Does Stuffy Air Make You Worse at Thinking?

Something is there, the famous numbers are almost certainly too big, and the effect appears mainly on one particular test. A 2016 controlled exposure study reported cognitive scores 61 to 101 per cent higher under better ventilation. A later double-blind crossover study at NASA’s Johnson Space Center found no clear dose-response pattern across four carbon dioxide concentrations. A critical review of 37 experimental studies concluded that pure carbon dioxide at indoor concentrations was only found to affect high-level decision-making measured by one specific test battery.

So the literature is genuinely unsettled and the striking figures come with an important asterisk.

The study everyone quotes

Allen and colleagues published a controlled exposure study in Environmental Health Perspectives in 2016. Twenty-four professional employees, described as architects, designers, programmers, engineers, creative marketing professionals and managers, worked for six days in an environmentally controlled office space while blinded to the test conditions.

Conditions varied daily: conventional, with high volatile organic compound concentrations; green, with low concentrations; green with enhanced outdoor air ventilation; and artificially elevated carbon dioxide independent of ventilation.

The headline: cognitive scores were 61 per cent higher on the green building day and 101 per cent higher on the two enhanced ventilation days than on the conventional day. Volatile organic compounds and carbon dioxide were independently associated with cognitive scores.

The paper also reports scores 20 per cent higher on enhanced ventilation days than on the moderate carbon dioxide day, and a 25 per cent average increase across eight of nine domains when outdoor air ventilation was increased.

Those are enormous numbers for an environmental intervention, and that is the first thing worth noticing about them.

Where the numbers come from

The outcome measure matters more than usual here.

The study scored participants using a battery covering nine cognitive domains, oriented towards decision-making. A national laboratory description of related work using the same instrument explains that it presents scenarios, such as being the manager of an organisation when a crisis hits, and scores participants on dimensions including how proactive they are and how they search for and use information.

This is not a reaction time test. It is a simulation of executive decision-making, originally developed for contexts including assessing cognitive deficits after traumatic brain injury.

That has two consequences. It may be more sensitive than standard psychometric tests, which is the reason its users give for detecting effects at low exposures. And percentage changes on a composite simulation score do not translate into percentage changes in anything you would recognise at your desk.

What happened at NASA

A double-blind crossover study published in npj Microgravity tested this directly, because carbon dioxide levels matter on the International Space Station.

Twenty-two participants at the Johnson Space Center were randomly assigned to groups, each exposed to a different sequence of four carbon dioxide concentrations: 600, 1200, 2500 and 5000 parts per million. They were assessed using the same decision-making battery as the earlier work plus a separate cognitive test suite.

The result: there were no clear dose-response patterns for performance on either battery.

The paper is candid about the field, noting that findings of effects of low concentrations of carbon dioxide on cognition are controversial and the literature is unsettled, and that no statistically significant effects on acute health symptoms or cognitive performance were seen when college students were exposed for over four hours to pure carbon dioxide at 1000, 3000 or 5000 parts per million in other work.

The review that locates the effect

A 2020 critical review in Indoor Air summarised 37 experimental studies that ran objective cognitive tests with manipulated carbon dioxide concentrations, either by adding pure carbon dioxide or by adjusting ventilation.

Its conclusion is the most useful single sentence in this literature. Among studies meeting objective criteria for strength and consistency, pure carbon dioxide at concentrations common indoors was only found to affect high-level decision-making measured by that one simulation battery in non-specialised populations. Lower ventilation and accumulation of indoor pollutants, including carbon dioxide, could reduce the speed of various functions but leave accuracy unaffected.

It names the confounds too: variation in cognitive assessment methods, study designs, individual and population differences, and uncertainties in exposure doses.

So carbon dioxide alone shows up reliably on one instrument and not others. Poor ventilation more broadly shows up as slower rather than less accurate performance. Both are real findings and neither is the doubling headline.

What to actually take from this

Ventilating a room you have been sitting in all day is free, harmless and supported by at least a speed effect. Nobody needs a research justification to open a window.

What is not supported is the version that circulates: that a stuffy room measurably halves your cognitive ability, or that a carbon dioxide monitor is a productivity instrument. The largest effects come from one sensitive decision-making simulation, in small samples, and a double-blind crossover study using the same instrument found no clear dose-response.

This is the same pattern as office temperature, where a striking headline number failed a reanalysis, and as office lighting, where brightness studies found much less than daylight studies did.

If you work in a room with no ventilation and it feels stale by the afternoon, the discomfort is real and worth fixing on its own terms. Persistent headaches, dizziness or breathlessness in a particular room warrant a doctor and a conversation about the building rather than a monitor. Dryness is a separate issue with its own mechanism, covered in dry air and your eyes.

Common questions

Does bad air actually make me stupider?

The evidence does not support that framing. A 2016 controlled exposure study of 24 office workers reported cognitive scores 61 per cent higher on a low-pollutant day and 101 per cent higher on enhanced ventilation days than on a conventional day. But a later double-blind crossover study of 22 participants at NASA’s Johnson Space Center, exposed to four carbon dioxide concentrations from 600 to 5000 parts per million, found no clear dose-response patterns on either of two test batteries. A 2020 critical review of 37 experimental studies concluded that pure carbon dioxide at common indoor concentrations was only found to affect high-level decision-making measured by one specific simulation battery.

Why are the reported effects so large?

Partly because of what was measured. The large percentages come from a battery covering nine domains oriented towards executive decision-making, which presents scenarios such as managing an organisation during a crisis and scores dimensions including proactivity and how information is searched for and used. It was developed for contexts including detecting cognitive deficits after traumatic brain injury, and its users argue its sensitivity is why it picks up effects at low exposures. A percentage change on this composite does not correspond to a percentage change in ordinary desk work. The 2020 review found the carbon dioxide effect appeared on this battery and not consistently elsewhere.

Should I buy a CO2 monitor for my office?

Nothing in this research tested one, and this site has tested no devices and names none. What the 2020 review found was that lower ventilation and accumulation of indoor pollutants could reduce the speed of various cognitive functions while leaving accuracy unaffected, which is a real but modest finding. It also named major confounds including variations in assessment methods, study designs, individual differences and uncertainties in exposure doses. Opening a window costs nothing and needs no justification. Treating a number on a monitor as a measure of your cognitive capacity goes well past what this literature shows.

Is ventilation worth bothering with at all?

Yes, on comfort grounds at minimum, and there is some evidence beyond that. The 2020 review of 37 studies found that lower ventilation and pollutant accumulation could reduce the speed of various functions, though accuracy was unaffected. The 2016 study found volatile organic compounds and carbon dioxide were independently associated with cognitive scores, and reported a 25 per cent average increase across eight of nine domains when outdoor air ventilation was increased, on the sensitive decision-making battery. If a room gives you persistent headaches, dizziness or breathlessness, that is a matter for a doctor and for whoever manages the building.

One instrument, three conclusions

The interesting thing about this literature is that a single measurement choice appears to carry most of the disagreement.

Studies using the sensitive decision-making simulation find large effects at low exposures. Studies using conventional cognitive batteries often find nothing. A double-blind crossover study using both found no clear dose-response on either. And a review of 37 studies concluded that pure carbon dioxide at indoor levels reliably affects performance on that one battery and not the others, while poorer ventilation generally slows people down without making them less accurate.

That is not a scandal and it is not a debunking. It is what an unresolved measurement question looks like, and the honest position is the one the NASA paper takes: the literature is unsettled.

Meanwhile the practical advice survives all of it, because it never needed the research. If the room feels stale, air it out. That was always free.