The impact of indoor carbon dioxide exposure on human brain activity: A systematic review and meta-analysis based on studies utilizing electroencephalogram signals

Abstract

BackgroundThe impact of indoor carbon dioxide (CO2) exposure on human brain activity has garnered scholarly attention, presenting conflicting findings. To reconcile these disparities, we conducted a systematic review and meta-analysis. MethodsThis study extensively searched Web of Science, Science Direct, and Google Scholar until October 2023. Employing a random effects model of meta-analysis, we aimed to elucidate the impact of indoor CO2 exposure levels (below 1000 ppm, 1000–2500 ppm, 2500–4000 ppm, and above 4000 ppm) on diverse aspects of brain activity (sleep/wakefulness, low/high-frequency) based on studies utilizing electroencephalogram (EEG) signals. Subgroup analysis accounted for variables such as EEG signal sources, test states, exposure time, and subject characteristics. ResultsThe systematic review included six studies with 106 subjects, revealing a discernible impact of CO2 exposure on brain activity. Elevated CO2 concentrations during sleep were associated with extended sleep onset latency (SOL) (standardized mean difference (95 % confidence interval) (SMD (95 % CI)): 1.03 (−0.98, 3.03)) and significantly reduced slow wave sleep (SWS) duration (SMD (95 % CI): −1.57 (−2.27, −0.87)). During wakefulness, low-frequency EEG oscillatory activity showed significant activation (SMD (95 % CI): 0.67 (0.22, 1.13)), while high-frequency EEG oscillatory activity experienced significant inhibition (SMD (95 % CI): −0.66 (−0.92, −0.40)). Single-domain EEG signals predominantly reflected low-frequency EEG oscillatory activity (SMD (95 % CI): 0.72 (0.12, 1.32)), whereas multi-domain signals accurately mirrored high-frequency EEG oscillatory activity (SMD (95 % CI): −0.83 (−3.50, 1.85)). ConclusionThe results indicate a noticeable influence of CO2 exposure on brain activity, emphasizing the effectiveness of EEG signals for assessment. Notably, single-domain low-frequency EEG signals and multi-domain high-frequency EEG signals, particularly in the frontal and parietal lobes, are identified as pivotal features. Future research should include task tests, long-term exposures, and diverse subject comparisons. To enhance impact precision, larger meta-analyses are warranted.