To the Editor.
—The Women's Health Study described by Dr Hennekens and colleagues1 reported a 24% increase in the risk of connective-tissue disease in women with breast implants. However, the authors failed to consider an important point—the possibility of selection bias into the study of only the healthiest women. The consent form of the study requires that the participants had to agree to not take aspirin or other drugs with aspirinlike effects for 5 years. Any women with connective-tissue disease or chronic pain would have difficulty agreeing to this study condition. Aching joints and muscles are frequent complaints of women with breast implants and are usually self-medicated with these drugs.The effect of this condition would be a strong self-selection bias out of the study. A woman may not return the questionnaire once she had read the consent form—indeed, less than 25% responded. A large underestimate of the actual
Karns ME, Cullison CA. Breast Implants and Connective-Tissue Disease. JAMA. 1996;276(2):101-102. doi:10.1001/jama.1996.03540020023017