- Email: [email protected]
Development of the PTSD-alcohol expectancy questionnaire
Available online at www.sciencedirect.com
Addictive Behaviors 33 (2008) 841 – 847
Short communication
Development of the PTSD-alcohol expectancy questionnaire Sonya B. Norman a,1 , Robyn K. Inaba b,2 , Tom L. Smith c,3 , Sandra A. Brown d,⁎ a
UCSD Department of Psychiatry and VA San Diego Healthcare Systems, 8810 Rio San Diego Drive, San Diego, CA 92108, USA b Cook County Juvenile Temporary Detention Center, Medical Department/Mental Health, 1100 S. Hamilton Ave., Chicago, IL 60612, USA c Department of Psychiatry (116A), University of California, San Diego, and the Veterans Affairs San Diego Healthcare System, 3350 La Jolla Village Drive, San Diego, California 92161-2002, USA d Department of Psychology and VA San Diego Healthcare Systems, University of California, San Diego, McGill Hall, 9500 Gilman Drive, La Jolla, CA 92093-0109, USA
Abstract Alcohol effect expectancies have important implications in our understanding drinking behavior and motivations for drinking. Several instruments have been developed to asses alcohol expectancies among various populations. Although co-occurrence of PTSD among those with alcohol use disorders is extremely common, there is no measure of PTSD-related alcohol expectancies. The Post-traumatic stress disorder-Alcohol Expectancy Questionnaire (P-AEQ) is a 27-item, self-report questionnaire that was developed to measure individuals' beliefs about the effects of alcohol with regard to symptoms of post-traumatic stress disorder. The P-AEQ was found to measure two primary dimensions, positive and negative alcohol effect expectancies. This instrument demonstrated internal consistency, reliability, and concurrent validity within the Alcohol Expectancies Questionnaire. In addition, the P-AEQ appears to be capable of differentiating AUD from non-AUD populations in a male veteran sample. © 2008 Elsevier Ltd. All rights reserved. Keywords: PTSD; Addiction; Comorbidity; Expectancies
⁎ Corresponding author. Tel.: +1 858 822 1887; fax: +1 858 822 1886. E-mail addresses: [email protected] (S.B. Norman), [email protected] (R.K. Inaba), [email protected] (T.L. Smith), [email protected] (S.A. Brown). 1 Tel.: +1 619 400 5198; fax: +1 619 400 5171. 2 Tel.: +1 312 433 7364; fax: +1 312 433 4600. 3 Tel.: +1 858 642 3883; fax: +1 858 552 7424. 0306-4603/$ - see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2008.01.003
842
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
1. Introduction The rate of posttraumatic stress disorder (PTSD) among individuals with an alcohol or substance use disorder (AUD or SUD) is 25–59% (Brown, Recupero, & Stout, 1995; Perkonigg, Kessler, Storz, & Wittchen, 2000; Norman, Tate, Anderson, & Brown, 2007). SUD patients with PTSD have worse clinical courses (Ouimette, Ahrens, Moos, & Finney, 1998) and more physical health problems (Tate, Norman, McQuaid, & Brown, 2007) than SUD patients without PTSD. Alcohol expectancies are “the effects attributed to alcohol that the individual anticipates experiencing while drinking” (Brown, Goldman, Inn, & Anderson, 1980). Alcohol expectancies can enhance the predictability of alcoholism treatment outcomes (Goldman, 2002). Although certain alcohol expectancies appear to be common among those with PTSD (e.g., tension and distress reduction; Ouimette et al., 1998; Simpson, 2003), alcohol expectancies specific to PTSD symptoms have yet to be measured. The goals of the present study were to 1) develop a measure of alcohol expectancies specific to PTSD symptoms, the PTSD-Alcohol Expectancy Questionnaire (P-AEQ), and 2) examine differences in PTSD-specific alcohol effect expectancies between veterans with and without PTSD.
