Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Aug 1.
Published in final edited form as: Rehabil Psychol. 2023 Jun 8;68(3):261–270. doi: 10.1037/rep0000507

Effects of Hypnosis, Mindfulness Meditation, and Education for Chronic Pain on Substance Use in Veterans: A Supplementary Analysis of a Randomized Clinical Trial

Aaron P Turner 1,2, Karlyn A Edwards 2,3, Mark P Jensen 1, Dawn M Ehde 1, Melissa A Day 1,4, Rhonda M Williams 1,2
PMCID: PMC10524362  NIHMSID: NIHMS1902112  PMID: 37289535

Abstract

Purpose/Objective:

To examine the impact of three behavioral interventions for chronic pain on substance use.

Research method/Design:

Participants were 328 Veterans with chronic pain receiving care at one of two Veterans Affairs Medical Centers in the northwest United States. Participants were randomly assigned to one of three 8-week manualized in-person group treatments: (1) Hypnosis (HYP), (2) Mindfulness Meditation (MM), or (3) Education Control (ED). Substance use frequency was assessed using ten individual items from the WHO-ASSIST, administered at baseline prior to randomization and at 3- and 6-months posttreatment.

Results:

Baseline substance use (i.e., any use) in the past 3 months was reported by 22% (tobacco), 27% (cannabis) and 61% (alcohol) of participants. Use of all other substances assessed was reported by <7% of participants. Results showed that MM, as compared to ED, significantly reduced risk of daily cannabis use by 85% and 81% at the 3- and 6-month posttreatment follow-ups, respectively, after adjusting for baseline use. HYP, as compared to ED, significantly reduced risk of daily cannabis use by 82% at the 6-month posttreatment follow-up after adjusting for baseline use. There was no intervention effect on tobacco or alcohol use at either posttreatment follow-up.

Conclusions/Implications:

HYP and MM for chronic pain may facilitate reductions in cannabis use, even when reducing such use is not a focus of treatment.

Keywords: Chronic pain, hypnosis, mindfulness meditation, substance use, alcohol, tobacco, cannabis

Introduction

Chronic pain is widely recognized as a significant public health problem, affecting up to 19% of US adults and resulting in up to $635 billion in health care costs and lost productivity (Kennedy, Roll, Schraudner, Murphy, & McPherson, 2014; Institute of Medicine, 2011). Prevalence rates and reported pain severity are often even higher in samples of Veterans of the US Armed Forces (Nahin, 2017). Adding to the complexity of chronic pain and its treatment in Veterans is the high rate of co-morbid mental health conditions, including substance use and use-related disorders (SUD).

Recent data suggests that Veterans report significantly higher rates of daily alcohol, cannabis, and tobacco use as compared to civilians (London, Wilmoth, Oliver, & Hausauer, 2020; Wagner et al., 2007). Daily substance use is associated with elevated risk of experiencing substance use-related problems, as well as worse physical and mental health (GBD 2016 Alcohol Collaborators, 2018; Verges, Ellingson, Schroder, Slutske, & Sher, 2018). Tobacco and alcohol have the highest substance attributable mortality rates across the world (Peacock et al., 2018). While cannabis is thought to have less risk of severe long-term physical health effects, it is associated with cognitive impairment, worse mental health, and risk of short-term health impacts such as nausea, vomiting, fatigue, and dizziness (Sherman & McRae-Clark, 2016; Whiting et al., 2015). Given the high rates of tobacco, alcohol, and cannabis use in Veterans, it is not surprising that the past year prevalence of SUD among Veterans has been found to be very high; specifically 17%, with a lifetime prevalence of SUD of 39%, both of which are estimated to be higher than rates in civilian populations (Boden & Hoggatt, 2018). Substance use in the context of persistent pain is prevalent, with frequent co-occurrence. For example, Veterans enrolled in a medical cannabis program endorse more than double the rate of past year medical cannabis use, and rates of medical cannabis use are even higher among Veterans with chronic pain (Davis et al., 2016)

Given that pain and substance use frequently co-occur, it is reasonable to hypothesize that pain may be an important factor in the development and exacerbation of problematic substance use, with both conditions potentially contributing to a reciprocal downward spiral of pain, substance use and poorer overall health. There is some evidence to support this idea, as Veterans with chronic pain who use cannabis to manage pain also endorse more frequent use of other substances, worse pain, more severe post-traumatic stress disorder (PTSD) symptoms, and worse sleep and physical health, as compared to those who do not use medical cannabis (Davis et al., 2016; Gros, Szafranski, Brady, & Back, 2015; Metrik, Bassett, Aston, Jackson, & Borsari, 2018). Further, qualitative interviews conducted with Veterans with chronic pain and SUD have highlighted that many endorse using substances to cope with both physical and emotional pain, and that pain is a common trigger for SUD relapse (Wyse et al., 2021). Taken together, these results suggest that pain and substance use may exacerbate each other, and also that treatments that improve one of these issues might also improve the other.

Consistent with this idea, early evidence suggests that, for individuals with co-occurring chronic pain and substance use, interventions specifically targeting both conditions may reduce use of some types of substances. For example, in individuals with chronic pain endorsing patterns of risky alcohol, cannabis or other drug use, brief counseling focused on the relationship between pain and substance use was shown to reduce both cannabis and alcohol use (Rosen et al., 2019). Similarly, in individuals receiving SUD treatment who also have chronic pain, brief counseling and cognitive behavioral therapy targeting both pain and substance use was shown to improve engagement in treatment (Cummins & Tobian, 2018) and reduce alcohol use (Ilgen et al., 2016). However, little is known about whether psychosocial treatments for chronic pain that do not focus on or even specifically mention substance use may improve substance use outcomes.

