By Nicola Twilley for The New Yorker
Dept. of Public Health
The Hidden Air Pollution in Our Homes
Outdoor air has been regulated for decades, but emissions from daily domestic activities may be more dangerous than anyone imagined. We spend most of our lives inside, where air quality has received little scrutiny.
Food magazines typically celebrate Thanksgiving in mid-July, bronzing turkeys and crimping piecrust four months in advance. By that time last year, Marina Vance, an environmental engineer at the University of Colorado Boulder, had already prepared two full Thanksgiving dinners for more than a dozen people. Vance studies air quality, and, last June, she was one of two scientists in charge of homechem, a four-week orgy of cooking, cleaning, and emissions measurement, which brought sixty scientists and four and a half million dollars’ worth of high-tech instrumentation to a ranch house on the engineering campus of the University of Texas at Austin. The two Thanksgiving dinners were the climax of the project and represented what Vance called a “worst-case scenario.” She suspected that the Pilgrims’ harvest celebration, as it is observed in twenty-first-century America, qualified as an airborne toxic event.Outdoor air has been regulated for decades, but emissions from daily domestic activities may be more dangerous than anyone imagined. We spend most of our lives inside, where air quality has received little scrutiny.
These days, a “very unhealthy” designation for outdoor air is rare. After the passage of the Clean Air Act, in 1963, and the creation of the Environmental Protection Agency, in 1970, the chemical composition of outdoor air became federally regulated, with penalties for polluters. Since the seventies, emissions of many harmful gases, such as carbon monoxide and sulfur dioxide, have fallen by half, and particulate counts by eighty per cent.
But this victory may be less significant than we assume, because, in America, we spend, on average, ninety per cent of our lives indoors. (By way of comparison, this means that humans spend more time inside buildings than sperm whales spend fully submerged in the ocean.) The statistic, from an E.P.A.-funded study conducted in 2001, might seem implausible, but it probably understates the case. More recent data, from the U.K., show that, on average, Britons are outside for just five per cent of the day—an hour and twelve minutes.
Unlike outdoor air, the air inside our homes is largely unregulated and has been all but ignored by researchers. We know barely the first thing about the atmospheres in which we spend the vast majority of our time. homechem—House Observations of Microbial and Environmental Chemistry—was the world’s first large-scale collaborative investigation into the chemistry of indoor air.
Thoroughly dissecting the data accumulated will take a couple of years, at least, and, even when the findings are published, no one will be able to state their public-health implications with certainty; homechem was designed to explore what the chemistry of indoor air is, not what it’s doing to us. But the experiment’s early results are just now emerging, and they seem to show that the combined emissions of humans and their daily activities—cooking, cleaning, metabolizing—are more interesting, and potentially more lethal, than anyone had imagined.
In September, 1776, Congress sent Benjamin Franklin and John Adams on an ultimately fruitless mission to Staten Island to negotiate peace with the British. One night, the two shared a room at a country inn, an adventure recorded in Adams’s diary. Adams, “who was an invalid and afraid of the Air in the night,” shut the window. To which Franklin responded, briskly, “The Air within this Chamber will soon be, and indeed is now worse than that without Doors: come! open the Window and come to bed, and I will convince you.”
According to the architectural historian David Gissen, debates about the relative dangers of household emissions versus urban emissions, and indoor air versus outdoor air, have swung back and forth between Franklin’s and Adams’s positions ever since, depending on each era’s prevailing beliefs and concerns. In 1867, inspired by the miasmatic tenements of America’s burgeoning cities, the engineer Lewis W. Leeds delivered a series of lectures under the title “Man’s Own Breath Is His Greatest Enemy.” He warned the unwary that “it is not in the external atmosphere that we must look for the greatest impurities, but it is in our own houses that the blighting, withering curse of foul air is to be found.” Half a century later, by contrast, the modernist architect Le Corbusier saw the indoor environments he designed as beneficent bubbles of man-made weather, shielded from the smog-choked city outside.
In mid-century America, cities such as Los Angeles and New York were repeatedly shrouded in thick brown fog—sometimes so lung-burningly toxic that it was mistaken for a chemical-weapon attack by a foreign power—and air pollution became an urgent issue. Legislation to curb it began appearing in the U.S. and other countries in the nineteen-fifties. After the passage of the Clean Air Act, government research dollars flowed to scientists looking to understand and to mitigate the sources and the health effects of air pollution.
But there was still almost no funding available for research into indoor air. Charles Weschler became one of just a few scientists in the field when he went to work for Bell Labs, in 1975, soon after completing a Ph.D. in chemistry. The company had noticed that the equipment in its telephone switching offices was failing faster than expected; it turned out that wire relays were being eaten away by an acidic, invisible indoor smog. Weschler told me that the little indoor-air research that was being done at the time was mostly geared not toward protecting people but toward preserving things.
