Incidence of HPV-related anogenital precancer and cancer in women with diabetes – a nationwide registry-based cohort study
Menée au Danemark à l'aide de données de registres portant sur 2 508 321 femmes nées entre 1916 et 2001, cette étude analyse l'association entre la présence d'une néoplasie intraépithéliale cervicale, vaginale, vulvaire ou anale de grade 2 à 3 et le risque de cancer, en fonction notamment du statut diabétique et de la vaccination contre le papillomavirus humain
In this register-based cohort study, we estimated the incidence of human papillomavirus (HPV)-related anogenital precancer and cancer in women with diabetes compared with women without diabetes. We followed all women living in Denmark born 1916–2001 (n=2,508,321) for individual-level information on diabetes (type 1 or 2), diagnoses of cervical, vaginal, vulvar and anal intraepithelial neoplasia grades 2–3 (IN2/3) and cancer, and other covariates from nationwide registries. We used Poisson regression to model the incidence rates of anogenital IN2/3 and cancer as a function of diabetes status, age, HPV vaccination, education, calendar year, and cervical cancer screening status. Incidence rate ratios (IRRs) were estimated for diabetes overall, and separately for type 1 and type 2 diabetes, compared with women without diabetes. Women with diabetes had higher rates of vulvar IN2/3 (IRR=1.63; 95% CI: 1.41–1.88), vulvar cancer (IRR=1.61; 95% CI: 1.36–1.91) and vaginal cancer (IRR=1.79; 95% CI: 1.27–1.91) than women without diabetes. Similar patterns were observed for anal IN2/3, anal cancer and cervical cancer, although not statistically significant. In contrast, women with diabetes had lower rates of cervical IN2/3 (IRR=0.74; 95% CI: 0.69–0.79) than women without diabetes. Patterns were generally similar in women with type 1 and type 2 diabetes, although cancer rates were higher in women with type 2 diabetes. In conclusion, the incidence of most anogenital precancers and cancers were increased in women with diabetes. However, women with diabetes had lower incidence of cervical precancer. Our findings could be explained by biological mechanisms and/or behavioral factors, such as smoking and less frequent cervical screening participation.