Music as an auditory cue for emotions and cravings in adults with substance use disorders

Abstract

Music is commonly found in substance using contexts yet little is known about whether music acts as an auditory cue for emotions and cravings that might lead to substance use. The current study addressed two questions: first, whether individuals in treatment for substance use disorders (SUD) show different emotional responses to music compared to matched controls, and second, whether music listening can increase and reduce cravings to use substances in individuals with SUD. Participants were 19 adults in residential treatment for SUD and 19 healthy adults matched for age and gender (both samples had a mean age of 31 years and 53% males). There were significant between-group differences in emotional response to relaxing, happy, and sad music – in particular, participants with SUD showed a dampened response to happy music. Furthermore, after listening to a participant-selected song related to their substance use, individuals with SUD experienced an increase in cravings, while after listening to a nominated abstinent song, there was a decrease in cravings. These results show that music may act as a mild auditory cue for emotions and cravings in adults with SUD. Potential uses of music in SUD treatment are discussed, such as musical stimuli for cue exposure.

Keywords

auditory cue cravings emotions music listening substance use disorders

The role of music as an important therapeutic tool is well established in the treatment of many psychological disorders including psychosis (Gold, Solli, Krüger, & Lie, 2009), depression (Maratos, Gold, Wang, & Crawford, 2008), and other internalising disorders such as anxiety and social withdrawal (Miranda, Gaudreau, Debrosse, Morizot & Kirmayer, 2012). In the arena of substance use disorders (SUD) however, practice and research regarding music is still in its infancy (Aldridge & Fachner, 2010; Dingle, Gleadhill, & Baker, 2008; Silverman, 2003). As a result, many fundamental questions about how individuals with substance use problems respond to music have not been answered. One of these basic questions is: how do substance users respond emotionally to music? If individuals in treatment for substance use disorders show a different emotional response to music compared to healthy controls, does this have implications for treatment of substance misuse?

Emotional responses to music

Strong emotions are one of the most widely cited maintaining factors for substance misuse in that the substance numbs or alleviates negative internal states – a process of negative reinforcement (Kassel, 2010; McCarthy, Curtin, Piper, & Baker, 2010). Furthermore, a large body of research demonstrates that music listening can both induce and help alleviate strong emotions (Juslin & Sloboda, 2010). Therefore, understanding how substance users emotionally respond to music has potentially important implications for substance use treatment. Of the published studies investigating the emotional responses of substance users to music, the majority have investigated substance users’ emotional responses to music therapy techniques (Baker, Gleadhill, & Dingle, 2007; de l’Etoile, 2002; Jones, 2005; Silverman, 2003, 2011; Soshensky, 2001). Only one study to the authors’ knowledge has explored the effects of everyday music listening on the emotional responses of substance users (Dingle, Kelly, Flynn, & Baker, under review). In this study, participants were asked to reflect on their substance using period prior to admission into treatment, and to describe how music affected their substance use experience and vice versa. While more than half of the participants chose not to answer these two questions, among those who responded, emotional effects were most noticeable. For example, music played during substance use was reported to enhance the euphoria from the substance use (41.2%) and also to enhance feelings of sadness and loss (23.5%). Music listening during substance use also influenced the vividness of dreams and memories (17.6%). Conversely, substance use while listening to music was reported to intensify both positive and negative emotional responses (21.4%). A minority (14%) indicated that the combination of music and substances made them feel worse. Although these preliminary findings do suggest that substance users have strong emotional responses when listening to music, it is unknown whether these effects occur once individuals have entered treatment and are abstinent from substance use.

The most well-established experimental protocol for examining listeners’ emotional responses to music has been to record individuals’ responses to specific musical stimuli, often in the form of short pieces of classical music which have been previously shown to induce specific emotions (see, e.g., Baumgartner, Esslen, & Jäncke, 2006; Etzel, Johnsen, Dickerson, Tranel, & Adolphs, 2006). The current study used this protocol to examine emotional responses to happy, sad, and relaxing music among individuals with SUD, and in particular, to determine if their emotional responses to music differ from that of matched control participants.

Craving responses to music

The second aim of the study was to examine the relationship between music listening and craving for substances in the SUD treatment sample. Craving has been described in the DSM-5 (American Psychiatric Association, 2013) as “a strong desire or urge to use the substance” and desire may be defined as “an affectively charged cognitive event in which an object or activity that is associated with pleasure or relief of discomfort is in focal attention” (Kavanagh, Andrade, & May, 2005, p. 447). Cue reactivity research demonstrates that visual and olfactory cues for substances can induce cravings (Loflin & Earleywine, 2014; Pulido, Brown, Cummins, Paulus, & Tapert, 2010). This occurs through a classical conditioning process of repeated exposure to environmental cues (such as the sight of alcohol or the smell of cannabis) paired with thoughts regarding the substance’s effect. Consequently in the presence of the same cue, the conditioned thoughts then drive urges to use the substance (Stacy, Widaman, & Marlatt, 1990). The links between cues, cravings and substance use behaviour are well established and therefore strategies to manage craving are commonly included in treatment programs for substance addiction (Addolorato, Abenavoli, Leggio, & Gasbarrini, 2005; Weiss et al., 2003).

