Prospective population-based cohort studies of white matter-lesion severity and incident stroke
| Study Group | Clinical Entry Criteria | Total Number of Patients | Follow-up (y) | Definition of WM Outcome Variable | % of Incident Strokes During Follow-up by WM Outcome Variable | Cumulative Stroke Incidence | Multivariable Adjusted Relative Risk (95% CI) | Subtype of Stroke | Subtype of Ischemic Stroke |
|---|---|---|---|---|---|---|---|---|---|
| Dutch TIA Trial Group7 | TIA or non-disabling ischemic stroke (Rankin ≤3) | 3017 | Not clearly defined | WMH vs No WMH | 15% vs 8% | 19% vs 10% (3 y) | 1.6 (1.2–2.2) | Ischemic 90% vs 89%; hemorrhage 6% vs 9%; unknown 4% vs 2% | Large artery 47% vs 45%; lacunar 33% vs 29%; vertebrobasilar 10% vs 15% |
| NASCET Group5 | Symptomatic carotid artery disease without WMH on entry CT | 596 | 6.0 (median) | Development of WMH vs no WMH | 36% vs 23.5% | Not defined | 1.5 (1.0–2.4) | Not defined | Large artery 66% vs 73%; lacunar 29% vs 21%; cardioembolic 5% vs 6% |
| ARIC Group10 | No history of stroke of TIA | 1684 | 4.7 (median) | Grade ≥3 vs grade <3* | 6.5% vs 1.3% | 6.8% vs 1.4% (5 y) | 3.4 (1.5–7.7) | Not subdivided by WM status: ischemic 78%; hemorrhage 16%; combined 6% | Not defined |
| Rotterdam Scan Study Group9 | No history of stroke or TIA | 1077 | 4.2 (mean) | Third tertile grade vs first tertile grade | Not defined | Not defined | 4.7 (2.0–11.2) (periventricular lesions); 3.6 (1.4–9.2) (subcortical lesions) | Not subdivided by WM status: ischemic 74%; hemorrhage 10%; unspecified 16% | Not defined |
| CHS Group8 | No history of stroke or TIA | 3293 | 7 (median) | Grade ≥5 vs grade ≤1† | 27.6% vs 6.0% | 2.8% vs 0.6% (annual risk) | 3.0 (1.9–4.7) | Ischemic (81%) | Lacunar (15%); cardioembolic (27%); atherosclerotic (5%); unknown (47%); mixed (6%) |
Note.— TIA indicates transient ischemic attack; WMH, white matter hypodensity; NASCET, North American Symptomatic Carotid Endarterectomy Trial Group; ARIC, Atherosclerosis Risk in Communities Study; CHS, Cardiovascular Health Study Collaborative Research Group; CI, confidence interval.
* WM grade 3 defined as thicker continuous periventricular rim with scattered patches of subcortical WM lesions.
† Grade 1 defined as discontinuous periventricular rim or minimal “dot”of subcortical disease; grade 5 defined as mild periventricular confluence surrounding the frontal and occipital horn.