Purpose: This article examined the therapeutic effectiveness of expressive writing (EW) for psychological health, physical health, and recovery among illicit drug users.
Methods: This review involved an extensive search of PubMed, PsychINFO, Web of Science, CINAHL, and the grey literature for randomized trials that evaluated the effect of EW on psychological health, physical health, and recovery of illicit drug users. Nine studies were included in this review. Cohen’s d was computed, effect sizes were combined, and trials’ methodological quality was evaluated.
Results: Pooled and non-pooled results indicated no effect of EW on all the outcome measures both at posttreatment and at follow up. The range of effect sizes (ESs) was small (0.02 to 0.35), and most ESs (SMD) were below 0.20 e.g., depression (0.16, N= 218) and drug use (0.14, N= 125). Baseline levels of distress and life change moderated the effect of EW.
Conclusion: The results failed to support the therapeutic effectiveness of EW among drug users. Examined studies were deficient—had high unclear or high risk of bias and heterogeneity of procedures. Rigorous studies are needed to examine the therapeutic effectiveness of EW among drug users.
This article examines the available randomized control trials that used expressive writing in drug using samples for effects of the intervention on psychological health, physical health, and recovery. In this review “expressive writing” is defined as a self-exposure writing intervention in which participants write their deepest thoughts and feelings related to traumatic, stressful or emotional events spontaneously and continuously without paying attention to spelling or grammar—usually for 15-30 minutes on 3 to 4 consecutive days or weeks. “Psychological health” refers to the level of emotional and behavioral adjustment (the intervention effect is reflected by changes in symptoms of mental distress and dysfunction such as depression and anxiety). “Physical health” is indicated by the body’s ability to function and its general condition (e.g., sleep, pain, and nutrition). “Recovery” from illicit drug dependence is defined as processes of change that allow individuals achieve abstinence (e.g., seeking treatment, decreased frequency and quantity of drug use).
Illicit drugs are widespread worldwide. Use of illicit drugs considerably increases the burden of disease in both developed and developing countries . Drug victims struggle with high stress and negative mood over the chronic and relapsing course of their disease . They face enormous debilitating consequences of the stigma associated with the status of illicit drug use [3, 4]. Further, they encounter co-morbidity with various psychopathologies [5-7]. More, they confront a trail of drug related problems: serious infections such as HIV, financial difficulties, disturbed relations, violence, criminality and incarceration, homelessness, poor quality of life, and low self-esteem—which all increase the risk of suicide in this group [8, 9]. Because drug users are socially restrained, they do not express distress . Instead, a majority of them tend to use cognitive avoidant coping , which stimulates chronic rumination of negative intrusive thoughts i.e., thinking cycles repeatedly around negative thoughts (a chronic drain of their psychological resources), which impedes the ability of problem solving, goal achievement, and elevates their levels of distress [12, 13]. Therefore, drug sufferers use more drugs to dismiss or control their distress [14, 15]. Most psychotherapies of drug use incorporate narrative expression of miserable experiences as an attempt to facilitate positive ways for change and recovery [16, 17]. Expressive writing (EW) is a cheap narrative intervention that involves writing details of traumatic experiences for 15-30 minutes on consecutive or separate days or weeks. EW may produce health benefits in different clinical populations . Various mechanisms have been proposed to illustrate how EW might exert beneficial effects. Expressiveness— full disclosure of deeply and strongly held thoughts and emotions pertaining to negative experiences—is essential for benefits of EW to occur . EW is a method of self-exposure; it functions through exposure to traumatic thoughts and negative emotions, which leads to habituation and extinction . Re-organization of the traumatic schema is another mechanism: when people create coherent narratives, they reappraise the traumatic events, find a meaning for their experiences, and gain a sense of perceived control [18, 20]. This process counteracts avoidant coping; therefore, intrusions drop and therapeutic effects occur (reduction of psychological and physical symptoms) . Higher levels of post writing distress and more use of emotion and cognition words are associated with increased benefits of EW. Thus, the technique was modified to foster the release of negative emotions e.g., response training  or facilitate the integration of traumas in the working memory e.g., cognitive-processing, self-regulation , and benefit finding  .