In cue exposure treatments, the participant is repeatedly exposed to stimuli relevant to the substance of abuse while the behaviour itself (use of the substance) is prevented. Over time, this process is expected to stop (extinguish) the substance use response to the cue. However, a meta-analysis of clinical studies using cue exposure treatment for a range of substances found it was ineffective (Conklin & Tiffany, 2002). The authors referred to evidence from animal studies to argue that cue exposure has not been effective because it is typically conducted in a setting that is isolated from the usual substance using context. Take, for example, a person who is alcohol dependent and undergoing cue exposure treatment in a rehabilitation centre where there is no alcohol. Conklin and Tiffany (2002) argue that the link between alcohol cues and drinking behaviour for this person will be stronger (and likely to be reinstated) in their home context where it was learned rather than in the rehabilitation context. This evidence indicates that cue exposure treatment could be improved by incorporating a broader range of contextual cues that approximate the participants’ typical substance using environment.

Despite the fact that music is commonly found in the environmental context of substance use, little is known about the potential for music to act as an auditory cue for substance cravings and use. For example, an individual who has repeatedly smoked cannabis while listening to reggae music may experience cravings for cannabis when s/he hears reggae music during a period of treatment. But is there any direct evidence that music can cue cravings for substances? Despite anecdotal evidence suggesting it can (Horesh, 2010), the question has received very little research attention. Dingle, Kelly, and colleagues (under review) explored this question by surveying 106 clients from six different residential rehabilitation services, and 43% of clients reported that listening to particular songs increased their urge to use substances. Although these results suggest that music can increase cravings (at least in some individuals) only one study has tested this link using experimental methods. In this study, Jansma, Breteler, Schippers, De Jong, and Van der Staak (2000) played participants a “depressing” music clip in order to test the influence of depressed mood on cravings. This study found no effect of the depressing music on cravings; however, the music was paired with the cues of pouring and sniffing a glass of alcohol so the specific effects of the music could not be determined.

If music can indeed increase cravings in substance users, the next logical question is can music also reduce these cravings? In a study exploring this question, Silverman (2011) assigned 141 patients in a detoxification unit to either a “rockumentary” music intervention (lyric analysis of the song “Under the Bridge” by the Red Hot Chili Peppers along with discussion of the band’s history and substance use and questions about how the lyrics relate to the clients’ own substance use and recovery), a recreation music therapy condition (“music bingo” plus discussion), or a (non-music related) verbal therapy condition and measured differences between groups in cravings. Results revealed no statistically significant differences between groups at post-test (post-test only design), though participants in both music therapy conditions had lower mean craving scores than participants in the verbal therapy condition. Overall, these results do suggest that music had a favourable impact on cravings; however, the study did not gather pre-test data and therefore it is unclear whether music did indeed decrease cravings from pre- to post-test. Furthermore, it is not possible to determine the influence of music listening in particular upon cravings from these results because several music therapy methods were used in combination in this study.

The current study uses an experimental design in which adults undergoing residential treatment for SUD (and a matched control sample) will be played experimenter-selected music that has been found in previous research to elicit emotions of relaxation, happiness, and sadness in listeners. Previous research suggests that adults in treatment for substance use disorders are emotionally sensitive to music, therefore hypothesis 1 is that there will be no difference between the treatment group and the control group in emotional responses to happy, sad, and relaxing music. In addition, we will investigate participants’ urge to use a substance before and after hearing a piece of music s/he nominates as being associated with previous substance use, and after hearing a second piece of music selected for helping him/her stay abstinent. In accordance with the principles of cue conditioning and earlier research suggesting that adults in treatment for substance use disorders experience an increase in urge to use substances when listening to personally relevant music, hypothesis 2 is that there will be an increase in urge to use substances after listening to a participant-selected “urge” song, and a decrease in urge to use substances after listening to a participant selected “clean” song.

Method

Participants

SUD sample

Nineteen participants were receiving treatment at a residential Therapeutic Community in Queensland, Australia. The sample had a mean age of 31.05 (SD = 5.94), nine were female, and 10 male. The treatment program was of six months duration and at the time of testing participants had been living at the Therapeutic Community an average of 64 days (SD = 48). The program requirements included complete abstinence from alcohol and drugs (besides prescription medications) and daily participation in therapeutic activities (e.g., relapse prevention groups). Residents were accepted into the program after a clinical interview had determined that they met DSM-IV criteria for substance dependence.¹ Participants’ most problematic substance was alcohol (42%), amphetamines (32%), marijuana (11%), prescription drugs (11%), and heroin (5%). All but one participant reported being a polysubstance user (more than one substance in a day), with the mean number of years of polysubstance use of 12.47 years (range: 1–24 years). Regarding musical involvement, 26% of the sample was currently singing in a choir/group or playing a musical instrument, 63% had done so at some point in their lives, with a mean number of years doing these activities of 7.11 years. Regarding levels of enjoyment for classical music, the SUD sample had a mean rating of 2.58 out of a possible 7 (1 = not at all to 7 = a great deal).

