Morgan T. Sammons, PhD

Continuing Education Information

Chronic pain, albeit a common medical complaint, is notoriously difficult to both define and study. Though as many as 35 percent of respondents to surveys report persistent pain on at least a weekly basis, pain is by definition subjective and difficult to characterize accurately. The experience of pain and the tolerance for pain vary substantially both within and between individuals. In general, pain can be classified into broad groupings of acute and chronic pain, and into rough categories of musculoskeletal or neuropathic pain. Pain that is musculoskeletal in origin includes syndromes of fibromyalgia, post-traumatic pain, arthritic pain, and head pain that is primarily muscular in origin. Of musculoskeletal pain, chronic low back pain and head pain tend to be best studied, so more evidence has accumulated for or against the use of psychotropics with these conditions. Neuropathic pain results from damage to the central or peripheral nervous system, either due to injury (stroke, traumatic injury), neurological degeneration associated with chronic diseases (multiple sclerosis, diabetic neuropathies) or that associated with infections processes (post-herpetic neuralgia). Other syndromes with uncertain etiology, like trigeminal neuralgia or complex regional pain syndrome (formerly known as reflex sympathetic dystrophy; Ribbers, et al., 2003) are notoriously difficult to treat and are the focus of much clinical attention in managing chronic pain. Gastrointestinal pain such as that associated with irritable bowel syndrome has been investigated as a target of treatment with psychotropic agents. Tricyclic antidepressants appear to have some efficacy here, other classes of antidepressants do not.

Head pain represents a special case in that it can be both muscular, vascular, and neuropathic in origin. Headache is the most frequent cause of disability associated with pain, and is most associated with lost work productivity. Headache is generally grouped into 4 types: migrainous or vascular head pain, muscle tension, cluster, and miscellaneous types including post-traumatic or post-infectious head pain (see Sammons, in press, for a review of treatment of headache). Although psychotropics again play an adjunctive role in managing headache, particularly since the advent of the triptan class of medications that are often successful in preventing the onset of head pain, antidepressants and mood stabilizers are often used for chronic head pain. As with other types of pain, the tricyclic antidepressants are in most instances found to be of some efficacy, as are anticonvulsant medications, with newer antidepressants showing less effectiveness. (Tomkins, Jackson, O'Malley, Balden, & Santoro, 2001).

Management of pain syndromes with non-opiate psychotropics (and for the purposes of this article opiate analgesics will be considered not to be psychotropics, in spite of their euphoriant and anxiolytic effects) has not proven to be an easy task. Opiates and the non-steroidal anti-inflammatory agents (NSAIDS, such as ibuprofen) remain the mainstay of treatment for much acute and chronic pain. Opiates are effective analgesic agents, but their association with dependence and abuse has led to the search for other effective agents, and fears of government prosecution for overprescription of opiates has led a number of physicians to be reluctant to prescribe them. Countering this fear has been the increasing advocacy of patient's rights groups who express concern that pain is routinely both underrecognized and undertreated. Reflective of this trend are activities by the Joint Commission on the Accreditation of Healthcare Organizations, which has recently proposed that assessment of pain be adopted as the "fifth vital sign" in hopes of providing effective pain management to all patients. The recent removal off the market of the COX-2 inhibitors such as Bextra (valdecoxib) and Vioxx (rofecoxib) due to their association with heart attacks has spurred renewed interest in finding other non-opiate agents that are both safe and effective in pain management.

Any discussion of the treatment of pain cannot avoid addressing the topic of the placebo response. A substantial placebo response to acute pain has been well known and convincingly demonstrated experimentally. Early clinical experiments showing that patients administered inert substances labeled as analgesics described a significant reduction in pain are well known. More recently, the placebo response has been convincingly demonstrated at the molecular level. Subjects with experimentally induced pain who were given a placebo with the expectation that it would cause pain relief reported lower pain scores and had an increased tolerance for painful stimuli. PET scanning of subjects' brains after administration of placebo showed activation of both cortical and subcortical opiate receptors in patterns expected after administration of opiate analgesics (Zubieta, et al., 2005).

