Should general psychiatry ignore somatization and hypochondriasis?

Abstract

This paper examines the tendency for general psychiatry to ignore somatization and hypochondriasis. These disorders are rarely included in national surveys of mental health and are not usually regarded as a concern of general psychiatrists; yet primary care doctors and other physicians often feel let down by psychiatry's failure to offer help in this area of medical practice. Many psychiatrists are unaware of the suffering, impaired function and high costs that can result from these disorders, because these occur mainly within primary care and secondary medical services. Difficulties in diagnosis and a tendency to regard them as purely secondary phenomena of depression, anxiety and related disorders mean that general psychiatry may continue to ignore somatization and hypochondriasis. If general psychiatry embraced these disorders more fully, however, it might lead to better prevention and treatment of depression as well as helping to prevent the severe disability that may arise in association with these disorders.

Somatoform disorders have been described recently as "one of the most controversial and challenging areas of modern psychiatry" (1). In spite of this claim, psychiatrists and health service planners tend to neglect these disorders (2,3). They are not usually included in national surveys of mental health and unmet need (4-6), and accurate estimates of burden have not been established (3). This brief review aims to examine some of the reasons underlying this state of affairs.

DIAGNOSTIC DIFFICULTIES

One main reason why somatization disorder has been omitted from population-based surveys is its apparent rarity. The original DSM-III definition of the disorder was so restrictive that only 15 people out of a general populationbased sample of 3798 were found to have the disorder (prevalence of 0.38%) (7). This, together with the prolonged questioning required to establish it, probably accounts for the fact that the diagnosis has been dropped from most subsequent national surveys.

In fact, a systematic review found that the prevalence of somatization disorder and hypochondriasis, as defined by DSM-III, was so low in population-based or primary care samples that researchers could not examine associated features reliably (8). Instead, many researchers abandoned the DSM-III diagnosis of somatization disorder in favour of more practical definitions such as abridged somatization (4 somatoform symptoms in men and 6 in women) or multisomatoform disorder (9,10). Subsequently, the DSMIV allowed a diagnosis of "undifferentiated somatoform disorder", in which one or more medically unexplained symptoms lead to distress or impairment.

This multiplicity of definitions has led to wide variation in the prevalence of somatization disorder. For example, in a primary care study, less than 1% of patients met DSM-IV criteria for somatization disorder, 6% met abridged criteria, 24% met the criteria for multisomatoform disorder and 79% met the criteria for undifferentiated somatoform disorder (11). These proportions are similar to those reported in two systematic reviews (8,12). Other studies have found a prevalence of ICD-10 somatization disorder of 2.7- 10.1% in primary care (13,14). This lack of consensus about the diagnosis and its prevalence does not encourage planners of population-based surveys to include somatization disorder. Much work remains to be done to establish the appropriate number of symptoms and associated impairment to define the disorder satisfactorily.

A similar situation pertains to the diagnosis of hypochondriasis. This also appears to be rare in non-specialist settings, so that alternative diagnostic criteria have been developed (8). One study found that 2% of primary care patients met diagnostic criteria for hypochondriasis, but 14% were bothered by the thought of having serious undiagnosed disease (15). In fact, disease conviction alone is an independent predictor of both impaired functioning and medical help seeking (16,17) and "abridged" hypochondriasis is no less disabling than the full diagnosis of hypochondriasis even after controlling for concurrent depression (17). On this basis, subthreshold hypochondriasis was included in a trial of cognitive behavioural therapy for hypochondriasis (18).

The relationship between illness worry and health anxiety is not clear, and some researchers have used a definition that includes both the number of somatic symptoms as well as disease fear, bodily preoccupation, and disease conviction (19).

The concept of "disorder" in relation to somatization and hypochondriasis has to be established in the light of the fact that both the number of bodily symptoms and illness worry are distributed continuously in primary care or population-based samples. There is a linear relationship between the number of bodily symptoms and impairment and healthcare use (15,20,21), though this may not be so for illness worry (21). In the light of present knowledge, the precise cut-off points which are clinically useful in primary and secondary care are not clear. Any new diagnostic classification (e.g., DSM-V) should encourage further research rather than, like the DSM-IV, artificially impose a cut-off which may later prove to be incorrect.

