Advertisement

Phantom Limb Pain

A Review of the Literature on Attributes and Potential Mechanisms

      Abstract

      This study presents a review of the literature on the attributes and potential mechanisms involved in phantom limb pain, encompassing studies describing pain in the residual limb, phantom sensation and phantom limb pain, and the difficulties that may arise when making these distinctions. A variety of theories have been proposed to explain causal mechanisms for phantom limb pain. Conceptually, research into phantom limb pain is informed by the particular theory of chronic pain that is dominant at the time the research is undertaken. For example, early physiological theories on the etiology of phantom limb pain were grounded in specificity or pattern theories of pain. Later physiological research was based on the framework provided by Gate Control Theory and focused on identifying peripheral, spinal, and central neural mechanisms. Psychological explanations were grounded in psychoanalytic or personality theories of chronic pain which propose that phantom limb pain results from pre-amputation psychological disturbance. Despite numerous studies examining phantom limb pain, much of this research has both conceptual and methodological shortcomings. As such, the application of these research findings to clinical practice has limited utility.

      Keywords

      Introduction

      The presence of a phantom limb is now seen as a natural consequence of amputation. This has not always been the case. Throughout history, phantom limbs have fascinated researchers and prompted numerous investigations in an attempt to define phantom phenomena and to understand why they occur. The presence of a phantom limb is seldom distressing for an amputee, particularly when they are informed about its likely presence prior to amputation.
      • Sherman R.A
      Stump and phantom limb pain.
      In fact, many amputees welcome a phantom limb as it allows them to use a prosthesis naturally. However, in addition to phantom sensation, a great many amputees suffer from phantom limb pain.
      Many authors have found it useful to distinguish between pain in the residual limb, phantom sensation, and phantom limb pain. Pain in the residual limb is defined as pain at the site of an extremity amputation. Phantom sensation is defined as any sensation in the absent limb except pain. Phantom limb pain is defined as painful sensations referred to the absent limb.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      • Ribbers G
      • Mulder T
      • Rijken R
      The phantom phenomenon a critical review.
      Although these are defined as discrete categories, there are many common features across these definitions. Given this lack of distinction, examination of the literature on phantom sensation and pain in the residual limb is also necessary to explore the phenomenon of phantom limb pain.
      This paper outlines the research that focuses on several specific issues. First, it defines phantom limb phenomena using the classification that distinguishes pain in the residual limb, phantom sensation, and phantom limb pain. Second, it examines the measurement of phantom limb pain and related phenomena. Third, it examines the literature on proposed causal mechanisms involved in phantom limb pain. The treatment of phantom limb pain is beyond the scope of this review.

      Phantom Phenomena

      Pain in the Residual Limb

      Pain in the residual limb is defined as pain at the site of an extremity amputation. Pain in the residual limb, not surprisingly, is particularly common in the early post-amputation phase. Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      noted that 50% of his nonselected sample of 46 new amputees reported pain in the residual limb in the first few weeks post-amputation, but that this declined to 13% at 13-month follow-up. Similarly, Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      note that 57% of their sample of 58 amputees reported pain in the residual limb immediately post-amputation. This fell to 21% at the 2-year follow-up.
      Pain in the residual limb also can persist beyond the stage of post-surgery healing. In common with many reports of immediate post-surgical pain, persistent pain in the residual limb is usually described as a “stabbing,” “shocking,” or “burning,” and is reported to occur at the lower end of the stump, close to the scar.
      • Browder J
      • Gallagher J.P
      Dorsal cordotomy for painful phantom limb.
      However, in contrast to immediate post-surgical pain, pain in the residual limb continues for many years after the surgical incision has healed and can occur in the absence of stimulation, or alternatively, in response to light stimulation of the stump.
      Occasionally, the intensity of pain experienced in the stump far exceeds any stimulation. This type of pain in the residual limb is described as a “nerve storm” and is characterized by spontaneous movement, cold surface temperature, sweating, and reduced blood flow to the stump.
      • Sunderland S
      Sliosberg
      • Silosberg A
      observed this condition in 50% of his sample of 251 amputees. Movements ranging from slight, hardly visible jerks, to severe contractions were noted in those who reported pain in the residual limb.
      Examination of stump frequently reveals pathology that may be related to the pain.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Sherman R.A
      • Sherman C.J
      A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
      This includes skin pathology, infection, bone spurs, and neuroma.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      There is also an association between stump pathology and increased levels of phantom limb pain.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Morgenstern F.S
      The effects of sensory input and concentration of post-amputation phantom limb pain.
      Moreover, when phantom limb pain and pain in the residual limb are both present, they are usually similar in nature and co-vary in intensity.
      • Sherman R.A
      However, it is important to note that both phantom limb and pain in the residual limb are present in many amputees who have no obvious stump pathology.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      • Sherman R.A
      • Sherman C.J
      A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
      There are several problems with the literature describing pain in the residual limb. Firstly, few studies make a distinction between stump and phantom pain. Of those that do, details are not given of how pain in the residual limb differs from phantom limb pain and of how these are measured. Secondly, studies that do examine the relationship between stump pathology and pain in the residual limb are generally retrospective. Because the development of pathology in relation to reported pain in the residual limb is not examined in these studies, any relationship found may be spurious.

      Phantom Sensation

      Prevalence

      As noted earlier, phantom sensation is defined as any sensation of the missing limb except pain and is experienced by virtually everyone who undergoes limb amputation.
      • Melzack R
      Phantom limbs.
      Ambrose Paré first reported this phenomenon in the literature as early as the 16th century. However, it was not until the late 19th century that a detailed description was published and the concept of “phantom limb” became part of the medical literature.
      • Mitchell S.W
      Phantom limbs.
      Once a body part has been excised, either by trauma or surgery, the feeling persists that the body part is still present.
      • Simmel M.L
      On phantom pain.
      In addition, phantom sensation is reported in the absence of amputation, for example, in patients with sensory loss due to spinal cord injury where normal sensation is absent.
      • Bors E
      Phantom limbs of patients with spinal cord injury.
      • Burke D.C
      • Woodward J.M
      Phantom movement and phantom feeling in complete paraplegic patients.
      • Conomy J.P
      Disorders of body image after spinal cord injury.
      • Davis R
      Pain and suffering following spinal cord injury.
      • Frisbie J.H
      • Aguilera E.J
      Chronic pain after spinal cord injury an expedient diagnostic approach.
      • Melzack R
      • Loeser J.D
      Phantom body pain in paraplegics evidence for a central “pattern generating mechanism” for pain.
      • Paeslack V
      • Spahn B
      • Sommer K
      Pain symptoms in paraplegic patients.
      Moreover, phantom limbs have been induced in experimental situations using an anesthetic block of an intact limb.
      • Melzack R
      • Bromage P.R
      Experimental phantom limbs.
      • Wall P.D
      • Nathan P.W
      • Noordenbos W
      Ongoing activity in peripheral nerves.
      For many years, it was thought that children born with limb deficiencies, that is, without all or part of a limb, did not experience a phantom.
      • Friedmann L.W
      • Kolb L
      • Simmel M.L
      The conditions of occurrence of phantom limbs.
      Given that the accepted basis of phantom sensation at that time was that prolonged sensory input from a limb is required for the formation of a cortical representation of the limb,
      • Simmel M.L
      The conditions of occurrence of phantom limbs.
      it was proposed that those with congenital limb absence would not experience a phantom. However, several studies have found that this is not the case. In an early study, Weinstein and Sersen
      • Weinstein S
      • Sersen E.A
      Phantoms in cases of congenital absence of limbs.
      described five children with congenital absence who experienced phantoms of a limb they had never had. In a follow-up study, Weinstein et al.
      • Weinstein S
      • Sersen E.A
      • Vetter R.J
      Phantoms and somatic sensation in cases of congenital aplasia.
      reported another 18 cases in which phantoms were experienced in congenitally absent limbs. More recent research has confirmed these findings.
      • Lacroix R
      • Melzack R
      • Smith D
      • Mitchell N
      Multiple phantom limbs in a child.
      • Saadah E.S.M
      • Melzack R
      Phantom limb experiences in congenital limb-deficient adults.

      Localization

      Immediately following amputation, the phantom limb resembles the pre-amputation limb in shape, length, and volume. In addition, it is reported to move in space and time in much the same way as the pre-amputation limb did.
      • Melzack R
      • Wall P.D
      The reality of a phantom limb is such that many amputees report trying to use the limb.
      • Friedmann L.W
      Despite the reality of a phantom limb, it does not remain a complete reproduction of the pre-amputation limb.

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      • Livingston W.K
      Over time, the proximal part of the phantom often fades. The remaining phantom is comprised of the distal portion of the limb, usually those parts that have the greatest representation in the somatosensory cortex. For example, in those with an upper limb amputation, the thumb is experienced more vividly than the remaining fingers and the balls of the fingers have greater clarity than the remaining part of the hand. Distal parts of the limb are felt more prominently than proximal parts. A similar distribution of phantom sensation is found in lower limb amputees.
      The length of the phantom may also change over time. In approximately one-third of amputees, a process of “telescoping” occurs where the phantom is gradually felt to approach the residual limb. Gradually, the distal part of the phantom may become attached to the residual limb or indeed be experienced within the residual limb.

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      • Haber W.B
      Effects of loss of limb on sensory functions.
      • Weiss S.A
      • Fishman S
      Extended and telescoped phantom limbs in unilateral amputees.
      Cronholm