Control sample

Nineteen age and gender matched controls were recruited through word of mouth via their acquaintance with the first author. The absence of a substance use disorder in each control participant was determined by means of a short interview with a qualified clinical psychologist. Regarding musical involvement, 26% of the sample was currently singing in a choir/group or playing a musical instrument, 84% had done so at some point in their lives, with a mean number of years doing these activities of 7.47 years. The mean rating of enjoyment for classical music was 4.16 out of a possible 7. There was no significant difference between the two samples in the percentage of the sample currently singing in a choir/group or playing a musical instrument: χ² (1, N = 38) = 1.00, p = .32, percentage who had done so at some point in their lives: χ² (1, N = 38) = 0.27, p = .60, or in the mean number of years doing these activities: F (1, 37) = 0.02, p = .89. There was a significant difference between groups in enjoyment for classical music: F (1, 37) = 7.67, p < .01, so this was entered as a covariate in the relevant analyses. A power analysis using the software package G*Power 3.1.9.2 (Faul, Erdfelder, Lang, & Buchner, 2007) calculated that for a repeated measures ANOVA with mixed within–between interactions, N = 38 with two groups and three measurements, a medium effect size (0.25) and alpha set at 0.05, the power is 0.92.

Measures

Genre preferences

Participants’ preferences for musical genres were assessed via 14 items each representing a different genre (blues, jazz, classical, folk, rock, alternative, heavy metal, country, soundtracks, pop, rap/hip-hop, RnB/soul/funk, electronic/dance, and religious) based on the most preferred music genres in previous research (Schäfer & Sedlmeier, 2009). Participants were asked to rate their “level of enjoyment for each musical style” on a 7-point Likert scale (1 = not at all to 7 = a great deal).

Emotional response

Emotional valence after each piece of music was assessed with the item “How pleasant/unpleasant do you currently feel?”, which listeners rated on a 7-point Likert scale from 1 = very unpleasant/negative to 7 = very pleasant/positive. Emotional arousal was assessed with the item “How calm or aroused do you currently feel?”, also rated on a 7-point scale from 1 = very calm to 7 = very aroused. The second measure of emotional response, the Geneva Emotions in Music Scale (GEMS-9; Zentner, Grandjean, & Scherer, 2008) is a 9-item self-report scale that measures the type and intensity of an individual’s emotional response to a musical stimulus. Specifically devised to measure musically evoked emotions directly following music listening, each of the 9 items represents an emotion label (e.g., “Peacefulness”), and within each emotion label are emotion adjectives which help to define each label (e.g., “Peacefulness: serene, calm, soothed”). Respondents rate the intensity with which they feel each emotion label on a 5-point Likert scale (1 = not at all to 5 = very much). For the purpose of the current study, only the three GEMS-9 items representative of the emotions induced by the three music pieces were analysed. These were “Peacefulness: serene, calm, soothed” (for the Relaxing music condition), “Joyful activation: joyful, amused, bouncy” (for the Happy music condition), and “Sadness: sad, sorrowful” (for the Sad music condition).

Craving measures

Craving to use substances was assessed after each music stimulus by a single item formulated for this experiment: “I have an urge to use/drink” on a 7-point Likert scale (1 = no urge to 7 = very strong urge). In addition, an adapted version of the Alcohol Urge Questionnaire (AUQ; Bohn, Krahn, & Staehler, 1995) was used. The AUQ is an 8-item measure of drinking urges and asks participants to endorse the extent to which they agree or disagree with statements relating to their desire to drink, expectation of a desired outcome from drinking, and inability to avoid drinking if alcohol was available. The scale has high internal consistency (α = 0.91), high test-retest reliability (1-day acute craving complete interval, 0.82), and good construct validity evidenced by strong correlations with measures of alcohol dependence severity and the Obsessive Compulsive Drinking Scale (Love, James, & Willner, 1998). Item wording was adjusted so that all items were applicable to both drug use as well as alcohol use.

Music stimuli

Three separate pieces of classical music were chosen which previous research had shown to elicit feelings of relaxation, happiness, and sadness in listeners (Sandstrom & Russo, 2013; Witvliet & Vrana, 1996). The relaxing music was Georges Bizet’s Carmen Suite no. 1: Intermezzo; the happy music was Mozart’s Horn Concerto No. 4 in E flat major, K495: Rondo, Allegro vivace; and the sad music was John Stanley’s Grand Voluntary in D major, Adagio. Witvliet and Vrana (1996) demonstrated that these three songs reliably induced the three different emotional states in two separate samples of undergraduate students (N = 85 and N = 67). Each piece was played for 60 seconds. The title and artist of the urge songs and clean songs nominated by participants in the treatment sample are listed in Table 1.

Table 1.

Urge songs and clean songs nominated by the adults in residential alcohol and drug treatment.

