Substance-Induced Depression and Independent Depression in ...

CONNER ET AL.

567

Substance-Induced Depression and Independent Depression in Proximal Risk for Suicidal Behavior KENNETH R. CONNER, PSY.D., M.P.H.,a,b,* STEPHANIE A. GAMBLE, PH.D.,a COURTNEY L. BAGGE, PH.D.,c HUA HE, PH.D.,a MARC T. SWOGGER, PH.D.,a ARTHUR WATTS, B.S.,a AND REBECCA J. HOUSTON, PH.D.d aUniversity

of Rochester Medical Center, Rochester, New York 2 Center of Excellence for Suicide Prevention, Canandaigua Veterans Affairs Medical Center, Canandaigua, New

bVeterans Affairs VISN

York cUniversity dResearch

of Mississippi Medical Center, Jackson, Mississippi Institute on Addictions, University at Buffalo, The State University of New York, Buffalo, New York

ABSTRACT. Objective: Major depressive episodes may be substance induced or occur independent of substance use. Studies of the roles of substance-induced depression (SID) and independent depression (IND) in suicidal behavior are limited to retrospective reports. The purpose of this study was to examine proximal (i.e., acute) risk for suicide attempts associated with SID and IND. Method: Individuals who had attempted suicide (n = 100) and nonsuicidal controls (n = 100) matched for site were recruited from residential substance use treatment programs. Participants were ages 18 and older and screened positive for potential alcohol use disorder. Validated semistructured interviews were used to assess SID, IND, and suicide attempts. Analyses of individual-level risk for attempts were based on multivariate logistic regression that adjusted for risk factors. Population-level attributable risk (PAR) frac-

tions for suicide attempts were also calculated to provide estimates of the percentage of attempts in the study population attributable to SID and IND, respectively. Results: SID was identified in 60% of attempters and 35% of controls and IND in 13% of attempters and 3% of controls. Both variables conferred risk for suicide attempt (SID: odds ratio [OR] = 3.73, 95% CI [1.84, 7.58]; IND: OR = 10.38, 95% CI [2.48, 43.49]. PAR for suicide attempts associated with SID and IND was 0.44 and 0.12, respectively. Conclusions: Both SID and IND confer proximal risk for suicide attempts after adjusting for other risk factors. SID also contributes substantial risk in this population overall. Future research should test the hypothesis that IND confers greater risk than SID at the individual level. (J. Stud. Alcohol Drugs, 75, 567–572, 2014)

A

havior among individuals with AUD (Conner et al., 2003; Murphy et al., 1992). Retrospective studies have examined SID and IND and lifetime suicide attempts among individuals with AUD (Preuss et al., 2002a, 2002b; Schuckit et al., 1997) and other substance use disorders (Aharonovich et al., 2002). These reports have shown that both depression subtypes are associated with increased risk for lifetime history of suicide attempt (Aharonovich et al., 2002; Preuss et al., 2002a). In an AUD sample, IND was associated with greater risk for a lifetime suicide attempt compared with SID (Schuckit et al., 1997), a result not shown in the substance use disorders sample (Aharonovich et al., 2002). In terms of characteristics of suicide attempts, IND was associated with greater suicide intent (i.e., intention to die) in the sample of substance use disorders (Aharonovich et al., 2002) but not in an AUD sample (Preuss et al., 2002b). An analysis of individuals with AUD also reported no difference in planning (i.e., forethought) of suicide attempts between IND and SID (Preuss et al., 2002b). Clarifying the role of the depression subtypes in the planning and intent of suicidal behavior is important because incidence of more planned acts of suicide tends to show higher intent and is generally more deadly (Conner, 2004). Overall, studies of depression subtypes in suicide attempts suggest that both SID and IND are associated with suicide attempts. There is also some support that IND is as-

