RESUMEN
BACKGROUND AND PURPOSE: Cross-sectional studies showed an inverse association between serum 25-hydroxyvitamin D (25OHD) and white matter hyperintensities (WMHs) whereas the few longitudinal studies did not. The association between baseline 25OHD and WMHs at 10-year follow-up in the Heinz Nixdorf Recall Study plus 1000BRAINS was investigated. METHODS: Data of 505 participants (49% women, 56.2 ± 6.6 years) with 25OHD at baseline (2000-2003) and WMH volume and grade of WMHs using the Fazekas classification at 10-year follow-up were analysed. The association between deseasonalized 25OHD and the base-10 logarithm of WMH volume was evaluated by multiple linear regression, adjusted for age, sex, education, smoking, alcohol consumption, sports, diabetes mellitus, systolic blood pressure and total cholesterol. ß-estimators were transformed back (10ß ). Using multiple logistic regression, odds ratios (ORs) and 95% confidence intervals (95% CI) were calculated to evaluate the association between deseasonalized 25OHD and Fazekas grades (0, absence and 1, punctate foci vs. 2, beginning and 3, large confluence). RESULTS: Mean 25OHD was 17.0 ± 8.2 ng/ml, and mean deseasonalized 25OHD was 16.9 ± 7.5 ng/ml. Mean WMH volume was 16.6 ± 17.4 ml, range 1-132 ml. Most grade 2-3 WMHs were found to be periventricular (39% of the participants), parietal (32%) and frontal (31%) (temporal 6%, occipital 3%). The linear regression showed an inverse association between 25OHD and WMH volume. On average, a 25OHD increase of 1 ng/ml was associated with a reduced WMH volume by a factor of 0.99 (95% CI 0.98; 1.00) (fully adjusted). There was also some indication for an inverse association between 25OHD and extent of periventricular (OR 0.98 [95% CI 0.96; 1.01]), frontal (0.99 [0.97; 1.02]) and parietal (0.98 [0.95; 1.00]) WMHs according to the Fazekas classification. CONCLUSIONS: Lower 25OHD may be a risk factor for the occurrence of WMHs.