Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 22.
Published in final edited form as: Am J Drug Alcohol Abuse. 2019 Jul 16;46(1):98–108. doi: 10.1080/00952990.2019.1624765

Fluctuations in Daily PTSD Symptoms Are Related to Proximal Alcohol Use: A Micro-Longitudinal Study of Women Victims of Intimate Partner Violence

Tami P Sullivan a, Stephen Armeli b, Howard Tennen c, Nicole H Weiss d, Nathan B Hansen e
PMCID: PMC12924698  NIHMSID: NIHMS2142207  PMID: 31311330

Abstract

Background.

One-third of women who experience intimate partner violence (IPV) are identified as having alcohol use problems. Yet, little research has examined factors that may increase the risk of alcohol use among this high-risk population.

Objectives.

This study overcomes limitations of previous research by using micro-longitudinal methods to examine how fluctuations in PTSD symptoms throughout the day are associated with proximal drinking behavior and whether these associations are related to individuals’ overall PTSD severity and race/ethnicity.

Methods.

Using phone-based interactive voice response, 244 female victims of current IPV reported their PTSD symptoms and drinking four times daily for 14 days.

Results.

Results indicated positive associations between PTSD symptom cluster severity and drinking level at the person, daily and within-day levels. The effects of within-person fluctuations in daily levels of PTSD severity on levels of drinking were stronger for individuals with lower PTSD severity. No evidence was found for within-person differences on time-lagged effects of PTSD on drinking or by racial/ethnic group. Further, in time-lagged models no evidence was found for reverse causation whereby alcohol use predicts increased PTSD symptom severity.

Conclusions.

Findings suggest that IPV-exposed women use alcohol to alleviate their PTSD symptoms at the micro-process level and that prevention and treatment efforts targeting PTSD symptoms may be useful in reducing alcohol use in this population. Further, these efforts should consider the overall severity of PTSD symptoms experienced given the differential findings among women with higher vs. lower PTSD symptom severity.

Keywords: posttraumatic stress disorder, intimate partner violence, alcohol use, women, micro-longitudinal data

Introduction

Women who experience intimate partner violence (IPV-exposed women) have rates of substance misuse up to seven times higher than women nationally (1, 2), with approximately one-third identified as having alcohol use problems (3). However, little research has examined factors that may explain victims’ increased risk of alcohol use. Posttraumatic stress disorder (PTSD) has emerged as one possible factor. PTSD is widespread among IPV-exposed women (4) with a weighted mean prevalence rate of 64% across 11 studies (5). Moreover, 72% of victims report functional impairment related to PTSD symptoms regardless of whether they meet full criteria for the disorder (6). PTSD is associated with alcohol use (7, 8), though PTSD-alcohol use findings are largely cross-sectional, which limits implications.

Consistent with self-medication (9, 10) and negative reinforcement (11) theories of alcohol use, IPV-exposed women may drink to mitigate the negative experiences of PTSD symptoms (12–14). PTSD may influence alcohol use temporally in at least two ways. The first is the general temporal order (i.e., at the macro level) whereby PTSD precedes alcohol misuse during a person’s life course. Research supports this sequencing in so far as victimization and PTSD tend to precede the development of alcohol use problems (15, 16). The second is a more discrete temporal order (i.e., at the micro process level) whereby an individual has a drink subsequent to experiencing a PTSD symptom(s) (17). For example, a woman may experience PTSD intrusion symptoms such as flashbacks in the early afternoon and drink nearly immediately or later that day to reduce/eliminate the effects of those symptoms.

Alternative theories to self-medication and negative reinforcement have been proposed to explain the PTSD-alcohol use relation, for example, mutual maintenance (18–20). Specific to PTSD-alcohol use, this model is both inclusive of and expands on self-medication and negative reinforcement. Mutual maintenance posits that PTSD symptoms contribute to alcohol use, which, in turn maintains PTSD symptom severity or even exacerbates it. Clearly, relations between PTSD symptoms and alcohol use may be complex. Thus, they require designs/methods that allow for the determination of their temporal association.

To test how processes unfold throughout the day to influence drinking, it is necessary to use micro-longitudinal methods. Such methods capture data frequently and in near real-time, and assess experiences and behaviors as they unfold in their natural environment, allowing the examination of within-person, proximal relationships (21). These benefits are critical to conducting a more sensitive test of models of alcohol use (at the micro-process level). Emerging research has begun to examine the micro-processes between PTSD and alcohol use, with many studies using micro-longitudinal methods in participants’ natural environments (e.g., 19, 22) to examine proximal relations and identify contingencies to be targeted in treatment. For example, PTSD symptom fluctuations were associated concurrently with alcohol and alcohol dependence symptoms (17, 20, 23), greater overall PTSD severity was associated with greater alcohol craving (22), and individuals were more likely to drink on days when they experienced greater behavioral avoidance and intrusion PTSD symptoms (19). A key study tested proximal relations of PTSD symptom severity to drinking among veterans with four times daily reporting (18). Results of concurrent effects models showed PTSD symptom severity was positively associated with number of drinks consumed, which provided support for the self-medication hypothesis. Results of time-lagged models revealed mixed support for the self-medication hypothesis, showing no association with a one-interval lag, but revealing an increase in drinking the next day. Regarding mutual maintenance, lagged effects findings did not support this model. On the contrary, increases in alcohol use were associated with decreases in PTSD symptom severity in the next interval. To our knowledge, this is the most sensitive test to date of the self-medication and mutual maintenance models of relations between PTSD and alcohol use.

Though this emerging body of research advances our understanding of the micro-processes between PTSD and alcohol use, no studies have tested the hypothesis that there are proximal relations between PTSD symptoms and alcohol use among IPV-exposed women in particular. Further, no studies have examined how these relations may differ based on the overall severity of an individual’s PTSD symptoms. For instance, participants in the studies cited above are no longer exposed to their referent traumatic events (i.e., combat veterans), unlike IPV victims who are chronically exposed to their trauma (i.e., experience repeated victimization) and traumatic stressors (e.g., abusive partner, environment in which IPV occurs) (24, 25).

