Table 3.

Interaction of Framingham Risk Score (FRS) components with Aβ in separate linear regressions with tau levels as outcome

Model	FRS term	Terms Estimate (SE), p
		FRS component	Aβ	Interaction of FRS component with Aβ
1	FRS-Age	0.358 (0.150), 0.017	1.413 (1.606), 0.379	−0.056 (0.107), 0.602
2	FRS-BMI	−0.018 (0.064), 0.785	0.607 (0.038), <0.001*	−0.057 (0.034), 0.096
3	FRS-SBP	0.030 (0.023), 0.187	0.654 (0.042), <0.001*	−0.033 (0.012), 0.009*
4	FRS-smoking	0.068 (0.077), 0.376	0.576 (0.027), <0.001*	−0.050 (0.040), 0.211
5	FRS-DM	0.000 (0.029), 0.998	0.575 (0.029), <0.001*	−0.003 (0.015), 0.867

Using 5 separate linear regressions with terms for FRS components, Aβ, and their interactions as predictors and tau as the outcome controlled for age at death, sex, and education (except model 1 controlled for sex and education only). In each of the 5 models, one of the FRS components has been examined: age, body mass index (BMI), systolic blood pressure (SBP), smoking, and diabetes mellitus (DM). Cells’ parameters are estimates (SE), p-values. SE, standard error. Aβ and tau square root transformed.

^*Statistically significant at a Bonferroni corrected alpha of 0.01.