LCOHOL USE DISORDER (AUD) and major depression commonly co-occur (Hasin et al., 2007; Kessler et al., 1997), particularly in clinical populations (Hesselbrock et al., 1985; Salloum et al., 1995). Among individuals with AUD, depressive episodes may occur in the context of substance use, referred to as substance-induced depression (SID), or during periods of abstinence or limited substance use, referred to as independent depression (IND) (Schuckit et al., 2007). Research highlights differences in SID and IND epidemiology (e.g., gender differences) and course (e.g., time to remission with abstinence) that suggest the value of tailored prevention and treatment strategies (Schuckit, 2006). There are few data on SID and IND and suicidal behavior, yet this is a critical issue because co-occurring AUD and major depression is the most common comorbidity among suicide decedents (Cavanagh et al., 2003; Yoshimasu et al., 2008), and major depressive episodes are a potent proximal risk factor for suicidal be-

Received: October 22, 2013. Revision: January 30, 2014. This study was supported by National Institute on Alcohol Abuse and Alcoholism Grants R01 AA016149 and K23 AA017246. *Correspondence may be sent to Kenneth R. Conner at the University of Rochester Medical Center, Psychiatry, 300 Crittenden Blvd, Rochester, NY 14642, or via email at: [email protected].

567

568

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / JULY 2014

sociated with greater risk for lifetime suicide attempts and attempts with greater suicide intent (but not planning). The aforementioned studies were limited by retrospective recall, reliance on single items to assess suicide attempt planning and intent, and uncertainty about the proximity between prior depressive episodes and suicide attempts. Most data on the proximal relationship between depressive episodes and suicidal behavior come from psychological autopsy studies (Cavanagh et al., 2003; Yoshimasu et al., 2008). However, the unavailability of people for direct interview in these studies makes the studies poorly suited to answer finer grained questions such as those regarding the role of SID and IND (Conner et al., 2011). Indeed, reliable subtyping of depression is challenging even with direct interviews of individuals (Hasin et al., 1998, 2006). Recent research has identified the utility of including population attributable risk (PAR) fractions in studies of suicidal behavior (Li et al., 2011; Tidemalm et al., 2008). PAR is a measure of the proportion of individuals in a population who experience a deleterious outcome because of a risk factor. PAR is important to consider when designing prevention strategies because it provides an estimate of how much an outcome (e.g., suicide) may be expected to be reduced by preventing or eliminating a given risk factor. PAR takes into consideration both the relative risk conferred by a variable and its prevalence in a population. For example, a meta-analysis examining PAR for suicide associated with mental disorders and low education concluded that these variables have similar PARs despite marked differences in prevalence (lower for mental disorders) and relative risk (higher for mental disorders) (Li et al., 2011). Prior research indicates that SID is more common than IND among suicide-attempting individuals with AUD (Preuss et al., 2002b) and among individuals with AUD generally (Schuckit et al., 1997). Preliminary data also suggest that IND confers greater relative risk for suicide attempts compared with SID among individuals with AUD (Schuckit et al., 1997). Such data make it unclear which depression subtype is associated with greater PAR for suicidal behavior. The purpose of this study was to examine SID and IND in proximal risk for suicide attempts among individuals with substance use disorders, most with AUD. We analyzed risk for an attempt associated with each subtype of depression and compared the degree of risk conferred by each subtype. We hypothesized that both subtypes of depression confer increased risk for a suicide attempt and that IND confers greater risk than SID. We further examined associations of the depression subtypes with the degree of planning and intent of the suicide attempts and hypothesized that IND is associated with greater intention to die. We also calculated PAR for suicide attempts associated with IND and SID.