Participant	Urge song title	Urge song artist	Clean song title	Clean song artist
1	Jar of Hearts	Christina Perri	Leaving on a Jet Plane	John Denver
2	Enter Sandman	Metallica	Buy Now Pay Later (Charlie No. 2)	The Whitlams
3	Signs	Bloc Party (Armand Van Helden remix)	Run to the Water	Live
4	Blister in the Sun	Violent Femmes	Blinded by the Lights	The Streets
5	Diamonds (The Ultimate Hip-Hop Collection)	e.g., I Got Them by Yo Gotti and Lil Wayne	Mockingbird	Eminem
6	Drugged and Driving	Dayglow Abortions	The Wall Pt 3	Pink Floyd
7	Friends for Life	Marshall & the Fro	Coming Home, I’m a Soldier	Dan Hannaford
8	Dead and Gone	T. I.	Ridin’ Solo	Jason Derulo
9	Back in Black	AC/DC	God Put a Smile Upon Your Face	Coldplay
10	Loving Cup	The Rolling Stones	The Ocean	Led Zeppelin
11	Piano Man	Billy Joel	Vincent	Don McLean
12	Toy Soldiers	Martika	Relaxation music	Stephen Halpern
13	Put Your Hands Up for Detroit	Fredde le Grand	Talking About A Revolution	Tracey Chapman
14	Fool’s Gold	Stone Roses	What You Gonna Do	Plan B
15	Stay Awake	Example	Blackbird	Beatles
16	Put It Down	Kottonmouth Kings	Skeleton Dance	Arch Enemy
17	My Happiness	Powderfinger	Wild Horses	Rolling Stones
18	Where’s the Weed At	Kottonmouth Kings	Rehab	Amy Winehouse
19	Changed the Way You Kiss Me	Example	Girl On Fire	Alicia Keys

Procedure

SUD sample

For each participant, testing was carried out in three phases. In the first phase, participants completed a 5–10 minute questionnaire which included items regarding participant demographics, substance use history, music involvement, and the music genre preference items. Couched within this questionnaire were two items which asked each participant to nominate an “urge song” (“Please write below a song that makes you want to use/drink?”), and a “clean song” (“Please write below a song that helps you stay clean/sober?”). The second and third phases were carried out one week later so as to reduce any effect of priming in the experimental procedure. In the second phase, participants completed baseline measures of their emotional state (valence, arousal, GEMS-9) and cravings (single-item, adapted AUQ). They were then asked to listen to the three pieces of classical music (relaxed, happy, and sad), and directly following each piece, participants rated their emotional response and cravings. Participants were given written instructions to rate how the music made them feel and not to describe the music or what emotion it might express. The order of presentation of the three musical stimuli was counterbalanced. In the third phase, participants listened to their selected “urge song” followed by their selected “clean song”. After each song, participants again rated their emotional response and cravings. All stimuli were presented and responses gathered on a computer interface using Qualtrics survey software. Music stimuli were downloaded from iTunes and embedded in the experimental procedure for each unique participant.

Control sample

The age and gender matched controls completed phase one and two of the experimental procedure described above (without the cravings measures). Participants in both groups were all required to listen to the music excerpts using headphones (Sennheiser HD215). All participants were also assured of voluntary participation and confidentiality, and each was debriefed as to the purpose of the study following completion. The methods were approved by the University of Queensland Ethics Committee, #201100953.

Results

Emotional responses to music

Valence

Mean valence ratings for each group across the three music emotions are presented in Figure 1. A 2 (Samples) × 3 (music emotions) mixed analysis of variance conducted on valence showed no main effect of group: F (1, 36) = 0.49, p = .491. There was a significant main effect of music emotion: F (1.65, 59.39) = 14.03, p < .001, $η_{p}^{2}$ = 0.28, moderated by a group × music emotion interaction: F (1.65, 59.39) = 4.68, p < .05, $η_{p}^{2}$ = 0.12. Tests of the simple effects revealed that the SUD sample rated the valence of the three pieces of music equally (slightly on the pleasant side of neutral): F (2, 36) = 1.42, p = .256; relaxing-happy: p = 1.000, relaxing-sad: p = .494, happy-sad: p = .837. In contrast, for the Control sample there were significant differences in valence ratings between stimuli: F (2, 36) = 16.45, p < .001, $η_{p}^{2}$ = 0.48, such that the happy and relaxing pieces were rated as significantly more pleasant than the sad music (relaxing-happy: p = 1.000, relaxing-sad: p < .01, happy-sad: p < .01). Hence, whilst Controls’ valence ratings fluctuated across the three pieces, SUD individuals’ valence remained relatively stable across music emotions. The same pattern of results was found when classical music preference ratings were entered as a covariate. Although enjoyment of classical music as a genre is not the same as the participants’ emotional response to the specific pieces used in the study, this finding makes it less likely that the group differences are due to differences in preference for classical music between the two samples.

Figure 1.

Mean emotional valence ratings for relaxing, happy, and sad music given by 19 adults with substance use disorders (diamonds) and 19 matched controls (squares).

Arousal

Mean arousal ratings for each sample across the three music emotions are presented in Figure 2. There is a clear divergence of arousal response between the two samples when listening to the happy music. A 2 × 3 mixed ANOVA showed no significant main effect of sample: F (1, 36) = 0.34, p = .564 or music stimulus: F (2, 72) = .51, p = .602. However, the group × music interaction was significant: F (2, 72) = 7.95, p < .001, $η_{p}^{2}$ = 0.01. Pairwise comparisons of arousal ratings confirmed that there was a significant difference between groups for the happy song: F (1, 108) = 7.96, p < .01, $η_{p}^{2}$ = 0.21, such that the Control sample showed higher levels of arousal than the SUD sample. There was no significant difference between groups in response to the relaxing music: F (1, 108) = 0.62, p = .433 or the sad music: F (1, 108) = 0.01, p = .921. Tests of the simple effects of music emotion showed that there was no significant difference in arousal ratings of the three songs for the SUD sample: F (1.31, 23.67) = 2.56, p = .115; relaxing-happy: p = .288, relaxing-sad: p = 1.000, happy-sad: p = .367. In contrast, the Control sample showed a varied pattern of arousal across music emotions: F (2, 36) = 5.51, p < .01, such that both the relaxing and sad songs were rated as significantly less arousing than the happy song (relaxing-happy: p < .05; relaxing-sad: p = 1.000, happy-sad: p < .05). Again these findings were robust to the addition of the covariate, preference for classical music.

