MRI white matter hyperintensities and neuropsychological performance in a clinical sample.

OBJECTIVE

White-matter hyperintensities (WMH) are commonly observed on MRI of non-demented patients. Findings are mixed regarding their association with neuropsychological test performance. The purpose of this study is to investigate the association of white-matter hyperintensities on routine clinical MRI and neuropsychological test performance in non-demented outpatients.

METHOD: Two groups were selected based on MRI results: (1) normal (n = 62, Mage = 50.21, Medu = 14.89) and (2) WMH without other MRI abnormality (n = 56, Mage = 55.43, Medu = 14.04). Neuropsychological tests assessed five cognitive domains for which index scores were calculated and categorized in the following clinical ranges: well below average, below average, low average, average, and above average.

RESULTS

Likelihood ratios comparing base rates for the WMH and normal groups across these clinical ranges revealed significant base rate differences only for attention/processing speed (Lχ2 = 16.47, df = 4, p < .01), with more WMH patients in the below average range and fewer above average. Odds ratios were calculated using two z-score cutoffs: -1.67 and -1.00. While patients with WMH were significantly more likely to have an index z-score ≤ -1.00 on attention/processing speed tests (OR = 3.62, 95% CI: 1.08-12.19) and an executive function test of reasoning (OR = 4.63, 95% CI: 1.18-18.19), there was no difference in the likelihood the groups would have a z-score ≤ -1.67 in any cognitive domain.

CONCLUSIONS

Taken together, these findings indicate that among referred outpatients without dementia, WMH on routine clinical MRI are associated with relatively mild decreased attention and processing speed.