In theory EW can be effective; however, the literature reports mixed results on its effectiveness [23-25]. To the author’s knowledge there are no reviews on EW among illicit drug users. Readers may be aware of a previous review about creative writing that focused on the therapeutic effect of art, poetry, singing to self, and EW. That study was narrative in nature, and it included only six studies as a whole (out of which 2 studies only used EW) . Hence, the effect of EW may not be clearly revealed apart from the effect of other writing therapies. Therefore, it is necessary to identify underlying factors that can be manipulated to heighten the success of EW in this population. To bridge this gap, a meta-analytic review was conducted to evaluate the effects of EW on psychological health, physical health, and recovery in alcohol and illicit drug using samples.
Materials and Methods
Data sources and inclusion criteria
A search of PubMed, PsychINFO, Web of Science, and CINAHL databases was conducted using a combination of keywords of expressive writing and drug use terms: expressive disclosure writing, expressive writing, writing disclosure, written emotional disclosure, structured writing, Pennebaker writing, therapeutic writing, linguistic analysis, drug depend*, drug misuse, drug abuse, drug addiction, illegal drug use, substance depend*, substance misuse, substance abuse, substance addiction, alcohol, substance-related disorders. The search was not limited by language, time, or study type. In addition, free hand search of Google Scholar and snowball backward search of the reference lists of selected publications were conducted. Eligible studies were: 1) randomized controlled trials that used EW as a primary or adjunctive intervention for alcohol or illicit drug users (not smokers), 2) in any setting 3) reporting outcomes of psychological health, physical health, or recovery from drug use. Studies with co-morbid conditions were included. Eligible studies for meta-analysis were trials that provided data for computing effect sizes (ESs) and confidence intervals (CIs).
Assessment of methodological quality
Two researchers independently evaluated selected studies for adequacy of the procedures of randomization, allocation concealment, blinding of participants and researchers to the group (the study aim/hypothesis) during intervention delivery. Detection bias was considered high if the interviewers or self-reporting participants were not blinded to the study aim/hypothesis. Attrition bias was considered low if drop out was < 30% and there were no between group differences in the characteristics of completers and non-completers. Meanwhile, attrition bias was considered high if drop out was > 30%, reasons for drop out were not mentioned, or characteristics of completers and non-completers varied between groups. Selective reporting was indicated high if an outcome measure in the methods section was not reported in the results. Other biases were considered e.g., non-compliance bias, sample size estimation, and selective publication bias.
Data extraction and analysis
Based on previous reviews a form was developed to extract data from eligible studies. Two researchers worked separately to screen selected studies and extract the following data: 1) study characteristics (aims, design, setting, participants’ characteristics, randomization procedures, inclusion and exclusion criteria, etc.), 2) outcome measures (instruments, measurement points, attrition, etc.), 3) writing procedures (number of sessions, duration of sessions, frequency of writing, writing instructions, etc.), and 4) study results (significant effects, sustainability, etc.). Authors of the selected studies were contacted in case of missing or unclear data. Whenever possible, Cohen’s d was computed. Effects sizes—the between-group standardized mean difference (SMD) with 95% confidence interval (CI)—were pooled in RevMan version 5.3.
Outcome of literature search
Databases search generated a total of 1,122 publications: CINAHL (N= 20), PubMed (N= 654), PsychINFO (N= 64), and Web of Science (N= 384). One study was identified by hand search. Of 1,123 obtained publications, 1,098 studies were excluded after initial screening of titles and abstracts—95 duplicate articles, 4 missing articles, and 999 irrelevant articles. The full texts of 25 articles were checked, and 16 studies were further excluded: 1) (N= 2) not involving drug using samples, 2) (N= 7) not EW intervention, 3) (N= 3) repeated reports from the same samples, 4) (N= 3) uncontrolled trials, and 5) (N= 1) not reporting the addressed outcomes. In final, this study comprised nine trials. Only four studies were eligible for meta-analysis (see Figure 1).
Table 1: Sociodemographic and clinical characteristics of participants of the included expressive writing trials (N= 9)
Characteristics of included studies
Table 1 shows the characteristics of the nine included studies. Samples varied between 23 and 429 participants, with a total of 1,188 participants. In seven studies, average age ranged from 18.2 to 48 years. Four studies recruited university students while education in the other 5 studies varied from some college in 72.7% of subjects  to elementary education in 32.4% of participants . Females constituted a majority in most studies. Of seven studies reporting ethnicity, in two studies 70% of the participants were African American [3, 15] while participants of five studies were multi-ethnic—Caucasian was the most noticeable ethnicity [2, 14, 22, 27, 28]. Four studies examined alcohol drinking college students [27-30], two studies examined multiple substance users in residential settings [14, 15] while three studies investigated hospitalized cocaine users , chronic pain conditions using opioids , and HIV positive gays using methamphetamine . Co-morbid psychopathologies (depression and post-traumatic stress disorder) were indicated in three studies [3, 14, 15]. As shown in Table 2, EW was delivered individually (in a lab setting, online, or at home) [2, 27-30], or in groups in a clinical setting [3, 14, 15], either standalone [3, 15, 28-30] or in combination with other techniques: explicit cognitive reappraisal , psychoeducation and positive writing , behavioral monitoring , and support group . The number of writing sessions ranged from one [28, 30] to 10 sessions , and it took place on consecutive days [2, 15, 29] and on separate days [3, 22] or separate weeks [14, 27]. Session length varied from 15 minutes  to whatever participants wanted [28, 30]. Writing instructions focused on alcohol related difficulties [27, 29Graduate School of Binghamton University], HIV related stress , pain , or any life trauma [3, 14, 15]. Three studies used two experimental groups—writing details of a negative/positive drinking experiences [28, 30] and standalone EW and EW combined with behavioral monitoring . Control groups were placebo writing in six studies e.g., time management, non-writing treatment as usual (TAU) group , behavioral monitoring group , or a mindfulness group . Three studies did not incorporate follow up [27, 28, 30], and the follow-up intervals ranged from 2 weeks  to 3 months post writing [3, 14, 22].
Table 2 shows that psychological health was assessed in five studies by eight indices [2, 3, 14, 15, 22]. Physical health was addressed in only two studies through Pennebaker Inventory of Limbic Languidness (PILL)  and the Brief Pain Inventory (BPI) . Recovery from drugs dependence was indicated through 10 indices (e.g., frequency of use, drug free days, amount used, craving intensity and frequency, diagnosis status, etc.) in eight studies [2, 3, 14, 22, 27, 29]. A range of secondary outcomes was covered: indirect measures of recovery from drug use such as drinking intention [28-30], linguistic analysis of written content [22, 29], immediate physical and emotional response to EW [3, 15, 28, 29], satisfaction and feedback on writing [15, 22], and miscellaneous measures e.g., follow up visits [2, 3, 22].
Quality assessment of identified studies
Not a single study met all the addressed criteria for high quality. Most studies were rated unclear risk for selection, performance and detection biases because of failure to offer information about the method of randomization (N= 4), allocation concealment (N= 7); blinding of participants (N= 6), therapists (N= 9), and investigators (N= 4). Attrition was high in 6 studies [2, 3, 14, 15, 27, 29]. Selective reporting was considered high risk in only two studies [29, 30]. Other sources of bias were noted e.g., power calculation of sample size was lacking in all the studies (see Figures 2 and 3, which demonstrate a summary of assessment of methodological quality and risk of bias).
Therapeutic effects of EW
As shown in Table 3 meta-analysis of four studies indicated no effect of EW; the range of ESs was small (0.02 to 0.35). Furthermore, most ESs (SMD) were below 0.20 e.g., symptoms of depression (N= 218: SMD= 0.16, 95% CI -0.14, 0.46, p=.23, I 2 =16%), trauma (N= 205: SMD= .02, 95% CI -.61, .64, p=.96, I 2 =68%), drug use (N= 125: SMD= 0.14, 95% CI -0.54, 0.83, p=.68, I 2 =68%), or physical health both at posttreatment and at follow up. Moreover, the high I2 % indicates that heterogeneity—which represents differences in ESs between groups—is highly attributed to systematic rather than random error. For details on individual studies’ weight and differences, check forest plots (Appendix). Reports from single studies go in line with aggregated ones— all values of Cohen’s d were small at posttreatment, and all reports were null at follow up (Table 2).
Moderators of the effect of expressive writing
Significant moderators of EW were indicated in two studies; findings implied that patients with high distress and life change at baseline had lower craving intensity  at posttreatment. Problem drinkers had lower intended heavy drinking occasions  i.e. the most severe cases might get the greatest benefit from EW.
Table 2: Characteristics of the included trials of expressive writing among alcohol and illicit drug users (N= 9)
Abbreviations:RCT: randomized control trial, EG: experimental group, CG: control group, P: primary outcome measures, S: secondary outcome measures, NR: not reported, min: minutes, hrs: hours, TLFB: Timeline Follow Back, RTCQ: Readiness to Change Questionnaire, LIWC: Linguistic Inquiry and Word Count Software, IES-R: HIV-related traumatic stress, IMC VAS: 0 to 100 Visual analogue scale on Intensity of methamphetamine craving, DDQ: Daily Drinking Questionnaire, PANAS: Positive and Negative Affect Scale, URICA: University of Rhode Island Change Assessment-Modified version, SOC: Stages of Change, MORE: Mindfulness-Oriented Recovery Enhancement, COMM: Current Opioid Misuse Measure, C–SOSI: Calgary Symptoms of Stress Inventory, BPI: The Brief Pain Inventory, FFMQ: Five Facet Mindfulness Questionnaire, CSQ: Coping Strategies Questionnaire, CERQ: Cognitive Emotion Regulation Questionnaire , AOD: Alcohol and Other Drugs, POMS: Profile of Mood States, BSI: Brief Symptoms Index, PSS: Perceived Stress Scale, BSCS: Brief Substance Craving Scale, SUD: Substance Use Disorders, PDS: Posttraumatic Stress Diagnostic Scale, CES-D: Center for Epidemiological Studies-Depression Scale, PILL: Pennebaker Inventory of Limbic Languidness, EEM: Essay Evaluation Measure, TOSCA-3 revised: Test of Self-Conscious Affect, J-PTSD: Jellinek-Posttraumatic Diagnostic Scale in SUD patients, SCID-I: Structured Clinical Interview for DSM-IV Axis I Disorders, AUDIT: Alcohol Use Disorders Identification Test.
Table 3: Pooled effects of expressive writing on psychological health, physical health, and recovery of drugs (N= 4 studies)
Immediate response to writing and participants’ feedback
Immediate response to writing refers to the emotional and physical reactions that people experience immediately after the end of the writing session. They occur as a result of re-exposure to the expressed unpleasant experiences. Four studies reported significant immediate increase in negative affect [15, 29], shame and guilt , paranoid ideation and blood pressure as well as lowered vigor  following initial or single sessions of EW. Participants significantly described EW as helpful; they continued journaling after sessions ended  and recommended EW to a friend .
In this review, nine trials were examined to determine the therapeutic potentials of EW among people who use drugs. Findings cannot support the effectiveness of EW. Pooled data indicated no evidence of immediate or remote effect of EW on depression, trauma or use of drugs whereas individual data on effects of EW were scattered, mostly null, or were short lived and faded away at follow up [3, 14, 15, 22]. The findings of this review are in accordance with results of former meta-analyses, which indicated no or minor and short-term effects of EW on physical and psychological health [13, 23-25, 31, 32]. Generalizability of conclusions is limited since samples were not representative of the drug using population: predominantly Caucasian, females, and alcohol drinking students. Because the meta-analysis comprised only 4 studies and samples were small, failure to capture effects of EW might be attributed to the low statistical power of the meta-analysis. In addition, pooled studies involved high co-morbidity with severe conditions such as chronic pain, PTSD, depression, and HIV [2, 14, 15, 22], which could have limited the effect of EW. Besides, reported findings should be taken up with regard to the validity threats ingrained in the reviewed studies. Further, studies had heterogeneity of writing instructions (general or drug related trauma), delivery (alone, in groups, computerized), number of sessions and interval between them, treatment setting, outcome measures, follow up, and participants (alcohol drinking students versus severe clinical patients). The effects of heterogeneity cannot be ignored as it could greatly alter findings .