There is increasing recognition of the value of hypnosis (HYP) and mindfulness meditation (MM) for the treatment of pain. Both HYP and MM teach self-management skills that, once acquired, can be used as needed and maintained with independent practice. Additionally, both HYP and MM have been shown to reduce pain, pain interference, and comorbid problems such as sleep disturbance and negative mood across a broad range of chronic pain conditions (Adachi, Fujino, Nakae, Mashimo, & Sasaki, 2014; Cherkin et al., 2016; Elkins, Jensen, & Patterson, 2007; Hilton et al., 2017; Jensen, 2009; Thompson et al., 2019). However, the impact of these pain-focused interventions on substance use remains unknown.

The current study examines an exploratory outcome measure – substance use – that was assessed in a randomized controlled trial (RCT) examining the comparative effectiveness of HYP, MM, and an education control intervention (ED) for the treatment of chronic pain in Veterans (Williams et al., 2022). The parent trial found that all three conditions were associated with significant reductions in pain intensity from pre-to post-treatment, and HYP and MM resulted in greater reductions in pain intensity than those in the ED condition at 3- and 6-months post-treatment. Further, those in the HYP and MM conditions demonstrated greater improvement in pain interference, depression, and worst pain intensity in the 6 months following treatment, compared to those in ED (Williams et al., 2022). In addition to the study’s primary and secondary outcomes, substance use was assessed at baseline and the post-treatment follow-ups as an exploratory outcome. The aim of the analyses presented here was to characterize substance use in a large sample of Veterans participating in treatment for chronic pain and, for substances with adequate prevalence, explore the potential impact of the HYP and MM treatment interventions on the use of those substances as compared to a rigorous ED treatment. The primary hypothesis was that Veterans participating in MM and HYP would experience greater reductions in substance use frequency at 3- and 6- months post treatment relative to ED.

Methods

Transparency and Openness Promotion

This study is a supplementary analysis of substance use outcomes in a two-site, three-arm, parallel group, randomized controlled trial. The primary outcome paper (Williams et al., 2022) and protocol paper (Williams et al., 2020), report sample size calculations, all data exclusions, all manipulations, and all measures that were included in the study; the paper adheres to CONSORT guideline for reporting randomized trials. Data for this study have not been made available to the public. Data were analyzed using SPSS version 26 (IBM Corp., 2019). The supplementary analyses reported here were not proposed a priori and were formulated after primary trial outcomes were known to the investigators. The parent trial was pre-registered on clinicaltrials.gov (NCT02653664).

Design

Following screening to determine eligibility, completion of informed consent for participation and baseline assessment, individuals were randomized to receive 8 weekly in-person group-based sessions of HYP, MM, or ED. Participants were individually randomized to the three interventions 1:1:1 in stratified blocks based on sex and pain type (neuropathic, non-neuropathic, mixed or undetermined). Study outcomes were assessed prior to treatment initiation (baseline), at weeks 2, 4, 6 (mid-treatment) and 8 (post treatment, primary trial endpoint), and 3- and 6-months follow-up. Substance use frequency – the focus of the current analyses – was measured only at baseline and 3- and 6-month follow-up. Participants were informed that all three treatments were forms of “pain self-management” and were unaware of the study hypotheses. Randomization was conducted by a central coordinating center external to the study. Research staff conducting recruitment did not have access to the randomization schedule to maintain concealment of allocation. All staff conducting outcome assessments were unaware of treatment assignment throughout the study. Additional details are reported in the protocol paper and primary outcome paper for this trial (Williams et al., 2022; Williams et al., 2020). The trial was approved by the Human Participants Review boards at VA Puget Sound and the University of Washington as well as the study sponsor.

Participants

Participants were 328 Veterans with chronic pain who received care at one of two study sites (Seattle VA Medical Center or American Lake VA Medical Center) between 2015 and 2019. Participants were recruited via clinician referrals, medical record review/mail approach, and self-referral. Inclusion criteria included: (1) Veteran status; (2) 18 years of age or older; and (3) self-reported presence of moderate or greater severity chronic pain, operationalized as average self-reported pain intensity rating of ≥ 3 on a 0–10 Numerical Rating Scale (NRS) in the last week, worst pain intensity of ≥ 5 on a 0–10 NRS in the last week, duration of chronic pain 3 months or more, and experience of pain at least 75% of the time in the last 3 months. Exclusion criteria included: (1) severe cognitive impairment defined as two or more errors on the Six-Item Screener (Callahan, Unverzagt, Hui, Perkins, & Hendrie, 2002); (2) psychiatric or behavioral conditions that would interfere with safe or effective group participation; (3) reported average daily use of >120 mg morphine equivalent dose (MED). For additional detail on recruitment, screening, and enrollment see the protocol paper (Williams et al., 2020) and the CONSORT diagram (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Consort Diagram of Participants in the Original Trial

Note. For additional details on ineligibility and withdrawals please see (Williams et al., 2022).

Interventions

All three interventions (HYP, MM, ED) were offered simultaneously three times per year at each study site for a total of 72 groups. Each intervention was delivered via eight 90-minute group-based sessions scheduled over 8–10 weeks. Each intervention was based on a manualized protocol that included a therapist guide and participant workbook, audio-recordings, and home practice.

In the HYP condition, group participants were invited to relax and listen to the clinician deliver a standardized script that included a hypnotic induction, a post hypnotic suggestion, and a post hypnosis alerting process. Inductions and suggestions varied by session but focused on promoting increased comfort, adaptive thoughts about pain, or improvement in pain-related experiences such as mood, relaxation, sleep quality, and optimism. Sessions concluded with a group discussion focused on enhancing home practice with and without recordings.

In the MM condition, group participants were guided through mindfulness meditation exercises based upon principles of Shamatha and Vipassana. Each session began with a mindfulness practice. The content for each session varied, however typically included bringing a present-focused awareness to the body, breath, sounds, or thoughts. Following each practice, the clinician facilitated a guided inquiry of participant experiences with meditation (both in-session as well as home practice), difficulties with practice, and encouragement of gentle persistence in home practice with or without recordings.

In the ED condition, group participants received education on topics related to chronic pain, such as the biopsychosocial model, strategies to manage mood and sleep, and enhance social support and communication skills. Sessions were interactive with most material shared via facilitated discussions. The ED condition was designed to be a credible intervention comparable to the HYP and MM interventions in terms of time, attention, delivery modality, therapeutic alliance and group cohesion and social support. Home practice was also encouraged as part of the ED condition, and included audio recordings of session content and affirming/supportive messages in addition to the participant workbook.