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      reported that, while the phantom may shorten or telescope, it never extends beyond the length of the pre-amputation limb. A telescoped phantom usually extends to the length and volume of the pre-amputation limb under certain circumstances, such as when the amputee wears a prosthesis. Furthermore, there is some evidence to suggest that only phantom sensation (not phantom pain) is experienced in a telescoped phantom limb and that the phantom extends beyond the stump to assume normal proportions when the individual experiences an episode of phantom pain.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Bors E
      Phantom limbs of patients with spinal cord injury.
      Melzack and Wall
      • Melzack R
      • Wall P.D
      proposed that changes in the “receptive fields” of dorsal horn neurons account for some aspects of phantom sensation. Expanded receptive fields of neurons in the spinal cord have been observed in a number of animal studies.
      • Melzack R
      • Wall P.D
      They propose that if a stimulus is applied close to the end of the residual limb, not only will it excite the cells in the dorsal horn relating to the precise area stimulated, but the stimulus will also activate nearby neurons with receptive fields that include the foot. Katz
      • Katz J
      Psychophysiological contributions to phantom limbs.
      suggested that the changes in receptive fields may also explain the phenomenon of telescoping. The perceived length and size of the phantom limb may be a “perceptual marker” of the extent to which cutaneous input from the stump and surrounding tissue has occupied the area of somatosensory cortex previously utilized by the amputated limb.
      Evidence of a link with cortical reorganization is demonstrated in the differences in sensory acuity found in stumps of amputees with either telescoped or extended phantom limbs.
      • Haber W.B
      Effects of loss of limb on sensory functions.
      In upper limb amputees, point localization is enhanced in those whose limb is experienced in a telescoped position. In addition, light touch and two-point discrimination thresholds also show greater stump sensitivity in those with telescoped as opposed to extended phantom limbs.
      • Haber W.B
      Effects of loss of limb on sensory functions.
      Katz
      • Katz J
      Psychophysiological contributions to phantom limbs.
      suggested that these findings support the hypothesis that the distal region of the stump takes over the tactile and sensory functions of the amputated limb.
      Further evidence of cortical reorganization following amputation comes from experimental studies with monkeys. In a microelectrode study, the area of somatosensory cortex previously occupied by a digit was shown to be taken over by cutaneous input from the stump and surrounding tissue following amputation.
      • Merzenich M.M
      • Nelson R.J
      • Stryker M.P
      Somatosensory cortical map changes following digit amputation in adult monkeys.
      Cells identified in the area of sensory cortex which originally had receptive fields that included the amputated digit responded, after amputation, to input of the adjacent digits, the palm of the hand, and the amputation stump. These observations are consistent with studies comparing the sensory acuity of the stump and the intact, contralateral limb in human amputees. For example, lower thresholds have been observed at the stump for light touch, two-point discrimination and point localization after amputation of an upper limb.
      • Haber W.B
      Effects of loss of limb on sensory functions.
      There is also some evidence to suggest that the plasticity of the somatosensory cortex is use-dependent, rather than a function of amputation. A microelectrode study of the area of somatosensory cortex occupied by a digit of a nonamputated monkey showed that the receptive field changed following prolonged non-noxious movement of the finger.
      • Jenkins W.M
      • Merzenich M.M
      • Ochs M.T
      • Allard T
      • Guic-Robles E
      Functional reorganization of primary somatosensory cortex in adult owl monkeys after behaviorally controlled tactile stimulation.
      These changes were indistinguishable from the changes in receptive fields following amputation.
      • Merzenich M.M
      • Nelson R.J
      • Stryker M.P
      Somatosensory cortical map changes following digit amputation in adult monkeys.
      This use-dependent aspect of cortical reorganization is not incompatible with the information available on telescoped phantom limbs. For example, Kallio
      • Kallio K.E
      Phantom limb of forearm stump cleft by kineplastic surgery.
      reported that telescoping is use-dependent in upper limb amputations. A number of below-elbow stumps were surgically cleaved so that, rather than use a prosthesis, the two branches of the residual limb could be used as pincers by the amputee. After extensive training, the branches of the stump could be opened and closed at a rate that approximated that of normal fingers. When patients were followed 2 to 6 years later, 36% reported a phantom in which some fingers had fused together and others had disappeared to accommodate the shape of the stump. Furthermore, the phantom arm had telescoped only to the extent that it shortened to fit the amputation stump. Movement of the branches of the residual limb resulted in similar movement being felt in the phantom hand.
      Beyond this, recent research by Flor and her colleagues
      • Flor H
      • Elbert T
      • Knecht S
      • Wienbruch C
      Phantom limb pain as a perceptual correlate of cortical reorganization following arm amputation.
      suggests that the plasticity of the somatosensory cortex is related to phantom limb pain rather than nonpainful phantom phenomena. A strong relationship was found between cortical reorganization (measured by a noninvasive neuroimaging technique) and the magnitude of phantom limb pain following arm amputation. No such relationship was found in relation to phantom sensation. Clearly, as technology improves, further research may be undertaken to tease out the link between various phantom phenomena and cortical reorganization.

      Quality

      The predominant description of phantom sensation is that of a mild tingling or tightness which is often described as “pins and needles.”
      • Friedmann L.W
      • Haber W.B
      Effects of loss of limb on sensory functions.
      Other qualities noted are touch, temperature, pressure, and itch.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      In addition, most amputees have a sense of the position, length, and volume of the phantom. Jensen and Rasmussen2 noted that the position of a phantom limb may be relaxed, fixed, or distorted. In upper limb amputees, the phantom arm is often experienced as hanging loosely at the side of the body, moving freely when the person walks. However, in other cases, the phantom arm is fixed, bent at the elbow and the fist is held clenched towards the chest.
      • Henderson W.R
      • Smyth G.E
      Phantom limbs.
      These features are also reported by lower limb amputees.

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      In some cases, a phantom is distorted in such a way that the position of the limb is the one that most closely resembles the immediate pre-amputation position.
      • Mitchell S.W
      Phantom limbs.
      • Riddoch G
      Phantom limbs and body shape.
      This most commonly occurs when a limb is traumatically lost and the limb was distorted by the accident. Distorted phantoms are also reported in cases of paraplegia following accidents in which the phantom limb is perceived to be in a different position from the “intact” limb.
      • Berger M
      • Gerstenbrand F
      Phantom illusions in spinal cord lesions.
      Although there is general agreement on the nature of phantom sensation, there are a number of problems in evaluating the literature within the definition of phantom sensation as “any sensation of the missing limb except pain.”
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      • Ribbers G
      • Mulder T
      • Rijken R
      The phantom phenomenon a critical review.
      The primary obstacle is that it is frequently confounded with phantom limb pain or pain in the residual limb. In part, this occurs because subjects are frequently drawn from populations seeking treatment for phantom limb pain or pain in the residual limb. Although these studies discuss phantom sensation, they seldom measure or distinguish it from other phantom phenomena. For example, if the subject has phantom limb pain, what aspect of their experience is defined as nonpainful? Moreover, does phantom sensation in those who experience phantom limb pain and/or pain in the residual limb differ from phantom sensation in those who do not? In addition, although recent research proposes that phantom sensation may be a feature of cortical reorganization following amputation, these findings are also difficult to evaluate. For example, because the distinction between phantom sensation and other phantom phenomena is not explicit in these studies, it is not clear whether cortical reorganization occurs only in those with phantom sensation or whether phantom limb pain or pain in the residual limb is also part of this process.

      Phantom Pain

      Prevalence

      In contrast with the literature on phantom sensation, there is little agreement concerning the prevalence of phantom limb pain among the amputee population. Substantial literature suggests that phantom limb pain is rare.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Kolb L
      • Henderson W.R
      • Smyth G.E
      Phantom limbs.
      • Abramson A.S
      • Feibel A
      The phantom phenomena its use and misuse.
      • Ewalt J.R
      • Randall G.C
      • Morris H
      The phantom limb.
      • Hermann L.G
      • Gibbs E.W
      Phantom limb pain.
      • Sternbach R.A
      However, many other studies indicate that 60–80% of amputees experience phantom limb pain.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Buchannan D.C
      • Mandel A.R
      The prevalence of phantom limb experience in amputees.
      • Krebs B
      • Jensen T.S
      • Kroner Nielsen J
      • Jorgensen H.S
      Phantom limb phenomena in amputees 7 years after limb amputation.
      • Solonen K.A
      The phantom phenomenon in Finnish war veterans.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      This lack of agreement has occurred, in part, because prevalence rates for phantom limb pain have been derived from research studies in which the patient’s request for treatment is the only indication of their pain status.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Kolb L
      • Henderson W.R
      • Smyth G.E
      Phantom limbs.
      • Abramson A.S
      • Feibel A
      The phantom phenomena its use and misuse.
      • Ewalt J.R
      • Randall G.C
      • Morris H
      The phantom limb.
      • Hermann L.G
      • Gibbs E.W
      Phantom limb pain.
      • Sternbach R.A
      When these figures are compared to those acquired from amputees who do not seek treatment for their pain, prevalence rates of phantom limb pain are estimated to be low. Recent studies indicate that a high percentage (54–85%) of amputees not seeking treatment for their pain also report significant levels of phantom limb pain.
      • Sherman R.A
      • Sherman C.J
      A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      • Sherman R.A
      • Sherman C.J
      Prevalence and characteristics of chronic phantom limb pain among American veterans.
      Sherman et al.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      have suggested that the difference in prevalence rates reported in many studies may also be a function of the amputee’s reluctance to report phantom limb pain to health care providers. Sherman and his colleagues reported that 69% of the 2694 veteran amputees responding to a detailed survey of phantom limb phenomena told the researchers that their physicians had directly stated, or clearly implied, that the pain was “just in their heads.” The great majority of amputees responding to this survey were afraid to tell their physicians that they had phantom pain for fear that the physician would think them insane or that they would jeopardize the relationship with their physician. This relationship is particularly important in terms of obtaining treatment for stump problems at a stage when verbal report is frequently the only evidence that such problems exist. Some support for this tendency to under report pain is found in a study by Kolb
      • Kolb L
      who noted that 0.5% of amputees in a large clinic spontaneously reported phantom pain, but 5% acknowledged it when asked.
      Sternbach et al.’s
      • Sternbach T
      • Nadvrona H
      • Arazi D
      A five year follow-up study of phantom limb pain in post-traumatic amputees.
      analysis of the literature suggests that a 0.5–10% incidence for chronic phantom pain is the accepted norm. However, it is likely that this figure represents substantial under reporting. In support of this, Sherman et al.’s
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      analysis showed that 85% of respondents reported experiencing significant levels of phantom pain. This survey, which was carried out by a team of researchers who had no connection with health care providers and no influence on treatment, had a 61% response rate. If all of the nonrespondents were pain-free, the prevalence of pain in this population would be 51%. These prevalence rates were found in a military population and were replicated in a nonveteran sample.
      • Sherman R.A
      • Sherman C.J
      Prevalence and characteristics of chronic phantom limb pain among American veterans.
      Other researchers have since confirmed similar rates of occurrence in nonmilitary populations.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Buchannan D.C
      • Mandel A.R
      The prevalence of phantom limb experience in amputees.
      • Krebs B
      • Jensen T.S
      • Kroner Nielsen J
      • Jorgensen H.S
      Phantom limb phenomena in amputees 7 years after limb amputation.
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      • Sternbach T
      • Nadvrona H
      • Arazi D
      A five year follow-up study of phantom limb pain in post-traumatic amputees.
      It is likely that the higher incidence rates reported in studies that ask directly about phantom limb pain are more representative of the general amputee population.

