Preventing somatization

Countless cross-sectional surveys of primary and secondary care have demonstrated the ubiquitous nature of medically unexplained symptoms. If grouping diverse symptoms under one heading is appropriate, they account for over half of all new presentations in secondary medical care (Nimnuan et al. 2001a) and a sizeable proportion of ‘frequent attenders ’ in secondary care have predominantly medically unexplained symptoms (Fink, 1992a ; Reid et al. 2002). Such symptoms are costly, persistent, and associated with significant disability and psychiatric disorder (Reid et al. 2001, 2003), but are generally ignored by mental health services (Bass et al. 2001). Current definitions of medically unexplained symptoms are best seen in the context of systems of medical care. In primary care, the concept of somatization as a process is widespread and defined by Lipowski (1988) as the tendency to experience and communicate somatic symptoms that are unaccounted for by pathological findings, to attribute these to physical illness and to seek medical help. Others have emphasized how medically unexplained symptoms are an expression of underlying mood disorder (Bridges et al. 1991). In secondary medical care specialities, a number of functional somatic syndromes have been described and have proliferated in step with the fracturing of general medicine into ever smaller components. Wessely et al. (1999) have argued that conditions such as irritable bowel syndrome in gastroenterology, atypical chest pain in cardiology and fibromyalgia in rheumatology each serve the corresponding speciality’s needs for a diagnostic label for patients with unexplained symptoms, but the overlap between such syndromes is so great (Nimnuan et al. 2001b) that they are better seen as a single underlying disorder. In psychiatry the somatoform disorders have been recognized in DSM and ICD, but are arbitrary and unreliable diagnoses (Escobar et al. 1998). Finally, from the patient’s perspective, the current vogue for referring to medically unexplained symptoms may itself have a limited life-span – patients find the term offensive and prefer the label of ‘functional ’ symptoms (Stone et al. 2002). It is tempting to view the experience and presentation of unexplained symptoms as learned behaviours. Many aspects of normal illness behaviour are learned (Pennebaker, 1982) and there is often a history of medically unexplained symptoms extending back to childhood. DSM definitions of somatization disorder described it as a life-long condition. If the behaviour is learned, this implies that certain early risk factors may be especially salient. For example, Craig et al. (1994) found that patients presenting to their GP with medically unexplained symptoms were more likely than either physically ill or depressed controls to report having suffered from a physical illness during their childhoods. Furthermore, parental illness during childhood was more frequently reported than in controls. Using data from the Medical Research Council National Survey of Health and Development (a birth cohort that followed individuals born in 1946), we demonstrated that the children of parents who rate their health as poor have more medically unexplained symptoms and hospitalizations in adult life (Hotopf et al. 1999, 2000). This study also indicated that medically unexplained symptoms, such as recurrent abdominal pain during childhood, were risk factors for subsequent difficulties. However, well-defined physical disease during childhood was not associated with adult unexplained symptoms. These epidemiological studies leave several questions unanswered. First, it is not clear what aspect of parental illness matters. Is the relationship a general one, in which all chronic illness in parents is associated with unexplained symptoms in the offspring? Or is it more specific – it could