New Insights on the Treatment of Hypochondriasis

Treating patients with hypochondriasis, who worry that they have a serious disease when in fact they do not, can be challenging. Many seek nonpsychiatric medical treatment, even though their illness fears and beliefs are unwarranted. Indeed, these patients have substantially elevated rates of medical care utilization, and psychosocial functioning is often impaired (1, 2).

In this issue, Fallon, Barsky, and colleagues—leading researchers on this disorder—report a landmark treatment study of this often difficult-to-treat condition (3). They randomly assigned 195 patients with DSM-IV hypochondriasis to one of four treatments: fluoxetine, cognitive-behavioral therapy (CBT), pill placebo, or joint treatment with both fluoxetine and CBT. This is the first study to examine the effects of combined CBT and a serotonin reuptake inhibitor (SRI) for hypochondriasis. CBT and SRIs have each been shown to be efficacious, although only two prior randomized controlled SRI trials have been done (2, 4).

The results supported the predicted pattern of response (p=0.036); response rates were 47.2% for joint therapy (fluoxetine plus CBT), 41.8% for single active treatment, and 29.6% for placebo. There was no significant difference in the proportion of responders across groups. However, secondary analyses showed that on one of two continuous measures of hypochondriasis, treatment group was significant overall (p=0.049); pairwise contrasts revealed that fluoxetine was significantly more effective than placebo and had a significantly faster rate of improvement than placebo. CBT was not more effective than placebo, however. When considering those participants who completed the study (a secondary analysis), the fluoxetine group had the highest response rate (81.3%) compared with 62.2% for joint treatment, 51.7% for CBT, and 44.0% for placebo.

It is somewhat surprising that joint treatment with both fluoxetine and CBT showed only a small incremental advantage over fluoxetine alone in intention-to-treat analyses and actually yielded a lower response rate than fluoxetine alone in secondary completer analyses. As the authors note, a possible explanation is that the mean endpoint fluoxetine dose in the joint treatment group was lower than that of the fluoxetine group. Consistent with this explanation, on one outcome measure improvement was significantly greater for those receiving ≥40 mg/day than for those receiving less than 40 mg/day.

It is possible that response rates would have been even higher if patients had received higher doses of fluoxetine. The mean doses of 40 mg/day in the fluoxetine group and 31 mg/day in the joint treatment group are lower than those used in a prior study of fluoxetine for hypochondriasis (4); they are also lower than doses often used and needed to successfully treat disorders with similarities to hypochondriasis, such as obsessive-compulsive disorder and body dysmorphic disorder (distressing or impairing preoccupation with nonexistent or slight defects in appearance) (5, 6). In fact, the American Psychiatric Association Practice Guideline for Obsessive-Compulsive Disorder states that doses as high as 120 mg/day of fluoxetine are occasionally used for obsessive-compulsive disorder (5), as is also the case for body dysmorphic disorder (6).

SRI dosing for obsessive-compulsive disorder and body dysmorphic disorder may be relevant to hypochondriasis because these three conditions share obsessional preoccupation and excessive repetitive behaviors (i.e., rituals, compulsions), such as checking one’s body for signs of illness in hypochondriasis (2, 7). Furthermore, hypochondriasis is often conceptualized as an obsessive-compulsive or related disorder (also known as an obsessive-compulsive spectrum disorder) (7). During the development of DSM-5, the controversial question of where to classify hypochondriasis was carefully examined (7). Some research evidence supported moving it to DSM-5′s new chapter “Obsessive-Compulsive and Related Disorders”; however, evidence for its placement in this chapter was more mixed and less persuasive than for other disorders, such as body dysmorphic disorder (7). Ultimately, hypochondriasis was not moved to this chapter, although the upcoming 11th edition of the International Classification of Diseases will likely classify hypochondriasis in a chapter on obsessive-compulsive and related disorders (8)—a reflection of these disorders’ similarities, including, perhaps, their treatment response.

Study results for CBT were relatively weak compared with those in prior studies (9). An alternative way of viewing the results, however—the “glass half full” perspective—is that 52% of study completers responded to CBT, a high proportion after only six sessions. As the investigators note, the relatively low response rate in intention-to-treat analyses may be explained by the fairly high dropout rate (one-third of participants did not complete at least six sessions) and the brevity of CBT treatment. Lengthening treatment, which clinicians would likely do if response were insufficient after six sessions, may have improved outcomes. Indeed, a meta-analysis found that more CBT sessions were associated with larger effect sizes at posttreatment (number of sessions varied from 3 to 16) (9). CBT could be lengthened and perhaps strengthened by further practicing already-learned CBT skills during additional sessions or by adding other cognitive or behavioral approaches. Additional approaches might include exposure to reduce avoidance of anxiety-provoking situations and ritual prevention to decrease repetitive behaviors such as body checking or reassurance seeking if present (approaches used for obsessive-compulsive disorder and body dysmorphic disorder) (2). The relatively high dropout rate suggests that strategies to enhance treatment retention might be valuable. Approaches such as motivational interviewing—developed for treatment of substance use disorders and subsequently adopted for a broad range of conditions—might facilitate patient engagement and retention in treatment (2, 10). Motivational interviewing may be especially well suited to hypochondriasis because insight—although not well studied in this condition—appears to often be poor, which may decrease patients’ willingness to receive psychiatric treatment.

DSM-5 made changes to hypochondriasis after this study began (11). DSM-5 does not use the name “hypochondriasis,” and it classifies DSM-IV hypochondriasis either as somatic symptom disorder or as illness anxiety disorder; both are in DSM-5′s chapter on somatic symptom and related disorders. Somatic symptom disorder highlights distressing or disruptive somatic symptoms (it also corresponds closely to DSM-IV’s somatization disorder). In contrast, illness anxiety disorder emphasizes preoccupation with having or acquiring a serious physical illness; it puts less emphasis on physical symptoms, requiring them to be only mild or not present at all. DSM-5′s definition of illness anxiety disorder overlaps with that of obsessive-compulsive disorder, especially when physical symptoms are absent, concerns focus on a fear of acquiring rather than actually having a serious illness, and excessive repetitive behaviors (compulsions) are present. This overlap suggests use of higher doses of fluoxetine (if needed and tolerated) for this form of illness anxiety disorder in particular. The study findings likely apply to illness anxiety disorder and somatic symptom disorder, but studies are needed to ascertain that this is the case.

No study is perfect, and the authors are to be congratulated for successfully conducting this important trial. Studies like this one—large randomized efficacy trials with multiple treatment arms that use state-of-the-art methodology—are difficult to do. Recruiting a large number of patients who meet all inclusion/exclusion criteria and are willing to accept their assigned treatment, retaining them in the study, and implementing and adhering to quality assurance procedures require tremendous effort and expertise. An additional recruitment and retention challenge is that patients with hypochondriasis may not readily recognize the potential value of psychiatric care (1, 2).

There is a pressing need for additional treatment studies of somatic symptom and related disorders. These conditions cause patients to suffer, and they challenge clinicians and health care systems. In the meantime, for patients with clinically significant illness worries, this study offers valuable insights about helpful treatment approaches.