Intervention Training

The interventions were conducted by health professionals employed at the study sites (N = 50) and included the following occupations: nursing, occupational therapy, physical therapy, speech pathology, physiatry, social work, chiropractic, psychology, and psychology trainees (predoctoral interns and postdoctoral fellows). All clinicians completed a 2-day in-person training conducted by subject matter experts in each of the 3 interventions that included in-vivo practice and additional self-study. All clinicians participated in twice-monthly supervision while delivering treatment. Clinicians were required to participate in at least 3 cohorts and deliver all 3 interventions in a counterbalanced order.

Intervention Fidelity

Intervention sessions were audio recorded, and 25% of sessions were reviewed by research staff using a standardized fidelity rating form. The average adherence ratings of essential prescribed elements covered in each session across intervention conditions was 97%. No proscribed elements were detected in fidelity review. There were no changes to the interventions during the study.

Measures

Demographics.

Demographic information including age, gender, race, education, employment status, living situation and pain intensity were collected for all participants by self-report at baseline.

Substance Use.

Substance use was measured using the individual substance frequency items from the World Health Organization Alcohol, Smoking and Substance Involvement Screening Test (WHO-ASSIST). The WHO-ASSIST is designed to screen for substance use over the past 3 months across 10 substance categories (cocaine, amphetamines, inhalants, sedatives, hallucinogens, opiates, tobacco, alcohol, cannabis, and other drugs) with 5 response options including “Never,” “Once or Twice,” “Monthly,” “Weekly” and “Daily or almost daily.” As described in the Results section, this study focused on only tobacco, alcohol, and cannabis use, because use for other substance categories was too infrequent (<7% endorsing any use) to allow for assessment over time. The WHO-ASSIST items have well-established reliability and validity (Who Assist Working Group, 2002; Humeniuk et al., 2008).

Data Analysis

Means and standard deviations for continuous study variables and counts and percentages for categorical study variables were computed at baseline to describe the sample. Effectiveness of randomization was established in the primary outcomes paper (Williams et al., 2022). The base rates of substance use across all substance categories were computed at baseline. The frequencies and distributions of the substance use variables were examined to determine whether parametric or non-parametric analyses would be conducted to test the study hypothesis regarding the effects of the active treatments. Chi-square analyses were used to examine proportional differences in baseline substance use frequency by intervention assignment. Finally, to ensure that baseline substance use did not contribute to study bias, attrition analyses were conducted which examined whether individuals with baseline substance use were more likely to drop out and thus not contribute data to follow-up.

For substances that approximated a binary distribution (i.e., tobacco, see Results), binary logistic regression analyses were conducted for each outcome period (3- and 6-month) studied. Binary logistic regression uses a logit transformation to examine how a group of predictor variables impact the likelihood of a binary outcome variable occurring. Coefficients are presented as odds ratios (ORs). ORs greater than 1.0 indicate a predictor increases the odds of the outcome occurring, and ORs less than 1.0 indicate a predictor deceases the odds of the outcome occurring (Nick & Campbell, 2007). The primary study hypothesis was that individuals in MM and HYP would have lower odds of use compared to ED at 3 months and 6 months following treatment, adjusting for gender, age and baseline use.

For substances that had adequate cell sizes across frequency categories (cannabis and alcohol, see Results), multinomial logistic regression analyses were conducted for each outcome period (3- and 6-month) studied. Multinomial logistic regression uses a similar approach to a binary logistic regression, however, this approach models a nominal outcome variable. Coefficients are presented as relative risk ratios (RRRs). RRRs are interpreted similarly to ORs, such that RRRs greater than 1.0 indicate a predictor increases the risk of an outcome occurring, and RRRs less than 1.0 indicate a predictor deceases the risk of an outcome occurring (Kwak & Clayton-Matthews, 2002). Again, the primary study hypothesis was that individuals in MM and HYP would have lower odds of use compared to ED at 3- and 6-months following treatment, adjusting for gender, age, and baseline substance use.

Results

Baseline Sample and Adequacy of Randomization

At baseline, all 328 randomized participants provided substance use data. Rates of participant retention were comparable across treatment arms. At 3-months follow-up, 259 of the 267 enrolled (97%) completed substance use assessments and at 6-months, 252 of 261 (97%) who remained enrolled completed substance use assessments. Demographic, pain, and substance use characteristics of the study sample by treatment arm are presented in Table 1. The majority of the sample identified as men (73%) and White (63%). Forty-five percent had a college degree and 83% were unemployed or retired. All analyses reported in the current study utilized all available data for each assessment point. Attrition analyses indicated that there were no significant proportional differences in baseline substance use frequency between those who had complete and missing data at the 3 month [all X2’s (3, 328) < 3.16, p’s >.37] and 6-month follow-up [all X2’s (3, 328) < 1.87, p’s >.60].

Table 1.

Demographic and clinical characteristics at baseline by treatment group

Demographic Characteristic Intervention Group P-value
Education (n=110) Hypnosis (n=110) Mindfulness (n=108)
Age in years, mean (SD) 53.5 (13.5) 51.0 (12.6) 55.0 (13.0) ns
Gender, % in category (n)
 Men 73 (80) 74 (81) 74 (80)
 Women 24 (27) 26 (29) 26 (28) ns
 Transgender 3 (3) 0 (0) 0 (0)
Race, % in category (n)
 Caucasian 62 (68) 65 (71) 63 (68)
 Black/African American 17 (19) 21 (23) 15 (16) ns
 Asian 3 (3) 2 (2) 4 (4)
 Otherc 17 (19) 12 (13) 18 (20)
Education, % in category (n)
 High school or less 10 (11) 8 (9) 6 (6)
 Some college/Technical 44 (48) 49 (54) 48 (52) ns
 College degree or higher 46 (51) 43 (47) 46 (50)
Employment status
 Unemployed, % yes (n) 49 (54) 44 (48) 33 (36) ns
 Retired, % yes (n) 41 (45) 38 (42) 44 (47) ns
 Employed full/part time, % yes (n) 24 (26) 29 (32) 28 (30) ns
 Student full/part time, % yes (n) 4 (5) 4 (4) 8 (9) ns
 Home maker, % yes (n) 6 (6) 6 (7) 4 (4) ns
Married/Living w/partner, % yes (n)f 61 (67) 55 (60) 52 (55) ns
Average Pain Intensity, mean (SD) 5.8 (1.6) 5.7 (1.8) 5.9 (1.6) ns
Open in a new tab

