020936 Zimerman, M. A., Copeland, L.A., Shope, J.T., & Dielman, T.E., (1997). A Longitudinal Study of Self-Estem: Implications for Adolescent Development. Journal of Youth and Adolescence, 26, 2, 117-41. doi:10.1023/A:1024596313925 ! 49! Apendix A: Tables ! 50! Table A1 Descriptive Statistics for al Variables for Regular Drinking Sample. Mean Std. Dev. Min. Max. Skew Kurtosis Total N Variable 295 Age 20.60 1.68 19.00 31.00 1.97 6.78 293 Year in school 2.91 1.16 1.00 5.00 -0.02 -0.90 295 Total drinks 29.41 26.18 4.00 115.28 1.67 2.70 279 Alcohol-related problems (YACQ, Frequency scale) 33.49 27.31 0.00 90.00 0.87 -0.38 279 SOC 4.99 3.93 0.00 22.00 1.19 1.55 279 SELF-P 2.54 2.94 0.00 13.00 1.41 1.62 280 SELF-C 4.36 5.34 0.00 31.00 1.93 4.76 280 ACC-OCC 2.15 2.68 0.00 12.00 1.64 2.75 280 RISK 3.99 4.27 0.00 21.00 1.43 0.15 280 BLKOUT 7.30 5.86 0.00 28.00 0.98 0.44 280 PHYS-DEP 1.18 1.88 0.00 10.00 2.38 6.18 280 CONTR 4.44 4.35 0.00 24.00 1.50 2.45 280 B-YACQ 11.25 6.35 0.00 24.00 0.18 -0.97 280 TB 0.98 1.02 0.00 4.07 1.22 0.77 274 PB 0.20 0.51 0.00 2.14 2.91 7.61 275 FAD 21.48 6.44 7.00 35.00 0.03 -0.44 276 Pain tolerance 29.34 6.87 9.00 49.00 -0.18 -0.07 274 P-SI 0.19 0.54 0.00 2.63 3.05 8.82 260 Coping motives 8.79 3.64 5.00 20.43 1.33 1.52 283 Conformity motives 6.98 2.50 5.00 14.73 1.35 1.04 283 Social motives 16.52 4.59 5.00 25.00 -0.07 -0.91 283 Enhancement motives 13.92 4.66 5.00 25.00 0.29 -0.65 283 Sadnes 0.69 0.77 0.00 3.00 1.04 0.84 264 Note. YAACQ= Young Adult Alcohol Consequences Questionaire; SOC= social problems subscale; SELF-P= self-perception subscale; SELF- C= self-care subscale; AC-OCC= academic/ ocupational subscale; RISK= risky behavior subscale; BLKOUT= blackout subscale; PHYS-DEP= physical-dependence subscale; CONTR=impaired control subscale; B-YAACQ= Brief Young Adult Alcohol Consequences Questionaire; FAD= fearlesnes about death; TB= thwarted belongingnes; PB= perceived burdensomenes; P-SI= pasive suicide ideation. ! 51! Table A2 Zero-Order Corelations betwen Study Variables in Regular Drinking Sample. Note. B-YAACQ= Brief Young Adult Alcohol Consequences Questionaire; TB=thwarted belongingnes; PB= perceived burdensomenes; FAD= fearlesnes about death; P-SI= pasive suicide ideation; YAACQ= Young Adult Alcohol Consequences Questionaire; BLKOUT= blackout subscale; RISK= risky behavior subscale; SELF-C= self-care subscale; SELF-P= self-perception subscale; AC-OCC= academic/ ocupational subscale; PHYS- DEP= physical-dependence subscale; SOC= social problems subscale; CONTR=impaired control subscale. *= p<.05. Measure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 1. Gender 1 2. Conformity Motives -.04 1 3. Coping Motives -.06 .44* 1 4. Social otives -.08 .42* .60* 1 5. Enhancement Motives -.20* .22* .53* .68* 1 6. Sadnes .04 .29* .37* .24* .18* 1 7. Total Drinks -.32* .16* .33* .34* .38* .06 1 8. B-YAACQ -.09 .41* .53* .46* .40* .31* .38* 1 