What is Mental Imagery?
Our conscious cognition can occur in mental imagery or verbal-linguistic formats. Mental imagery refers to perceptual experiences based on information stored in memory, in the absence of external sensory input,Reference Kosslyn, Ganis and Thompson 1 commonly referred to as seeing in the “mind’s eye,” hearing in the “mind’s ear,” etc.Reference Kosslyn, Ganis and Thompson 1 Mental imagery can be multi-sensory and occur in a variety of forms, from fleeting sensory impressions and static scenes (mental pictures) to detailed re-play or pre-play of events situated in a specific space, time, and context (mental movies). Mental imagery is involved in many everyday cognitive functionsReference Kosslyn, Seger, Pani and Hillger 2 such as spatial navigationReference Byrne, Becker and Burgess 3 and language comprehension.Reference Dils and Boroditsky 4 Mental imagery also plays a special role in representing our past and future experiencesReference Holmes and Mathews 5 , Reference Kosslyn 6 that is believed to facilitate adaptive decision-making, planning, and self-regulatory behavior.Reference Taylor, Pham and Rivkin 7 , Reference Seligman, Railton, Baumeister and Sripada 8 Importantly, the ability for mental imagery to emulate real-life experienceReference Moulton and Kosslyn 9 can powerfully impact emotionReference Holmes and Mathews 5 , Reference Di Simplicio, McInerney, Goodwin, Attenburrow and Holmes 10 and motivation.Reference May, Kavanagh and Andrade 11
While mental imagery dysfunction linked to neurocognitive impairment are implicated in mental disorders such as the dementiasReference Irish and Piolino 12 and schizophrenia,Reference Brébion, Ohlsen, Pilowsky and David 13 anomalies in the occurrence of mental imagery-based representations of salient stimuli is also prevalent across mental disorders and contribute to emotional, motivational, and behavioral dysregulation. Through examples of mood and anxiety disorders, self-harm and suicidality, and addiction, the present review aims to characterize the conceptual and clinical significance of mental imagery-based cognition as a tool for assessing psychopathology, and as a potential treatment target and a source of treatment innovation.
Approaches to Measuring Mental Imagery
When practitioners ask patients “what was running through your mind?” at times of reported distress, the assumption is that the content of cognition would assist in identifying factors that may be contributing to the distress. The present review contends that, in addition to the content of cognition, the representation format of such content (imagery or verbal) holds additional information about the impact of cognition. In other words, inviting patients to report on whether their cognition involved mental imagery may provide additional insight into the nature of psychopathology.Reference Di Simplicio, McInerney, Goodwin, Attenburrow and Holmes 14 Here, we move away from the standard clinical interview to possible research tools, too, some of which may also be useful in clinical practice.
Reporting naturally occurring mental imagery
Mental imagery can be consciously experienced and reported to another person, subject to the individual’s capacity to translate such experience into words or ratings and their willingness to communicate it. Clinically, people can report on their imagery in an assessment once given a definition of what is meant by mental imagery (eg, for intrusive memories in post-traumatic stress disorder [PTSD]). Retrospective reporting of mental imagery can introduce memory biases.Reference Kahneman 15 Thus, it may be desirable to minimize these effects in retrospective reports (eg, focusing on salient imagery or recent imagery rather than usual or average ratings), particularly for research purposes. Momentary assessments are now possible using mobile technologies, which can prompt the recording of naturally occurring cognition. If problematic emotion or behavior are more likely during certain times or occasions, prompts to record imagery can be scheduled accordingly. As such recording still relies on the person’s adherence, random sampling or assessment over selected periods may minimize the task burden. Initial tests of the feasibility of such methods are already underway in depression and anxiety research, with promising results.Reference Boukhechba, Gong and Kowsari 16 , Reference Chow, Fua and Huang 17
Questionnaire measures of mental imagery
Standardized measures of mental imagery ability tend to ask participants to generate imagery of everyday scenarios and, then, immediately report the phenomenology of such imagery, in either one sensory modality (Vividness of Visual Imagery Questionnaire)Reference Marks 18 or multiple sensory modalities (The Plymouth Sensory Imagery Questionnaire).Reference Andrade, May, Deeprose, Baugh and Ganis 19 Alternatively, they may ask respondents to state how frequently they experience imagery during specific daily situations (Spontaneous Use of Imagery Scale).Reference Reisberg, Pearson and Kosslyn 20 Such instruments attempt to assess generalized imagery ability, rather than assessing idiosyncratic imagery that may be rich in emotional content and personal relevance, the type that is of particular relevance to psychopathology. In some domains, psychometric measures have been developed to assess recent experiences of specific mental imagery, such as the Craving Experience QuestionnaireReference May, Andrade and Kavanagh 21 and the Motivation Thought Frequency and State Motivation Scales,Reference Kavanagh, Robinson, Connolly, Connor, Andrade and May 22 which assess the frequency and intensity of imagery associated with craving or motivation.
Laboratory assessments of mental imagery
Within the laboratory, mental imagery can be evoked and assessed in real time under controlled task conditions. Early experimental studies sought to understand the nature of mental imagery by devising problem-solving tasks that required the use of mental imagery, broken down into distinct components, from generation, maintenance, and inspection to transformation.Reference Kosslyn 6 Examples of such tasks include mentally visualizing clocks and reporting on the angle between clock hands for various times.Reference Paivio 23 Such tasks have the advantage of offering objective tests of reaction time performance that help to identify the underlying cognitive processes involved.