The study of the PTSD-alcohol association, whether at the cross-sectional level or micro-longitudinal level, has focused largely on PTSD symptom severity. However, some research shows that associations between PTSD and substance use can differ by PTSD symptom cluster experienced (i.e., intrusion, avoidance, negative alterations in cognition and mood, alterations in arousal and reactivity) and class of substance used (i.e., alcohol, drugs, tobacco) (19, 26–28), suggesting that alcohol may be selected because of its depressant and anxiolytic effects (29). Cross-sectional research among IPV victims has revealed that clusters have differential relations with alcohol use, for example that avoidance and numbing symptoms are more frequently experienced among alcohol users than re-experiencing symptoms (28). Micro-longitudinal research among sexually assaulted college students has also revealed differential relations whereby intrusive and behavioral avoidance PTSD symptoms (but not dysphoric or hyperarousal PTSD symptoms) were related to drinking level at the day level(19). These findings demonstrate the need to understand the role of each symptom cluster in relation to alcohol use at the daily level, among IPV victims.

Further, extant literature reveals some differences in rates of IPV, PTSD, and alcohol use by racial/ethnic group among Blacks, Whites and Latinas. Differences regarding IPV and PTSD most often show that rates are highest among Blacks, though there is variability by traumatic event exposure (1, 30–33). Differences regarding alcohol use reveal nuanced differences depending on how alcohol use is assessed. For example, a review of data from the National Survey on Drug Use and Health (34) reports binge drinking rates are highest among Latinas and lowest among Blacks, whereas consumption measured in ounces is highest among Blacks and lowest among Latinas. However, to our knowledge, the role of race/ethnicity in the PTSD-alcohol use relation at the micro-process level has not been examined. Studies of relationships among these variables can inform the development of interventions that are culturally-tailored to the processes that may be unique to each racial/ethnic group.

The purposes of this study are to:

  1. Examine associations among PTSD symptom clusters and drinking across multiple levels of analysis such as at the between-person level (i.e., are average levels of PTSD symptom cluster severity associated with average levels of drinking?), the within-person day level (i.e., are daily fluctuations from women’s overall mean levels of PTSD symptom cluster severity associated with daily drinking level?), and within-day interval level (i.e., are within-day fluctuations from the day’s overall mean levels of PTSD symptom cluster severity related to average levels of drinking during that within-day period?; concurrent-time model).

  2. Examine time-lagged associations between PTSD symptom cluster severity and drinking level within-day (i.e., are within-day relative levels of PTSD symptom cluster severity related to changes in drinking level in the subsequent interval; time-lagged model).

  3. Examine whether the aforementioned associations vary as a function of overall PTSD severity, race/ethnicity and a woman’s mean levels of PTSD symptom cluster severity during the daily reporting phase.

  4. Test an alternative model, i.e., mutual maintenance, whereby within-day relative levels of drinking are related to changes in PTSD symptom cluster severity in the subsequent interval.

Methods

Sample

Two hundred seventy-nine women were recruited from two counties in New England using three methods. 1) We placed flyers with tear-off sheets with the study phone number in community establishments such as grocery stores, pizza shops, community kiosks, libraries, and selected state offices such as the Departments of Education and Employment as well as in waiting rooms, bathrooms, and exam rooms of urban-area primary care clinics. We also targeted neighborhoods and clinics including community health centers where minority racial/ethnic group members were highly represented to stratify participation by racial/ethnic group (Black, Latina, White). 2) We posted advertisements on the interior of mass transit vehicles that serve the cities and surrounding towns. 3) We posted information on community websites such as Craigslist. All materials were posted in English and Spanish. Eligibility was determined via a phone screen.

Inclusion and exclusion criteria were based on women’s self-report and include: (a) involvement in a heterosexual intimate relationship (because the dynamics in same-sex IPV-relationships contradict theories that have been developed in the context of heterosexual relationships) (35), (b) the presence of physical IPV victimization in the last three months in a relationship of at least six months duration, (c) the use of any amount of alcohol or drugs in the last three months; (d) continuous partner contact (i.e., saw partner at least twice weekly with no more than two weeks apart during the past month); (e) age 18 or older; (f) the ability to speak English or Spanish; and (g) and self-reported race/ethnicity as Black, White, or Latina, as the study was designed to stratify the sample and test for moderation. Exclusion criteria were a woman’s current psychiatric instability based on self-reported inpatient psychiatric hospitalization within the last year.

Because this study is identifying the extent to which PTSD symptoms are related to alcohol use at a micro-process level, we retained for analyses only those participants who provided daily data past the first three days of the study; this was the initial period after which we contacted participants to determine their interest/willingness/ability to continue to participate in the study’s daily component. See Table 1 for demographic information.

Table 1.

Descriptive information on demographics and key variables at the person-level.

Test Statistic
Demographic Variables
Age M = 37.2, SD = 12.9
Racial/ethnic background
 Black 43.4%
 Latina 27.2%
 White 26.5%
 More than one race/ethnicity or other 2.9%
Educational attainment (in years) M = 12.3, SD = 2.0
Past-year household income M = $19,003.7, SD = $18,299.5
Employment status
 Employed full-time 11.8%
 Employed part-time 25.4%
 Not in the labor force (e.g., student) 4.7%
 Unemployed 58.1%
Cohabitating 67.4%
Relationship length (in years) M = 7.3, SD = 7.2
Past-month intimate partner violence (IPV) severity
 Physical IPV M = 10.5, SD = 11.8
 Psychological IPV M = 39.9, SD = 12.0
 Sexual IPV M = 5.1, SD = 8.9
Diagnostic Data
 Probable posttraumatic stress disorder 35.5%
 Alcohol disorder 12.3%
 Drug use disorder 19.7%
Open in a new tab

Procedure

There were three study components; a baseline interview, daily data collection, and a follow-up interview. For the baseline interview, prospective participants met with a trained female interviewer in a private office who obtained informed consent and administered self-report measures in English or Spanish using computer-assisted interviewing (Questionnaire Development System; 36). During this interview session, subsequent to the completion of the self-report measures, participants were trained to use the interactive voice response (IVR) telephone system to record their information daily with training procedures modeled after those detailed in Stone and Shiffman (37). Participants were provided with wallet-size cards (that appeared to be for a Breast Cancer Awareness Study so as not to increase risk of IPV by their partners) with information to assist them in completing IVR calls. Daily participation consisted of calling the IVR system four times a day. Finally, a follow-up interview was scheduled with each participant regardless of the number of completed daily calls.

