Mild Cognitive Impairment: Statistical Models of Transition Using Longitudinal Clinical Data
Table 4
Estimated relative risks and 95% confidence intervals for one-step transitions to test-based amnestic MCI (aMCITB), test-based mixed MCI (mMCITB), or clinical consensus MCI (MCICC) versus the base state of normal cognition (bolding denotes statistical significance).
Risk factor*
aMCITB versus Normal
mMCITB versus Normal
MCICC versus Normal
Age
1.02 (1.01–1.04)
1.07 (1.05–1.08)
1.12 (1.09–1.15)
Female sex (versus male)
0.77 (0.62–0.95)
1.01 (0.82–1.24)
0.71 (0.46–1.09)
Family history of dementia (yes versus no)
0.81 (0.65–1.00)
0.76 (0.62–0.94)
1.04 (0.66–1.64)
≥one APOE-4 allele (versus none)
1.04 (0.83–1.31)
1.12 (0.89–1.40)
1.89 (1.21–2.95)
≤12 years of education (versus >12 years)
1.24 (0.89–1.74)
1.79 (1.33–2.42)
2.20 (1.24–3.91)
History of hypertension (yes versus no)
0.95 (0.76–1.18)
1.04 (0.84–1.28)
0.79 (0.42–1.49)
aMCITB at prior assessment (versus normal)
1.15 (0.91–1.45)
1.00 (0.77–1.29)
2.28 (1.30–4.00)
mMCITB at prior assessment (versus normal)
0.76 (0.57–1.02)
4.51 (3.63–5.61)
4.80 (2.94–7.81)
*As risk factors do not depend on the prior state, covariate effects are the same regardless of whether transitions occur from a prior state of normal cognition, aMCITB, or mMCITB.