Few moderators were explored, and diaries were underutilized. All trials failed to investigate roles of gender, expressiveness, word use, and writing on the same topic or different topics as moderators, despite the documented association between these variables with the occurrence of benefits of EW . One study (a single EW session) attempted to explain how EW works; it suggested that guilt that followed writing increased readiness to change (RTC), which in turn affected drinking intentions . Conversely, another study with a longer course of EW also reported increased RTC after EW, but RTC was not associated with changes of cognitive words use or drinking related measures —which renders the role of guilt in benefits of EW mute. The current study has some strength: PRISMA guidelines were followed, a comprehensive search was conducted, and ESs were computed. Still, limitations exist; the paucity of sound RCTs forced the author to include flawed unpublished PhD theses. In some studies, it was difficult to compute ESs because of lack of necessary data, despite attempts to contact authors. Similarly, few studies were eligible for meta-analysis, and the analysis was limited: it was not possible to perform subgroup analysis to explore the effect of doses of EW, co-morbidity, gender, abuse severity, etc. More, the high heterogeneity of study designs and outcome measures lowers the credibility of reports.
Findings of this review do not uphold the helpfulness of EW for psychological and physical wellbeing or for recovery of drug use among people using illicit drugs. However, results should be considered with caution because EW studies among drug users are few, methodologically limited, and highly heterogeneous. More rigorous studies involving enhanced EW procedures are needed to identify specific subgroups that can benefit from EW.
The author thanks Yumi Asai and Amin Hendawy for their help with data extraction and quality assessment as well as Dr. Toshi Aki and Prof. Norito Kawakami for their support.
Conflict of interest
The author has no conflict of interest to disclose. This review did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Appendix: Forest plots.
1. Wu L-T. Substance abuse and rehabilitation: responding to the global burden of diseases attributable to substance abuse. Subst Abuse and Rehabil 2010; 1: 5–11.
2. Garland EL, Manusov EG, Froeliger B, et al. Mindfulness-oriented recovery enhancement for chronic pain and prescription opioid misuse: Results from an early-stage randomized controlled trial. J Consult Clin Psychol 2014; 82(3): 448–459.
3. Grasing K, Mathur D, Desouza C. Written emotional expression during recovery from cocaine dependence: group and individual differences in craving intensity. Subst Use Misuse 2010; 45(7-8): 1201–1215.
4. Ali AM. Stigma and Psychological Distress among Egyptian Patients with Substance Use Disorders. Ann Psychiatry Ment Health 2018; 6(3): 1136.
5. 5Kay-Lambkin FJ, Baker AL, Kelly BJ, et al. It’s worth a try: The treatment experiences of rural and urban participants in a randomized controlled trial of computerized psychological treatment for comorbid depression and alcohol/other drug use. Journal of Dual Diagnosis 2012; 8(4): 262–276.
6. Roberts NP, Roberts PA, Jones N, et al. Psychological interventions for post-traumatic stress disorder and Comorbid substance use disorder: A systematic review and meta-analysis. Clin Psychol Rev 2015; 38: 25–38.
7. Ali AM, Ahmed A, Sharaf A, et al. The Arabic Version of The Depression Anxiety Stress Scale-21: Cumulative scaling and discriminant-validation testing. Asian J Psychiatr 2017; 30: 56-58.
8. Naji L, Dennis BB, Bawor M, et al. A prospective study to investigate predictors of relapse among patients with opioid use disorder treated with methadone. Subst Abuse 2016; 10: 9–18.
9. Mowbray O, Scott JA. The effect of drug use disorder onset, remission or persistence on an individual’s personal social network. Am J Addict 2015; 24(5): 427–434.
10. Ali AM. The relationship between emotional intelligence, distress disclosure and psychological distress among egyptian illicit substance users. J Addict Recovery 2018; 1(1002): 1-5.
11. Mohebi MD, Ayubi E, Azmoodeh A, et al. The relationship between identity styles and addiction vulnerability: A cross-sectional study among medical students in Zahedan, South Eastern Iran. Psychiatry Res 2018; 268: 184-188.
12. Stokes C, Hirsch CR. Engaging in imagery versus verbal processing of worry: Impact on negative intrusions in high worriers. Behav Res Ther 2010; 48(5): 418–423.