      Intensity

      Several studies report that “severe” phantom limb pain occurs in only 0.5 to 5% of all amputees.
      • Melzack R
      • Loeser J.D
      Phantom body pain in paraplegics evidence for a central “pattern generating mechanism” for pain.
      • Henderson W.R
      • Smyth G.E
      Phantom limbs.
      • Ewalt J.R
      • Randall G.C
      • Morris H
      The phantom limb.
      These figures contrast with a study of 2694 amputees, which reported that 51% experienced phantom limb pain “severe” enough to hinder lifestyle on more than 6 days per month.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      Twenty-seven percent of this sample experienced phantom limb pain for more than 15 hours each day and a further 21% reported daily pain over a 10- to 14-hour period. Clearly, differences in the definition of “severe” phantom limb pain account for the differences among these studies.
      The literature that discusses the intensity of phantom limb pain is difficult to evaluate because: (a) they are frequently case studies that do not describe how phantom limb pain is measured;
      • Browder J
      • Gallagher J.P
      Dorsal cordotomy for painful phantom limb.
      • Morgenstern F.S
      The effects of sensory input and concentration of post-amputation phantom limb pain.
      • Almagor M
      • Jaffe Y
      • Lomranz J
      The relation between limb dominance, acceptance of disability, and the phantom limb phenomenon.
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      • Dawson L
      • Arnold P
      Persistent phantom limb pain.
      • Duane L.T.C
      • Howard L
      Group therapy for amputees in a ward setting.
      • Marsland A.R
      • Weekes J.W.N
      • Atkinson R.L
      • Leong M.G
      Phantom limb pain a case for beta blockers.
      • Minichiello W.E
      Treatment of hyperhidrosis of amputation site with hypnosis and suggestions involving classical conditioning.
      • McGrath P.A
      • Hillier L.M
      Phantom limb sensations in adolescents a case study to illustrate the utility of sensation and pain logs in pediatric clinical practice.
      • Siegel E.F
      Control of phantom limb pain by hypnosis.
      • Stannard C.F
      • Porter G.E
      Ketamine hydrochloride in the treatment of phantom limb pain.
      or (b) they describe how phantom limb pain is measured but the sample population comprises only those seeking treatment for their pain.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Kolb L
      • Henderson W.R
      • Smyth G.E
      Phantom limbs.
      • Abramson A.S
      • Feibel A
      The phantom phenomena its use and misuse.
      • Ewalt J.R
      • Randall G.C
      • Morris H
      The phantom limb.
      • Hermann L.G
      • Gibbs E.W
      Phantom limb pain.
      • Sternbach R.A
      One study assesses pain using the McGill Pain Questionnaire and found that phantom limb pain is similar in intensity to chronic low back pain, nonterminal cancer pain, and labor pain.
      • Dubuisson D
      • Melzack R
      Classification of clinical pain descriptions by multiple group discriminant analysis.
      More recent studies that used this measure reported similar intensities.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Katz J
      • Melzack R
      Referred sensations in chronic pain patients.
      • Melzack R
      The McGill Pain Questionnaire major properties and scoring methods.
      However these studies comprised patients selected from those attending pain management programs and the results cannot be generalized to amputees not seeking treatment for pain. Nonetheless, a recent study using nonselected subjects reported similar phantom limb pain intensity in a small sample of Scottish male amputees.
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.

      Localization

      Like phantom sensation, which predominates in distal portion of the phantom over time, phantom limb pain is also primarily localized to the distal part of the missing limb. In upper limb amputees, phantom pain is normally felt in the fingers, palm of the hand, and occasionally the wrist and in lower limb amputees, phantom pain is generally experienced in the toes, ball of the foot, instep, top of the foot and ankle.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Melzack R
      Phantom limbs.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      Given the similarity in location, it is possible that the changes in receptive fields and cortical reorganization observed following limb amputation are related to both phantom limb pain and phantom sensation.

      Duration

      There is a great deal of confusion concerning the duration of phantom limb pain. Many studies suggest that phantom limb pain either diminishes or disappears during the first 2 years post-amputation. For example, Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      found that 84% of amputees experienced phantom pain immediately post-amputation but only 61% continued to experience some phantom pain. Of these cases, only 30% reported that their pain was “moderate to severe,” and 30% reported “mild” phantom pain. Similarly, in a prospective study, Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      found that a small decline in pain prevalence from 72% 8 days post-amputation to 67% at a 6-month follow-up was accompanied by a significant reduction in intensity to 50%. Both these studies were carefully conducted, prospective studies following the course of phantom limb pain over the 2-year period following amputation. In contrast, other research has shown that phantom pain may be present in those who were amputated up to 30 years previously.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      In the study by Sherman and his colleagues, 44% reported that their phantom pain had not diminished over a 30-year period. The latter studies were also carefully conducted but differed from those reported by Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      and Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      in a number of ways. First, the study populations varied. Studies that reported a decline in pain over the first 2 years post-amputation primarily described elderly amputees. The latter studies were conducted on a younger population of amputees.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      Second, the studies differed in the reasons for amputation. In the Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      and Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      studies, the majority of subjects lost their limbs as a result of vascular disease, whereas in the Sherman et al.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      study, the majority of the population underwent traumatic amputation following injury. Only Hill’s
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      study comprised subjects who underwent amputation for a variety of reasons but the numbers in this study were small and the subjects were all male. Finally, differences in the research design of these studies also limit comparisons. Both Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      and Jensen and colleagues
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      conducted prospective studies, collecting data at several time points over a 2-year period, whereas both Hill
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      • Hill A
      • Niven C.A
      • Knussen C
      The role of coping in adjustment to phantom limb pain.
      and Sherman et al.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      employed a retrospective design that collected data at only one time point.

      Quality

      Two of the most common descriptors applied to phantom limb pain are “burning” and “cramping.” Other terms are also used. Phantom limb pain has been described as “numb,” “smarting,” “stinging,” “throbbing,” “piercing,” and “tearing.”
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Sherman R.A
      • Sherman C.J
      A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
      • Morgenstern F.S
      The effects of sensory input and concentration of post-amputation phantom limb pain.
      • Friedmann L.W
      • Livingston W.K
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      • Katz J
      Psychophysical correlates of phantom limb experience.
      • Shukla G.D
      • Sahu C
      • Tripathi R.P
      • Gupta D
      Phantom limbs a phenomenological study.
      Evaluation of the qualities associated with phantom limb pain is difficult because a variety of methods have been used to generate the descriptors. In many cases, the descriptors are spontaneously reported in single case studies. In others, the clinician, who may be familiar with the literature, prompts the patient (i.e., “is it a burning type of pain”?)
      • Friedmann L.W
      Some studies provide a list of descriptors from which the patient selects those that are appropriate.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      • Katz J
      Psychophysical correlates of phantom limb experience.
      • Katz J
      • Melzack R
      Auricular TENS reduces phantom limb pain.
      These studies too, are difficult to evaluate, as the lists vary from study to study. In addition, many of these studies include patients seeking treatment for their pain, and the resulting description may not be applicable to the general amputee population. Finally, phantom limb pain descriptors may change over time. In some of the above studies, descriptors are sought for immediate post-amputation pain,
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      whereas in others, the patient may have been experiencing phantom limb pain for many years.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Katz J
      Psychophysical correlates of phantom limb experience.
      • Katz J
      • Melzack R
      Auricular TENS reduces phantom limb pain.
      Many patients report pain that resembles pre-amputation pain both in quality and location. For example, in a study of “somatosensory pain memories,” Katz and Melzack
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      noted that amputees report pain that has similar qualities and is experienced in the same location as pre-amputation pain from surgical incisions, wounds, bedsores, ingrown toenails, ulcers, arthritis, corns, and calluses. This is graphically illustrated in a case study conducted by Bailey and Moersch
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      that reported on a male patient who had undergone amputation 22 years prior to the study. The patient had an accident that left a painful sliver under his fingernail. One week later, his arm was torn off in a machine accident at work. For 2 years following this accident, the patient experienced pain of the same quality and in the same location as that experienced when he had the sliver under his fingernail. A more recent case study reported an individual who experienced recurrences of pain during dressings of a wound prior to amputation.
      • Hill A
      • Niven C.A
      • Knussen C
      Pain memories in phantom limbs a case study.

      The Role of Patient Characteristics in Phantom Limb Pain

      In general, studies of phantom limb pain presuppose that amputees are a homogeneous group. Therefore, little is known about variation within this population.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      • Ribbers G
      • Mulder T
      • Rijken R
      The phantom phenomenon a critical review.
      Although some research has examined the relationship between patient characteristics (age, gender, duration of pain, reason for amputation, site of amputation, etc.) and levels of phantom limb pain, many yielded mixed results because of differences in sample selection, sample size, and study methods.
      Bailey and Moersch
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      found that the incidence of phantom limb pain is greater in male than female amputees, but other studies did not identify a difference.
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      Similar difficulties can be found in studies investigating the role of age and medical status. For example, Buchannan and Mandel
      • Buchannan D.C
      • Mandel A.R
      The prevalence of phantom limb experience in amputees.
      found that older amputees report the presence of phantom limb pain more often than younger amputees, but Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      found no such difference. Kashani et al.,
      • Kashani J.H
      • Frank R.G
      • Kashani S.R
      • Wonderlich S.A
      • Reid J.C
      Depression among amputees.
      Kegel et al.,
      • Kegel B
      • Carpenter M.L
      • Burgess E.M
      Functional capabilities of lower extremity amputees.
      Parkes,
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      and Parkes and Napier
      • Parkes C.M
      • Napier M.M
      Psychiatric sequelae of amputation.
      all report that the presence of concurrent medical conditions, such as arthritis or diabetes, predict increased phantom limb pain, whereas Morgenstern
      • Morgenstern F.S
      The effects of sensory input and concentration of post-amputation phantom limb pain.
      found that medical status is not related to phantom limb pain. The evaluation of this literature is difficult because phantom limb pain is often inferred rather than measured
      • Buchannan D.C
      • Mandel A.R
      The prevalence of phantom limb experience in amputees.
      • Kashani J.H
      • Frank R.G
      • Kashani S.R
      • Wonderlich S.A
      • Reid J.C
      Depression among amputees.
      • Kegel B
      • Carpenter M.L
      • Burgess E.M
      Functional capabilities of lower extremity amputees.
      and when phantom limb pain is measured, different instruments are used. For example, Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      used an interview method whereas Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      used both interviews and a visual analogue scale (VAS) to measure phantom limb pain.