2. Method 2.1. Instrument development: PTSD-Alcohol Expectancies Questionnaire (P-AEQ) The P-AEQ (Table 1) is a 27-item, self report questionnaire that examines beliefs about the effects of alcohol in relationship to symptoms of PTSD. Respondents rate each item on a 5-point likert scale, indicating the extent to which they agree or disagree. Items for the initial version of the P-AEQ were written based on the DSM-IV criteria for PTSD. For each of the 20 symptoms of PTSD, one positive expectancy item and one negative expectancy item was included in the original measure. Items were reviewed by two clinical psychologists with experience in the assessment and treatment of addictive disorders and PTSD. A qualitative pilot study with 16 male veterans, ages 21 to 70, with DSM IV diagnosis of PTSD and/or Alcohol Dependence who were receiving psychiatric and/or substance abuse treatment (mean age = 47.2, 69% unmarried, 62.5% Caucasian) completed the protocol materials and found that item content, vocabulary, reading level and clarity of instructions were adequate. 2.2. Subjects and procedures Participants were 120 male veterans who were receiving medical and/or psychiatric treatment at the VA San Diego Healthcare System (VASDHS; Table 2). Medical patients with a documented history of alcohol or drug problems, psychiatric disorder, or prohibitive medical condition (e.g., life threatening illness, impending surgery) were excluded. Psychiatric patients with a documented history of DSM IV PTSD and/ or Alcohol Dependence with no active psychotic symptoms were selected. Veterans unable to read, comprehend questions, or provide informed consent were excluded. Of 150 medical and psychiatric patients deemed eligible by in person screening, 33 medical patients and 93 psychiatric patients agreed to participate. Six veterans were dropped from the study. Five were discharged prior to protocol completion and one subject produced inconsistent PTSD information. Participants were classified into one of four diagnostic groups: alcohol dependence without PTSD, with PTSD, PTSD without current alcohol
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
843
Table 1 Final 27-item P-AEQ items and item loadings Item Positive factor (Eignvalue = 8.46, accounted for 31.3% of variance) After a few drinks, my past traumatic experiences would feel less real After a few drinks I would forget things about my past traumatic experiences After a few drinks, memories of traumatic experiences would not bother me as much After a few drinks, I am less likely to be bothered by cues or triggers that remind me of past traumatic experiences After a few drinks my flashbacks about past traumatic events would decrease After a few drinks would numb painful feelings and thoughts about past traumatic events After a few drinks my illusions (visual memories) about past traumatic events would decrease My bad dreams would decrease after a few drinks After a few drinks I would feel calm even when I am around things that remind me of past traumatic experiences After a few drinks my hallucinations about traumatic past events would decrease After a few drinks I would be less startled by things After a few drinks, I would be less angry or on edge After a few drinks I would feel closer to others Negative factor (Eignvalue = 6.45, accounted for 23.9% of variance) After a few drinks my flashbacks about traumatic past events would increase A few drinks would increase painful feelings and thoughts about traumatic events After a few drinks I would be bothered by more memories of traumatic experiences After a few drinks my traumatic past experiences would feel more real After a few drinks my illusions (visual memories) about past traumatic events would increase After a few drinks I would be more likely to be bothered by cues or triggers that remind me of past traumatic experiences After a few drinks I would remember more about my past traumatic experiences After a few drinks my hallucinations about traumatic past experiences would increase My bad dreams would increase after a few drinks After a few drinks I would be more angry or on edge After a few drinks I would be more on guard After a few drinks my heart races, my palms sweat or I have difficulty breathing when I am around things that remind me of past traumatic experiences After a few drinks I would be more easily startled by things After a few drinks I would feel more alone and distant from others
Factor loading .79 .78 .77 .77 .76 .73 .72 .66 .65 .63 .59 .57 .53 .88 .88 .85 .84 .82 .81 .79 .78 .76 .72 .68 .64 .63 .58
dependence, and medical patients with no psychiatric diagnosis. Medical patients did not differ significantly from the other groups on demographic variables. After a minimum of five days of inpatient medical or psychiatric service, participants completed a diagnostic interview conducted by a trained research assistant and self-report measures. All participants signed and received a copy of the informed consent form for VASDHS and the University of California, San Diego. 2.3. Measures In addition to the P-AEQ, participants completed the Structured Clinical Interview for DSM-IV (SCIDIV; First, Spitzer, Gibbon, & Williams, 1997) for PTSD, alcohol, and substance use disorders, the Alcohol
844
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
Table 2 Participant demographic characteristics
N Age (mean)1 Ethnicity African American Hispanic Caucasian Other Marital status2 Never married Married Widowed, separated, divorced Education3 Less than grade 12 Grade 12 or GED 1 + years of college Current income per month4 N$800 $800–600 N600 Impact of events scale score (range = 0–60)5 Average # of drinks in 90 days pre-treatment6 1
Alcohol use disorder
PTSD
Alcohol and PTSD
Medical controls
Total
45 48.02 (9.07)
15 49.60 (3.68)
30 48.93 (3.83)
30 55.47 (9.19)
50.40 (8.34)
9 (7.5%) 2 (1.7%) 33 (27.5%) 1 (0.8%)
0 (0.0%) 2 (1.7%) 12 (10.0%) 1 (0.8%)
7 (5.8%) 7 (5.8%) 14 (11.7%) 2 (1.7%)
8 (6.7%) 3 (2.5%) 17 (14.2%) 2 (1.7%)
24 (20.0%) 14 (11.7%) 76 (63.3%) 6 (5.0%)
11 (9.2%) 3 (2.5%) 31 (25.8%)
1 (0.8%) 6 (5.0%) 8 (6.6%)
1 (0.8%) 6 (5.0%) 23 (19.1%)
5(4.2%) 16 (13.3%) 9 (7.4%)
18 (15.0%) 31 (25.8%) 71 (59.2%)
3 (2.5%) 18 (15.0%) 24 (20%)
1 (0.8%) 3 (2.5%) 11 (9.2%)
4 (3.3%) 9 (7.5%) 17 (15.2%)
1 (0.8%) 10 (8.3%) 19 (15.8%)
9 (7.5%) 40 (33.3%) 71 (59.2%)
27 (22.5%) 12 (10.0%) 6 (5.0%)
2 (1.7%) 2 (1.7%) 11 (9.2%)
12 (10.01%) 6 (5.0%) 12 (10.0%)
10 (8.3%) 8 (6.7%) 12 (9.9%)
51 (42.5%) 28 (23.3%) 41 (34.2%)
31.95 459.06
47.66 6.40
46.50 319.26
24.80 8.66
35.76 254.93
Medical patients were significantly older than the other three groups.