P-value derived from One-way ANOVA for continuous variables and Fisher exact test for categorial variables. No between-group comparisons produced p<.05. Race and marital status variables contained small amounts of missing data (N<4) each. For employment status individuals could endorse more than one category. Ns = nonsignificant.

Substance Use at Baseline

At baseline, any use of seven types of substances was endorsed by 7% or less of the sample [cocaine (<1%), amphetamines (1%), sedatives (7%), hallucinogens (<1%), opiates (1%), and other drugs (<1%)], therefore, these substances were excluded from the outcome analyses. Three categories, however, (tobacco, alcohol, and cannabis) demonstrated enough baseline use (22% to 61% of the sample endorsed use) to allow for an analysis of change in use over time and were therefore included in the subsequent analyses.

See Table 2 for a description of tobacco, alcohol, and cannabis use at each timepoint. As shown, at baseline, 22% of the sample endorsed tobacco use, 27% endorsed cannabis use, and 61% of the sample endorsed alcohol use in the past 3 months. The response categories “once or twice” and “monthly” use in the past 3 months were collapsed because they represented only a one-time difference in use in the past 3-months, leaving four response categories to describe tobacco, cannabis, and alcohol use. Tobacco use approximated a binary distribution. It was therefore collapsed into a binary variable representing any use and no use. Alcohol and cannabis use were deemed to have appropriate cell sizes in all four response categories and were therefore examined as nominal variables.

Table 2.

Frequency of tobacco, alcohol, and cannabis use at baseline, 3 months, and 6 months across the entire sample.

Frequency of use in the past 3 months
Timepoint Substance Never N (%) Once, Twice, or Monthly N (%) Weekly N (%) Daily/Almost daily N (%) Any Use N (%)
Baseline (N = 328) Tobacco 256 (78%) 17 (5%) 6 (2%) 50 (15%) 73 (22%)
Alcohol 127 (39%) 128 (39%) 51 (15%) 23 (7%) 202 (61%)
Cannabis 241 (73%) 39 (12%) 22 (7%) 27 (8%) 88 (27%)
3 months (N = 259) Tobacco 204 (79%) 10 (4%) 8 (3%) 37 (14%) 55 (21%)
Alcohol 102 (40%) 100 (39%) 45 (17%) 12 (5%) 157 (60%)
Cannabis 193 (75%) 19 (7%) 15 (6%) 32 (12%) 66 (25%)
6 months (N = 252) Tobacco 200 (79%) 12 (5%) 7 (3%) 33 (13%) 52 (21%)
Alcohol 92 (37%) 101 (40%) 46 (18%) 13 (5%) 160 (63%)
Cannabis 189 (75%) 19 (8%) 16 (6%) 28 (11%) 63 (25%)
Open in a new tab

Note. Values measured using the WHO-ASSIST. Any use proportion was utilized for the analysis of tobacco. Categorical responses were utilized for Alcohol and Cannabis. WHO-ASSIST = World Health Organization Alcohol, Smoking and Substance Involvement Screening Test.

Chi-square analyses indicated that there were no proportional differences in tobacco use frequency (no use vs any use) by intervention at baseline [X2 (2, N = 328) = 2.44, p = .30]. There were also no proportional differences in cannabis use frequency by intervention at baseline [X2 (6, N = 328) = 7.62, p = .27]. Some differences in alcohol use frequency by intervention at baseline were observed [X2 (6, N = 328) = 12.61, p = .05], such that there was a marginally higher proportion of individuals who endorsed weekly alcohol use in the ED condition as compared to those in HYP and MM.

Effect of Interventions on Substance Use

Tobacco use.

Results from the binary logistic regression analyses examining change in tobacco use from baseline to the 3-month and baseline to the 6-month follow-up can be found in Table 3. As shown, baseline use was significantly associated with higher odds of endorsing tobacco use at both follow-up endpoints (3-month OR = 381.73, 95% CI = 89.07 – 1636.03; 6-month OR = 271.17, 95% CI = 76.07 – 966.64). However, there were no significant effects of age, gender, or intervention group on tobacco use at either follow up period.

Table 3.

Binary logistic regressions examining treatment condition and likelihood of tobacco use at 3 months and 6 months post treatment.

3 months 6 months
Predictor OR 95% CI OR 95% CI
Age 1.04 .98 – 1.09 1.00 95 – 1.05
Gender 1.97 .54 – 7.14 .50 10 – 2.56
Baseline use 381.73*** 89.07 – 1636.03 271.17*** 76.07 – 966.64
Hypnosisa .52 .11 – 2.51 .52 11 – 2.51
Mindfulnessa 1.12 .26 – 4.86 1.12 26 – 4.86
Open in a new tab
***

p < .001. OR = Odds Ratio. CI = 95% Confidence Interval.

a

= Reference group is the pain education intervention (ED).

Alcohol Use.

Results from the multinomial logistic regression analyses examining alcohol use can be found in Table 4. Baseline use was significantly associated with a higher risk of endorsing more frequent alcohol use across all response categories at both the 3- and 6-month follow-up assessments. However, there were no significant effects of age, gender, or intervention group on alcohol use at either follow-up assessment, with the exception of age at the 3-month follow up period. Results indicated that those who were younger were at a slightly higher risk (3%) of using alcohol once, twice, or monthly as compared to no use in the past 3-months (RRR = .97, 95% CI = .94 - .99).

