Do air purifiers actually improve health, and which types should you avoid?

This page covers what purifiers do for health. For how HEPA and carbon filters work, see Air Purifiers, Filters, and Ventilation. For smoke and cooking sources, see Protecting Yourself from Wildfire Smoke and Are Gas Stoves Bad for Your Health?.

Particles drop, but health effects are small. Across blood-pressure trials, purifiers removed about 60% of indoor PM2.5 on average, with a range of 23% to 82%⁶.

Blood pressure. A pooled analysis of 17 true-versus-sham trials found systolic pressure 2.35 mm Hg lower, and 3.62 mm Hg lower once high-risk-of-bias studies were removed⁶. In older adults in Detroit, three days of filtration lowered systolic pressure by 3.2 mm Hg⁸. In Shanghai students, a 57% particle cut over 48 hours lowered blood pressure and several inflammatory markers⁷.

Asthma. Results conflict. HEPA cleaners cut unscheduled asthma visits in children exposed to secondhand smoke³, and a package of alerts plus HEPA at home and school improved asthma control in a dust-storm region¹⁰. But classroom HEPA filters did not significantly reduce symptom-days⁴, and home cleaners that cut PM2.5 by 64% did not improve asthma outcomes⁵.

Sleep. A small two-week crossover pilot found about 12 more minutes of sleep a night, but more time awake after falling asleep⁹.

Choose by CADR. The clean air delivery rate (CADR) rates particle removal per minute. The EPA's sizing chart works out to a CADR of about two-thirds of the room's floor area, such as 195 for a 300 ft² room with 8-foot ceilings¹. Higher fan speeds and longer run times filter more air¹. CADR does not rate gas removal¹.

Types to avoid. Ozone generators deliberately make ozone, a lung irritant. They can exceed health standards even when used as directed, and they do not remove particles². Electrostatic precipitators, ionizers, some UV units, and plasma cleaners may also emit ozone¹. In real homes, an ozone generator near a pine-oil cleaner raised fine particles¹¹, and an ion generator raised ozone and, near an air freshener, ultrafine particles¹².

Blood pressure trials. Chen and colleagues ran a randomized, double-blind crossover trial in 35 healthy Shanghai students who alternated true and sham purifiers for 48 hours with a two-week washout⁷. True purification cut PM2.5 by 57% (96.2 to 41.3 µg/m³) and lowered systolic pressure by 2.7%, diastolic by 4.8%, and exhaled nitric oxide by 17%, along with interleukin-1β (68%) and myeloperoxidase (33%)⁷. Lung-function gains were not significant⁷. Morishita and colleagues studied 40 older adults in a Detroit senior building across three 3-day exposures: sham, low-efficiency HEPA-type, and true HEPA⁸. Personal PM2.5 fell from 15.5 to 10.9 and 7.4 µg/m³⁸. Any filtration lowered systolic pressure by 3.2 mm Hg (95% CI -6.1 to -0.2) and diastolic by 1.5 mm Hg (-3.3 to 0.2); most secondary outcomes did not improve, and the true-HEPA effect alone (-2.9, -6.2 to 0.5) was not significant⁸. Faridi and colleagues pooled 17 trials with about 880 participants: -2.35 mm Hg systolic (-4.5 to -0.2) and -0.81 diastolic (-1.86 to 0.24)⁶. Apart from the Chinese studies, trials were in low-pollution settings⁶.

Asthma trials. Lanphear and colleagues randomized 225 children aged 6 to 12 with asthma who lived with smokers to active or inactive HEPA cleaners³. The active group had 42 fewer unscheduled asthma visits (P = .043), but parent-reported symptoms, exhaled nitric oxide, air nicotine, and cotinine did not differ³. In the School Inner-City Asthma Intervention Study, 236 students in 41 schools were randomized to classroom HEPA or sham filters⁴. Symptom-days per two weeks were 1.6 versus 1.8 (incidence rate ratio 1.47, 95% CI 0.79 to 2.75), a null result; a school pest-management arm was also null⁴. The MEDEA trial randomized 182 children in Cyprus and Greece to no intervention, alerts plus advice to stay indoors, or that plus HEPA purifiers at home and school¹⁰. After three months, the combined arm improved on the childhood Asthma Control Test by 2.63 points (0.72 to 4.54) and FEV1 by 4.26 percentage points (0.54 to 7.99) versus controls¹⁰. The package cannot separate the filter from the advice. AIRWEIGHS randomized 164 children to HEPA or filterless units: PM2.5 fell 64.4%, yet there was no improvement in primary or secondary asthma outcomes at 12 weeks⁵.

Sleep. Lamport and colleagues tested 30 healthy adults in a double-blind crossover, two weeks per condition⁹. The HEPA filter increased total sleep time by about 12 minutes and time in bed by 19 minutes a night, but wake time after sleep onset was higher, and many outcomes did not change⁹.

Ozone-producing devices. Hubbard and colleagues released terpenes from a pine-oil cleaner in three homes while an ozone generator ran; particle number and mass rose when both were present¹¹. Waring and Siegel tested an ion generator in a 27 m³ room: ozone rose in every configuration, formaldehyde and nonanal rose (within uncertainty in some cases), and ultrafine particles rose with an air freshener present; they advised limiting ion-generator use¹². The EPA states that ozone can cause chest pain, cough, and shortness of breath and may worsen asthma².

Null and conflicting results. Two of the four asthma trials were null^4,5, and the positive ones paired filters with other measures or changed visits more than symptoms^3,10. The investigators of the school trial noted that baseline allergen, particle, and symptom levels may matter⁴. The blood pressure effect is small and was not significant in some single comparisons⁸, though it grew when biased studies were removed⁶. None of these trials lasted long enough to show whether purifiers prevent heart attacks or new asthma.

Bottom line (confidence 3 of 5). Purifiers reliably remove particles and probably lower blood pressure by a few mm Hg. For asthma and sleep the evidence is mixed. Avoid ozone-emitting designs.