Longitudinal association of edentulism with cognitive impairment, sarcopenia and all-cause mortality among older Chinese adults

Table 2 Average changes in cognitive function during 2013 to 2018 waves by edentulism in CHARLS using mixed-effects linear regression models

Variables	β-coefficient (95% confidence) Interval), cognitive function during Wave 2 to Wave 4
Variables	Model 1 ^a	Model 2 ^a	Model 3 ^a
Total sample
Edentulism
No	Reference	Reference	Reference
Yes	-0.50 (-68, -0.33) ^¶	-0.75 (-1.13, -0.38) ^¶	-0.70 (-1.09, -0.31) ^¶
Age	-0.11 (-0.12, -0.10) ^¶	-0.12 (-0.14, -0.11) ^¶	-0.12 (-0.14, -0.11) ^¶
Female	-0.53 (-0.65, 0.42) ^¶	-1.02 (-1.31, -0.74) ^¶	-1.00 (-1.31, -0.70) ^¶
Education
Primary or less	Reference	Reference	Reference
Lower secondary	2.92 (2.78, 3.06) ^¶	3.04 (2.74, 3.35) ^¶	3.15 (2.84, 3.47) ^¶
Upper secondary	4.41 (4.23, 4.59) ^¶	4.50 (4.05, 4.94) ^¶	4.64 (4.17, 5.11) ^¶
Tertiary	5.79 (5.38, 6.21) ^¶	6.27 (4.99, 7.54) ^¶	6.31 (4.89, 7.73) ^¶
Participants without stroke or heart diseases at Wave 2
Edentulism
No	Reference	Reference	Reference
Yes	-0.56 (-0.89, -0.23) ^§	-0.70 (-1.12, -0.28) ^§	-0.68 (-1.10, -0.26) ^§

^a Model 1 was adjusted for age, sex, household annual income and education; Model 2 was further adjusted for smoking, alcohol drinking, sleep duration, body mass index and dentures; in Model 3, hypertension, diabetes, respiratory illnesses, stroke and heart disease and other chronic illnesses at Wave 4 were added

ISSN: 1472-6831