Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Dec 1.
Published in final edited form as: Addict Behav. 2019 Aug 5;99:106079. doi: 10.1016/j.addbeh.2019.106079

Development and Initial Validation of the Alcohol-Induced Blackout Measure

Mary Beth Miller a,b,*, Angelo M DiBello b, Jennifer E Merrill b, Kate B Carey b
PMCID: PMC6791777  NIHMSID: NIHMS1538882  PMID: 31442787

Abstract

Background:

Blackouts are common among young adults and predict alcohol-related harm. However, existing measures do not capture the range of alcohol-induced memory impairment involved in blackout experiences and do not differentiate between fragmentary and en bloc blackouts. This study aimed to develop and validate a brief, reliable measure of alcohol-induced blackouts among young adults.

Methods:

College students reporting alcohol-induced memory impairment in the past year were recruited via Qualtrics to participate in an online survey (N=350, 56% female). A subsample (n=109, 67% female) completed a one-month follow-up. Principal component analysis was used to determine the structure of the Alcohol-Induced Blackout Measure (ABOM), which was designed to reflect two components (fragmentary and en bloc blackouts). The reliability and validity of the total ABOM score was assessed.

Results:

The final five items fit in a two-component scale structure; however, a single principal component accounted for 73% of variance in blackout items, all of which demonstrated high component loadings and communalities. The total blackout score demonstrated strong internal consistency, test-retest reliability, and convergent and incremental validity. ABOM scores predicted alcohol-related consequences at baseline and one-month follow-up.

Conclusions:

The ABOM is a brief and reliable, self-report measure that quantifies the frequency of a range of blackout experiences in the past 30 days. Accounting for this range of experiences improved predictive validity over single-item blackout measures. Blackout frequency is a strong, unique predictor of alcohol-related problems.

Keywords: alcohol, drinking, amnesia, college students, young adults

1. Introduction

Blackouts are periods of alcohol-induced anterograde amnesia, in which individuals are unable to form new long-term memories while intoxicated but still able to recall short-term memories and memories formed prior to intoxication. As a result, intoxicated individuals participate in events that they later cannot remember (White, 2003). This experience differs from “passing out,” in that individuals are conscious and engage with the environment, making it difficult for others to determine that they are blacked out (Wetherill & Fromme, 2016). Blackouts occur primarily as a function of rapid increase in blood alcohol concentration (White, 2003) and have been associated with consequences ranging from truancy to overdose (Hingson, Zha, Simons-Morton, & White, 2016).

Approximately half of young adult drinkers report a lifetime history of blackout (Wetherill & Fromme, 2016). Given the risks associated with blackouts, such high incidence is concerning. However, the term blackout encompasses varying levels of memory impairment. Goodwin and colleagues (1969) documented two forms of blackout: “en bloc” blackouts (EB), or complete memory loss for events that occur while intoxicated, and “fragmentary” blackouts (FB), or partial memory loss that can later be pieced together, often with cues (Goodwin et al., 1969). FB are more common than EB, and young adults evaluate them less negatively (Hartzler & Fromme, 2003; Miller, Merrill, DiBello, & Carey, 2018; White, Signer, Kraus, & Swartzwelder, 2004). Thus, it seems they are distinct phenomena.

Despite young adults’ differential evaluations of these experiences, the literature lacks a standardized blackouts assessment (Wetherill & Fromme, 2016). Instead, research has relied on blackout measures derived either by the author or from select (often single) items from a larger scale (see Wetherill & Fromme, 2016 for a review). For example, Marino and Fromme (2018) define blackouts as “difficulty remembering things you said or did, or events that happened, while you were drinking.” This definition characterizes both FB and EB but does not differentiate between the two.

Blackout assessments derived from existing scales also typically fail to capture the range of alcohol-related memory impairment. For example, the Alcohol Use Disorders Identification Test (AUDIT; Saunders, Aasland, Babor, de la Fuente, & Grant, 1993) includes a single item assessing how often a participant has “been unable to remember what happened the night before because you had been drinking.” Another commonly used scale, the Rutgers Alcohol Problem Index (RAPI; White & Labouvie, 2000), assesses how often one has “suddenly found yourself in a place that you could not remember getting to.” The Young Adult Alcohol Consequences Questionnaire (Read, Kahler, Strong, & Colder, 2006) is rare in its inclusion of a ‘blackouts’ subscale; however, this subscale includes hangover, passing out, and throwing up as ‘blackout’ experiences. Moreover, none of these scales differentiate blackout experiences of varying severity. Differentiating between degrees of blackout is expected to advance research by improving understanding of the prevalence, predictors, and consequences of a broader range of alcohol-induced memory impairment. To achieve these goals, a valid and reliable measure of alcohol-induced blackout is needed.

This study aimed to develop and examine the preliminary validity of a blackouts measure that quantifies the frequency of both fragmentary and en bloc blackout experiences. Consistent with Goodwin and colleagues’ (1969) characterization of blackouts, we defined EB as complete memory loss for drinking events and FB as partial memory loss that can later be pieced together, often with cues. We examined the factor structure, internal and test-retest reliability, and construct and incremental validity of the newly-developed Alcohol-Induced Blackout Measure (ABOM). Based on previous research (Goodwin et al., 1969), we hypothesized that the ABOM would have a two-factor structure reflecting the “fragmentary” and “en bloc” nature of blackout experiences. Because blackouts are expected to increase with heavier drinking (Wetherill & Fromme, 2016), we hypothesized that greater drinking quantity would be associated with greater FB and EB frequency. Moreover, given strong associations between blackouts and alcohol problems (Hingson et al., 2016), we hypothesized that greater EB and FB frequency would predict alcohol-related consequences, both cross-sectionally and prospectively.