2
AUD patients with and without PTSD were more likely to be separated or divorced than others. 3 AUD patients with and without PTSD had lower education than non-AUD patients. 4 AUD patients with and without PTSD had lower income than non-AUD patients. 5 PTSD patients with and without AUD had had higher Impact of Events Scale scores than patients without PTSD. 6 AUD patients with and without PTSD had greater number of drinks in the 90 days preceding treatment than patients without AUD.
Expectancy Questionnaire — Revised (AEQ-R; Brown, Christiansen, & Goldman, 1987), the Timeline Follow-back Procedure (TLFB; Sobell & Sobell, 1992) for recent alcohol use, and The Impact of Events Scale (Horowitz, Wilner, & Alvarez, 1979) for PTSD symptoms.
3. Results 3.1. P-AEQ development, reliability, and validity Principal components analysis as suggested by Briggs and Cheek (1986) was used to articulate the underlying factor structure. Results suggested 27 retainable items with two factors: positive PTSD alcohol expectancies and negative PTSD alcohol expectancies (Table 1). Cronbach's alphas were .91 and .94 for the positive and negative scales. Concurrent validity was evidenced by the correlation between the positive items on the P-AEQ and the AEQ (r = .58, p b .001).
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
845
Table 3 Group differences in P-AEQ positive factor Sources
Mean
Alcohol PTSD Alcohol X PTSD Residual Total
SS
df
MS
F
Sig.
Eta2
2283.145 15.638 167.966 14978.923 17481.242
1 1 1 116 119
2283.145 15.638 167.996 129.129 146.901
17.681 .121 1.301
.000 .728 .256
.106 .004 .011
Table 4 Group differences in P-AEQ negative factor Sources
SS
df
MS
F
Sig.
Eta2
Covariate: age Alcohol PTSD Alcohol × PTSD Residual Total
1124.782 1070.217 2288.684 12.161 19450.228 24083.125
1 1 1 1 115 119
1124.782 1070.217 2288.684 12.161 169.133 202.379
6.650 6.328 13.532 .072
.011 .013 .000 .789
.047 .10 .001
3.2. Alcohol expectancies in AUD and PTSD ANCOVAs were used to assess the hypothesis that P-AEQ alcohol expectancies would be greater in those with co-occurring PTSD and alcohol use disorders than in those with PTSD without alcohol use disorders. The positive and negative P-AEQ scales were evaluated separately. Age, marital status, education, amount consumed in the 3 months prior to assessment, and income were included as covariates (Table 2). For both the positive and negative scale, a 2 (AUD vs. non-AUD) by 2 (PTSD vs. non-PTSD) analysis of variance was conducted. Only the main effect for AUD was significant for the positive scale (Table 3). Both the main effect for alcohol dependence and the main effect for PTSD were significant for the negative scale (Table 4).
4. Discussion The study demonstrates that alcohol effect expectancies can be assessed in relation to symptoms of PTSD. The P-AEQ was found to measure two primary dimensions, positive and negative PTSD symptom related alcohol effect expectancies. This instrument demonstrated in a preliminary way that internal consistency and concurrent validity with the P-AEQ are adequate. The negative expectancies scale was found to differentiate veterans with PTSD from non-PTSD adults. In addition, the P-AEQ appears to be capable of differentiating AUD from non-AUD populations. Veterans with PTSD did not endorse more symptoms of positive alcohol expectancies related to PTSD symptoms than other veterans. Thus, a self-medication hypothesis for alcohol use among those with PTSD symptoms was not supported. These veterans were receiving treatment for PTSD and/or alcohol use, and treatment interventions may have influenced their beliefs about the effects of alcohol on PTSD symptoms.