With the increasing availability of neuroimaging methods such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), researchers have been able to more directly compare brain activation during actual sensory perception and the mental imagery of the same information, such as visually presented faces and houses.Reference O’Craven and Kanwisher 24 However, these tasks are limited in their capacity to investigate mental imagery representations of real-world experiences that have emotional implications for a particular individual. To meet this need, laboratory imagery tasks were developed to assess mental imagery of emotion-provoking situations in everyday life, administered as paper-based instruments (eg, Prospective Imagery TaskReference Holmes, Lang, Moulds and Steele 25 ) or as computer tasks that can be completed in the laboratory or inside a brain scanner.Reference Szpunar and Schacter 26 , Reference Di Simplicio, Alfarano, Ji, Suri, Visser and Holmes 27 Importantly, spontaneous imagery can be assessed under standardized laboratory conditions using adapted mind-wandering paradigms,Reference Ji, Holmes, MacLeod and Murphy 28 which have potential to give further insight into the neurocognitive basis of associative imagery—a form of imagery that has not received sufficient attention to date. For a recent review of mental imagery measures, see Pearson et al., 2013Reference Pearson, Deeprose, Wallace-Hadrill, Burnett Heyes and Holmes 29 .
The Neurocognitive Basis of Mental Imagery
The nature of mental imagery has been the subject of a long-standing debate in psychological science,Reference Galton 30 and accounts of the main theoretical debates have been covered extensively.Reference Pearson and Kosslyn 31 – Reference Pylyshyn 33 Neuroimaging studies have shown that both visual imagery generation and visual perception involves retinotopic activation of the primary visual cortex.Reference Dijkstra, Zeidman, Ondobaka, van Gerven and Friston 34 , Reference Slotnick, Thompson and Kosslyn 35 In addition to the visual modality, shared neural activation between actual sensory perception and mental imagery has been found for the auditory and somatosensory modalities.Reference Ehrsson, Geyer and Naito 36 , Reference Hubbard 37 Further, neuronal signals produced by imagined stimuli can integrate with signals generated by real stimuli of a different sensory modality,Reference Berger and Ehrsson 38 indicating common cross-modal sensory integration processes between perception and imagery. In consequence, mental imagery can function as a form of “weak perception.”Reference Pearson, Naselaris, Holmes and Kosslyn 39 Such neurocognitive findings help to explain why mental imagery can simulate past and future experiences in an as-if-real manner.Reference Ji, Heyes, MacLeod and Holmes 40 This key feature of mental imagery is central to its impact on emotion and motivation, which will be discussed later in this article.
While brain regions involved in actual sensory perception and mental imagery overlap substantially, perception is associated with bottom-up “feed-forward” projections from the primary sensory areas, and imagery generation and maintenance are associated with top-down “feedback” projections from the prefrontal cortex and superior parietal areas involved in cognitive control.Reference Stokes, Thompson, Cusack and Duncan 41 , Reference Mechelli, Price, Friston and Ishai 42 In addition, mental imagery also recruits neural regions involved in the retrieval of information from long-term memory stores and the maintenance or manipulation of such information in working memory.Reference Schacter, Addis and Buckner 43 , Reference Baddeley and Andrade 44 A corollary is that neural damage to neural regions involved in cognitive control and memory will impair mental imagery functioning.
Impairments in General Mental Imagery Function
Congenital impairments in imagery generation
While most individuals can experience mental imagery, reports have emerged in recent years of healthy adults discovering for the first time that, unlike others, they cannot experience mental imagery. Such individuals exhibit severe deficiencies in the ability to remember the past as perceptual experiencesReference Palombo, Alain, Söderlund, Khuu and Levine 45 or to construct perceptual simulations of hypothetical future events.Reference Zeman, Dewar and Della Sala 46 Researchers coined new terms to describe such conditions, namely “aphantasia,”Reference Zeman, Dewar and Della Sala 46 and severely deficient autobiographical memory (SDAM).Reference Palombo, Alain, Söderlund, Khuu and Levine 45 Recent psychophysical evidence indicates that such individuals fail to show mental imagery-induced binocular rivalry, suggesting that the observed deficiencies in mental imagery reflect an absence of mental imagery, rather than alternative possibilities such as impaired metacognitive awareness of intact mental imagery.Reference Keogh and Pearson 47 Interestingly, such individuals report no major functional impairments in daily life and can perform highly on learning and memory tasks using compensatory strategies involving semantic memoryReference Palombo, Alain, Söderlund, Khuu and Levine 45 or spatial representation, which may be unimpaired despite the absence of visual images.Reference Keogh and Pearson 48 Neurocognitive research into aphantasia is at an early stage, and its structural or neuro-functional bases remains to be identified.