Method Procedure Study procedures have been described elsewhere (Conner et al., 2012) and were approved by the research review boards of each of the participating institutions. Briefly, people who attempted suicide (referred to hereafter as “cases”) were recruited from residential substance use disorder treatment programs if they reported a suicide attempt within 90 days of admission and scored an 8 or higher on the Alcohol Use Disorders Identification Test (AUDIT; Allen et al., 1997). For each case, a control subject at the same treatment site with AUDIT score of 8 or more was recruited. Participants who were controls had no self-reported history of suicidal thoughts or attempts. We did not match controls to cases on age, sex, or other variables, given potential biases associated with such matching (Rothman and Greenland, 1998; Vandenbroucke et al., 2007). Measures Outcomes. All questions were assessed by interview. Suicide attempts were identified with the item, “Have you ever tried to kill yourself or attempt suicide?” which has shown high test–retest reliability, κ = .82 (Conner et al., 2007). The attempts were dated with use of a calendar. The assessment of suicide planning was based on a fouritem scale with a possible range of 0–15 that showed solid test–retest reliability (r = .64, p < .001) and internal consistency (α = .84) (Conner et al., 2012). An example item is, “How long did you think about it before you made the suicide attempt?” Subjective intent of suicide attempts was based on six items with a possible range of 0–12 from the Suicide Intent Scale (Beck et al., 1974). The scale showed high internal consistency (α = .86). An example item is, “When you did this, how likely did you think it was that you would die?” Depression subtypes. The Psychiatric Research Interview for Substance and Mental Disorders (PRISM) is a semistructured, diagnostic interview that measures the major Axis I Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994), diagnoses of alcohol, other drug, and psychiatric disorders (Hasin et al., 1998, 2006). It reliably distinguishes between depressive symptoms that are the expected result of intoxication or withdrawal from those that constitute major depression and allows for further classification of major depressive episodes into SID and IND subtypes. In the current study, select portions of the PRISM were administered to assess major depressive disorder, to establish SID or IND for current depressive episodes, and to assess symptoms of alcohol dependence. Diagnoses were considered current if they occurred during the 90-day period before treatment

CONNER ET AL. admission, comparable to the period during which suicide attempts were studied. Covariates. We selected covariates of age, sex, alcoholrelated severity, other drug–related severity, aggression, child sexual abuse history, and major interpersonal life events. These variables were chosen because they have consistently been shown to be associated with suicidal behavior in individuals with AUD and other substance use populations (Conner and Ilgen, 2011). Alcohol-related severity was assessed with a composite measure (Conner et al., 2012) that was created by combining assessments of (a) alcohol use based on AUDIT items 1–3, (b) alcohol-related difficulties based on AUDIT items 4–10, and (c) alcohol dependence symptom count using DSM-IV criteria assessed with the PRISM (α = .74). Other drug–related severity was assessed with a 10-item version of the Drug Abuse Screening Test (Cocco and Carey, 1998) (α = .87). Aggression was assessed with the five-item subscale of the Lifetime History of Aggression Questionnaire (Coccaro et al., 1997) (α = .76). Child sexual abuse history (presence/absence) was assessed using two standard screening items (Dube et al., 2003). A reliable, semistructured stressful life-event interview was used to assess events occurring during the 90-day pre-admission period (Conner et al., 2012). Specifically, we covaried major interpersonal events in the analysis because these events alone conferred risk for suicide attempt in a prior analysis of this sample (Conner et al., 2012). Analyses. One case and corresponding control were dropped because of missing data, yielding a sample of 100 cases and 100 controls. We examined the association of SID and IND with suicide attempts using a multivariate unconditional logistic regression model (Agresti, 2002). Hosmer and Lemeshow (2000) tests were used to assess model fit. In the primary model, SID and IND were compared with a nondepressed reference group. We also redefined the reference group in the depression variable to be SID and reran the TABLE 1.

569

analysis to directly compare risk associated with SID versus IND. We calculated the PAR fractions for suicide attempt associated with IND and SID using a formula that considers the adjusted odds ratios (point estimates) for suicide attempt associated with these variables and their prevalence in cases (Kleinbaum et al., 1982; Schlesselman, 1982). In sensitivity analyses, we reran the multivariate model and the estimates of PAR using conditional logistic regression analyses that considered the case–control pairings (Hosmer and Lemeshow, 2000). Last, in analyses limited to cases, we used multivariate linear regression models to examine the associations of the depression variables with the planning and intent of suicide attempts. All analyses were conducted using SAS statistical software Version 9.1 (SAS Institute Inc., Cary, NC). Results Most participants (88% of cases, 81% of controls) met DSM-IV alcohol dependence criteria, with dependence symptom count averaging (SD) 5.4 (2.1) in cases and 4.8 (2.2) in controls. There was weekly or greater use of 1.6 (1.3) different categories of drugs (cannabis, cocaine, etc.) among cases and 1.1 (1.1) among controls. Suicide attempts were most often by drug ingestion (67%) or cutting (21%). Reattempts were common; 80% of cases had made at least one prior suicide attempt, with mean (SD) lifetime number of attempts (including the index attempt) of 3.9 (4.1) and a median of 3. Results of the analyses are presented in Table 1. In the univariate logistic regression analysis, both depression types conferred increased risk (SID: odds ratio [OR] = 3.94, 95% CI [2.13, 7.28]; IND: OR = 9.95, 95% CI [2.62, 37.78]). The multivariate results yielded similar outcomes (SID: OR = 3.73, 95% CI [1.84, 7.58]; IND: OR = 10.38, 95% CI [2.48, 43.49]). The multivariate model showed acceptable

