The DSM-III-(R) definition of somatization disorder is too restrictive
for use in general practice. A more comprehensive definition, the ''s
omatic symptom index'' (SSI) has shown good validity in open populatio
ns. However, a definition has to differentiate validly within a popula
tion of frequent attenders to be a useful diagnostic instrument in gen
eral practice. We studied a threshold of five complaints (nearly ident
ical to the SSI) in 80 Dutch general practice patients. Patients were
selected on age (20-44 years), history of back, neck or abdominal comp
laints, and on frequency of consultation-at least 12 consultations in
the previous 3 years, corrected for consultations with compelling soma
tic reason for encounter. Prevalence of somatization in this group was
45%. Women had a 2 times higher risk of somatization. A relation with
age was not found. Somatization was related to depressive complaints
(relative risk 2.5) and probably also to anxiety. Somatizing patients
consulted their general practitioner more often and had more health pr
oblems (especially psychic problems) than non-somatizers. These result
s support the validity of this definition. The distinction between our
definition of somatization and somatization defined as a symptom of p
sychiatric (e.g. depressive or anxiety) disorder is emphasized.