Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jan 1.
Published in final edited form as: AIDS Behav. 2016 Jan;20(0 1):19–39. doi: 10.1007/s10461-015-1108-9

Alcohol Use Predicts Sexual Decision-Making: A Systematic Review and Meta-Analysis of the Experimental Literature

Lori A J Scott-Sheldon a,b,c, Kate B Carey d, Karlene Cunningham a,b,e, Blair T Johnson e, Michael P Carey a,b,c; the MASH Research Teama,1
PMCID: PMC4683116  NIHMSID: NIHMS710232  PMID: 26080689

Abstract

Alcohol is associated with HIV and other sexually transmitted infections through increased sexual risk-taking behavior. Establishing a causal link between alcohol and sexual behavior has been challenging due to methodological limitations (e.g., reliance on cross-sectional designs). Experimental methods can be used to establish causality. The purpose of this meta-analysis was to evaluate the effects of alcohol consumption on unprotected sex intentions. We searched electronic bibliographic databases for records with relevant keywords; 26 manuscripts (k = 30 studies) met inclusion criteria. Results indicate that alcohol consumption is associated with greater intentions to engage in unprotected sex (d+s = 0.24, 0.35). The effect of alcohol on unprotected sex intentions was greater when sexual arousal was heightened. Alcohol consumption is causally linked to theoretical antecedents of sexual risk behavior, consistent with the alcohol myopia model. Addressing alcohol consumption as a determinant of unprotected sex intentions may lead to more effective HIV interventions.

Keywords: alcohol, sexual behavior, unprotected sex, experimental methods, meta-analysis

INTRODUCTION

Alcohol consumption and related alcohol problems have been implicated as prominent risk factors in the transmission of HIV and other sexually transmitted infections (STIs) through increased sexual risk behaviors. A large body of scientific research shows that consuming alcohol is associated with increased sexual activity, heightened sexual arousal, and increased sexual risk behaviors.1-4 Most of these studies correlate typical patterns of alcohol use and sexual behavior (global association studies) or patterns of behavior during certain time periods (situational association studies), finding strong support for the association. These studies, however, only provide indirect support of the alcohol-risky sex association as causality cannot be established due to methodological limitations (e.g., cross-sectional designs, general measures of alcohol and sex, retrospective data).4,5 Studies measuring alcohol consumption during specific sexual events (i.e., event-level studies) establish a temporal link between alcohol and sexual risk behavior, representing a methodological improvement over global and situational association studies, but also rely largely on retrospective measures of behavior. The most recent advancements in the measurement of alcohol use and sexual risk behavior (i.e., ecological momentary assessments6) assess alcohol and sex behaviors in real time, providing more accurate assessments of alcohol use in the context of sex; however, few studies use real-time assessments of alcohol and sexual risk behavior. Thus, establishing a causal link between alcohol consumption and sexual risk behavior using the current scientific literature, which relies largely on retrospective reports, is challenging. Laboratory studies that experimentally manipulate alcohol consumption can improve our ability to establish causality.

Two theoretical models explain how alcohol consumption causally disinhibits sexual behaviors: the alcohol myopia model and outcome expectancy theory. The alcohol myopia model7 posits that alcohol consumption restricts cognitive capacity such that, when intoxicated, people are less able to attend to, or process, multiple situational cues; instead, intoxicated persons focus solely on the most salient cues. When exposed to a sexual stimulus, for example, the most salient cues are likely to involve sexual arousal; when individuals are under the influence of alcohol, more subtle cues of potential sexual risk, such as a partner's level of disease risk, may not be considered. Therefore, the salient cues of sexual arousal may circumvent the subtle cues to use condoms.

In contrast to the pharmacological effects of alcohol emphasized by alcohol myopia theory, alcohol expectancy theory emphasizes the psychological effects of alcohol consumption on sexual risk behavior. Alcohol expectancy theory proposes that behavior during or following alcohol consumption is strongly influenced by outcome expectancies associated with consuming alcohol (i.e., a self-fulfilling prophecy).8,9 In an experimental study, intentions to engage in unprotected sex may be influenced by perceived alcohol consumption (i.e., told they would receive alcohol but did not) if, for example, an individual holds positive expectancies regarding alcohol's ability to enhance a sexual situation.

Experimental methods can test both the pharmacological (myopia) and psychological (expectancy) effects of alcohol. Experimental studies administer beverages, often using a placebo or a balanced placebo design10,11 to randomly assign participants to an alcohol, placebo, or no alcohol control condition. Some participants are told that they will receive alcohol but are either given alcohol (both expect and receive alcohol) or not (expect alcohol, receive placebo). Another group of participants are told that they will receive a non-alcoholic beverage and are indeed given a non-alcoholic beverage (control). These experimental designs allow for a test of the pharmacological (myopia) or psychological (expectancy) effects of alcohol consumption by comparing the (a) alcohol vs. placebo or no alcohol control, and the (b) placebo vs. no alcohol control groups, respectively. Random assignment of participants to alcohol conditions strengthens causal inferences.5

Experimental studies have distinctive aspects (e.g., alcohol dose, sexual arousal manipulations) but use the same basic procedures. The procedures typically involve alcohol administration as described above, exposure to a sexual stimulus (e.g., written vignette, video), and measures of the perceived likelihood or intentions to engage in unprotected sex with a hypothetical partner portrayed in the sexual stimulus. Although experimental designs cannot establish a direct link between alcohol use and enacted sexual behavior, alcohol consumption can be experimentally manipulated and intentions to engage in unprotected sexual behavior can be measured contemporaneously. We can infer a causal link between alcohol use and sexual behavior using experimental approaches given the large body of research demonstrating that intentions to use a condom is one of the strongest predictors of subsequent condom use.12,13 Furthermore, experimental methods can manipulate and control for “third variables” (i.e., moderators of the alcohol-risky sex relation such as gender and alcohol dose) that may moderate the association.5,14 For example, a previous study manipulated partner risk and found no interaction between alcohol condition (alcohol, placebo, or control) and perceived partner riskiness (low- vs. high-risk) on unprotected sex intentions.15 Thus, experimental approaches establish a direct link between alcohol consumption and theoretical antecedents of sexual behavior (i.e., intentions to engage in unprotected sex) while controlling for potential moderators of the relation.

Prior reviews of the experimental literature generally support a causal relation between alcohol consumption and theoretical antecedents of sexual behavior (e.g., sexual arousal, likelihood of unprotected sex).5,8,9,16,17 A recent meta-analysis of 12 experimental studies also found support for a causal association between alcohol consumption and intentions for unprotected sex but limited their review to studies that randomly assigned participants to alcohol and placebo conditions and assessed intentions to engage in unprotected sex.18 (A single study was excluded from their analyses due to insufficient statistical information.15) By excluding other comparison conditions (i.e., no alcohol control), Rehm et al.18 could not test the pharmacological (myopia) and psychological (expectancy) effects of alcohol. Furthermore, the previous meta-analysis assessed only two potential moderators (i.e., gender, sample source), neither of which significantly moderated the relation between alcohol and unprotected sex intentions, and did not assess potential moderators of the alcohol effects (i.e., experimental manipulations of the alcohol dose). Because the previous meta-analysis was based on a small sample of studies, it may have been underpowered to detect moderators of intervention efficacy.19 Therefore, the primary purpose of the present systematic review and meta-analysis was to evaluate the experimental literature assessing the effects of alcohol consumption on sexual decision-making more broadly (i.e., unprotected sex intentions, sexual communication and negotiation skills). A secondary goal was to assess whether alcohol consumption influenced other antecedents of sexual behavior (i.e., sexual arousal, sexual communication and negotiation intentions). We extend the prior meta-analytic review by including studies assessing a no alcohol control condition (to evaluate the effects of expectancies), evaluating other antecedents of sexual behavior (i.e., sexual arousal, sexual communication and negotiation skills and intentions), and examining alcohol administration procedures as potential moderators of the alcohol effects.

We included in our review all experimental studies that manipulated alcohol consumption by randomly assigning participants to an alcohol condition (i.e., alcohol, placebo, no alcohol control), administering alcohol prior to the sexual stimulus (e.g., role-plays, erotic video), and assessing sexual outcomes (e.g., intentions to engage in unprotected sex, sexual communication and negotiation skills) following exposure to the sexual stimulus. We evaluated both the pharmacological (myopia) and psychological (expectancy) effects of alcohol by creating three group contrasts: (a) alcohol vs. placebo, (b) alcohol vs. no alcohol control, and (c) placebo vs. control. Consistent with alcohol myopia theory, participants who consumed alcohol should report stronger sexual arousal, weaker sexual communication and negotiation skills, and stronger intentions to engage in unprotected sex compared to participants exposed to a placebo or no alcohol control. Consistent with alcohol expectancy theory, participants who received a placebo should report stronger sexual arousal, weaker sexual communication and negotiation skills, and stronger intentions to engage in unprotected sex compared to participants in a no alcohol control group. Thus, the two theories lead to different hypotheses regarding the relationship of the placebo group to the no alcohol control group. Finally, we examined the extent to which participant characteristics or experiment features moderated the causal relation between alcohol consumption and sexual risk-taking.

METHODS

Overview

This systematic review and meta-analysis follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.20 The PRISMA Checklist can be found in the Supplemental Digital Content.

