Skip to main content

Table 1 Comparison of age-period-cohort submodels for the incidence of non-Hodgkin's lymphoma to separate contributions from each of the time variables (age group 20-89 years, 1980-2005, Doubs region, France).

From: Age-period-cohort modelling of non-Hodgkin's lymphoma incidence in a French region: a period effect compatible with an environmental exposure

Terms in model

Da (df)

Effect

ΔD (Δdf)

p value

Age

1848 (1812)

-

-

-

Age + Drift

1764 (1811)

δb|A

84 (1)

< 10-15

Age + Period

1727 (1805)

Pc|A

37 (6)

< 10-5

Age + Period + Cohort

1721 (1799)

Cc|A, P

6 (6)

0.46

Age + Cohort

1753 (1805)

Pc|A, C

32 (6)

< 10-4

Age + Drift

1764 (1811)

Cc|A

11 (6)

0.08

  1. a deviance
  2. b drift or linear secular trend
  3. c curvature or non-linear effect
  4. d the submodels are arranged in a sequence that gives all the relevant tests as comparisons between adjacent lines (using the difference in deviances and the F test). The successive tests (all adjusted for age) refer therefore to the drift, the non linear effect of period, the non linear effect of cohort (adjusted for period), the non linear effect of period (adjusted for cohort) and the non linear effect of cohort.
Back to article page

Environmental Health

ISSN: 1476-069X