      Measurement of Phantom Limb Pain

      As discussed in the preceding sections, much of what we know about phantom limb pain has been derived from surveys of treatment outcomes in which the presence of phantom limb pain is inferred from the patient’s request for treatment rather than measured.
      • Browder J
      • Gallagher J.P
      Dorsal cordotomy for painful phantom limb.
      • Morgenstern F.S
      The effects of sensory input and concentration of post-amputation phantom limb pain.
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      • Duane L.T.C
      • Howard L
      Group therapy for amputees in a ward setting.
      • Marsland A.R
      • Weekes J.W.N
      • Atkinson R.L
      • Leong M.G
      Phantom limb pain a case for beta blockers.
      • Minichiello W.E
      Treatment of hyperhidrosis of amputation site with hypnosis and suggestions involving classical conditioning.
      • Siegel E.F
      Control of phantom limb pain by hypnosis.
      • Stannard C.F
      • Porter G.E
      Ketamine hydrochloride in the treatment of phantom limb pain.
      Other surveys of patients requesting treatment have attempted to measure the intensity of phantom limb pain using VAS
      • Almagor M
      • Jaffe Y
      • Lomranz J
      The relation between limb dominance, acceptance of disability, and the phantom limb phenomenon.
      • McGrath P.A
      • Hillier L.M
      Phantom limb sensations in adolescents a case study to illustrate the utility of sensation and pain logs in pediatric clinical practice.
      or verbal descriptor scales.
      • Dawson L
      • Arnold P
      Persistent phantom limb pain.
      Studies designed to describe, rather than treat, phantom limb pain have used a variety of measures which assess both pain intensity and quality.
      • Melzack R
      • Katz J
      Pain measurement in persons in pain.
      Several studies have used the McGill Pain Questionnaire
      • Hill A
      The use of pain coping strategies by patients with phantom limb pain.
      • Dubuisson D
      • Melzack R
      Classification of clinical pain descriptions by multiple group discriminant analysis.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Katz J
      • Melzack R
      Referred sensations in chronic pain patients.
      • Hill A
      • Niven C.A
      • Knussen C
      The role of coping in adjustment to phantom limb pain.
      and others have used structured or semi-structured interviews to quantify various aspects of phantom limb pain.
      • Sherman R.A
      Stump and phantom limb pain.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      • Marshall M
      • Helme S.E
      • Deathe A.B
      A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      • Bach S
      • Noreng M.F
      • Tjéllden N.U
      Phantom limb pain in amputees during the first 12 months following limb amputation, after preoperative lumbar epidural blockade.
      • Parkes C.M
      Components of the reaction to loss of a limb spouse or home.
      These studies have yielded some useful data on the relationship between phantom limb pain and other variables, particularly the initiating and exacerbating factors involved in episodes of phantom limb pain. However, each of these measures has limitations. In addition, the diversity of measures used in the literature does not permit meaningful comparison of research findings on phantom limb pain across studies.
      A number of problems arise when making a distinction between phantom sensation, phantom limb pain, and pain in the residual limb. For instance, it is difficult to see how phantom sensation can be compared across subjects or studies, given that the definition calls for the amputee to decide which sensations are not painful. As discussed earlier, pain perception is multifaceted and therefore directly related to the individual’s unique history. While one amputee might report “tingling” as a sensation, another might define the same word as painful. This makes it impossible to determine whether it is the amputee’s interpretation of the word or some difference in the quality of the sensation that leads to different classifications. Given these problems, it is not surprising that few studies make the distinction between phantom limb pain and phantom sensation explicit.
      Similarly, amputees are not always able to distinguish between stump pain and phantom pain. Although sources of pain in the residual limb are frequently obvious (for example, lesions or scar tissue), this is not always the case.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      There is great variation in reported incidence of pain in the residual limb whereas Cronholm

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      and Abramson and Feibel
      • Abramson A.S
      • Feibel A
      The phantom phenomena its use and misuse.
      reported that 15% of their respective study populations experienced pain in the residual limb, Sliosberg
      • Silosberg A
      noted a prevalence of 27% and Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      reported a prevalence of 50%. Given these findings, it is possible that the disparity in prevalence rates for pain in the residual limb and for phantom limb pain result from the amputee’s inability to distinguish pain in the residual limb from phantom limb pain. It is interesting to note that the prevalence rate of pain in the residual limb in these studies is akin to the prevalence rate of phantom limb pain. Cronholm

      Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

      and Abramson and Feibel
      • Abramson A.S
      • Feibel A
      The phantom phenomena its use and misuse.
      report that a small proportion of their study populations experience pain in the residual limb (15% respectively) and a small proportion also experience phantom limb pain (35% and 2%, respectively). In contrast, Sliosberg
      • Silosberg A
      and Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      report that a large percentage of their study populations have pain in the residual limb (27% and 50%, respectively) and a large percentage also have phantom limb pain (72.5% and 62%, respectively).
      A further difficulty in disengaging stump pain from phantom limb pain comes from studies that have examined the difference in temperature between the stump and contralateral limb. In amputees with phantom limb pain, the temperature of the stump is significantly lower than in the contralateral limb.
      • Livingston W.K
      • Sherman R.A
      Direct evidence of a link between burning phantom pain and stump blood circulation a case report.
      • Sherman R.A
      • Bruno G.M
      Concurrent variation of burning phantom limb and stump pain with near surface blood in the stump.
      Sherman and Bruno
      • Sherman R.A
      • Bruno G.M
      Concurrent variation of burning phantom limb and stump pain with near surface blood in the stump.
      found a consistent inverse relationship between phantom limb pain and the temperature of the residual limb relative to the contralateral limb. This relationship has been demonstrated for throbbing, burning, and tingling descriptions of phantom limb pain, but not for any other descriptions (e.g., tearing, cramping). Using thermography to examine blood flow to the stump, Kristen et al.
      • Kristen H
      • Lukeschitsch G
      • Plattner F
      • Sigmund R
      • Resch P
      Thermography as a means for quantitative assessment of stump and phantom pains.
      recorded a significant difference between stump temperature (28°C) and the temperature of the contralateral limb (31°C). This study found that phantom limb pain is more likely to occur if the distribution of stump temperature is patchy. Similarly, Katz
      • Katz J
      Psychophysical correlates of phantom limb experience.
      found a significant difference in temperature between the stump (30°C) and the same area on the contralateral limb (32°C) in the group with phantom limb pain, and in a group with phantom sensation. No significant difference in temperature was found in the group who reported neither phantom sensation nor phantom limb pain. Given the problems in classifying various aspects of phantom phenomena outlined in preceding sections, the findings of Katz
      • Katz J
      Psychophysical correlates of phantom limb experience.
      showing differences between those reporting phantom sensation, phantom limb pain, and no phantom limb pain should be viewed with some caution. Nonetheless, the research providing concrete evidence of different physiological mechanisms underlying different qualities of phantom limb pain advances our understanding of this phenomenon.

      Limitations of Research in Phantom Limb Pain

      A number of criticisms apply to studies concerned with pain in the residual limb, phantom sensation, and phantom limb pain. First, sample populations frequently comprise military veterans, primarily young men who have lost limbs as a result of trauma.
      • Sherman R.A
      Stump and phantom limb pain.
      • Solonen K.A
      The phantom phenomenon in Finnish war veterans.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      • Sherman R.A
      • Bruno G.M
      Concurrent variation of burning phantom limb and stump pain with near surface blood in the stump.
      • Frank R.G
      • Kashini J.H
      • Kashini S.R
      • Wonderlich S.A
      • Umlaf R.L
      • Ashkanazi G.S
      Psychological response to amputation as a function of age and time since amputation.
      • Gerhards F
      • Florin I
      • Knapp T
      The impact of medical, reeducational, and psychological variables on rehabilitation outcomes in amputees.
      • Sherman R.A
      • Sherman C.J
      • Bruno G.M
      Psychological factors influencing phantom limb pain an analysis of the literature.
      Other studies assess elderly amputees, who customarily have a primary diagnosis of peripheral vascular disease.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      • Bach S
      • Noreng M.F
      • Tjéllden N.U
      Phantom limb pain in amputees during the first 12 months following limb amputation, after preoperative lumbar epidural blockade.
      • Frank R.G
      • Kashini J.H
      • Kashini S.R
      • Wonderlich S.A
      • Umlaf R.L
      • Ashkanazi G.S
      Psychological response to amputation as a function of age and time since amputation.
      • Lindesay J
      Validity of the general health questionnaire in detecting psychiatric disturbance in amputees with phantom limb pain.
      • Williamson G.M
      • Schulz R
      • Bridges M.W
      • Behan A.M
      Social and psychological factors in adjustment to phantom limb pain.
      Given that elderly amputees with vascular disease make up the largest proportion of the overall amputee population, this is not surprising. However, as this particular group has additional problems associated with age and medical conditions, the findings of these studies cannot be generalized to younger amputees. A final group consists of those requesting treatment for phantom limb pain. Again, it is difficult to generalize findings about pain and distress in this patient group to those amputees who are not actively seeking treatment. Nonetheless, research findings from these limited populations are habitually generalized to the amputee population as a whole.
      • Winchell E
      A second criticism is that even in those studies that involve a more general amputee population, sample sizes are too small to draw conclusions about the variables of interest. For example, of the studies in which gender was the variable of interest, few had a sample of more that 50 subjects. Of these, only 5% to 7% were female.
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      • Dawson L
      • Arnold P
      Persistent phantom limb pain.
      • Tebbi C.K
      • Mallon J.C
      Long term psychosocial outcome among cancer amputees in adolescence and early adulthood.
      A final criticism is that studies of phantom phenomena do not distinguish between immediate post-amputation and chronic phantom phenomena. This creates particular problems in that immediate and chronic phantom limb pain will clearly be very different, as are acute and chronic pain. For example, in terms of Gate Control Theory, pain is conceived of as a perceptual experience whose quality and intensity are influenced by the unique history of the individual, by the meaning they give to the pain-producing situation, the meaning they give to the consequences of pain, and by the individual’s state of mind.
      • Melzack R
      • Wall P.D
      Pain mechanisms a new theory.
      In addition to ongoing physiological changes, clearly these features of pain perception will vary tremendously as a function of adjustment to amputation.

