Skip to main content

Table 3 Potential confounders stratified by donepezil use

From: Frameworks for estimating causal effects in observational settings: comparing confounder adjustment and instrumental variables

 

Donepezil (\(n=134\))

Control (\(n=483\))

SMD\(^a\)

Cardiovascular Issues - Yes (n (%))\(^{c}\)

96 (71.6)

342 (70.8)

0.018

Neurological Issues - Yes (n (%))\(^{c}\)

43 (32.1)

171 (35.4)

0.070

Renal Issues - Yes (n (%))\(^{c}\)

67 (50.0)

210 (43.5)

0.131

Age (mean (SD))

74.92 (6.69)

72.96 (7.71)

0.271

Sex - Male (n (%))

81 (60.4)

282 (58.4)

0.042

Years Education (mean (SD))

15.81 (2.65)

16.23 (2.67)

0.158

APOE4 Count (n (%))

  

0.306

  0

60 (44.8)

283 (58.6)

 

  1

56 (41.8)

166 (34.4)

 

  2

18 (13.4)

34 (7.0)

 

Time in ADNI (mean (SD))

1.19 (2.03)

0.41 (1.50)

0.437

A\(\beta\) (mean (SD))\(^{b,c}\)

781.21 (354.35)

1096.50 (441.29)

0.788

P-tau (mean (SD))\(^{b,c}\)

29.61 (12.61)

26.17 (12.63)

0.272

Starting ADAS-cog (mean (SD))\(^{c}\)

11.90 (4.46)

8.88 (3.96)

0.714

  1. \(^a\)Standardized mean difference across groups
  2. \(^b\)Measured at ADNI entry - may not coincide with their baseline start date
  3. \(^c\)Includes imputed values from one of the five imputed datasets