2. Method

2.1. Participants and Procedure

All procedures were approved by Brown University’s Institutional Review Board. Young adults participated in an anonymous online survey using Qualtrics survey panels. To minimize self-selection bias, the recruitment email did not include details about survey content. Eligibility criteria included 18-29 years of age, college enrollment, and inability “to remember events that happened while you were drinking” in the past 12 months. Interested panelists (N=521) provided informed consent and completed a screening survey. Of those, 402 met screening criteria and completed the baseline survey. Approximately one month later (M=25.6, SD=12.4 days), participants were re-contacted by Qualtrics to complete a follow-up. Due to budgetary constraints, follow-up was open to the first 120 participants to complete it, 109 of whom met eligibility criteria. Participants received incentives ranging in value from $2.25 to $5.25.

Overall, 350 participants met all eligibility criteria, 109 of whom also completed the follow-up (see Table 1). Participants who completed both baseline and follow-up (n=109) were more likely than those who completed only the baseline assessment (n=241) to be female [67% vs. 51%; χ2(1)=7.35, p=.01] and non-White [34% vs. 24%; χ2(1)=4.05, p=.04], and they reported fewer drinks per week [M=10.04, SD=8.49 vs. M=12.50, SD=11.32; t(272)=2.26, p=03]. Baseline and follow-up samples did not differ significantly in age, t(348)=1.32, p=.19; alcohol-related consequences, t(348)=0.03, p=.98; or blackout frequency, t(2491)=1.27, p=.20.

Table 1.

Demographic characteristics of college students reporting alcohol-induced memory impairment in the past year.

  Baseline sample
(N=350)		   Follow-up sample
(n=109)
n % n %
Female birth sex 197 (56%) 73 (67%)
Race (White vs. non-White) -- --
 White/Caucasian 256 (73%) 72 (66%)
 Black/African American 42 (12%) 14 (13%)
 Asian 39 (11%) 23 (21%)
 Native American or Native Alaskan 5 (1%) 1 (1%)
 Native Hawaiian or Pacific Islander 1 (<1%) 0 (0%)
 Other 22 (6%) 5 (5%)
Hispanic/Latino 62 (18%) 16 (15%)
Year in school -- --
 Freshman 64 (18%) 15 (14%)
 Sophomore 76 (22%) 18 (17%)
 Junior 98 (28%) 33 (30%)
 Senior 112 (32%) 43 (39%)
Type of institution -- --
 4-year (vs. 2-year) 326 (93%) 102 (94%)
 Public (vs. private) 258 (74%) 75 (69%)
Housing -- --
 Dormitory 115 (33%) 39 (36%)
 Off-campus residence 157 (45%) 43 (39%)
 With family/spouse 68 (19%) 24 (22%)
 Fraternity/sorority house 10 (3%) 3 (3%)
Fraternity/sorority member 106 (30%) 27 (25%)
M (SD) M (SD)
Age 21.78 (2.83) 21.49 (2.56)
Drinks per week at BL 11.73 (10.57) 10.04 (8.49)
 Drinks per week at 1mo --- 9.33 (8.77)
Frequency of HID at BL 0.34 (1.00) 0.21 (0.68)
Alcohol consequences at BL1 9.05 (5.67) 9.00 (5.37)
 Alcohol consequences at 1mo1 --- 8.17 (5.87)
AUDIT score1 12.38 (6.80) 10.97 (5.42)
FB frequency at BL 4.11 (2.65) 3.53 (3.34)
 FB frequency at 1mo --- 3.64 (2.76)
EB frequency at BL 3.89 (3.86) 3.89 (2.38)
 EB frequency at 1mo --- 3.05 (2.76)
Total blackout frequency at BL 8.01 (6.19) 7.42 (5.39)
 Total blackout frequency at 1mo --- 6.69 (5.98)
Open in a new tab

Note. 1mo=1 month. BL=baseline.

1

The blackouts item was removed from this variable. AUDIT=Alcohol Use Disorders Identification Test. Mo=month.

2.2. Measures

2.2.1. ABOM Item Pool

The initial item pool included 10 items (see Table 2). The authors generated seven items based on previous descriptions of blackouts (Hartzler & Fromme, 2003; White et al., 2004), and three items were drawn from existing scales (AUDIT, RAPI, and BYAACQ). On a scale from 0 (never) to 4 (twice a week or more), participants indicated how often in the past 30 days they experienced each form of alcohol-induced memory impairment. As noted in Table 3, items 1-4 were designed as ‘fragmentary’ blackout items, and items 5-10 were piloted as ‘en bloc’ blackout items. Items were assessed at baseline and follow-up.

Table 2.

Incidence of Alcohol-Induced Blackout Measure items among college students reporting alcohol-induced memory impairment in the past year (N=350).