      Causal Explanations for Phantom Limb Pain

      Physiological Mechanisms

      A number of physiological mechanisms have been offered to explain the development of phantom phenomena. Early theories of phantom limb pain are grounded in “specificity” theories of pain and explain phantom limb pain entirely in terms of peripheral factors. For example, based solely on the observation that manipulation of the stump influences phantom limb pain, a number of researchers proposed that stimulation of the nerve endings in the amputation stump transmits information to the brain that is interpreted as phantom limb pain.
      • Riddoch G
      Phantom limbs and body shape.
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      However, this formulation does not explain the reason that manipulation of the stump results in the perception of pain rather than sensation, or the reason that the output of this manipulation should be experienced as a phantom rather than at the site of the manipulation.
      Recent research has also highlighted the role of peripheral nerve fibers in the explanation of phantom limb pain. Following amputation, fibers from the cut end of nerves grow into nodules (neuromas) which generate abnormal impulses. These impulses activate central nervous system neurons and may result in the perception of phantom pain (ref. 12, page 123; ref. 94). This hypothesis receives support from examination of stumps, which frequently reveal pathological findings (skin pathology, circulatory disturbances, infection, bone spurs, or neuroma). Phantom pain is reported more frequently by patients with observable stump pathology
      • Sherman R.A
      Stump and phantom limb pain.
      and co-occur in terms of frequency and intensity with pain in the residual limb.
      • Sherman R.A
      • Sherman C.J
      A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
      • Sherman R.A
      • Sherman C.J
      • Parker L
      Chronic phantom and stump pain among American veterans results of a survey.
      Moreover, surgical removal of neuromas sometimes provides relief from phantom limb pain.
      • Jensen T.S
      • Rasmussen P
      Phantom pain and other phenomena after amputation.
      Mechanical, chemical, and electrical irritants applied to the nerve ending in the stump have been shown to exacerbate phantom limb pain, and local anesthesia has been shown to eliminate phantom limb pain.
      • Livingston W.K
      Although these studies show that peripheral factors undoubtedly play a role in phantom limb pain, there is evidence to suggest that they are not the primary eliciting factor. Pain can also occur in the absence of stump pathology2 and surgical revision of the stump, including removal of neuromas, has only limited success in alleviating phantom pain.
      • Sunderland S
      Phantom pain can occur in the absence of nerve damage, such as when a limb is congenitally absent, and when information from the periphery is blocked, such as when there has been a complete transection of the spinal cord.
      • Melzack R
      Phantom limbs.
      • Harwood D.D
      • Hanumanthu S
      • Stoudemire A
      Pathophysiology and management of phantom limb pain.
      The observation that one injection of local anesthetic will occasionally eliminate phantom limb pain beyond the active life of the anesthetic is inconsistent with the notion that phantom limb pain is caused by chronic activity in peripheral nerves.
      • Melzack R
      Phantom limb pain implications for treatment of pathologic pain.
      Moreover, the observation that phantom sensation and phantom limb pain may be present immediately after amputation rule out the causal role of neuromas, as these have not yet formed in the amputation stump.
      As early as the 1940s, a number of authors were proposing that phantom limb pain could not be explained by peripheral mechanisms alone.
      • Livingston W.K
      • Bailey A.A
      • Moersch F.P
      Phantom limb.
      This led to a search for alternative mechanisms. Livingston proposed that phantom limb pain can be attributed to abnormal firing patterns in the internuncial neurons in the spinal cord.
      • Livingston W.K
      In a study of 36 upper limb amputees suffering from phantom limb pain, he injected a local anesthetic into the sympathetic ganglia in the spinal cord. Although nine of these patients reported a permanent reduction in their pain, more than two-thirds reported a temporary reduction in pain. Livingston suggested that “closed, self-sustaining, reverberating circuits” are set up by chronic peripheral irritation or by the release of spinal cord cells from inhibitory control through the loss of afferent input. When these abnormal impulses reach the brain, they are experienced as painful. Furthermore once these circuits are established surgical removal of the peripheral source has no effect on them and, therefore, will not abolish the pain.
      Other authors have supported this hypothesis. For example, Carlen and his colleagues
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      proposed that peripheral and spinal factors alone account for phantom limb pain. These researchers argue that if higher central mechanisms, released from the inhibition of peripheral input, were critical to the experience of phantom limb pain, it would follow that phantoms experienced by paraplegics should be more vivid as they have lost more input than amputees. The literature suggests that this is not the case, as paraplegics report fewer and less vivid phantom phenomena than amputees.
      • Carlen P.L
      • Wall P.D
      • Nadvorna H
      • Steinbach T
      Phantom limbs and related phenomena in recent traumatic amputations.
      The mechanisms by which disinhibition in the spinal cord results in phantom limb pain have been examined in detail by Wall,
      • Wall P.D
      On the origin of pain associated with amputation.
      who proposed that the sudden lack of afferent input following amputation results in a number of changes at both the peripheral and spinal level. Furthermore, Wall proposed that there are both immediate and chronic changes. The process described by Wall
      • Wall P.D
      On the origin of pain associated with amputation.
      suggests that the effects of peripheral nerve lesions spread beyond the damaged cells into the spinal cord itself. As such, it addresses some of the observations on phantom phenomena not explained by peripheral theories.
      Although disinhibited spinal cord neurons may be responsible for phantom sensation, it is unlikely that this is the primary cause of phantom limb.
      • Ribbers G
      • Mulder T
      • Rijken R
      The phantom phenomenon a critical review.
      • Harwood D.D
      • Hanumanthu S
      • Stoudemire A
      Pathophysiology and management of phantom limb pain.
      Wall
      • Wall P.D
      On the origin of pain associated with amputation.
      confined his analysis of phantom phenomenon to those that occur following amputation. However, as noted above, phantom limb pain has also been reported in cases where there is no nerve damage and in cases where there has been a complete transection of the spinal cord.
      • Melzack R
      Phantom limbs.
      Accordingly, some researchers have gone on to postulate a role for higher brain centers in phantom limb pain. For example, Melzack
      • Melzack R
      Phantom limb pain implications for treatment of pathologic pain.
      and Melzack and Loeser
      • Melzack R
      • Loeser J.D
      Phantom body pain in paraplegics evidence for a central “pattern generating mechanism” for pain.
      have proposed that the reticular activating system plays an important role in phantom limb pain. These studies propose that when peripheral fibers are destroyed, thereby reducing input, inhibition is decreased and synchronous, self-sustaining activity develops at all neural levels. Thus, lack of input from the periphery following amputation will result in disinhibition not only at the spinal level described above, but also at the cortical level. This model would account for some of the puzzling aspects of phantom limb pain. It has not been empirically tested.

      Psychological Explanations for Phantom Limb Pain

      Some researchers propose that the puzzling aspects of phantom limb pain can be explained by looking at the psychological makeup of the amputee. Psychological theories have attributed chronic pain to personality disorders, masked depression, guilt, childhood deprivation or trauma, defense against loss, or repressed hostility and aggression.
      • Blumer D
      • Heilbronn M
      Chronic pain as a variant of depressive disease the pain-prone disorder.
      • Bond M.R
      • Pearson I.B
      Psychological aspects of pain in women with advanced cancer of the cervix.
      • Engel G.L
      ‘Psychogenic’ pain and the pain prone patient.
      • Hughes M
      • Zimm R
      Children with psychogenic abdominal pain and their families.
      • Lesse S
      Atypical facial pain of psychogenic origin a masked depressive syndrome.
      • Merskey H
      • Boyd D
      Emotional adjustment and chronic pain.
      • Swanson D.W
      Chronic pain as a third pathologic emotion.
      • Violon A
      The process of becoming a chronic pain patient.
      Preexisting personality has also been cited as an important feature in the development of phantom limb pain. For example, a study by Parkes
      • Parkes C.M
      Factors determining the persistence of phantom pain in the amputee.
      found that those who have persistent phantom pain scored highly on a personality measure of “compulsive self-reliance” and “rigidity.” Parkes proposed that those with a “rigid” personality dislike and resist change and therefore experienced persistent phantom pain because they found it difficult to deal with the changes that are an inevitable consequence of amputation. In addition, Parkes noted those who were “compulsively self-reliant” were also likely to experience persistent pain because of the helplessness that often accompanies amputation. If these individuals have to rely on others for things they previously did themselves, they become distressed and this leads to pain.
      Parkes and Napier
      • Parkes C.M
      • Napier M.M
      Psychiatric sequelae of amputation.
      concluded that the “denier” or “defiant type” of amputee has a “compulsive need to do everything at least as well as he could before the operation and if possible, better, as if to convince himself and everyone else that he is not incapacitated at all” (p. 443). These authors believed that if this defense mechanism is disrupted, for example, by post-amputation disability, psychological consequences, such as phantom pain would follow. In a similar vein, Friedmann
      • Friedmann L.W
      reported that “sensitive,” “intelligent,” and “neurotic” individuals are more prone to phantom sensation and phantom pain than less imaginative and more emotionally balanced people. It is interesting to note that this entire body of literature reports on “personality types” that are associated with increased phantom limb pain, whereas none discuss personality types that are less likely to experience phantom limb pain.
      The assumption in these retrospective studies is that emotional disturbance precedes the onset of phantom limb pain and is, therefore, significant in its etiology. This assumption is questionable. Sherman et al.
      • Sherman R.A
      • Sherman C.J
      • Bruno G.M
      Psychological factors influencing phantom limb pain an analysis of the literature.
      proposed that psychological explanations of phantom limb pain have less to do with personality and more to do with the post-amputation experience of many amputees suffering from phantom limb. For example, research has shown that despite numerous treatments for phantom limb pain, few are successful.
      • Sherman R.A
      • Tippens J.K
      Suggested guidelines for treatment of phantom limb pain.
      Sherman and his colleagues suggested that this low success rate of treatment for phantom limb pain will ultimately deter all but the most “persistent,” “self-reliant individuals” from continuing to insist on treatment. However, by continuing to seek treatment, these individuals are more likely to have their pain labeled as “psychogenic” and be referred to mental health clinics.
      • Sherman R.A
      • Tippens J.K
      Suggested guidelines for treatment of phantom limb pain.
      As many studies of the relationship between personality and phantom pain recruit subjects from mental health clinics, this self-selection bias results in a tendency for individuals with these characteristics to dominate the clinical picture of a typical patient with phantom limb pain.
      Studies that utilize representative samples of subjects do not support the hypothesis that patients with phantom limb pain are more likely to have personality problems. For example, Shukla et al.
      • Shukla G.D
      • Sahu C
      • Tripathi R.P
      • Gupta D
      Phantom limbs a phenomenological study.
      found no difference on a number of personality measures between those with and without phantom limb pain. Other recent studies indicate that among nonselected samples of amputees, those with phantom limb pain cannot be distinguished from those with phantom limb sensation or no pain by their scores on a questionnaire designed to measure psychological “rigidity.”
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      • Katz J
      • Melzack R
      Auricular TENS reduces phantom limb pain.
      In the chronic pain literature, study of the role of personality types has been replaced, to a degree, by examination of the role of appraisal in pain perception. These studies have better explanatory power in relation to the emotions and behavior of individuals with chronic pain. Unfortunately, the role of appraisal has, to date, not been examined in studies of individuals with phantom limb pain.
      Another class of psychological explanation proposes that phantom limb pain results from the use of defense mechanisms such as “denial” or “repression.”
      • Kolb L
      • Simmel M.L
      The conditions of occurrence of phantom limbs.
      • Weiss S.A
      • Fishman S
      Extended and telescoped phantom limbs in unilateral amputees.
      • Frazier S.H
      • Kolb L
      Psychiatric aspects of pain and the phantom limb.
      Weiss and Fishman
      • Weiss S.A
      • Fishman S
      Extended and telescoped phantom limbs in unilateral amputees.
      suggested that the phantom results from the individual’s denial of limb loss. Phantom pain is explained as functional in this situation, as it serves to reinforce the presence of the limb. In contrast, Simmel
      • Simmel M.L
      The conditions of occurrence of phantom limbs.
      proposed that phantom sensation was not caused by denial, but was the focus for it. Phantom pain occurs because the individual is overwhelmed by the emotions and anxiety associated with loss of a body part. Kolb,
      • Kolb L
      who proposed that phantom sensation was, in fact, a healthy response to amputation, advanced another denial mechanism. Those individuals who do not experience a phantom are denying their loss in contrast to those who do experience phantom limb pain. Kolb suggests that the more important the body part is to the individual (emotionally), the more likely they will be to deny their loss. Although he attributes the lower incidence of phantom breast to this mechanism of denial, recent studies have shown that the incidence of phantom sensation following mastectomy actually is as high as it is following limb amputation.
      • Simmel M.L
      The conditions of occurrence of phantom limbs.
      • Weinstein S
      • Sersen E.A
      • Vetter R.J
      Phantoms and somatic sensation in cases of congenital aplasia.
      • Bressler B
      • Cohen S.I
      • Magnussen F
      The problem of phantom breast and phantom pain.
      • Jamieson K
      • Wellisch D.K
      • Katz R.L
      Phantom breast syndrome.
      • Jarvis J.H
      Post-mastectomy breast phantom.
      It is possible that “denial” is a feature of the immediate post-amputation process,
      • Simmel M.L
      A study of phantoms after amputation of the breast.
      • Jamieson K
      • Wellisch D.K
      • Katz R.L
      Phantom breast syndrome.
      which has been likened to the grieving process associated with any loss,
      • Parkes C.M
      Components of the reaction to loss of a limb spouse or home.
      but it is difficult to see how a concept such as “denial” is empirically testable.
      Loss of a limb for whatever reason is a major event with profound implications for the psychological health of the individual involved. Given this, it is not surprising that 20–60% of amputees attending surgical or rehabilitation clinics are assessed as being clinically depressed.
      • Shukla G.D
      • Sahu C
      • Tripathi R.P
      • Gupta D
      Phantom limbs a phenomenological study.
      • Kashani J.H
      • Frank R.G
      • Kashani S.R
      • Wonderlich S.A
      • Reid J.C
      Depression among amputees.
      • Randall G
      • Ewalt J
      • Blair H
      Psychiatric reaction to amputation.
      However, there are major difficulties in establishing a causal link between psychological distress, such as depression, and phantom limb pain. Few studies actually measured psychological distress in amputees and those that do seldom examine distress in relation to phantom limb pain. In the studies described above, depression was examined in relation to amputee status rather than phantom pain.
      • Shukla G.D
      • Sahu C
      • Tripathi R.P
      • Gupta D
      Phantom limbs a phenomenological study.
      • Kashani J.H
      • Frank R.G
      • Kashani S.R
      • Wonderlich S.A
      • Reid J.C
      Depression among amputees.
      • Randall G
      • Ewalt J
      • Blair H
      Psychiatric reaction to amputation.
      Other reports of psychological distress come from pain management programs where those seeking treatment for phantom limb pain may be expected to experience more distress.
      • Dawson L
      • Arnold P
      Persistent phantom limb pain.
      • Frierson R.L
      • Lippmann S.B
      Psychiatric consultation for acute amputees.
      • Tomaszek D.E
      • Buckwalter J.A
      Above-knee amputations in psychiatric inpatients.
      Lindesay
      • Lindesay J
      Multiple pain complaints in amputees.
      divided his study population into those who requested treatment for phantom pain and those who did not and found that those who requested treatment for phantom pain were significantly more depressed than those who did not.
      A further difficulty in establishing the role of psychological distress comes from the items that appear in many of the instruments used to measure psychological distress. For example, difficulty in getting to sleep is regarded as a symptom of depression. It is also a feature of chronic pain. Similarly, endorsement of items relating to fatigue is indicative of depression, but is also associated with chronic pain in the absence of depression.
      • Haythornthwaite J.A
      • Sieber W.J
      • Kerns R.D
      Depression and the chronic pain experience.
      The relationship between phantom pain and psychological distress is also confounded by a variety of post-amputation factors. Parkes
      • Parkes C.M
      Psychosocial transitions comparison between reactions to loss of a limb and loss of a spouse.
      noted that the psychological distress that occurs following loss of a limb is similar to that noted following loss of a spouse. Others have not specifically reported grief, but have suggested there are a number of psychological adjustments to be made in the immediate post-amputation phase.
      • Friedmann L.W
      One consequence of these confounding factors is that they will vary as a function of time since amputation. A study by Frank et al.
      • Frank R.G
      • Kashini J.H
      • Kashini S.R
      • Wonderlich S.A
      • Umlaf R.L
      • Ashkanazi G.S
      Psychological response to amputation as a function of age and time since amputation.
      suggested that a significant proportion of amputees succumb to depression only after they have left the medical system. However, psychological distress often has been measured in the early post-amputation phase while amputees are still being seen regularly at rehabilitation clinics. Research on the relationship between psychological distress and phantom limb pain has been retrospective and cross sectional, rather than longitudinal. As such, nothing is known about the variation and/or changes in psychological distress from the early to the chronic post-amputation period. The complexity of this literature makes it difficult not only to determine causal connections, but also to determine how prevalent psychological distress is in amputees.
      Thus the literature may present a misleading picture of the psychological status of amputees with phantom limb pain. This situation has arisen, in part, because many researchers based their conclusions on unintentionally biased samples drawn from those amputees requesting treatment for phantom pain. This has resulted in unsubstantiated assumptions being made about the general population of amputees. These difficulties are further exacerbated by the failure to differentiate clearly between acute adjustment reactions following amputation and chronic problems.
      • Sherman R.A
      • Sherman C.J
      • Bruno G.M
      Psychological factors influencing phantom limb pain an analysis of the literature.
      For example, it is common to find reactions of shock, grief, anger, frustration, or denial in those who have recently undergone amputation.
      • Friedmann L.W
      • Parkes C.M
      Components of the reaction to loss of a limb spouse or home.
      Indeed, the enormity of amputation is such, that these symptoms of distress will be expected in the immediate post-amputation phase. More recent research suggests that psychological factors do not play a causal role in phantom limb pain.
      • Sherman R.A
      • Sherman C.J
      • Bruno G.M
      Psychological factors influencing phantom limb pain an analysis of the literature.
      • Arena J.G
      • Sherman R.A
      • Bruno G.M
      • Smith J.D
      The relationship between situational stress and phantom limb pain cross-lagged correlational data from six month pain logs.
      Rather, these studies suggest that episodes of phantom limb pain can be exacerbated by stress, fear, fatigue, and insomnia as may occur in other types of chronic pain. However, these are the only studies to date that have examined the role of psychological factors in ongoing phantom limb pain and both have utilized samples comprised of military veterans.

