Humming, and the Gas in the Sinuses

Svātmārāma's instruction for bhrāmarī is one verse, and it is mostly sound:

वेगाद्घोषं पूरकं भृङ्ग-नादं भृङ्गी-नादं रेचकं मन्द-मन्दम् । योगीन्द्राणामेवमभ्यास-योगाच्चित्ते जाता काचिदानन्द-लीला ॥ ६८॥

vegād ghoṣaṃ pūrakaṃ bhṛṅga-nādaṃ bhṛṅgī-nādaṃ recakaṃ manda-mandam | yogīndrāṇām evam abhyāsa-yogāc citte jātā kācid ānanda-līlā ||

Inhale quickly, with a sound — the note of the male bee. Exhale slowly, slowly — the note of the female bee. By this practice there arises in the heart of the great yogins a certain play of bliss.

Note what is not in this verse: the ears. Plugging them with the fingers is standard instruction almost everywhere bhrāmarī is taught today, and it is not in the Haṭha Yoga Pradīpikā. It comes from the Gheraṇḍa Saṃhitā, where bhrāmarī is a different practice altogether — after midnight, in silence, ears closed with the hands, listening for an internal sound that goes from cricket to flute to bell to drum. