Impact of age and mean intracranial pressure on the morphology of intracranial pressure waveform and its association with mortality in traumatic brain injury

Abstract

BACKGROUND: Morphological analysis of intracranial pressure (ICP) pulse waveforms provides indirect information on cerebrospinal compliance, which might be reduced by space-occupying lesions but also by intracranial hypertension and aging. This study investigates the impact of age and mean ICP on the shape and amplitude of ICP pulse waveform in traumatic brain injury (TBI). Additionally, it explores the association between morphological parameters and mortality after TBI. METHODS: ICP recordings from 183 TBI patients (median age: 50 (30, 61) years) from the CENTER-TBI database were retrospectively analyzed. ICP morphology was assessed using the artificial intelligence-based pulse shape index (PSI) and peak-to-peak amplitude of ICP pulse waveform (AmpICP). The impact of mean ICP, age, and their interaction on PSI and AmpICP were estimated using factorial ANOVA. To account for influence of disturbance in the intracranial volume on AmpICP and PSI, a multiple regression analysis was performed using age, mean ICP, and the Rotterdam CT score as explanatory variables. The associations of AmpICP and PSI with six-month mortality were assessed using the area under the ROC curve (AUC). RESULTS: Age had a predominant influence on PSI (p < 0.01), accounting for 33.1% of its variance, while mean ICP explained 6.6% (p < 0.01). Conversely, mean ICP primarily affected AmpICP (p < 0.01), explaining 22.8% of its variance, with age contributing 8.0% (p < 0.01). A combined effect of age and mean ICP on AmpICP (p = 0.01) explained 11.7% of its variance but did not influence PSI. After accounting for Rotterdam CT score, the results remained consistent, indicating that advanced age has the strongest impact on PSI (β = 0.342, p < 0.01) while elevated mean ICP has dominant influence on AmpICP (β = 0.522, p < 0.01). Both AmpICP and PSI were moderately associated with mortality (AUC: 0.76 and 0.71, respectively). CONCLUSIONS: AmpICP and PSI capture distinct aspects of cerebrospinal compliance. PSI appears to reflect age-related stiffening of the cerebrovascular system, while AmpICP, influenced by mean ICP, indicates acute volume compensatory changes. Combined, they provide a more comprehensive assessment of cerebrospinal volume-pressure compensation. Both morphological metrics are associated with mortality after TBI. As cerebrospinal compliance declines with age, older TBI patients become more susceptible to uncontrolled rises in ICP, which can worsen their outcome.