Resistant somatoform symptoms: Try CBT and antidepressants
Preferred strategy for ‘mismatched’ category of illnesses.
Treatment-resistant somatoform disorders are chronic (duration >1 year), can cause significant functional impairment, and respond poorly to routine care.
In the somatoform category, DSM-IV-TR includes diverse diagnoses such as conversion disorder, hypochondriasis, pain disorder, and body dysmorphic disorder. But like mismatched shoes, these disorders do not fit together well—one reason they are often misdiagnosed and ineffectively treated. This article describes:
- debate about how to categorize somatoform disorders—as psychological or physiologic
- evidence supporting psychotherapy and antidepressants to help patients with treatment-resistant somatoform disorders.
Somatoform disorders: Interacting psychiatric and biologic processes
Psychobiologic causes of somatoform disorders are poorly understood. In a recent review, Rief and Barsky1 emphasized that somatoform symptoms such as abdominal pain, headaches, or dizziness “are not strictly mental events, but are associated with a diversity of biological processes.” They propose that the following factors might contribute to somatoform disorders.
Autonomic physiologic arousal may lead patients to misperceive the meaning of normal bodily symptoms, but most studies have been equivocal or correlate closely with changes in the cardiovascular system. For example, patients with somatoform spectrum disorders who performed mentally distressing tasks did not have the same decrease in heart rate after completing the task as normal controls did, suggesting a deficit in autonomic reactivity.
Hypothalamic-pituitary-adrenal (HPA) axis studies also have been equivocal. Some have found low cortisol in patients with somatoform disorders—suggesting commonalities with conditions such as posttraumatic stress disorder—but other studies have found normal or even elevated cortisol. Although a relationship between the HPA axis and somatoform disorders is likely, its nature remains unclear or may be indirect.
Serotonin is known to alter pain perception in major depressive disorder, so this neurotransmitter also probably plays a role in somatoform disorders. Low serotonin—mediated in part by alterations in branched-chain amino acid concentration—may be linked to increased pain perception.
Perception and filtering of body signals. A signal-filtering model of somatoform symptoms proposes that physical sensations enter consciousness influenced by numerous factors. These signals are then sent to a filter system, which itself is subject to factors that may decrease its activity. Cortical perception of distress may occur and symptoms begin to manifest if enough factors come into play.
Somatoform disorders are common in primary care. A medical utilization survey of 1,500 primary care patients found somatization symptoms in >20%.3 Controlling for comorbid psychiatric or medical illness did not change the study’s findings, which suggests that somatization is a distinct entity and not a symptom of another underlying disorder.
Little is known about somatoform disorders’ pathophysiology (Box 1),1 but their unifying theme is that psychological factors contribute to, amplify, or alter the presentation of physical illness. Not only do these disorders not form a coherent DSM category, but—as described by Mayou et al2—the lack of clearly defined thresholds between normal and pathologic behaviors is one of numerous problems that complicate diagnosis and treatment (Box 2).
Psychosomatic diad. Despite DSM-IV’s claims to etiologic neutrality, the origin of somatoform disorders’ physical symptoms clearly is meant to be psychological. As Lipowski4 said, somatization is “a tendency to experience and express somatic distress and symptoms unaccounted for by pathological findings, to attribute them to physical illness, and to seek medical help for them. It is often assumed that somatization becomes manifest in response to psychosocial stress brought about by life events that are personally stressful to the individual.”
- the difficulty in knowing when a physical symptom truly is unexplained, especially in patients with comorbid medical illness5
- the instability of somatoform diagnoses (in a cohort examined with the same questionnaire 12 months apart, 43% of “lifetime somatic symptoms” patients reported at the first screening were not reported at the second).6
Kroenke5 suggests using “physical symptom disorder” as an etiologic-neutral descriptor of unexplained physical symptoms. He would place this category on Axis III and shift the causal emphasis from psychological to unexplained. This category would replace somatization disorder, undifferentiated somatoform disorder, and pain disorder in DSM.
Similarly, Mayou et al2 contend that because most patients with somatoform disorders are treated by primary care physicians, having their disorders understood as psychiatric does not serve them well.