Search Strategy

We searched multiple electronic reference databases (PubMed, PsycINFO, ProQuest Dissertations and Theses Full Text, CINAHL, ERIC, Global Health, SocIndex, The Cochrane Library, and Web of Science [social sciences and science citation indices]) using a Boolean search strategy: (alcoholic OR alcoholic beverages OR alcohol OR ethanol OR intoxi* OR drink* OR drunk* OR binge) AND (unprotected OR condom OR sex OR sexu* OR “sexual behavior” OR “sexual behaviour” OR risky OR “risky sex” OR “risk behavior” OR “risky behavior” OR “risk behaviour” OR “risky behaviour”) AND ((experiment* OR laboratory* OR manipulat* OR expect* OR myopi*) AND (theor* OR hypoth*). Our search statement was developed with the assistance of an expert health sciences librarian in the Alpert Medical School of Brown University. Because many electronic databases have specific search methods (e.g., Medical Subject Heading [MeSH] terms used in PubMED are not available in other databases such as PsycINFO), our basic search strategy was modified based on the specific search requirements for each electronic database. (The full details of the search are available upon request from the first author.) No language restrictions were applied. All electronic reference database searches were conducted in July 2014 and updated through December 2014. The search was conducted by two authors [LAJSS, PW].

Eligibility Criteria and Study Selection

All records (i.e., titles and abstracts) retrieved from our electronic database searches were initially screened for inclusion by three authors [LAJSS, PW, AC]. Studies were included if participants (1) were randomly assigned to an alcohol condition and a comparison group (i.e., placebo or no alcohol control) and (2) exposed to a sexual stimulus (e.g., role-play, videos, vignette); and the studies (3) evaluated sexual decision-making (e.g., likelihood of unprotected sex, objective measure behavioral skills in communicating about sex and/or condoms); (4) provided sufficient information to calculate effect sizes; and (5) were available (including electronic publications and dissertations) through December 2014. If insufficient information was available to calculate effect sizes, emails requesting the missing data were sent to the study author(s). (Of the six study authors contacted, five provided the requested information.) Full-text manuscripts of potentially relevant records were retrieved and reviewed for final inclusion. Reference sections of relevant manuscripts (including published reviews obtained through the electronic reference database searches) were also reviewed. Full manuscripts (or the relevant studies within manuscripts) that fulfilled the inclusion criteria were included. When authors reported details, ancillary information (e.g., results from the pilot study), and/or outcomes of a study in multiple manuscripts, the manuscripts were linked in the database and represented as a unit. The manuscript reporting the most complete study sample was selected as the primary manuscript. The final sample included 26 manuscripts reporting 30 studies (Figure 1).15,21-45

Figure 1.

Figure 1

Study Selection and Retrieve Process for the Meta-Analysis.

Coding and Reliability

Two independent coders [AC, C-WL, BM, or PW] extracted study information (e.g., publication year), sample characteristics (e.g., gender, ethnicity), design (e.g., recruitment strategy) and measurement specifics, and alcohol condition procedures (e.g., timing of blood alcohol concentration [BAC] assessment, target BAC) from each experiment. Experimental procedures were assessed using 14 items obtained from a prior meta-analytic review of the experimental literature evaluating the effects of alcohol consumption on aggression.46 Methodological quality was assessed using 17 items (e.g., random assignment) from validated measures;47-49 total possible quality score is 25. If data from the same study sample appeared in multiple publications, the most comprehensive manuscript was coded while supplementing missing data from the other manuscript(s). Inter-rater reliability was assessed for all study, sample, and methodological variables. For the categorical variables, raters agreed on 89% of the judgments (mean Cohen's κ = .80). Reliability for the continuous variables yielded an average intra-class correlation coefficient (ρ) of 0.71 across categories (median = 0.96). Disagreements between pairs of coders were resolved through discussion.

Study Outcomes and Effect Size Calculations

Effect sizes were calculated from studies’ post-experiment assessments of sexually-related outcomes (i.e., decision-making, arousal). Sexual decision-making included (a) likelihood of unprotected sex (i.e., intentions to have [unprotected] sex, likelihood of [unprotected] sex) and (b) sexual communication and negotiation (i.e., objectively measured behavioral skills or self-reported intentions) with a hypothetical partner portrayed in the sexual stimuli. We also calculated effect sizes for (c) sexual arousal (i.e., self-reported and physiological genital response). All study experiments assessed sexual decision-making (29: intentions or likelihood of unprotected sex; 4: sexual communication and negotiation skills; 6: intentions to communicate about sex). Of the fifteen studies reporting sexual arousal, all included a self-report measure of arousal and four of these studies also measured physiological genital responses.

Effect sizes (d) were calculated as the mean difference between the alcohol and comparison group (placebo or no alcohol control) divided by the pooled standard deviation.50 In the absence of means and standard deviations, other statistical information (e.g., t- or F-values) were used.19,51 For studies reporting a correlation coefficient (e.g., point-biserial or product-moment correlation) between two variables of interest (e.g., received alcohol [dichotomous] and likelihood of unprotected sex [continuous]), we computed the standardized mean difference using published formulas.19 If a study reported dichotomous outcomes (e.g., frequencies), we calculated an odds ratio and transformed it to d using the Cox transformation.52 If no statistical information was available (and could not be obtained from the authors) and the study reported a non-significant or significant between-group difference, we estimated that effect size to be zero or calculated an effect size based on the minimum statistically significant P-value (i.e., P = .05), respectively.19 (Only 6% of the effect sizes were estimated. Analyses excluding the estimated effect sizes [data not shown] revealed the same pattern of results; therefore, these studies were retained in the final analyses.) All effect sizes were corrected for sample size bias (Hedges’ g);53 positive effect sizes indicated that participants in the alcohol condition were more likely to intend to engage in unprotected sex, had lower sexual communication and negotiation skills or intentions, and were more sexually aroused compared to controls. Effect sizes for each study were calculated by two independent coders [LAJSS and KC or XT], reviewed, and discrepancies were resolved through discussion and final calculations.

Statistical Analyses

All analyses evaluated the pharmacological (myopia) and psychological (expectancy) effects of acute alcohol consumption. Contrasts were created to indicate when participants received alcohol: (1) Alcohol vs. Placebo, (2) Alcohol vs. No Alcohol Control, (3) Alcohol vs. Placebo/No Alcohol Control and, when they expected alcohol: (4) Placebo vs. No Alcohol Control. Therefore, all analyses were conducted separately for the four contrasts (i.e., Alcohol vs. Placebo, Alcohol vs. No Alcohol Control, Alcohol vs. Placebo/No Alcohol Control, and Placebo vs. No Alcohol Control) when available.

Estimation of the Overall Mean Effect Sizes

Data analyses were conducted with Stata 12.154 using published macros.19,55 Weighted mean effect sizes were calculated using random-effects procedures.19 The 95% confidence intervals (CIs) surrounding the weighted mean effect size were calculated; CIs indicate the degree of precision as well as the significance of the mean effect size.19 Heterogeneity in effect sizes was identified by computing Q and the associated degrees of freedom; a significant Q indicates a lack of homogeneity and an inference of heterogeneity. To assess the extent to which outcomes were consistent across studies, the I2 index and its corresponding 95% CIs were calculated.56,57 I2 varies between 0 (homogeneous) and 100% (heterogeneous).58 If the CIs around I2 include a zero, the set of effect sizes is considered homogeneous.

Moderator Tests

To explain variability in the effect sizes, moderator analyses were conducted using two methods: (a) a modified weighted regression analysis (using the metareg macro in Stata, following methods of moment [mm] procedures to estimate the additive [between study] component of variance)59 or (b) a meta-analytic analogue to the ANOVA (using the metaf macro in Stata, following mm procedures to estimate the between-study variance)19 with weights equivalent to the inverse of the variance plus the random variance component for each effect size.19,59,60. Regression analyses examined a priori determined moderators. Participant (proportion women, age, and current drinking as determined by the number of drinks consumed per week) and experimental (e.g., alcohol dose, target BAC; for complete list, see Table II) characteristics were examined as potential moderators of the weighted mean effect sizes by contrast.

Table II.

Characteristics of the 37 alcohol conditions included in the meta-analysis

Variable K % or M (SD)
Fasted from alcohol and/or other substances 37 73
Hours fasted from alcohol and/or other substances 26 24 (0)
Fasted from food 37 76
Hours fasted from food 28 3.21 (0.42)
Time of day alcohol consumed 3
    Afternoon 67
    Afternoon and Evening 33
Alcohol Dose
    Units (g/kg) of absolute alcohol administered 28 .69 (.16)
    Percent absolute alcohol in beverage 31 50 (16)
    Ratio of alcohol to non-alcohol 34
        1:2 12
        1:3 18
        1:4 65
        1:6 6
Type of alcohol beverage 31
    Absolute alcohol 10
    Vodka 87
    Other distilled spirits 3
Time allotted for alcohol ingestion (minutes) 37 20.84 (17.40)
Time allotted for alcohol absorption (minutes) 22 20.27 (25.84)
Target BAC 35 .08 (.01)
Actual BAC 32 .07 (.01)
BAC assessed 37
    None/not reported 3
    Before task/manipulation 3
    After task/manipulation 16
    Before and after task/manipulation 78
Environment 37
    Laboratory 95
    Contrived Bar 5
Surveillance during alcohol consumption 30
    Participant alone 27
    Multiple participants 20
    Experimenter/RA present 53
Surveillance during cue manipulation/task 35
    Participant alone 83
    Multiple participants 11
    Experimenter/RA present 6
Participant distracted while consuming beverages 37 32
Beverage manipulation successful 37 51

Note. k, number of alcohol conditions; BAC, blood alcohol concentration; RA, research assistant.

Assessing Publication Bias

Asymmetries in the distributions of effect sizes, indicating a possible reporting bias,61 were examined by inspecting funnel plots62 and assessing the degree of asymmetry using Begg's63 and Egger's64 techniques. Trim and fill procedures65,66 were used to estimate and correct for the possibility of missing studies (based on a rank-based data augmentation procedure) if publication bias was detected using the aforementioned funnel plot asymmetry tests.63,64

RESULTS

Study and Sample Characteristics

Study, sample, and experiment details of the 26 included manuscripts (k = 30 studies) are provided in Table I. Most manuscripts were published in journals (98%) between 1996 and 2015 (median publication date = 2009). Samples were typically recruited from universities (33%), communities (23%), or both universities and communities (43%) in the United States (81%; 15% Canada, 4% United Kingdom). Of the 3,964 individuals who consented to participate in the studies, less than half were women (42%), most were White (72%; k = 24), and averaged 24 years of age (range = 20-32; k = 25). Most participants were sexually active (M = 98%, k = 20). Of the 15 studies reporting the quantity of alcohol consumption, participants drank an average of 11 (SD = 3; range = 6-16) drinks per week.