9. TB -.14* .20* .26* .15* .07 .42* .01 .18* 1 10. PB -.05 .20* .17* .09 .07 .48* -.01 .15* .55* 1 11. Pain Tolerance -.33* -.12 -.05 -.02 .05 -.11 .03 -.07 .05 -.07 1 12. FAD -.30* -.14* -.12* -.10 -.05 -.05 -.05 -.10 .11 .01 .32* 1 13. P-SI -.13* .07 .10 .06 .03 .32* -.03 .10 .40* .42* .01 .12 1 14. YACQ-BLKOUT -.13* .33* .40* .48* .41* .21* .42* .82* .09 .10 .04 -.01 .07 1 15. YACQ-RISK -.15* .40* .45* .44* .41* .24* .33* .86* .14* .17* .03 -.06 .04 .81* 1 16. YACQ-SELF-C -.07 .38* .50* .40* .33* .25* .36* .81* .15* .10 -.04 -.04 .09 .69* .75* 1 17. YACQ-SELF-P -.05 .40* .37* .24* .15* .28* .16* .74* .22* .15* -.12 -.11 .10 .59* .69* .74* 1 18. YACQ-ACC-OCC -.02 .35* .50* .37* .32* .27* .30* .83* .15* .15* -.08 -.08 .12 .69* .74* .79* .67* 1 19. YACQ-PHYS-DEP -.16* .23* .44* .36* .35* .20* .35* .74* .20* .15* -.01 -.01 .06 .65* .67* .57* .52* .59* 1 20. YACQ-SOC -.05 .38* .43* .46* .37* .24* .29* .82* .15* .13* -.01 -.01 .03 .80* .82* .71* .72* .70* .62* 1 21. YACQ-CONTR -.06 .39* .52* .42* .38* .26* .35* .84* .15* .16* -.02 -.02 .06 .77* .80* .81* .74* .73* .66* .78* 1 ! 52! Table A3 Standardized and Unstandardized Parameter Estimates for Path Model 1. ! ! Parameter Estimated B (SE) t ? p P-SI on TB 0.12 0.05 2.53 0.24 <.01 PB 0.29 0.12 2.50 0.28 <.01 Alcohol-related Problems 0.00 0.00 -0.12 -0.01 .90 Gender -0.08 0.06 -1.30 -0.08 .19 PB on Alcohol-related Problems 0.01 0.01 1.76 0.12 .07 Gender -0.03 0.06 -0.43 -0.03 .67 TB on Alcohol-related Problems 0.02 0.01 2.34 0.15 .02 Gender -0.25 0.12 -2.07 -0.12 .04 FAD on Alcohol-related Problems -0.15 0.06 -2.55 -0.14 .01 Gender -4.22 0.75 -5.61 -0.33 .00 Pain Tolerance on Alcohol-related Problems -0.10 0.06 -1.56 -0.09 .12 Gender -4.70 0.79 -5.98 -0.34 <.01 PB with TB 0.27 0.06 4.92 0.54 <.01 FAD 0.04 0.22 0.16 0.01 .87 Pain Tolerance -0.24 0.19 -1.28 -0.07 .19 TB with FAD 0.58 0.39 1.47 0.10 .13 Pain Tolerance 0.11 0.36 0.30 0.02 .77 FAD with Pain Tolerance 9.28 2.56 3.62 0.24 <.01 Alcohol-related Problems with Gender -0.53 0.18 -2.88 -0.17 <.01 Note. Gender coded as male=0, female =1. P-SI= pasive suicide ideation. TB= thwarted belongingnes; PB= perceived burdensomenes; FAD= fearlesnes about death. ! 53! Table A4 Indirect Efects in Path Model 1. Efect Tested Estimate (95% CI) Standardized Estimate (95% CI) p Alcohol related problems on P-SI through TB 0.00 (0.00-0.01) 0.04 (-0.00-0.07) .05 Alcohol related problems on P-SI ideation through PB 0.00 (0.00-0.01) 0.03(-0.02-0.08) .15 Note. P-SI= pasive suicide ideation. TB= thwarted belongingnes; PB= perceived burdensomenes. ! 