Figure 2.

Mean emotional arousal ratings for relaxing, happy, and sad music given by 19 adults with substance use disorders (diamonds) and 19 matched controls (squares).

GEMS items

Group means on the GEMS items for each music piece are presented in Figure 3. Results of three separate one-way ANOVAs revealed that there was a significant difference between groups for the happy music: F (1, 36) = 14.00, p < .01, $η_{p}^{2}$ = 0.28, such that the Control group showed higher intensity of joyful activation than the SUD group. There was however no significant difference between groups on each respective GEMS item for the relaxing music: F (1, 36) = 2.83, p = .101, or the sad music: F (1, 36) = 2.31, p = .137. When each of the three ANOVAs was run with classical genre preference entered as a covariate, identical statistically significant results were observed.

Figure 3.

Group means on the Geneva Emotion in Music Scale items for each music stimulus.

Music and cravings

Single craving item

A one-way repeated measures ANOVA revealed a significant effect of Time on craving: F (153, 27.46) = 19.01, p < .001, $η_{p}^{2}$ = 0.51. Follow-up pairwise comparisons of cravings across times revealed a significant increase in craving from baseline to urge song (p < .01), and a significant decrease in craving from urge song to clean song (p < .001). There was no significant difference in levels of craving between baseline and clean song (p = .412). No Figure is included for this measure as it was so close to the Adapted AUQ results shown in Figure 4.

Figure 4.

Mean ratings on the adapted Alcohol Urge Questionnaire at baseline and after listening to music in 19 adults with substance use disorders.

Adapted AUQ

A similar pattern of results was found on the adapted AUQ (see Figure 4). A one-way repeated measures ANOVA revealed a significant effect of Time on craving: F (2, 36) = 15.44, p < .001, $η_{p}^{2}$ = 0.46. Follow-up pairwise comparisons of cravings between time points revealed a significant increase in craving from baseline to urge song (p < .01), and a significant decrease in craving from urge song to clean song (p < .001). There was no significant difference in levels of craving between baseline and the clean song (p = .488). It is important to note for both cravings measures, although craving significantly increased after the urge song, the mean rating did not go above the midpoint on either scale, suggesting that only a mild craving response was experienced.

Discussion

The current study aimed to answer some fundamental questions regarding how individuals in treatment for substance use disorders respond to music. We first explored whether people with SUD have different emotional responses when listening to experimenter-selected relaxing, happy, and sad music compared to healthy controls. Overall, results revealed that compared to controls, the participants in the therapeutic community showed significantly less variation in emotional response across relaxing, happy, and sad music. The group difference was most pronounced for happy music whereby the SUD individuals’ ratings of pleasure (valence), intensity of joyful activation and emotional arousal were significantly lower than controls’. The results cannot be attributed to differences between groups in musical involvement and training, as the groups had comparable musical backgrounds. In terms of different preferences for classical music, Table 1 shows that the participants in the SUD sample listened to music from genres such as hip-hop, rock, pop, and RnB but did not list any classical music pieces as either their “urge” song or their “clean” song. Furthermore, their mean preference rating for classical music was significantly lower than that for the Control sample. Experimenter-selected classical music excerpts were used in the study because they have been found in previous research to evoke the desired emotional states and we wanted experimental control over the music stimuli. However, it is quite possible that the pattern of emotional responses was influenced by the experimenter-selected music and further research is required to understand how SUD participants respond emotionally to personally meaningful happy, sad, and relaxing music.

Another possible explanation for the between-group effects – in response to both happy and sad music – was the likely presence of depression in the SUD group. Separate research conducted at the same therapeutic community revealed high rates of depression amongst residents, particularly in first few months after entering the program: 44% had experienced serious depression in the month before entering the community, and 22% had suicidal thoughts (Dingle, Stark, Cruwys, & Best, 2014); however, depression symptoms were not measured in participants of the current study. The SUD sample rated the sad music as more pleasant and less sad than the Control sample; and they rated the happy music as less pleasant and less joyful than the Control group. Assuming that participants experienced some depressed mood, this pattern of responses to happy and sad music may reflect a preference for music that is mood-congruent, as has been found in previous research (Papinczak, Dingle, Stoyanov, Hides & Zelenko, 2015; van den Tol & Edwards, 2013). Furthermore, previous research has revealed that people with depression show more blunted emotional responses to musical excerpts representative of a range of emotions (Al’tman, Alyanchikova, Guzikov, & Zakharova, 2000; Osuch et al., 2009; Punkanen, Eerola, & Erkkilä, 2011).