In the eighties, amid emerging concerns about “sick-building syndrome,” a nonspecific malaise reported by occupants of the era’s new, more tightly sealed buildings, the E.P.A. started measuring indoor concentrations of known toxins, such as formaldehyde and asbestos, and assessing where they came from (paint, floor coverings, upholstery, particleboard). Researchers found that concentrations of these compounds were consistently higher indoors than they were outdoors, and some states began regulating consumer products containing the contaminants.
But it wasn’t until the aftermath of 9/11, with its heightened fear of airborne biological attacks, that indoor-air research finally attracted some funding—from the Alfred P. Sloan Foundation, one of the largest private grant-making nonprofits in the U.S. (Among its many grantees is a podcast I produce.) Through a program managed by Paula Olsiewski, a biochemist by training, Sloan began supporting research into H.V.A.C. filtration systems. Olsiewski identified a major difficulty in detecting traces of biological weapons: a complete lack of knowledge about the typical, baseline conditions inside buildings. As she put it to me, “If the biological threat was a needle in the haystack, what’s in the haystack? What microbes are in the air, and in the rooms, and on the surfaces?” She launched a multimillion-dollar program to investigate the microbiology and, later, the chemistry of our built environment.
Because there were so few specialists in the area, she decided to use Sloan money to lure eminent atmospheric chemists indoors. Delphine Farmer, a chemist based at Colorado State University, told me that, when she was invited to attend a workshop on indoor air chemistry in France, in 2015, her initial reaction was “You know, sure, I’ll take a free trip to France.” Farmer had spent the bulk of her career developing ways to accurately measure extremely tiny amounts of very complicated airborne molecules. She knew little about indoor air, but assumed that it wouldn’t be of interest.
Outdoors, primary emissions—whether from tailpipes, factories, or fertilizer-laden farms—undergo near-constant transformation into new combinations of chemicals through a cascading sequence of reactions. Indoor atmospheres were widely assumed to be far more static. But Farmer was captivated by the presentations that she heard. “I realized that we know nothing about indoors from a chemistry perspective,” she told me. “It was very clear that it was an area that was ripe for study, and that the indoor community just hadn’t had the resources we have in outdoor atmospheric chemistry.”
Olsiewski asked Farmer to lead an initiative to develop new instruments and databases for the study of indoor atmospheric chemistry. She recruited Marina Vance around the same time, hoping that the pair could build networks among researchers in the field. Vance and Farmer decided that the best way to achieve both goals was to initiate a large field study. Collaborative field studies are common in outdoor atmospheric research, because capturing the diversity and the complexity of the chemistry involved requires more instruments and more varied expertise than one lab can muster, but nothing of this scale had ever been undertaken indoors. Farmer and Vance gathered twenty research groups from thirteen universities, and homechem was launched.
At the University of Texas at Austin, the UTest House sits in a corner of the J. J. Pickle Research Campus, a scrubby four-hundred-and-seventy-five-acre plot of land dotted with radio antennae, a prototype nuclear reactor, and one of the nation’s largest nonmilitary computers. Atila Novoselac, the building engineer who runs the house, drove me there, pointing out the local landmarks before parking next to a jumble of weathered concrete chunks, which a structural-engineering lab was using to study the aging of pillars that support bridges and highway overpasses.
The house, a twelve-hundred-square-foot prefab that cost sixty thousand dollars, has been on the campus since 2006. Novoselac signed the contract to buy it on a Monday, and the house, delivered in two halves that were then glued together, was ready by the end of the week, complete with kitchen cabinets, bathroom fixtures, vinyl flooring, and curtains so ugly that he removed them immediately. In the years since then, for various research projects, Novoselac and his colleagues have cut the house open, studded it with thermal sensors, and pumped it full of gases.
Novoselac says that although it is fully operational as a house, he doesn’t think of it as one: “It’s a tool, a piece of equipment—the same as a screwdriver or a sensor.” Nonetheless, over the years it has been decorated with a doormat that says “hello” in looping cursive, a wobbly floor lamp, and a selection of scientific posters detailing research that has been conducted there.
By the time that Delphine Farmer and Marina Vance were looking for a site to host homechem, the UTest House was so dilapidated that Novoselac was contemplating scrapping it. But, as one of only a handful of full-scale test homes in the country, it was a perfect place to conduct a full-scale simulation of human inhabitation. Vance and Farmer devised a schedule that would organize real-life activities—cooking, cleaning, and simply hanging out—into a series of controlled, sequential experiments.
