The Science of Air Purifiers and Better Health: Is There Any?

happy kid in the field

What’s the hottest gift this Chinese new year? Perhaps instead of moon cakes and red envelopes, you could give “the gift that keeps on giving”: an indoor air purifier. They certainly are all the rage in China since last year, with skyrocketing sales and sold-out inventories after the trio of highly publicized airpocalyptic crises. I think this is a good turn of events: plenty of independent testing, including mine, has documented that a good air purifier can dramatically improve your indoor PM2.5 by 80% or more. But is there any good data that proves that this actually makes you healthier? It seems logical, of course, that decreasing exposure to pollution would decrease harmful health effects. But medical history is filled with tales of common sense and tradition that later turn out to be worthless or harmful — like bloodletting, or the more modern tradition of multivitamins. A big percentage of people reading this article take a daily multivitamin, assuming it’s “healthier” to do so, but the best evidence shows they are worthless, and possibly harmful. Could air purifiers be the same?

In theory and in testing, a good purifier should improve a room’s pollution levels more than 80%; this 80% reduction is also what the private Association of Home Appliance Manufacturers (AHAM) uses in their clean air delivery rate (CADR) tests, which are widely cited in comparison charts of air purifiers. So let’s say you’ve installed a top-of-the-line purifier in your living room, feeling quite safe and cozy. But how much of your time is actually in that filtered room? Or maybe the purifier is too small for that room size, or the filters are old, or the fan speed is too low, or the windows are open? Even this commonly cited CADR test is just a lab test for only 20 minutes — what about in the real world? I want to take this conversation to the next level, seeking out proof that your health will improve when using these machines. I want to be able to tell my patients and readers that there are published research studies which followed people over many months or even years, compared them to a control group not using air purifiers, and measured their health to see if there was any improvement in heart and lung disease, cancers and death rates. Are there any such studies?

I searched the Pubmed scientific database to find the best studies, and I was disappointed but not surprised to find very little strong data. A properly designed research project like this would be very difficult and expensive. But there are a few attempts, especially studies looking at using HEPA filters to help children with asthma. One was a systematic review published in 2002, which found that air filters helped to improve asthma symptoms — but the effect was small, and there was wide variation between studies which made conclusive assessments difficult. A more recent, very well designed study published in Pediatrics in 2011 followed two hundred children with asthma who also were exposed to secondhand smoke at home, and gave half of the kids a true HEPA purifier and the other half a fake purifier for their bedrooms. After a year, the HEPA group of children had less doctor visits for asthma flares, which possibly — but not conclusively — could be due to the 25% decrease in PM2.5 in their homes.

Other studies have focused on allergies, including an interesting study from 2008 which assessed children with documented pet allergies, following them over a year and recording lung function and blood markers. After a year, those who used HEPA air purifiers showed no clear difference in lung function, use of allergy medicines, or blood markers of allergies. Another study back in 1990 was a bit more impressive, showing not only a 70% reduction in indoor PM0.3 but also improved patient symptoms of allergies.

All of these hint at health benefits, but they still dance around the edges of what I want to know for us in China and the developing world. In the USA, most of the air purifier marketing and testing focuses on allergies and asthma. But here in the developing countries, the air pollution is much more severe and thus the health risks are far more serious. We are worried about pollution’s long-term risks of death, heart and lung disease and cancer. These studies I just mentioned still aren’t answering that deeper question: can long-term use of indoor air purifiers prevent death, heart and lung disease, and cancer?

Happy Family Riding Bicycle With Cloud Background

The best study I found was published in January 2013 in Indoor Air. It was very well designed for this complicated type of study, being a randomized double-blind crossover study of 20 homes over three weeks, using an air purifier or a placebo purifier. Their main goal in this remote First Nations community in Canada was to assess whether air purifiers could improve cardiorespiratory health. As their abstract says,

“…each home received an electrostatic air filter and a placebo filter for 1 week in random order, and lung function, blood pressure, and endothelial function measures were collected at the beginning and end of each week… On average, air filter use was associated with a 217-ml increase in forced expiratory volume in 1 second, a 7.9-mm Hg decrease in systolic blood pressure, and a 4.5-mm Hg decrease in diastolic blood pressure. Consistent inverse associations were also observed between indoor PM2.5 and lung function. In general, our findings suggest that reducing indoor PM2.5 may contribute to improved lung function in First Nations communities.”

