The epidemiology of multiple somatic symptoms

The risk factors for a high total somatic symptom count are unclear; and it is not known whether total somatic symptoms count is a predictor of impaired health status.

A prospective population-based cohort study in North West England. Randomly sampled residents (1443 participants; 58% response) completed questionnaires to determine number of somatic symptoms (SSI), health status and a wide range of risk factors; 741 completed questionnaires 1 year later. We used logistic regression to identify risk factors for high SSI at follow-up and for persistently high SSI. We used ANCOVAR and multiple regression to assess whether baseline SSI predicted health status at follow-up.

Twenty-one percent of participants scored over 25 on the Somatic Symptom Inventory (SSI) at baseline and 14% at both baseline and follow-up. Risk factors for a persistent high SSI were: fewer than 12 years of education, separated, widowed or divorced status, reported psychological abuse during childhood, co-existing medical illnesses, anxiety and depression. In multivariate analysis baseline SSI predicted health status (SF12 physical component score and health-related quality of life (EuroQol)) 12 months later. Persistent high SSI was a clinically meaningful predictor of these outcomes.

Our data support a biopsychosocial approach to somatic symptoms rather than the dualistic approach of identifying "medically unexplained" symptoms. The risk factors for total somatic symptom count were those associated with psychiatric disorders including physical illness. A persistent high somatic symptom count provides a readily measured dimension of importance in epidemiology as a predictor of health status.