Background: White matter hyperintensities (WMH) feature as a marker of long-term white matter
(WM) degeneration, caused by a broad range of pathologies (see Maillard et al., 2012). WMH can
be detected on T2-weighted fluid attenuated inversion recovery (FLAIR) magnetic resonance imaging
(MRI) scans and progressively increase with age (de Leeuw et al., 2001). Age-related WMH
are strongly associated with vascular risk factors and are assumed to cause a global decline in cognitive
performance. The progression of WMH majorly affects attention and executive functioning
(Kloppenburg et al., 2014). However, a comprehensive, domain-specific characterization of the neuropsychological
features associated with WMH is needed to validate recent findings and to further
identify cognitive patterns associated with WMH.
Here, we aim at the identification of neuropsychological characteristics associated with WMH in a
large, non-clinical sample.
Methods: We included 850 persons participating in the study of the Leipzig Research Centre for
Civilization Diseases at the University of Leipzig, Germany (M = 60 years, SD = 13.1 years, range:
21-79 years).
MRI was acquired for every participant with a 3 Tesla MRI-scanner. The amount of WMH on the
individual FLAIR sequences was quantified on the 4-stage Fazekas scale (Fazekas, 1987) by
experienced neuroradiologists. The sample was categorized in 4 Fazekas groups. Age, sex and
education were identified as confounding factors.
The neuropsychological test battery included the Trail Making Test (parts A and B), the Stroop Test,
the Consortium to Establish a Registry of Alzheimer’s Disease (CERAD) test battery and the
20-item Dysexecutive Questionnaire (DEX) of the Behavioral Assessment of the Dysexecutive
Syndrome (BADS). Neuropsychological subtests were assigned to the cognitive domains attention,
executive function, memory, learning, language, verbal fluency and perceptual-motor abilities,
under well-established theoretical considerations (Beck et al., 2014; Madureira, 2006) and with
respect to the clinical diagnostic criteria for mild and major neurocognitive disorder (DSM-5,
American Psychiatric Association, 2013).
Individual test results were age-standardized (M = 0, SD = 1) to the mean of the corresponding age
group (<65 y, 66-69 y, 70-74 y, 75+ y). Age-standardized scores corresponding to the same cognitive
domain were averaged to seven cognitive domain scores indicative of objective cognitive
performance (see Madureira, 2006).
A content based categorization of DEX items to five cognitive domains (attention, executive
function, memory, language and social cognition) was applied. The sum score for every domain was
age-standardized to the mean of the corresponding age group.
Differences between the four Fazekas groups in (1) objective cognitive performance (7 measures)
and (2) subjective cognitive complaint (5 measures) were analyzed with the nonparametric
rank-sum test with data-alignment (critical alpha level: p < .05; see Bortz et al., 2008). Data was
aligned for the effects of sex and education and their interaction with Fazekas score. Data was
corrected for identical ranks. Group differences were examined using the single comparison
algorithm proposed by Schaich & Hamerle (1984; see Bortz et al., 2008). For single comparisons,
the alpha level was adjusted due to multiple comparisons (critical alpha level: 0.005).
Results: A general decline in cognitive performance is associated with a higher Fazekas score. The
decline in cognition becomes evident with higher lesion load (starting at Fazekas score 2). Fazekas
groups significantly differed in performance measures of attention (χ²(3,846) = 13.77, p < .01) and
executive functions (χ²(3,846) = 12.84, p < .01). Analyses yielded marginally significant group
differences in objective measures of memory (χ²(3,438) = 7.74, p = .05) and visuoconstructive
abilities (χ²(3,434) = 7.73, p = .05). In all four cognitive domains, performance significantly
declines with larger WM lesion load (Fazkas score 2), while low lesion load (Fazekas score 1) was
not associated with a decline in cognitive performance.
Fazekas groups significantly differed in measures of memory complaint (χ²(3,796) = 18.8, p < .001)
and complaint on executive function (χ²(3,796) = 39.88, p < .001). For both measures, persons with
low lesion load (Fazekas score 1) worried significantly less than healthy persons (Fazekas 0) and
persons with larger lesion load (Fazekas score 2, 3).
Conclusion: Results show a substantial decline in cognitive performance with larger lesion load,
which becomes evident at progressed stages of WMH (starting from Fazekas score 2). In line with
recent literature (Kloppenborg, 2014), attention and executive function are assumed to be major
dimension of cognitive decline. Furthermore, results suggest that small, punctual lesions (Fazekas
score 1) are less likely to be associated with impaired cognitive performance. A similar relation has
also been found for subjective cognitive performance. Memory and executive functions are majorly
affected by WMH and correspond to the results of objective cognitive performance. Surprisingly,
persons with small WM lesions tend to be less worried about their cognitive performance than
healthy persons. However, larger lesion load (Fazekas score 2 & 3) was associated with increased
subjective complaint.
This large cohort study contributes to the framework of age related WM changes and their
association with domain specific cognitive performance and subjective cognitive complaints.