54! Table A5 Standardized and Unstandardized Parameter Estimates for Path Model 2. (continued) Parameter Estimated B (SE) t ? p P-SI on TB 0.11 0.05 2.33 0.21 <.01 PB 0.24 0.12 2.06 0.23 .03 Alcohol-related Problems 0.00 0.01 -0.03 0.00 .98 Gender -0.10 0.07 -1.51 -0.09 .13 Sadnes 0.10 0.05 2.04 0.14 .04 Coping Motives -0.01 0.01 -0.63 -0.05 .53 PB on Alcohol-related Problems 0.00 0.01 -0.18 -0.01 .85 Gender -0.07 0.06 -1.20 -0.06 .23 Sadnes 0.33 0.06 5.54 0.49 <.01 Coping Motives 0.00 0.01 -0.07 -0.01 .95 TB on Alcohol-related Problems 0.00 0.01 -0.28 -0.02 0.78 Gender -0.32 0.11 -2.94 -0.16 <.01 Sadnes 0.51 0.09 5.44 0.38 <.01 Coping Motives 0.03 0.02 1.47 0.11 .15 FAD on Alcohol-related Problems -0.10 0.07 -1.43 -0.10 .15 Gender -4.18 -0.74 -5.70 -0.32 <.01 Sadnes 0.11 0.59 0.19 0.01 .85 Coping Motives -0.14 0.12 -1.19 -0.08 .24 Pain Tolerance on Alcohol-related Problems -0.07 0.07 -0.10 -0.07 .32 Gender -4.61 0.79 -5.87 -0.32 <.01 Sadnes -0.67 0.62 -1.08 -0.07 .85 Coping Motives 0.00 0.13 -0.03 0.00 .24 PB with TB 0.18 0.04 4.75 0.43 <.01 FAD 0.04 0.18 0.21 0.01 .84 Pain Tolerance -0.13 0.17 -0.77 -0.04 .44 TB with FAD 0.61 0.33 1.88 0.11 .06 Pain Tolerance 0.30 0.33 0.91 0.05 .37 FAD with Pain Tolerance 9.28 2.54 3.66 0.24 <.01 Alcohol-related Problems with Gender -0.52 0.18 -2.82 -0.17 <.01 ! 55! Parameter Estimated B (SE) t ? p Sadnes 1.29 0.31 4.16 0.27 <.01 Coping Motives 12.71 1.46 8.73 0.55 <.01 Gender with Sadnes 0.01 0.02 0.50 0.03 .62 Coping Motives -0.14 0.11 -1.29 -0.07 .20 Sadnes with Coping Motives 0.99 0.22 4.52 0.36 <.01 Note. Gender coded as male=0, female=1. P-SI= pasive suicide ideation; TB= thwarted belongingnes; PB= perceived burdensomenes; FAD= fearlesnes about death. ! 56! Table A6 Indirect Efects in Path Model 2. Efect Tested Estimate (95% CI) Standardized Estimate (95% CI) p Alcohol related problems on P-SI through TB 0.00 (0.00-0.00) 0.00 (-0.07-0.05) .78 Alcohol related problems on P-SI through PB 0.00 (0.00-0.00) 0.00 (-0.04-0.03) .85 Coping motives on P-SI through TB 0.00 (0.00-0.01) 0.02 (-0.01-0.06) .17 Coping motives on P-SI through PB 0.00 (-0.01-0.01) 0.00 (-0.04-0.04) .95 Note. P-SI= pasive suicide ideation. TB= thwarted belongingnes; PB= perceived burdensomenes. ! 57! Table A7 Standardized and Unstandardized Parameter Estimates for Path Model 3. Parameter Estimated B (SE) t ? p P-SI on TB 0.11 0.05 2.31 0.21 <.01 PB 0.25 0.12 2.10 0.24 .03 Sadnes 0.09 0.05 1.75 0.12 .08 Gender -0.09 0.06 -1.45 -0.09 .15 PB on Sadnes 0.32 0.06 5.65 0.48 <.01 Gender -0.06 0.05 -1.12 -0.06 .26 TB on Sadnes 0.56 0.09 6.48 0.42 <.01 Gender -0.33 0.11 -2.96 -0.16 <.01 FAD on Sadnes -0.37 0.56 -0.66 -0.04 .51 Gender -3.87 0.73 -5.29 -0.30 <.01 Pain Tolerance on Sadnes -0.84 0.60 -1.41 -0.09 .16 Gender -4.44 0.77 -5.78 -0.32 <.01 PB with TB 0.18 0.04 4.81 0.43 <.01 FAD 0.05 0.18 0.27 0.02 .79 Pain Tolerance -0.12 0.16 -0.77 -0.04 .45 TB with FAD 0.55 0.34 1.66 0.10 .09 Pain Tolerance 0.28 0.33 0.86 0.05 .39 FAD with Pain Tolerance 9.63 2.57 3.75 0.24 <.01 Gender with Sadnes 0.01 0.02 0.44 0.03 .66 Note. Gender coded as male=0, female =1. P-SI= pasive suicide ideation. TB= thwarted belongingnes; PB= perceived burdensomenes; FAD= fearlesnes about death. ! 58! Apendix B: Figure ! 59! ! 60! ! 61!