846
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
Expectancies at onset of problem drinking may be different than those that maintain drinking behavior. Most of the Vietnam Veterans in this study had been drinking continuously since their combat experience. It is possible that over time their expectancies or beliefs about the effects of alcohol may no longer be symptom specific, but more general in terms of the effects that alcohol may produce. Individuals who are alcohol dependent may not evaluate potential alcohol-related outcomes, but instead drink as an unconditioned automatic response to negative internal states. The sample was limited to male veterans with chronic combat-related PTSD. Future research is necessary to determine differences between clinical and non-clinical samples regarding the P-AEQ. Additional research in different populations with varied types of trauma is needed. A limitation of both the AEQ and the P-AEQ is that respondents are asked to consider a moderate dose of alcohol (e.g. “a few drinks”). Prior research has found that expectancies vary with the quantity of alcohol considered (Southwick, Steele, Marlatt, & Lindell, 1981) and clinical samples generally drink beyond “moderate” levels. The results of this initial study suggest that the P-AEQ may be a useful research and clinical tool. The P-AEQ is one of the first measures developed specifically to measure cognitive mechanisms underlying alcohol use in those with PTSD. The P-AEQ may have utility in the assessment and treatment of comorbid PTSD and AUD/SUD. Acknowledgments This research was supported by a V.A. Merit Award to Dr. Sandra Brown and K23-AA015707 awarded to Dr. Sonya Norman.
References Briggs, S. R., & Cheek, J. M. (1986). The role of factor analysis in the development and evaluation of personality scales. Journal of Personality, 54, 106−148. Brown, P. J., Recupero, P. R., & Stout, R. (1995). PTSD substance abuse comorbidity and treatment utilization. Addictive Behavior, 20, 251−254. Brown, S. A., Christiansen, B. A., & Goldman, M. S. (1987). The Alcohol Expectancy Questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. Journal of Studies on Alcohol, 48, 483−491. Brown, S. A., Goldman, M. S., Inn, A., & Anderson, L. R. (1980). Expectations of reinforcement from alcohol: Their domain and relation to drinking patterns. Journal of Consulting and Clinical Psychology, 45, 219−224. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Structured clinical interview for DSM-IVAxis I Disorders Patient edition (SCID-I/P, Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute. Goldman, M. S. (2002). Expectancy and risk for alcoholism: The unfortunate exploitation of a fundamental characteristic of neurobehavioral adaptation. Alcohol Clinical Experience Research, 26, 737−746. Horowitz, M., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of subjective stress. Psychosomatic Medicine, 41, 209−218. Norman, S. B., Tate, S. R., Anderson, K. G., & Brown, S. A. (2007). Do trauma history and PTSD symptoms influence addiction relapse context? Drug and Alcohol Dependence, 33(1), 25−32. Ouimette, P. C., Ahrens, C., Moos, R. H., & Finney, J. W. (1998). During treatment changes in substance abuse patients with posttraumatic stress disorder. The influence of specific interventions and program environments. Journal of Substance Abuse Treatment, 15, 555−564. Perkonigg, A., Kessler, R. C., Storz, S., & Wittchen, H. U. (2000). Traumatic events and post-traumatic stress disorder in the community: Prevalence, risk factors and comorbidity. Acta Psychiatrica Scandinavia, 101, 46−59.
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
847
Simpson, T. L. (2003). Childhood sexual abuse, PTSD, and the functional roles of alcohol use among women drinkers. Substance Use and Misuse, 38, 249−270. Sobell, L. C., & Sobell, M. B. (1992). Timeline follow-back: A technique for assessing self-reported alcohol consumption. In R. L. Litten & J.P. Allen (Eds.), Measuring Alcohol Consumption: Psychosocial and Biochemical Methods New Jersey: Humana Press. Southwick, L., Steele, C. M., Marlatt, G. A., & Lindell, M. (1981). Alcohol-related expectancies: Defined by phase of intoxication and drinking experience. Journal of Consulting and Clinical Psychology, 49, 713−721. Tate, S. R., Norman, S. B., McQuaid, J. R., & Brown, S. A. (2007). Health problems of substance-dependent veterans with and those without trauma history. Journal of Substance Abuse Treatment, 33(1), 25−32.