Table 4.

Multinomial logistic regressions examining treatment condition and alcohol use frequency at 3 months and 6 months post treatment.

Never vs. Once, Twice, or Monthly Never vs. Weekly Never vs. Daily/Almost Daily
Time-point Predictor RRR 95% CI RRR 95% CI RRR 95% CI
3 mo Age .97* .94 – .99 .99 .94 – 1.03 .92 .82 – 1.02
Gender .76 .35 – 1.67 1.00 .35 – 2.81 .13 .01 – 2.28
Baseline use 15.48*** 7.69 – 31.18 93.21*** 35.11 – 247.44 3311.29*** 271.27 – 40419.36
Hypnosisa .63 .25 – 1.55 .67 .20 – 2.25 .04 .002 – 1.04
Mindfulnessa .71 .28 – 1.80 .50 .13 – 1.88 .31 .02 – 4.70
6 mo Age 1.00 .97 – 1.03 1.00 .96 – 1.05 .96 .89 – 1.02
Gender .91 .42 – 1.97 1.60 .60 – 4.24 .48 .07 – 3.02
Baseline use 10.86*** 5.64 – 20.88 47.99*** 20.12 – 114.50 142.10*** 41.33 – 488.57
Hypnosisa 2.04 .86 – 4.87 1.59 .48 – 5.25 .53 .07 – 4.11
Mindfulnessa 1.43 .60 – 3.41 1.79 .55 – 5.86 1.59 .27 – 9.41
Open in a new tab
*

p < .05.

***

p < .001. RRR = Relative Risk Ratio. CI = 95% Confidence Interval.

a

= Reference group is the pain education intervention (ED).

Cannabis Use.

Results from the multinomial logistic regression analyses examining cannabis use can be found in Table 5. Notably, there was a significant effect of intervention, such that MM, as compared to ED, reduced risk of daily cannabis use by 85% at the 3-month follow-up (RRR = .15, 95% CI = .03 - .87), and by 81% at the 6-month follow-up (RRR = .19, 95% CI = .04 - .99) after adjusting for baseline use. Further, HYP, as compared to ED, reduced risk of daily cannabis use by 82% at the 6-month follow-up (RRR = .18, 95% CI = .04 - .92) after adjusting for baseline use. There was also a significant effect of age at the 6-month follow-up, which indicated that those who were younger were at higher risk (9%) for using cannabis once, twice, or monthly as compared to no use in the past 3-months (RRR = .91, 95% CI = .87 - .95). Baseline use was significantly associated with higher risk of endorsing more frequent cannabis use at the 3- and 6-month follow up across all response categories. There was no significant effect of gender on cannabis use at either follow up period.

Table 5.

Multinomial logistic regressions examining treatment condition and cannabis use frequency at 3 months and 6 months post treatment.

Never vs. Once, Twice, or Monthly Never vs. Weekly Never vs. Daily/Almost Daily
Time-point Predictor RRR 95% CI RRR 95% CI RRR 95% CI
3 mo Age .96 .92 – 1.01 .99 .94 – 1.05 .98 .93 – 1.03
Gender .95 .31 – 2.96 .67 .16 – 2.78 .62 .16 – 2.44
Baseline use 13.89*** 5.66 – 34.07 25.46*** 9.45 – 68.56 52.21*** 18.54 – 147.05
Hypnosisa 1.58 .39 – 6.30 .29 .06 – 1.50 .29 .06 – 1.38
Mindfulnessa 1.22 .28 – 5.39 .33 .06 – 1.67 .15* .03 – .87
6 mo Age .91*** .87 – .95 .97 .92 – 1.02 .95 .90 – 1.01
Gender .79 .26 – 2.41 .25 .03 – 2.01 1.25 .34 – 4.66
Baseline use 5.95*** 2.69 – 13.16 11.37*** 5.08 – 25.45 25.25*** 10.36 – 61.54
Hypnosisa .58 .16 – 2.08 .32 .07 – 1.48 .18* .04 – .92
Mindfulnessa .79 .22 – 2.83 .30 .06 – 1.43 .19* .04 – .99
Open in a new tab
*

p < .05.

***

p < .001. RRR = Relative Risk Ratio. CI = 95% Confidence Interval.

a

= Reference group is the pain education intervention (ED).

Discussion

The current study is a supplementary analysis of a clinical trial that examined HYP and MM vs ED for the treatment of chronic pain. Analyses for this study focused on changes in substance use, which was an exploratory outcome measure included in the parent trial. Results in the present study mirrored those in the main trial outcomes, where HYP and MM were found to be associated with greater improvements at 3- and 6-months post intervention than ED. Among Veterans with chronic pain, use of tobacco, alcohol, and cannabis was common at baseline (22%, 61%, and 27% reported any use, respectively, in the past 3-months). Use of other substances (cocaine, amphetamines, inhalants, sedatives, hallucinogens, opiates, other drugs) occurred infrequently (any use of these substances was reported by <7% of the sample). Not surprisingly, the three substances used most frequently were all legal and easily available without prescription in the state of Washington where the trial was conducted.

Participants who received HYP and MM treatment, as compared to ED, were at lower risk for daily cannabis use at most follow-up assessments, even after adjusting for baseline use. Specifically, MM reduced risk of daily cannabis use by 85% and 81% at the 3- and 6-month posttreatment follow-ups, respectively, and HYP reduced risk of daily cannabis use by 82% at the 6-month posttreatment follow-up. This finding is particularly noteworthy, as neither HYP nor MM specifically addressed substance use in any of the treatment sessions. It is also important to note that mindfulness-based interventions have been shown to be efficacious in treating substance use disorder (Bowen et al., 2014). Although MM for chronic pain and substance use share some components, such as the development of a self-regulatory practice to improve overall distress tolerance, the treatment targets and skill applications differ. MM for substance use targets craving, relapse, and impulsivity, and encourages use of skills to reduce reliance on substances (Brewer, Elwafi, & Davis, 2013). MM for pain targets a shift in the relationship to pain, pain-related distress, and modulation of pain, and encourages the use of skills to improve function despite the presence of pain (Zeidan & Vago, 2016).