Descriptive data and univariate and multivariate logistic regression results

Predictor Age, in years Female LHAQ Child sex abuse Alcohol severity composite DAST-10 Major interpersonal stressful life event Independent depression Substance-induced depression

100 cases n (%) or M (SD)

100 controls n (%) or M (SD)

OR [95% CI]

AOR [95% CI]

39.0 (9.7) 41.(41%) 16.0 (5.6) 50.(50%)

40.7 (10.6) 21.(21%) 12.7 (4.9) 26.(26%)

0.98 [0.96, 1.01] 2.61 [1.40, 4.88] 1.12 [1.06, 1.19] 2.85 [1.57, 5.16]

0.99 [0.95, 1.02] 2.25 [1.09, 4.65] 1.09 [1.02, 1.16] 1.91 [0.92, 3.96]

11.1 (2.9) 7.5 (2.8)

10.2 (2.9) 6.1 (3.1)

1.12 [1.01, 1.24] 1.18 [1.07, 1.30]

1.11 [0.98, 1.25] 1.08 [0.96, 1.22]

21.(21%)

5.(5%)

5.05 [1.82, 14.01]

5.30 [1.59, 17.69]

13.(13%)

3.(3%)

9.95 [2.62, 37.78]

10.38 [2.48, 43.49]

60.(60%)

35.(35%)

3.94 [2.13, 7.28]

3.73 [1.84, 7.58]

Notes: Independent depression and substance-induced depression have a nondepressed common reference group. OR = odds ratio; AOR = adjusted odds ratio; CI = confidence interval; LHAQ = Lifetime History of Aggression Questionnaire; DAST-10 = 10-item Drug Abuse Screening Test.

570


fit, χ2(8) = 4.71, p = .79. We reran the analysis to compare the depression types directly, and no statistically significant differences were found: IND did not confer greater risk for attempt than SID in the univariate analysis (OR = 2.53, 95% CI [0.67, 9.49]) or in the multivariate analysis (OR = 2.78, 95% CI [0.74, 13.80]). With the exception of age, in univariate analyses all the covariates were associated with increased risk for suicide attempt. In the multivariate analysis, child sexual abuse, alcohol-related severity, and other drug–related severity became statistically nonsignificant. PAR for suicide attempts associated with SID and IND was 0.44 and 0.12, respectively. Sensitivity analyses that considered the case–control pairings yielded similar results. In univariate linear regression analyses of cases, the depression variables were not associated at a statistically significant level with the intent of suicide attempts (SID: p = .325; IND: p = .130) or the planning of attempts (SID: p = .826; IND, p = .811). Similar results were obtained in the multivariate models. Discussion We examined proximal risk for suicide attempts associated with SID and IND in an adult clinical substance use disorders population, with more than 80% of cases and controls meeting DSM-IV alcohol-dependence criteria. Consistent with our hypotheses, both depression subtypes were associated with increased proximal risk after adjustment for risk factors. We did not determine that IND confers greater risk for attempt than SID, as hypothesized. Results also did not support our hypotheses that IND is associated with greater suicide planning and suicide intent compared with SID. Although depression in general is associated with suicidal acts of greater planning (Bagge et al., 2013b; Conner, 2004), it may not be the case that proximal planning is greater for one subtype of depression versus another. However, compared with previous studies on this topic, the current study had a small number of participants, limiting statistical power to detect effects. Notably, we observed a statistical trend (p = .17) for IND having greater proximal risk for suicide attempt compared with SID, consistent with the findings of Preuss and colleagues (2002a). Conjecture regarding observed relations between suicide intent, IND, and SID groups is a bit more complex. Aharonovich et al. (2002) demonstrated that IND was associated with greater suicide intent (compared with nondepressed suicide attempters), and the current study observed a statistical trend in this direction using the same comparison group (IND vs. nondepressed; p = .13). Given our relatively small sample of suicide attempters (i.e., cases), we were unable to test suicide intent differences between IND and SUD groups. Thus, future larger studies should examine whether suicide intent does not differ between SID and IND groups, as was found in a much larger sample of lifetime suicide attempters (Preuss et al., 2002b).