Addictive Behaviors 33 (2008) 841 – 847
Short communication
Development of the PTSD-alcohol expectancy questionnaire Sonya B. Norman a,1 , Robyn K. Inaba b,2 , Tom L. Smith c,3 , Sandra A. Brown d,⁎ a
UCSD Department of Psychiatry and VA San Diego Healthcare Systems, 8810 Rio San Diego Drive, San Diego, CA 92108, USA b Cook County Juvenile Temporary Detention Center, Medical Department/Mental Health, 1100 S. Hamilton Ave., Chicago, IL 60612, USA c Department of Psychiatry (116A), University of California, San Diego, and the Veterans Affairs San Diego Healthcare System, 3350 La Jolla Village Drive, San Diego, California 92161-2002, USA d Department of Psychology and VA San Diego Healthcare Systems, University of California, San Diego, McGill Hall, 9500 Gilman Drive, La Jolla, CA 92093-0109, USA
Abstract Alcohol effect expectancies have important implications in our understanding drinking behavior and motivations for drinking. Several instruments have been developed to asses alcohol expectancies among various populations. Although co-occurrence of PTSD among those with alcohol use disorders is extremely common, there is no measure of PTSD-related alcohol expectancies. The Post-traumatic stress disorder-Alcohol Expectancy Questionnaire (P-AEQ) is a 27-item, self-report questionnaire that was developed to measure individuals' beliefs about the effects of alcohol with regard to symptoms of post-traumatic stress disorder. The P-AEQ was found to measure two primary dimensions, positive and negative alcohol effect expectancies. This instrument demonstrated internal consistency, reliability, and concurrent validity within the Alcohol Expectancies Questionnaire. In addition, the P-AEQ appears to be capable of differentiating AUD from non-AUD populations in a male veteran sample. © 2008 Elsevier Ltd. All rights reserved. Keywords: PTSD; Addiction; Comorbidity; Expectancies
⁎ Corresponding author. Tel.: +1 858 822 1887; fax: +1 858 822 1886. E-mail addresses: [email protected] (S.B. Norman), [email protected] (R.K. Inaba), [email protected] (T.L. Smith), [email protected] (S.A. Brown). 1 Tel.: +1 619 400 5198; fax: +1 619 400 5171. 2 Tel.: +1 312 433 7364; fax: +1 312 433 4600. 3 Tel.: +1 858 642 3883; fax: +1 858 552 7424. 0306-4603/$ - see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2008.01.003
842
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
1. Introduction The rate of posttraumatic stress disorder (PTSD) among individuals with an alcohol or substance use disorder (AUD or SUD) is 25–59% (Brown, Recupero, & Stout, 1995; Perkonigg, Kessler, Storz, & Wittchen, 2000; Norman, Tate, Anderson, & Brown, 2007). SUD patients with PTSD have worse clinical courses (Ouimette, Ahrens, Moos, & Finney, 1998) and more physical health problems (Tate, Norman, McQuaid, & Brown, 2007) than SUD patients without PTSD. Alcohol expectancies are “the effects attributed to alcohol that the individual anticipates experiencing while drinking” (Brown, Goldman, Inn, & Anderson, 1980). Alcohol expectancies can enhance the predictability of alcoholism treatment outcomes (Goldman, 2002). Although certain alcohol expectancies appear to be common among those with PTSD (e.g., tension and distress reduction; Ouimette et al., 1998; Simpson, 2003), alcohol expectancies specific to PTSD symptoms have yet to be measured. The goals of the present study were to 1) develop a measure of alcohol expectancies specific to PTSD symptoms, the PTSD-Alcohol Expectancy Questionnaire (P-AEQ), and 2) examine differences in PTSD-specific alcohol effect expectancies between veterans with and without PTSD.