      Multicausal Theories

      Both physiological and psychological theories may explain some aspects of phantom limb pain. Even combined however, they do not provide a comprehensive explanation of phantom phenomena, including phantom limb pain. This is, perhaps, not surprising given the multicausal nature of pain proposed by the Gate Control Theory. Beyond a multicausal explanation, Sherman
      • Sherman R.A
      Stump and phantom limb pain.
      argues that phantom limb pain itself is not a unitary syndrome but a number of symptom classes. Each class is qualitatively different and is subserved by different etiological mechanisms. To support this argument, Sherman and his coworkers draw on the results of recent research showing that one class of phantom limb pain, which is characterized by “thermal” qualities, is associated with decreased blood flow to the stump and is described as burning phantom limb pain.
      • Sherman R.A
      Direct evidence of a link between burning phantom pain and stump blood circulation a case report.
      • Sherman R.A
      • Bruno G.M
      Concurrent variation of burning phantom limb and stump pain with near surface blood in the stump.
      • Kristen H
      • Lukeschitsch G
      • Plattner F
      • Sigmund R
      • Resch P
      Thermography as a means for quantitative assessment of stump and phantom pains.
      A separate class of phantom limb pain characterized by a “cramping” quality is associated with spike activity recorded from electromyography (EMG) of muscles within the stump. In the same study, no association was found between EMG spike activity and “thermal” descriptors of phantom limb pain.
      • Sherman R.A
      Stump and phantom limb pain.
      Further evidence that phantom limb pain may comprise a variety of symptom classes comes from a study of “somatosensory pain memories” in phantom limb pain.
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      Somatosensory pain memories are characterized by the qualities of pain experienced in the intact limb prior to amputation. Many authors have reported that some amputees experience phantom limb pain that closely resembles pain that was experienced in the limb prior to amputation (e.g., ref. 57). For example, phantom limb pain may be experienced as similar in quality and location to a painful wound being dressed, an ingrown toenail, a painful foot ulcer, or pain resulting from deep tissue injuries in the limb prior to amputation. Amputees assert that it is real pain that they are experiencing, not merely a recollection of the pain they had prior to amputation. Moreover, this type of phantom limb pain differs from that which is experienced in the distal part of the phantom limb.
      • Hill A
      • Niven C.A
      • Knussen C
      Pain memories in phantom limbs a case study.
      The pain experienced in a limb prior to amputation also may influence the course of phantom limb pain many months later. For example, pain in the limb prior to amputation predicts phantom limb pain at a 6 month follow-up.
      • Jensen T.S
      • Krebs B
      • Nielsen J
      • Rasmussen P
      Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
      However, if pain is relieved by the administration of a continuous epidural block for 3 days prior to amputation, the incidence of phantom limb pain is decreased at 6 months follow-up.
      • Bach S
      • Noreng M.F
      • Tjéllden N.U
      Phantom limb pain in amputees during the first 12 months following limb amputation, after preoperative lumbar epidural blockade.
      Jensen et al.
      • Jensen T.S
      • Krebs B
      • Nielson J
      • Rasmussen P
      Immediate and long-term phantom limb pain in amputees clinical characteristics and relationship to pre-amputation limb pain.
      reported that 74% of the patients in their study had pain in a similar location to pre-amputation pain 8 days after amputation. Forty-five percent still had pain in the same location 2 years later. The character of their phantom limb pain was similar to pre-amputation pain in 53% and 35% of patients after 8 days and 2 years, respectively. When both location and quality were examined, phantom limb pain resembled pre-amputation pain in 36% of patients at 8 days and 10% of patients after 2 years. Similarly, Katz and Melzack
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      report that 42% of their sample had a “somatosensory pain memory” that resembled the quality and location of pre-amputation sensations. Recent findings suggest that the reason for these “somatosensory pain memories” may lie in functional or structural change within the central nervous system in response to noxious somatosensory input.
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      • McQuay H.J
      • Dickenson A.H
      Implications of nervous system plasticity for pain management.