Participants completed IVR surveys for 30 days. A priori examination of the data revealed drop-off in adherence to the daily reporting protocol after day 14. Therefore, to examine the relationships between PTSD and alcohol use without bias introduced by an increase in missing data, we examine relationships in the first 14 days. The number and timing of the surveys was informed by recommendations for micro-longitudinal designs in the behavioral sciences (38, 39). Surveys took place between 7:00 a.m. and 10:00 a.m. (wake-up), 11:45 a.m. and 1:00 p.m. (mid-day), 4:45 p.m. and 6:00 p.m. (early evening), and 8:45 p.m. and 10:00 p.m. (nighttime). For each survey, participants were asked to report their PTSD symptoms and alcohol use since the previous call. Calls could be initiated by the participant or the participant could have elected to have the IVR system initiate a reminder call to her.

Women were remunerated up to $325, which included payments for baseline and follow-up interviews, and graduated payments with bonuses for compliance with daily surveys. Participants were provided with a list of community resources. Assistance with referrals was provided upon participant request. The interviewer also offered to develop a unique safety plan with the participant. The PI of the project, a licensed psychologist, was available on-call if participants required additional trauma-related support.

Measures

Daily PTSD symptom cluster severity.

Participants responded to questions regarding the 17 DSM-IV PTSD symptoms (40). To the extent possible, they were instructed to consider the IPV with their partner as the referent traumatic event. Consistent with recommendations (41, 42), we differentiated avoidance and numbing as separate clusters (i.e., re-experiencing (five symptoms); avoidance (three symptoms); numbing (four symptoms); and arousal (five symptoms)). Responses were rated on a scale of 1-5 where 5 indicated the highest level of symptom severity. If the participant did not experience a symptom, she was instructed to enter 0. Continuous cluster severity scores were created for each of the four clusters by averaging together their respective items. Cronbach alphas calculated on the interval-level data were .92 for re-experiencing, .84 for avoidance, .90 for numbing, and. 91 for arousal.

Daily alcohol use.

Participants reported the number of standard drinks they consumed. To avoid undue influence from extreme values, we Winsorized extreme values (43) by recoding values greater than 10 to 11 to maintain rank value while minimizing skew; less than 0.3% of the responses reported more than 10 standard drinks during a survey interval.

Person-level (baseline) PTSD.

The Posttraumatic Stress Diagnostic Scale (PDS; 44) was administered to assess the DSM-IV PTSD symptoms in relation to IPV victimization by a current male partner. Responses were rated on a scale from 0 (not at all, or only one time) to 3 (5 or more times a week, or almost always). A total symptom severity score was created by summing all 17 symptom severity items (Cronbach’s α = .88). Additionally, to characterize the sample, a probable PTSD diagnosis was calculated based on the presence of a Criterion A traumatic event; endorsement of at least one re-experiencing symptom, three avoidance/numbing symptoms, and two hyperarousal symptoms; duration of at least 1 month; and impairment in at least two areas of functioning.

Person-level (baseline) descriptive information on key variables.

To further characterize substance use, we administered the SCID substance use module to determine current alcohol or drug use disorder (45). Interviews were administered by master- or doctoral-level female research associates or postdoctoral fellows trained by the principal investigator. All interviews were reviewed by a PhD level clinician, with diagnoses confirmed in consensus meetings.

Data analysis

We tested the central questions with multilevel regression models using HLM software (46). Our design produced data corresponding to a 3-level structure, with level 1 being the four within-day intervals, level 2 being the day-level, and level 3 being the person-level. Given that the wake-up survey included information about PTSD and alcohol use that occurred the previous evening (after 10:00 p.m.) and the overnight period, we restructured the data so that these values became the late-night/overnight record for the previous day (i.e., the last assessment in temporal order for the previous day).

We tested concurrent-time models and time-lagged interval models. In the concurrent-time models, we examined the associations between the PTSD symptom cluster severity scores and drinking level during the same interval. This approach used all available data. For the time-lagged interval models, we first modeled the association between the PTSD symptom cluster severity in interval t as a predictor of the drinking occurring in interval t+1 (i.e., the subsequent interval) controlling for drinking level in interval t. For these models, we predicted drinking in the early-evening (4:45 to 6:00 p.m.), nighttime (8:45 to 10:00 p.m.) and the late-night/overnight (after 10:00 p.m.) intervals from the PTSD symptom cluster severity in the prior intervals. To test the reverse causal pattern, i.e., mutual maintenance, we then switched each of the PTSD symptom cluster severity scores to the dependent variable (i.e., PTSD in interval t+1 as a function of drinking level in interval t controlling for PTSD in interval t). For all time-lagged models, we had one fewer within-day record for analysis given the lag structure.

Because drinking level is a count, we used HLMs non-linear modeling specification when examining it as the dependent variable. Specifically, we estimated a Poisson sampling model with over-dispersion and a log-link function and we interpreted the unit-specific fixed effects. For models predicting PTSD symptom cluster severity, we used a linear model setup. For all models, we report robust standard errors. We mean-centered the predictors to obtain unbiased estimates of associations at each level of analyses. For the 3-level structure, the within-day interval (level 1) effects correspond to deviations (i.e., fluctuations) from the mean levels of that day; the effects for the day-level (level 2) predictors correspond to fluctuations from individuals’ overall mean level (i.e., reflects an increase relative to a participant’s mean); and the person-level effects (level 3) – predictors at this level were grand-mean centered – correspond to associations between overall mean levels (i.e., between person effects). In the person level portion of all models we included PTSD symptom cluster severity means, a baseline PTSD severity score, and two dummy codes for race/ethnicity (white was the reference group). For all models, we estimated random intercepts and slopes at level 1; if variance components for slopes were non-significant, we fixed them to zero.

Results

Descriptive statistics

We had 9,616 reporting intervals nested within 3,191 days nested within 244 persons for the cross-sectional models. Participants contributed a mean of 39.4 within-day intervals (SD = 15.1) out of a possible 56; this corresponded to daily adherence of 70.4%. Three-fourths (75.8%) of the sample completed at least 50% of the interval assessments. Descriptive information at the person-level (i.e., based on baseline data) regarding IPV severity and substance use are reported in Table 1.

Participants reported drinking during 22.3% of the within-day intervals. The nighttime interval (approximately covering the time between 6:00 p.m. and 10:00 p.m.) had the highest drinking rate (28.4%) and the mid-day interval (approximately covering the time between 10:00 a.m. and 1:00 p.m.) had the lowest drinking rate (14.7%). Participants drank on 37.6% of the reporting days. On average, individuals drank 2.64 drinks (SD = 2.04) within intervals when they drank and 4.76 drinks (SD = 4.67) on drinking days.

