Imagery is both a mental process (as in imagining) and a wide variety of procedures used in therapy to encourage changes in attitudes, behavior, or physiological reactions. As a mental process, it is often defined as "any thought representing a sensory quality" (Horowitz, 1983). It includes, as well as the visual, all the senses-aural, tactile, olfactory, proprioceptive, and kinesthetic. Imagery is often used synonymously with visualization; this use is misleading, because the latter refers only to seeing something in the mind's eye, whereas imagery can mean imagining through any sense, as through hearing or smell.

Imagery is a common ingredient in many behavioral therapies not specifically labeled imagery. Since it often involves directed concentration, it can also be thought of as a form of meditation (see the "Meditation" section). Imagery can be taught either individually or in groups, and the therapist often uses it to affect a particular result, such as quitting smoking or bolstering the immune system to attack cancer cells.

Practices that have a component of imagery are almost ubiquitous. They include, among many others, biofeedback, desensitization and counterconditioning, psychosynthesis, neurolinguistic programming, gestalt therapy, rational emotive therapy, and hypnosis (see the "Hypnosis" section). Any therapy that relies on imagery or fantasy to motivate, communicate, solve problems, or evoke heightened awareness and sensitivity could be described as a form of imagery. Forms of meditation that involve repeating a sound or mantra (e.g., TM) or focusing attention on an object that has no concurrent external referent (such as a whale in the ocean) could also be developed as aspects of imagery. Likewise, relaxation techniques that involve instruction (e.g., 'Your hands are heavy"), such as autogenic training, have an imagery component.

Whether imagery differs from hypnosis in terms of purpose and state of consciousness is currently debated. Hypnotherapists, particularly those who train clients in methods of selfhypnosis, are often indistinguishable from practitioners of imagery. What has been agreed on is that there is a correlation between the ability to image and the capacity to enter into an altered state of consciousness, including the hypnotic state (Barber, 1984; Hilgard, 1974; Lynn and Rhue, 1987).

Numerous studies indicate that mental imagery can bring about significant physiological and biochemical changes. These findings, which have encouraged the development of imagery as a health care tool, include its capacity to affect the following: oxygen supply in tissues (Olness and Conroy, 1985); cardiovascular changes (Barber, 1969); vascular or thermal change (Green and Green, 1977); the pupil and the cochlear reflex (Luria, 1968); heart rate and galvanic skin response (Jordan and Lenington, 1979); salivation (Barber et al., 1964; White, 1978); gastrointestinal activity (Barber, 1978); increase in breast size (Barber, 1984); the Mantoux reaction (Black et al., 1963); and blood glucose levels (Stevens, 1983). Several hundred studies using biofeedback, which Green and Green (1977) refer to as an "imagery trainer," expand the list considerably, running the gamut from effects on the firing of single motorneurons (Basmajian, 1963) to brain wave alterations (Brown, 1977).

Some of these findings are from well-controlled studies, but the vast majority represent reports of single cases or small studies that have not been replicated. Nevertheless, the overriding conclusion is that there is a relationship between imagery of bodily change and actual bodily change. Without question, imagery calls for further and more precise investigation.

Clinical applications. Procedures for imagery fall into at least three major categories: (1) evaluation or diagnostic imagery, (2) mental rehearsal, and (3) therapeutic intervention.

Techniques used in evaluation or diagnostic imagery involve asking the person to describe his or her condition in sensory terms. The therapist gathers information regarding the disease, the effect of treatment, and any natural inner healing resources the person might be sensing. The patient is asked, literally, "How do you feel?" In psychotherapy settings, dreams or fantasies might be used in this way, as a means to gaining insight or control over a situation.

Evaluation imagery is usually done early in a therapy session and serves as a format for designing both mental rehearsal and therapeutic intervention strategies. It also is an indicator of the person's understanding of the mechanisms of health and disease and provides opportunity for patient education.

Mental rehearsal is an imagery technique used before medical techniques, usually in an attempt to relieve anxiety, pain, and side effects, which are exacerbated by heightened emotional reactions. Surgery or a difficult treatment is rehearsed before the event so that the patient is prepared and is rid of any unrealistic fantasies.

Typically, a relaxation strategy is taught, then the treatment and recovery period are described in sensory terms as the patient is taken on a guided imagery "trip." Care is taken to be factual without using emotion-laden or fear-provoking words, and the medical procedure is reframed in a positive way whenever possible. The patient is taught coping techniques such as distraction, mental dissociation, muscle relaxation, and abdominal breathing.

Published results with mental rehearsals (or sensory education) are almost uniformly positive and often dramatic. Effects include reduced pain and anxiety; decreased length of hospital stay; the use of fewer pain medicines, barbiturates, tranquilizers, and other medications; and reduced treatment side effects. Mental rehearsal is a cornerstone of certain natural childbirth practices. It has also been tested in burn debridement (Kenner and Achterberg, 1983) and as a preparation for spinal surgery (Lawlis et al., 1985), cholecystectomy, pelvic examination, cast removal, and endoscopy (Johnson et al., 1978). In each of these instances, rehearsal through imagery has been found to diminish pain and discomfort and to reduce side effects.

Imagery as a therapeutic intervention is based on the idea that the images have either a direct or an indirect effect on health. Therefore, either the patients are shown how to use their own flow of images about the healing process or, alternatively, they are guided through a series of images that are intended to soothe and distract them, reduce any sympathetic nervous system arousal, or generally enhance their relaxation. The practitioner may also use "end state" types of imagery, having patients imaging themselves in a state of perfect health, well-being, or successfully achieved goals.

