Does stale indoor air affect your thinking and sleep?

We breathe out CO2, so in a closed room it climbs until fresh air dilutes it. Typical levels, in parts per million (ppm):

CO2 (ppm)	Where you meet it
400–600	Outdoor air and well-ventilated rooms, usually treated as normal7
About 717	Average bedroom with a window or door open1
About 1,000–1,150	Average bedroom with window and door shut1; decision-making dipped in one lab test at 1,0003
2,400–2,600	Student bedrooms with the window shut2

Is CO2 itself the culprit? Maybe not. One chamber study showed thinking slipping at 1,000 ppm³, but volunteers breathing pure CO2 up to 3,000 ppm did fine. At the same level from stuffy, people-filled air they reported headache, fatigue and foggy thinking⁴. CO2 may be a stand-in for poor ventilation.

What ventilation does for health. In offices, cutting ventilation from 10 to 5 litres per second per person (about 21 to 11 cubic feet per minute) raised sick-building symptoms by roughly 23%, and raising it to 25 litres per second cut them about 29%⁸. A multi-expert review found respiratory infections, asthma symptoms and sick days tended to rise as ventilation fell, and in Nordic homes, more than 0.5 air changes per hour went with fewer allergic problems in children⁵. In California classrooms, each extra litre per second per person cut illness absence by 1.6%⁹. In a randomized trial in Nunavut homes, a heat recovery ventilator brought CO2 into range and was linked to less wheeze and rhinitis in young children¹⁰.

Sleep. In two small trials, lower bedroom CO2, from an open window or a fan vent, went with better measured sleep quality and sharper concentration the next day². In 17 young adults, an open window or door cut average bedroom CO2 by about 430 ppm, and reported depth of sleep differed¹.

Ranked fixes, cheapest first

1. Open windows or doors. Cross-ventilate stuffy rooms, and crack a bedroom window or door at night if noise and outdoor air allow^1,2.
2. Add a quiet fan vent to a bedroom you cannot open. In one trial a fan that switched on above 900 ppm cut average CO2 from 2,395 to 835 ppm².
3. In a tight home, a heat recovery ventilator supplies steady fresh air without wasting heat¹⁰.
4. Use a direct-reading (NDIR) CO2 monitor to see how fast a room stuffs up.
5. Purifiers and houseplants do not replace fresh air: purifiers filter particles, not CO2 (see air purifiers and ventilation).

Does CO2 itself impair thinking? The studies disagree. Satish and colleagues exposed 22 people, blinded, to 600, 1,000 and 2,500 ppm in an office-like chamber (three 2.5-hour sessions in one day). Against 600 ppm, six of nine decision-making scales fell moderately at 1,000 ppm and seven fell sharply at 2,500 ppm (raw score ratios 0.06 to 0.56)³. Zhang and colleagues put 25 people through five 255-minute exposures. Adding pure CO2 to reach 1,000 or 3,000 ppm changed nothing significant, but 3,000 ppm that came from restricted ventilation lowered perceived air quality, raised headache, fatigue and sleepiness, and slowed several tasks⁴. Rodeheffer and colleagues randomized 36 submariners to 600, 2,500 or 15,000 ppm and found no difference on any of nine decision-making measures⁷. Du and colleagues reviewed 37 experimental studies: at levels common indoors, pure CO2 affected only high-level decision-making in non-specialized people, while poorer ventilation could slow responses without hurting accuracy. Methods, subjects and exposure doses varied widely¹¹.

Realistic offices. Allen and colleagues had 24 people spend six full workdays in a controlled office, blinded to conditions. Cognitive scores were 61% higher on a low-VOC "Green" day and 101% higher on two "Green+" days with extra outdoor air than on a "Conventional" day, and CO2 and VOCs were independently associated with scores¹². The conditions were simulated, and the group was small. Laurent and colleagues followed 302 office workers in six countries for 12 months with app-based tests. Each interquartile rise in CO2 (315 ppm) went with a 0.85% longer Stroop response time (95% CI 0.32 to 1.39), and each 8.8 µg/m³ rise in fine particles with 0.82% longer⁶. Those effects are small and observational, and well-ventilated buildings differ in other ways.

Sleep. Strøm-Tejsen and colleagues ran two field experiments in single-occupancy student rooms, one week per condition in balanced order. In a pilot of 14, opening a window gave average CO2 of 660 rather than 2,585 ppm. In a second experiment of 16, an intake fan that ran above 900 ppm gave 835 rather than 2,395 ppm. Wrist-actigraph sleep quality and perceived freshness improved significantly at the lower CO2, as did next-day sleepiness, concentration and a logical-thinking test². Mishra and colleagues studied 17 healthy volunteers with the window or door open versus closed; average CO2 was 717 versus 1,150 ppm. Of all sleep measures analyzed, only questionnaire-reported depth of sleep (P = 0.002) and an actigraphy sleep-phase measure (P = 0.003) differed¹. These are small samples of young, healthy adults over a week, with actigraphy rather than sleep-lab recordings, and CO2 moved together with temperature, noise and other pollutants.

Ventilation and illness. Fisk and colleagues combined published office studies into curves, and the 10-to-5 and 10-to-25 litre-per-second changes quoted above come from them; they warn that the average relationship may differ in a specific building⁸. Sundell and colleagues judged 27 papers informative and found consistent links between lower ventilation and worse outcomes, though the data on infections, asthma and sick leave are limited⁵. Mendell and colleagues tracked 162 classrooms in 28 schools for two years, estimating ventilation from daily CO2. All three districts had median ventilation below the 7.1 litres-per-second state standard, and the 1.6% drop in absence per extra litre assumes the link is causal and still needs confirming⁹. In the Nunavut trial, 68 ventilators were placed and 51 homes analyzed over six months: the odds of reported wheeze fell 12.3% per week (95% CI 1.9 to 21.6%) and rhinitis odds were lower in months 1 and 4, but health-centre visits did not fall significantly¹⁰.

Where this leaves us. Chamber results do not replicate cleanly: effects at 1,000 ppm³ vanished when CO2 was added alone⁴, and sailors showed none even at 15,000 ppm⁷. Ventilation-linked findings (symptoms, absence, wheeze) are more consistent but mostly observational, and no long trial has tested whether a bedroom CO2 monitor changes sleep or health.

Bottom line (emerging evidence). Better ventilation is linked to fewer symptoms and possibly to a little better sleep and concentration, but the case that CO2 itself is the cause is weak. Opening a window costs nothing, so the practical advice is easy.