Antidepressants and chronic pain

The pain control mechanism of action of antidepressants is not well described but is almost certainly separate from their antidepressant mechanism of action. In spite of the recognized associations among depression, anxiety and chronic pain, effective treatment of depression contributes relatively little to the overall amelioration of chronic pain symptoms. Both depressed and non-depressed patients may show some response to antidepressants, and it is generally the case that in pain management most antidepressants are not used at doses considered to be effective for depression. Some theorize that the activity of antidepressants at multiple receptor sites for norepinephrine and serotonin give antidepressants efficacy as antinociceptives. These drugs, however, are also active at the n-methyl-d-aspartate (NMDA) receptor and at certain subtypes of opiate receptors (Colluzi & Mattia, 2005). Their activity at the alpha-adrenoceptor may also be associated with pain control (Bixquert-Jimenez & Bixquert-Pla, 2005), though this is by no means clear.

Though clarity on their mechanism of action is lacking, antidepressants have been used for decades as adjunctive agents in the management of chronic pain, both of musculoskeletal and neuropathic origin, and a reasonable body of evidence has accumulated that support their use, at least in an adjunctive fashion. Most of the evidence in this realm pertains to the use of older agents, notably the tricyclic antidepressants (TCAs). These are generally used in lower doses, and while some neurobiological explanation for their specific use in pain syndromes is extant in the literature, they are by no means the cornerstone of treatment. Tricyclic antidepressants have a fairly wide spectrum of action and influence neurotransmission at a number of receptor sites, including those for norepinephrine, dopamine, and serotonin, as well as activity at cholinergic and histaminergic receptors. As noted above, they also have lesser activity at other receptors, such as opiate receptor sites, but it is unknown which if any of these actions is responsible for their ability to modulate neuropathic or musculoskeletal pain. What does seem clear is that this effect is not related to their ability to block serotonin reuptake, since more potent serotonin blockers are less successfully used in pain management. Clinically, it is commonly assumed that the TCAs of the tertiary amine class, such as imipramine, amitriptyline, and doxepin may have slightly more analgesic potency than their metabolites (e.g., the secondary amines such as nortriptyline and desipramine). Whether this is because the tertiary TCAs are more sedating and in general have more anticholinergic activity than the secondary amines is also unknown. In any event, the TCAs must be used judiciously due to their many side effects (weight gain, sedation, dry mouth, urinary retention, blurred vision, and cardiac conduction problems among them). This significantly limits the use of these agents in the elderly and those with concomitant medical conditions - that is, those who may be most prone to chronic pain syndromes. Doses of these drugs, as stated earlier, rarely approaches true antidepressant dose levels, but is generally in the range of 25-100 mgs, usually given at bedtime to improve sleep.

Evidence for the use of the Serotonin Reuptake Inhibitors (SRIs) is in general not encouraging, and these have not achieved a foothold in the clinical management of pain. After their introduction in the late 1980s and demonstrated efficacy in the management of depression and anxiety there was interest in expanding their use to treat pain syndromes. As with depression and anxiety disorders, this interest was spurred by the desire to find alternatives to the TCAs that were both less toxic and lacking the problematic side effect profile of the earlier agents. While some early studies suggested promise in the management of head pain and fibromyalgia, more careful investigation has not resulted in encouraging findings. Head to head comparisons of SRIs like fluoxetine (Prozac and others), citalopram (Celexa) and sertraline (Zoloft) against the TCAs generally yield results in favor of the TCAs.

Some of the newer "mixed-function" or serotonin and norepinephrine reuptake inhibitors, (SNRIs) like venlafaxine (Effexor), duloxetine (Cymbalta), milnacipran (not yet available in the United States), and mirtazepine (Remeron) seem to have greater efficacy than the SRIs in managing chronically painful conditions. Although some authors believe that the SNRIs are as effective as the TCAs in pain syndromes and should be preferentially used because of a more benign side effect profile, as of yet this has not been conclusively shown.

