Objectives: The purpose of this analysis was to study temporal variation in
oral health in an adult population. The cohort analysis comprised 1) estim
ation of longitudinal, cross-sectional, and time-lag differences in caries
and treatment experience of the adult population in Trondelag from 1973 to
1983 and to 1994; 2) assessment of which of the effects of age, period and
cohort was the most plausible explanation for the observed differences and
3) discussion of causal and other factors underlying the effects of age, pe
riod and cohort. Methods: Cohort analysis was used to study changes in oral
health over time. Study participants were selected by random sampling in 1
973, 1983, and 1994 (n=1959, n=3195, n=2341 respectively). The methods of d
ata collection comprised standardised clinical measurements and a self-admi
nistered questionnaire. Cohort tables were established to analyse changes i
n DFT from 1973 to 1983 and DFT and DFS from 1983 to 1994. Results: Four pa
tterns of changes in oral health were identified from adolescence to older
adulthood. They point in different directions which indicate that complex p
rocesses are present. There were two and three significant changes in the s
ubsets of analysis. Assumptions and conditions that allow identification of
separate cohort, period, and age effects appeared not to be present. The p
attern of changes was consistent with main effects of both age, period, and
cohort. The effects were discussed from the perspective of the chosen fram
e of explanatory factors. Conclusions: The changes in oral health between p
opulations in 1973, 1983 and 1994 were the combined consequences of simulta
neously occurring factors that account for both age, period and cohort effe
cts.