SID was the more common depression subtype, identified in 60% of attempters (and 35% of controls) compared with 13% of cases (and 3% of controls) with IND. The differing prevalence rates figured prominently in the PAR calculations, estimated to be 0.44 for SID and 0.12 for IND. In other words, the analysis suggests that in this clinical sample, approximately 44% of suicide attempts were attributable to SID and 12% to IND. These results underscore the importance of addressing comorbid depression, particularly SID, in the design of suicide-prevention interventions in substance use disorder treatment settings. Further, patients with SID who present for psychiatric evaluation should be carefully assessed; one study reported that 40% of hospitalized suicidal inpatients had a substance-induced mental disorder (Ries et al., 2009). Research indicates that untreated depression increases risk for drinking following alcohol treatment (Kodl et al., 2008) and that individuals who conclude treatment with higher levels of depressive symptoms drink more often and more intensively in the year following treatment than patients with fewer symptoms (Gamble et al., 2010). As a consequence, effective adjunctive interventions that address depressive symptoms in the context of addiction treatment are needed to reduce risk for suicidal behavior and to improve longer-term abstinence and overall functioning. There were limitations to the study. The moderate overall sample size and small samples of cases and controls with IND limited statistical power and precision. Accordingly, the nonsignificant result in the comparison of individuallevel risk associated with IND versus SID is interpreted with caution. Indeed, the large difference in the point estimates between these variables suggests that IND is the more potent risk factor at the individual level, a hypothesis to test in future research. We studied suicide attempts in a generally severe adult substance use disorders treatment sample, and, therefore, generalization to suicide deaths and to other study populations or age groups is unclear. There was also the potential for memory decay and biases in recall because of the retrospective examination of suicide attempts up to 90 days preceding treatment admission. There were also noteworthy strengths. This was a rare study of proximal risk for suicide attempts in an AUD population. To our knowledge, it was the first study of any population to examine proximal risk for suicide attempts associated with SID and IND. We used rigorous assessments of suicide attempts, intent of the attempts, and assessments of SID and IND. Analyses were adjusted for carefully selected risk factors for suicide attempt. We obtained similar results in sensitivity analyses. Research examining the role of subtypes of depression in suicidal behavior among individuals with AUD and other substance use disorders is at a nascent stage. Advancements can be made with the use of larger, more powerful studies. Alternative recruitment settings that facilitate interviews with suicide attempters within a short period of the attempt, for example emergency departments and other hospital set-