Response for past 30 days, n (%)
Items: In the past 30 days, how
often have you…		 Never 1
time		 2-3
times		 Weekly 2x/week
or more		 Overall
incidence
1. Had fuzzy memories of events that occurred while you were drinking 54
(15%)		 144
(41%)		 109
(31%)		 35
(10%)		 8
(2%)		 296
(85%)
2. Been unable to remember things that happened while you were drinking 77
(22%)		 128
(37%)		 108
(31%)		 27
(8%)		 10
(3%)		 273
(78%)
3. Had memories that became clear only when someone/something gave you cues or reminded you later 77
(22%)		 129
(37%)		 96
(27%)		 37
(11%)		 11
(3%)		 273
(78%)
4. Been unable to remember things that happened while you were drinking until someone reminded you about it the next day 89
(25%)		 123
(35%)		 103
(29%)		 27
(8%)		 8
(2%)		 261
(75%)
5. Been told about things you said/did that you don’t remember at all, even when people tried to remind you 120
(34%)		 121
(35%)		 65
(19%)		 32
(9%)		 12
(3%)		 230
(66%)
6. Been unable to remember an hour or more of a drinking event 121
(35%)		 114
(33%)		 82
(23%)		 25
(7%)		 8
(2%)		 229
(65%)
7. Wondered or suspected that things happened while you were drinking that you don’t remember 128
(37%)		 109
(31%)		 72
(21%)		 32
(9%)		 9
(3%)		 222
(63%)
8. Been unable to remember what happened the night before because you had been drinking (AUDIT) 138
(39%)		 111
(32%)		 61
(17%)		 29
(8%)		 11
(3%)		 212
(61%)
9. Not been able to remember large stretches of time while drinking heavily (BYAACQ) 158
(45%)		 93
(27%)		 53
(15%)		 37
(11%)		 9
(3%)		 192
(55%)
10. Suddenly found yourself in a place you don’t remember getting to (RAPI) 200
(57%)		 68
(19%)		 53
(15%)		 23
(7%)		 6
(2%)		 150
(43%)
Open in a new tab

Note. AUDIT=Alcohol Use Disorders Identification Test. BYAACQ=Brief Young Adult Alcohol Consequences Questionnaire. RAPI=Rutgers Alcohol Problem Index.

Table 3.

Results of principal component analyses for the Alcohol-Induced Blackout Measure (N=350).

Column A Column B Column C
Fixed 2
components
10 items)		 Fixed 2 components
(5 items)		 Unrestricted
(5 items)
PM PM PM PM
Abbreviated item (EB) (FB) h2 (EB) (FB) h2 CM h2 α-less
1. Fuzzy memories of drinking events −.02 .90 .78 .01 .91 .84 .83 .69 .89
2. Unable to remember things that happened .21 .69 .74 --- --- --- --- --- ---
3. Memories became clear only after cues −.04 .93 .80 −.01 .92 .85 .83 .68 .90
4. Unable to remember until reminded .46 .45 .74 --- --- --- --- --- ---
5. Told about things you don’t remember .62 .31 .77 --- --- --- --- --- ---
6. Unable to remember an hour or more .42 .48 .73 --- --- --- --- --- ---
7. Wondered or suspected .74 .14 .74 --- --- --- --- --- ---
8. Unable to remember what happened the night before .90 .01 .82 .92 .002 .84 .88 .77 .88
9. Not able to remember large stretches of time .85 .08 .82 .86 .08 .84 .89 .79 .88
10. Suddenly found yourself in a place .97 −.14 .76 .93 −.05 .80 .85 .71 .89
Cronbach’s α .94 .90 .81 .89 .91
Eigenvalues 7.2 0.6 3.7 0.5 3.7
SSL % of variance 71.5 5.5 73.0 10.3 73.0
Correlation between components .78 .74 ---
Open in a new tab

Note. Items 1-4 were designed as ‘fragmentary’ blackout items, and items 5-10 were piloted as ‘en bloc’ blackout items. h2=communality. α-less=α if item deleted. CM=component matrix. EB=en bloc blackout subscale. FB=fragmentary blackout subscale. PM=pattern matrix coefficient

2.2.2. Alcohol use

Typical alcohol use was assessed at baseline and follow-up using the Daily Drinking Questionnaire (Collins, Parks, & Marlatt, 1985). On a seven-day grid, participants estimated the number of standard drinks (e.g., 12oz regular beer) they consumed on each day of a typical week in the past month. Responses were summed to estimate number of drinks consumed per week. For descriptive purposes, responses were also used to estimate frequency of binge (4/5+ drinks for women/men) and high-intensity drinking (8/10+ drinks) in a typical week at baseline.

2.2.3. Alcohol-related consequences

Alcohol-related consequences were assessed at baseline and follow-up using the Brief Young Adult Alcohol Consequences Questionnaire (BYAACQ) (Kahler, Hustad, Barnett, Strong, & Borsari, 2008). Participants indicated (yes/no) if they had experienced 24 consequences, such as feeling sick to one’s stomach, as a result of drinking in the past month. The ‘blackouts’ item (i.e., not been able to remember large stretches of time while drinking heavily) was removed from the scale to avoid confounds between predictor and outcome variables. Participant responses to other items were summed (total scores range 0-23). This scale has been validated among college students (Kahler et al., 2008). Without the blackouts item, it demonstrated strong reliability at baseline (α=.89) and follow-up (α=.90).

