Neurology

Exercise and intensive vascular risk reduction did not improve 2‑year cognition in at‑risk older adults

In this 24‑month, 4‑arm randomized trial, 513 adults aged 60–85 years at risk for dementia received aerobic exercise, intensive vascular risk reduction (IRVR), both, or usual care. Global cognition by PACC score improved modestly in all groups, with no significant differences between exercise and no‑exercise arms. Similarly, IRVR targeting systolic blood pressure <130 mm Hg and LDL lowering with atorvastatin produced no additional cognitive benefit over non‑IRVR care. These findings suggest that, over 2 years, structured aerobic exercise and pharmacologic vascular intensification alone do not meaningfully alter global cognition in non‑demented, at‑risk older adults. 1

rTMS and TBS provide the strongest noninvasive stimulation benefits for post‑stroke unilateral spatial neglect

This systematic review and meta-analysis included 17 randomized or crossover trials (425 patients) testing rTMS, TBS, tDCS, or tACS for post‑stroke unilateral spatial neglect. rTMS significantly improved line bisection, star cancellation, and Catherine Bergego Scale scores with large effect sizes versus control. TBS also improved star cancellation and Catherine Bergego Scale scores, while tDCS had modest benefits and tACS showed no significant effect. Subgroup analyses suggested greater benefit when stimulation was delivered in the acute phase and when using excitatory protocols over ipsilesional posterior parietal cortex. Overall, rTMS and TBS appear most promising for clinically meaningful neglect reduction and ADL gains, warranting protocol‑driven use in stroke rehabilitation programs. 2

Machine learning using central vein sign, cortical, and rim lesions markedly improves MRI-only MS diagnosis

This multicenter retrospective study applied 71 machine learning models to MRI biomarkers in 322 MS/mimic and 84 prodromal/non‑MS adults. Models combined central vein sign, cortical lesions, and paramagnetic rim lesions and were compared with McDonald dissemination-in-space criteria. Fifty‑one models significantly outperformed dissemination-in-space, improving balanced accuracy by up to 13 percentage points. The best simplified logistic regression model achieved about 95% balanced accuracy and generalized robustly across two external test sets, including prodromal presentations. A web-based implementation using simplified biomarker assessments could help clinicians reduce MS misdiagnosis using routine advanced MRI sequences. 3

Post‑thrombectomy systolic blood pressure drops independently predict worse 3‑month functional outcome

This registry-based cohort included 300 patients with successful mechanical thrombectomy (mTICI ≥2b) and high‑resolution BP monitoring for 24 hours post‑procedure. A systolic blood pressure drop ≥40 mm Hg within 1 hour occurred in 41% of patients, often multiple times. BP drops independently predicted worse 3‑month modified Rankin Scale outcomes, with an adjusted common odds ratio of 1.66 for any drop. Each additional drop further worsened functional outcome, and drops doubled the odds of early neurological deterioration. BP drops were not associated with mortality, suggesting a specific impact on disability rather than survival. These data support tight avoidance of abrupt post‑thrombectomy hypotension in stroke unit and ICU protocols. 4