2. Method 2.1. Instrument development: PTSD-Alcohol Expectancies Questionnaire (P-AEQ) The P-AEQ (Table 1) is a 27-item, self report questionnaire that examines beliefs about the effects of alcohol in relationship to symptoms of PTSD. Respondents rate each item on a 5-point likert scale, indicating the extent to which they agree or disagree. Items for the initial version of the P-AEQ were written based on the DSM-IV criteria for PTSD. For each of the 20 symptoms of PTSD, one positive expectancy item and one negative expectancy item was included in the original measure. Items were reviewed by two clinical psychologists with experience in the assessment and treatment of addictive disorders and PTSD. A qualitative pilot study with 16 male veterans, ages 21 to 70, with DSM IV diagnosis of PTSD and/or Alcohol Dependence who were receiving psychiatric and/or substance abuse treatment (mean age = 47.2, 69% unmarried, 62.5% Caucasian) completed the protocol materials and found that item content, vocabulary, reading level and clarity of instructions were adequate. 2.2. Subjects and procedures Participants were 120 male veterans who were receiving medical and/or psychiatric treatment at the VA San Diego Healthcare System (VASDHS; Table 2). Medical patients with a documented history of alcohol or drug problems, psychiatric disorder, or prohibitive medical condition (e.g., life threatening illness, impending surgery) were excluded. Psychiatric patients with a documented history of DSM IV PTSD and/ or Alcohol Dependence with no active psychotic symptoms were selected. Veterans unable to read, comprehend questions, or provide informed consent were excluded. Of 150 medical and psychiatric patients deemed eligible by in person screening, 33 medical patients and 93 psychiatric patients agreed to participate. Six veterans were dropped from the study. Five were discharged prior to protocol completion and one subject produced inconsistent PTSD information. Participants were classified into one of four diagnostic groups: alcohol dependence without PTSD, with PTSD, PTSD without current alcohol
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
843
Table 1 Final 27-item P-AEQ items and item loadings Item Positive factor (Eignvalue = 8.46, accounted for 31.3% of variance) After a few drinks, my past traumatic experiences would feel less real After a few drinks I would forget things about my past traumatic experiences After a few drinks, memories of traumatic experiences would not bother me as much After a few drinks, I am less likely to be bothered by cues or triggers that remind me of past traumatic experiences After a few drinks my flashbacks about past traumatic events would decrease After a few drinks would numb painful feelings and thoughts about past traumatic events After a few drinks my illusions (visual memories) about past traumatic events would decrease My bad dreams would decrease after a few drinks After a few drinks I would feel calm even when I am around things that remind me of past traumatic experiences After a few drinks my hallucinations about traumatic past events would decrease After a few drinks I would be less startled by things After a few drinks, I would be less angry or on edge After a few drinks I would feel closer to others Negative factor (Eignvalue = 6.45, accounted for 23.9% of variance) After a few drinks my flashbacks about traumatic past events would increase A few drinks would increase painful feelings and thoughts about traumatic events After a few drinks I would be bothered by more memories of traumatic experiences After a few drinks my traumatic past experiences would feel more real After a few drinks my illusions (visual memories) about past traumatic events would increase After a few drinks I would be more likely to be bothered by cues or triggers that remind me of past traumatic experiences After a few drinks I would remember more about my past traumatic experiences After a few drinks my hallucinations about traumatic past experiences would increase My bad dreams would increase after a few drinks After a few drinks I would be more angry or on edge After a few drinks I would be more on guard After a few drinks my heart races, my palms sweat or I have difficulty breathing when I am around things that remind me of past traumatic experiences After a few drinks I would be more easily startled by things After a few drinks I would feel more alone and distant from others
Factor loading .79 .78 .77 .77 .76 .73 .72 .66 .65 .63 .59 .57 .53 .88 .88 .85 .84 .82 .81 .79 .78 .76 .72 .68 .64 .63 .58
dependence, and medical patients with no psychiatric diagnosis. Medical patients did not differ significantly from the other groups on demographic variables. After a minimum of five days of inpatient medical or psychiatric service, participants completed a diagnostic interview conducted by a trained research assistant and self-report measures. All participants signed and received a copy of the informed consent form for VASDHS and the University of California, San Diego. 2.3. Measures In addition to the P-AEQ, participants completed the Structured Clinical Interview for DSM-IV (SCIDIV; First, Spitzer, Gibbon, & Williams, 1997) for PTSD, alcohol, and substance use disorders, the Alcohol
844
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
Table 2 Participant demographic characteristics
N Age (mean)1 Ethnicity African American Hispanic Caucasian Other Marital status2 Never married Married Widowed, separated, divorced Education3 Less than grade 12 Grade 12 or GED 1 + years of college Current income per month4 N$800 $800–600 N600 Impact of events scale score (range = 0–60)5 Average # of drinks in 90 days pre-treatment6 1
Alcohol use disorder
PTSD
Alcohol and PTSD
Medical controls
Total
45 48.02 (9.07)
15 49.60 (3.68)
30 48.93 (3.83)
30 55.47 (9.19)
50.40 (8.34)
9 (7.5%) 2 (1.7%) 33 (27.5%) 1 (0.8%)
0 (0.0%) 2 (1.7%) 12 (10.0%) 1 (0.8%)
7 (5.8%) 7 (5.8%) 14 (11.7%) 2 (1.7%)
8 (6.7%) 3 (2.5%) 17 (14.2%) 2 (1.7%)
24 (20.0%) 14 (11.7%) 76 (63.3%) 6 (5.0%)
11 (9.2%) 3 (2.5%) 31 (25.8%)
1 (0.8%) 6 (5.0%) 8 (6.6%)
1 (0.8%) 6 (5.0%) 23 (19.1%)
5(4.2%) 16 (13.3%) 9 (7.4%)
18 (15.0%) 31 (25.8%) 71 (59.2%)
3 (2.5%) 18 (15.0%) 24 (20%)
1 (0.8%) 3 (2.5%) 11 (9.2%)
4 (3.3%) 9 (7.5%) 17 (15.2%)
1 (0.8%) 10 (8.3%) 19 (15.8%)
9 (7.5%) 40 (33.3%) 71 (59.2%)
27 (22.5%) 12 (10.0%) 6 (5.0%)
2 (1.7%) 2 (1.7%) 11 (9.2%)
12 (10.01%) 6 (5.0%) 12 (10.0%)
10 (8.3%) 8 (6.7%) 12 (9.9%)
51 (42.5%) 28 (23.3%) 41 (34.2%)
31.95 459.06
47.66 6.40
46.50 319.26
24.80 8.66
35.76 254.93
Medical patients were significantly older than the other three groups.
