Cerebral infarct site and affected vascular territory as factors in breathing weakness in patients with subacute stroke

Objective: A better understanding of factors influencing breathing weakness in stroke survivors would help in planning rehabilitation therapies. The main objective of this study was to determine whether the location of cerebral infarct is associated with breathing weakness in patients with subacute stroke.

Design: Cross-sectional analysis of a prospective cohort.

Patients: Consecutive patients admitted to a neurology rehabilitation unit with first-time ischaemic stroke (n = 170).

Methods: Breathing weakness was defined as > 70% reduction in maximal inspiratory and expiratory pressures (PImax and PEmax, respectively) compared with reference values. Computed tomography and magnetic resonance imaging were used to locate stroke lesions, which were classified as cortical, subcortical, cortico-subcortical, brainstem, or cerebellum. The affected cerebrovascular territory was identified to classify stroke subtype. The association between maximal respiratory pressure and affected brain area was studied using median regression analysis.

Results: Breathing weakness was detected in 151 (88.8%) patients. Those with cortical and cortico-subcortical stroke location had the lowest PImax and PEmax values (median 33 cmH2O). This value differed significantly from maximal respiratory pressures of patients with strokes located in the brainstem and the cerebellum, with PImax median differences (β) of 16 cmH2O (95% confidence interval (95% CI) 4.1-27.9) and 27 cmH2O (95% CI 7.8-46.2), respectively, and PEmax median differences of 27 cmH2O (95% CI 11.4-42.7) and 49 cmH2O (95% CI 23.7-74.3), respectively, both of which remained significant after adjustments.

Conclusion: The prevalence of breathing weakness was very high in stroke patients admitted to a neurorehabilitation ward, being more severe in cortical or cortico-subcortical stroke.

Keywords: rehabilitation; stroke assessment; stroke subtype; breathing weakness.