Abstract
BACKGROUND: The purpose of this study is to investigate whether changes in brain microstructure, detected by magnetization transfer imaging, are associated with cognition in older subjects at increased risk for vascular disease. METHODS: One hundred ninety three nondemented subjects (105 men, mean age 77 ± 3 years) from the Prospective Study of Pravastatin in the Elderly at Risk were included. To assess cross-sectional associations between magnetization transfer ratio (MTR) peak height and cognitive test scores, general linear model multivariate analysis was performed. Models were adjusted for age, sex, education level, vascular risk factors, individual white matter lesion volume, and brain atrophy. A repeated measures general linear model was used to investigate whether MTR peak height relates to cognitive test performance at baseline and 3.3-year follow-up. RESULTS: Cross-sectionally, MTR peak height was associated with performance on the STROOP test (unstandardized β = -0.27, p = 0.045), delayed Picture Word Learning (PWL) test (β = 0.48, p = 0.007), and the Letter Digit Coding test (β = 1.1, p = 0.006). Repeated measures general linear model analysis showed that individuals with low MTR peak height at baseline performed worse on the STROOP test compared to subjects with intermediate MTR peak height (mean time to complete the test at baseline and follow-up, lower versus middle tertile of MTR peak height: 61.6 versus 52.7 s, p = 0.019) or compared to subjects with high MTR peak height (p = 0.046). Similarly, low MTR peak height was associated with worse performance on the immediate (lower versus middle tertile, p = 0.023; lower versus higher tertile, p = 0.032) and delayed PWL test (lower versus middle, p = 0.004; lower versus higher, p = 0.012) at baseline and follow-up testing. CONCLUSIONS: MTR peak height is associated with cognitive function in older subjects at increased risk for vascular disease.