Psychiatric component. Conversely, patients with somatization disorder often have psychological symptoms, and many have personality disorders. The number of somatic symptoms with unexplained cause may be a normally distributed trait, with somatization disorders at the extreme end of the spectrum. Thus:
- Hypochondriasis could be reconsidered as health anxiety disorder because it features anxiety about potential illness.2
- Conversion disorders might be regrouped with other disorders focused on dissociation.2
- Body dysmorphic disorder might be regrouped with obsessive-compulsive disorder.7
These changes would shift focus away from the disorders’ physiologic presentations, emphasize the psychiatric disorders to which they likely are related, and provide insight into treatments and clinical investigations.
Pain disorder could be removed from DSM because of persistent concerns about the validity of this diagnostic category. Tyrer8 reviewed his clinical experience and reported shifting from a view that people with excessive pain had a psychiatric disorder to the view that living with chronic pain produces a profile similar to that of a person with a psychiatric disorder.
Problems with DSM categorization of somatoform disorders
- Somatoform disorders lack clearly defined thresholds that establish a difference between normal and pathologic behaviors
- Somatoform disorders do not form a coherent category, and exclusion criteria are ambiguous
- By existing, the category suggests that some disorders are physical and others are mental, leaving little room for intermediate or mixed conditions
- Patients reject the term “somatoform” because it conveys doubt about the reality of their conditions
- Somatoform disorders are incompatible with some cultures’ views of mental illness (for example, the DSM translation used in China does not include the somatoform category)
- Nonspecific somatoform illness subcategories cannot achieve established reliability standards in studies examining diagnoses
- Medical-legal cases and insurance entitlements are complicated by unclear descriptions of somatoform disorders
Physiologic component. Others recommend caution before radically altering DSM’s categorizations. Rather than shift symptoms to Axis III—as Kroenke suggests—Starcevic9 would use unexplained physical symptoms as an organizing principle and group disorders with common features, such as somatization disorder, conversion disorder, pain disorder, and undifferentiated somatoform disorder. Body dysmorphic disorder and hypochondriasis—focusing on dysfunctional appraisal of physical symptoms—would likely move elsewhere.
Treatment approach to patients with somatoform disorders
- Carefully evaluate for mood, anxiety, and substance use disorders
- Assess over-the-counter, prescription, and illicit drugs the patient may be using to control pain and other physical symptoms
- Review medical and psychiatric records, laboratory and radiographic findings
- Discuss the patient’s case with his or her primary care physician, and provide formal feedback or a consultation letter
- Treat comorbid psychiatric disorders with medications such as selective serotonin reuptake inhibitors or venlafaxine, which have improved somatoform symptoms in randomized, controlled trials
- Offer the patient cognitive-behavioral therapy (10 to 20 sessions) that focuses on reducing somatoform symptoms, and advise the primary care provider to schedule monthly follow-up intervals
Hiller and Rief10—who advocate strongly for keeping somatoform disorders in DSM—suggest 4 categories: monosymptomatic, polysymptomatic, hypochondriasis, and body dysmorphic disorder. They believe grouping diagnoses in this way would improve and refine existing nosology.
New treatment approaches
As the categorization debate continues, a treatment approach is developing that includes cognitive-behavioral therapy (CBT) and antidepressants to address the psychological and physiologic effects of resistant somatoform disorders (Box 3).
Consultation letters. Sending a consultation letter to the patient’s primary care physician is considered the standard of care (Box 4).11 In the study that introduced the consultation letter,12 patients with somatization disorder were randomly assigned to treatment (a consultation letter) or control (treatment as usual). Health care utilization costs declined approximately 50%—largely because of decreased hospitalization—when patients’ physicians received consultation letters, compared with no change for usual treatment.
Consultation letters may reduce health care spending but are less effective in improving symptoms. Evidence is changing treatment as psychotherapies have been found to help patients with somatoform disorders.
