TABLE 2.

Log odds and standard errors of the effects of twin type and postnatal covariates on dementia risk in female twins

	Model 1	Model 2	Model 3
Testing from age 60
Twin type	−0.19 (0.13)		−0.15 (0.13)
Education		−0.04 (0.01)	−0.04 (0.01)
Exercise		−0.002 (0.08)	0.00 (0.08)
Vascular risk		0.03 (0.02)	0.03 (0.02)
Postnatal hormone exposure		0.01 (0.002)	0.01 (0.002)
Testing from age 70
Twin type	−0.44 (0.15)		−0.40 (0.15)
Education		−0.04 (0.01)	−0.04 (0.01)
Exercise		0.01 (0.08)	0.02 (0.08)
Vascular risk		0.04 (0.02)	0.04 (0.02)
Postnatal hormone exposure		0.01 (0.002)	0.01 (0.002)
Testing from age 80
Twin type	−0.96 (0.29)		−0.94 (0.30)
Education		−0.06 (0.01)	−0.06 (0.01)
Exercise		0.05 (0.12)	0.05 (0.12)
Vascular risk		0.02 (0.03)	0.02 (0.03)
Postnatal hormone exposure		0.002 (0.003)	0.002 (0.003)

Note. Bolded values indicate P < 0.05. Parameter estimates are pooled estimates across 25 imputed data sets. Birth year and the interaction between birth year and twin type were statistically adjusted for in all models.