2
AUD patients with and without PTSD were more likely to be separated or divorced than others. 3 AUD patients with and without PTSD had lower education than non-AUD patients. 4 AUD patients with and without PTSD had lower income than non-AUD patients. 5 PTSD patients with and without AUD had had higher Impact of Events Scale scores than patients without PTSD. 6 AUD patients with and without PTSD had greater number of drinks in the 90 days preceding treatment than patients without AUD.
Expectancy Questionnaire — Revised (AEQ-R; Brown, Christiansen, & Goldman, 1987), the Timeline Follow-back Procedure (TLFB; Sobell & Sobell, 1992) for recent alcohol use, and The Impact of Events Scale (Horowitz, Wilner, & Alvarez, 1979) for PTSD symptoms.
3. Results 3.1. P-AEQ development, reliability, and validity Principal components analysis as suggested by Briggs and Cheek (1986) was used to articulate the underlying factor structure. Results suggested 27 retainable items with two factors: positive PTSD alcohol expectancies and negative PTSD alcohol expectancies (Table 1). Cronbach's alphas were .91 and .94 for the positive and negative scales. Concurrent validity was evidenced by the correlation between the positive items on the P-AEQ and the AEQ (r = .58, p b .001).
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
845
Table 3 Group differences in P-AEQ positive factor Sources
Mean
Alcohol PTSD Alcohol X PTSD Residual Total
SS
df
MS
F
Sig.
Eta2
2283.145 15.638 167.966 14978.923 17481.242
1 1 1 116 119
2283.145 15.638 167.996 129.129 146.901
17.681 .121 1.301
.000 .728 .256
.106 .004 .011
Table 4 Group differences in P-AEQ negative factor Sources
SS
df
MS
F
Sig.
Eta2
Covariate: age Alcohol PTSD Alcohol × PTSD Residual Total
1124.782 1070.217 2288.684 12.161 19450.228 24083.125
1 1 1 1 115 119
1124.782 1070.217 2288.684 12.161 169.133 202.379
6.650 6.328 13.532 .072
.011 .013 .000 .789
.047 .10 .001
3.2. Alcohol expectancies in AUD and PTSD ANCOVAs were used to assess the hypothesis that P-AEQ alcohol expectancies would be greater in those with co-occurring PTSD and alcohol use disorders than in those with PTSD without alcohol use disorders. The positive and negative P-AEQ scales were evaluated separately. Age, marital status, education, amount consumed in the 3 months prior to assessment, and income were included as covariates (Table 2). For both the positive and negative scale, a 2 (AUD vs. non-AUD) by 2 (PTSD vs. non-PTSD) analysis of variance was conducted. Only the main effect for AUD was significant for the positive scale (Table 3). Both the main effect for alcohol dependence and the main effect for PTSD were significant for the negative scale (Table 4).
4. Discussion The study demonstrates that alcohol effect expectancies can be assessed in relation to symptoms of PTSD. The P-AEQ was found to measure two primary dimensions, positive and negative PTSD symptom related alcohol effect expectancies. This instrument demonstrated in a preliminary way that internal consistency and concurrent validity with the P-AEQ are adequate. The negative expectancies scale was found to differentiate veterans with PTSD from non-PTSD adults. In addition, the P-AEQ appears to be capable of differentiating AUD from non-AUD populations. Veterans with PTSD did not endorse more symptoms of positive alcohol expectancies related to PTSD symptoms than other veterans. Thus, a self-medication hypothesis for alcohol use among those with PTSD symptoms was not supported. These veterans were receiving treatment for PTSD and/or alcohol use, and treatment interventions may have influenced their beliefs about the effects of alcohol on PTSD symptoms.