Table I.

Characteristics of the Studies, Samples, and Alcohol Conditions Included in the Meta-Analysis

Citation Studies and Samples Setting Alcohol Administration M (SD) Comparison Conditions Arousal Manipulation Outcome Measure(s) Available Comparisons Available for Analysis
Abbey et al.21 N =180; 50% F; 65% W; 18% B; Mage = 24; 100% sexually active; M drinks in past month = 33 (SD = 40) USA; large urban university Target BAC .08%
Actual BAC .069% (.018)
Placebo
Sober
Vignette Likelihood of having unprotected sex Received: A vs. P/C
Expected: P vs. C
Abbey et al.15 N = 120; 50% F; 57% W; 24% B; Mage = 24; M drinks in past month = 25 (SD = 26) USA; large urban university Target BAC .08%
Actual BAC .072% (.019)
Placebo
Sober
Video Sexual arousal (self-reported)
Likelihood of having unprotected sexa
Likelihood of discussing pregnancy/STIs with a partner (communication)a
Received: A v. P
Received: A vs. C
Expected: P vs. C
Cho and Span22 N = 200; 50% F; 83% W; Mage = 22; 100% sexually active; M drinks in past week = 9 (SD = 4) USA; university and community Target BAC .08%
Actual BAC NR
Placebo
Sober
Vignette Likelihood of not using a condomb Received: A vs. P
Received: A vs. C
Expected: P vs. C
Conner et al.23 Study 2
N = 60; 50% F; 87% W; Mage = 21
UK; university Target BAC .09% men .07% women
Actual BAC NR
Placebo Vignette Intentions to have (unprotected) sexb Received: A vs. P
Davis24 N = 124; 0% F; 70% W; 6% B; Mage = 24; M drinks per week = 16 (SD = 11) USA; community Target BAC .08%
Actual BAC .056% (.012)
Control Vignette Likelihood of engaging in unprotected sex with an unwilling partner Received: A vs. C
Davis et al.25 N = 150; 49% F; 75% W; 6% B; Mage = 25; M drinks per week = 11 (SD = 9) USA; large urban university and community Target BAC .08%
Actual BAC E1: Ascending BAC Limb .073% (.011)
E2: Descending BAC Limb .072% (.009)
Control Vignette Sexual arousal (self-reported)e
Likelihood of engaging in unsafe sex (3 items)e
Received: A vs. C
Davis et al.26 N = 61; 49% F; 74% W; 0.3% B; Mage = 24; M drinks per week = 13 (SD = 9) USA; community Target BAC .10%
Actual BAC .092%(.011)
Control Vignette Likelihood of having sex without a condom Received: A vs. C
Davis et al.27
Linked Studies75,76
N = 448; 100% F; 72% W; 6% B; Mage = 25; 100% sexually active; M drinks per week = 14 (SD = 8) USA; urban community Target BAC .10%
Actual BAC .08% (.01)
Control Vignette Sexual desire/arousal (self-reported)
Likelihood of negotiating condom use
Likelihood of risky sex (3 items)
Received: A vs. C
Ebel-Lam et al.28 N = 94; 0% F; Mage = 20; 100% sexually active CAN; two universities Target BAC .08%
Actual BAC E1: Aroused .079% (.013)
E2: Neutral .079% (.013)
Placebo
Sober
Vignette and Video Likelihood of having (unprotected) sexd Received: A vs. P/C
George et al.29
Linked Studies77
Study 2
N = 165; 0% F; 76% W; 4% B; Mage = 25; M drinks per week = 16 (12)
USA; university and community Target BAC .06%, .08%, .10%
Actual BAC E1: .06% .048% (.012)
E2: .08% .060 (.010)
E3:. 10% .070 (.011)
Control Video and Vignette Sexual arousal (self-reported)e
Sexual arousal (genital response)e
Likelihood of unprotected sex (4 items)e
Received: A vs. C
Study 3
N = 173; 0% F; 75% W; 6% B; Mage = 25; 7% WSW; M drinks per week = 11 (8)
USA; university and community Target BAC .06%, .08%, .10%
Actual BAC E1: .06% .048% (.012)
E2: .08% .060 (.010)
E3: .10% .070 (.011)
Control Video and Vignette Sexual arousal (self-reported)e
Sexual arousal (genital response)e
Likelihood of unprotected sex (4 items)e
Received: A vs. C
Gilmore et al.30
Linked Studies78
N = 144; 100% F; 69% W; 9% B; Mage = 25; 71% sexually active; M drinks per week = 12 (SD = 8) USA; university and community Target BAC .10%
Actual BAC ≥.06%
Control Video and Vignette Sexual arousal (self-report)
Sexual arousal (genital response)
Intentions to have (unprotected) sex
Received: A vs. C
Gordon et al.31
Linked Studies79,80
N = 60; 0% F; 87% W; 8% B; Mage = 25; 100% sexually active Syracuse, NY, USA; community Target BAC .065
Actual BAC .068% (.014)
Placebo
Sober
Audio-interactive role plays Safer sex and refusal of unsafe sex communication and negotiation skills (objective measure) Received: A vs. C
Received: A vs. P
Expected: P vs. C
Kruse and Fromme32 Study 1
N = 80; 0% F; 77% W; 3% B; Mage = 23
USA; community Target BAC NR
Actual BAC: .06% (.01)
Placebo Model photographs Perceived partner desirability [arousal]
Intentions to discuss risks
Intentions to use condoms
Received: A vs. P
Study 2
N = 60; 0% F; 60% W; 3% B; Mage = 23
USA; community Target BAC NR
Actual BAC .05% (.01%)
Placebo Model photographs Perceived partner desirability [arousal]
Intentions to discuss risks
Intentions to use condoms
Received: A vs. P
MacDonald et al.33 Study 1
N = 65; 0% F; 18 – 25 yrs;100% sexually active
Alberta, CAN; university Target BAC .080%
Actual BAC E1: Impelling .091% (.043)
E2: Inhibiting .106% (.033)
Placebo
Sober
Video Likelihood of having (unprotected) sex Received: A vs. P/C
Study 4
N = 44; 0% F; 18 – 25 years; 100% sexually active
Alberta, CAN; university Target BAC .080%
Actual BAC E1: Impelling .072% (.013)
E2: Inhibiting .081% (.013)
Placebo
Sober
Video Likelihood of having (unprotected) sex Received: A vs. P/C
MacDonald et al.34 N = 358; 0% F; 100% sexually active Alberta and Ontario, CAN; university Target BAC .080%
Actual BAC .075% (.015)
Placebo
Sober
Video Sexual arousal (self-report)
Intentions to engage in unprotected sexd
Received: A vs. P/C
MacDonald et al.35
Linked Studies81
Study 2
N = 54; 0% F; ≥19 yrs; 100% sexually active
Waterloo, Ontario, CAN; university Target BAC .080%
Actual BAC .084% (.040)
Sober Video Sexual arousal (self-report)
Intentions to engage in (unprotected) sex
Received: A vs. C
Study 3
N = 55; 0% F; ≥19 yrs; 100% sexually active
Waterloo, Ontario, CAN; university Target BAC .080%
Actual BAC NR
Placebo
Sober
Video Sexual arousal (self-report)
Intentions to engage in (unprotected) sex
Received: A vs. P
Received: A vs. C
Expected: P vs. C
Maisto et al.37 N = 102; 100% F; 78% W; Mage = 24; 100% sexually active Syracuse, NY, USA; university and community Target BAC .075%
Actual BAC .060% (.014)
Placebo
Sober
Audio-Video
Role-Play
Scenarios
Likelihood of having unprotected sex Received: A vs. C
Received: A vs. P
Expected: P vs. C
Maisto et al.36 N = 67; 100% F; 78% W; Mage = 23; 100% sexually active; 60% heavy drinkers USA; university and community Target BAC E1. Low Dose .035%
E2. Moderate Dose .070% - .075%
Actual BAC E1. Low Dose .039% (.010)
E2. Moderate Dose .067% (.016)
Placebo
Sober
Audio-Video
Role-Play
Scenarios
Likelihood of having unprotected sexc
Safer sex and refusal of unsafe sex communication and negotiation skills (objective measure)c
Received: A vs. C
Received: A vs. P
Expected: P vs. C
Maisto et al.38 N = 53; 0% F; 87% W; Mage = 24; 100% sexually active; 75% heavy drinkers USA; university and community Target BAC .070% - .075%
Actual BAC .059% (.011)
Placebo
Sober
Audio-Video
Role-Play
Scenarios
Safer sex and refusal of unsafe sex communication and negotiation skills (objective measure)
Likelihood of having sex without a condom
Received: A vs. C
Received: A vs. P
Expected: P vs. C
Maisto et al.39 N = 171; 0% F; 58% W; Mage = 31; 100% sexually active; 100% MSM Syracuse, NY and Boston, MA, USA; community Target BAC .07%
Actual BAC E1. Aroused .068% (.016)
E2. Non-Aroused .061% (.019)
Placebo
Control
Audio-Video
Role-Play
Scenarios
Safer sex and refusal of unsafe sex communication and negotiation skills (objective measure)d
Intentions to have unprotected anal sexd
Received: A vs. C
Received: A vs. P
Expected P vs. C
Norris et al.40
Linked Studies82-84
N = 173; 100% F; 65% W; 14% B; Mage = 25; 100% sexually active; M drinks per week = 10 (SD=8) Large West Coast city, USA; university and community Target BAC E1. Low Dose .04%
E2. High Dose .08%
Actual BAC E1. Low Dose .034% (.008)
E2. High Dose .064% (.009)
Placebo
Sober
Vignette Sexual arousal (self-report)e
Intentions to communicate about condomse
Likelihood of having unprotected sexe
Received: A vs. C
Expected: P vs. C
Purdie et al.41 N = 234;100% F; 62% W; 11% B; Mage = 25; 100% sexually active; M drinks per week = 11 (SD = 10) USA; community Target BAC E1. Low Dose .04%
E2. High Dose .08%
Actual BAC E1. Low Dose .034% (.008)
E2. High Dose .062% (.007)
Placebo
Control
Vignette: Low, High, and Unknown Partner Risk Intentions for condom communicationse,f
Likelihood of unprotected sex (3 items)e,f
Received: A vs. C
Expected: P vs. C
Schacht et al.42
Linked Studies85
N = 64; 100% F; 81% W; 3% B; Mage = 27; 100% sexually active; M drinks per week = 10 (SD = 7) USA; university and community Target BAC .06%, .08%, .10%
Actual BAC NR
Sober Video and Vignette Sexual arousal (self-report)e
Sexual arousal (genital response)e
Likelihood of engaging in unprotected sex (vaginal/oral)e
Received: A vs. C
Stoner43
Linked Studies29,86
N = 160; 51% F; 75% W; 3% B; Mage = 24; 96% sexually active; 7% bisexual; M drinks per week = 6.1 (SD = 5.5) Washington, USA; university and community Target BAC E1. Low Dose .04%
E2. High Dose .08%
Actual BAC E1. Low Dose .036% (.009)
E2. High Dose .082% (.013)
Sober Vignette Sexual arousal (self-report)c
Likelihood of unprotected sex (5 items)c
Received: A vs. C
Wray et al.44
Linked Studies87
N = 113; 0% F; 87% W; 6% B; Mage = 22; 100% sexually active; 87% heavy drinkers Vermillion, South Dakota, USA; university and community Target BAC .08%
Actual BAC E1. Alcohol-High Autonomic Arousal .064% (.13)
E2. Alcohol-Low Autonomic Arousal .071% (.11)
Placebo
Control
Video Sexual arousal (self-report)d
Intentions to have unprotected vaginal sexd
Received: A vs. C
Received: A vs. P
Expected: P vs. C
Zawacki45 N = 132; 100% F; 46% W, Hispanic; 12% B; Mage = 23; M drinks per week = 10 (SD = 8) Southwestern City, USA; large public university Target BAC .08%
Actual BAC .085% (.012)
Placebo
Sober
Social Interaction Likelihood of having unprotected sex Received: A vs. P
Received: A vs. C
Expected: P vs. C

