Reliability, reproducibility and validity of dynamic cerebral autoregulation in a large cohort with transient ischaemic attack or minor stroke

Abstract

Objective: Cerebral autoregulation (CA) is critical to maintenance of cerebral perfusion but its relevance to the risk of stroke and dementia has been under-studied due to small study sizes and a lack of consensus as to the optimal method of measurement. We determined the reliability and reproducibility of multiple CA indices and the effect of intensive data-processing in a large population with transient ischaemic attack or minor stroke. Approach: Consecutive, consenting patients in the population-based Oxford Vascular Study (OXVASC) Phenotyped cohort underwent up to 10-min supine continuous blood pressure monitoring (Finometer) with bilateral middle cerebral artery (MCA) transcranial ultrasound (DWL-Dopplerbox). Un-processed waveforms (Un-A) were median-filtered, systematically reviewed, artefacts corrected and their quality blindly graded (optimal (A) to worst (E)). CA metrics were derived in time-domain (autoregulatory index (ARI), Pearson’s Mx, Sx, Dx) and in very-low (VLF) and low-frequency (LF) domains (WPS-SI: wavelet phase synchronisation, transfer function analysis), stratified by recording quality. Reliability and reproducibility (Cronbach’s alpha) were determined comparing MCA sides and the first vs. second 5-min of monitoring. Main results: In 453 patients, following manual data-cleaning, there was good reliability of indices when comparing MCA sides (Mx: 0.77; WPS-SI-VLF: 0.85; WPS-SI-LF 0.84), or repeated five minute epochs (Mx: 0.57; WPS-SI-VLF: 0.69; WPS-SI-LF 0.90), with persistently good reliability between sides even in lower quality Groups (Group D: Mx: 0.79; WPS-SI-VLF: 0.92; WPS-SI-LF: 0.91). Reliability was greatest for Pearson’s Mx and wavelet synchronisation index, with reasonable reliability of transfer function analyses, but ARI was prone to occasional, potentially defective, extreme estimates (left vs right MCA: 0.68). Significance: Resting-state measures of CA were valid, reproducible and robust to moderate noise, but require careful data-processing. Mx and wavelet synchronisation index were the most reliable indices for determining the prognostic value of CA in large epidemiological cohorts and its potential as a treatment target.

Reliability of the mean flow index (Mx) for assessing cerebral autoregulation in healthy volunteers

“…The reliability of Mxa has previously been assessed in healthy volunteers in several studies, which have reported highly variable repeatability and reproducibility ranging from poor to excellent (Chi et al, 2018;Lee et al, 2020;Lorenz et al, 2007;Mahdi, Nikolic, Birch, Olufsen, et al, 2017), and from poor to good (Lorenz et al, 2008;Ortega-Gutierrez et al, 2014;Riberholt et al, 2021), respectively. As a potential explanation, these studies utilized short recordings, often shorter than 6 min, the minimum duration necessary for Mxa to stabilize according to one study .…”

Reliability and validity of the mean flow index (Mx) for assessing cerebral autoregulation in humans: A systematic review of the methodology

“…24 However, another study including 167 recordings from patients with transient ischaemic attack used a similar methodology with the comparison of the first and last half of a 10-minute recording with the full recording showed a significant difference between the full recording and the last half for both sides. 25 Comparing two consecutive 5-minute recordings, in 46 healthy volunteers, resulted in an ICC of 0.39 (95%CI: 0.08-0.63) for the native recordings. 26 In another study, repeated recordings were obtained in twenty healthy volunteers during 60 s of sitting and 60 s of free-standing.…”

“…28 The actual block and epoch size are not described in the study, but the authors conclude that recordings shorter than six minutes should not be utilized. Nonetheless, all but one 29 of the studies addressing reliability used recordings shorter than these six minutes as at least one of the comparators, 5,24–27,30 and one study even used 60 s recordings. 27 This may have contributed to the poor repeatability.…”

Cerebral Vasoreactivity Changes Over Time in Patients With Different Clinical Manifestations of Cerebral Small Vessel Disease

“…Despite the multitude of alternatives to the gas challenges, existing literature lacks definitive conclusions regarding the best practices for the vasoactive modulation and associated analysis protocols for the TCD studies resulting in numerous metrics of cerebral autoregulation (Valdueza et al, 2008;Sanders et al, 2018). Regardless of methodology, an assessment of cerebrovascular autoregulation is prone to moderate noise and artifact, with low reliability and reproducibility (Lee et al, 2020). In our study, CVR was assessed using TCD which additionally suffers from a limited field of view comparing with the MRI techniques and thus it may be less reflective of the local changes in the tissue blood supply.…”