Table 3.

Associations of nocturnal oxygen desaturation with brain structural measurements (n=660).

	Gray matter (ml)	White matter (ml)	White matter lesions (SD)a
	B (95% CI), p	B (95% CI), p	B (95% CI), p
Oxygen desaturation (n/min)
Age, sex adjusted	−4.3 (−11.0; 2.3), 0.20	−10.4 (−18.0;−2.8), 0.01	0.04 (−0.14; 0.21), 0.67
Multivariable adjustedb	−4.6 (−11.7; 2.7), 0.22	−8.3 (−16.7;−0.02), 0.049	−0.004 (−0.20; 0.19), 0.97
Mediator modelc	−4.6 (−11.8; 2.7), 0.22	−8.4 (−16.7;−0.03), 0.049	−0.004 (−0.20; 0.19), 0.97
Low oxygen desaturation, n=220	0 (reference)	0 (reference)	0 (reference)
Moderate oxygen desaturation, n=221
Age, sex adjusted	−1.4 (−6.6; 3.8), 0.60	−1.5 (−7.5; 4.6), 0.63	−0.09 (−0.23; 0.05), 0.19
Multivariable adjusted	−1.1 (−6.5; 4.2), 0.67	−0.7 (−6.8; 5.4), 0.83	−0.11 (−0.25; 0.04), 0.14
High oxygen desaturation, n=219
Age, sex adjusted	−2.8 (−8.1; 2.4), 0.29	−6.1 (−12.2;−0.1), 0.047	−0.00 (−0.14; 0.14), 0.99
Multivariable adjusted	−2.6 (−8.2; 3.1), 0.37	−4.2 (−10.7; 2.3), 0.20	−0.03 (−0.18; 0.12), 0.70

Abbreviations: AHI, apnea-hypopnea index; CI, confidence interval; SD, standard deviation. Values represent difference in brain tissue volume per unit increase in oxygen desaturation. Linear regression analyses adjusted for intracranial volume.

^aNormalized and standardized.

^bAdditionally adjusted for body mass index, education, smoking, alcohol use, diabetes, myocardial infarction and interval between brain scan and polysomnography study.

^cAdditionally adjusted for systolic blood pressure and total cholesterol.