Although the effect of music on emotions in this study was modest, it is consistent with a larger body of research on the uses of music to influence mood and/or behaviour, both in the addiction treatment context (Baker et al., 2007), and other settings such as music mood induction during laboratory studies (Etzel et al., 2006), music for anxiety and pain management in dental and medical settings (Bernatzky, Presch, Anderson, & Panksepp, 2011; Thoma et al., 2014) and music to influence emotions and performance in sporting contexts (Lane, Davis, & Devonport, 2011).

The second research question explored whether music listening could influence cravings for substances in people with SUD undergoing rehabilitation. To the authors’ knowledge, this was the first study to empirically test whether music listening alone can influence self-reported cravings to use substances using an experimental pre–post music listening design. Figure 4 shows that at baseline, participants’ cravings and urges to use substances were rated as low. Listening to their nominated urge song significantly increased ratings of cravings and urges to use substances, and these ratings returned to baseline following listening to their nominated clean song. Although these changes in cravings were not dramatic, they were observed for the majority of participants. It may be that music stimuli evoke a milder craving response than other stimuli directly associated with the substance (such as the sight or smell of drug taking equipment), although a direct comparison of various types of stimuli is needed in order to verify this. The modest changes in cravings may be explained by the fact that participants were already settled in rehabilitation, were stabilised in terms of physiological withdrawal, and had already learned several strategies and techniques in the treatment program to control and contain their cravings. A limitation of the study was that we played the urge song before the clean song in all cases to ensure that participants would not complete the study with an increased urge to use substances. Unfortunately this means the effect of listening to the “clean” song on its own remains uncertain. Nevertheless, this finding suggests that clients could listen to their clean song/s while experiencing a craving to experience a modest decrease in craving.

The clinical implications of the findings are that individuals in SUD treatment should be strategic in their choice of music to evoke the effects they want. For instance, it may be useful to evoke strong emotions during therapy so that clients can explore and learn to tolerate these emotions in a substance-free environment (Baker et al., 2007). Given their blunted emotional response, clinicians may need to work with clients to select pieces of music that induce high levels of emotional valence and arousal, and additional techniques may be required to strengthen the intensity of their emotional experience (e.g., imagery – see Holmes & Matthews, 2010) in order for emotions to reach a threshold with which they can be effectively explored.

There are several important clinical implications of the findings regarding cravings. First, clinicians could educate clients in SUD treatment about the potential effect of their everyday music listening on cravings. An analysis of the nominated urge songs and clean songs is beyond the scope of the present study; however, it can be seen from the song list in Table 1 that many of the urge songs contained themes about substance use and some lyrics normalise or even glorify substance use (for example, Blister in the Sun by the Violent Femmes contains the lyrics “… yeah I’m so strung out, I’m high as a kite I just might stop to check you out”; while the song Drugged and Driving by Dayglow Abortions includes the lyrics “I just ate 25 speeders and I could feel the adrenaline … I stepped on the gas and went with the power”). The messages in such songs are working against the messages that participants receive in treatment and this is an issue that clinicians could bring to their clients’ attention during a discussion of their music listening choices. Clients could be encouraged to review the music in their personal music players to identify songs that trigger and reduce urges to use, and when to listen to them. Simply identifying these songs and raising clients’ awareness of their effects may constitute a useful intervention.

The current findings also highlight the potential for music to be included as part of cue exposure treatment for substance use disorders. Cue exposure therapy is gaining acceptance as a useful treatment for substance use disorders, although there is mixed scientific evidence regarding its efficacy (Carter & Tiffany, 1999; Conklin & Tiffany, 2002). The cues traditionally employed in cue exposure therapy treatment have been alcoholic beverages, drug using equipment, and external environments previously associated with drinking and using. Clients in early abstinence may find such cues extremely confronting and as a result cue exposure therapy has been associated with high drop-out rates (Marissen, Franken, Blanken, van den Brink, & Hendriks, 2007). Music listening may constitute a more tolerable and context relevant stimulus for use in cue exposure procedures and could even be used as a first step in a cue exposure hierarchy.

Future research could investigate whether the blunted emotional response to music of SUD individuals is also found on physiological measures. Although there would be some practical and logistical considerations to overcome, it would also be interesting to conduct brain imaging of people in SUD treatment while they are listening to their urge song and their clean song, to further examine which brain regions may be implicated in these responses to musical stimuli. Additional contextual influences could be explored to ascertain if the effects found in the current study change when participants listen to music in groups (vs. alone), and when living in the community (vs. in residential treatment), as previous research suggests that emotional responses vary depending on the listening context (Gabrielsson, 2001; Scherer & Zentner, 2001). As mentioned above, it would also be worthwhile investigating whether the results found in the current study generalise to relaxing, happy, and sad music selected by participants, and of different musical genres.

In conclusion, this study found preliminary evidence that music listening can act as an auditory cue for emotions and cravings among adults in treatment for substance use disorders. Music listening holds potential as stimuli for personalised cue exposure therapy for these individuals.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Addolorato

Abenavoli

Leggio

Gasbarrini

(2005). How many cravings? Pharmacological aspects of craving treatment in alcohol addiction: A review. Neuropsychobiology, 51(2), 59–66. doi: 10.1159/000084161

Al’tman

Y. A.

Alyanchikova

Y. O.

Guzikov

Zakharova

(2000). Estimation of short musical fragments in normal subjects and patients with chronic depression. Human Physiology, 26(5), 553–557. doi: 0362–1197/00/2605–0553

Aldridge

Fachner

(2010). Music therapy and addictions. London, UK: Jessica Kingsley.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders: Fifth Edition (DSM-5). Washington, DC: American Psychiatric Publishing.