2.3. Data Screening and Analysis

Of the 402 individuals who consented to participate, 18 were missing >70% data, 3 indicated random responding, and 31 did not meet inclusion criteria. These participants were excluded from analysis, resulting in a baseline sample of 350 and a follow-up sample of 109 (see Table 1 for demographics and Table 2 for blackout incidence). Outliers for baseline drinks per week (n=4) were replaced with the value three standard deviations plus one integer above the mean (Tabachnick & Fidell, 2007). After recoding outliers, skewness and kurtosis estimates for outcome variables fell within the normal range.

Data analysis was conducted in SPSS Statistics 25. The Kaiser-Meyer-Olkin (KMO) statistic and Bartlett’s test of sphericity were used to determine the appropriateness of the factor analysis. Principal component analysis (PCA) with principal components extraction was used to determine the factorial validity of the 10 ABOM items (see Table 3). Because we hypothesized a two-factor structure, the model was fixed to extract two components. Oblique (direct oblimin) rotation was used to allow the two components to correlate. Items with relatively strong cross-loadings (>.3) were removed (Costello & Osborne, 2005). To avoid redundancy, items demonstrating highly overlapping incidence rates (i.e., similar proportions of participants experiencing items never, one time, 2-3 times, weekly, or twice a week or more; see Table 2) were also removed. Consistent with recommendations (Lim & Jahng, 2019; Zwick & Velicer, 1986), parallel analysis (Horn, 1965) using 1000 datasets and the 95th percentile estimates was used to determine the number of factors to retain.

The internal reliability of items was assessed using Cronbach’s alpha, and alpha-less item analyses were used to determine if removal of any item from the total score resulted in increases in alpha. To determine the test-retest reliability of the scales, baseline scores were correlated with scores assessed one month later. Evidence for the convergent validity of the scales was derived from bivariate correlations with drinking variables (see Table 4). In addition, because blackouts increase with heavier drinking (Wetherill & Fromme, 2016), hierarchical linear regression was used to examine drinking quantity as a cross-sectional and prospective predictor of ABOM scores, controlling for potential variations in rates of heavy drinking based on age, sex, and race (Schulenberg, Johnston, O'Malley, Miech, & Patrick, 2017).

Table 4.

Bivariate correlations among study variables (N=350).

1. 2. 3. 4. 5. 6. 7.
1. Age ---
2. Female birth sex −0.39*** ---
3. White race −0.13* 0.19*** ---
4. Drinks per week (BL) 0.20*** −0.23*** 0.03 ---
5. Consequences (BL) 0.25*** −0.19*** −0.11* 0.38*** ---
6. Consequences (1mo) 0.01 −0.10 −0.13 0.16 0.59*** ---
7. ABOM score (BL) 0.26*** −0.26*** −0.14* 0.36*** 0.58*** 0.37*** ---
8. ABOM score (1mo) 0.03 −0.08 −0.08 0.29** 0.43*** 0.62*** 0.65***
Open in a new tab

Note. ***p<.001, **p<.01, *p<.05. 1mo=1 month. ABOM = alcohol-induced blackout measure. BL=baseline.

To provide evidence for incremental validity, hierarchical linear regression was used to assess the utility of ABOM scores in predicting alcohol-related consequences at baseline and one-month follow-up. Age, female versus male birth sex, White versus non-White race, and drinks per week were included as covariates in Step 1. To determine the utility of ABOM scores above and beyond the influence of existing single-item measures of alcohol-induced blackout, the BYAACQ and AUDIT blackout items were added simultaneously as predictors of alcohol-related consequences in Step 2. ABOM scores were added to each model in Step 3 (see Table 5).

Table 5.

Incremental validity of ABOM scores (N=350).

Outcome: BL BYAACQ B β p 95% CI Adj. R2 (ΔR2)
 N=350 Lower Upper
Step 1 covariates 0.18*
 Age 0.32(0.11) 0.16 .003 0.11 0.53
 Female sex −0.40 (0.62) −0.04 .52 −1.62 0.82
 White race −1.13 (0.64) −0.09 .08 −2.38 0.13
 Drinks per week 0.18 (0.03) 0.34 <.001 0.13 0.24
Step 2 predictors 0.37(0.20*)
 BYAACQ blackouts item 4.16(0.51) 0.37 <.001 3.15 5.16
 AUDIT blackouts item 1.21 (0.29) 0.20 <.001 0.64 1.78
Step 3 predictors
 ABOM score 0.45 (0.07) 0.36 <.001 0.31 0.59 0.44 (0.07*)
Outcome: 1mo
BYAACQ		 B β p 95% CI Adj. R2 (ΔR2)
 n=109 Lower Upper
Step 1 covariates 0.03
 Age −0.16(0.23) −0.07 .50 −0.61 0.30
 Female sex −0.90(1.21) −0.07 .46 −3.30 1.50
 White race −2.31 (1.29) −0.19 .08 −4.87 0.24
 Drinks per week 0.14(0.07) 0.20 .04 0.01 0.28
Step 2 predictors 0.08 (0.07*)
 BYAACQ blackouts item 1.61 (1.12) 0.14 .15 −0.62 3.84
 AUDIT blackouts item 1.70 (0.84) 0.21 .045 0.04 3.36
Step 3 predictors
 ABOM score 0.40 (0.18) 0.26 .03 0.05 0.75 0.10(0.03*)
Open in a new tab

Note. *significant R2 value at p<.05. AUDIT=Alcohol Use Disorders Identification Test. ABOM = Alcohol-Induced Blackout Measure. BYAACQ=Brief Young Adult Alcohol Consequences Questionnaire. CI=confidence interval.