Group psychotherapy. In a controlled trial, primary care patients with somatization disorder received short-term group CBT or treatment as usual, with follow-up 6 months later. Those in the CBT group—who had received patient education and relaxation training—showed moderate but significant improvement in physical illness and somatic preoccupation, hypochondriasis, and medication use. Usual-care patients did not improve.13
CBT vs relaxation. A group of 191 inpatients described as “highly impaired” by somatization syndrome—≥8 DSM-IV somatoform symptoms—was evaluated for psychopathology, subjective health status, and life satisfaction. They then were randomly assigned to somatization-focused CBT (“soma”) or relaxation training and compared with 34 control patients. At 1-year follow-up, doctor visits had declined significantly in patients who received CBT (“soma”), and their somatoform symptoms were reduced compared with controls’.11
Psychotherapy vs listening. In a randomized, controlled trial, 102 patients with chronic refractory irritable bowel syndrome were assigned to receive exploratory psychotherapy or supportive listening. After 12 weeks, psychotherapy was more effective in improving physical and psychological symptoms, although the difference was statistically significant only in women. After 1 year, patients who received psychotherapy remained well and control patients who declined psychotherapy had relapsed.14
CBT vs usual treatment. In a randomized controlled trial, 84 patients with somatization disorder received 10 CBT sessions or treatment as usual. CBT’s goals were to:
- reduce physiologic arousal though relaxation techniques
- enhance activity regulation through increasing exercise and meaningful pleasurable activities and pacing activities
- increase awareness of emotions
- modify dysfunctional beliefs
- enhance communication of thoughts and emotions
- reduce spousal reinforcement of illness behavior.
The Clinical Global Impression Scale for Somatization Disorder showed significantly greater improvement in the group receiving CBT. Somatic symptoms as measured by patients’ self-ratings also improved more in the CBT group.15
Psychotherapy’s success in these and other studies supports the idea that somatoform spectrum disorders resemble other conditions—such as mood and anxiety disorders—that respond to psychological treatment.
Controlled trials also have shown that some antidepressants are more effective than placebo in improving somatoform symptoms.
St. John’s wort. In a randomized, placebo-controlled, double-blind trial, 184 patients with somatoform disorders but not major depression received St. John’s wort extract, 300 mg bid, or placebo. After 6 weeks, 45% of patients responded to St. John’s wort, compared with 21% for placebo (P=0.0006). Six measures determined response; St. John’s wort and placebo were equally well tolerated.16
Consultation letter for somatization: Discourage saying ‘it’s in your head’
Describe somatoform disorder, its relapsing course, and low morbidity and mortality rates
Encourage the primary care physician to:
- serve as the patient’s primary doctor and avoid fragmented care from numerous sources
- schedule regular appointments with the patient
- perform physical exams at each visit
- eliminate unnecessary tests or hospitalizations
- avoid statements such as “it’s all in your head” when medical tests are negative
Extended-release venlafaxine. A pilot study enrolled 112 adult primary care patients with multisomatoform disorder (≥3 medically unexplained, bothersome physical symptoms plus ≥2-year history of somatization) and comorbid major depressive disorder, generalized anxiety disorder, or social anxiety disorder. Patients were randomly assigned to double-blind treatment with venlafaxine ER, ≤225 mg/d (n=55), or placebo (n=57).
Primary outcome was change in the 15-item Patient Health Questionnaire (PHQ-15) somatic symptom severity score. After 12 weeks, PHQ-15 scores declined significantly (P <0.0001) in both groups but did not improve significantly more with venlafaxine ER than with placebo (-8.3 vs -6.6, respectively, P=0.097). Among secondary measures, venlafaxine ER was more effective than placebo in improving bodily pain (P=0.03), physical symptoms (P=0.02), and anxiety (P=0.02).17
Citalopram. In an 8-week trial, investigators compared the efficacy of a selective serotonin reuptake inhibitor (SSRI) and a selective noradrenaline reuptake inhibitor (SNRI) on pain symptoms in 35 patients with somatoform pain disorder. Patients were randomly assigned to double-blind treatment with the SSRI citalopram, 40 mg/d (n=17), or the SNRI reboxetine, 8 mg/d (n=18).