Skip to main content
. Author manuscript; available in PMC: 2011 Jun 1.
Published in final edited form as: Arch Neurol. 2009 Mar;66(3):343–348. doi: 10.1001/archneur.66.3.343

Table 2.

Separate GEE models measuring decline in cognitive function associated with vascular risk factors

Predictors Outcome: Composite Cognitive z-score
Age-Adjusted Multivariable-Adjusted*
B p-value β p-value
Total Cholesterol 0.000 0.95 −0.001 0.31
Time*total cholesterol −0.001 0.04 −0.001 0.03
HDL (log transformed) 0.345 0.02 0.219 0.15
Time*HDL −0.024 0.61 −0.057 0.20
Triglycerides (log transformed) −0.165 0.15 −0.146 0.22
Time*triglycerides −0.013 0.66 −0.006 0.85
LDL 0.001 0.70 −0.001 0.31
Time*LDL −0.001 0.04 −0.001 0.045
Diabetes −0.105 0.37 −0.083 0.46
Time*Diabetes −0.051 0.03 −0.050 0.045
Hypertension −0.207 0.16 −0.210 0.14
Time*Hypertension 0.012 0.77 0.000 0.99
History of Stroke −0.026 0.88 −0.052 0.77
Time*Stroke −0.045 0.29 −0.043 0.31
History of Heart Disease 0.049 0.69 0.023 0.91
Time*Heart Disease −0.072 0.07 −0.058 0.16
Ever Smoked (baseline) −0.106 0.28 −0.035 0.68
Time*Ever Smoked 0.027 0.36 0.029 0.33

All models also included the following covariates, entered simultaneously: time, age at incidence, sex, race/ethnicity, education (years), and study cohort. Models for lipid variables additionally controlled for and time between the lipid draw and dementia incidence. Time was expressed in years.

Significant (p<0.05) coefficients are bolded.