      Neuromatrix Theory

      Given the diversity of phantom limb phenomena, no one causal mechanism can explain phantom limb pain. Melzack
      • Melzack R
      Phantom limbs.
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      has proposed that this diversity can be better explained using the concept of a neuromatrix.
      The neuromatrix is defined as a “network of neurons that extends throughout widespread areas of the brain, composing the anatomical substrate of the physical self” (ref. 71, p. 91). It is suggested that the neuromatrix extends to at least three major neural circuits. One is the sensory pathway that travels through the thalamus to the somatosensory cortex. It is this pathway that primarily carries information from the periphery. A second is the pathway that goes through the reticular formation to the limbic system. This system is critical for emotion and motivation and may account for the affective descriptions of their phantoms used by paraplegics who have complete transection of the spinal cord and psychological distress that is commonly seen in patients with phantom limb pain.
      • Lindesay J
      Validity of the general health questionnaire in detecting psychiatric disturbance in amputees with phantom limb pain.
      • Arena J.G
      • Sherman R.A
      • Bruno G.M
      • Smith J.D
      The relationship between situational stress and phantom limb pain cross-lagged correlational data from six month pain logs.
      Manipulating the limbic system in an animal model can relieve pain.
      • Coderre T.J
      • Vaccarino A.L
      • Melzack R
      1990). Central nervous system plasticity in the tonic response to subcutaneous formalin injection.
      A third circuit incorporates the parietal lobe, an area that is significant in evaluating sensory signals and in the recognition of the self. The importance of the parietal lobe in the sense of self has been shown in studies of brain-damaged patients. For example, Sacks
      • Sacks O
      reported that patients with damage to the parietal lobe often refuse to accept that their limb is part of them. The limb itself is not damaged, but the patient does not accept it as part of the body. A brain-damaged patient cried out in pain when his leg was strongly pinched but still would not accept that it was his leg that was pinched.
      Input to the above systems, either from the periphery or from within the neuromatrix itself, is processed simultaneously and then shared with other brain systems, producing output which is transformed into conscious awareness. Melzack describes the basic output from the neuromatrix as a “neurosignature” which is particular to the individual. It is thought to be determined by the pattern and strength of connections between neurons within the neuromatrix. It is this pattern that indicates that “the body is intact and unequivocally one’s own” (ref. 71, p. 91). In normal circumstances, sensory signals are processed by these systems then modulated by the ongoing pattern of the neuromatrix (the neurosignature). The resulting output of the system contains information about the sensory input and the conviction that the sensation is occurring in the body.
      Melzack suggests that this postulated process is much like the working of “cell assemblies” proposed by Hebb.
      • Hebb D.O
      Hebb argued that the strength of the connection between any two cells simultaneously activated by sensory input will become stronger. This in turn will lead to an assembly of neurons in which a signal going into one part of the group will spread through the rest. Hebb’s model suggests that cells activated by environmental stimuli will become part of the cell assembly when paired with other active cells. Melzack
      • Melzack R
      Phantom limbs.
      argues that, in terms of the neuromatrix, the primary components must be genetically prewired, although experience will “add or delete, strengthen or weaken existing synapses” (p. 124). He cites cases where a limb is congenitally absent yet the individual still experiences a vivid phantom as evidence of the genetic or “hard wired” basis of the neuromatrix. In these cases, the limb itself has never provided input to the matrix (ref. 13, p. 124).
      This theory proposes that abnormal input to the neuromatrix following amputation alters the pattern generated by the neuromatrix and results in output which is experienced as a painful phantom. Abnormal input can either result from lack of normal sensory input following amputation or from high levels of input caused by excessive firing in damaged nerves. The phantom itself appears to be felt because of the basic pattern of the neuromatrix, the neurosignature. The painful aspect of the output may occur for a variety of reasons. For example, the cramping or shooting quality of phantom limb pain might occur because limb movement will be a prewired aspect of the neuromatrix. Following amputation, the neuromatrix no longer receives signals from the periphery that the limb is moving. The output from the neuromatrix then will include the basic neurosignature which has been modulated to include strong messages for the limb to move. This results, not only in the report of a cramping type of pain, but also in the EMG spike activity associated with this aspect of phantom limb pain.
      • Sherman R.A
      Stump and phantom limb pain.
      • Melzack R
      Phantom limbs.
      One illustration of the way that experience can shape the pattern produced by the neuromatrix is highlighted by research on somatosensory pain memories.
      • Katz J
      • Melzack R
      Pain “memories” in phantom limbs review and clinical observations.
      It is suggested that a neural representation of the pre-amputation pain is formed subsequent to one very intense pain experience or is formed and gradually strengthened as a result of multiple occurrences of pain. Given that “pain memories” are experienced as both sensory and affective events, it is reasonable to assume that both these aspects of the original experience have been encoded within the neuromatrix. This is supported in a recent study by Hill et al.
      • Hill A
      • Niven C.A
      • Knussen C
      The role of coping in adjustment to phantom limb pain.
      Because the neuromatrix extends over such a wide area of the brain, this is a difficult theory to test. Some work is currently being undertaken.
      • Coderre T.J
      • Vaccarino A.L
      • Melzack R
      1990). Central nervous system plasticity in the tonic response to subcutaneous formalin injection.
      • Coderre T.J
      • Katz J
      • Vaccarino A.L
      • Melzack R
      Contribution of central neuroplasticity to pathological pain review of clinical and experimental evidence.
      • Vaccarino A.L
      • Melzack R
      The role of the cingulum bundle in self-mutilation following peripheral neurectomy in the rat.
      • Vaccarino A.L
      • Melzack R
      Temporal processes of formalin pain differential role of the cingulum bundle, fornix pathway and medial bulboreticular formation.
      Nonetheless, it does provide a framework that is consistent with peripheral, spinal, and central mechanisms identified in the literature and incorporates many previously unexplained research findings. Notably, Neuromatrix Theory implies that many facets of an amputee’s experience might contribute to the quality and intensity of his or her phantom limb pain by initiating activity within the neuromatrix. In particular, this theory suggests that in addition to sensory input triggering the neuromatrix, psychological or social factors may also produce input that activates the matrix and results in the experience of phantom limb pain. For example, research showing that episodes of phantom limb pain can be exacerbated by stress, fear, fatigue, and insomnia
      • Sherman R.A
      • Sherman C.J
      • Bruno G.M
      Psychological factors influencing phantom limb pain an analysis of the literature.
      • Arena J.G
      • Sherman R.A
      • Bruno G.M
      • Smith J.D
      The relationship between situational stress and phantom limb pain cross-lagged correlational data from six month pain logs.
      might be explained by this theory. However, this possibility is not explicit in the exposition of this theory to date and no research has been conducted examining the role of psychological factors in phantom limb pain within this framework.
      In many respects, Neuromatrix Theory is similar to Gate Control Theory and could be applied to chronic pain in general, rather than phantom limb pain in particular. Where it differs from Gate Control Theory is in the postulation of a neural network across widespread areas of the brain, in which activity of one area will result in output from the entire network. As such, this theory provides a mechanism to explain how numerous factors can result in the perception of pain. Gate Control Theory proposes that pain is a perception whose quality and intensity are influenced by the unique history of the individual, the meaning they give to the pain-producing situation, the meaning given to the consequences of pain, and the individual’s state of mind. This theory does not, however, explain how these factors influence pain perception. In contrast, Neuromatrix Theory not only suggests a possible mechanism to explain the diversity of factors involved in the perception of pain, but also provides a plausible explanation for much of literature on the psychological management of pain.
      Neuromatrix Theory can be criticized on several grounds. First, because the neuromatrix is postulated to extend over such a diffuse area of the brain, the proposed mechanisms are almost impossible to operationalize and test. For example, it would be unworkable to isolate and examine any one proposed activating stimulus when all aspects of experience are described as potential stimuli within this matrix. Animal models of phantom limb pain may be informative.
      • Coderre T.J
      • Vaccarino A.L
      • Melzack R
      1990). Central nervous system plasticity in the tonic response to subcutaneous formalin injection.
      • Coderre T.J
      • Katz J
      • Vaccarino A.L
      • Melzack R
      Contribution of central neuroplasticity to pathological pain review of clinical and experimental evidence.
      • Vaccarino A.L
      • Melzack R
      The role of the cingulum bundle in self-mutilation following peripheral neurectomy in the rat.
      • Vaccarino A.L
      • Melzack R
      Temporal processes of formalin pain differential role of the cingulum bundle, fornix pathway and medial bulboreticular formation.
      Second, Neuromatrix Theory provides a plausible explanation for phantom sensation (particularly in the case of congenital limb absence where there is no damage to the peripheral nervous system). However, it is not clear why phantom limb pain rather than phantom sensation should occur as a response to the output from this matrix. Moreover, this theory provides no explanation of why some amputees do not experience phantom limb pain. Future research is required to develop the prediction that mechanisms can be operationalized and tested in amputees experiencing phantom limb pain.
      • Melzack R
      Phantom limbs and the concept of a neuromatrix.

      References

        • Sherman R.A
        Stump and phantom limb pain.
        Neurol Clin. 1989; 7: 249-264
        • Jensen T.S
        • Rasmussen P
        Phantom pain and other phenomena after amputation.
        in: Wall P.D Melzack R Textbook of pain. 3. Churchill Livingstone, London1994: 651-683
        • Ribbers G
        • Mulder T
        • Rijken R
        The phantom phenomenon.
        Int J Rehabil Res. 1989; 12: 175-186
        • Parkes C.M
        Factors determining the persistence of phantom pain in the amputee.
        J Psychosom Res. 1973; 17: 97-108
        • Jensen T.S
        • Krebs B
        • Nielsen J
        • Rasmussen P
        Phantom limb, phantom pain and stump pain in amputees during the first six months following limb amputation.
        Pain. 1983; 17: 243-256
        • Browder J
        • Gallagher J.P
        Dorsal cordotomy for painful phantom limb.
        Ann Surg. 1948; 128: 457-469
        • Sunderland S
        Nerves and nerve injuries. E. and S. Livingstone, Edinburgh1978
        • Silosberg A
        Les algies des amputés. Masson, Paris1948
        • Jensen T.S
        • Krebs B
        • Nielson J
        • Rasmussen P
        Immediate and long-term phantom limb pain in amputees.
        Pain. 1985; 21: 407-414
        • Sherman R.A
        • Sherman C.J
        A comparison of phantom sensations among amputees whose amputations were of civilian and military origins.
        Pain. 1985; 21: 91-97
        • Morgenstern F.S
        The effects of sensory input and concentration of post-amputation phantom limb pain.
        J Neurol Neurosurg. 1964; 27: 58-65
        • Sherman R.A
        Phantom pain. Plenum Press, New York1997
        • Melzack R
        Phantom limbs.
        Sci Amer. 1992; April: 120-126
        • Mitchell S.W
        Phantom limbs.
        Lippincott’s Magazine of Popular Literature and Science. 1871; 8: 563-569
        • Simmel M.L
        On phantom pain.
        Arch Neurol Psychiatry. 1956; 75: 637-647
        • Bors E
        Phantom limbs of patients with spinal cord injury.
        Arch Neurol Psychiatry. 1951; 66: 610-631
        • Burke D.C
        • Woodward J.M
        Phantom movement and phantom feeling in complete paraplegic patients.
        Clin Neurol. 1976; 26: 489-499
        • Conomy J.P
        Disorders of body image after spinal cord injury.
        Neurology. 1973; 23: 842-850
        • Davis R
        Pain and suffering following spinal cord injury.
        Clin Orthop Rehabil Res. 1975; 112: 78-80
        • Frisbie J.H
        • Aguilera E.J
        Chronic pain after spinal cord injury.
        Paraplegia. 1990; 28: 460-465
        • Melzack R
        • Loeser J.D
        Phantom body pain in paraplegics.
        Pain. 1978; 4: 195-210
        • Paeslack V
        • Spahn B
        • Sommer K
        Pain symptoms in paraplegic patients.
        Fortschrider Medicine. 1990; 108: 57-58
        • Melzack R
        • Bromage P.R
        Experimental phantom limbs.
        Exp Neurol. 1973; 39: 261-269
        • Wall P.D
        • Nathan P.W
        • Noordenbos W
        Ongoing activity in peripheral nerves.
        I. Exp Neurol. 1973; 38: 90-98
        • Friedmann L.W
        The psychological rehabilitation of the amputee. Charles C. Thomas, Springfield, Illinois1978
        • Kolb L
        The painful phantom. Charles C. Thomas, Springfield, IL1954
        • Simmel M.L
        The conditions of occurrence of phantom limbs.
        Proc Am Philos Soc. 1958; 102: 492-500
        • Weinstein S
        • Sersen E.A
        Phantoms in cases of congenital absence of limbs.
        Neurology. 1961; 10: 905-1111
        • Weinstein S
        • Sersen E.A
        • Vetter R.J
        Phantoms and somatic sensation in cases of congenital aplasia.
        Cortex. 1964; 1: 276-290
        • Lacroix R
        • Melzack R
        • Smith D
        • Mitchell N
        Multiple phantom limbs in a child.
        Cortex. 1992; 28: 503-507
        • Saadah E.S.M
        • Melzack R
        Phantom limb experiences in congenital limb-deficient adults.
        Cortex. 1994; 30: 479-485
        • Melzack R
        • Wall P.D
        The challenge of pain. 2. Penguin Books, London1988
      1. Cronholm B. Phantom limbs in amputees: a study of changes in the integration of centripetal impulses with special reference to referred sensations. Acta Psychiatr Neurol Scandinav 1951(suppl 72).