A major and serious criticism of imagery literature (as well as hypnosis literature) is that clinic protocols are seldom provided. Therefore, it is impossible to know what type of therapeutic strategy was used, and of course it cannot be replicated. Some practitioners even regard their protocols as trade secrets and refuse to divulge them.

Whether imagery is merely an antidote to feelings of helplessness or whether the image itself has the capacity to induce the desired physical effect is still unclear. Existing research suggests both conclusions are justified, depending on the situation in question.

Imagery has been successfully tested as a strategy for alleviating nausea and vomiting associated with chemotherapy in cancer patients (Frank, 1985; Scott et al., 1986), to relieve stress (Donovan, 1980), and to facilitate weight gain in cancer patients (Dixon, 1984). It has been successfully used and tested for pain control in a variety of settings; as adjunctive therapy for several diseases, including diabetes (Stevens, 1983); and with geriatric patients to enhance immunity (Kiecolt-Glaser et al., 1985).

Imagery is usually combined with other behavioral approaches. It is best known in the treatment of cancer as a means to help patients mobilize their immune systems (Borysenko, 1987; Siegel, 1986; Simonton et al., 1978), but it also is used as part of a multidisciplinary approach to cardiac rehabilitation (Ornish, 1990; Ornish et al., 1983) and in many settings that specialize in treating chronic pain.

In a survey of alternative techniques used by cancer patients (Cassileth et al., 1984), imagery was cited as the fourth most frequently used. And 46 percent of the respondents listed "self" as practitioner, indicating that imagery is often used as a self-help tool.

Imagery assessment tools. The measurement of imagery as a mental process is fraught with the same problems faced in measuring any other so-called hypothetical construct, including learning, motivation, and perception. So far, psychology has risen to the occasion and developed reliable and meaningful measurement strategies.

A number of instruments with varying purposes, degrees of validity, and reliability are currently in use for measuring imagery. Sheikh and Jordan (1983) have reviewed the imagery test used for psychological diagnosis. Imagery of cancer, diabetes, and spinal pain have been specifically analyzed by Achterberg and Lawlis, using a protocol to elicit sensory information on healing mechanisms, treatment, and the disease itself (Achterberg and Lawlis, 1984). These tests have been found to be accurate predictors of treatment outcome in a number of clinics and rehabilitation facilities.

Research accomplishments. Recent studies suggest a direct impact or correlation between imagery (both as a mental process and a set of procedures) and immunology. These findings include the following:

• Correlations between various types of leukocytes and components of cancer patients' images of their disease, treatment, and immune system (Achterberg and Lawlis, 1984).
• Increased phagocytic activity following biofeedback-assisted relaxation (Peavey et al., 1985).
• Enhanced natural killer cell function following a relaxation and imagery training procedure with geriatric patients (Kiecolt-Glaser et al., 1985) and in adult cancer patients with metastatic disease (Gruber et al., 1988).
• Changes in lymphocyte reactivity following hypnotic procedures (Hall, 1982-83) and instruction in relaxation and imagery in adult cancer patients with metastatic disease.
• Altered neutrophil adherence or margination, as well as white blood cell count, following an imagery procedure (Schneider et al., 1983).
• Increased secretary immunoglobulin A (IgA) (significantly higher than control group) following training in location, activity, and morphology of IgA and 6 weeks of daily imaging.
• The specificity of imagery training was suggested by a study on training patients in cell-specific imagery of either T lymphocytes or I neutrophils. The effects of training, which were assessed after 6 weeks, were statistically associated with the type of imagery procedure employed (Achterberg and Rider, 1989).

Research issues. Although this early research is very promising, further investigations are badly needed. Longitudinal studies are virtually nonexistent. Consequently, the major question remains: Will the physiological-biochemical changes noted in imagery studies have an ultimate impact on health or on the course of the disease?

Distinguishing clinical from statistical significance is critical. Relying on statistical significance alone may obscure much valuable information, such as the few outstanding cases in which the methods were remarkably successful.

For complex clinical research, innovative research paradigms and statistical treatments are needed. Traditional research methodology is based on the idea of a univariate, linear model, which is rare (if not completely absent) in the real world. The spirit of discovery is not served by clinging to models that obscure much of the richness of the human condition. Furthermore, there are a number of complex variables that need to be accounted for in developing a research design. The following are examples:

• The randomized control group design is often impossible, impractical, and unnecessary. Its general efficacy and the ethics of its application are now being seriously challenged (Rider et al., 1990). Other designs should be considered.
• Participant and therapist-researcher motivation and belief are critical and significant variables to consider in this type of behavioral research and should serve as factors in group selection and measurement.
• Studies should be designed to maximize the possibility of good outcome on health and well-being.
• Research into the relationship between imagery and biological parameters-particularly those related to immunology-is hindered by the state of the art in that area. For instance, normative data are often absent, and reliability of assay procedures is questionable. Clinical significance of any changes may or may not be known. The specific impact of diet, season, environment, age, mood, or even the time of day on many of the immune assays is not well studied.

Research needs and opportunities. Existing data suggest at least two major research directions:

1. The impact of imagery as part of a multimodal treatment with conditions such as cancer, AIDS, or autoimmune disorders. The research should include repeat immunologic testing and follow-up. Specific studies could be embedded within the overall design; for example, studies on the effect of imagery specifically designed to enhance medical treatment, the relationship between imagery and outcome of disease, types of patients who respond to imagery, and so on.

2. Replication and expansion of earlier intriguing-but small or poorly controlled-studies that indicated a direct effect of imagery on biologic function.