A number of potential mechanisms might explain the results and could serve as the basis for future studies of the mediators underlying changes in cannabis use. For example, both of the active interventions encouraged the development of a purposeful and intentional self-regulatory practice. It is possible that this practice and its effects generalized to other areas of adaptive self-care, such as reducing cannabis use (Nagaya, Yoshida, Takahashi, & Kawai, 2007; Unger, 1996).

Another possibility is that the reductions in pain intensity associated with HYP and MM resulted in a reduced perceived need for cannabis use as a means of controlling pain. It is also possible that HYP and MM effectively impacted other important pain mechanisms that may not have been affected by the ED condition (Day, Jensen, Ehde, & Thorn, 2014; Jensen & Patterson, 2014), such as pain catastrophizing or pain acceptance, which may themselves influence changes in cannabis use. Future research on HYP and MM could examine these and other possible mechanisms in relation to changes in substance use.

Another potential explanation for the improvements observed with HYP and MM is that these interventions may produce improvements in individuals’ ability to self-regulate their emotional responses in a way that might be perceived as similar to the effects of cannabis on emotional responses. Thus, use of these self-regularly practices might result in reductions in symptoms of depression, anxiety or PTSD, without the aid of substances. Evidence supporting the beneficial effects of HYP and MM alone or in combination with other treatment modalities on these psychological function domains continues to grow (Banks, Newman, & Saleem, 2015; Goldberg et al., 2019; Hammond, 2010; Hofmann, Sawyer, Witt, & Oh, 2010; Rotaru & Rusu, 2016; Shih, Yang, & Koo, 2009; Valentine, Milling, Clark, & Moriarty, 2019). In support of improved emotion regulation as a potential mechanism underlying reduced cannabis use are findings from a recent systematic review and meta-analysis that examined the effects of cannabis use on experimental pain among healthy adults. The study found that acute administration of cannabis improved pain tolerance and reduced pain unpleasantness, although it did not have a reliable effect on pain intensity itself (De Vita, Moskal, Maisto, & Ansell, 2018). This suggests that cannabis may influence affective processes (i.e., pain unpleasantness), which may similarly be influenced by MM and HYP, thereby facilitating a reduction in the perceived need for cannabis. This might also explain why the ED condition did not evidence similar changes in cannabis use, although it was associated with similar reductions in pain intensity as HYP and MM at post-treatment.

The findings of this study do not provide specific insights into why participants in the MM and HYP conditions were more likely to reduce cannabis use than tobacco and alcohol use. One possibility is that physical dependence associated with cannabis use is generally thought to be lower than that of alcohol or tobacco (Anthony, Warner, & Kessler, 1994), and thus once other drivers of use (e.g., negative reinforcement associated with pain relief or reduced emotional distress) are addressed, reductions might be easier to make. This hypothesis would need to be tested in future studies.

Limitations

The study and the parent trial from which it is derived have a number of limitations that should be considered when interpreting the results. Perhaps the most important limitation is the study’s exploratory nature, given that we did not make any a priori predictions of the effects of the treatments on substance use. Therefore, the findings should be viewed as tentative and in need of replication before definitive conclusions regarding the effects of MM and HYP on substance use can be made. Second, approximately 21% to 23% of the study sample did not complete follow-up assessments. While attrition analyses indicated that missingness was not associated with substance use frequency at baseline, it remains possible that attrition related to other aspects of substance use may have occurred and impacted the findings. Third, participants with high levels of opioid medication use (greater than 120 daily MED) were excluded from the parent study. This not only affected the rates of opioid use reported, but may have also limited the ability to examine changes in opioid use over the course of the study.. Fourth, substance use was only characterized by frequency of use due to limitations (and the need to minimize assessment burden) in the parent study assessment battery; it is not known if the interventions were effective in reducing other SUD outcome domains, such as craving, substance quantity, or the negative consequences associated with use. We were also not able to control for some potential moderators, such as prior substance use treatment and cognitive functioning, which may have impacted our findings. It is therefore unclear how the present findings might generalize to Veterans with a confirmed SUD diagnosis. Fifth, given that our analyses examined the relative risk of an outcome occurring as compared to a reference group (in our case the ED intervention), we are not able to describe absolute changes in substance use frequency that occurred within any intervention condition, including ED. Sixth, the fact that substance use was not measured at the treatment midpoints or at immediate posttreatment made it difficult to look at potential mediation of effects. Seventh, this study was conducted in a state in which recreational cannabis use is currently legal, and circumstances of use may have been different than those experienced by individuals in states where use is not legal. Finally, this study was conducted in a sample of primarily white, male, treatment-seeking Veterans with chronic pain. The results might not generalize to more diverse, non-treatment seeking, or non-Veteran populations.

Despite the study’s limitations and exploratory nature, the findings suggest the intriguing possibility that MM and HYP treatment, even when provided for addressing pain and not specifically targeting substance use, might effectively reduce substance use (here specifically, cannabis). Additional research to examine this possibility is warranted. Such research should also test for a number of possible mechanisms of these effects, including via the effects of these treatments on pain intensity, emotion regulation, pain-related thoughts and cognitive processes such as self-efficacy and catastrophizing. If these findings are found to be reliable, this would offer individuals with chronic pain and problematic cannabis use additional options for managing this use.

Supplementary Material

Supplemental Material

Impact.

  • Many investigations have examined the impact of psychosocial treatments for chronic pain. Comparatively little information exists about the extent to which these treatments provide the collateral benefit of contributing to reductions in substance use.

  • In this study, reductions in cannabis use among individuals with chronic pain were observed across multiple treatment modalities (mindfulness and hypnosis) relative to a rigorous control.

  • Both mindfulness and hypnosis for chronic pain (and co-occurring substance use) can be delivered efficiently in a group-based format and may have benefits extending beyond the treatment of pain itself.