3. Results

3.1. Factorial Validity

We first ran a PCA, fixed to extract two components, on the full ten items. The KMO statistic indicated that an adequate proportion of the variance among the variables was attributable to common variance (.96); and Bartlett’s test of sphericity indicated reliable associations between variables, χ2(45)=3037.14,p<001. As noted in Table 3, the two-component, 10-item solution accounted for 77.0% of total variance, with the majority of variance (71.5%) attributed to the first component. In contrast to hypotheses, parallel analysis and Kaiser’s eigenvalues >1 criterion indicated that a single- (rather than two-) component solution best fit the data. However, 3 of the 10 items demonstrated moderate to strong cross-loadings (.31-.46) across components. Similarly, items 2 and 3 and items 7 and 8 demonstrated highly overlapping incidence rates (see Table 2). Based on these response patterns, items 2, 4, 5, 6, and 7 were removed from the scale. Items 2 and 7 were removed in favor of items 3 and 8 because items 3 and 8 demonstrated stronger correlation coefficients.

We re-ran the PCA with the remaining five items (see Table 3). The two-component solution accounted for 83.3% of total variance, again with the majority of variance (73.0%) attributed to the first component. As noted in Table 3, the pattern matrix differentiated between the two hypothesized components for FB and EB. All items demonstrated strong correlations with their hypothesized component (loadings≥.70) and low cross-loadings on the second component (≤08). However, the two components were highly correlated (r=.74), and again, parallel analysis and eigenvalues (see Table 3) indicated a single-component solution. Results indicated that only one component represented eigenvalues larger than those that might occur by chance. Thus, we ran another PCA that did not restrict the number of components to extract.

In the third PCA (see Table 3), KMO (.87) and Bartlett’s statistics, χ2(10)=1128.81, p<.001, were again within the appropriate range. Results and parallel analysis suggested a single component accounting for 73.0% of variance. All items demonstrated strong correlations with the principal component (loadings≥83), and a large portion of variance in each item was explained by common variance (all h2≥.68). Because a single component best fit the data, the total score (calculated by summing all five items of the scale) was used in subsequent analyses.

3.2. Reliability and Validity

3.2.1. Internal consistency

The five ABOM items demonstrated strong internal consistency at baseline (α=.91) and follow-up (α=.89). Reliability did not increase with removal of any item (see α-less estimates, Table 3).

3.2.2. Test-retest reliability

In the subsample who completed the one-month follow-up (n=109), ABOM scores at baseline (M=5.11, SD=3.87) were moderately correlated with ABOM scores one month later (M=4.72, SD=4.27), r(109)=0.65, p<.001, and did not significantly differ, t(108)=1.21, p=.23.

3.2.3. Convergent validity

ABOM scores were significantly correlated with drinking quantity and alcohol-related consequences at baseline and one month (see Table 4). ABOM scores were also correlated with indicators of heavy drinking, specifically frequency of binge [r(350)=0.23, p<.001] and high-intensity drinking [r(350)=0.21, p<.001]. Multivariate models examining drinks per week as a predictor of ABOM score were significant, both cross-sectionally, F(4,345)=20.61, p<.001, Adj. R2=.18, and prospectively, F(4,104)=3.33, p=.01, Adj. R2=.08. In the cross-sectional model, greater drinks per week (B=0.13, SE=0.02, p<.001) and older age (B=0,22, SE=0.08, p=.01) were associated with more frequent blackouts, while female birth sex (B=−1.07, SE=0.49, p=.03) and White race (B=−1.03, SE=0.50, p=.04) were associated with less frequent blackouts. In the prospective model, drinks per week was the only significant predictor of ABOM score at one-month follow-up (B=0.17, SE=0.05, p=.001).

3.2.4. Incremental validity

Regression coefficients and adjusted R2 values for models examining ABOM score as a predictor of alcohol-related consequences are presented in Table 5. ABOM score was a cross-sectional predictor of alcohol-related consequences. Step 1 covariates accounted for a significant amount of variance in alcohol-related consequences at baseline, F(4,345)=19.71, p<.001, Adj. R2=.18. In Step 2, existing single blackout items accounted for an additional 20% of variance in baseline consequences, F(6,343)=35.28, p<.001, Adj. R2=.37. In Step 3, ABOM score accounted for an additional 7% unique variance, F(7,342)=39.95, p< 001, Adj. R2=.44.

ABOM score at baseline was also a unique predictor of alcohol-related consequences at one-month follow-up (see Table 5). In contrast to previous models, Step 1 predictors did not account for a significant amount of variance in alcohol-related consequences at one-month follow-up, F(4,104)=1.75, p=.15, Adj. R2=.03. In Step 2, existing blackout items accounted for 7% unique variance, F(6,102)=2.49, p=.03, Adj. R2=.08. In Step 3, ABOM score accounted for an additional 4% unique variance, F(7,101)=2.94, p=.01, Adj. R2=.11.

4. Discussion

The Alcohol-Induced Blackout Measure (ABOM) is a brief and reliable measure of alcohol-induced memory impairment that quantifies the frequency of both fragmentary and en bloc blackout experiences. It contributes to the literature by improving the predictive validity of single-item blackout measures and capturing a range of blackout experiences. Although items did load on two separate scale components that were consistent with empirically-derived characterizations of FB and EB (Goodwin et al., 1969), a single-component model best fit the data. We speculate that this discrepancy occurred because heavy-drinking young adults describe individual blackout events as more ‘fragmentary’ or ‘en bloc’ in nature but have a range of blackout experiences over a 30-day period. A single-component solution is also consistent with young adults’ self-reported perceptions of blackout experiences. Specifically, in a qualitative study examining young adults’ blackout experiences, college students reported perceiving blackouts as occurring “along a continuum,” with EB “at the extreme” and FB encompassing “any other memory loss that’s not complete” (Miller, Merrill, et al., 2018). Thus, the unidimensional structure of the ABOM may reflect the dimensional versus categorical nature of these experiences.

Internal consistency and test-retest reliability estimates indicate that, among young adults with a history of blackout, scale items are also reliable and relatively stable over time. This may seem counterintuitive, given the relative infrequency of blackouts in a 30-day period (Wetherill & Fromme, 2016). However, results from this study suggest that the blackout items used in previous measures (i.e., AUDIT, BYAACQ, RAPI) likely capture EB, as opposed to FB, experiences. This may result in an underestimation of blackouts, as FB are much more common (Hartzler & Fromme, 2003; Miller, Merrill, et al., 2018). Because the ABOM assesses both FB and EB experiences, it may be a more sensitive measure of blackouts over time.

ABOM scores also demonstrated construct validity, as higher scores were positively associated with drinking quantity. This is consistent with research indicating that likelihood of blackouts should parallel drinking (Wetherill & Fromme, 2016; White, 2003). Demographic correlates should be interpreted cautiously, as demographic variables were concurrent but not prospective predictors of blackout frequency in this sample. However, in this sample, male birth sex was a concurrent predictor of blackout frequency in the past 30 days. Research regarding the prevalence of blackouts among men and women has been mixed, with some studies failing to find sex differences (Barnett et al., 2014; Merrill et al., 2016; Mundt, Zakletskaia, Brown, & Fleming, 2012; Wilhite & Fromme, 2015), some reporting higher rates of blackouts among women than men (Bonar et al., 2019; Hingson et al., 2016; Schuckit et al., 2015), and some reporting higher rates among men than women (Chartier, Hesselbrock, & Hesselbrock, 2011; Nelson et al., 2004). Inconsistent findings may be due in part to lack of uniformity in blackout assessments. Alternatively, there is some evidence that women are more likely than men to reduce their drinking as a result of blackout experiences (Read, Wardell, & Bachrach, 2013). In that case, women who reported a blackout at baseline may experience fewer blackouts over the next 30 days because they are more likely than men to reduce their drinking as a result of those experiences. Future research is needed to support or refute this hypothesis.

Consistent with hypotheses, blackout frequency as measured by the ABOM was a predictor of alcohol-related consequences, above and beyond the influence of drinking quantity and existing single-item measures of blackout. This is consistent with previous studies (Hingson et al., 2016; Mundt et al., 2012) in demonstrating something unique about blackouts – beyond heavy drinking – that contributes to alcohol-related problems. Importantly, our study confirms this association both cross-sectionally and prospectively. It is possible that, in the moment, the inability to retain memories for more than a few minutes limits individuals’ awareness and/or comprehension of danger in their environments. Over time, this may also limit their ability to learn about factors leading to negative outcomes, increasing their risk of negative consequences. It is also possible that items assessing blackouts (as opposed to drinking quantity) capture the “speed” of drinking that is not reflected in estimates of drinking quantity alone. Alternatively, a number of behaviors increase risk for blackout (e.g., use of other drugs) and may inherently increase risk for other alcohol-related consequences, independent of drinking quantity. Research is needed to determine the extent to which alcohol-induced memory impairment is a proxy for other high-risk drinking behaviors versus a risk factor on its own.

Given the association between blackouts and other alcohol-related consequences, blackout assessments may be a valuable screening tool in clinical settings. Several researchers have suggested that blackout screening items be used to identify individuals at risk for other alcohol-related problems (Hingson et al., 2016; Merrill et al., 2016; Mundt et al., 2012; Wilhite & Fromme, 2015). Research also suggests that young adults who have recently experienced memory loss for large stretches of time respond more favorably to brief alcohol interventions than those who have not (Miller, DiBello, Carey, & Pedersen, 2018; Miller, DiBello, Meier, et al., 2018). Given the strong association between blackout frequency and alcohol problems in this and other samples, additional research examining the cut-off score that most accurately identifies high-risk individuals is encouraged. However, until such research is conducted, we recommend that the scale be used to quantify frequency of blackout experiences.

4.1. Limitations

Several limitations of this study are worth noting. First, analyses were based on self-reported data, where social desirability or confidentiality could be of concern. However, anonymous data collection minimizes these issues. Second, given our sample size and inclusion of only 10 survey items, we were unable to conduct more advanced analyses (e.g., item response theory) that would allow for examination of item functioning characteristics. This study was also not designed to replicate our factor structure in an independent sample (Floyd & Widaman, 1995); thus, cross-validation using confirmatory factor analyses would enhance confidence in the generalizability of this factor structure. Because our sample consisted of college students who had a history of alcohol-induced blackout, cross-validation is also needed before the scale is used in more general samples. Similarly, our follow-up sample contained a disproportionate number of females and non-White participants; thus, results should be replicated in more diverse samples. The timeframe for the measure (past 30 days) may also need to be extended for samples containing lighter drinkers, as only 20% of young adults who drink report “forgetting where you were or what you did while drinking” in the past 6 months (Hingson et al., 2016). Finally, this study did not assess contextual factors that may influence both blackouts and other alcohol-related consequences (e.g., other drug use, food consumption, cognitive impairment). Future research is needed to determine the extent to which these factors modify the association between blackouts and alcohol-related harm.

4.2. Conclusion

This study fills a gap in the literature by developing a brief and reliable measure of alcohol-induced memory impairment that will allow researchers to quantify the frequency of a range of blackout experiences in the past 30 days. Accounting for this range of experiences improved predictive validity over single-item blackout measures. Heavier drinking was associated with greater blackout frequency, which in turn was associated with alcohol-related consequences. Moreover, associations between blackout frequency and alcohol-related consequences occurred independent of demographic covariates and drinking quantity, indicating that there is something unique about blackouts that contributes to alcohol problems. Additional research examining the cut-off score that most accurately identifies individuals in need of intervention is warranted.

Table 6.

The Alcohol-Induced Blackout Measure (ABOM).

In the past 30 days, as a result of alcohol use, how often have you:
Never 1 time 2-3
times		 Weekly Twice a week
or more
1. Had fuzzy memories of events that occurred while you were drinking? 0 1 2 3 4
2. Had memories that became clear only when someone/something gave you cues or reminded you later? 0 1 2 3 4
3. Been unable to remember what happened the night before because you had been drinking? 0 1 2 3 4
4. Not been able to remember large stretches of time while drinking heavily? 0 1 2 3 4
5. Suddenly found yourself in a place you don’t remember getting to? 0 1 2 3 4
Open in a new tab

Highlights.

  • The ABOM is a reliable measure of alcohol-induced memory impairment.

  • A single-component scale solution best fit the data.

  • ABOM scores were positively associated with drinking quantity.

  • ABOM scores predicted alcohol consequences, both concurrently and prospectively.

Acknowledgments

This work was supported in part by a Research Excellence Award (PI Miller) from the Center for Alcohol and Addiction Studies at Brown University. Jennifer Merrill’s contribution to this project was supported by the National Institute on Alcohol Abuse and Alcoholism (grant number K01AA022938). NIH had no role in study design; data collection, analysis, or interpretation; manuscript preparation; or the decision to submit the paper for publication. The authors have no conflicts of interest to report.

Footnotes

Conflict of Interest

The authors have no conflicts of interest.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Barnett NP, Clerkin EM, Wood M, Monti PM, Tevyaw TOL, Corriveau D, … Kahler CW (2014). Description and predictors of positive and negative alcohol-related consequences in the first year of college. Journal of Studies on Alcohol and Drugs, 75(1), 103–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bonar EE, Goldstick JE, Cunningham RM, Fernandez AC, Davis AK, Ilgen MA, & Walton MA (2019). Individual and social factors related to trajectories of blackouts among underage drinkers in the emergency department. Alcohol and Alcoholism, 1–8. doi: 10.1093/alcalc/agy087 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chartier KG, Hesselbrock MN, & Hesselbrock VM (2011). Alcohol problems in young adults transitioning from adolescence to adulthood: The association with race and gender. Addictive Behaviors, 36, 167–174. doi: 10.1016/j.addbeh.2010.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Collins RL, Parks GA, & Marlatt GA (1985). Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology, 53, 189–200. doi: 10.1037/0022-006X.53.2.189 [DOI] [PubMed] [Google Scholar]
  5. Costello AB, & Osborne JW (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research, & Evaluation, 10, 1–9. [Google Scholar]
  6. Fabrigar LR, & Wegener DT (2012). Exploratory Factor Analysis: Understanding Statistics. New York, NY: Oxford University Press. [Google Scholar]
  7. Floyd FJ, & Widaman KF (1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7, 286–299. doi: 10.1037/1040-3590.7.3.286 [DOI] [Google Scholar]
  8. Goodwin DW, Crane JB, & Guze SB (1969). Phenomenological aspects of the alcoholic 'blackout'. The British Journal of Psychiatry, 115, 1033–1038. doi: 10.1192/bjp.115.526.1033 [DOI] [PubMed] [Google Scholar]
  9. Hartzler B, & Fromme K (2003). Fragmentary and en bloc blackouts: Similarity and distinction among episodes of alcohol-induced memory loss. Journal of Studies on Alcohol, 64, 547–550. [DOI] [PubMed] [Google Scholar]
  10. Hingson RW, Zha W, Simons-Morton B, & White A (2016). Alcohol-induced blackouts as predictors of other drinking related harms among emerging young adults. Alcoholism: Clinical and Experimental Research, 40, 776–784. doi: 10.1111/acer.13010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Horn JL (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 32, 179–185. doi: 10.1007/BF02289447 [DOI] [PubMed] [Google Scholar]
  12. Kahler CW, Hustad J, Barnett NP, Strong DR, & Borsari B (2008). Validation of the 30-day version of the Brief Young Adult Alcohol Consequences Questionnaire for use in longitudinal studies. Journal of Studies on Alcohol and Drugs, 69, 611–615. doi: 10.15288/jsad.2008.69.611 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Lim S, & Jahng S (2019). Determining the number of factors using parallel analysis and its recent variants. Psychological Methods, Advance online publication. doi: 10.1037/met0000230 [DOI] [PubMed] [Google Scholar]
  14. Marino EN, & Fromme K (2018). Alcohol-induced blackouts, subjective intoxication, and motivation to decrease drinking: Prospective examination of the transition out of college. Addictive Behaviors, 80, 89–94. doi: 10.1016/j.addbeh.2018.01.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Merrill JE, Treloar H, Fernandez AC, Monnig MA, Jackson KM, & Barnett NP (2016). Latent growth classes of alcohol-related blackouts over the first two years of college. Psychology of Addictive Behaviors, 30, [epub ahead of print]. doi: 10.1037/adb0000214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Miller MB, DiBello AM, Carey KB, & Pedersen ER (2018). Blackouts as a moderator of young adult Veteran response to personalized normative feedback for heavy drinking. Alcoholism: Clinical and Experimental Research, 42, 1145–1153. doi: 10.111/acer.13637 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Miller MB, DiBello AM, Meier E, Leavens ELS, Merrill JE, Carey KB, & Leffingwell TR (2018). Alcohol-induced amnesia and personalized drinking feedback: Blackouts predict intervention response. Behavior Therapy, 50, 25–35. doi: 10.1016/j.beth.2018.03.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Miller MB, Merrill JE, DiBello AM, & Carey KB (2018). Distinctions in alcohol-induced memory impairment: A mixed methods study of en bloc versus fragmentary blackouts. Alcoholism: Clinical and Experimental Research, 42, 2000–2010. doi: 10.1111/acer.13850 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mundt MP, Zakletskaia LI, Brown DD, & Fleming MF (2012). Alcohol-induced memory blackouts as an indicator of injury risk among college drinkers. Injury Prevention, 18, 44–49. doi: 10.1136/ip.2011.031724 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Nelson EC, Heath AC, Bucholz KK, Madden PAF, Fu Q, Knopik VS, … Martin NG (2004). Genetic epidemiology of alcohol-induced blackouts. Archives of General Psychiatry, 61, 257–263. [DOI] [PubMed] [Google Scholar]
  21. Read JP, Kahler CW, Strong DR, & Colder CR (2006). Development and preliminary validation of the young adult alcohol consequences questionnaire. Journal of Studies on Alcohol, 67, 169–177. [DOI] [PubMed] [Google Scholar]
  22. Read JP, Wardell JD, & Bachrach RL (2013). Drinking consequence types in the first college semester differentially predict drinking the following year. Addictive Behaviors, 38, 1464–1471. doi: 10.1016/j.addbeh.2012.07.005 [DOI] [PubMed] [Google Scholar]
  23. Saunders JB, Aasland OG, Babor TF, de la Fuente JR, & Grant M (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption. II. Addiction, 88, 791–804. doi: 10.1111/j.1360-0443.1993.tb02093.x [DOI] [PubMed] [Google Scholar]
  24. Schuckit MA, Smith TL, Heron J, Hickman M, Macleod J, Munafo MR, … Davey-Smith G (2015). Latent trajectory classes for alcohol-related blackouts from age 15 to 19 in ALSPAC. Alcoholism: Clinical and Experimental Research, 39, 108–116. doi: 10.1111/acer.12601 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Schulenberg JE, Johnston LD, O'Malley PMB, J. G., Miech RA, & Patrick ME (2017). Monitoring the Future national survey results on drug use, 1975-2016: Volume II, college students and young adults ages 19-55. Ann Arbor: Institute for Social Research: The University of Michigan. [Google Scholar]
  26. Tabachnick BG, & Fidell LS (2007). Using Multivariate Statistics (5th Ed.). New York, NY: Haper and Row. [Google Scholar]
  27. Wetherill RR, & Fromme K (2016). Alcohol-induced blackouts: A review of recent clinical research with practical implications and recommendations for future studies. Alcoholism: Clinical and Experimental Research, 40, 922–935. doi: 10.1111/acer.13051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. White AM (2003). What happened? Alcohol, memory blackouts, and the brain. Alcohol Research & Health, 27, 186–196. [PMC free article] [PubMed] [Google Scholar]
  29. White AM, Signer ML, Kraus CL, & Swartzwelder HS (2004). Experiential aspects of alcohol-induced blackouts among college students. The American Journal of Drug and Alcohol Abuse, 30, 205–224. doi: 10.1081/ADA-120029874 [DOI] [PubMed] [Google Scholar]
  30. White HR, & Labouvie EW (2000). Longitudinal trends in problem drinking as measured by the Rutgers Alcohol Problem Index. Alcoholism: Clinical and Experimental Research, 24(76A). [Google Scholar]
  31. Wilhite ER, & Fromme K (2015). Alcohol-induced blackouts and other negative outcomes during the transition out of college. Journal of Studies on Alcohol and Drugs, 76, 516–524. doi: 10.15288/jsad.2015.76.516 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Zwick WR, & Velicer WF (1986). Comparison of five rules for determining the number of components to retain. Psychological Bulletin, 99(432–442). [Google Scholar]

RESOURCES