Preserved Structural Brain Health Reduces the Probability of Clinical Symptoms in Acute Ischemic Stroke
Erik Lindgren1, Kenda Alhadid2, Christina Jern3, Arne Lindgren4, Jane Maguire5, Robert Regenhardt2, Natalia Rost2, Markus Schirmer2
1University of Gothenburg, Region Vastra Gotaland and Massachusetts General Hospital, 2Massachusetts General Hospital, 3University of Gothenburg AND Region Vastra Gotaland, 4Skane University Hospital, 5University of Technology Sydney
Objective:
To assess whether structural brain health influences the probability of clinical symptoms at acute hospital admission in ischemic stroke (IS).
Background:
Structural brain health may facilitate resilience to detrimental functional consequences of a parenchymal injury, but this relationship has not been investigated in acute IS.
Design/Methods:
We included patients with imaging-verified IS from the GASROS (single center; 2003-2011) and MRI-GENIE (international multicenter; 2003-2011) cohorts. From MRI obtained at hospital admission, white matter hyperintensity (WMH) and brain volume were measured on T2-FLAIR, and acute infarct volume on DWI. Infarct and WMH volumes were normalized by brain volume, creating lesion load and WMH load for each patient. We quantified structural brain health using effective reserve (eR), defined as a latent variable based on age, WMH load, and normal appearing brain volume. Clinical symptoms were assessed at acute hospital admission using the NIHSS. In logistic regression, we model the probability of symptoms (NIHSS ≥1) as a function of infarct load across groups categorized as low (Q1), medium (Q2-Q3) and high (Q4) brain health based on eR values from the entire cohort. We calculated estimated marginal means and Tukey-adjusted pairwise comparisons to test group differences.
Results:
We included 1,036 patients with median age 67 (IQR 56-77) years (62% male). At admission, median NIHSS was 3 (IQR 1-7) and 904 (87%) patients had clinical symptoms (NIHSS ≥1). The mean predicted probability of clinical symptoms was significantly lower in the group with high brain health (0.81, 95%CI 0.76-0.86), compared to the medium (0.90, 95%CI 0.87-0.92, p<0.001) and low brain health groups (0.89, 95%CI 0.85-0.93, p=0.002). We observed no difference in probability of symptoms between the low and medium brain health groups.
Conclusions:
High structural brain health, quantified using the biomarker eR, decreases the probability of parenchymal injury being expressed as clinical symptoms at acute hospital admission in patients with IS.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.