Participants reported experiencing PTSD symptoms during 69.4% of the within-day intervals. The afternoon interval had the fewest occurrences of symptoms (68.4%) and the late night/overnight interval had the greatest (71.4%). Participants had PTSD symptoms on 81.5% of reporting days. On average, they experienced symptoms from 3.11 (SD = 1.11) clusters within an interval when any symptoms were experienced.

Table 2 shows the descriptive statistics and correlations for the aggregated daily data variables and person variables. At the between-person level of analysis, PTSD symptom cluster severity scores were highly correlated (all correlations > .78). All of the PTSD symptom clusters showed moderate positive correlations with the aggregate levels of drinking. Person-level (baseline) PTSD severity showed strong associations with the aggregate daily PTSD clusters and weak associations with the aggregate drinking. Aggregate daily PTSD variables were moderately correlated with aggregate drinking.

Table 2.

Descriptive statistics and correlations

M SD 1 2 3 4 5
1 PTSD Severity 19.28 11.80
2 Re-experiencing 0.97 1.17 .628**
3 Avoidance 1.07 1.23 .611** .912**
4 Numbing 1.24 1.36 .599** .781** .820**
5 Arousal 1.14 1.25 .640** .874** .876** .905**
6 Number drinks 0.61 0.92 .163* .241** .249** .259** .298**
Open in a new tab

Note. N = 244. Values for all variables except Mean PTSD severity correspond to within-day interval means. PTSD severity is calculated at the person level.

**

p < .01,

*

p < .05

Multilevel regressions

Concurrent-time models

Results for concurrent-time models predicting the number of drinks consumed in the four within-day intervals are shown in Table 3. We examined the PTSD symptom clusters separately given their high correlations; each set of columns in the table corresponds to a model focusing on a specific cluster as the independent variable. We first modeled the additive effects of the predictors (results shown in first step); next, we incorporated the interaction effects (results shown in second step).

Table 3.

Concurrent time multilevel regression results predicting number of drinks.

Reexperiencing Avoidance Numbing Arousal
Step Predictor b p exp(b) b p exp(b) b p exp(b) b p exp(b)
1 Black .541 .026 1.718 .570 .021 1.768 .700 .007 2.014 .649 .009 1.913
Latina −.299 .269 .742 −.353 .189 .702 −.338 .226 .713 −.289 .296 .749
PTSD Severity −.012 .253 .988 −.012 .250 .988 −.014 .223 .986 −.022 .044 .978
Mean PTSD (Cluster) .345 .002 1.412 .363 .001 1.437 .329 .002 1.390 .459 <.001 1.583
Day PTSD (Cluster) .305 <.001 1.356 .213 .001 1.238 .201 .004 1.223 .341 <.001 1.407
Interval PTSD (Cluster) .183 .002 1.201 .184 .001 1.202 .423 <.001 1.526 .305 <.001 1.356
2 Black × Day cluster −.253 .169 .776 −.018 .907 .982 .152 .277 1.164 .045 .813 1.046
Latina × Day cluster −.181 .377 .834 .014 .932 1.014 .266 .111 1.305 .088 .696 1.093
PTSD Severity × Day cluster .007 .324 1.007 .003 .651 1.003 .007 .257 1.007 .005 .470 1.005
Mean PTSD (Cluster) × Day cluster −.164 .011 .849 −.160 .002 .852 −.171 .001 .843 −.107 .060 .898
Black × Interval cluster −.161 .248 .851 −.058 .664 .944 .009 .958 1.009 .108 .535 1.114
Latina × Interval cluster −.217 .218 .805 .088 .558 1.092 −.127 .513 .881 −.051 .795 .950
PTSD Severity × Interval cluster .002 .806 1.002 −.001 .840 .999 −.006 .399 .994 .002 .819 1.002
Mean PTSD (Cluster) × Interval cluster .000 .995 1.000 −.085 .056 .919 −.102 .047 .903 −.213 .001 .808
Open in a new tab

Note. Black = 1, Whites = 0; Latina: Latina = 1, Whiles = 0. Mean PTSD (cluster) = Overall person means for relevant PTSD cluster; Day PTSD (cluster) = daily means for relevant PTSD cluster; Interval PTSD (cluster) = within-day interval levels of relevant PTSD cluster. Exp(b) = increase in the rate of drinking for a unit increase in the predictor.

Results from the additive portion of the model indicate that Black participants had a greater number of drinks overall. We include the exponentiated slope (exp[b]) as an index of effect size. So, for example, the value for the Black contrast in the re-experiencing model indicates that controlling for the other predictors, the rate of drinking for Blacks, compared to whites, was 1.718 times higher. 1 In addition, we found that overall mean severity levels of the PTSD symptom clusters were associated with higher mean number of drinks consumed. We found similar positive associations at the day- and interval-level of analysis. Specifically, on days when PTSD symptom cluster severity levels were relatively higher than their overall mean levels (i.e., day PTSD cluster), individuals drank more. Finally, during within-day intervals when PTSD symptom cluster severity levels were relatively higher than that day’s overall mean severity level (i.e., interval PTSD cluster), individuals drank more.

In the multiplicative portion of model (step 2), we found no evidence that the within-person daily- and interval-level associations between PTSD symptom cluster severity and drinking varied across racial/ethnic groups. In contrast, we found a consistent moderating effect of overall mean levels of the PTSD symptom cluster severity on the associations between relative daily levels of PTSD symptom cluster severity and daily drinking. The form of this interaction was similar across all clusters and is illustrated in Figure 1; This shows the association between relative daily levels of re-experiencing (the x-axis), with low and high levels corresponding to the 5th and 95th percentiles for fluctuations from mean daily levels and, daily drinking levels (y-axis)2 as a function of overall mean levels of re-experiencing (with low and high values corresponding to the mean values in the lower and upper quartile of mean re-experiencing scores3). Specifically, Figure 1 shows that individuals with high mean severity levels for re-experiencing, compared to individuals with low mean severity levels, drank more overall, but showed a weaker positive association between fluctuations in daily symptom levels and daily drinking.

Figure 1.

Figure 1.