Note. N, number of participants who consented; F, female; W, White; B, Black/African American; MSM, men who have sex with men; WSW, women who have sex with women; A, alcohol; C, no alcohol control; P, placebo.

a

Analyses stratified by cognitive reserve (low, high) as measured by the Wide Range Achievement Test 3 (WRAT3).88

b

Analyses stratified by gender.

c

Analyses stratified by alcohol dose.

d

Analyses stratified by arousal manipulation.

e

Analyses collapsed across alcohol dose.

f

Analyses stratified by hypothetical partner risk.

Alcohol Administration

Of the 30 studies included in the meta-analysis, 37 separate alcohol administration conditions were assessed: 22 studies did not manipulate alcohol dose; 5 studies manipulated alcohol dose (e.g., low and high levels) representing 9 alcohol administration conditions.25,29,36,43 Three studies manipulated alcohol dose but subsequently collapsed analyses across conditions.40-42 (For these three studies, the mean target and actual BACs for the alcohol administration conditions were used in our analyses.)

Participants were typically asked to abstain from alcohol and/or other substances for 24 hours (73%) or fast from food for a median of 3 hours (76%) prior to their appointment. Alcohol administration procedures involved administering an average of 0.69 g/kg (SD = 0.16, range = 0.35 – 1.02) of 100-proof alcoholic beverage (M% absolute alcohol = 50, SD = 16). Drinks typically consisted of one part alcohol (e.g., vodka) to four parts other beverages (e.g., water, fruit juice). Participants were given a median of 15 minutes to drink their alcohol beverages (M = 21, SD = 17, range = 9 to 60 minutes) and a median of 15 minutes for alcohol absorption (M = 20, SD = 26, range = 3 to 103 minutes). The mean target BAC for the alcohol administration conditions was 0.08% (SD = 0.01; range = 0.04 - 0.10%) and the achieved BAC was 0.07% (SD = 0.01; range = 0.04 – 0.10%). A research assistant (RA) was typically present during alcohol consumption (53%) and participant's BAC levels were most often assessed before and after the alcohol administration (78%).

Methodological Quality

Methodological quality of the 30 studies ranged from 12 to 20 (out of 25) with a median score of 18 (M = 17.32, SD = 2.33). Overall, the studies satisfied an average of 69% (SD = 0.09) of the criteria, indicating moderate to strong methodological quality. The total methodological quality score was not correlated with any outcome with two exceptions: sexual communication and negotiation skills (Placebo vs. No Alcohol Control Contrast: r = 0.90, p = .04) and sexual communication or negotiation intentions (Alcohol vs. No Alcohol Control contrast: r = −0.84, p =.04). Skills in sexual communication and negotiation were stronger for the placebo group (vs. control) when methodological quality was high; intentions to communicate and negotiate about sex were higher among participants in the alcohol condition (vs. control) if the study was of lower methodological quality.

Overall Weighted Mean Effect Sizes Stratified by Type of Contrast

Weighted mean effect sizes and homogeneity analyses are presented in Table III. These analyses are stratified by the four contrast types: Alcohol vs. Placebo, Alcohol vs. No Alcohol Control, Placebo vs. No Alcohol Control, and Alcohol vs. Placebo/No Alcohol Control (when analyses were collapsed across placebo and no alcohol controls). Prior to conducting the analyses, all dependent variables by contrast were examined for extreme outliers.67 Only six outliers were detected: likelihood for unprotected sex (Alcohol vs. Placebo15,23,39; Placebo vs. Control39), sexual communication and negotiation skills (Alcohol vs. Placebo36), and self-reported intentions to communicate/negotiate regarding sex (Placebo vs. Control41). Therefore, the six effect sizes were removed from the overall weighted mean effect sizes. Forest plots for each dependent variable by contrast can be found in Supplemental Digital Content A.

Table III.

Weighted mean effect sizes and homogeneity tests by alcohol contrasts

A vs. P A vs. C

Dependent Variable k d+ (95% CI) Q (p-value) I2 (95% CI) k d+ (95% CI) Q (p-value) I2 (95% CI)
Unprotected sex intentions 14 0.35 (0.20, 0.51) 2.90 (1.00) 0 (0, 100) 28 0.24 (0.15, 0.32) 28.24 (.40) 4.39 (0, 34)

Sexual Communication and negotiation
    Behavioral skills 5 −0.43 (−0.73, −0.13) 0.58 (.97) 0 (0, 38) 6 −0.47 (−0.87, −0.08) 9.76 (.08) 49 (0, 80)
    Self-reported 3 −0.25 (−0.53, 0.03) 2.22 (.33) 10 (0, 48) 6 0.06 (−0.07, 0.20) 2.86 (.72) 0 (0, 71)

Sexual arousal
    Genital response 0 -- -- -- 4 −0.15 (−0.41, 0.11) 5.71 (.13) 47 (0, 83)
    Self-reported 6 0.19 (−0.03, 0.41) 1.88 (.86) 0 (0, 56) 18 0.31 (0.13, 0.50) 59.96 (<.01) 72 (54, 82)
P vs. C A vs. P/C

k d+ (95% CI) Q (p-value) I2 (95% CI) k d+ (95% CI) Q (p-value) I2 (95% CI)
Unprotected sex intentions 17 −0.10 (−0.23, 0.03) 9.41 (.90) 0 (0, 27) 9 0.17 (−0.13, 0.46) 23.36 (<.01) 66 (30, 83)

Sexual Communication and negotiation
    Behavioral skills 5 −0.13 (−0.53, 0.28) 7.43 (.12) 46 (0, 80) 0 -- -- --
    Self-reported 4 0.09 (−0.15, 0.33) 0.5 (.92) 0 (0, 65) 0 -- -- --

Sexual arousal
    Genital response 0 -- -- -- 0 -- -- --
    Self-reported 5 0.07 (−0.39, 0.53) 14.72 (<.01) 73 (32, 89) 1 -- -- --

Intentions to Engage in Unprotected Sex

Participants who consumed alcohol were more likely to report intentions to engage in unprotected sex with a potential sexual partner relative to participants assigned to a placebo (d+random = 0.35, 95% CI = 0.20, 0.51) or no alcohol condition (d+random = 0.24, 95% CI = 0.15, 0.32). There were no differences in reported likelihood of unprotected sex between the placebo vs. no alcohol control (d+random = −0.10, 95% CI = −0.23, 0.03) or the alcohol vs. placebo/no alcohol control (d+random = 0.17, 95% CI = −0.13, 0.46) conditions. The hypothesis of homogeneity was supported for all contrasts except for the alcohol vs. placebo/no alcohol control contrast. Examination of the effect sizes for the alcohol vs. placebo/no alcohol control contrast revealed that 4 of the 6 studies manipulated sexual arousal (e.g., enhancing impelling cues and suppressing inhibiting cues).28,33,34 Experimental manipulations of arousal were analyzed separately (i.e., aroused, non-aroused). Compared to placebo/no alcohol controls, participants who expected alcohol (placebo) were more likely to report unprotected sex intentions with a potential partner if sexual arousal was heightened (d+random = 0.55, 95% CI = 0.30, 0.80, k = 4) vs. non-aroused (d+random = −0.23, 95% CI = −0.48, 0.02, k = 4), QB (1) = 24.79, p <.001.