CONCURRENT ANXIETY AND DEPRESSIVE DISORDERS

Although there is good evidence that depression and anxiety are closely associated with the number of somatic symptoms (22), two large studies in primary and secondary care found that medically unexplained symptoms as a whole were not associated with these psychiatric disorders (23,24). Even one third of those patients attending primary care who report five or more medically unexplained symptoms do not have a psychiatric disorder (25). In secondary care medical clinics, where medically unexplained symptoms are more common (35-50% of all attendees), two-thirds do not have anxiety, depressive or related disorders (21,24,26-28). So, it is quite wrong to think of numerous somatic symptoms solely as a manifestation of depressive or anxiety disorders, and these data do not support the notion that somatization disorder does not really exist and can be subsumed into anxiety and depression disorders (29). Even more compelling is the fact that successful treatment of somatization or hypochondriasis with cognitive behavioural therapy or antidepressants is not necessarily accompanied by a significant reduction of anxiety and depression (30-32).

This notion came about because an early definition of somatization required the presence of a psychiatric disorder underlying the somatic symptoms (33) and influential studies of somatization in primary care identified patients with depressive and anxiety disorder who presented somatic symptoms to the general practitioner (33,34). This method of patient recruitment does not include the many primary care patients who have "medically unexplained" symptoms but do not have anxiety and depressive disorders (12).

The real importance of concurrent anxiety and depressive disorders in relation to somatization and hypochondriasis is that the associated impairment of functioning is greater when both occur concurrently - the two types of disorder appear to be additive in their effect on functioning (35). If the concurrent depressive disorder is successfully treated, functioning improves (36).

It must be recognized, therefore, that the relationship of anxiety and depression with somatization or hypochondriasis is complex. We can recognize at least three groups of patients with somatization or hypochondriasis. In one group the bodily symptoms or preoccupation are accompanied by anxiety or depression, and when the latter is successfully treated the somatic symptoms improve. A second group also has concurrent anxiety or depression, but successful treatment of the latter may lead to improved functioning while the somatic symptoms change little (30,32,36,37). A third group have definite somatization or hypochondriasis, but do not have concurrent anxiety or depression; the cognitive or perceptual changes which underlie somatization or hypochondriasis occur alone.

WHY SHOULD GENERAL PSYCHIATRY PAY MORE ATTENTION TO SOMATIZATION?

There are several reasons why general psychiatry should be more interested in somatization and hypochondriasis.

First, some general psychiatrists and public health doctors are concerned about the high prevalence and poor treatment of depressive disorder. They need to be aware that depressive disorder occurring in conjunction with numerous somatic symptoms is less likely to be recognized and treated than depression that presents psychologically (38,39). Secondly, there is some evidence that somatization may precede the onset of depressive disorder (40), and paying more attention to marked somatization in primary care may provide us with a rare opportunity to prevent the onset of depressive disorder. Third, somatization and illness worry are independent predictors of frequent attendance at primary care - even after depression and anxiety are controlled (16,41,42). We need to understand the reasons for this as it may help reduce the burden of these disorders on the healthcare system.

Lastly, the aetiologies of somatization and depression are closely allied but not identical. Therefore, research about the causes of somatization and depression together may be more informative than research into depression alone. For example, there are preliminary findings that somatization and depression may be genetically and environmentally distinct, but with overlapping genetic predisposition (43) and a different relationship with stress - adversity being associated with somatization and depression with other life events (44). New onsets of somatoform conditions, which include somatization, have been associated with female gender, lower social class, the experience of any substance use, anxiety and affective disorder, and traumatic sexual and physical threat events (45). Understanding in more detail the independent and overlapping risk factors for depression and somatization will help us to develop more effective interventions.