CONNER ET AL. tings (Bagge et al., 2013a, 2013b), can lend even greater precision in the study of proximal risk. Intervention studies designed to treat SID and IND and/or to prevent suicidal behavior in individuals with these disorders are needed. References Agresti, A. (2002). Categorical data analysis. Hoboken, NJ: Wiley & Sons. Aharonovich, E., Liu, X., Nunes, E., & Hasin, D. S. (2002). Suicide attempts in substance abusers: Effects of major depression in relation to substance use disorders. American Journal of Psychiatry, 159, 1600–1602. Allen, J. P., Litten, R. Z., Fertig, J. B., & Babor, T. (1997). A review of research on the Alcohol Use Disorders Identification Test (AUDIT). Alcoholism: Clinical and Experimental Research, 21, 613–619. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Bagge, C. L., Lee, H.-J., Schumacher, J. A., Gratz, K. L., Krull, J. L., & Holloman, G., Jr. (2013a). Alcohol as an acute risk factor for recent suicide attempts: A case-crossover analysis. Journal of Studies on Alcohol and Drugs, 74, 552–558. Bagge, C. L., Littlefield, A. K., & Lee, H. J. (2013b). Correlates of proximal premeditation among recently hospitalized suicide attempters. Journal of Affective Disorders, 150, 559–564. Beck, A. T., Schuyler, D., & Herman, I. (1974). Development of suicidal intent scales. In A. T. Beck, C. L. Resnick, & D. Lettieri (Eds.), The prediction of suicide (pp. 45–56). Bowie, MD: Charles Press. Cavanagh, J. T. O., Carson, A. J., Sharpe, M., & Lawrie, S. M. (2003). Psychological autopsy studies of suicide: A systematic review. Psychological Medicine, 33, 395–405. Coccaro, E. F., Berman, M. E., & Kavoussi, R. J. (1997). Assessment of life history of aggression: Development and psychometric characteristics. Psychiatry Research, 73, 147–157. Cocco, K. M., & Carey, K. B. (1998). Psychometric properties of the Drug Abuse Screening Test in psychiatric outpatients. Psychological Assessment, 10, 408–414. Conner, K. R. (2004). A call for research on planned vs. unplanned suicidal behavior. Suicide & Life-Threatening Behavior, 34, 89–98. Conner, K. R., Beautrais, A. L., Brent, D. A., Conwell, Y., Phillips, M. R., & Schneider, B. (2011). The next generation of psychological autopsy studies. Part I. Interview content. Suicide & Life-Threatening Behavior, 41, 594–613. Conner, K. R., Beautrais, A. L., & Conwell, Y. (2003). Risk factors for suicide and medically serious suicide attempts among alcoholics: analyses of Canterbury Suicide Project data. Journal of Studies on Alcohol, 64, 551–554. Conner, K. R., Britton, P. C., Sworts, L. M., & Joiner, T. E., Jr. (2007). Suicide attempts among individuals with opiate dependence: The critical role of belonging. Addictive Behaviors, 32, 1395–1404. Conner, K. R., Houston, R. J., Swogger, M. T., Conwell, Y., You, S., He, H., . . . Duberstein, P. R. (2012). Stressful life events and suicidal behavior in adults with alcohol use disorders: Role of event severity, timing, and type. Drug and Alcohol Dependence, 120, 155–161. Conner, K. R., & Ilgen, M. A. (2011). Substance use disorders and suicidal behaviour. In R. O’Connor, S. Platt, & J. Gordon (Eds.), The international handbook of suicide prevention (pp. 93–107). Oxford, England: Wiley-Blackwell. Dube, S. R., Felitti, V. J., Dong, M., Chapman, D. P., Giles, W. H., & Anda, R. F. (2003). Childhood abuse, neglect, and household dysfunction and the risk of illicit drug use: The adverse childhood experiences study. Pediatrics, 111, 564S–572. Gamble, S. A., Conner, K. R., Talbot, N. L., Yu, Q., Tu, X. M., & Connors, G. J. (2010). Effects of pretreatment and posttreatment depressive symp-

