Oxygen supplementation and cognitive function in long-COVID

by Christine Gagnon, Thomas Vincent, Louis Bherer, Mathieu Gayda, Simon-Olivier Cloutier, Anna Nozza, Marie-Claude Guertin, Patricia Blaise, Isabelle Cloutier, Alan Kamada, Stanislav Glezer, André Denault, Jean-Claude Tardif

Patients can experience persistent cognitive complaints and deficits in long-COVID. Inflammation and capillary damage may contribute to symptoms by interfering with tissue oxygenation.

This was an exploratory pilot crossover study designed to describe the effects of supplemental oxygen (portable oxygen concentrator, POC) on cognitive performance and peripheral and cerebral oxygen saturation at rest and exercise. Participants with long-COVID (n = 21) were randomized 1:1 to: 1) POC (3h/day) for 2 weeks followed by standard of care (Control) for 2 weeks or 2) Control for 2 weeks then POC (3h/day) for 2 weeks, with a 1-week washout. Cognitive assessment (global cognition [Montreal Cognitive Assessment, MoCA], episodic memory [Hopkins], working memory [Digit Span], executive function [Verbal fluency]) was performed at baseline and after each treatment period. Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder-7 were completed. Peripheral and cerebral oxygen saturation were measured at rest and exercise (treadmill) at baseline and after each treatment period. Statistical analyses were descriptive without formal testing.

MoCA scores were similar under POC (26.45±2.31) and Control (26.37±2.85); overall POC-Control difference was -0.090 (95% CI [-1.031, 0.850]). Because of a learning effect, post-hoc analyses were performed for Period 1, where the MoCA score difference was 1.705 [0.140, 3.271]. MoCA subscores suggested better performance with POC for Visuospatial/executive (0.618 [-0.106, 1.342]) and Attention (0.975 [0.207, 1.743]). POC trended to have better scores on Digit Span backward (difference: 0.822 [-0.067, 1.711]) and self-reported depressive symptoms (difference: -1.335 [-3.166, 0.495]). For specific PHQ-9 items, POC tended to have lower (better) scores for Q1 (Little interest/pleasure) and Q7 (Trouble concentrating). Cerebral oxygen saturations at end of exercise showed no difference between POC and Control. Peripheral saturations during exercise were similar under POC and Control (difference: 0.519% [-1.675, 2.714]).

An advantage of POC over Control was observed for global cognition, attention, visuospatial/executive performance and depressive symptoms. Results need to be validated in a larger study.