An example is found in reported history of sexual abuse. This is said to be a risk factor for somatization as well as depression, but we do not know whether it is a risk factor for "pure" somatization, i.e. that which occurs without concurrent depression. On the other hand, it does appear that somatization mediates the association between reported history of sexual abuse and frequent healthcare use more precisely than depression, suggesting that there may be a specific link between abuse and sensitivity to bodily symptoms (46,47). Furthermore, we found that both reported abuse and depression independently predicted increased pain threshold when improvement in severe irritable bowel syndrome patients occurred following psychotherapeutic or antidepressant treatment (48). Many of these patients fulfilled the criteria for somatization disorder.

In fact, the evidence concerning childhood adversity as a risk factor for somatization or hypochondriasis is conflicting (49,50), and such adversity may predict a person's response to bodily symptoms (leading to persistent symptoms and marked functional disability) rather than their onset (51).

RECENT RESEARCH FINDINGS

Several myths regarding somatization have being overturned by recent research, indicating the need for a fresh look at the phenomenon.

Previous research has considered somatization primarily in relation to bodily symptoms which are "medically unexplained, disabling and/or lead to medical help seeking". This alone could have accounted for the association with functional impairment and healthcare use of many previous studies. In fact, the number of bodily symptoms and degree of illness worry are associated with impairment and healthcare use whether they are medically explained or not (16,21,52,53); the relationship is a general one and not confined to patients with medically unexplained symptoms.

It has long been thought that somatization disorder and hypochondriasis were chronic disorders, but in primary care both disorders resolve in approximately 50% of people over a year (8). As mentioned above, resolution may be associated with reduction of anxiety and depressive symptoms, but this is not a consistent finding (8,54).

The traditional idea that somatization is more common in developing, rather than developed, cultures has not stood up to scrutiny, with high incidence rates being reported from European and South American centres in a World Health Organization study (54).

Patients with somatization or related disorders differ from those with an anxiety or depressive disorder in the attribution of their symptoms to "organic", "bodily" or external causes, such as a germ, virus, environment, ageing or genetic vulnerability (55-58). Patients with both somatoform and depressive or anxiety disorders have both psychological and "organic" attributions. A change from somatic to psychological attribution has been associated with improved outcome (59), whereas persisting dysfunctional somatic attributions in irritable bowel syndrome patients have been associated with continuing frequent medical consultations (60). If we understood better how these attributions develop, we would be in a much better position to treat people successfully.

Recent qualitative research suggests that general practitioners' responses to patients' symptoms, rather than patients' demands, may underlie the ordering of investigations for organic disease, which carry the potential to reinforce somatic attributions (61). This may provide an opportunity to prevent persistent somatization.

CONCLUSION

The material reviewed above supports the notion that somatoform disorders are one of the most controversial and challenging areas of modern psychiatry (1). These disorders are particularly challenging to psychiatrists exposed only to patients who have become severely disabled by chronic disorder. Under these circumstances psychiatrists often protest that early intervention might have prevented numerous secondary problems. Such early intervention will not be achieved unless general psychiatrists who also work in primary or secondary general medical care encourage medical and nursing staff to adopt early and appropriate management of all patients who present numerous bodily symptoms or marked illness worry.

The challenges presented to research are also formidable. Longitudinal, population-based studies are necessary if we are to understand the risk factors for these disorders and their associated features of impaired functioning and high healthcare use. At present such studies are very limited. Better diagnostic accuracy is essential, and it is to be hoped that the authors of DSM-V draw upon the appropriate expertise and most up-to-date research in formulating sensible criteria to avoid the mistakes of DSM-IV (62). Further treatment studies are needed, but these need to be carefully designed to account for the different degrees of severity seen in primary and secondary care (63) and in patients with and without concurrent depressive disorders. It is to be hoped that the growth of knowledge in this area will encourage more multidisciplinary research groups to include these disorders in their research endeavours.

Until the clinical and research effort is greatly increased, somatization and hypochondriasis will remain a major public health concern because of considerable unmet need. It will be unfortunate if these disorders only receive attention because of the high health costs they incur - the suffering experienced by so many people deserves help in its own right. No other group is likely to research and improve the care of this group of people. Therefore psychiatrists, who understand much about their unmet need, should take up the challenge.