When I visited the house, two doctoral students, Catherine Masoud and Kanan Patel, made us all stir-fry for lunch, while Novoselac used an ultrafine-particle counter to monitor the air quality. The instrument, which looked like a clunky blue plastic car phone from the eighties, complete with telescoping aerial, recorded a background level of around two thousand particles per centimetre cubed before the cooking began. Patel pulled up the project’s stir-fry spreadsheet on her laptop, turned the right-front burner to its highest setting, and set rice to boil in a small pot. Masoud warmed up a couple of tablespoons of oil in a wok, dumped in two bags of frozen vegetables, and stirred them over a high heat. As the broccoli and the sugar snap peas started to caramelize, the smell from the kitchen made my stomach rumble, and Novoselac’s particle counter emitted an accelerating series of beeps. “That sound means we’ve hit the limit of the instrument,” he said. “Above a hundred thousand, it’s unreliable.” As the beeps blurred together into a shrill lament, Masoud doused the vegetables in three-quarters of a cup of Baby Ray’s Sweet Teriyaki Sauce and a tablespoon of sriracha, and lunch was served.
Stir-fries were a staple of homechem life, because Vance and Farmer had determined that dishes cooked at high heat would produce the most interesting organic aerosols. When Vance first arrived in Austin, she went to a grocery store, picked out four brands of teriyaki sauce, made four stir-fries, and served them to the students. “We were, like, Which one do you want to eat for a month?” she told me. “There was no science in the choosing of the stir-fry sauce—we just wanted it to taste good.”
In the early phases of homechem, the researchers ran into difficulties. Their instruments, designed for outdoor atmospheric measurement, had to be recalibrated to deal with the much higher concentrations that build up indoors. A bigger obstacle was controlling for the irreducible complexity of human behavior. On the first full day of the study, student volunteers mopped the floor with Pine-Sol five times in a row, airing out the house completely between each cleaning. The goal was to create an emissions inventory for mopping: a consistent signature that would allow the team members to isolate its chemical contribution from the noise of everyday activity. Unfortunately, the initial data revealed that some students mopped less thoroughly than others, giving rise to noticeably uneven emissions of organic chemicals. As a result, the most diligent mopper—subsequently christened Mr. Clean—was appointed mop boss, in charge of enforcing quality control for the month.
Once the baseline emissions for things like cooking a stir-fry or an English breakfast and mopping with Pine-Sol or a bleach-based cleaner had been established, the scientists began to layer activities together. A group of volunteers would spend the day in the house, cooking breakfast and lunch, checking their e-mail, cleaning up, making dinner, and running the dishwasher, in order to see whether, say, the emissions from frying vegetables in teriyaki sauce would react with the bleach fumes from mopping the kitchen floor afterward. Farmer told me that, based on her preliminary data, it seems as though they did, producing temporary spikes of chloramines, a class of chemicals that are known to inflame airway membranes. Another product of the marriage of bleach-based mopping and gas-burner ignition is nitryl chloride, a compound that is known to atmospheric chemists for its role in coastal smog formation. No one had expected to find it indoors.
Cooking a stir-fry on a gas burner rather than on an electric hot plate produced much higher emissions over all, primarily because of the additional products of combustion. Meanwhile, the two Thanksgiving experiments provided some hints that cooking meat produces different atmospheric chemistry than cooking vegetarian dishes does: one group has been analyzing ammonia concentrations that they believe came from the breakdown of proteins in the turkey. Indeed, Atila Novoselac told me that, although homechem was not designed to study this, it’s entirely possible that different dietary regimes or national cuisines could result in quite different emissions inventories. Spices have varying levels of reactivity to ozone, a key ingredient in smog formation; for example, star anise, which contains high levels of volatile sesquiterpenes, might act as an ozone sink, reducing its levels and improving air quality.
Cooking and cleaning are thought to be the main activities through which humans add chemicals to the indoor environment, but we also create emissions by simply existing. Exhaled breath contains carbon dioxide and also a host of organic chemicals such as isoprene, acetone, and acetaldehyde. When I asked, Caleb Arata told me that his gauges are capable of registering the gaseous signature of a fart, though he discreetly declined to confirm or deny the presence of flatulence in the data. And squalene, a primary ingredient in skin oil, is extremely reactive with ozone—a fact that may explain why air travel, which exposes us to the higher ozone concentrations of the upper atmosphere, often leaves us feeling dirty.