Acquired impairments in imagery generation due to neurological disease: the dementias
Unlike congenital aphantasia, acquired damage to the capacity to generate mental imagery has been observed in neurocognitive disorders. For example, patients with dementia can experience impaired ability to engage in mental imagery simulations of hypothetical future experiences. Dementia reflects a degeneration of neural regions that support the ability to remember past experiences as perceptually and experientially coherent events (episodic memory), including the medial prefrontal and parietal cortices, and temporal regions.Reference Schacter, Addis and Buckner 43 , Reference Buckner, Snyder and Shannon 49 – Reference Szpunar, Watson and McDermott 53
The capacity to imagine perceptually and experientially coherent hypothetical future experiences (episodic prospection) relies on the ability to extract and recombine relevant contextual and sensory features from memory.Reference Addis and Schacter 54 In consequence, both episodic memory and episodic simulation are impaired in dementia.Reference Irish and Piolino 12 For example, when asked to recall or imagine an event in as much detail as possible in a three-minute time window, people with Alzheimer’s disease (AD) generated fewer event-related visuospatial and contextual perceptual details and experiential details (emotions and thoughts) relative to age-matched healthy controls.Reference Addis, Sacchetti, Ally, Budson and Schacter 55 – Reference Irish, Hodges and Piguet 58
Impaired capacity to imagine future events appears to be ubiquitous across dementia subtypes.Reference Irish and Piolino 12 For example, patients with semantic dementia, a subtype of frontotemporal dementia that compromises memory of facts and knowledge rather than memory for events (episodic memory), also exhibit impaired ability to imagine hypothetical future events,Reference Irish and Piguet 59 , presumably because semantic information aids the recall of associated imagery elements and supports the construction of coherent episodic imagery. An impaired ability to imagine future events can cause significant functional impairments to daily life, including impaired decision-making because of insensitivity to future consequencesReference Bechara, Tranel and Damasio 60 and a loss of temporally coherent self-identity,Reference Addis, Sacchetti, Ally, Budson and Schacter 55 , Reference Irish, Addis, Hodges and Piguet 57 which has been linked to suicidality in patients with semantic dementia.Reference Hsiao, Kaiser, Fong and Mendez 61 However, the impact of impaired ability to imagine the future on emotional dysregulation and self-regulation in populations with dementia is still not well understood.Reference Irish and Piolino 12
Imagery anomalies associated with schizotypy
There has been significant research interest in the potential relationship between imagery and hallucinations in schizophrenia.Reference Seal, Aleman and McGuire 62 People who score higher on schizotypy questionnaires report greater levels of imagery use in daily lifeReference Aynsworth, Nemat, Collerton, Smailes and Dudley 63 and greater levels of reliving when asked to retrieve mental imagery of past events.Reference Winfield and Kamboj 64 Furthermore, people with higher schizotypy scores also report a greater tendency to experience imagery-based intrusions after a stressor.Reference Holmes and Steel 65 , Reference Steel, Fowler and Holmes 66 Additionally, people who experience hallucinationsReference Barrett 67 or are at high genetic risk for schizophrenia report greater ease of imagery generation and higher vividness of imagery generated in the laboratory.Reference Crespi, Leach, Dinsdale, Mokkonen and Hurd 68 However, other studies have either found no schizotypy-related differences in the vividness of imagery when assessed in the laboratoryReference Aynsworth, Nemat, Collerton, Smailes and Dudley 63 or have only observed such effects for imagery of specific sensory modalitiesReference Winfield and Kamboj 64 or emotional tone.Reference Aynsworth, Nemat, Collerton, Smailes and Dudley 63
In tasks in which pictures and written words are presented for encoding, followed by a surprise test of recognition memory, people who experience visual hallucinations were more likely to misremember verbally presented information as having been visually presented.Reference Aynsworth, Nemat, Collerton, Smailes and Dudley 63 This error has been observed in patients with schizophrenia who experience visual hallucinations, relative to patients with schizophrenia who do not experience visual hallucinationsReference Brébion, Ohlsen, Pilowsky and David 13 , Reference Brébion, Ohlsen, Bressan and David 69 or healthy individuals.Reference Brébion, Ohlsen, Pilowsky and David 13 Such results have been attributed to the presence of vivid visual imagery representations of the information presented, leading to impaired encoding or retrieval of the source information. An argument has been made that confusion between internal and external stimuli may also occur during perception, contributing to the experience of hallucinations in both schizophrenia Reference Bentall, Baker and Havers 70 and other conditions such as Parkinson’s disease (PD).Reference Barnes, Boubert, Harris, Lee and David 71 While this hypothesis is plausible, direct evidence for the confusion between mental imagery and actual sensory perception is still required.
Heightened imagery strength appears unlikely to be a contributor to sensory hallucinations. Consistent with the presence of deficits in executive function in schizophrenia,Reference Keefe and Harvey 72 the retrieval and generation of imagery are often impaired.Reference D’Argembeau, Raffard, Van and der Linden 73 , Reference Steel, Wykes, Ruddle, Smith, Shah and Holmes 74 Studies using objective measures of mental imagery ability (such as mental rotation) have found little relationship between auditory hallucinations and imagery vividness in schizophrenia.Reference Seal, Aleman and McGuire 62 Importantly, relative to healthy controls, the higher vividness of deliberately generated mental imagery appears to be unrelated to hallucination frequency in patients with schizophrenia,Reference Sack, van de Ven, Etschenberg, Schatz and Linden 75 , Reference Oertel, Rotarska-Jagiela and van de Ven 76 relatives of these patients, and people with high schizotypy.Reference Oertel, Rotarska-Jagiela and van de Ven 76 However, it may be premature to say that the issue is completely resolved. For example, the heightened vividness of deliberately generated imagery may not relate to hallucination frequency, but the heightened vividness of spontaneously generated imagery may contribute to reality-monitoring errors. Future studies could investigate schizotypy-linked individual differences in the vividness of spontaneously generated imagery that is assessed under standardized laboratory conditions (eg, using adapted mind-wandering paradigmsReference Ji, Holmes, MacLeod and Murphy 28 ).
