Table 3. Results of cohort analyses: changes † in activities of daily living(%) between two interviews among Beijing elderly:

	1992(n=3257)	Change in 1994 (n=2703)			Change in 1997 (n=2002)			Change in 2000 (n=1581)
	Baseline	Imp ‡	Stable §	Dec ¶	Imp ‡	Stable §	Dec ¶	Imp ‡	Stable §	Dec ¶
Total	3.9	0.6	1.1	3.9	0.7	1.5	5.1	1.5	1.8	5.3
Age group
55-64	0.7	0.1	0.3	1.5**	0.4	0.9	2.0 **	1.2	1.0	2.5 **
65-74	2.7	0.5	0.7	3.2	0.9	1.3	3.6	0.3	1.5	6.1
75-	8.0	1.3	2.7	7.8	0.9	3.1	14.2	4.9	5.3	11.5
Sex
Male	3.3	0.4	0.6	3.3**	0.6	1.4	5.1	1.3	1.7	4.8
Female	4.4	0.8	1.6	4.5	0.8	1.5	5.2	1.6	1.9	5.7
Region
Urban	3.4	0.6	1.2	3.2	0.6	1.0	4.3 **	1.3	2.0	4.4**
Suburban	5.2	0.7	1.4	5.5	0.7	2.4	8.9	2.0	1.7	4.8
Hilly	3.9	0.3	0.3	5.1	1.1	2.6	3.6	1.5	1.0	8.0
Marital Status
Married	2.3	0.5	0.6	3.0**	0.5	1.4	3.4 **	1.3	1.4	3.9 **
Unmarried	6.8	0.9	2.2	5.6	1.2	1.8	8.5	1.8	2.7	7.7
Education
Illiterate	1.4	1.0	1.6	5.1**	0.9	2.3	6.6 **	1.6	2.0	7.0
Primary	3.1	0.4	0.7	3.0	0.3	0.7	3.7	1.4	1.4	4.5
≥ Secondary	5.3	0.0	0.8	2.6	0.8	1.0	4.1	1.2	2.0	3.2

† : changes means transition in ADLs from one statue to another statue between two interviews.

‡ : Improvement (restoration from disability compared the first interview ); § : stability (disability continued without improvement)

¶ : declination (from ability to disability).

** Compared among each stratified group within each year, χ² test P< .01.