Acknowledgments

This work was supported by the National Institutes of Health, National Center for Complementary and Integrative Health (Grant # 1R01AT008336-01, awarded to Co-Principal Investigators Mark P. Jensen, PhD, and Rhonda M. Williams, PhD). ClinicalTrials.gov, Identifier NCT02653664

Footnotes

The authors identify no conflicts of interest.

References

  1. Anthony JC, Warner LA, & Kessler RC (1994). Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalants: Basic findings from the national comorbidity survey. Experimental and Clinical Psychopharmacology, 2(3), 244–268. [Google Scholar]
  2. Banks K, Newman E, & Saleem J. (2015). An Overview of the Research on Mindfulness-Based Interventions for Treating Symptoms of Posttraumatic Stress Disorder: A Systematic Review. Journal of Clinical Psychology, 71(10), 935–963. doi: 10.1002/jclp.22200 [DOI] [PubMed] [Google Scholar]
  3. Boden MT, & Hoggatt KJ (2018). Substance Use Disorders Among Veterans in a Nationally Representative Sample: Prevalence and Associated Functioning and Treatment Utilization. Journal of Studies on Alcohol and Drugs, 79(6), 853–861. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/30573015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bowen S, Witkiewitz K, Clifasefi SL, Grow J, Chawla N, Hsu SH, … Larimer ME. (2014). Relative efficacy of mindfulness-based relapse prevention, standard relapse prevention, and treatment as usual for substance use disorders: a randomized clinical trial. JAMA Psychiatry, 71(5), 547–556. doi: 10.1001/jamapsychiatry.2013.4546 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brewer JA, Elwafi HM, & Davis JH (2013). Craving to quit: psychological models and neurobiological mechanisms of mindfulness training as treatment for addictions. Psychology of Addictive Behaviors, 27(2), 366–379. doi: 10.1037/a0028490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Callahan CM, Unverzagt FW, Hui SL, Perkins AJ, & Hendrie HC (2002). Six-item screener to identify cognitive impairment among potential subjects for clinical research. Medical Care, 40(9), 771–781. doi: 10.1097/00005650-200209000-00007 [DOI] [PubMed] [Google Scholar]
  7. Collaborators GBDA (2018). Alcohol use and burden for 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet, 392(10152), 1015–1035. doi: 10.1016/S0140-6736(18)31310–2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cummins DM, & Tobian R. (2018). Motivational Enhancement Therapy for Veterans with Chronic Pain and Substance Use. Health and Social Work, 43(4), 269–273. doi: 10.1093/hsw/hly026 [DOI] [PubMed] [Google Scholar]
  9. Davis AK, Bonar EE, Ilgen MA, Walton MA, Perron BE, & Chermack ST (2016). Factors associated with having a medical marijuana card among Veterans with recent substance use in VA outpatient treatment. Addictive Behaviors, 63, 132–136. doi: 10.1016/j.addbeh.2016.07.006 [DOI] [PubMed] [Google Scholar]
  10. Day MA, Jensen MP, Ehde DM, & Thorn BE (2014). Toward a theoretical model for mindfulness-based pain management. The Journal of Pain, 15(7), 691–703. doi: 10.1016/j.jpain.2014.03.003 [DOI] [PubMed] [Google Scholar]
  11. De Vita MJ, Moskal D, Maisto SA, & Ansell EB (2018). Association of Cannabinoid Administration With Experimental Pain in Healthy Adults: A Systematic Review and Meta-analysis. JAMA Psychiatry, 75(11), 1118–1127. doi: 10.1001/jamapsychiatry.2018.2503 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Goldberg SB, Tucker RP, Greene PA, Davidson RJ, Kearney DJ, & Simpson TL (2019). Mindfulness-based cognitive therapy for the treatment of current depressive symptoms: a meta-analysis. Cognitive Behaviour Therapy, 48(6), 445–462. doi: 10.1080/16506073.2018.1556330 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gros DF, Szafranski DD, Brady KT, & Back SE (2015). Relations Between Pain, PTSD Symptoms, and Substance Use in Veterans. Psychiatry, 78(3), 277–287. doi: 10.1080/00332747.2015.1069659 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Group WAW (2002). The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): development, reliability and feasibility. Addiction, 97(9), 1183–1194. doi: 10.1046/j.13600443.2002.00185.x [DOI] [PubMed] [Google Scholar]
  15. Hammond DC (2010). Hypnosis in the treatment of anxiety- and stress-related disorders. Expert Review of Neurotherapeutics, 10(2), 263–273. doi: 10.1586/ern.09.140 [DOI] [PubMed] [Google Scholar]
  16. Hofmann SG, Sawyer AT, Witt AA, & Oh D. (2010). The effect of mindfulness-based therapy on anxiety and depression: A meta-analytic review. Journal of Consulting and Clinical Psychology, 78(2), 169–183. doi: 10.1037/a0018555 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Humeniuk R, Ali R, Babor TF, Farrell M, Formigoni ML, Jittiwutikarn J, … Simon S. (2008). Validation of the Alcohol, Smoking And Substance Involvement Screening Test (ASSIST). Addiction, 103(6), 1039–1047. doi: 10.1111/j.1360-0443.2007.02114.x [DOI] [PubMed] [Google Scholar]
  18. Ilgen MA, Bohnert AS, Chermack S, Conran C, Jannausch M, Trafton J, & Blow FC (2016). A randomized trial of a pain management intervention for adults receiving substance use disorder treatment. Addiction, 111(8), 1385–1393. doi: 10.1111/add.13349 [DOI] [PubMed] [Google Scholar]
  19. Jensen MP, & Patterson DR (2014). Hypnotic approaches for chronic pain management: clinical implications of recent research findings. American Psychologist, 69(2), 167–177. doi: 10.1037/a0035644 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kennedy J, Roll JM, Schraudner T, Murphy S, & McPherson S. (2014). Prevalence of persistent pain in the U.S. adult population: new data from the 2010 national health interview survey. J Pain, 15(10), 979–984. doi: 10.1016/j.jpain.2014.05.009 [DOI] [PubMed] [Google Scholar]