This same Canadian research team had earlierpublished a similar study, testing 45 non-smokers for 7 days in 20 homes that used wood stoves, comparing health effects with or without HEPA purifiers. The people using the filters showed improved endothelial function and biomarkers of inflammation such as CRP. As most pollution researchers now see pollution as a pro-inflammatory disease, testing for such biomarkers could indeed be an accurate surrogate for later health problems. This approach is also being used in studies of air pollution masks, which I recently reviewed.

My take from these studies? Firstly, they all confirm what we already know: air purifiers can reduce the levels of indoor PM2.5, but with a wide range of effectiveness. Secondly are the more important results looking at health markers. I think the most encouraging finding was the First Nation study showing improvement in lung function, even in such a short amount of time (less than a month). Their data was a bit less convincing on blood pressure improvements, but perhaps a larger study would help confirm their initial findings of a slight improvement.

None of these studies are slam-dunk proof for me, but I honestly don’t know whether we ever will get many more well designed studies like these, unless governmental researchers or Gates-type philanthropists fund them. But until better studies come along, we must rely on what we do know:

  • Air pollution contains many chemicals, but PM2.5 is considered to be the most harmful to health.
  • There is no such thing as a “safe” level of  PM2.5. Lower is always better.
  • Worsening PM2.5 causes deaths from all causes, especially heart and lung diseases and cancers. Many studies have shown this, including this 2013 meta-analysis of the population in China.
  • On the brighter side, long-term improvements in PM2.5 do help to decrease mortality. The best study was a huge epidemiological analysis of entire populations in American cities as the air improved from the 1970’s to 1990’s. Lifespans improved for everyone, for a multitude of reasons, and they estimate that 15% of the improved life expectancy was due to cleaner air.
  • Shorter studies have also shown improvements in health from better outdoor air pollution. The best designed study I’ve seen on this happened right here in Beijing, during the 2008 Olympics. A team of researchers followed 125 healthy young doctors before, during and after the Olympics, and found improved blood pressure, heart rate and other biomarkers of inflammation during those lovely days of improved air pollution. Another encouraging study followed pregnant women and their babies in Tongliang, China both before and after a heavily polluting coal-fired power plant was forced to shut down in 2004, and found improved neurodevelopmental scores in newborns at age 2 years.

Is all of this enough to convince you to use an indoor purifier? For me, I was already convinced years ago — it’s not just common sense, it actually makes biochemical sense and also perfectly fits with the precautionary principle: “When an activity raises threats of harm to the environment or human health, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically.”


Follow me on:
Twitter @RichardStCyrMD
Facebook @BainbridgeBabaDoc
Photography: richardsaintcyr.com

9 thoughts on “The Science of Air Purifiers and Better Health: Is There Any?”

  1. I appreciate the attempt here to look scientifically at all this. It is disappointing that there haven’t been some more long-term tests since, as you mention, there are important factors that the studies here aren’t able to answer. Common sense would say air filters are making a difference, but as you say science and medicine are rife with examples of how “common sense” has fallen short.

    Which is why I can’t possibly agree with your final sentence, as it is stated there. The law of unintended consequences is harsh and unforgiving. I think immediately of George McGovern’s senate committee in the 60s pushing doctors to tell them what to tell people to eat, and when doctors and scientists pleaded with them not to make pronouncements to the public based on untested associations replying that “senators don’t have the luxury of waiting,” thus fulfilling the prophecies of the very doctors there in the room about the danger of enrolling the entire American public in a giant, uncontrolled experiment. We all did enroll, changed up the macronutrient composition of our diet and are all now fattening and sickening like 150-day grain fed cattle.

    Or there’s the laws requiring that folic acid be sprayed on all processed foods to make sure everyone “gets enough.” Seemed like a great idea at the time it was invented (folic acid of course being a chemical compound processed by the liver and folate being the actual vitamin that is absorbed in the intestines) and now all the concern is the strong associations between too much folic acid and colon cancer and other problems–plus new discoveries that (SURPRISE!) a lab made chemical compound doesn’t have all the effects of the real vitamin. It has proved somewhat useful in reducing fetal deformities, but not in other areas that folate is necessary for.

