1. ¹Enheten för klinisk epidemiologi, Institutionen för medicin vid Karolinska sjukhuset, Karolinska Institutet, Stockholm, Sweden
2. ²Institutionen för medicinsk epidemiology, Karolinska Institutet, Stockholm, Sweden
3. ³Centrum för allergiforskning & Institutet för miljömedicin, Karolinska Institutet, Stockholm, Sweden

Correspondence: Dr S M Montgomery, Enheten för klinisk epidemiologi, Karolinska sjukhuset M9:01, SE – 171 76 Stockholm, Sweden; E-mail: Scott.Montgomery@medks.ki.se

Received 19 April 2002; Revised 31 July 2002; Accepted 9 August 2002.

Abstract

Common warts could indicate cervical cancer susceptibility, as both are caused by human papillomavirus (HPV). Eczema was also investigated, as atopic eczema has been negatively associated with warts, but non-atopic eczema may be associated with compromised host defences, as observed in patients with HIV, suggesting increased susceptibility to HPV infection and cervical cancer. 'Cervical cancer' was self-reported during an interview by 87 of 7594 women members of two longitudinal British birth cohorts. The accuracy of the diagnoses is limited by lack of confirmation using medical records. Odds ratios are adjusted for common warts and eczema in childhood; and cigarette smoking, number of cohabiting partners and social class in early adult life. The odds ratios of warts and eczema with cervical cancer are 2.50 (95% confidence interval 1.14–5.47) and 3.27 (1.95–5.49), respectively. The association of eczema with cervical cancer is independent of hay fever as a marker of atopy, suggesting the importance of non-atopic eczema. Both heavier smoking compared with non-smoking and four or more cohabiting partners compared with one/none have odds ratios for cervical cancer of 8.26 (4.25–15.10) and 4.89 (1.39–17.18), respectively. Common warts in childhood may indicate cervical cancer susceptibility; this and the relationship with eczema deserves investigation.