Open in a new tab

The association between relative levels of daily re-experiencing and daily drinking as a function of overall mean levels of re-experiencing

We also found interactions involving interval level (level 1) numbing and arousal and their corresponding mean levels. The form of the effect for arousal is illustrated in Figure 2 (the form for numbing was similar); we again used the 5th and 95th percentile for the low and high x-axis values and the mean of the lower and upper quartiles scores for low and high mean arousal levels, respectively. As shown, individuals with higher overall symptom severity levels of arousal drank more overall and showed no interval-level association. In contrast, individuals with lower overall arousal mean symptom severity showed a positive association between within-day variation in arousal and drinking level.

Figure 2.

Figure 2.

Open in a new tab

The association between relative levels of within-day interval levels of arousal and drinking as a function of overall mean levels of arousal

Lagged-interval models.

Results for the lagged-interval models are shown in Table 4. For these models, we had 5,834 level 1 observations nested within 2,472 days nested within 242 persons (two individuals did not have consecutive intervals, thus were omitted from these analysis). Given our focus was on possible lagged effects of PTSD clusters severity and drinking, and to reduce type 1 error inflation, we did not include cross-level interactions in these models. At the top portion of the table are the results for the models predicting drinking level, i.e., tests of self-medication; not shown are the results for person- and day-level predictors of drinking since these were reported in the concurrent-time models with the full sample of observations. Of central interest, in these models were the effects of interval t PTSD symptom cluster severity on interval t+1 drinking. We found no evidence of positive associations between PTSD symptom cluster severity and level of drinking using this time-lagged specification. On the contrary, we found that intervals characterized by relatively higher within-day levels of all of the PTSD symptom clusters (except re-experiencing) were associated with lower drinking levels in the subsequent interval.

Table 4.

Lag-time multilevel regression results

Drinking as Outcome at t+1 Reexperiencing Avoidance Numbing Arousal
Predictors b p exp(b) b p exp(b) b p exp(b) b p exp(b)
Interval t PTSD (cluster) −.129 .174 .879 −.170 .011 .844 −.248 .006 .780 −.198 .041 .820
Interval t drinking −.093 .001 .911 −.091 .004 .913 −.094 .002 .910 −.090 .003 .914
PTSD Cluster as Outcome at t+1 b p β b p β b p β b p β
Interval t Drinking .018 .153 .018 −.003 .780 −.002 −.003 .807 −.002 .007 .428 .007
Interval t PTSD (cluster) −.260 <.001 −.258 −.301 <.001 −.300 −.274 <.001 −.276 −.268 <.001 −.267
Open in a new tab

Note. Models predict outcomes at interval t+1. Interval t PTSD (cluster) = within-day interval levels of relevant PTSD cluster. Exp(b) = increase in the rate of drinking for a unit increase in the predictor. β = standardized slope.

Results for the lagged-interval models whereby alcohol use predicts PTSD cluster severity, i.e., test of the mutual maintenance model, are shown at the bottom of Table 4.4 Again, for parsimony, we report only the results for the level 1 (interval level) predictors.5 We found that across participants, there was no overall effect of relative drinking levels on later PTSD cluster severity. We also found a negative relationship for the lagged effect of PTSD cluster severity (for all types) whereby intervals characterized by relatively higher levels of PTSD severity were followed by intervals with lower levels of cluster severity.

Discussion

To our knowledge, this is the first study to examine the relations of PTSD symptoms and alcohol use among IPV-exposed women across multiple levels of analysis, including at the micro-process level, whereby both concurrent and lagged effects could be tested. Results from our concurrent time models are generally consistent with both self-medication (9, 10) and negative reinforcement (11) theories in that elevated levels of PTSD symptom cluster severity were associated with greater likelihood and amount of drinking. These effects were revealed at three levels of analysis: (1) Within day intervals with relatively higher levels of PTSD symptom cluster severity (e.g., afternoon, evening, late night) were characterized by a greater likelihood of drinking and greater number of drinks. (2) Similarly, days characterized by relatively higher levels of PTSD symptom cluster severity were characterized by a greater likelihood of drinking and a greater number of drinks. (3) Finally, individuals with higher overall levels of PTSD symptom cluster severity drank more often and in greater amounts. Of note, women’s average number of drinks on a drinking day (4.67) puts them in the range of hazardous alcohol use, which regardless of diagnostic criteria or relation to PTSD symptom severity, puts them at risk for negative consequences (47). Though findings revealed that Black participants drank more overall, there was no evidence that within-person processes between PTSD symptom cluster severity and alcohol behaviors differ by racial/ethnic group.

Further, some of the within-person associations between relative levels of PTSD symptom cluster severity and drinking varied as a function of overall levels of PTSD symptom cluster severity. In general, this trend indicated stronger associations in the positive direction among individuals with lower overall mean levels of the PTSD clusters. Our results show that individuals with more severe PTSD are more likely to drink, and drink in greater quantities, across all days and intervals. In other words, symptom fluctuations did not influence level of alcohol use among those with high PTSD severity. This might be due to the more chronic nature of their symptoms and, perhaps, the habitual use of alcohol as a strategy to alleviate symptoms. In contrast, individuals with lower overall levels of PTSD severity might only be at risk for drinking and drinking a greater number of drinks when symptoms “flare up,” meaning that they experience symptoms more intensely than is typical for them and drink to mitigate these symptoms. This finding also suggests that, in this latter group, alcohol use may be effective at reducing symptoms. We note, however, that our findings at this level are cross-sectional.

Results from models testing self-medication/negative reinforcement time-lagged effects did not support a positive association between fluctuations in PTSD symptom cluster severity on drinking in subsequent time periods. In fact, support for the opposite direction was found. Specifically, we found that relatively higher levels of numbing, arousal, and avoidance were associated with lower drinking levels in subsequent periods. It should be noted that these models controlled for the effects of drinking from the prior interval, which our concurrent time models indicated were positively related to PTSD symptom cluster severity. This is possibly an artifact of how quickly drinking occurs in response to elevated levels of PTSD symptoms and that drinking levels will regress to lower levels after periods of increased drinking. This pattern also is seen in the negative association between the within-day lagged effects of the drinking outcomes.

Results from our time-lagged models provided little evidence for the mutual maintenance model, i.e., drinking predicting subsequent increases in PTSD symptom cluster severity. On average, there was no association between relative increases in drinking and subsequent PTSD cluster severity. Again, however, the null time-lagged findings do not rule out the mutual maintenance model (or the reverse causal framework) as findings from both the concurrent-time and aggregate day portion of the first set of models tested could be interpreted in either causal direction, i.e., drinking exacerbating PTSD cluster severity or vice versa.

