Pd. Wang et Rs. Lin, AGE-PERIOD-COHORT ANALYSIS OF CERVICAL-CANCER INCIDENCE IN TAIWAN - DIFFERENCES BETWEEN SQUAMOUS-CELL CARCINOMA AND ADENOCARCINOMA, International journal of gynecological cancer, 7(2), 1997, pp. 106-111
To examine the etiologic distinction between squamous cell carcinoma a
nd adenocarcinoma of the cervix, the relationship between the incidenc
e of both histotypes of cancer and age at diagnosis, time period at di
agnosis and birth cohort was analyzed using data from the Taiwan Cance
r Registry. Included in the study were all cases of both histotypes oc
curring during the period 1979-90 within a population of Taiwanese wom
en aged 25 to 75 years. A log-linear model modified from the method of
Osmond and Gardner was used for the analyses. Age-period-cohort analy
ses of age effect indicated that, prior to the age group of 49-51, the
re is an almost identical incidence of both histotypes with an approxi
mate linear trend of age effect but that, after this time, there is a
divergence, with the age effect of increase slightly declining with ag
e for squamous cell carcinoma and clearly declining for adenocarcinoma
. In regard to period and cohort effects, a substantial moderation of
squamous cell carcinoma risk was exhibited with both advancing recent
calendar periods and birth cohorts, while adenocarcinoma risk also has
a moderation but to a lesser extent. The model also identified the ch
anges in female sex hormones after the menopause as a determinant of t
he differing age effects, the efficiency of Pap smear screening practi
ces as a determinant of the differing period effects, and changes in r
eproductive patterns as a determinant of the differing cohort effects.
These findings may provide clues with which to develop etiologic hypo
theses and support the contention of etiologic distinctions between bo
th histologic types.