Participants�� blood pressure (2 times) was also measured during selleck chemical Wortmannin the interview. The average of the two blood pressure measures were used in our analyses. Blood spots on filter paper were collected from all study participants at the end of the interview. Apolipoprotein E (APOE) genotype was determined by eluting DNA from a dried blood spot  followed by Hhal digestAPOEion of amplified products . Statistical analysis In order to capture potential nonlinear association between baseline cognitive function and mortality, we divided the study population into quintiles according to baseline cognitive z-score levels. BMI was modeled using the World Health Organization criteria into four groups: underweight (BMI<18.5), normal weight (18.5��BMI<25), overweight (25��BMI<30), and obese (BMI��30) .
The following variables were considered to be potential confounders possibly related to mortality: age at baseline, gender, education (whether the participant attended school), marital status, household composition, alcohol consumption and smoking history, systolic and diastolic blood pressure measures, APOE genotype (?4 carriers vs. non-carriers) and history of comorbidities, such as cancer, Parkinson��s disease, diabetes, hypertension, stroke, heart attack, head injury and fracture. Analysis of variance models (ANOVA) were used to compare differences in continuous variables, and chi-square tests were used to compare differences in categorical variables across the five quintile groups defined by cognitive z-scores.
Kaplan-Meier estimator was used to estimate survival probabilities for the cohort stratified by cognitive quintile and BMI groups, respectively. Univariate Cox��s proportional hazard model was used to examine the association between each covariate and mortality risk. Multivariate Brefeldin_A Cox��s proportional hazard model was used to explore the association between cognitive scores, BMI and mortality adjusting for significant covariates identified in univariate models. Results Of the 2000 participants enrolled at baseline, 144 died prior to the 2.5 year follow-up evaluation and 329 participants died between the 2.5 year and the 7 year follow-up evaluations. In Table 1, we present baseline characteristics of study participants by cognitive quintile groups. Age, gender, education, marital status, household composition, smoking, consuming alcohol, BMI, history of hypertension and heart attack were significantly different among cognitive quintile groups.