Neurovascular coupling on trial: How the number of trials completed impacts the accuracy and precision of temporally derived neurovascular coupling estimates.

Abstract

Standard practices for quantifying neurovascular coupling (NVC) with transcranial Doppler ultrasound (TCD) require participants to complete one-to-ten repetitive trials. However, limited empirical evidence exists regarding how the number of trials completed influences the validity and reliability of temporally derived NVC metrics. Secondary analyses was performed on 60 young healthy participants (30 females/30 males) who completed eight cyclical eyes-closed (20-seconds), eyes-open (40-seconds) NVC trials, using the "Where's Waldo?" visual paradigm. TCD data was obtained in posterior and middle cerebral arteries (PCA and MCA, respectively). The within-day (n = 11) and between-day (n = 17) reliability were assessed at seven- and three-time points, respectively. Repeat testing from the reliability aims were also used for the concurrent validity analysis (n = 160). PCA metrics (i.e., baseline, peak, percent increase, and area-under-the-curve) demonstrated five trials produced excellent intraclass correlation coefficient (ICC) 95% confidence intervals for validity and within-day reliability (>0.900), whereas between-day reliability was good-to-excellent (>0.750). Likewise, 95% confidence intervals for coefficient of variation (CoV) measures ranged from acceptable (<20%) to excellent (<5%) with five-or-more trials. Employing fewer than five trials produced poor/unacceptable ICC and CoV metrics. Future NVC, TCD-based research should therefore have participants complete a minimum of five trials when quantifying the NVC response with TCD via a "Where's Waldo?" paradigm.