571

toms on alcohol consumption following treatment in Project MATCH. Journal of Studies on Alcohol and Drugs, 71, 71–77. Hasin, D., Samet, S., Nunes, E., Meydan, J., Matseoane, K., & Waxman, R. (2006). Diagnosis of comorbid psychiatric disorders in substance users assessed with the Psychiatric Research Interview for Substance and Mental Disorders for DSM-IV. American Journal of Psychiatry, 163, 689–696. Hasin, D., Trautman, K., & Endicott, J. (1998). Psychiatric research interview for substance and mental disorders: Phenomenologically based diagnosis in patients who abuse alcohol or drugs. Psychopharmacology Bulletin, 34, 3–8. Hasin, D. S., Stinson, F. S., Ogburn, E., & Grant, B. F. (2007). Prevalence, correlates, disability, and comorbidity of DSM-IV alcohol abuse and dependence in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 64, 830–842. Hesselbrock, M. N., Meyer, R. E., & Keener, J. J. (1985). Psychopathology in hospitalized alcoholics. Archives of General Psychiatry, 42, 1050–1055. Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression. New York, NY: Wiley and Sons. Kessler, R. C., Crum, R. M., Warner, L. A., Nelson, C. B., Schulenberg, J., & Anthony, J. C. (1997). Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey. Archives of General Psychiatry, 54, 313–321. Kleinbaum, D. G., Kupper, L. L., & Morgenstern, H. (1982). Epidemiologic research. Belmont, CA: Lifetime Learning Publications. Kodl, M. M., Fu, S. S., Willenbring, M. L., Gravely, A., Nelson, D. B., & Joseph, A. M. (2008). The impact of depressive symptoms on alcohol and cigarette consumption following treatment for alcohol and nicotine dependence. Alcoholism: Clinical and Experimental Research, 32, 92–99. Li, Z., Page, A., Martin, G., & Taylor, R. (2011). Attributable risk of psychiatric and socio-economic factors for suicide from individual-level, population-based studies: A systematic review. Social Science & Medicine, 72, 608–616. Murphy, G. E., Wetzel, R. D., Robins, E., & McEvoy, L. (1992). Multiple risk factors predict suicide in alcoholism. Archives of General Psychiatry, 49, 459–463. Preuss, U. W., Schuckit, M. A., Smith, T. L., Danko, G. P., Buckman, K., Bierut, L., . . . Reich, T. (2002a). Comparison of 3190 alcohol-dependent individuals with and without suicide attempts. Alcoholism: Clinical and Experimental Research, 26, 471–477. Preuss, U. W., Schuckit, M. A., Smith, T. L., Danko, G. P., Dasher, A. C., Hesselbrock, M. N., . . . Nurnberger, J. I., Jr. (2002b). A comparison of alcohol-induced and independent depression in alcoholics with histories of suicide attempts. Journal of Studies on Alcohol, 63, 498–502. Ries, R. K., Yuodelis-Flores, C., Roy-Byrne, P. P., Nilssen, O., & Russo, J. (2009). Addiction and suicidal behavior in acute psychiatric inpatients. Comprehensive Psychiatry, 50, 93–99. Rothman, K. J., & Greenland, S. (1998). Matching. In K. J. Rothman & S. Greenland (Eds.), Modern epidemiology (pp. 147–162). Philadelphia, PA: Lippincott-Raven. Salloum, I. M., Mezzich, J. E., Cornelius, J., Day, N. L., Daley, D., & Kirisci, L. (1995). Clinical profile of comorbid major depression and alcohol use disorders in an initial psychiatric evaluation. Comprehensive Psychiatry, 36, 260–266. Schlesselman, J. J. (1982). Case-control studies: Design, conduct, analysis. New York, NY: Oxford University Press. Schuckit, M. A. (2006). Comorbidity between substance use disorders and psychiatric conditions. Addiction, 101, Supplement 1, 76–88. Schuckit, M. A., Smith, T. L., Danko, G. P., Pierson, J., Trim, R., Nurnberger, J. I., Jr., . . . Hesselbrock, V. (2007). A comparison of factors

572


associated with substance-induced versus independent depressions. Journal of Studies on Alcohol and Drugs, 68, 805–812. Schuckit, M. A., Tipp, J. E., Bergman, M., Reich, W., Hesselbrock, V. M., & Smith, T. L. (1997). Comparison of induced and independent major depressive disorders in 2,945 alcoholics. American Journal of Psychiatry, 154, 948–957. Tidemalm, D., Långström, N., Lichtenstein, P., & Runeson, B. (2008). Risk of suicide after suicide attempt according to coexisting psychiatric disorder: Swedish cohort study with long term follow-up. BMJ, 337, a2205.

Vandenbroucke, J. P., von Elm, E., Altman, D. G., Gøtzsche, P. C., Mulrow, C. D., Pocock, S. J., . . . Egger, M. (2007). Strengthening the reporting of observational studies in epidemiology (STROBE): Explanation and elaboration. PLoS Med 4(10): e297. Retrieved from doi:10.1371/journal. pmed.0040297 Yoshimasu, K., Kiyohara, C., Miyashita, K., & The Stress 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, 243–256.