On top of these involuntary emissions, many humans also intentionally apply a cocktail of chemical compounds in the form of personal-care products. Three days of the homechem experiment were dedicated to studying their effects: on the first day, student volunteers were asked to use only minimal skin and hair care; on the second, they were allowed to follow their normal routine; and, on the third, they were encouraged to go to town with scented body sprays, lotions, mousses, and mists. This data set has yet to be analyzed, but earlier research hints at its potential. Novoselac and another homechem researcher, Richard Corsi, recently collaborated on a separate study of nearby high schools and found that the highest emission levels were always of the same two chemicals, found in exactly the same ratio at every location. After a little bit of detective work, they identified the culprit: Axe body sprays, which the teen-age boys of Texas apparently apply lavishly in classrooms between periods.
By the end of a month of stir-fries, mopping, and antiperspirant, even researchers who’d doubted whether indoor air would be interesting had come around. One converted skeptic, Philip Stevens, an atmospheric chemist at Indiana University Bloomington, had contributed an instrument designed to measure the hydroxyl radical, a compound so reactive that it is known as “the Pac-Man of the atmosphere.” Hydroxyl radicals drive much of outdoor atmospheric chemistry, and are a mixed blessing from a health point of view: they break down VOCs but also react with nitrogen oxides to produce ozone, making smog formation more likely. Stevens was surprised when his readings registered their presence, because their production requires sunlight, and a house’s walls and windows block much of the sun’s energy. Like many researchers, he’d assumed that indoor air, lacking sunlight and, thus, hydroxyl radicals, wouldn’t yield the kind of rapid photochemical reactions that atmospheric scientists like to study. But his results, supported by a colleague’s measurements of light intensity inside the house, have convinced him that afternoon sunshine filtered through a window, combined with emissions from a gas stove, is sufficient to produce chemical reactions “similar to what you might find outside on a smoggy urban day.”
Dozens of the chemicals measured by the homechem team are known to be harmful, and, as every scientist I spoke with mentioned, we spend almost all our time indoors, breathing them. Nonetheless, it is outdoor air-pollution levels that have been firmly linked to public health. In 2016, the World Health Organization attributed 4.2 million premature deaths to outdoor air. The associations between outdoor air pollution and heart disease, lung disease, and cancer have been well documented; more recent research has suggested connections to low birth weight, diabetes, and even cognitive damage.
“So there is a big question here,” Marina Vance pointed out. “If all these studies have found an association between outdoor air pollution and a decrease in life quality and life expectancy, but we’re not outside, how does that relationship still hold?”
One possibility is that the brief moments we spend outdoors have an outsized impact on our health. Another consideration is that outdoor pollutants can and do come inside. But one homechem researcher, Allen Goldstein, recently co-authored a paper that suggests a fascinating inversion. The dominant source of VOCs in Los Angeles is now emissions from consumer products, including toiletries and cleaning fluids. In other words, vehicle emissions have been controlled to such an extent that, even in the most car-clogged city in America, indoor air that has leaked outdoors may create more smog than transportation does.
The scientists involved in homechem wonder about these questions, though they are cautious about speculating too freely. So far, Delphine Farmer told me, it’s safe to say that levels of many traditional air pollutants are lower indoors—until you do something like cook a stir-fry, at which point some of those levels will briefly reach peaks that are ten times the maximum observed outdoors. Other, more complex organic molecules seem to always be more plentiful indoors. There’s also evidence that outdoor particles can get coated with gases when they come indoors, which could potentially provide a different pathway for them to penetrate your lungs.
Simply measuring concentrations of a chemical in a test house is not enough to infer potential exposure, however. John Balmes, a pulmonologist at the Human Exposure Laboratory, at the University of California, San Francisco, told me, “Going from chemistry to epidemiology is a big leap.” To gauge the varying levels of each compound that the Thanksgiving cooks and their guests likely inhaled would require precise readings at various heights in various rooms, correlated with activity patterns. Still, when I told Balmes that the carbon-dioxide reading for Thanksgiving had peaked at four thousand parts per million, he was taken aback. “Wow,” he said. “Those kinds of levels will lower your cognitive functioning, at least in the short term. Whether it has any long-term effect, we don’t know.”
Similarly, when I told Francesca Dominici, a biostatistician at Harvard, that the Thanksgiving levels of fine particulate matter had reached two hundred and eighty-five micrograms per cubic metre, she responded with shock. “Even short-term increases of just ten micrograms per cubic metre from one day to the next will increase hospital admission rates and mortality in the over sixty-fives,” she said.