Mood and anxiety disorders
Impaired general imagery generation ability in anxious individualsReference Morrison, Amir and Taylor 77 and reduced imagery generation and manipulation ability in depression have been demonstrated.Reference Zarrinpar, Deldin and Kosslyn 78 A more recent comprehensive assessment revealed no deficits in the ability to generate, manipulate, or recall images in people with bipolar disorder, major depression, or anxiety disorders relative to healthy controls matched on socio-demographic factors.Reference Di Simplicio, Renner and Blackwell 79 More isolated imagery generation deficits have been reported in anxiety disorders, concerning the ability to generate event-related details of future simulations.Reference Wu, Szpunar, Godovich, Schacter and Hoffman 80 However, mental imagery-linked dysfunction in mood and anxiety disorders appears to be primarily linked to its emotional amplification effects when the content of cognition is emotion laden, which is discussed in more detail in the following section.
The Impact of Mental Imagery on Emotion and Motivation
In addition to impairments in the general capacity to generate mental imagery or to discriminate it from actual sensory perception, anomalies in imagery-based emotion-laden cognition are of importance to mental disorders because of its impact on emotion and motivation.
Mental imagery can evoke strong emotional responses
Many experimental studies have demonstrated that mental imagery can evoke emotional responses at subjective and physiological levels.Reference Ji, Heyes, MacLeod and Holmes 40 For example, in healthy people, imagining fear and anger-related situations leads to heart rate accelerationReference Vrana and Lang 81 , Reference Witvliet and Vrana 82 ; increased skin conductanceReference Lang, Levin, Miller and Kozak 83 , Reference Weerts and Lang 84 and potentiation of the startle blink reflexReference Vrana and Lang 81 , Reference Cuthbert, Lang, Strauss, Drobes, Patrick and Bradley 85 relative to imagining emotionally neutral stimuli.
At the neural level, fMRI studies on the visual imagery of emotional stimuli have found patterns of activation akin to those found during visual perception of the same stimuli. For example, the nucleus accumbens (NAcc) and medial prefrontal cortex (mPFC) are selectively activated during both visual perceptionReference Sabatinelli, Bradley, Lang, Costa and Versace 86 and visual imagery of emotionally positive scenes.Reference Costa, Lang, Sabatinelli, Versace and Bradley 87 The amygdala can also be activated during both visual perceptionReference Kim, Kim and Kim 88 and visual imageryReference Costa, Lang, Sabatinelli, Versace and Bradley 87 , Reference Kim, Kim and Kim 88 of emotionally positive and negative scenes or faces. Finally, fMRI studies using real-time neural activation feedback show that, over successive trials, participants were able to use idiosyncratic mental imagery of positive and negative scenarios to regulate activation of emotion processing regions such as the insula.Reference Lee, Ruiz, Caria, Veit, Birbaumer and Sitaram 89
The functional impact of imagery on emotion
To investigate whether imagery plays a functional role in eliciting emotional responses, researchers have interfered with imagery generation and tested whether such interference leads to reductions in imagery vividness and emotional impact. Typically, studies have employed a concurrent task to reduce the availability of the visuospatial working memory resources required for visual mental imagery generation. For example, Andrade and colleagues had participants generate mental imagery in response to emotional cues, and those who engaged in simultaneous finger tapping of a spatial pattern reported lower vividness of mental imagery than those who did not perform a concurrent task.Reference Andrade, Kavanagh and Baddeley 90 Importantly, the reduction in imagery vividness accompanied a corresponding decline in the intensity of the emotional response to the negative imagery. This result was replicated using concurrent eye movements, instead of finger tapping, during repeated emotional memory imagery generation.Reference Kavanagh, Freese, Andrade and May 91 Other researchers have likewise shown that similar types of concurrent task interference can reduce the emotional impact of imagery representing feared future events in healthy and clinical samples.Reference Engelhard, van den Hout, Janssen and van der Beek 92 – Reference Lilley, Andrade, Turpin, Sabin-Farrell and Holmes 94
Researchers have also tested the impact of visuospatial interference on the frequency of negative intrusive mental imagery. There is growing evidence that playing the visuospatial computer game, Tetris,Reference Lau-Zhu, Holmes, Butterfield and Holmes 95 following exposure to analog trauma film clips depicting traumatic scenes reduces the frequency of subsequent involuntary memory imagery of the trauma film.Reference James, Lau-Zhu, Clark, Visser, Hagenaars and Holmes 96 For example, in an early study, healthy people either played Tetris or not after viewing the traumatic film and reported intrusive imagery of the film clip content in a diary over the following week.Reference Holmes, James, Coode-Bate and Deeprose 97 Those who played Tetris after the film reported experiencing fewer imagery intrusions and distress than those who had not.Reference Holmes, James, Coode-Bate and Deeprose 97 Recent proof of concept studies moving from the lab to the clinic show that a visuospatial interference task, relative to a control task, led to reductions in the frequency of negative intrusive mental imagery when administered soon after a road traffic accidentReference Iyadurai, Blackwell and Meiser-Stedman 98 or traumatic childbirth.Reference Horsch, Vial and Favrod 99
Mental imagery as an emotional amplifier: a comparison with verbal-linguistic cognition
The abovementioned findings showed that mental imagery can evoke emotional responses may be because of the emotional content of imagined stimuli, rather than because of the imagery representation of such stimuli. Several studies have sought to address this question by comparing the relative emotional impact of imagery-based and verbal-linguistic representations of the same information. In one study, healthy participants were instructed to either generate imagery or focus on the semantic meaning of initially ambiguous auditory scenarios that ended negatively (eg, “You are at work when you hear the fire alarms go off. You run to the exit to discover that it is . . . for real”).Reference Holmes and Mathews 5 Participants in the imagery condition reported greater increases in self-reported anxiety than the verbal condition.Reference Holmes and Mathews 5 This finding has since been replicated by subsequent studies that found imagery representations to have evoked greater emotional impact than did verbal-linguistic representations of negative informationReference Holmes, Lang and Shah 100 – Reference Nelis, Holmes, Palmieri, Bellelli and Raes 102 and positive information.Reference Holmes, Lang and Shah 100 , Reference Holmes, Coughtrey and Connor 103 Researchers have, therefore, proposed that mental imagery can amplify the impact of cognition on emotion and exacerbate mood instability in both negative and positive directions, such as in bipolar disorder.Reference Holmes, Geddes, Colom and Goodwin 104
The impact of emotional, mental imagery on motivation
Functional theories of emotion postulate that emotions serve to motivate approach and avoidance behavior in response to appetitive and aversive cues, where cues can be internal cognition or stimuli in the environment.Reference Carver 105 , Reference Lang and Bradley 106 For appetitive or aversive stimuli, the moment-to-moment experience of being motivated to approach or avoid such stimuli involves emotionally charged cognition.Reference Kavanagh, Andrade and May 107 Mental imagery representations of appetitive stimuli can evoke intense desires or craving for, and motivation to acquire, such stimuli.