Baker

F. A.

Gleadhill

L. M.

Dingle

G. A.

(2007). Music therapy and emotional exploration: Exposing substance abuse clients to the experiences of non-drug-induced emotions. The Arts in Psychotherapy, 34(4), 321–330. doi: 10.1016/j.aip.2007.04.005

Baumgartner

Esslen

Jäncke

(2006). From emotion perception to emotion experience: Emotions evoked by pictures and classical music. International Journal of Psychophysiology, 60(1), 34–43. doi: 10.1016/j.ijpsycho.2005.04.007

Bernatzky

Presch

Anderson

Panksepp

(2011). Emotional foundations of music as a non-pharmacological pain management tool in modern medicine. Neuroscience & Biobehavioral Reviews, 35(9), 1989–1999. doi: 10.1016/j.neubiorev.2011.06.005

Bohn

M. J.

Krahn

D. D.

Staehler

B. A.

(1995). Development and initial validation of a measure of drinking urges in abstinent alcoholics. Alcoholism: Clinical and Experimental Research, 19(3), 600–606. doi: 10.1111/j.1530–0277.1995.tb01554.x

Carter

B. L.

Tiffany

S. T.

(1999). Meta-analysis of cue-reactivity in addiction research. Addiction, 94(3), 327–340. doi: 0965–2140/99/030327

10.

Conklin

C. A.

Tiffany

S. T.

(2002). Applying extinction research and theory to cue-exposure addiction treatments. Addiction, 97, 155–167. doi: 10.1046/j.1360–0443.2002.00014.x

11.

de l’Etoile

S.K

. (2002). The effectiveness of music therapy in group psychotherapy for adults with mental illness. The Arts in Psychotherapy, 27(1), 31–56. doi: 10.1016/S0197–4556(02)00139–9

12.

Dingle

G. A.

Gleadhill

L. M.

Baker

F. A.

(2008). Can music therapy engage patients in group cognitive behaviour therapy for substance abuse treatment? Drug and Alcohol Review, 27(2), 190–196. doi: 10.1080/09595230701829371

13.

Dingle

G. A.

Kelly

P. J.

Flynn

L. M.

Baker

F. A.

( under review). The influence of music on emotions and cravings in clients in addiction treatment: A mixed methods study of two samples [Manuscript submitted for publication].

14.

Dingle

G. A.

Stark

Cruwys

Best

(2014). Breaking good: Breaking ties with social groups may be good for recovery from substance misuse. British Journal of Social Psychology. Advance online publication. doi: 10.1111/bjso.12081

15.

Etzel

J. A.

Johnsen

E. L.

Dickerson

Tranel

Adolphs

(2006). Cardiovascular and respiratory responses during musical mood induction. International Journal of Psychophysiology, 61(1), 57–69. doi: 10.1016/j.ijpsycho.2005.10.025

16.

Faul

Erdfelder

Lang

A.-G.

Buchner

(2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191.

17.

Gabrielsson

(2001). Emotions in strong experiences with music. In Juslin

P. N.

Sloboda

(Eds.), Music and emotion: Theory and research. (pp. 123–147). Oxford, UK: Oxford University Press.

18.

Gold

Solli

H. P.

Krüger

Lie

S. A.

(2009). Dose-response relationship in music therapy for people with serious mental disorders: Systematic review and meta-analysis. Clinical Psychology Review, 29(3), 193–207. doi: 10.1016/j.cpr.2009.01.001

19.

Holmes

E. A.

Mathews

(2010). Mental imagery in emotion and emotional disorders. Clinical Psychology Review, 30(3), 349–362. doi: 10.1016/j.cpr.2010.01.001

20.

Horesh

(2010). Drug addicts and their music: A story of a complex relationship. In Aldridge

Fachner

(Eds.), Music therapy and addictions (pp. 57–74). London, UK: Jessica Kingsley.

21.

Jansma

Breteler

M. H. M.

Schippers

G. M.

De Jong

C. A. J.

Van Der Staak

C. P. F.

(2000). No effect of negative mood on the alcohol cue reactivity of in-patient alcoholics. Addictive Behaviors, 25(4), 619–624. doi: 10.1016/S0306–4603(99)00037–4

22.

Jones

. (2005). A comparison of songwriting and lyric analysis techniques to evoke emotional change in a single session with people who are chemically dependent. Journal of Music Therapy, 42(2), 94–110.

23.

Juslin

P. N.

Sloboda

(2010). Handbook of music and emotion: Theory, research, applications. Oxford, UK: Oxford University Press.

24.

Kassel

J. D.

(Ed.). (2010). Substance abuse and emotion. Washington, DC: American Psychological Association.

25.

Kavanagh

D. J.

Andrade

May

. (2005). Imaginary relish and exquisite torture: The elaborated intrusion theory of desire. Psychological Review, 112(2), 446–467. doi: 10.1037/0033–295X.112.2.446

26.

Lane

A. M.

Davis

P. A.

Devonport

T. J.

(2011). Effects of music interventions on emotional states and running performance. Journal of Sports Science & Medicine, 10(2), 400–407.

