Anuloma Viloma Timer
Across six randomised trials and 525 people, alternate nostril breathing lowered systolic blood pressure by about 7 mmHg more than control. That headline comes from a 2024 systematic review and meta-analysis in Complementary Medicine Research by Tae Gwang Nam and colleagues (Karger). The pooled difference was 7.16 mmHg for the top number and 5.16 mmHg for the bottom one.
It is a good result, and it is the one that travels. The number worth sitting with is a different one in the same paper: 93%.
That is the I squared for systolic pressure, a statistic that measures how much the individual trials disagree with each other beyond what chance would produce. For diastolic pressure it was 87%. Anything above 75% is conventionally called high, and the authors flagged it themselves. Put plainly: some of the six trials found large drops and some found small ones, and the 7 is an average across studies that were not really doing the same thing.
Most coverage treats heterogeneity as the small print. This week it read more like the main finding, because every other study in the week’s research, and every viral breathing trick in the week’s feeds, pointed at the same explanation for it. The trials disagree because “alternate nostril breathing” is not one protocol. It is a family of them, and the details change the result.
Which nostril, and it is not the one the internet says
The clearest demonstration came from a 2025 randomised controlled trial by Satyam Tiwari and Arnav Bhavsar, presented at the BIOSIGNALS biomedical engineering conference and published in its proceedings (SciTePress). Ninety healthy adults were randomised to ten minutes of left-nostril breathing, right-nostril breathing, or unstructured breathing.
The left-nostril group’s heart rate variability fell: SDNN down 27.0% and RMSSD down 25.1%, with stress and sympathetic indices rising. The right-nostril group saw systolic pressure fall by 5.5 mmHg and diastolic by 3.3 mmHg, with little change in the other autonomic markers.
Set that next to the trick that circulated all week: close the right nostril, breathe through the left, fall asleep faster. On these numbers, ten minutes of left-only breathing moved the heart in the opposite direction to the one a sleep hack would want.
Caveats first, because they matter. This is a conference paper, the abstract is what is publicly visible, the participants were healthy, and it measured one ten-minute session rather than weeks of practice. It does not prove that the left nostril is “activating”. What it does show is that two practices which look almost identical from the outside, one hand on the nose and slow breathing through one side, produced measurably different results depending on a single choice.
The hatha yoga tradition the hack is borrowed from never asked anyone to make that choice. The Hatha Yoga Pradipika, compiled by Svatmarama in the 15th century, calls the left nostril chandra, the moon, and the right one surya, the sun. Its central breathing practice does not pick one. It breathes in through the left and out through the right, then in through the right and out through the left, round after round. The point of the alternation was that neither side runs the session. The viral version kept the moon and dropped the instruction.
How long, and the dose is not trivial
The second variable that differs between trials is time. A 2022 study in Sensors tested it directly (PMC). Forty healthy adults (41 recruited) practised slow deep breathing for 5, 15 and 30 minutes on separate days, in random order, with heart rate variability recorded afterwards.
Five minutes produced few significant changes. SDNN rose by roughly 8 to 13% after five minutes, 17 to 23% after fifteen, and 18 to 32% after thirty. The authors chose the 30-minute dose for the patient arm of the study.
Two details are easy to miss. The breathing in that study was 4 seconds in and 6 seconds out, six breaths a minute, so the “slow breathing” being dosed already had the longer exhale built in. And the gap between 15 and 30 minutes was much smaller than the gap between 5 and 15. If the trials in a blood pressure meta-analysis range from a few minutes a day to half an hour, some of that 93% is simply dose.
How fast, and the dramatic version did not win
The third variable is pace. The Airways to Alteration (A2A) study, published in Frontiers in Psychology in June 2026 (Frontiers), compared fast, high-ventilation conscious connected breathing with a body scan meditation in 24 experienced adults. It was open-label and the authors describe it as proof of concept. Both groups improved on stress, anxiety, depression and wellbeing, with no significant difference between them.
