Modern buildings often have automatized ventilation systems to reduce energy consumption, which leads to a higher air tightness of the building envelope and limit the control humans have over the ventilation. Therefore, there is an increasing research field on the impact of indoor air quality on human cognition and health. The indoor concentration of carbon dioxide is of particular interest because it can be used as a proxy for indoor air quality. However, past studies suggested that carbon dioxide itself can be a contributor to impaired cognition, but the findings are inconclusive due to several limitations [1]. Either a carbon dioxide exposure level higher than usually occurring indoors was used or the exposure time was short. Furthermore, studies differ in their applied cognition tests.

To cover these limitations, this interdisciplinary study used a cross-over design, in which 20 healthy office workers were exposed to two test days of eight hours to either 800 ppm carbon dioxide or 3,000 ppm carbon dioxide in a respiration chamber. Volatile Organic Compounds were filtered out from the air. Cognitive performance was measured using the Cambridge Neuropsychological Test Automated Battery (CANTAB). Additionally, multiple price lists from economics literature were used to measure the risk and time preferences when faced with a financial decision-making problem [2]. A heuristics battery from psychology literature was used to measure sensitivity to bias behaviour in decision-making [3]. Lastly, subjects’ satisfaction about the indoor air quality and their belief to which degree the air quality hinders their ability to answer the questionnaire was recorded on a 7-point Likert scale. Physiological parameters including oxygen consumption, heart rate, heart rate variability, respiration rate, blood carbon dioxide concentration, blood pressure, and skin temperature were measured continuously to investigate possible mechanisms. A linear model with group fixed effects on test subject and clustering of standard errors on the subject level was used. As robustness checks, the same linear model but with bootstrapped standard errors was estimated and a mixed linear model with a random intercept on the subject and a random slope on the carbon dioxide condition was applied. Multiple hypotheses testing has been applied to derive corrected p-values [4].

The statistical analysis indicated no significant effect of carbon dioxide on any test results of either the CANTAB test battery, the economic preferences, or the heuristics battery after applying multiple hypotheses testing. Furthermore, no significant effect on the physiological parameters could be found after correcting for multiple hypotheses testing. The same insignificance was derived using the other regression models in the robustness checks. Also, there was no significant difference in satisfaction levels with the air quality. However, subjects rated that in the high carbon dioxide condition the air quality hindered them less to answer the questionnaire (p < 0.05, after multiple hypotheses testing).

We could not replicate the negative effects of pure carbon dioxide exposure found in past studies. It seems that carbon dioxide is not causing the impaired cognition. The absence of a physiological response cannot confirm any adaptive behaviour which could explain why cognition is unaffected.