Table 2.

Assessment of the relative contributions of birth cohort and calendar period on NHL incidence (1964-2003)

ModelAIC SwedenAIC DenmarkAIC FinlandDEV SwedenDEV DenmarkDEV Finland
Age + sex4,0022,7983,4022,7841,6792,292
Age + sex + linear1,6261,4241,428405302315
Age + sex + cohort1,5801,3861,408328232264
Age + sex + period1,4501,3681,342218234217
Age + sex + period +cohort1,4261,3201,342160154185
  • NOTE: Comparisons of Poisson models, including sex, age (5-year bands, 30-84 years of age), calendar period, and birth cohort (overlapping 10-year bands). Age, calendar period, and birth cohort were all treated as categorical variables. Linear denotes a continuous linear trend not specified as of birth cohort or calendar period origin (or both). Aikake Information Criterion was used as a qualitative measure of model goodness of fit that accommodates for differences in model complexity, with lower values indicating better fit. Deviance (−2 times the log likelihood ratio) was used as a formal measure for goodness of fit, with lower values indicating better fit.

    Abbreviations: AIC, Akaike Information Criterion; DEV, deviance.