Some cluster level differences were detected among the concurrent-time and lagged-interval models regarding self-medication/negative reinforcement. Prior evidence suggests differential associations among the specific PTSD symptom clusters and alcohol use (26–28). Specifically, alcohol, which is associated with depressant and anxiolytic effects (29), is more frequently used among individuals who experience higher levels of PTSD arousal symptoms (48, 49). Such findings provide support for a functional association among PTSD and alcohol use, underscoring relations among PTSD symptom clusters and substances with properties that counter those symptoms. Notably, in extending cross-sectional work in this area, our findings provide evidence for a more complex understanding. For instance, at the between-person level of analysis, all of the PTSD symptom clusters showed moderate positive correlations with the aggregate levels of alcohol use, whereas at the within-person level of analysis findings were mixed for the PTSD symptom clusters. In particular, we found interaction effects such that fluctuations in numbing and arousal cluster symptom severity (with a trend effect for avoidance) were associated with drinking level in the concurrent time models, but only for individuals with lower overall PTSD symptom severity. This suggests that, as noted above, individuals may select alcohol because of its depressant and anxiolytic effects to target these symptoms. Our findings are consistent with results of certain studies showing that emotional numbing is associated with alcohol misuse (26) but inconsistent with others, that revealed drinking increased on days where individuals experienced avoidance and intrusion symptoms (19). One possibility for the divergent findings are differences in the samples relative to their referent traumatic events for PTSD and their ongoing exposure to their traumatic stressors (i.e., combat veterans and sexually assault college students, respectively compared to IPV victims). Additional research is needed to better understand these findings.

The current study is not without limitations. Our measure of PTSD was based on the DSM-IV-TR classification. The completion rate of daily assessments in our study is slightly lower than certain other studies using daily assessment methods (e.g., 50, 51, 52); this likely is related to the fact that we opted for the most generalizable sample and, therefore, did not exclude women with significant substance use problems or who are unstably housed as other studies have. Findings cannot be assumed to generalize to non-IPV/substance using populations and require replication across a more diverse group of individuals who experience IPV (e.g., men, women in same-sex intimate relationships). Our null findings for lagged effects of PTSD symptom cluster severity on drinking (or vice versa) does not rule out this possibility. Our models were somewhat constrained by the within day reporting windows we chose. It is possible that such effects might unfold across shorter periods and/or might vary in terms the lag duration. Future studies using more fine-grained reporting strategies could further test this possibility. Finally, given the prevalence of drug use disorders in this sample, it is important to understand how the PTSD-alcohol use association differs among those who do vs. do not use drugs. Future research should examine these relations as the complexity of the models analyzed here makes inclusion of such analyses beyond the focus of this manuscript.

Despite limitations, our results extend extant research on the role of PTSD symptoms in alcohol use generally, and among IPV-exposed women specifically. Our findings suggest that IPV-exposed women drink to alleviate their PTSD symptoms at the micro-process level. These results have important implications for clinical practice. For instance, evidence of a functional association between PTSD symptoms and alcohol use suggests the importance of assessing PTSD symptoms among alcohol-using populations (53), and alcohol use among populations characterized by PTSD symptoms (54). Further, results indicate that behavioral and/or pharmaceutical prevention and treatment efforts targeting PTSD symptoms may be useful in reducing alcohol use in this population, consistent with etiological and treatment research (55–57). Intervention research that enhances strategies for coping with PTSD (58) could be a valuable starting point.

Acknowledgments

This project was supported by Grant No. R01DA031275, awarded by the National Institute on Drug Abuse to Tami Sullivan/Yale University. Support was also provided by NIH grants K23DA039327 (NHW) and L30DA038349 (NHW).

Footnotes

1.

For quantitative predictors, the exp(b) coefficient corresponds to the increase in the rate of drinking for a unit increase in the predictor. Values for exp(b) less than 1.0 correspond to the rate decrease.

2.

The Y-axis in this figure corresponds to the mean drinking levels across the 4 within-day intervals, not the total day sum.

3.

For this interaction and the one shown in Figure 2, we used these values instead of plus/minus 1 SD from the mean because the moderators were positively skewed and using the conventional minus 1 SD below the mean to convey low levels was out of the range of the observed values.

4.

These models were identical to the drinking level models, except PTSD cluster severity at interval t+1 is the dependent variable and we replaced mean levels of PTSD clusters severity as a level 3 predictor with mean drinking level.

5.

For descriptive purposes, we report standardized partial slopes as an effect size. We calculated these values by standardizing the data prior to model estimation. All significance tests are based on the unstandardized slopes.