Sexual Communication and Negotiation

Participants who consumed alcohol were less likely to perform objectively measured sexual communication and negotiation skills compared to those in the placebo condition (d+random = −0.43, 95% CI = −0.73, 0.13) and no alcohol controls (d+random = −0.47, 95% CI = −0.87, −0.08). There were no differences in communication and negotiation skills between the placebo vs. no alcohol control (d+random = −0.13, 95% CI = −0.53, 0.28) conditions. The hypothesis of homogeneity was supported for all contrasts, ps >.08. There were no differences in self-reported intentions to communicate or negotiate about sex for any contrast: alcohol vs. placebo (d+random = −0.25, 95% CI = −0.53, 0.03), alcohol vs. no alcohol control (d+random = 0.06, 95% CI = −0.07, 0.12), or placebo vs. no alcohol control (d+random = 0.09, 95% CI = −0.15, 0.33). The hypothesis of homogeneity was supported for all contrasts, ps >.56.

Sexual Arousal

Four studies reported measuring sexual arousal using physiological genital responding.29,30,42 Participants in the alcohol vs. no alcohol control conditions did not differ (d+random = −0.15, 95% CI = −0.41, 0.11). No other contrasts were available for analysis. Participants who consumed alcohol were more likely to self-report being aroused by the sexual stimuli than those in the no alcohol control group, d+random = 0.31, 95% CI = 0.13, 0.50. There were no differences in reported sexual arousal for the alcohol vs. placebo and placebo vs. no alcohol control contrasts. The hypothesis of homogeneity was not supported for the (a) alcohol vs. no alcohol control or (b) placebo vs. no alcohol control contrasts.

Comparisons between Experimental Contrast Pairs

The between-group homogeneity test, QB, was conducted using random-effects assumptions to determine if there were any differences in weighted mean effects sizes between contrasts pairs (i.e., alcohol vs. placebo, alcohol vs. no alcohol control, placebo vs. control, and alcohol vs. placebo/no alcohol control).

Intentions to Engage in Unprotected Sex

There was a significant difference between the four contrasts on unprotected sex intentions, QB (3) = 30.45, p <.001. The weighted mean effect size for the alcohol vs. placebo contrast was larger than the weighted mean effect size for the placebo vs. no alcohol control (p <.001) and alcohol vs. placebo/no alcohol control (p = .034) contrasts. The weighted mean effect size for the alcohol vs. no alcohol control contrast and the alcohol vs. placebo/no alcohol control contrast was larger than the weighted mean effect size for the placebo/no alcohol control contrast (ps ≤.003).

Sexual Communication and Negotiation

A significant difference was found in weighted mean effect sizes between the three contrast pairs on sexual communication and negotiation skills, QB (2) = 55.09, p <.001. The overall effect size for the placebo vs. no control contrast was smaller than the overall effect sizes for the alcohol vs. placebo (p <.001) and alcohol vs. no alcohol control (p <.001) contrasts. Also, a significant difference in the weighted mean effect sizes between contrasts pairs was seen for self-reported intentions for sexual communication and negotiation, QB (2) = 46.30, p <.001. Analyses indicated that the weighted mean effect size for the alcohol vs. placebo contrast was larger than the alcohol vs. no alcohol control (p <.001) and the placebo vs. no alcohol control (p <.001) contrasts.

Sexual Arousal

Between-group homogeneity tests revealed a significant difference in the weighted mean effect sizes between contrast pairs for self-reported sexual arousal, QB (2) = 10.81, p <.001. The weighted mean effect sizes for the placebo vs. no alcohol control contrast was smaller than the weighted mean effect sizes for the alcohol vs. placebo (p = .018) or the alcohol vs. no alcohol control (p = .001) contrasts.

Moderators of Intentions to Engage in Unprotected Sex

Meta-regression (continuous variables) and the meta-analytic analogue to the ANOVA (categorical variables) homogeneity analysis (following random-effects assumptions) were conducted to examine potential moderators of intentions to engage in unprotected sex. We first examined participant characteristics (i.e., gender, age, and current drinking pattern) as potential moderators of unprotected sex intentions across all four contrasts. These analyses were followed by moderator tests of the alcohol administration procedures on alcohol condition effect sizes (i.e., alcohol vs. no alcohol control and alcohol vs. placebo).

Participant Characteristics

Age and current drinking pattern (i.e., number of drinks per week) did not moderate any intentions to engage in unprotected sex (data not shown). Participants who consumed alcohol (compared to no alcohol controls) were more likely to report unprotected sex intentions when more men (vs. women) were sampled (B = 0.21, SE = 0.10, p = .04, k = 28). Gender was not a significant moderator of unprotected sex intentions for any other contrast.

Alcohol Administration Procedures

Only three variables moderated the effects of alcohol consumption on unprotected sex intentions for the alcohol vs. no alcohol control contrast: fasting from alcohol and/or other drugs, BAC assessment, and presence of an experimenter during alcohol consumption (Table IV). Participants who consumed alcohol were more likely to report greater intentions to engage in unprotected sex when they had fasted from alcohol and/or other drugs (QB [1] = 4.06, p =.043) or BAC was assessed at one time point (i.e., before OR after the task/manipulation) (QB [1] = 5.73, p =.017).

Table IV.

Impact of Alcohol Consumption on Unprotected Sex Intentions Stratified by Alcohol Administration Procedures

Alcoho vs. Placebo Alcohol vs. No Alcohol Control

k d+ (95% CI) k d+ (95% CI)

Fasted from alcohol and/or other substances
    Yes 11 0.39 (0.22, 0.57) 21 0.30 (0.19, 0.40)
    No/not reported 3 0.23 (−0.09, 0.55) 7 0.14 (−0.06, 0.33)

Hours fasted from food
    3 hours 6 0.50 (0.21, 0.78) 19 0.27 (0.17, 0.38)
    4 hours 4 0.25 (0.01, 0.49) 4 0.06 (−0.31, 0.44)

Alcohol Dose
    Units of alcohol administered
        <.70 g/kg 6 0.27 (0.05, 0.49) 6 0.29 (0.11, 0.47)
        ≥.70 g/kg 8 0.44 (0.22, 0.65) 22 0.23 (0.13, 0.33)
    Percent absolute alcohol in beverage
        40 10 0.40 (0.20, 0.60) 9 0.19 (−0.04, 0.42)
        50 50 0.25 (−0.02, 0.52) 13 0.22 (0.11, 0.33)
        95 0 -- 3 0.31 (0.01, 0.61)
    Ratio of alcohol to non-alcohol content
        1:4 8 0.44 (0.21, 0.66) 18 0.24 (0.13, 0.36)
        Other ratio 6 0.28 (0.06, 0.49) 10 0.26 (0.08, 0.43)

Type of alcohol beverage
    Vodka 12 0.35 (0.19, 0.52) 22 0.21 (0.11, 0.31)
    Other alcoholic beverage 0 -- 3 0.31 (0.01, 0.61)

Time allotted for alcohol ingestion (minutes)
    <20 7 0.38 (0.15, 0.60) 7 0.46 (0.21, 0.70)
    ≥20 7 0.34 (0.12, 0.55) 21 0.21 (0.12, 0.29)

Time allotted for alcohol absorption (minutes)
    <20 9 0.40 (0.19, 0.61) 18 0.22 (0.08, 0.36)
    ≥20 5 0.30 (0.07, 0.53) 10 0.24 (0.13, 0.35)

Target BAC
    <.08 4 0.48 (0.13, 0.82) 10 0.18 (0.01, 0.35)
    ≥.08 10 0.32 (0.15, 0.50) 18 0.26 (0.16, 0.37)

Actual BAC
    <.07 7 0.39 (0.16, 0.62) 15 0.25 (0.13, 0.37)
    ≥.07 7 0.33 (0.11, 0.54) 13 0.25 (0.10, 0.40)

BAC assessed
    Before or after task/manipulation 1 -- 3 0.54 (0.28, 0.81)
    Before and after task/manipulation 12 0.35 (0.19, 0.52) 25 0.20 (0.11, 0.29)

Environment
    Laboratory 12 0.38 (0.20, 0.55) 28 0.24 (0.15, 0.32)
    Contrived Bar 2 0.27 (−0.07, 0.60) 0 --

Experimenter present during alcohol consumption
    Yes 4 0.25 (0.01, 0.49) 13 0.19 (0.06, 0.32)
    No/not reported 10 0.43 (0.22, 0.63) 15 0.26 (0.15, 0.38)

Experimenter present during cue manipulation
    Yes 3 0.30 (−0.01, 0.62) 3 0.13 (−0.18, 0.44)
    No/not reported 11 0.37 (0.19, 0.55) 25 0.25 (0.16, 0.34)

Participant distracted while consuming beverages
    Yes 9 0.37 (0.17, 0.56) 11 0.26 (0.10, 0.42)
    No/not reported 5 0.33 (0.08, 0.59) 17 0.24 (0.12, 0.36)

Beverage manipulation successful
    Yes 13 0.36 (0.20, 0.52) 17 0.17 (0.03, 0.31)
    No/not reported 1 -- 11 0.28 (0.17, 0.38)