For example, higher levels of vividness of food imagery are associated with more intense craving for the imagined food item,Reference Tiggemann and Kemps 108 and such effects are amplified in dieters.Reference Harvey, Kemps and Tiggemann 109 The vividness of imagery concerning a wide range of other substances and activities are also positively related to the intensity of desires to consume the imagined substances or to engage in the imagined activities.Reference May, Andrade, Kavanagh and Penfound 110 Similarly, in cigarette smokers, instructions to generate smoking-related imagery is reliably associated with significant increases in self-reported craving,Reference Maude-Griffin and Tiffany 111 , Reference Tiffany and Drobes 112 comparable to those from in-vivo exposure to cigarettes.Reference Drobes and Tiffany 113 However, direct comparisons of mental imagery and smoking cues show stronger effects from lighting a cigarette than from smoking-related imagery,Reference Heishman, Lee, Taylor and Singleton 114 consistent with the idea that mental imagery functions as a weak form of actual sensory perception.Reference Pearson, Naselaris, Holmes and Kosslyn 39
Physiologically, evoking mental imagery of food has been shown to modulate the gustatory salivary reflex (salivary pH levels) in a similar way to having participants process the sensory qualities of real food.Reference Vanhaudenhuyse, Bruno, Brédart, Plenevaux and Laureys 115 Furthermore, repetitive mental imagery concerning food consumption has been shown to produce satiation.Reference Morewedge, Huh and Vosgerau 116
Neuroimaging evidence also corroborates the role of mental imagery in addiction-related processes. For example, in individuals with cocaine addiction, neural activation during mental imagery of personalized drug use, as compared with anger-related mental imagery, is associated with greater activation in the limbic, paralimbic, and striatal brain regions, which are involved in reward learning, wanting, and anticipation.Reference Kilts, Schweitzer and Quinn 117
In the natural environment, craving-related imagery is typically evoked by associated cues in the form of internal cognition or environmental stimuli. The anticipated pleasure or relief provided by imagery of the desired substance leads to the further elaboration of such imagery,Reference May, Kavanagh and Andrade 11 , Reference Kavanagh, Andrade and May 107 which may reduce the level of cognitive resources available for performing other cognitive tasks, particularly those requiring visuospatial working memory.Reference Tiggemann, Kemps and Parnell 118
The functional impact of imagery on motivation
A range of concurrent visuospatial tasks has been used to reduce craving by disrupting imagery of desired stimuli, including lateral eye movements,Reference Littel, van den Hout and Engelhard 119 the elicitation of competing mental imagery,Reference Schumacher, Kemps and Tiggemann 120 playing Tetris,Reference Skorka-Brown, Andrade, Whalley and May 121 and modeling clay.Reference Andrade, Pears, May and Kavanagh 122 Reductions in craving have also been observed from the use of more passive tasks such as viewing changing visual patterns on a screen such as “dynamic visual noise”Reference Kemps and Tiggemann 123 , Reference May, Andrade, Panabokke and Kavanagh 124 or scanning bodily sensations.Reference Hamilton, Fawson, May, Andrade and Kavanagh 125 Such results provide evidence that imagery representations of desired stimuli have a causal impact on approach motivation for such stimuli.
Contexts in which concurrent task interference do not reduce craving are also illuminating. For example, focusing on abdominal sensations,Reference May, Andrade, Batey, Berry and Kavanagh 126 inhaling a sweet scent during imagery of desired food items,Reference Firmin, Gillette, Hobbs and Wu 127 or completing a concurrent verbal working memory task Reference Andrade, Pears, May and Kavanagh 122 do not reduce craving. Successful disruption of cravings induced by imagining the desired stimuli appears to require tasks that tax the same sensory domains as those involved in cravings induced by actual encounters with the stimuli. In the case of coffeeReference Kemps and Tiggemann 128 and cigarettes,Reference Versland and Rosenberg 129 both concurrent visual and olfactory imagery are effective at reducing craving. However, concurrent auditory imagery does not interfere with cravings for foodReference Kemps and Tiggemann 130 or coffeeReference Kemps and Tiggemann 128 that is consistent with observations that sound is not a prominent feature in desires for these targets.Reference May, Andrade, Kavanagh and Penfound 110
In summary, evidence suggests that mental imagery-based cognitions are involved in emotional and motivational processing, and comparative data with verbal-linguistic cognition suggest that imagery has a greater impact on emotional and motivational outcomes. Interfering with imagery of desired stimuli weakens its emotional and motivational impact, providing persuasive evidence of a causal relationship between mental imagery and its emotional and motivational response.