  21. Kwak C, & Clayton-Matthews A. (2002). Multinomial logistic regression. Nursing Research, 51(6), 404–410. doi: 10.1097/00006199-200211000-00009 [DOI] [PubMed] [Google Scholar]
  22. London AS, Wilmoth JM, Oliver WJ, & Hausauer JA (2020). The Influence of Military Service Experiences on Current and Daily Drinking. Substance Use and Misuse, 55(8), 1288–1299. doi: 10.1080/10826084.2020.1735438 [DOI] [PubMed] [Google Scholar]
  23. Medicine I. o. (2011). Relieving pain in america: A blueprint for transforming prevention, care, education, and research. In. Washington, DC: The National Academies Press. [PubMed] [Google Scholar]
  24. Metrik J, Bassett SS, Aston ER, Jackson KM, & Borsari B. (2018). Medicinal versus Recreational Cannabis Use among Returning Veterans. Translational Issues in Psychological Science, 4(1), 6–20. doi: 10.1037/tps0000133 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nagaya T, Yoshida H, Takahashi H, & Kawai M. (2007). Cigarette smoking weakens exercise habits in healthy men. Nicotine and Tobacco Research, 9(10), 1027–1032. doi: 10.1080/14622200701591575 [DOI] [PubMed] [Google Scholar]
  26. Nahin RL (2017). Severe Pain in Veterans: The Effect of Age and Sex, and Comparisons With the General Population. The Journal of Pain, 18(3), 247–254. doi: 10.1016/j.jpain.2016.10.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nick TG, & Campbell KM (2007). Logistic Regression. In Ambrosius WT(Ed.), Topics in Biostatistics. Methods in molecular biology (Vol. 404): Humana Press. [DOI] [PubMed] [Google Scholar]
  28. Peacock A, Leung J, Larney S, Colledge S, Hickman M, Rehm J, … Degenhardt L. (2018). Global statistics on alcohol, tobacco and illicit drug use: 2017 status report. Addiction, 113(10), 1905–1926. doi: 10.1111/add.14234 [DOI] [PubMed] [Google Scholar]
  29. Rosen MI, Becker WC, Black AC, Martino S, Edens EL, & Kerns RD (2019). Brief Counseling for Veterans with Musculoskeletal Disorder, Risky Substance Use, and Service Connection Claims. Pain Medicine, 20(3), 528–542. doi: 10.1093/pm/pny071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rotaru TS, & Rusu A. (2016). A Meta-Analysis for the Efficacy of Hypnotherapy in Alleviating PTSD Symptoms. International Journal of Clinical and Experimental Hypnosis, 64(1), 116–136. doi: 10.1080/00207144.2015.1099406 [DOI] [PubMed] [Google Scholar]
  31. Sherman BJ, & McRae-Clark AL (2016). Treatment of Cannabis Use Disorder: Current Science and Future Outlook. Pharmacotherapy, 36(5), 511–535. doi: 10.1002/phar.1747 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shih M, Yang YH, & Koo M. (2009). A meta-analysis of hypnosis in the treatment of depressive symptoms: a brief communication. International Journal of Clinical and Experimental Hypnosis, 57(4), 431–442. doi: 10.1080/00207140903099039 [DOI] [PubMed] [Google Scholar]
  33. Unger JB (1996). Stages of change of smoking cessation: relationships with other health behaviors. American Journal of Preventive Medicine, 12(2), 134–138. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/8777067 [PubMed] [Google Scholar]
  34. Valentine KE, Milling LS, Clark LJ, & Moriarty CL (2019). The Efficacy of Hypnosis as a Treatment for Anxiety: A Meta-Analysis. International Journal of Clinical and Experimental Hypnosis, 67(3), 336–363. doi: 10.1080/00207144.2019.1613863 [DOI] [PubMed] [Google Scholar]
  35. Verges A, Ellingson JM, Schroder SA, Slutske WS, & Sher KJ (2018). Intensity of Daily Drinking and Its Relation to Alcohol Use Disorders. Alcoholism: Clinical and Experimental Research, 42(9), 1674–1683. doi: 10.1111/acer.13812 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wagner TH, Harris KM, Federman B, Dai L, Luna Y, & Humphreys K. (2007). Prevalence of substance use disorders among veterans and comparable nonveterans from the National Survey on Drug Use and Health. Psychological Services, 4(3), 149–157. [Google Scholar]
  37. Whiting PF, Wolff RF, Deshpande S, Di Nisio M, Duffy S, Hernandez AV, … Kleijnen J. (2015). Cannabinoids for Medical Use: A Systematic Review and Meta-analysis. JAMA, 313(24), 2456–2473. doi: 10.1001/jama.2015.6358 [DOI] [PubMed] [Google Scholar]
  38. Williams RM, Day MA, Ehde DM, Turner AP, Ciol MA, Gertz KJ, … Jensen MP. (2022). Effects of Hypnosis vs Mindfulness Meditation vs Education on Chronic Pain Intensity and Secondary Outcomes in Veterans: A Randomized Clinical Trial. Pain, 63(10), 1905–1918. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Williams RM, Ehde DM, Day M, Turner AP, Hakimian S, Gertz K, … Jensen MP. (2020). The chronic pain skills study: Protocol for a randomized controlled trial comparing hypnosis, mindfulness meditation and pain education in Veterans. Contemporary Clinical Trials, 90, 105935. doi: 10.1016/j.cct.2020.105935 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wyse JJ, Lovejoy J, Holloway J, Morasco BJ, Dobscha SK, Hagedorn H, & Lovejoy TI (2021). Patients’ perceptions of the pathways linking chronic pain with problematic substance use. Pain, 162(3), 787–793. doi: 10.1097/j.pain.0000000000002077 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zeidan F, & Vago DR (2016). Mindfulness meditation-based pain relief: a mechanistic account. Annals of the New York Academy of Sciences, 1373(1), 114–127. doi: 10.1111/nyas.13153 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplemental Material
NIHMS1902112-supplement-Supplemental_Material.pdf (183.9KB, pdf)

RESOURCES