Predicting Successful Aging in a Population-Based Sample of Georgia Centenarians

Table 4

Age at onset of chronic disease and lifetime prevalence in control group of 80 octogenarians versus 244 centenarians. To correct for right censoring due to GCS study termination a stepwise Cox Regression [17] from survival analysis was performed to estimate average age at onset in centenarians. The average age at onset from the Cox regression [17] was computed at the mean of the covariates.

(a)

Disease	Average Age at Onset	𝑛	Average Age at Onset from Cox regression	𝑛
Cancer∗	78±3	22	60±0.5	238
Cardiovascular	83±1	181	77±0.3	226

∗Many cancers are absent, and skin cancers were excluded because of their later onset. The 𝑛 above differs from the totals in (b) due to missing data.

(b)

Disease	No cancer	Cancer	Total
Octogenarians	58	22 (28%)	80
Centenarians	171	73 (30%)	244
Total	229	95 (29%)	324

A Fisher’s exact test [14] is not significant at 𝛼=0.05, nor is a 𝑧-test, significant, with 𝑧=0.90 on the proportions, 0.28 and 0.30.

(c)

Disease	No Cardiovascular	Cardiovascular	Total
Octogenarians	19	61 (76%)	80
Centenarians	48	196 (80%)	244
Total	67	257 (79%)	324

A Fisher’s exact test [14] is not significant at 𝛼=0.05, nor is a 𝑧-test, significant, with 𝑧=1.80 on the proportions, 0.76 and 0.80.