Emotional, Mental imagery-linked Dysfunction in Psychopathology
Given the capacity of mental imagery to impact emotion and motivation, anomalies in the occurrence of emotional mental imagery may contribute to psychopathology in a transdiagnostic manner. This section reviews current evidence linking anomalies in the occurrence of mental imagery to emotional, motivational, and behavioral dysfunction through the example of several mental disorders.
Anxiety disorders
Elevated anxiety appears to be linked to elevated levels of vividness of negative future imageryReference Morina, Deeprose, Pusowski, Schmid and Holmes 131 , Reference Pile and Lau 132 and a greater ease of generating negative future imagery relative to non-negative future imagery.Reference Wu, Szpunar, Godovich, Schacter and Hoffman 80 In a sample of people with mood and anxiety disorders, higher anxiety levels corresponded with higher levels of negative future imagery vividness and higher sensations of “as-if-real” pre-experiencing of the imagined negative future event.Reference Di Simplicio, Renner and Blackwell 79
Unwanted mental imagery that comes to mind unbidden is prevalent in anxiety disorders.Reference Hirsch and Holmes 133 For example, intrusions of trauma-related scenes in PTSD,Reference Association 134 typically triggered by sensory stimuli associated with the trauma, are a core symptom that causes significant distress, anxiety, and avoidance behaviors that impair daily functioning.Reference Ehlers, Hackmann and Michael 135 , Reference Grey and Holmes 136 Similarly, mental imagery of feared contamination scenes in obsessive-compulsive disorder (OCD) is highly prevalent and is associated with distress, anxiety, and compulsive washing.Reference Coughtrey, Shafran, Lee and Rachman 137 In social anxiety, individuals often experience mental imagery of themselves committing a social gaffe or appearing anxious to observers, with such imagery eliciting subjective anxiety and distress.Reference Hirsch, Clark, Mathews and Williams 138
Generalized anxiety disorder (GAD) is characterized by repetitive self-referential negative verbal styles of thinking, such as worry.Reference Borkovec, Alcaine and Behar 139 Researchers have proposed that people with GAD worry in verbal form to avoid the highly emotional and physiologically arousing negative imagery of feared outcomes.Reference Stöber 140 Experimental evidence provides support for this account, as inducing repeated verbal thinking of hypothetical future events leads increased abstractness of thought content and reduced imagery over time.Reference Behar, DiMarco, Hekler, Mohlman and Staples 141 However, while repetitive abstract verbal thinking about feared outcomes lowers fear and anxiety initially, it is followed by increases in anxiety that coincide with the occurrence of imagery.Reference Behar, DiMarco, Hekler, Mohlman and Staples 141
Another way in which mental imagery may contribute to dysfunctional avoidance behaviors that sustain anxiety, such as in phobias, is through aversive conditioning. Dadds and colleagues investigated individual differences in imagery ability and the tendency to report specific aversions such as phobias of certain foods.Reference Dadds, Hawes, Schaefer and Vaka 142 Results indicated that higher imagery ability was associated with higher levels of aversion, even after controlling for potential confounders such as proneness to disgust.Reference Dadds, Hawes, Schaefer and Vaka 142 Presumably, a greater ability to engage in the imagery of food consumption may strengthen the learned associations between the food stimuli and disgust,Reference Dadds, Bovbjerg, Redd and Cutmore 143 contributing to avoidance of the food item and, thereby, maintaining the phobia.
In summary, emotional, mental imagery could maintain anxiety either directly through vivid imagery of aversive scenarios or intrusive traumatic imagery or indirectly through avoidance of aversive images that sustains avoidance behaviors, maintaining anxiety.
