Developmental switch in prediction and adaptation to pain in human neonates

Abstract

Habituation to recurrent non-threatening or unavoidable noxious stimuli is an important aspect of adaptation to pain and indicates the ability of the brain to encode expectation of imminent nociception. However, it is not known whether the newborn brain can predict and habituate to recurrent noxious inputs. We used electroencephalography to investigate changes in cortical microstates, which represent the complex sequential processing of noxious inputs, following repeated clinically-required heel lances in term and preterm infants. Noxious stimulus repetition decreased the engagement of early sensory-related microstates and associated behavioural and physiological responses in term infants, while preterm infants did not show signs of adaptation. Nevertheless, both groups displayed a switch between different microstates at longer latencies. These data suggests that the preterm brain is capable of encoding high-level contextual differences in pain, but cannot update its prediction, which allows for adaptation, emphasising the vulnerability of this population to recurrent pain.

Clinical thresholds in pain-related facial activity linked to differences in cortical network activation in neonates

“…Microstate analysis was used in place of the traditional event-related potential (ERP) analysis at the vertex electrode, due to the additional information which can be obtained. 39 Importantly, using information from a full scalp electrode array can detect differences which are otherwise missed using ERP analysis. Microstate analysis was conducted using Ragu 23 and custom-written MATLAB scripts.…”

“…Microstate analysis was conducted using Ragu 23 and custom-written MATLAB scripts. For a full description of the analysis, see Rupawala et al 38 …”

“…Processing of a noxious stimulus requires the parallel and sequential activation of a distributed network of brain regions [5,61], with different regions subserving distinct cortical processes that are engaged hierarchically [3,13,50,51]. Whereas previous assessment of the relationship between behaviour and cortical activity has been limited to the nociceptive event-related potential (N3P3) at a single electrode (vertex) and time-point (500-700ms poststimulus), the use of global event-related topographical analysis (microstates) can capture this complex serial activation [38,55], allowing for the examination of potentially distinct cortical functions related to subclinical and clinical facial expression changes. An important consideration in this relationship is the rapid development of the neonatal brain over the final trimester of gestation [53], with different brain regions and networks developing at different rates [10,17,18].…”

“…Microstate analysis was conducted using Ragu 23 and customwritten MATLAB scripts. For a full description of the analysis, see Rupawala et al 38 First, for each group average, we calculated the global field power (GFP, standard deviation of the cortical activity across all electrodes at every timepoint). The GFP represents the overall amount of activity at each time point considering all recording electrodes simultaneously.…”

“…These differences were statistically compared with a null distribution that was obtained by generating 2 random groups of surrogate time-series data (as described in Ref. 38), refitting the microstates to the new averages, and extracting the difference parameters (5000 iterations).…”