Note. k, number of alcohol condition; BAC, blood alcohol concentration

Publication Bias

Both published and unpublished manuscripts were included in the meta-analysis to avert the possibility of publication bias; we also examined our data using graphical and statistical tools to test for the possibility of publication bias (funnel plots and results of the statistical tests are available in Supplemental Digital Content B). Results from Begg's test 63 and Egger's regression asymmetry test64 revealed the possibility of publication bias for two variables: unprotected sex intentions (Alcohol vs. Placebo) and self-reported sexual arousal (Alcohol vs. Control). Trim and fill procedures 65,66 were used to estimate an overall effect size for these variables after accounting for the bias. These analyses indicated that six studies assessing unprotected sex intentions may be missing; the pooled estimate was similar to the original estimate (original d+random = 0.36, 95% CI = 0.20, 0.51; estimated d+random =0.28, 95% CI = 0.15, 0.42) suggesting that adding the additional studies would not change our conclusions (i.e., greater intentions to engage in unprotected sex reported by participants in the alcohol vs. placebo conditions). Trim and fill procedures for self-reported sexual arousal indicated that nine studies may be missing; the interpretation of our findings would differ if those studies were included (original d+random = 0.31, 95% CI = 0.13, 0.50, k = 18; estimated d+random =−0.00, 95% CI = −0.20, 0.20, k = 27). Significant heterogeneity was observed (original: Q [17] = 59.97, p <.001; estimated: Q [26] = 149.69, p <.001). The findings for self-reported sexual arousal should be interpreted with caution as performance of the trim-and-fill method is poor when significant heterogeneity is present.68-70

DISCUSSION

Experimental research assessing the effects of alcohol consumption on sexual decision-making has proliferated since the mid-to-late 1980s due to the increasing scientific knowledge that alcohol use is a critical factor in the HIV epidemic.5 Use of experimental alcohol administration methods require assessment of theoretical antecedents of sexual behavior (i.e., intentions to use condoms), rather than risk behaviors per se. Nonetheless, methodologies involving random assignment to alcohol conditions improve our ability to make causal inferences about the association between alcohol consumption, sexual risk behavior, and HIV. The current meta-analysis examined 30 experimental studies (N = 3,964) that manipulated alcohol consumption by randomly assigning participants to an alcohol condition (e.g., alcohol, placebo, no alcohol control), administering alcohol prior to a sexual stimulus (e.g., role-plays, erotic video), and assessing sexual decision-making (e.g., intentions to engage in unprotected sex, sexual communication and negotiation skills) following exposure to the sexual stimulus. Our findings indicated that alcohol consumption – but not alcohol expectancies -- is causally related to sexual outcomes, providing support that alcohol use is an important risk factor for HIV.

The weighted mean effect sizes for unprotected sex intentions in both the alcohol vs. placebo (d+ = 0.35) and the alcohol vs. no alcohol control (d+ = 0.24) contrasts were significant. We found no significant differences in unprotected sex intentions when studies collapsed their analyses across placebo and no alcohol control conditions (i.e., alcohol vs. placebo/no alcohol control), d+ = 0.17 (95% CI = -0.13, 0.46). However, the effects of alcohol (vs. placebo/no alcohol control) on intentions to engage in unprotected sex was stronger when sexual arousal was heightened (p <.001). The magnitude of the effect sizes for unprotected sex intentions was small to medium indicating that the influence of alcohol on unprotected sex intentions is robust. Our findings are consistent with prior research using meta-analysis to examine only experimental studies that randomize participants to alcohol and placebo conditions.18

We evaluated both the pharmacological (myopia) and psychological (expectancy) hypotheses about the effects of alcohol. Consistent with the alcohol myopia perspective, the overall pattern of findings indicates that participants who actually consumed alcohol reported stronger intentions to engage in unprotected sex, weaker sexual communication and negotiation skills, and higher levels of sexual arousal compared to participants in either placebo or no alcohol control groups. Contrary to the predictions of expectancy theory, we found no differences in sexual outcomes between placebo and no alcohol controls. Between-contrast comparisons also indicated that the weighted mean effect sizes were significantly larger for the alcohol vs. placebo contrast relative to the placebo vs. no alcohol control contrast for all outcomes measures (i.e., unprotected sex intentions, sexual communication and negotiation skills and intentions, and sexual arousal). Thus, our findings support alcohol myopia theory as the most relevant theoretical explanation for the influence of alcohol consumption on sexual outcomes.

The hypothesis of homogeneity was supported for most contrasts, which limited our ability examine patterns in the effect sizes for unprotected sex intentions. This was notable for the alcohol vs. placebo contrast in which we found the overall weighted mean effect size was robust across the studies included in the analysis (Q [13] = 2.90, p = 1.00; I2 = 0, 95% CI = 0, 100). We did observe modest differences in the overall weighted mean effect size for the alcohol vs. control contrast (Q [27] = 28.24, p = 0.40; I2 = 4.39, 95% CI = 0, 34). Meta-regression analyses indicated that participants who consumed alcohol (vs. no alcohol controls) were less likely to report unprotected sex intentions if they were female (d+random = 0.16, 95% CI = 0.05, 0.27; k = 12) vs. male (d+random = 0.37, 95% CI = 0.21, 0.53; k = 10). Consistent with alcohol myopia theory, both men and women are less likely to attend to inhibitory cues and more likely to attend to impelling cues when intoxicated. Nonetheless, inhibitory cues may be more salient for women as women express somewhat greater sexual concerns than do men (e.g., general attitudes toward sexual permissiveness, attitudes toward extramarital sex).71 Moderator tests also indicated that participants who consumed alcohol (vs. no alcohol controls) were more likely to report unprotected sex intentions if (a) they had fasted from alcohol and/or other drugs before their research appointment or (b) BAC was assessed only once (rather than both before and after the task/manipulation). These findings highlight the fact that experimental procedures can substantially influence the effects of intoxication on participants.

Limitations

Several limitations should be considered when interpreting these findings. First, experimental methods examine theoretical antecedents of sexual behavior (i.e., intentions to engage in unprotected sex). Although intention to engage in sexual behavior is a significant predictor of enacted sexual behavior, we cannot assert that alcohol consumption is directly related to sexual risk behaviors based on our findings. Second, all outcomes, except for sexual communication and negotiation skills and genital responses to assess sexual arousal, involve self-reports, which are vulnerable to measurement (e.g., text, formatting, and context) and social (e.g., self-presentation) biases.46,72 Self-report is imperfect, but most researchers employed methods designed to optimize data quality. Third, few studies sampled those who are most affected by HIV in the United States, namely, men who have sex with men (MSM) or racial/ethnic minorities (especially blacks/African Americans).73 Laboratory-based studies have relied almost exclusively on heterosexuals (or individuals indicating their interest in opposite-sex relationships) with only a single study sampling 100% MSM.40 Most participants were white with only 8% of the study samples included blacks/African Americans. The reliance on heterosexuals and whites limits the generalizability of our findings. Fourth, a direct measure of alcohol expectancies (e.g., Sex-Related Alcohol Expectancies Questionnaire74) was measured in only a subset (27%) of the studies included in the meta-analysis. Only three of these studies provided a correlation between alcohol expectancies (broadly defined) and unprotected sex intentions.24,39,45 These studies measured various sex-related alcohol expectancies (e.g., aggression, enhancement, riskiness). Therefore, the available primary level studies do not allow strong tests of the effects of specific sex-related alcohol expectancies on unprotected sex intentions. Finally, our moderator tests were limited to the data available in the individual studies. Some details regarding the alcohol administration procedures were reported infrequently (e.g., time of day alcohol was consumed) and thus were omitted from our moderator analyses. We were also unable to test for individual- and contextual-level characteristics (e.g., sensation seeking, partner type) that may explain the relation between alcohol consumption and unprotected sex intentions.14 Future research is needed to fully investigate potential moderators of the relation between alcohol consumption and unprotected sex intentions.

CONCLUSIONS

Alcohol consumption directly affects sexual-decision making (i.e., intentions to engage in unprotected sex) which is associated with increased sexual risk behavior and incident HIV. Addressing alcohol consumption as a key component in sexual risk-taking may lead to more effective HIV interventions.

Supplementary Material

Supplemental Digital Content A
Supplemental Digital Content B

ACKNOWLEDGEMENTS

Research reported in this paper was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under award number R01 AA021355 to Lori A. J. Scott-Sheldon. Karlene Cunningham was supported by the Child/Adolescent Biobehavioral HIV Research Training Grant (T32 MH078788) from the National Institute of Mental Health of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

We thank the following study authors who provided additional information and/or data: Young-Hee Cho, PhD, Amanda K. Gillmore, PhD, Tara K. MacDonald, PhD, Stephen A. Maisto, PhD, and Tyler B. Wray, PhD. We also thank Erica Sevetson, MLS, for assistance with our electronic bibliographic database search statement, and Karla Pineda Dvgas, BA, for translating our abstract from English to Spanish.

