Age-incidence relationships and time trends in cervical cancer in Sweden

Age-incidence relationships are informative of carcinogenic mechanisms. The se have been previously assessed for cervical squamous cell carcinoma (SCC) but not for adenocarcinoma. The aim was to assess by means of age-, period - and cohort- specific analyses and Poisson regression modelling whether th e two types of cervical cancer show an age-incidence maximum at a relativel y young age, as shown in cross-sectional analyses. The Swedish Family-Cance r Database was used to analyse age-incidence relationships in cervical SCC and adenocarcinoma diagnosed in years 1958-1996, including a total of 15,11 8 and 1866 cases, respectively. Area of residence and socio-economic status were included in analyses because they were risk factors of cervical cance r. The analysis of cervical SCC confirmed an incidence maximum at ages 35-3 9 years. The data for adenocarcinoma also suggested a similar early age max imum but the curves differed extensively by birth cohort. The incidence of adenocarcinoma increased substantially at young age groups towards the end of follow-up. Endometrial adenocarcinoma and vaginal and vulvar SCC, which share some risk factors with cervical cancer, did not show an early age inc idence maximum. The results also showed that there was a decrease in the in cidence of cervical SCC around year 1960, almost 10 years before the organi zed population screening, probably due to introduced opportunistic pap test ing. The benefits of the organized screening were observed as a further dec line in the incidence rates. The unique age-incidence relationships in cerv ical cancer call for biological explanations.