Article Text

Download PDFPDF

Breast implants and illness: a model of psychological illness
  1. M Ahern,
  2. M Smith,
  3. H Chua,
  4. P Youssef
  1. Department of Rheumatology, Repatriation General Hospital, Adelaide, South Australia
  1. Correspondence to:
    Professor M Ahern, Department of Rheumatology, Repatriation General Hospital, Flinders University of South Australia, Daws Road, Daw Park, South Australia 5041, Australia;

Statistics from

We read with interest the hypothesis of Dush in the Annals of the Rheumatic Diseases1 and provide some evidence to support it. Dush suggested that many of the symptoms of women with silicone breast implants might be attributed to somatisation, stress, and mass somatisation. We recently reviewed 179 women with silicone breast implants who were involved in product liability and litigation.

Their ages ranged from 29 to 74 years (mean (SD) 46.8 (8.1)). The indications for surgery were cosmetic in 146 (82%), followed by cancer in 17 (9%), fibrocystic disease in 12 (7%), and congenital hypoplasia in 4 (2%) women. The most common symptoms were burning breast pain in 142 (79%), chronic fatigue in 142 (79%), arthralgia in 134 (75%), sleep disturbance in 127 (71%), cognitive dysfunction in 102 (57%), sicca symptoms in 100 (56%), night sweats in 97 (54%), and myalgia in 91 (51%). Findings on clinical examination were few, including chest wall abnormalities in 60 (34%), tender trigger points in 31 (17%), and carpal tunnel syndrome in 6 (3%) women. Sixty five (36%) women had radiological and/or surgical proof of implant leakage or rupture and 61 (34%) had signs of implant contractures. We found no evidence of increased occurrence of any connective tissue disorder such as rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), or Sjögren's syndrome.

The women were asked to complete two questionnaires and return them by mail: (a) the General Health Questionnaire (GHQ), possibly the most widely used and extensively validated screening test for functional psychiatric illness—a score above 12 denoting significant psychiatric morbidity; (b) the Speilberger State-Trait Anxiety Inventory (STAI). This comprises two self report scales for measuring two distinct anxiety concepts: (i) state anxiety (A state), which may vary over time and (ii) trait anxiety (A trait), a relatively enduring personality characteristic which does not tend to change much over time.

Of the 179 questionnaires sent, 117 were returned, giving a response rate of (65%). The women had a mean (SD) score of 18.3 (8.2) on the GHQ; 75% of women having scores above 12. Table 1 shows the means (SD) of the A state and A trait.

Patients with breast implant had significantly higher A state and A trait scores than female undergraduate students (p<0.001) and medical/surgical inpatients (p<0.001) and as much anxiety reaction as psychiatric patients with anxiety disorders.

Thus these women were found to have significant psychiatric morbidity and to show significant state and trait anxiety. The causes for these high anxiety levels may be related to the reasons these women sought breast implants—for example, poor self esteem, interpersonal and psychological problems. Possibly, these high anxiety levels are exacerbated by the litigation and media attention. The raised trait-anxiety levels maybe a risk factor for somatisation.

We attempted to reassure these women that there was no evidence of any underlying connective tissue disorder or systemic disease but could not allay their fears. Nevertheless, we identified the fact that these woman were distressed and required medical and psychological rehabilitation. Unfortunately, many of these women had been “dismissed” and their symptoms not taken seriously by their treating doctors because epidemiological research had not shown any association with recognisable systemic disease.

Because of the uncontrolled nature of our observations we can only propose a hypothesis for the causes of their symptoms. As rheumatologists encountering women with silicone breast implants, we need to reassure them that they are unlikely to have an underlying connective tissue disorder. We also must provide access to treatments that will alleviate their anxiety and distress, as suggested by Dush.

Table 1

A state and A trait scores, mean (SD)


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.