References

  • 1.Substance Abuse and Mental Health Services Administration. Results from the 2008 National Survey on Drug Use and Health: National Findings, DHHS Publication No. SMA 09-4434; Rockville, MD: Office of Applied Studies, 2009; [Google Scholar]
  • 2.Devries KM, Child JC, Bacchus LJ, Mak J, Falder G, Graham K, Watts C, Heise L. Intimate partner violence victimization and alcohol consumption in women: a systematic review and meta-analysis. Addiction 2014; 109 (3), 379–391. [DOI] [PubMed] [Google Scholar]
  • 3.Sullivan TP, Cavanaugh CE, Buckner JD, Edmondson D. Testing posttraumatic stress as a mediator of physical, sexual, and psychological intimate partner violence and substance problems among women. Journal of Traumatic Stress 2009; 22 (6), 575–584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Spencer C, Mallory AB, Cafferky BM, Kimmes JG, Beck AR, Stith SM. Mental health factors and intimate partner violence perpetration and victimization: A meta-analysis. Psychology of violence 2019; 9 (1), 1–17. [Google Scholar]
  • 5.Golding JM. Intimate partner violence as a risk factor for mental disorders: A meta-analysis. J Fam Violence 1999; 14 (2), 99–132. [Google Scholar]
  • 6.Hellmuth JC, Jaquier V, Swan SC, Sullivan TP. Elucidating Posttraumatic Stress Symptom Profiles and Their Correlates Among Women Experiencing Bidirectional Intimate Partner Violence. Journal of Clinical Psychology 2014; 70 (10), 1008–1021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sullivan TP, Ashare RL, Jaquier V, Tennen H. Risk factors for alcohol-related problems among victims of partner violence. Subst Use Misuse 2012; 47 (6), 673–685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Sullivan TP, Weiss NH, Flanagan JC, Willie TC, Armeli S, Tennen H. PTSD and Daily Co-Occurrence of Drug and Alcohol Use Among Women Experiencing Intimate Partner Violence. Journal of Dual Diagnosis 2016; 12 (1), 36–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Khantzian EJ. The self-medication hypothesis of substance use disorders: a reconsideration and recent applications. Harvard Review of Psychiatry 1997; 4 (5), 231–244. [DOI] [PubMed] [Google Scholar]
  • 10.Khantzian EJ. The self-medication hypothesis revisited: The dually diagnosed patient. Primary Psychiatry 2003; 10 (9), 47–48, 53–54. [Google Scholar]
  • 11.Baker TB, Piper ME, McCarthy DE, Majeskie MR, Fiore MC. Addiction Motivation Reformulated: An Affective Processing Model of Negative Reinforcement. Psychological Review 2004; 111 (1), 33–51. [DOI] [PubMed] [Google Scholar]
  • 12.Kaysen D, Dillworth TM, Simpson T, Waldrop A, Larimer ME, Resick PA. Domestic violence and alcohol use: Trauma-related symptoms and motives for drinking. Addict Beh 2007; 32 (6), 1272–1283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cohen LR, Field C, Campbell ANC, Hien DA. Intimate partner violence outcomes in women with PTSD and substance use: A secondary analysis of NIDA Clinical Trials Network “Women and Trauma” Multi-site Study. Addict Beh 2013; 38 (7), 2325–2332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Murdoch L, Jina P, Jitender S, M. BJ The use of alcohol and drugs to self-medicate symptoms of posttraumatic stress disorder. Depression and anxiety 2010; 27 (8), 731–736. [DOI] [PubMed] [Google Scholar]
  • 15.Brady KT, Back SE. Childhood trauma, posttraumatic stress disorder, and alcohol dependence. Alcohol research : current reviews 2012; 34 (4), 408–413. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Cross D, Crow T, Powers A, Bradley B. Childhood trauma, PTSD, and problematic alcohol and substance use in low-income, African-American men and women. Child Abuse & Neglect 2015; 44, 26–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ouimette P, Read JP, Wade M, Tirone V. Modeling associations between posttraumatic stress symptoms and substance use. Addict Beh 2010; 35 (1), 64–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Possemato K, Maisto SA, Wade M, Barrie K, McKenzie S, Lantinga LJ, Ouimette P. Ecological momentary assessment of PTSD symptoms and alcohol use in combat veterans. Psychology of addictive behaviors 2015; 29 (4), 894–905. [DOI] [PubMed] [Google Scholar]
  • 19.Kaysen D, Atkins DC, Simpson TL, Stappenbeck CA, Blayney JA, Lee CM, Larimer ME. Proximal relationships between PTSD symptoms and drinking among female college students: Results from a daily monitoring study. Psychology of Addictive Behaviors 2014; 28 (1), 62–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Simpson TL, Stappenbeck CA, Luterek JA, Lehavot K, Kaysen DL. Drinking motives moderate daily relationships between PTSD symptoms and alcohol use. Journal of Abnormal Psychology 2014; 123 (1), 237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bolger N, Laurenceau J. Intensive longitudinal methods: an introduction to diary and experience sampling research. New York: Guilford; 2013. [Google Scholar]
  • 22.Simpson TL, Stappenbeck CA, Varra AA, Moore SA, Kaysen D. Symptoms of posttraumatic stress predict craving among alcohol treatment seekers: results of a daily monitoring study. Psychology of Addictive Behaviors 2012; 26 (4), 724–733. [DOI] [PubMed] [Google Scholar]
  • 23.Simons JS, Simons RM, Keith JA, Grimm KJ, Stoltenberg SF, O’Brien C, Andal K. PTSD symptoms and alcohol-related problems among veterans: Temporal associations and vulnerability. Journal of Abnormal Psychology 2018; 127 (8), 733–750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Cattaneo LB, Goodman LA. Risk Factors for Reabuse in Intimate Partner Violence: A Cross-Disciplinary Critical Review. Trauma Violence Abuse 2005; 6 (2), 141–175. [DOI] [PubMed] [Google Scholar]
  • 25.Kuijpers KF, Van der Knaap LM, Winkel FW. Victims’ Influence on Intimate Partner Violence Revictimization:An Empirical Test of Dynamic Victim-Related Risk Factors. J Interpers Violence 2012; 27 (9), 1716–1742. [DOI] [PubMed] [Google Scholar]
  • 26.Jakupcak M, Tull MT, McDermott MJ, Kaysen D, Hunt S, Simpson T. PTSD symptom clusters in relationship to alcohol misuse among Iraq and Afghanistan war veterans seeking post-deployment VA health care. Addict Beh 2010; 35 (9), 840–843. [DOI] [PubMed] [Google Scholar]
  • 27.Shipherd JC, Stafford J, Tanner LR. Predicting alcohol and drug abuse in Persian Gulf War veterans: What role do PTSD symptoms play? Addict Beh 2005; 30 (3), 595–599. [DOI] [PubMed] [Google Scholar]
  • 28.Sullivan TP, Holt LJ. PTSD symptom clusters are differentially related to substance use among community women exposed to intimate partner violence. Journal of Traumatic Stress 2008; 21 (2), 173–180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Eckardt MJ, File SE, Gessa GL, Grant KA, Guerri C, Hoffman PL, Kalant H, Koob GF, Li T-K, Tabakoff B. Effects of Moderate Alcohol Consumption on the Central Nervous System*. Alcoholism: Clinical and Experimental Research 1998; 22 (5), 998–1040. [DOI] [PubMed] [Google Scholar]