REFERENCES

  • 1.Shuper PA, Joharchi N, Irving H, Rehm J. Alcohol as a correlate of unprotected sexual behavior among people living with HIV/AIDS: review and meta-analysis. AIDS Behav. 2009 Dec;13(6):1021–1036. doi: 10.1007/s10461-009-9589-z. [DOI] [PubMed] [Google Scholar]
  • 2.Shuper PA, Neuman M, Kanteres F, Baliunas D, Joharchi N, Rehm J. Causal considerations on alcohol and HIV/AIDS--a systematic review. Alcohol and alcoholism (Oxford, Oxfordshire) 2010 Mar-Apr;45(2):159–166. doi: 10.1093/alcalc/agp091. [DOI] [PubMed] [Google Scholar]
  • 3.Halpern-Felsher BL, Millstein SG, Ellen JM. Relationship of alcohol use and risky sexual behavior: A review and analysis of findings. Journal of Adolescent Health. 1996;19(5):331–336. doi: 10.1016/S1054-139X(96)00024-9. 11// [DOI] [PubMed] [Google Scholar]
  • 4.Leigh BC, Stall R. Substance use and risky sexual behavior for exposure to HIV: Issues in methodology, interpretation, and prevention. American Psychologist. 1993;48(10):1035–1045. doi: 10.1037//0003-066x.48.10.1035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Hendershot CS, George WH. Alcohol and sexuality research in the AIDS era: Trends in publication activity, target populations and research design. AIDS & Behavior. 2007;11(2):217–226. doi: 10.1007/s10461-006-9130-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Shiffman S, Stone AA, Hufford MR. Ecological momentary assessment. Annu. Rev. Clin. Psychol. 2008;4:1–32. doi: 10.1146/annurev.clinpsy.3.022806.091415. [DOI] [PubMed] [Google Scholar]
  • 7.Steele CM, Josephs RA. Alcohol myopia. Its prized and dangerous effects. Am Psychol. 1990 Aug;45(8):921–933. doi: 10.1037//0003-066x.45.8.921. [DOI] [PubMed] [Google Scholar]
  • 8.Crowe LC, George WH. Alcohol and human sexuality: Review and integration. Psychological Bulletin. 1989;105(3):374–386. doi: 10.1037/0033-2909.105.3.374. [DOI] [PubMed] [Google Scholar]
  • 9.Lang AR. The social psychology of drinking and human sexuality. Journal of Drug Issues. 1985;15(2):273–289. [Google Scholar]
  • 10.Rohsenow DJ, Marlatt GA. The balanced placebo design: methodological considerations. Addictive Behaviors. 1981;6(2):107–122. doi: 10.1016/0306-4603(81)90003-4. [DOI] [PubMed] [Google Scholar]
  • 11.Marlatt GA, Rohsenow DJ. Advances in substance abuse: Behavioral and biological research. JAI Press; Greenwich, CT: 1980. Cognitive processes in alcohol use: Expectancy and the balanced placebo design. pp. 159–199. [Google Scholar]
  • 12.Albarracin D, Johnson BT, Fishbein M, Muellerleile PA. Theories of reasoned action and planned behavior as models of condom use: a meta-analysis. Psychological Bulletin. 2001 Jan;127(1):142–161. doi: 10.1037/0033-2909.127.1.142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sheeran P, Abraham C, Orbell S. Psychosocial correlates of heterosexual condom use: a meta-analysis. Psychol Bull. 1999 Jan;125(1):90–132. doi: 10.1037/0033-2909.125.1.90. [DOI] [PubMed] [Google Scholar]
  • 14.Cooper ML. Does Drinking Promote Risky Sexual Behavior?: A Complex Answer to a Simple Question. Current Directions in Psychological Science. 2006;15(1):19–23. [Google Scholar]
  • 15.Abbey A, Saenz C, Buck PO, Parkhill MR, Hayman Jr LW. The effects of acute alcohol consumption, cognitive reserve, partner risk, and gender on sexual decision making. Journal of Studies on Alcohol and Drugs. 2006;67(1):113. doi: 10.15288/jsa.2006.67.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Hull JG, Bond CF., Jr. Social and behavioral consequences of alcohol consumption and expectancy: a meta-analysis. Psychological Bulletin. 1986 May;99(3):347–360. [PubMed] [Google Scholar]
  • 17.George WH, Stoner SA. Understanding acute alcohol effects on sexual behavior. Annu Rev Sex Res. 2000;11:92–124. [PubMed] [Google Scholar]
  • 18.Rehm J, Shield KD, Joharchi N, Shuper PA. Alcohol consumption and the intention to engage in unprotected sex: systematic review and meta-analysis of experimental studies. Addiction. 2012 Jan;107(1):51–59. doi: 10.1111/j.1360-0443.2011.03621.x. [DOI] [PubMed] [Google Scholar]
  • 19.Lipsey MW, Wilson DB. Practical meta-analysis. Sage; Thousand Oaks, CA: 2001. [Google Scholar]
  • 20.Moher D, Liberati A, Tetzlaff J, Altman DG, The PG. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS medicine. 2009;6(7):e1000097. doi: 10.1371/journal.pmed.1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Abbey A, Saenz C, Buck PO. The cumulative effects of acute alcohol consumption, individual differences and situational perceptions on sexual decision making. Journal of Studies on Alcohol. 2005;66(1):82–90. doi: 10.15288/jsa.2005.66.82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Cho Y-H, Span SA. The effect of alcohol on sexual risk-taking among young men and women. Addictive Behaviors. 2010;35(8):779–785. doi: 10.1016/j.addbeh.2010.03.007. [DOI] [PubMed] [Google Scholar]
  • 23.Conner M, Sutherland E, Kennedy F, Grearly C, Berry C. Impact of alcohol on sexual decision making: Intentions to have unprotected sex. Psychology and Health. 2008;23(8):909–934. doi: 10.1080/08870440701596551. [DOI] [PubMed] [Google Scholar]
  • 24.Davis KC. The influence of alcohol expectancies and intoxication on men's aggressive unprotected sexual intentions. Experimental and Clinical Psychopharmacology. 2010;18(5):418–428. doi: 10.1037/a0020510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Davis KC, George WH, Norris J, et al. Effects of alcohol and blood alcohol concentration limb on sexual risk-taking intentions. Journal of Studies on Alcohol and Drugs. 2009;70(4):499. doi: 10.15288/jsad.2009.70.499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Davis KC, Hendershot CS, George WH, Norris J, Heiman JR. Alcohol's Effects on Sexual Decision Making: An Integration of Alcohol Myopia and Individual Differences. Journal of Studies on Alcohol and Drugs. 2007;68(6):843–851. doi: 10.15288/jsad.2007.68.843. [DOI] [PubMed] [Google Scholar]
  • 27.Davis KC, Masters NT, Eakins D, et al. Alcohol intoxication and condom use self-efficacy effects on women's condom use intentions. Addictive Behaviors. 2014;39(1):153–158. doi: 10.1016/j.addbeh.2013.09.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ebel-Lam AP, MacDonald TK, Zanna MP, Fong GT. An experimental investigation of the interactive effects of alcohol and sexual arousal on intentions to have unprotected sex. Basic and Applied Social Psychology. 2009;31(3):226–233. [Google Scholar]
  • 29.George WH, Cue Davis K, Norris J, et al. Indirect effects of acute alcohol intoxication on sexual risk-taking: The roles of subjective and physiological sexual arousal. Archives of Sexual Behavior. 2009;38(4):498–513. doi: 10.1007/s10508-008-9346-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Gilmore AK, George WH, Nguyen HV, Heiman JR, Davis KC, Norris J. Influences of Situational Factors and Alcohol Expectancies on Sexual Desire and Arousal Among Heavy-Episodic Drinking Women: Acute Alcohol Intoxication and Condom Availability. Archives of Sexual Behavior. 2013:1–11. doi: 10.1007/s10508-013-0109-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Gordon CM, Carey MP, Carey KB. Effects of a drinking event on behavioral skills and condom attitudes in men: Implications for HIV risk from a controlled experiment. Health Psychology. 1997;16(5):490–495. doi: 10.1037//0278-6133.16.5.490. [DOI] [PubMed] [Google Scholar]
  • 32.Kruse MI, Fromme K. Influence of physical attractiveness and alcohol on men's perceptions of potential sexual partners and sexual behavior intentions. Experimental and Clinical Psychopharmacology. 2005;13(2):146–156. doi: 10.1037/1064-1297.13.2.146. [DOI] [PubMed] [Google Scholar]
  • 33.MacDonald TK, Fong GT, Zanna MP, Martineau AM. Alcohol myopia and condom use: Can alcohol intoxication be associated with more prudent behavior? Journal of Personality and Social Psychology. 2000;78(4):605–619. doi: 10.1037//0022-3514.78.4.605. [DOI] [PubMed] [Google Scholar]
  • 34.MacDonald TK, MacDonald G, Zanna MP, Fong G. Alcohol, sexual arousal, and intentions to use condoms in young men: Applying alcohol myopia theory to risky sexual behavior. Health Psychology. 2000;19(3):290–298. [PubMed] [Google Scholar]
  • 35.MacDonald TK, Zanna MP. Why common sense goes out the window: Effects of alcohol on intentions to use condoms. Personality and Social Psychology Bulletin. 1996;22:763–775. [Google Scholar]
  • 36.Maisto SA, Carey MP, Carey KB, Gordon CM, Schum JL. Effects of alcohol and expectancies on HIV-related risk perception and behavioral skills in heterosexual women. Experimental and Clinical Psychopharmacology. 2004;12(4):288–297. doi: 10.1037/1064-1297.12.4.288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Maisto SA, Carey MP, Carey KB, Gordon CM. The effects of alcohol and expectancies on risk perception and behavioral skills relevant to safer sex among heterosexual young adult women. Journal of Studies on Alcohol. 2002;63(4):476–485. doi: 10.15288/jsa.2002.63.476. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Maisto SA, Carey MP, Carey KB, Gordon CM, Schum JL, Lynch KG. The relationship between alcohol and individual differences variables on attitudes and behavioral skills relevant to sexual health among heterosexual young adult men. Archives of Sexual Behavior. 2004;33(6):571–584. doi: 10.1023/B:ASEB.0000044741.09127.e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Maisto SA, Palfai T, Vanable PA, Heath J, Woolf-King SE. The effects of alcohol and sexual arousal on determinants of sexual risk in men who have sex with men. Archives of Sexual Behavior. 2012;41(4):971–986. doi: 10.1007/s10508-011-9846-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Norris J, Stoner SA, Hessler DM, et al. Cognitive Mediation of Alcohol's Effects on Women's in-the-Moment Sexual Decision Making. Health Psychology. 2009;28(1):20–28. doi: 10.1037/a0012649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Purdie MP, Norris J, Davis KC, et al. The Effects of Acute Alcohol Intoxication, Partner Risk Level, and General Intention to Have Unprotected Sex on Women's Sexual Decision Making With a New Partner. Experimental and Clinical Psychopharmacology. 2011;19(5):378–388. doi: 10.1037/a0024792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Schacht RL, George WH, Davis KC, et al. Sexual abuse history, alcohol intoxication, and women's sexual risk behavior. Archives of Sexual Behavior. 2010;39(4):898–906. doi: 10.1007/s10508-009-9544-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Stoner SA. Alcohol and sexual disinhibition among college students [PhD] Psychology, University of Washington; US: 2003. [Google Scholar]