Katherine Hammond, an exposure scientist at U.C. Berkeley’s School of Public Health, was particularly struck by the holiday’s high levels of ultrafine particles. As little as a nanometre in diameter, ultrafines are small enough to pass through into the bloodstream with ease. “They’re tiny, but we think they may have a disproportionate health effect,” she told me. Some of her previous research has scrutinized emissions from self-cleaning ovens. “When the ultrafines soared, you could feel it in your eyes and in your throat,” she said. “There are even some theories that they can go from your nose directly into your brain, following the olfactory nerve.” All the same, she was careful not to sound alarmist. “The point of an experiment like this is that you start raising questions and figuring out how to go further into the detail,” she said. “But you can’t take this data and convert it to a health risk.” Right now, as Balmes pointed out, scientists don’t even know whether all particles of the same size are created equal. “Is inhaling particles from diesel engines worse than inhaling particles from fried foods?” he said. “The research hasn’t been done.”
None of the homechem researchers consider the health risks of cooking emissions to be worrying enough to forgo the benefits of a delicious, home-cooked meal, and they agree that we’re still a long way from being able to predict exactly which combinations of activities and environmental conditions might create harmful indoor air. As the pace of indoor research increases, however, we will soon know enough to do so—at which point, the question will be how to make indoor air healthier. Many of the primary sources of indoor emissions are resistant to regulation. A move to outlaw toasters or bleach-based floor cleaners seems unlikely to succeed. “You can even light a candle at home!” Marina Vance pointed out, in a tone that expressed her horror at the atmospheric implications of doing such a thing.
On the other hand, once the main risk factors are identified, pollution may prove easier to curb indoors than out, precisely because the space is more confined. It’s taken nearly fifty years for the United States to successfully reduce outdoor ozone and particulate concentrations, whereas much of the particulate matter from cooking can probably be removed by measures as simple as investing in a decent exhaust hood and replacing its filter frequently. John Balmes already has data suggesting that running an exhaust hood while cooking correlates with a steep reduction both in household particulate levels and in childhood asthma attacks. Benjamin Franklin’s advice to open the window is worth following, too, particularly when mopping or making toast—provided, of course, that the air outside is clean.
The Thanksgiving guests began to arrive at 3:35 p.m., exactly as they had done the previous week. They brought with them a whoosh of dust particles from the outside, and also their own personal emissions—lactic acid from sweat, squalene from skin oil, and carbon dioxide. The diners, all homechem researchers, had been excluded from the house while the cooking was happening, because their presence would have skewed the data—scientists can calculate how many people have entered a space from the rise in carbon-dioxide levels alone. Crammed around a couple of folding tables, some eating from bowls, owing to a shortage of plates, the group offered toasts to the cooks and to the experiment.
Whatever the air quality, the atmosphere was lively. “That’s obviously a combination of grad students and free food,” Caleb Arata told me. “But then, also, you know, it wasn’t loaded with family dynamics.” Instead of awkward discussions about children, careers, and politics, conversation revolved around such matters as the likely source of the limonene that one of the instruments had registered at lunchtime. (It had been produced by a spritz of lime added to guacamole.)
One of the scientists, Lea Hildebrandt Ruiz, said that conditions inside the house had briefly exceeded those of the world’s most polluted city—“and I can say that,” she added, “because I have a monitoring program in New Delhi.” According to the World Health Organization, the Indian capital’s air quality is the worst of any major city. During the dirtier winter months, levels of fine particulate matter in the air there typically hover at around two hundred and twenty-five micrograms per cubic metre. That’s still significantly lower than the two hundred and eighty micrograms per cubic metre that was reached during the final, frenzied hour of cooking. Everyone had expected Thanksgiving to be bad, but no one had expected it to be that bad—a finding that was alarming but also, from a research point of view, thrilling.
Curiously, although an understanding of indoor air is still in its infancy, scientifically speaking, it’s something that we’re all equipped to detect. “Your nose is a pretty good chemical instrument,” Farmer told me, giving the example of cutting onions. Slicing through an onion’s cell walls causes them to emit syn-propanethial-S-oxide, a VOC responsible for a temporary but powerful shift in the indoor atmosphere—and for the resulting tears. But our noses can lead us astray, too. Vance pointed out that food smells most delicious when it is browning, in a process called the Maillard reaction, yet the compounds emitted as steaks sear and bread toasts include brown carbon (a form of particulate matter) and VOCs from incomplete combustion. “I used to think, Wow, this house smells so good—it smells like Thanksgiving,” Caleb Arata said. But, this past November, he told me, as he prepared his third turkey dinner of the year, he thought instead, “This house smells so good—I wonder what I’m inhaling?” ♦
This article appears in the print edition of the April 8, 2019, issue, with the headline “Home Smog.”
Nicola Twilley is a contributing writer to The New Yorker. She is a co-host of the podcast “Gastropod” and is at work on two books: one about refrigeration and one about quarantine. Read more »