Mood disorders: major depressive disorder
Depression has been associated with the elevated occurrence of emotionally distressing mental imagery and reduced occurrence of positive mental imagery.Reference Holmes, Blackwell, Burnett Heyes, Renner and Raes 144 For example, in community samples, depressive symptoms are related to higher frequencies of negative mental imagery and less frequent experiences of positive imagery.Reference Weßlau, Cloos, Höfling and Steil 145 When instructed to imagine negative and positive personal future events within different time periods, individuals with major depression are found to generate fewer positive, but not negative, future events relative to healthy controls.Reference MacLeod and Conway 146 – Reference MacLeod, Tata, Tyrer, Schmidt, Davidson and Thompson 148
Investigators have also examined depression-linked biases in phenomenological aspects such as imagery vividness. When instructed to generate imagery in response to verbal sentences describing positive and negative future situations, people dysphoria,Reference Holmes, Lang, Moulds and Steele 25 , Reference Szőllősi, Pajkossy and Racsmány 149 major depression,Reference Kemps and Tiggemann 130 and major depression with acquired brain injury (ABI)Reference Murphy, Veers, Blackwell, Holmes and Manly 150 report lower vividness for imagery of positive relative to negative future events, as compared with healthy controls. In patients with ABI, a reduced capacity to anticipate positive relative to negative future events and reduced imagery vividness of positive relative to negative future events, as compared with healthy controls, was uniquely found in patients with ABI and depression, but not patients with ABI without major depression.Reference Murphy, Veers, Blackwell, Holmes and Manly 150 Further, reduced positive imagery vividness appears to be uniquely associated with depression, not anxiety, in adolescence.Reference Pile and Lau 132
In addition to anomalies in deliberately generated emotional imagery, patients with depression tend to report frequent experiences involving unwanted mental imagery in the form of spontaneously occurring emotionally aversive mental imagery of past and future experiences.Reference Patel, Brewin, Wheatley, Wells, Fisher and Myers 151 Recent laboratory research has shown that higher scores on the Beck Depression Inventory-II (BDI-II) are associated with lower tendencies to spontaneously imagine positive relative to negative aspects of the future during mind-wandering.Reference Ji, Holmes, MacLeod and Murphy 28 Such anomalies in mental imagery–based cognition are postulated to maintain the depression-linked elevation of negative emotional states and attenuation of positive emotional states.Reference Holmes, Blackwell, Burnett Heyes, Renner and Raes 144
At present, there is little research linking impoverished positive mental imagery to anhedonia (loss of interest and motivation in rewarding experiences), a core symptom of depression. However, results from a treatment study suggested that, for people with clinical levels of depression, four weeks of positive mental imagery training resulted in a faster and greater recovery of re-engagement in daily activities than in a non-imagery active control condition,Reference Renner, Ji, Pictet, Holmes and Blackwell 152 indicating that positive mental imagery–based manipulations may enhance behavioral activation in depression.
Mood disorders: bipolar disorder
Patients with bipolar disorder also have highly vivid negative future imagery, and vividness is higher in those with a more unstable course of the disorder.Reference Hales, Deeprose, Goodwin and Holmes 153 Unlike major depression, bipolar disorder is associated with the presence of vivid positive imagery at times of high positive affect.Reference Ivins, Di Simplicio, Close, Goodwin and Holmes 154 Hypomania is associated with intrusive positive imagery about future eventsReference Gregory, Brewin, Mansell and Donaldson 155 and a greater tendency to use imagery in daily life.Reference Meyer, Finucane and Jordan 156 In a laboratory study, hypomanic traits predicted greater emotional response to positive mental imagery in adolescents.Reference O’Donnell, Di Simplicio, Brown, Holmes and Burnett Heyes 157 The bipolar phenotype appears to be associated with greater vividness and emotional response to both negative and positive mental imagery, which is postulated to exacerbate mood oscillationsReference Holmes, Coughtrey and Connor 103 and maladaptive approach-related behaviors in bipolar disorder.Reference O’Donnell, Di Simplicio, Brown, Holmes and Burnett Heyes 157 While preliminary evidence suggests that mood lability is associated with greater vividness and self-involvement of mental imagery across mood and anxiety disorders,Reference Di Simplicio, Renner and Blackwell 79 further experimental research is required to investigate possible downstream impacts on behavior in bipolar disorder.
In summary, anomalies in the strength and occurrence of positive and negative mental imagery can contribute to mood dysregulation and mood instability in a transdiagnostic manner.
Non-suicidal self-injury and suicidal behavior
Mental imagery of self-harm has been linked to non-suicidal self-injury (NSSI). One study surveyed NSSI cognition and behavior in college students and found the occurrence of NSSI mental imagery of past NSSI acts frequently preceded NSSI behavior, particularly while the urge to self-harm is strong.Reference Hasking, Di Simplicio, McEvoy and Rees 158 Specific characteristics of self-harm imagery may relate to acting or not acting on the self-harm urges, depending on whether the imagery was appraised as aversive or comforting.Reference Hasking, Di Simplicio, McEvoy and Rees 158 These findings indicate that emotion-laden mental imagery of self-harm may play a role in motivating approach behaviors toward, or avoidance behaviors from, self-harm.Reference McEvoy, Hayes, Hasking and Rees 159
In addition to NSSI, compelling intrusions of imagined suicidal acts (suicide “flash-forwards”)Reference Holmes, Crane, Fennell and Williams 160 are often reported by people with major depressionReference Holmes, Crane, Fennell and Williams 160 , Reference Crane, Shah, Barnhofer and Holmes 161 and individuals with bipolar disorder.Reference Hales, Deeprose, Goodwin and Holmes 153 Consistent with the Interpersonal Theory of Suicide,Reference Van Orden, Witte, Cukrowicz, Braithwaite, Selby and Joiner 162 graphic mental imagery of suicide may facilitate suicidal behavior by contributing to the process of habituating to the painful and fearful aspects of death by suicide and by facilitating planning, rehearsal, and action readiness.Reference Crane, Shah, Barnhofer and Holmes 161 Accordingly, cross-sectional research in college samples has reported associations between frequent and vivid suicide-related mental imagery and higher levels of suicidal intentReference Selby, Anestis, Joiner, Selby, Anestis and Joiner 163 or past suicidal behavior.Reference Holaday and Brausch 164 Importantly, in a recent longitudinal study in a Chinese in-patient sample, the occurrence of suicidal “flash-forwards” significantly predicted suicidal intent at both baseline and 7 weeks,Reference Ng, Di Simplicio, McManus, Kennerley and Holmes 165 and suicidal imagery was among the factors differentiating suicidal intent from suicidal attempts in a representative sample of young adults.Reference Wetherall, Cleare and Eschle 166