Of the SNRIs, only two are available at this time in the US: Venlafaxine (Effexor) and Cymbalta (duloxetine). Venlafaxine is available in a number of dose strengths and extended and immediate release preparations, making dose adjustments relatively facile. Cymbalta is a newer agent and is, as of this writing, available only in immediate release form. Duloxetine has, however, been specifically studied in one type of neuropathic pain (diabetic neuropathy) and has received an FDA indication for this condition, making it the only newer antidepressant with a specific indication for pain. A recent limited (Phase 2) study of a similar drug, milnacipran, in treating patients with fibromyalgia found greater efficacy than placebo at doses of 200 mg daily for 3 months for symptoms of pain, fatigue, and a measure of global well-being. The authors of this drug company sponsored trial reported that the observed effect sizes were similar to those of the TCAs (Gendreau, et al., 2005). These findings await further replication, but it seems to be reasonably well accepted that the SNRIs have greater pain control capability than SRIs, with those that have a lower ability to block serotonin reuptake (e.g., milnacipran and duloxetine) being more effective at pain control than more selective agents (e.g. venlafaxine; Stahl, Grady, Moret, & Briley, 2005).

Bupropion (Wellbutrin) is an antidepressant with a mechanism of action different from that of SRIs and the TCAs, having more activity at dopamine and norepinephrine receptors than with serotonin or cholinergic receptor subtypes. It has not been extensively studied in pain syndromes. One placebo controlled trial of sustained release bupropion did not find it more effective than placebo in reducing symptoms on commonly used pain scales (the McGill Pain Questionnaire and others; Katz, Pennella-Vaughan, Hetzel, Kanazi, & Dworkin, 2005), but at least one other study has found it to be of use in chronic pain.
Neuropathic pain, as stated above, may result from either nervous system insult such as stroke, trauma, infection, or neurodegenerative processes. As is the case with fibromyalgia, muscle tension head pain, and other musculoskeletal pain syndromes, the TCAs have some utility in the adjunctive management of neuropathic pain, and are of superior efficacy than the SRIs. Anticonvulsants, however, may have a more central role in managing neuropathic pain.
Perhaps most commonly employed in chronic neuropathic pain syndromes are the anticonvulsants (Maizels & McCarberg, 2005), which in addition to their antiseizure activity also have an established role in the management of cycling mood disorders like bipolar disorder. Like the TCAs, their exact mechanism of action is unknown, but is likely related to their ability to reduce neuronal excitation. Gabapentin is a close analog of the inhibitory neurotransmitter GABA (gamma-amino-butyric acid) but does not interact directly at GABA receptors in brain. Of these, gabapentin (Neurontin) and similar gabapentinoid compounds are frequently prescribed in pain syndromes. Gabapentin is specifically indicated for the treatment of postherpetic neuralgia. There is some evidence that its combination with opiate analgesics can assist in reducing doses of both medications while maintaining adequate pain control, a particular benefit in those for whom the sedation associated with higher doses of gabapentin is intolerable. Gabapentin is not metabolized, so drug-drug metabolic interactions are minimal, however, it is quite sedating and is associated with psychomotor impairment, and patients must be warned of synergistic interactions with other sedating drugs. Excess sedation is among the most common reasons for discontinuation of gabapentin. Gabapentin is generally given in doses of 300-1200 mg daily, with the bulk of the dose often given at bedtime to reduce side effects of sedation and drowsiness. Sedation and psychomotor impairment can be present even at low doses, however, and patients must be warned against driving or engaging in other potentially hazardous activities while initiating treatment.