        • Livingston W.K
        Pain mechanisms. Macmillan, New York1943
        • Haber W.B
        Effects of loss of limb on sensory functions.
        J Psychol. 1995; 40: 115-123
        • Weiss S.A
        • Fishman S
        Extended and telescoped phantom limbs in unilateral amputees.
        J Abnormal Soc Psychol. 1963; 66: 489-497
        • Katz J
        Psychophysiological contributions to phantom limbs.
        Canad J Psychiatry. 1992; 37: 282-297
        • Merzenich M.M
        • Nelson R.J
        • Stryker M.P
        Somatosensory cortical map changes following digit amputation in adult monkeys.
        J Comp Neurol. 1984; 224: 591-605
        • Jenkins W.M
        • Merzenich M.M
        • Ochs M.T
        • Allard T
        • Guic-Robles E
        Functional reorganization of primary somatosensory cortex in adult owl monkeys after behaviorally controlled tactile stimulation.
        J Neurophysiol. 1990; 63: 82-104
        • Kallio K.E
        Phantom limb of forearm stump cleft by kineplastic surgery.
        Acta Chir Scand. 1950; 99: 121-132
        • Flor H
        • Elbert T
        • Knecht S
        • Wienbruch C
        Phantom limb pain as a perceptual correlate of cortical reorganization following arm amputation.
        Nature. 1995; 375: 482-484
        • Henderson W.R
        • Smyth G.E
        Phantom limbs.
        J Neurol Neurosurg Psychiatry. 1948; 11: 88-112
        • Riddoch G
        Phantom limbs and body shape.
        Brain. 1941; 64: 197-222
        • Berger M
        • Gerstenbrand F
        Phantom illusions in spinal cord lesions.
        in: Siegfried J Zimmermann M Phantom and stump pain. Springer Verlag, Verlag, Berlin1981: 66-73
        • Abramson A.S
        • Feibel A
        The phantom phenomena.
        Bull NY Acad Med. 1981; 57: 99-122
        • Ewalt J.R
        • Randall G.C
        • Morris H
        The phantom limb.
        Psychosom Med. 1947; 9: 118-123
        • Hermann L.G
        • Gibbs E.W
        Phantom limb pain.
        Am J Surg. 1945; 67: 168-180
        • Sternbach R.A
        Pain. Academic Press, New York1968
        • Buchannan D.C
        • Mandel A.R
        The prevalence of phantom limb experience in amputees.
        Rehabil Psychol. 1986; 31: 183-188
        • Krebs B
        • Jensen T.S
        • Kroner Nielsen J
        • Jorgensen H.S
        Phantom limb phenomena in amputees 7 years after limb amputation.
        in: Fields H.L Dubner H.L Cervero F Advances in Pain Research and Therapy, Vol. 9. Raven Press, New York1984: 425-429
        • Solonen K.A
        The phantom phenomenon in Finnish war veterans.
        Acta Orthop Scand. 1962; 54: 1-119
        • Sherman R.A
        • Sherman C.J
        • Parker L
        Chronic phantom and stump pain among American veterans.
        Pain. 1984; 18: 83-95
        • Hill A
        The use of pain coping strategies by patients with phantom limb pain.
        Pain. 1993; 55: 347-353
        • Sherman R.A
        • Sherman C.J
        Prevalence and characteristics of chronic phantom limb pain among American veterans.
        Am J Phys Med. 1983; 62: 227-238
        • Sternbach T
        • Nadvrona H
        • Arazi D
        A five year follow-up study of phantom limb pain in post-traumatic amputees.
        Scand J Rehabil Med. 1982; 14: 203-207
        • Almagor M
        • Jaffe Y
        • Lomranz J
        The relation between limb dominance, acceptance of disability, and the phantom limb phenomenon.
        J Abnormal Psychol. 1978; 87: 377-379
        • Bailey A.A
        • Moersch F.P
        Phantom limb.
        Can Med Assoc J. 1941; 45: 37-42
        • Dawson L
        • Arnold P
        Persistent phantom limb pain.
        Percept Motor Skills. 1981; 53: 135-138
        • Duane L.T.C
        • Howard L
        Group therapy for amputees in a ward setting.
        Military Med. 1983; 148: 678-680
        • Marsland A.R
        • Weekes J.W.N
        • Atkinson R.L
        • Leong M.G
        Phantom limb pain.
        Pain. 1982; 12: 295-297
        • Minichiello W.E
        Treatment of hyperhidrosis of amputation site with hypnosis and suggestions involving classical conditioning.
        Int J Psychosom. 1987; 34: 7-8
        • McGrath P.A
        • Hillier L.M
        Phantom limb sensations in adolescents.
        J Pain Symptom Manage. 1992; 7: 46-53
        • Siegel E.F
        Control of phantom limb pain by hypnosis.
        J Consult Clin Psychol. 1979; 21: 285-287
        • Stannard C.F
        • Porter G.E
        Ketamine hydrochloride in the treatment of phantom limb pain.
        Pain. 1993; 54: 227-230
        • Dubuisson D
        • Melzack R
        Classification of clinical pain descriptions by multiple group discriminant analysis.
        Exp Neurol. 1976; 51: 480-487
        • Marshall M
        • Helme S.E
        • Deathe A.B
        A comparison of psychosocial functioning and personality in amputee and chronic pain populations.
        Clin J Pain. 1992; 8: 351-357
        • Katz J
        • Melzack R
        Referred sensations in chronic pain patients.
        Pain. 1987; 28: 5-59
        • Melzack R
        The McGill Pain Questionnaire.
        Pain. 1975; 1: 277-299
        • Carlen P.L
        • Wall P.D
        • Nadvorna H
        • Steinbach T
        Phantom limbs and related phenomena in recent traumatic amputations.
        Neurology. 1978; 28: 211-217
        • Katz J
        • Melzack R
        Pain “memories” in phantom limbs.
        Pain. 1990; 43: 319-336
        • Melzack R
        Phantom limbs and the concept of a neuromatrix.
        Trends Neurosci. 1990; 13: 88-92
        • Hill A
        • Niven C.A
        • Knussen C
        The role of coping in adjustment to phantom limb pain.
        Pain. 1995; 62: 79-86
        • Katz J
        Psychophysical correlates of phantom limb experience.
        J Neurol Neurosurg Psychiatry. 1992; 55: 811-821
        • Katz J
        • Melzack R
        Auricular TENS reduces phantom limb pain.
        J Pain Symptom Manage. 1991; 6: 73-83
        • Shukla G.D
        • Sahu C
        • Tripathi R.P
        • Gupta D
        Phantom limbs.
        Br J Psychiatry. 1982; 141: 54-58
        • Hill A
        • Niven C.A
        • Knussen C
        Pain memories in phantom limbs.
        Pain. 1996; 66: 381-384
        • Kashani J.H
        • Frank R.G
        • Kashani S.R
        • Wonderlich S.A
        • Reid J.C
        Depression among amputees.
        J Clin Psychiatry. 1983; 44: 256-258
        • Kegel B
        • Carpenter M.L
        • Burgess E.M
        Functional capabilities of lower extremity amputees.
        Arch Phys Med Rehabil. 1978; 59: 109-120
        • Parkes C.M
        • Napier M.M
        Psychiatric sequelae of amputation.
        Br J Psychiatry. 1975; 9: 440-446
        • Melzack R
        • Katz J
        Pain measurement in persons in pain.
        in: Melzack R Wall P.D Textbook of pain. 3. Churchill Livingstone, Edinburgh1994: 337-351
        • Bach S
        • Noreng M.F
        • Tjéllden N.U
        Phantom limb pain in amputees during the first 12 months following limb amputation, after preoperative lumbar epidural blockade.
        Pain. 1988; 33: 297-301
        • Parkes C.M
        Components of the reaction to loss of a limb spouse or home.
        J Psychosom Res. 1972; 17: 97-98
        • Sherman R.A
        Direct evidence of a link between burning phantom pain and stump blood circulation.
        Orthopedics. 1984; 7: 19-20
        • Sherman R.A
        • Bruno G.M
        Concurrent variation of burning phantom limb and stump pain with near surface blood in the stump.
        Orthopedics. 1987; 10: 395-401
        • Kristen H
        • Lukeschitsch G
        • Plattner F
        • Sigmund R
        • Resch P
        Thermography as a means for quantitative assessment of stump and phantom pains.
        Prosthet Orthot Int. 1984; 8: 76-81
        • Frank R.G
        • Kashini J.H
        • Kashini S.R
        • Wonderlich S.A
        • Umlaf R.L
        • Ashkanazi G.S
        Psychological response to amputation as a function of age and time since amputation.
        Br J Psychiatry. 1984; 144: 493-497
        • Gerhards F
        • Florin I
        • Knapp T
        The impact of medical, reeducational, and psychological variables on rehabilitation outcomes in amputees.
        Int J Rehabil Res. 1984; 7: 379-388
        • Sherman R.A
        • Sherman C.J
        • Bruno G.M
        Psychological factors influencing phantom limb pain.
        Pain. 1987; 28: 285-295
        • Lindesay J
        Validity of the general health questionnaire in detecting psychiatric disturbance in amputees with phantom limb pain.
        J Psychosom Res. 1986; 30: 277-281
        • Williamson G.M
        • Schulz R
        • Bridges M.W
        • Behan A.M
        Social and psychological factors in adjustment to phantom limb pain.
        J Soc Behav Personal (special publication). 1994; 9: 249-269
        • Winchell E
        Coping with limb loss. Avery Publishing Group, New York1995
        • Tebbi C.K
        • Mallon J.C
        Long term psychosocial outcome among cancer amputees in adolescence and early adulthood.
        J Psychosoc Oncol. 1988; 5: 69-82