That is not a finding about alternate nostril breathing, and it should not be borrowed as one. It is a useful counterweight to the idea that more intensity buys more effect. In a small trial, the most dramatic breathwork on offer did no better than lying still and paying attention to your body. Speed is a protocol choice, and it is not obviously the one doing the work.
The sentence that already had the checklist
Here is what ties the week together. When a meta-analyst finds 93% heterogeneity, the standard advice is better reporting: future trials should specify exactly what participants did, for how long, at what count. That advice is roughly 1,600 years old.
Yoga Sutra 2.50, in Patanjali’s short text on the breath, names three movements: the breath going out, the breath coming in, and the breath held. It then says each is regulated by three things: desha, place; kala, time; and sankhya, number. Place, time and count. It is a protocol specification in one line, and it is the exact list of variables the modern studies above turned out to be sensitive to: which nostril, how long, what rhythm.
The Hatha Yoga Pradipika adds the fourth variable, progression. Verse 2.15 says the breath should be brought under control “by slow degrees”, the way lions, elephants and tigers are tamed. Nothing in the chapter rewards intensity on day one. The count lengthens because the shorter count became easy first.
Even box breathing, which spent the week being sold as a modern performance drill, fits inside this. Its four sides (in, hold, out, hold) are all in that list from 2.50. What the square changes is the ratio: it makes every side equal. The classical alternate nostril count is 1:4:2, which for a beginner without the hold becomes 1:2, four in and eight out. And bhramari, the humming “bee breath” that the Hatha Yoga Pradipika lists among its practices, is another way of doing the same thing: giving the exhale a sound so you can hear when it is being rushed.
The Global Wellness Institute’s Breathe Initiative trends for 2026 names “personalization of breathing” and “micro-dosing breath” among the year’s trends. Read generously, that is the market rediscovering the same point: the variables matter, so choose them deliberately. Read less generously, a lot of “personalisation” is a reason to sell you a protocol you could have written yourself.
What this means for your next session
You cannot fix a meta-analysis. You can make your own practice repeatable, so that whatever it does for you, it does consistently and you can tell. Use the sutra’s three words as a checklist.
Place: both nostrils, alternating. Not the left alone. In through the left, out through the right, in through the right, out through the left. That is one round.
Count: twice as long on the way out. Four counts in, eight out. If eight is a struggle, use three and six. The ratio matters more than the numbers. Leave holds out until the rhythm is effortless, and leave them out entirely if you have high blood pressure, a heart condition or are pregnant, unless your doctor agrees.
Time: start at five minutes, aim for fifteen. Five minutes is a real start and the easiest habit to keep. The Sensors data suggest most of the extra benefit arrives somewhere between five and fifteen minutes, so that is the first extension worth earning, by slow degrees, over weeks.
And keep it the same. Same time of day, same count, same length, for a few weeks before you change anything. A protocol you vary every night is an unblinded trial with no control.
If you are practising for blood pressure, keep your medicine. None of the six trials tested breathing as a replacement for treatment, and the 93% is a reason not to count on any particular number. The sensible reading is that slow, even, alternate nostril breathing is a low-cost practice that may help alongside what your doctor has already advised, and is worth mentioning to them.
Sharad Navratri ends on Monday 19 October, with Dussehra on the 20th. If you have kept a short practice going through the nine nights, this is the moment to write your protocol down, rather than let it drift.
This week’s podcast episode, From Mouth Tape to Moon Nostrils: The Science of Extended Exhales, covers the long exhale itself and why it leans the nervous system toward rest. This piece is the other half: whatever the exhale does, you only find out if you do it the same way enough times to notice.
The Anuloma Viloma Timer holds your count for you, sets the ratio and length, and keeps a streak so the “same way, every day” part takes care of itself. It is free on the App Store, no subscription necessary.
The honest summary
Alternate nostril breathing lowered blood pressure on average across six trials, and the trials disagreed with each other almost as much as they could. The week’s other studies suggest why: one nostril versus the other, five minutes versus thirty, and fast versus slow each moved the results. An ancient sentence asked for place, time and count to be fixed before anything else. On this week’s evidence, that is still the first thing to get right.