846
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
Expectancies at onset of problem drinking may be different than those that maintain drinking behavior. Most of the Vietnam Veterans in this study had been drinking continuously since their combat experience. It is possible that over time their expectancies or beliefs about the effects of alcohol may no longer be symptom specific, but more general in terms of the effects that alcohol may produce. Individuals who are alcohol dependent may not evaluate potential alcohol-related outcomes, but instead drink as an unconditioned automatic response to negative internal states. The sample was limited to male veterans with chronic combat-related PTSD. Future research is necessary to determine differences between clinical and non-clinical samples regarding the P-AEQ. Additional research in different populations with varied types of trauma is needed. A limitation of both the AEQ and the P-AEQ is that respondents are asked to consider a moderate dose of alcohol (e.g. “a few drinks”). Prior research has found that expectancies vary with the quantity of alcohol considered (Southwick, Steele, Marlatt, & Lindell, 1981) and clinical samples generally drink beyond “moderate” levels. The results of this initial study suggest that the P-AEQ may be a useful research and clinical tool. The P-AEQ is one of the first measures developed specifically to measure cognitive mechanisms underlying alcohol use in those with PTSD. The P-AEQ may have utility in the assessment and treatment of comorbid PTSD and AUD/SUD. Acknowledgments This research was supported by a V.A. Merit Award to Dr. Sandra Brown and K23-AA015707 awarded to Dr. Sonya Norman.
References Briggs, S. R., & Cheek, J. M. (1986). The role of factor analysis in the development and evaluation of personality scales. Journal of Personality, 54, 106−148. Brown, P. J., Recupero, P. R., & Stout, R. (1995). PTSD substance abuse comorbidity and treatment utilization. Addictive Behavior, 20, 251−254. Brown, S. A., Christiansen, B. A., & Goldman, M. S. (1987). The Alcohol Expectancy Questionnaire: An instrument for the assessment of adolescent and adult alcohol expectancies. Journal of Studies on Alcohol, 48, 483−491. Brown, S. A., Goldman, M. S., Inn, A., & Anderson, L. R. (1980). Expectations of reinforcement from alcohol: Their domain and relation to drinking patterns. Journal of Consulting and Clinical Psychology, 45, 219−224. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Structured clinical interview for DSM-IVAxis I Disorders Patient edition (SCID-I/P, Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute. Goldman, M. S. (2002). Expectancy and risk for alcoholism: The unfortunate exploitation of a fundamental characteristic of neurobehavioral adaptation. Alcohol Clinical Experience Research, 26, 737−746. Horowitz, M., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of subjective stress. Psychosomatic Medicine, 41, 209−218. Norman, S. B., Tate, S. R., Anderson, K. G., & Brown, S. A. (2007). Do trauma history and PTSD symptoms influence addiction relapse context? Drug and Alcohol Dependence, 33(1), 25−32. Ouimette, P. C., Ahrens, C., Moos, R. H., & Finney, J. W. (1998). During treatment changes in substance abuse patients with posttraumatic stress disorder. The influence of specific interventions and program environments. Journal of Substance Abuse Treatment, 15, 555−564. Perkonigg, A., Kessler, R. C., Storz, S., & Wittchen, H. U. (2000). Traumatic events and post-traumatic stress disorder in the community: Prevalence, risk factors and comorbidity. Acta Psychiatrica Scandinavia, 101, 46−59.
S.B. Norman et al. / Addictive Behaviors 33 (2008) 841–847
847
Simpson, T. L. (2003). Childhood sexual abuse, PTSD, and the functional roles of alcohol use among women drinkers. Substance Use and Misuse, 38, 249−270. Sobell, L. C., & Sobell, M. B. (1992). Timeline follow-back: A technique for assessing self-reported alcohol consumption. In R. L. Litten & J.P. Allen (Eds.), Measuring Alcohol Consumption: Psychosocial and Biochemical Methods New Jersey: Humana Press. Southwick, L., Steele, C. M., Marlatt, G. A., & Lindell, M. (1981). Alcohol-related expectancies: Defined by phase of intoxication and drinking experience. Journal of Consulting and Clinical Psychology, 49, 713−721. Tate, S. R., Norman, S. B., McQuaid, J. R., & Brown, S. A. (2007). Health problems of substance-dependent veterans with and those without trauma history. Journal of Substance Abuse Treatment, 33(1), 25−32.