Effects of tilt on cerebral hemodynamics measured by NeoDoppler in healthy neonates

Abstract

Background: Today, there are conflicting descriptions of how neonates respond to tilt. Examining physiologic responses of cerebral blood flow velocities (BFVs) in challenging situations like a tilt requires equipment that can cope with positional changes. We aimed to characterize how healthy term neonates respond to mild cerebral hemodynamic stress induced by a 90° tilt test using the recently developed NeoDoppler ultrasound system.

Methods: A small ultrasound probe was fixated to the neonatal fontanel by a cap, and measured cerebral BFV in healthy neonates during and after a 90° head-up tilt test, five min in total, at their first and second day of life. Unsupervised k-means cluster analysis was used to characterize common responses.

Results: Fifty-six ultrasound recordings from 36 healthy term neonates were analyzed. We identified five distinct, immediate responses that were related to specific outcomes in BFV, heart rate, and pulsatility index the next two min. Among 20 neonates with two recordings, 13 presented with different responses in the two tests.

Conclusions: Instant changes in cerebral BFV were detected during the head-up tilt tests, and the cluster analysis identified five different hemodynamic responses. Continuous recordings revealed that the differences between groups persisted two min after tilt.

Impact: NeoDoppler is a pulsed-wave Doppler ultrasound system with a probe fixated to the neonatal fontanel by a cap that can measure continuous cerebral blood flow velocity. Healthy neonates present with a range of normal immediate cerebral hemodynamic responses to a 90° head-up tilt, categorized in five groups by cluster analysis. This paper adds new knowledge about connection between immediate responses and prolonged responses to tilt. We demonstrate that the NeoDoppler ultrasound system can detect minute changes in cerebral blood flow velocity during a 90° head-up tilt.

Fig. 1. Impact of tilt and the NeoDoppler probe.

a–c The 90° tilt test was performed by lifting the neonate from the supine position and up towards the chest of the tester. d Illustration of the impact of a head-up tilt. In the supine position, the heart and brain are in the same height. After a head-up tilt, a height difference (h) is introduced between the heart and brain resulting in a hydrostatic pressure because of the gravitational force $\left(\overrightarrow{g}\right)$. The corresponding hydrostatic pressure can be calculated from the heart–brain distance, gravitational constant, and the specific weight of whole blood and is approximately 15 mmHg in this study. e Attachment of the probe to the neonate. Each component is shown in the illustration to the right. The soft hat and the probe holder can be seen in a–c. The image to the left is by Inventas (Trondheim, Norway).

Fig. 2. Ultrasound recording with time windows.

Upper panel: Unsupervised cluster analysis revealed five distinct, initial responses to tilt (A–E). The dotted lines indicate the tilt. The tilt was performed within 2 s. Colored lines are individual recordings and the black lines are cluster means. Lower panel: A typical recording of blood flow velocity (BFV) per heartbeat during the test. The four gray boxes show the time segments (S₁–S₄) used for further analysis, and the green box the time window used for cluster analysis. The neonate’s position is illustrated by the silhouettes.

Fig. 3. Absolute parameter estimates during the test.

Linear mixed models were used to estimate mean blood flow velocity (BFV, a), pulsatility index (PI, b), heart rate (HR, c) in beats per min (BPM), and mean arterial pressure (MAP, d) at group level during four different time segments of the test (Fig. 2). The position of the neonate at each time segment is indicated by the silhouettes.

Fig. 4. Parameter estimates during the test relative to baseline.

Linear mixed models were used to estimate mean blood flow velocity (BFV, a), pulsatility index (PI, b), heart rate (HR, c), and mean arterial pressure (MAP, d) at group level during four different time segments of the test (Fig. 2). Here, the change in each parameter is shown relative to the baseline value prior to tilt (time segment S₁). The position of the neonate at each time segment is indicated by the silhouettes.