27.

Loflin

M. J.

Earleywine

(2014). Do I smell pot? A trial of an alternative paradigm of olfactory cue-reactance to marijuana. Addiction Research and Theory, 22(3), 210–217. doi: 10.3109/16066359.2013.805750

28.

Love

James

Willner

(1998). A comparison of two alcohol craving questionnaires. Addiction, 93(7), 1091–1102. doi: 10.1046/j.1360–0443.1998.937109113.x

29.

Maratos

Gold

Wang

Crawford

(2008). Music therapy for depression. Cochrane Database of Systematic Reviews, 1, CD004517.

30.

Marissen

M. A.

Franken

I. H.

Blanken

van den Brink

Hendriks

V. M.

(2007). Cue exposure therapy for the treatment of opiate addiction: Results of a randomized controlled clinical trial. Psychotherapy and Psychosomatics, 76(2), 97–105. doi: 10.1159/000097968

31.

McCarthy

D. E.

Curtin

J. J.

Piper

M. E.

Baker

T. B.

(2010). Negative reinforcement: Possible clinical implications of an integrative model. Washington, DC: American Psychological Association.

32.

Miranda

Gaudreau

Debrosse

Morizot

Kirmayer

L. J.

(2012). Music listening and mental health: Variations on internalizing psychopathology. In MacDonald

Kreutz

Mitchell

(Eds.), Music, health, and wellbeing. Oxford, UK: Oxford University Press.

33.

Osuch

E. A.

Bluhm

R. L.

Williamson

P. C.

Théberge

Densmore

Neufeld

R. W. J.

(2009). Brain activation to favorite music in healthy controls and depressed patients. Neuroreport, 20(13), 1204–1208. doi: 10.1097/WNR.0b013e32832f4da3

34.

Papinczak

Z. E.

Dingle

G. A.

Stoyanov

Hides

Zelenko

(2015). Young people’s use of music for wellbeing. Journal of Youth Studies. doi: 10.1080/13676261.2015.1020935

35.

Pulido

Brown

S. A.

Cummins

Paulus

M. P.

Tapert

S. F.

(2010). Alcohol cue reactivity task development. Addictive Behaviors, 35(2), 84–90. doi: 10.1016/j.addbeh.2009.09.006

36.

Punkanen

Eerola

Erkkilä

(2011). Biased emotional recognition in depression: Perception of emotions in music by depressed patients. Journal of Affective Disorders, 130(1), 118–126. doi: 10.1016/j.jad.2010.10.034

37.

Sandstrom

G. M.

Russo

F. A.

(2013). Absorption in music: Development of a scale to identify individuals with strong emotional responses to music. Psychology of Music, 41(2), 216–228. doi: 10.1177/0305735611422508

38.

Schäfer

Sedlmeier

(2009). From the functions of music to music preference. Psychology of Music, 37, 279–300. doi: 10.1177/0305735608097247

39.

Scherer

K. R.

Zentner

M. R.

(2001). Emotional effects of music: Production rules. In Juslin

P. N.

Sloboda

(Eds.), Music and emotion: Theory and research (pp. 361–392). Oxford, UK: Oxford University Press.

40.

Silverman

M. J.

(2003). Music therapy and clients who are chemically dependent: A review of literature and pilot study. Arts in Psychotherapy, 30, 273–281. doi: 10.1016/j.aip.2003.08.004

41.

Silverman

M. J.

(2011). Effects of music therapy on change readiness and craving in patients on a detoxification unit. Journal of Music Therapy, 48(4), 509–531.

42.

Soshensky

(2001). Music therapy and addiction. Music Therapy Perspectives, 19(1), 45–52. doi: 10.1093/mtp/19.1.45

43.

Stacy

A. W.

Widaman

K. F.

Marlatt

G. A.

(1990). Expectancy models of alcohol use. Journal of Personality and Social Psychology, 58(5), 918–928. doi: 10.1037/0022–3514.58.5.918

44.

Thoma

M. V.

Zemp

Kreienbühl

Hofer

Schmidlin

P. R.

Attin

… Nater

U. M.

(2014). Effects of music listening on pre-treatment anxiety and stress levels in a dental hygiene recall population. International Journal of Behavioral Medicine. Advance online publication. doi: 10.1007/s12529–014–9439-x

45.

Van den Tol

A. J. M.

Edwards

. (2013). Exploring a rationale for choosing to listen to sad music when feeling sad. Psychology of Music, 4(4), 440–465. doi: 10.1177/0305735611430433

46.

Weiss

R. D.

Griffin

M. L.

Mazurick

Berkman

Gastfriend

D. R.

Frank

A., …

Moras

(2003). The relationship between cocaine craving, psychosocial treatment, and subsequent cocaine use. American Journal of Psychiatry, 160(7), 1320–1325.

47.

Witvliet

C. V.

Vrana

(1996). The emotional impact of instrumental music on affect ratings, facial EMG, autonomic response, and the startle reflex: Effects of valence and arousal [Supplement]. Psychophysiology, 33, S91.

48.

Zentner

Grandjean

Scherer

K. R.

(2008). Emotions evoked by the sound of music: Characterization, classification, and measurement. Emotion, 8(4), 494. doi: 10.1037/1528–3542.8.4.494