    The literature decades ago was rife with untested, unproved hypothesis showing strong associations “proving” women treated with HRT didn’t get as many heart attacks. So we started prescribing it anytime a women over 40 started to feel a little warm. When someone finally bothered to do an actual controlled study later they found exactly the opposite. The final answer on that still hasn’t been resolved.

    You yourself mention the multivitamin issue–multivitamins that people’s doctors and governments advise them to take and are just now backing off of in the face of conflicting evidence that suggests they do as much harm as good.

    Sadly, I could go on and on: studies published in the Lancet analyzing that data and showing that seat belts haven’t improved vehicular death rates and have in fact increased the death rate for those outside vehicles during accidents, and providing of free condoms has no provable association with decreased incidents of STD transmission.

    There’s always something like this; some problem that’s just “so important” we have to have a law or a regulation or a strong suggestion at least. We “have to act”, even though we’re acting in ignorance. And once the law or regulation or strong suggestion is in place, the motivation to do the hard science necessary to get the real evidence beyond worthless associations and correlations is often lost. And we HAVE to get past associations and correlations, as John Ionides has shown in his very careful and comprehensive analysis of observational studies that have gone on for further testing, showing that any association or correlation identified is far more likely to ultimately be proved wrong that right. I believe the statistics he found were that 80% of associations and correlations turn out to be false.

    I realize I’ve gotten a bit away from air pollution here, but I was just a bit horrified at that last sentence.

    (Lancet study, if you’re interested: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.169.4005&rep=rep1&type=pdf

    John Ionides’ study: http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020124)

    1. Hi Miriam, thanks for the long response! But I’m a bit confused why you would disagree with the well known public health concept of the precautionary principle. The first rule for health is to “do no harm”. The precautionary principle is quite clear: if we are unsure about health effects of something, then it’s not a bad idea to avoid exposure to that “something” until evidence comes up later to prove good or bad. So this perfectly applies to air pollution — we have unequivocal proof that air pollution and PM2.5 harm your health. So avoiding that risk factor is a smart idea — and if air purifiers decrease levels of PM2.5, then it’s perfectly normal assumption to apply. Plus, it does no harm (except on your wallet!).

      1. I don’t disagree with the concept of acting when there is real harm. What I disagree with is the logic here, and the concept of acting when the causes and effects are not well understood. It is really the second part of that principle that I find horrifying, and honestly I’m not sure that it being a “well known” principle means anything, would it? To refuse to disagree with a principle just because many people think it’s true is the antithesis of science, and the logical fallacy argumentum ad populum. The second half of that principle can also come into direct conflict with the other, separate principle you stated: first do no harm. It is especially dangerous when the actions taken are either required or promoted by authorities.

        The example of multivitamins illustrates this. A threat of harm was identified (vitamin or mineral deficiencies), a precautionary measure was taken (advise and promote the use of multivitamins as well as spraying vitamins all over processed food) when the cause and effect relationships were not fully established scientifically. Multivitamins were widely advised and the fortifying of processed foods was required by law when the long term effects were utterly unknown. But who could argue with giving people healthy vitamins? Promoting the use of multivitamins was perfectly in accord with the principle you’ve stated, yet utterly in violation of the more important principle to do no harm.

        Of course I realize that the air filter situation differs somewhat. For one thing, installing an air filter is more about avoiding a known harm than added a perceived good (like a multivitamin). For another, authorities at this time are not making air filter purchase mandatory or even strongly advising everyone to get one. But even with those two caveats an analogous situation should make us cautious. That would be the use of antibacterial soaps and hand sanitizers. They certainly are very valuable in certain situations, like surgery, or certain times, such as during an epidemic. But they have been promoted and suggested to people such that everyone now uses them for everything. And they certainly do fulfill their intended consequence of reducing an individual’s exposure to viruses and bacteria, but they also have had unintended consequences.

        They are also, like air filters, about avoiding something bad: bacteria and viruses. I personally can’t in my wildest dreams think of any way that avoiding air pollution could possibly have a downside. I agree with you that it’s wise to limit exposure to something known to be harmful. But I also am aware that those who invented and promoted antibacterial cleaners must surely have had the same thought. How on earth could it possibly be bad to avoid something like a germ? And yet now these items are being more thoroughly investigated as to their implication in precipitous rises in childhood asthma, poorly developed immune systems and even a possible role in the rise of superbugs.

