. Author manuscript; available in PMC: 2014 Jun 12.

Published in final edited form as: Arch Intern Med. 2010 Jul 12;170(13):1135–1141. doi: 10.1001/archinternmed.2010.173

Table 3. Logistic Regression Models for Relative Risk of 6-Year Substantial Cognitive Decline in Nondemented Older Persons by Serum 25(OH)D Level.

Measure of Substantial Cognitive Decline^a	Serum 25(OH)D, nmol/L^b				P Value for Linear Trend
Measure of Substantial Cognitive Decline^a	≥75	≥50 to <75	≥25 to <50	<25	P Value for Linear Trend
MMSE
Unadjusted model^c	1 [Reference]	1.28 (0.94-1.65)	1.25 (0.94-1.65)	1.85 (1.49-2.17)	<.001
Fully adjusted model^d	1 [Reference]	1.22 (0.86-1.63)	1.18 (0.80-1.47)	1.64 (1.20-2.05)	.02
Trails A
Unadjusted model^c	1 [Reference]	0.91 (0.54-1.43)	1.29 (0.88-1.78)	1.27 (0.78-1.87)	.11
Fully adjusted model^d	1 [Reference]	0.94 (0.54-1.50)	1.19 (0.76-1.73)	1.16 (0.65-1.85)	.41
Trails B
Unadjusted model^c	1 [Reference]	0.94 (0.69-1.17)	1.07 (0.86-1.26)	1.27 (1.01-1.46)	.04
Fully adjusted model^d	1 [Reference]	0.99 (0.73-1.23)	1.11 (0.87-1.32)	1.32 (1.03-1.51)	.04

Abbreviations: 25(OH)D, 25-hydroxyvitamin D; MMSE, Mini-Mental State Examination²⁹ (range, 0-30; higher score represents better function); Trails, Trail Making Tests (range, 0-300; higher score represents worse function).

Substantial cognitive decline was defined as 3 or more points on the MMSE and the worst 10% of cognitive decline or test discontinued for the Trails A and B.

Unless otherwise indicated, data are reported as relative risk (95% confidence interval).

Adjusted for baseline cognitive score only.

Adjusted for age, sex, education, baseline cognitive score, season tested, alcohol consumption, current smoking status, depressive symptoms, body mass index, total energy intake, serum vitamin E level (alpha tocopherol), and impaired mobility.