Carbamazepine (Tegretol and others) is an anticonvulsant indicated specifically for the treatment of trigeminal neuralgia but which is also used in the treatment of other types of neuropathic pain. Carbamazepine is linked to the production of fetal malformations and its use is contraindicated in pregnancy. Becase carbamazepine is also a relatively potent inducer of the hepatic enzyme that is also responsible for the metabolism of oral contraceptives (OCPs), and can thereby lessen the contraceptive efficacy of OCPs, women taking carbamazepine should be counseled to rely on other mechanisms of birth control. Carbamazepine is more rarely associated with anemia and serious reductions in white blood cells, and with a variety of dental and dermatological side effects, including gum hyperplasia and sensitivity to sunlight. Careful dose monitoring is required for carbamazepine, as toxicity can be fatal. Monitoring of serum levels may be indicated in some instances, particularly if toxicity is suspected. Typical doses range from 400-1200 mg daily and usual serum levels for carbamazepine are from 4-12 micrograms per liter, but it should be noted that these figures relate to its use as an anticonvulsant, and standard serum levels for its use in pain have not been established. A long acting form of carbamazepine may reduce the need for multiple daily doses, however, requirements for serum monitoring and other cautions that apply to its use must still be observed. Like other anticonvulsants, carbamazepine is associated with sedation and cognitive slowing even at therapeutic doses, and interactions with other sedating agents can be problematic.

Oxcarbazepine, a metabolite of carbamazepine, is a potential alternative to carbamazepine, and presents some clinical advantages, chiefly in that blood draws for serum levels are not required. But it presents with some of the same drug-drug interaction concerns as carbamazepine, and has not been well studied in chronic pain conditions.
Pregabalin (Lyrica) is a recently approved agent indicated for the treatment of epilepsy and chronic pain in adults. Its mechanism of action is somewhat unclear. It does not directly bind to receptors for the inhibitory neurotransmitter GABA but binds to subreceptors on certain calcium channel in such a fashion as to reduce the release of several neurotransmitters. It has FDA approval for neuropathic pain, in particular postherpetic neuralgia and pain secondary to diabetic neuropathy. Pregabalin is minimally metabolized in the liver, so drug-drug metabolic interactions are less of a concern, but synergistic interactions with alcohol and sedatives are likely. It is, however, primarily excreted via the kidney, so dose adjustments are necessary in the elderly or those on dialysis. Doses in the range of 150-300 mg daily have generally been found to be efficacious in neuropathic pain. However, dizziness and somnolence are common side effects that worsen as the dose increases. Weight gain, edema, and blurred vision are also reported side effects that should be discussed with patients. Lyrica may be subject to abuse and is a Schedule V controlled substance.

A number of both older and newer anticonvulsants are encountered relatively frequently in pain management, but have not been as well studied. Phenytoin (Dilantin and generic) has not been systematically investigated, and results are mixed for two newer agents, lamotrigine (Lamictal) and zonisamide (Zonergan; Backonja, 2002). As with many similar agents, including most antidepressants, it is important to remember that these drugs are not FDA indicated for chronic pain. Clinicians must be aware of the medico-legal risks that attend the off-label use of such agents.

Like other areas of the psychopharmacological literature, long term systained improvement outside of clinical trials has not been convincingly demonstrated.


Psychotropics, primarily the TCAs and some anticonvulsants, are rather routinely employed in the management of pain. They appear to have some efficacy, but most often are used in the context of a multiple drug regime, including NSAIDS or opiate analgesics. While they have been demonstrated to be superior to placebo in clinical trials, most patients who take them experience numerous side effects and rarely achieve adequate pain control. Their exact mechanism of action remains poorly understood. They may act at certain receptor sites associated with pain reduction, such as opiate receptors, or they may manage pain by slowing activity in neurons responsible for transmission of pain impulses. Also at the receptor level, they may potentiate the activity of other analgesics. They may also act as anxiolytics, sedatives, and sleep promoters, all qualities likely to be appreciated by patients suffering from acute or chronic pain. While their use should not be overlooked in devising a comprehensive pain control regimen, patients should be carefully educated as to the limits of pain control likely to be achieved, and should be instructed in comprehensive regimens including exercise as tolerated, psychological interventions such as relaxation therapies or cognitive behavioral therapies, and other rehabilitative strategies like physical therapy. But patients should be advised that unimodal therapies are not likely to bring complete remission of pain. Psychotherapy should not be expected to serve as an adequate unimodal treatment for chronic pain.