13. Mogk C, Otte S, Reinhold-Hurley B, et al. Health effects of expressive writing on stressful or traumatic experiences - a meta-analysis. Psychosoc Med 2006; 3.
14. van Dam D, Ehring T, Vedel E, et al. Trauma-focused treatment for posttraumatic stress disorder combined with CBT for severe substance use disorder: a randomized controlled trial. BMC Psychiatry 2013; 13: 172.
15. Meshberg-Cohen S, Svikis D, McMahon TJ. Expressive writing as a therapeutic process for drug dependent women. Subst Abus 2014; 35(1): 80–88.
16. Weegmann M. Just a story? Narrative approaches to addiction and recovery. Drugs Alcohol Today 2010; 10(3): 29-38.
17. Stephenson GM, Laszlo J, Ehmann B, et al. Diaries of significant events: socio-linguistic correlates of therapeutic outcomes in patients with addiction problems. Journal of Community & Applied Social Psychology 1997; 7(5): 389-411.
18. Baikie K, Wilhelm K. Emotional and physical health benefits of expressive writing. Adv Psychiatr Treat 2005; 11(5): 338-346.
19. Merz EL, Fox RS, Malcarne VL. Expressive writing interventions in cancer patients: a systematic review. Health Psychol Rev 2014; 8(3): 339-361.
20. Konig A, Eonta A, Dyal SR, et al. Enhancing the benefits of written emotional disclosure through response training. Behav Ther 2014; 45(3): 344–357.
21. Nazarian D, Smyth JM. An experimental test of instructional manipulations in expressive writing interventions: examining processes of change. Journal of Social and Clinical Psychology 2013; 32(1): 71-96.
22. Carrico AW, Nation A, Gómez W, et al. Pilot trial of an expressive writing intervention with HIV-positive methamphetamine-using men who have sex with men. Psychol Addict Behav 2015; 29(2): 277–282.
23. Zachariae R, O’Toole MS. The effect of expressive writing intervention on psychological and physical health outcomes in cancer patients--a systematic review and meta-analysis. Psychooncology 2015; 24(11): 1349- 1359.
24. Frisina PG, Borod JC, Lepore SJ. A meta-analysis of the effects of written emotional disclosure on the health outcomes of clinical populations. J Nerv Ment Dis 2004; 192(9): 629-634.
25. Zhou C, Wu Y, An S, et al. Effect of expressive writing intervention on health outcomes in breast cancer patients: A systematic review and meta-analysis of randomized controlled trials. PLoS One 2015; 10(7): e0131802.
26. Evans B, Pakstis D, Snead B, et al. The Use of Creative Writing Interventions in Substance Abuse Treatment. Therapeutic Recreation Journal 2015; XLIX(3): 179- 182.
27. Ackermann ME. The importance of gender and readiness to change in the prediction of drinking and negative consequences of first-year student drinkers. Virginia: Doctorate Dissertation. Old Dominion University; 2007.
28. Rodriguez LM, Young CM, Neighbors C, et al. Evaluating guilt and shame in an expressive writing alcohol intervention. Alcohol 2015; 49(5): 491–498.
29. Gallo LA. Investigating written emotional disclosure as an intervention for college student drinking. New York: Doctorate Thesis. State University of New York, Graduate School of Binghamton University; 2007.
30. Young CM, Rodriguez LM, Neighbors C. Expressive writing as a brief intervention for reducing drinking intentions. Addict Behav 2013; 38(12): 2913-2917.
31. Nyssen OP, Taylor SJ, Wong G, et al. Does therapeutic writing help people with long-term conditions? Systematic review, realist synthesis and economic considerations. Health Technol Assess 2016; 20(27).
32. Travagin G, Margola D, Revenson TA. How effective are expressive writing interventions for adolescents? A meta-analytic review. Clin Psychol Rev 2015; 36: 42–55.
33. Knaevelsrud C, Böttche M. Writing therapy after traumatic events: Therapeutic approaches and mechanisms of change. Psychother Psychosom Med Psychol 2013; 63(09-10): 391-397.
34. Hevey D, Wilczkiewicz E. Changes in language use mediate expressive writing’s benefitson health-related quality of life following myocardial infarction. Health Psychol Behav Med 2014; 2(1): 1053–1066.