        I wouldn’t anticipate negative long term consequences in a place like Beijing even if everyone in the city had an air purifier – (except of course for the rise in electricity demand, the question of how to dispose of so many filters full of heavy metals and the question of public and state motivation to do something to lower pollution levels if everyone can be protected against them in their home, office and car anyway.) After all, if the AQI outside is 500 and your air filter cleans out 80% that’s still an indoor AQI of 100, which is still worse than Salt Lake City on a really bad day. It’s not as if the air cleaners in Beijing are going to be scrubbing the air to absolute zero.

        But what possible consequences might there be when everyone in the world has an air purifier, and everyone runs them no matter what the AQI? Are there any possible consequences to me in using one? What if I stay in Beijing for five more years than I would have otherwise, because I think of myself as “safe” with my lovely air purifier? What are the consequences of going suddenly from from such clean to such dirty air suddenly multiple times a day? I can’t think of what the downside would be; but that doesn’t mean there won’t be one.

        That’s all I’m objecting to–solely the second half of that concept you mention at the end of your post: that we (especially those in authority) should act when cause and effect have not been established. That’s leaping from a frying pan without knowing what’s underneath.

      2. I’m a little late to comment, but the precautionary principal never requires. It merely blocks action. It’s the same as first, do no harm. While this is indeed a weaker definition, there is still a strong a weak version of that. A quick look a Wikipedia will show that many actually disagree with the principal, especially when adopted by governments.

        In the end though, we are talking about air cleaners. They are pretty simple machines and it’s hard to imagine that are doing harm. Unknown risks could include the noise, the charged particles in the air, and the possible benefits of airborne gasses and particles. All I know though, is that when I have my Sharp on during the night I down wake up with that tightness in my chest. And when I have the humidifier on, I don’t wake up with a sore throat.

        Thank you for all the reviews and information.

  2. Yes! This is an excellent point. Back before I started Smart Air, I was thinking that we need a really good study where we randomly assign filters to people in China–maybe visiting expats? Now that I’ve got bulk orders of filters, I’d love to team up with a public health researcher to do a real-world experiment. My hunch is that I am more likely to get a cold when the air is bad, but I don’t have any good empirical evidence on that.

    You also point out a problem I found when I was researching filters: so many of the good Western studies focus on “first-world problems:” pollen and pet dander. It’s not clear to me that we can generalize those studies to industrial air pollution.

    I think of it like:

    1. There’s lots of air pollution in China.

    2. Studies have linked air pollution to health problems.

    3. Purifiers reduce air pollution.

    ***4. Purifiers reduce health problems.

    It’s #4 that is the missing link. That study you point to is promising, but we need more!

    1. Yes, we definitely need more studies. I’ll try to hook you up with the researchers I know around China. And I hear that students at ISB are looking for science projects — all of these could make fascinating studies!

  3. Dr. Saint Cyr – What is the rule of thumb for how much outdoor 2.5 pollution permeates indoors, in general? That is, if it’s say 200 outside, what would it be indoors – with closed doors and windows, with no air filters or advanced HVAC?

    I’ve heard that windows and walls block out a lot of pmi 2.5. Don’t know if it’s true or not.

    I’m concerned about the air quality at indoor gyms in malls and such.

    1. Good question. In my own testing, indoor air was almost always better than outdoor air with a typical range 50-80%. But Thomas from the DIY team did a report and found even better indoor air, usually in the 30% range. I think a lot of indoor places would have other problems if the air doesn’t move much, especially VOC, formaldehyde, benzene etc…

  4. Your post puzzles me: air filters are not medicines, or something that you assume into your body, and whose main interaction is chemical, like multivitamin, and not physical, so the comparison is overstretched.

    Does the opening of an umbrella make you more or less wet when raining? If less it is useful

    You do not need too much testing to prove that working in a coal mine is dangerous to health, if it can be shown that air purifiers decrease the percentage of particulate in the air I do not see need of detailed studies.

    Obviously, it is different in case of allergies and other contact illnesses, but as far as dust is concerned I wouldn’t worry too much.

Leave a Reply