  • 44.Wray TB, Simons JS, Maisto SA. Effects of alcohol intoxication and autonomic arousal on delay discounting and risky sex in young adult heterosexual men. Addictive Behaviors. 2015 doi: 10.1016/j.addbeh.2014.10.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Zawacki T. Effects of Alcohol on Women's Risky Sexual Decision Making during Social Interactions in the Laboratory. Psychology of Women Quarterly. 2011;35(1):107–118. [Google Scholar]
  • 46.Bushman BJ, Cooper HM. Effects of alcohol on human aggression: an integrative research review. Psychological Bulletin. 1990 May;107(3):341–354. doi: 10.1037/0033-2909.107.3.341. [DOI] [PubMed] [Google Scholar]
  • 47.Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. Journal of Epidemiology & Community Health. 1998;52(6):377–384. doi: 10.1136/jech.52.6.377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Fowkes FG, Fulton PM. Critical appraisal of published research: introductory guidelines. British Medical Journal. 1991;302(6785):1136–1140. doi: 10.1136/bmj.302.6785.1136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Miller WR, Brown JM, Simpson TL, Handmaker NS, Bien TH, Luckie LF. What works? A methodological analysis of the alcohol treatment outcome literature. In: Hester RK, Miller WR, editors. Handbook of alcoholism treatment approaches: Effective alternatives. 2nd ed. Allyn & Bacon; Needham Heights, MA: 1995. pp. 12–44. [Google Scholar]
  • 50.Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Erlbaum; New York: 1998. [Google Scholar]
  • 51.Johnson BT, Eagly AH. Meta-Analysis of Research in Social and Personality Psychology. In: Reis HT, Judd CM, editors. Handbook of research methods in social and personality psychology. 2nd ed. Cambridge University Press; London: 2014. pp. 675–707. [Google Scholar]
  • 52.Sanchez-Meca J, Marin-Martinez F, Chacon-Moscoso S. Effect-size indices for dichotomized outcomes in meta-analysis. Psychological Methods. 2003;8(4):448–467. doi: 10.1037/1082-989X.8.4.448. [DOI] [PubMed] [Google Scholar]
  • 53.Hedges LV. Distribution theory for Glass's estimator of effect size and related estimators. Journal of Educational and Behavioral Statistics. 1981;6:107–128. [Google Scholar]
  • 54.Stata/SE [computer program]. Version 12.1 for Windows. StataCorp LP; College Station, TX: 2013. [Google Scholar]
  • 55.Meta-analysis macros for SAS, SPSS, and Stata [computer program] 2001 [Google Scholar]
  • 56.Huedo-Medina TB, Sanchez-Meca J, Marin-Martinez F, Botella J. Assessing heterogeneity in meta-analysis: Q statistic or I2 index? Psychological Methods. 2006;11(2):193–206. doi: 10.1037/1082-989X.11.2.193. [DOI] [PubMed] [Google Scholar]
  • 57.Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine. 2002;21(11):1539–1558. doi: 10.1002/sim.1186. [DOI] [PubMed] [Google Scholar]
  • 58.Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. British Medical Journal. 2003;327(7414):557–560. doi: 10.1136/bmj.327.7414.557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Harbord RM, Higgins JPT. Meta-regression in Stata. Stata Journal. 2008;8(4):493–519. [Google Scholar]
  • 60.Hedges LV. Fixed effects models. In: Cooper H, Hedges LV, editors. The Handbook of Research Synthesis. Russell Sage Foundation; New York: 1994. pp. 285–299. [Google Scholar]
  • 61.Rosenthal R. The “file-drawer” problem and tolerance for null results. Psychological Bulletin. 1979;86:638–641. [Google Scholar]
  • 62.Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. Journal of Clinical Epidemiology. 2001;54(10):1046–1055. doi: 10.1016/s0895-4356(01)00377-8. [DOI] [PubMed] [Google Scholar]
  • 63.Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–1101. [PubMed] [Google Scholar]
  • 64.Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. British Medical Journal. 1997;315(7109):629–634. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56(2):455–463. doi: 10.1111/j.0006-341x.2000.00455.x. [DOI] [PubMed] [Google Scholar]
  • 66.Borenstein M. Software for Publication Bias. In: Rothstein H, Sutton AJ, Borenstein M, editors. Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments. Wiley; West Sussex, United Kingdom: 2005. [Google Scholar]
  • 67.Emerson JD, Strenio J. Boxplots and Batch Comparisons. In: Hoaglin DC, Mosteller F, Tukey JW, editors. Understanding Robust and Exploratory Data Analysis. Wiley; New York: 1983. pp. 58–96. [Google Scholar]
  • 68.Terrin N, Schmid CH, Lau J, Olkin I. Adjusting for publication bias in the presence of heterogeneity. Statistics in Medicine. 2003;22(13):2113–2126. doi: 10.1002/sim.1461. [DOI] [PubMed] [Google Scholar]
  • 69.Lau J, Ioannidis JP, Terrin N, Schmid CH, Olkin I. The case of the misleading funnel plot. British Medical Journal. 2006;333(7568):597–600. doi: 10.1136/bmj.333.7568.597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Sutton AJ. Publication Bias. In: Cooper H, Hedges LV, Valentine JC, editors. The Handbook of Research Synthesis and Meta-Analysis. 2nd ed. Sage; New York: 2009. pp. 435–452. [Google Scholar]
  • 71.Petersen JL, Hyde JS. Gender Differences in Sexual Attitudes and Behaviors: A Review of Meta-Analytic Results and Large Datasets. Journal of Sex Research. 2011;48(2/3):149–165. doi: 10.1080/00224499.2011.551851. [DOI] [PubMed] [Google Scholar]
  • 72.Schroder KE, Carey MP, Vanable PA. Methodological challenges in research on sexual risk behavior: II. Accuracy of self-reports. Annals of Behavioral Medicine. 2003 Oct;26(2):104–123. doi: 10.1207/s15324796abm2602_03. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Centers for Disease Control and Prevention HIV in the United States: At A Glance. [May 8, 2015];Fast Facts. 2015 http://www.cdc.gov/hiv/pdf/statistics_basics_ataglance_factsheet.pdf.
  • 74.Dermen KH, Cooper ML. Sex-related alcohol expectancies among adolescents: I. Scale development. Psychology of Addictive Behaviors. 1994;8(3):152. [Google Scholar]
  • 75.George WH, Davis KC, Masters NT, et al. Sexual Victimization, Alcohol Intoxication, Sexual-Emotional Responding, and Sexual Risk in Heavy Episodic Drinking Women. Archives of Sexual Behavior. 2013 doi: 10.1007/s10508-013-0143-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Masters NT, George WH, Davis KC, et al. Women's unprotected sex intentions: roles of sexual victimization, intoxication, and partner perception. Journal of Sex Research. 2014;51(5):586–598. doi: 10.1080/00224499.2012.763086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Schacht RL, George WH, Heiman JR, et al. Effects of alcohol intoxication and instructional set on women's sexual arousal vary based on sexual abuse history. Archives of Sexual Behavior. 2007;36(5):655–665. doi: 10.1007/s10508-006-9147-y. [DOI] [PubMed] [Google Scholar]
  • 78.Gilmore AK, Schacht RL, George WH, et al. Assessing women's sexual arousal in the context of sexual assault history and acute alcohol intoxication. Journal of Sexual Medicine. 2010;7(6):2112–2119. doi: 10.1111/j.1743-6109.2010.01786.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Gordon CM. Alcohol's influence on requisites for HIV-risk reduction [Dissertation] Psychology, Syracuse University; Syracuse, NY: 1996. [Google Scholar]
  • 80.Gordon CM, Carey MP. Alcohol's effects on requisites for sexual risk reduction in men: An initial experimental investigation. Health Psychology. 1996;15(1):56–60. doi: 10.1037//0278-6133.15.1.56. [DOI] [PubMed] [Google Scholar]
  • 81.MacDonald TK, Fong GT. The effects of alcohol on intentions to use condoms.. Paper presented at: 102nd Annual Meeting of the American Psychological Association; Los Angeles, CA.. August 12-16, 1994. [Google Scholar]
  • 82.Stoner SA, Norris J, George WH, et al. Women's condom use assertiveness and sexual risk-taking: effects of alcohol intoxication and adult victimization. Addictive Behaviors. 2008;33(9):1167–1176. doi: 10.1016/j.addbeh.2008.04.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Norris J, Stoner SA, Hessler DA, et al. Influences of Sexual Sensation Seeking, Alcohol Consumption, and Sexual Arousal on Women's Behavioral Intentions Related to Having Unprotected Sex. Psychology of Addictive Behaviors. 2009;23(1):14–22. doi: 10.1037/a0013998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84.Zawacki T, Norris J, Hessler DM, et al. Effects of Relationship Motivation, Partner Familiarity, and Alcohol on Women's Risky Sexual Decision Making. Personality and Social Psychology Bulletin. 2009;35(6):723–736. doi: 10.1177/0146167209333043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 85.Schacht RL. Women's Sexual Behavior, Attitudes, and Alcohol-Related Sex Risk following Sexual Assault in Childhood, Adulthood, or Both [Ph.D.] University of Washington; Ann Arbor: 2011. [Google Scholar]
  • 86.Stoner SA, George WH, Peters LM, Norris J. Liquid courage: alcohol fosters risky sexual decision-making in individuals with sexual fears. AIDS and Behavior. 2007;11(2):227–237. doi: 10.1007/s10461-006-9137-z. [DOI] [PubMed] [Google Scholar]
  • 87.Wray TB. Delay discounting and risky sexual intentions: A test of the direct and indirect effects of alcohol intoxication and physiological arousal [Ph.D.] University of South Dakota; Ann Arbor: 2013. [Google Scholar]
  • 88.Wilkinson GS. The Wide Range Achievement Test (WRAT3) Administration Manual. Wide Range; Wilmington, DE: 1993. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Digital Content A
Supplemental Digital Content B

RESOURCES