Abnormalities of cerebral white matter and stroke risk in the Framingham Offspring Study
Background:MRI Abnormalities of cerebral white matter (WMH) are associated with multiple cerebral vascular risk factors, such as hypertension. The association between WMH and risk for stroke, however, has been investigated only in a limited fashion. The Framingham stroke risk profile (FSRP), based on information from the original Framingham Study Cohort (FHS), has been shown to estimate age- and gender specific probability of stroke in a number of populations. To test the utility of WMH to predict future stroke, we related the FSRP to WMH in 909 offspring of the original FHS (FHSo) who were free of clinical stroke. Methods:Quantification of MRI WMH was performed in the years 1999–2000 on 919 individuals of the FHSo who had taken Exam 5 during 1991–95 when they were ages 35–74 (mean age 55 ±9 years). None were demented. Data from 10 individuals with prior stroke were excluded. FSRP was computed using Exam 5 risk factor data. Using linear regression, age-related differences in WMH volume, corrected for differences in head size, were computed. WMH volumes >1 sd above the age-predicted mean were considered large. Analyses employed T-test comparison according to WMH category (large vs normal) as well as multiple logistic regression analysis to predict category. Results:43 individuals (4.7%) were identified as having large volumes of WMH. These individuals had significantly higher SBP (132 ±19 mmHg vs 124 ±18 mmHg; p=0.003) and significantly higher FSRP (0.070 ±0.057 versus 0.040 ±0.043; p=0.002) at Exam 5. Using logistic regression, the odds of having large WMH volumes increased by 1.26 (95% CL 1.08–1.48; p=0.004) for each 10mmHg increase of SBP and 1.09 (95% CL 1.05–1.14; p=0.001) for each 1% increase in FSRP. Analysis by gender showed a similar pattern although SBP did not reach statistical significance for WMH in women (128.4 +19.1 versus 121.7 +18.7, p=0.068). Conclusion:Large WMH are seen in men and women with previously measured, significantly increased predicted stroke risk. Further prospective evaluation will be important to determine if the extent of WMH is a predictor of risk for stroke as suggested by the increased values of FSRP amongst those with large WMH.