Cognitive-Behavioral Therapy has been demonstrated to be of only modest efficacy in managing chronic pain (Butler, et al, 2005), underscoring the difficulty in managing this complex condition with unimodal therapies. Both patients and clinicians should have a reasonable expectation of the efficacy of proposed treatments prior to initiating treatment, and unfortunately, enhancing adaptation to chronic pain, rather than abolishing it, should in most instances be the agreed upon goal of therapy.


Morgan T. Sammons, Ph.D.1, is Director, Clinical Operations, U. S. Navy Bureau of Medicine and Surgery, Washington, DC. Dr. Sammons is President /Chair of the National Register Board of Directors. He is the author of a forthcoming volume entitled Psychopharmacology: An Integrated Approach, published by Oxford.

1The opinions expressed by this author are his own and do not represent the opinions or official positions of the US Navy or Department of Defense.


  • Backonja, M. M. (2002). Use of anticonvulsants for treatment of neuropathic pain Neurology, 59 (Suppl 2) S14-S-17.
  • Bixquert-Jimenez M, Bixquert-Pla L. (2005). Antidepressant therapy in functional gastrointestinal disorders. Gastroenterologia y hepatologia¸28 485-492.
  • Butler AC, Chapman JE, Forman EM, Beck AT.(2005). The empirical status of cognitive-behavioral therapy: A review of meta-analyses. Clinical Psychology Review
  • Coluzzi F, Mattia C. (2005).Mechanism-based treatment in chronic neuropathic pain: the role of antidepressants. Current Pharmacological Des 11, 2945-2960
  • Gendreau RM, Thorn MD, Gendreau JF, Kranzler JD, Ribeiro S, Gracely RH, Williams DA, Mease PJ, McLean SA, Clauw DJ. (2005). Efficacy of milnacipran in patients with fibromyalgia. Journal of Rheumatology, 32, 1975-1985.
  • Katz J, Pennella-Vaughan J, Hetzel RD, Kanazi GE, Dworkin RH. (2005). A randomized, placebo-controlled trial of bupropion sustained release in chronic low back pain. Journal of Pain, 6, 656-661.
  • Maizels, M., & McCarberg, B. (2005). Antidepressants and antiepileptic drugs for chronic non-cancer pain. American Family Physician, 71, 483-490.
  • Sammons, M. T. (2004). Pharmacological management of chronic pain: I. Fibromyalgia and neuropathic pain Professional Psychology: Research and Practice, 35, 206-210.
  • Sammons, M. T. (in press). Psychopharmacology Update: Treatment of head pain with psychotropics Professional Psychology: Research and Practice.
  • Tomkins, G. E., Jackson, J. L., O'Malley, P. G., Balden, E., & Santoro, J. E. (2001). Treatment of chronic headache with antidepressants: a meta-analysis. American Journal of Medicine, 111, 54-63.
  • Stahl SM, Grady MM, Moret C, Briley M. (2005). SNRIs: Their Pharmacology, Clinical Efficacy, and Tolerability in Comparison with Other Classes of Antidepressants. CNS Spectrum, 10, 732-47.
  • Zubieta, J. K., Bueller, J. A., Jackson, L. R., Scott, D. J., Xu, Y., Koeppe, R. A., Nichols, T. E., & Stohler, C. S. (2005). Placebo effects medicated by endogenous opioid activity on mu-opiate receptors. The Journal of Neuroscience, 25, 7754-7762.