  • 30.Roberts AL, Gilman SE, Breslau J, Breslau N, Koenen KC. Race/ethnic differences in exposure to traumatic events, development of post-traumatic stress disorder, and treatment-seeking for post-traumatic stress disorder in the United States. Psychological Medicine 2010; First View, 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hatch S, Dohrenwend B. Distribution of Traumatic and Other Stressful Life Events by Race/Ethnicity, Gender, SES and Age: A Review of the Research. American Journal of Community Psychology 2007; 40 (3), 313–332. [DOI] [PubMed] [Google Scholar]
  • 32.Sullivan TP, Cavanaugh CE, Ufner MJ, Swan SC, Snow DL. Relationships Among Women’s Use of Aggression, Their Victimization, and Substance Use Problems: A Test of the Moderating Effects of Race/Ethnicity. Journal of Aggression, Maltreatment & Trauma 2009; 18 (6), 646–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Caetano R, Cunradi CB, Schafer J, Clark CL. Intimate partner violence and drinking patterns among white, black, and Hispanic couples in the U.S. J Subst Abuse 2000; 11 (2), 123–138. [DOI] [PubMed] [Google Scholar]
  • 34.Vaeth PAC, Wang-Schweig M, Caetano R. Drinking, Alcohol Use Disorder, and Treatment Access and Utilization Among U.S. Racial/Ethnic Groups. Alcoholism: Clinical and Experimental Research 2017; 41 (1), 6–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Johnson MP, Ferraro KJ. Research on domestic violence in the 1990s: Making distinctions. Journal of Marriage & the Family 2000; 62 (4), 948–963. [Google Scholar]
  • 36.NOVA Research Company. Questionnaire Development System TM User’s Manual. 2.0 Ed.; Bethesda, MD: NOVA Research Company, 2003. [Google Scholar]
  • 37.Stone AA, Shiffman S. Capturing momentary, self-report data: A proposal for reporting guidelines. Annals of Behavioral Medicine 2002; 24 (3), 236–243. [DOI] [PubMed] [Google Scholar]
  • 38.Stone AA, Shiffman S. Ecological momentary assessment (EMA) in behavorial medicine. Annals of Behavioral Medicine 1994; 16 (3), 199–202. [Google Scholar]
  • 39.Hektner JM, Schmidt J,A, Csikszentmihalyi M Experience sampling method: Measuring the quality of everyday life; Thousand Oaks: Sage Publications, Inc., 2007. [Google Scholar]
  • 40.American Psychiatric Association. Diagnostic and statistical manual of mental disorders (4th ed.); Washington, DC: American Psychiatric Publishing, Inc., 1994. [Google Scholar]
  • 41.King DW, Leskin GA, King LA, Weathers FW. Confirmatory factor analysis of the clinician-administered PTSD scale: Evidence for the dimensionality of posttraumatic stress disorder. Psychological Assessment 1998; 10 (2), 90–96. [Google Scholar]
  • 42.Yufik T, Simms LJ. A meta-analytic investigation of the structure of posttraumatic stress disorder symptoms. Journal of abnormal psychology 2010; 119 (4), 764–776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Liao H, Li Y, Brooks G. Outlier impact and accommodation methods: Multiple comparisons of Type I error rates. Journal of Modern Applied Statistical Methods 2016; 15 (1), 23. [Google Scholar]
  • 44.Foa EB, Cashman L, Jaycox L, Perry K. The validation of a self-report measure of posttraumatic stress disorder: The Posttraumatic Diagnostic Scale. Psychological Assessment 1997; 9 (4), 445–451. [Google Scholar]
  • 45.First MB, Spitzer RL, Miriam G, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P). New York: Biometrics Research, New York State Psychiatric Institute, 2002. [Google Scholar]
  • 46.Raudenbush S, Bryk A, Cheong YF, & Congdon R . HLM 6: Hierarchical Linear and Nonlinear Modeling. lincolnwood,IL: Scientific Software International, 2004. [Google Scholar]
  • 47.Babor TF, Higgins-Biddle JC, Saunders JB, Monteiro MG. The Alcohol Use Disorders Identification Test. 2nd Ed.: World Health Organization, 2001. [Google Scholar]
  • 48.Stewart SH, Conrod PJ, Pihl RO, Dongier M. Relations between posttraumatic stress symptom dimensions and substance dependence in a community-recruited sample of substance-abusing women. Psychology of Addictive Behaviors 1999; 13 (2), 78–88. [Google Scholar]
  • 49.Saladin ME, Brady KT, Dansky BS, Kilpatrick DG. Understanding comorbidity between PTSD and substance use disorders: two preliminary investigations. Addict Beh 1995; 20 (5), 643–655. [DOI] [PubMed] [Google Scholar]
  • 50.Searles JS, Helzer JE, Rose GL, Badger GJ. Concurrent and retrospective reports of alcohol consumption across 30, 90 and 366 days: Interactive voice response compared with the timeline follow back. Journal of Studies on Alcohol 2002; 63 (3), 352–362. [DOI] [PubMed] [Google Scholar]
  • 51.Searles JS, Perrine MW, Mundt JC, Helzer JE. Self-report of drinking using touch-tone telephone: Extending the limits of reliable daily contact. Journal of Studies on Alcohol 1995; 56 (4), 375–382. [DOI] [PubMed] [Google Scholar]
  • 52.Collins RL, Kashdan TB, Gollnisch G. The Feasibility of Using Cellular Phones to Collect Ecological Momentary Assessment Data: Application to Alcohol Consumption. Experimental & Clinical Psychopharmacology 2003; 11 (1), 73–78. [DOI] [PubMed] [Google Scholar]
  • 53.Brady KT, Killeen T, Saladin ME, Dansky B, et al. Comorbid substance abuse and posttraumatic stress disorder: Characteristics of women in treatment. Am J Addict 1994; 3 (2), 160–164. [Google Scholar]
  • 54.Brown PJ, Stout RL, Mueller T. Substance use disorder and posttraumatic stress disorder comorbidity: Addiction and psychiatric treatment rates. Psychology of Addictive Behaviors 1999; 13 (2), 115–122. [Google Scholar]
  • 55.Brady KT, Back SE, Coffey SF. Substance abuse and posttraumatic stress disorder. Current Directions in Psychological Science 2004; 13 (5), 206–209. [Google Scholar]
  • 56.Ouimette P, Read JP. Trauma and substance abuse: Causes, consequences, and treatment of comorbid disorders, 2nd ed; Washington, DC, US: American Psychological Association, 2014. [Google Scholar]
  • 57.Roberts NP, Roberts PA, Jones N, Bisson JI. Psychological interventions for post-traumatic stress disorder and comorbid substance use disorder: A systematic review and meta-analysis. Clinical Psychology Review 2015; 38, 25–38. [DOI] [PubMed] [Google Scholar]
  • 58.Sullivan TP, Weiss NH, Price C, Pugh N, Hansen NB. Strategies for coping with individual PTSD symptoms: experiences of African American victims of intimate partner violence. Psychological trauma: theory, research, practice, and policy 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES