Table 1.

Estimable APC parameters and functions calculated by the Web toola

NomenclatureInterpretation
Net driftAPC analog of the EAPC of the ASR, log-linear component of Embedded Image, Embedded Image, and Embedded Image
CAT = LAT − net driftCross-sectional age trend; log-linear trend in Embedded Image
LAT = CAT + net driftLongitudinal age trend; log-linear trend in Embedded Image
Age deviations, Embedded ImageNonlinear age effects incorporated into Embedded Image, Embedded Image, and Embedded Image; orthogonal to the linear trend in age
Period deviations, Embedded ImageNonlinear period effects incorporated into Embedded Image and Embedded Image; orthogonal to the linear trend in period
Cohort deviations, Embedded ImageNonlinear cohort effects incorporated into Embedded Image and Embedded Image; orthogonal to the linear trend in cohort (over the entire rate matrix)
Fitted temporal trends, Embedded ImageFitted rates in reference age group Embedded Image adjusted for cohort deviations; APC analog of the ASR
Cross-sectional age curve, Embedded ImageFitted cross-sectional age-specific rates in reference period Embedded Image adjusted for cohort deviations
Longitudinal age curve, Embedded ImageFitted longitudinal age-specific rates in reference cohort Embedded Image adjusted for period deviations
Ratio of longitudinal versus cross-sectional age curves Embedded ImageQuantifies influence of net drift on age-associated natural history
Period rate ratios, Embedded ImageRatio of age-specific rates in period Embedded Image relative to reference period Embedded Image
Cohort rate ratios, Embedded ImageRatio of age-specific rates in cohort Embedded Image relative to reference cohort Embedded Image
Local drifts, Embedded ImageEstimated annual percentage change over time specific to age group Embedded Image
  • aFor the APC model defined over Embedded Image age groups and Embedded Image calendar periods with equal intervals. In the Web tool, the central age group, calendar period, and birth cohort define the reference values Embedded Image, respectively. When there is an even number of age, period, or cohort categories, the reference value is the lower of the two central values. In the R package, reference values can be set to any observed age, period, and cohort = period–age in the rate matrix.