. Author manuscript; available in PMC: 2010 Sep 1.

Published in final edited form as: J Geriatr Psychiatry Neurol. 2008 Dec 10;22(3):188–195. doi: 10.1177/0891988708327888

Table 3.

Logistic Regression Models Illustrating the Odds of Cognitive Impairment (95% Confidence Intervals) by Quartiles of Serum 25-hydroxyvitamin D [25(OH)D] for Men and Women Separately^a

	Men (n = 708)			Women (n = 1058)
	Unadjusted Model	Basic Adjusted Model^b	Fully Adjusted Model^c	Unadjusted Model	Basic Adjusted Model^b	Fully Adjusted Model^c
Serum 25(OH)D (nmol/L)
Fourth (66–170)	Ref	Ref	Ref	Ref	Ref	Ref
Third (45–65)	2.86 (0.92–8.92)	2.54 (0.83–7.80)	2.60 (0.84–8.07)	0.97 (0.52–1.81)	0.85 (0.45–1.59)	0.86 (0.47–1.58)
Second (31–44)	6.91 (2.45–19.55)	6.01 (2.00–18.11)	6.83 (2.24–20.80)	1.16 (0.62–2.19)	0.82 (0.42–1.60)	0.88 (0.46–1.66)
First (8–30)	11.37 (4.16–31.09)	9.39 (3.30–26.67)	9.57 (3.18–28.79)	2.73 (1.57–4.76)	1.61 (0.88–2.95)	1.43 (0.78–2.61)
P for linear trend	<.001	<.001	<.001	<.001	.056	.139

Population weights are used to adjust for the sampling design. Cognitive impairment was defined as a score <8 on the Abbreviated Mental Test.²³

Adjusted for age, education, ethnicity, and season tested.

Adjusted for age, education, ethnicity, season tested, current smoking status, alcohol consumption, psychiatric morbidity, impaired mobility, hypoalbuminemia, and medical history (stroke, heart disease, diabetes, and hypertension).