Substance-related and addictive disorders
As previously noted, more vivid imagery is associated with more intense craving—a diagnostic feature of addictions (DSM-V).Reference Association 134 During treatment for alcohol abuse or dependence, a study found that 81% of participants reported experiencing imagery associated with craving.Reference Kavanagh, May and Andrade 167 Importantly, this mental imagery was not just visual in modality, as taste imagery was also prominent, providing a unique contribution to the prediction of craving strength.Reference Kavanagh, May and Andrade 167
Mental imagery can also amplify cravings elicited by other factors such as physiological withdrawal. For example, smoking urges in both cigarette-deprived and non-deprived groups are accentuated by experimentally inducing mental imagery of cigarettes.Reference Drobes and Tiffany 113 However, if the person is attempting to abstain from acquisition or consumption of the desired stimuli, imagery-induced craving also elicits anxiety.Reference Fox, Bergquist, Hong and Sinha 168
The relationship between the imagery of appetitive stimuli and approach behavior toward such stimuli is complex. While simple actions such as lighting a cigarette, or more complex habitual chains of actions, may become automated and directly activated by environmental cues without conscious cognition,Reference Tiffany 169 most cases of substance use involve some degree of conscious planning or context-specific modification of behavior (eg, finding a tobacco outlet).Reference Kavanagh, Andrade and May 107 Conscious planning and behavioral modification are more likely to occur when craving intensity is high, that is, in turn, associated with more vivid and frequent imagery of the desired stimuli.Reference May, Kavanagh and Andrade 11
Of course, the effects of desires on behavior are constrained by additional factors such as stimuli availability, self-control, and the strength of countervailing goals desires.Reference Kavanagh, Andrade and May 107 Despite such constraints, the frequency of alcohol imagery accounted for 16% of the variance in monthly alcohol consumption in young adults.Reference Connor, Kavanagh and Andrade 170 The frequency of craving imagery also predicts relapse and treatment dropout in inpatients being treated for alcohol dependence.Reference Coates, Gullo and Feeney 171 Therefore, the occurrence of craving imagery holds important implications for high-risk approach behaviors in addiction.
Future Directions
Evidence of the link between mental imagery-based cognition and emotional, motivational, and behavioral dysfunction across several categories of mental disorders strongly suggests that clinical assessment should focus on the content and format of cognition.
One important reason for assessing both the content of cognition and the format of cognition is that frequent and vivid mental imagery may be a key factor behind the progression from ideation to action. The clinical implications of not assessing imagery have been highlighted in the field of suicidal behavior, in which missed identification of imagery of suicidal action means neglecting an index of progression to a potential attempt.Reference Wetherall, Cleare and Eschle 166 Future research should further investigate the link between mental imagery and emotionally charged risky behaviors such as NSSI, mania, aggression, and addiction.
A second reason to assess mental imagery is that we need to obtain a better understanding of the neurocognitive bases of imagery representations of salient stimuli that contribute to psychopathology. Initial advances have been made in exploring the basis of trauma flashbacksReference Clark, Holmes, Woolrich and Mackay 172 and also of elevated emotional imagery in bipolar disorder.Reference Di Simplicio, Alfarano, Ji, Suri, Visser and Holmes 27 Future research should continue to focus on delineating the precise neural substrates and cognitive processes involved in emotionally charged mental imagery and how such processes are disrupted in psychopathology.
While this review has focused on mental imagery representations that play a role in emotional and motivational dysfunction, mental imagery can also be a powerful lever toward wellbeing and adaptive behavior.Reference Holmes, Arntz and Smucker 173 There has been recent growing interest in developing imagery-based interventions that could train individuals to use positive imagery to improve mood,Reference Blackwell, Browning and Mathews 174 and the use of soothing imagery is already part of cognitive therapy approaches for depression, trauma, and personality disorders.Reference Wilson, Schwannauer, McLaughlin, Ashworth and Chan 175 , Reference Naismith, Mwale and Feigenbaum 176 Research on positive imagery has also demonstrated the importance of fostering motivation for functional goals such as weight reduction.Reference Solbrig, Whalley and Kavanagh 177 The emotional and motivational power of mental imagery suggests that these examples will be precursors of other, highly innovative applications of mental imagery to treatment.
Conclusions
Mental imagery–based cognition is of significant conceptual and clinical relevance to psychiatry and allied disciplines. Mental imagery is core to our capacity to re-experience past events and pre-experience future ones, affording it the power to impact emotion, motivation, and behavior. Through the examples of mood and anxiety disorders, self-injury, and addiction, this review aimed to characterize mental imagery–linked dysfunction as extending beyond the contexts of neurocognitive impairment. When problematic cognition occurs in mental imagery form, its impact on emotion, motivation, and behavior is amplified, which serves to maintain and exacerbate psychopathology and dysfunction. Therefore, mental imagery is an important target for clinical assessment across mental disorders, and a deeper understanding of mental imagery-linked dysfunction has the potential to drive the development of algorithms to aid treatment decision-making and inform transdiagnostic treatment innovation.
Disclosure
Julie Ji was supported by the Forrest Research Foundation. Emily Holmes was supported by the Karolinska Institutet and the Swedish Research Council (VR). Colin MacLeod, David Kavanagh, and Martina Di Simplicio have nothing to disclose.
Open access
