The term somatization has been used to refer to a variety of clinical phenomena. One traditional view defines somatization as an inability or unwillingness to express emotional distress,⁽¹⁾ so that somatic symptoms are an alternative ‘idiom of distress’. An alternative view defines somatization as the presentation of somatic complaints to medical providers in the presence of an occult anxiety or depressive disorder.⁽²⁾ A third view defines somatization as somatic symptoms, which have no clear medical explanation.⁽³⁾ While these definitions appear closely related, they identify somewhat different groups of patients. The third definition (presentation of unexplained somatic symptoms) is used by official systems of classification and by most epidemiological studies, so this review will focus on that phenomenon.

Both the ICD and DSM classification systems define somatization disorder as a chronic condition characterized by the reporting of numerous unexplained somatic symptoms.^{(4, 5)} Recent versions of both classification systems identify a core syndrome of somatization (a persistent tendency to report multiple unexplained somatic symptoms) using a simplified set of diagnostic criteria.

The reported prevalence of well-defined somatization disorder appears to depend significantly on the method used for assessment. Community and primary care surveys have typically relied on structured interviews to assess the lifetime prevalence of unexplained somatic symptoms. Community surveys in North America⁽⁶⁾ and Western Europe^{(7, 8)} have found prevalence rates of less than 2 per cent with primary care surveys finding only slightly higher prevalence rates.⁽⁹⁾ Data from the World Health Organization (WHO) multicentre primary care survey indicate that recall during structured interviews may significantly underestimate the lifetime prevalence of somatization symptoms.⁽¹⁰⁾ More accurate recall of lifetime symptoms (by either repeated assessments or the use of medical records) might yield significantly higher prevalence rates.

The prevalence of somatization disorder and unexplained somatic symptoms is typically twice as high in women as in men,^{(11, 12)} and this difference appears at time of menarche.⁽¹³⁾ Community and primary care surveys demonstrate a substantial overlap between somatization disorder and anxiety and depressive disorders.^{(7, 14, 15)} Anxiety and depressive disorders also predict the subsequent onset of somatization disorder.⁽¹⁶⁾

Available data show a mixed picture regarding cross-national or cross-cultural differences in the prevalence of somatization. Studies of clinical samples find that somatic symptoms are a common accompaniment of depressive and anxiety disorders worldwide.^{(17, 18 and 19)} The WHO primary care survey documented large differences in the prevalence of unexplained somatic symptoms with a markedly higher prevalence in South America than in Europe or the United States.⁽⁹⁾ That same study, however, found that the association between unexplained symptoms and symptoms of depression or anxiety was similar across a wide range of cultures and levels of economic development.⁽¹⁴⁾ One explanation for these apparently disparate findings is that the prevalence of unexplained somatic symptoms (like the prevalence of anxiety or depressive disorder) varies widely across nations and cultures, but the association between somatic and psychological distress is universal. Countries or cultures with higher rates of anxiety or depressive disorders would be expected to have higher prevalence of somatization disorder and other somatization syndromes. Given the consistent overlap between somatization disorders and other common mental disorders, some have questioned whether these conditions actually belong in a distinct category.^{(20, 21)}

Available data do not support a specific diagnostic threshold based on the number or distribution of unexplained somatic symptoms. An increasing number of somatic symptoms is consistently associated with increases in comorbid mood or anxiety disorder, functional impairment, and use of health services.^(14,15) Mindful of this continuum, both Escobar et al.⁽²²⁾ and Kroenke et al.⁽²³⁾ have described less restrictive somatization syndromes, which, despite their higher prevalence, are strongly associated with impairment and the use of health services. Both the ICD and DSM classification systems describe subthreshold or less extreme forms of this condition characterized by a smaller number of medically unexplained symptoms.^{(5, 24)}

Longitudinal data raise questions about the presumed stability or chronicity of somatization disorder or medically unexplained somatic symptoms. Traditional descriptions of somatization disorder emphasize its stability and chronicity. Data from the WHO primary care survey, however, suggest that individual somatization symptoms vary considerably over time.⁽⁹⁾ While the syndrome of somatization seemed somewhat more stable than anxiety and depressive disorders (typically regarded as episodic), only half of the primary care patients, satisfying Escobar’s criteria for somatization syndrome at the baseline assessment, continued to meet the criteria one year later.

Both the ICD and DSM classification systems define hypochondriasis by the triad of disease conviction, functional impairment, and refusal to accept appropriate reassurance.

Attempts to estimate the prevalence of hypochondriasis have been limited by the absence of proven standardized methods for standardized assessment. Community surveys find prevalence rates of 1 per cent or less,⁽²⁵⁾ while primary care surveys typically find rates of approximately 5 per cent,^(26,27) while the WHO multicentre primary care survey⁽²⁸⁾ found an overall prevalence of only 0.8 per cent. In reviewing data from the WHO survey, Gureje et al.⁽²⁸⁾ found that a less restrictive definition more than doubled the prevalence rate (to 2.2 per cent). Cases added by this relaxed definition did not differ significantly from those satisfying CIDI/ICD criteria, suggesting that CIDI/ICD criteria may be somewhat too restrictive.

Despite the variation in prevalence, primary care surveys yield similar results regarding demographic correlates of hypochondriasis. The prevalence of hypochondriasis is 1.5 to 2 times as great in women as men but does not appear to vary significantly with age.⁽²⁸⁾

While the ICD and DSM classification systems suggest that hypochondriasis is distinct from anxiety and depressive disorders, available data suggests considerable overlap. In every sample examined, hypochondriasis is strongly associated with major depression, panic disorder, and generalized anxiety disorder.^{(26,28, 29 and 30)} Among those with hypochondriasis, clinical features do not clearly distinguish those with and without a comorbid psychiatric diagnosis.⁽³⁰⁾ In addition, changes over time in anxiety or depression are consistently associated with parallel changes in symptoms of hypochondriasis.⁽³¹⁾ As with somatization disorders, some have recently argued that hypochondriasis be re-classified as a form of anxiety disorder.^(20,21)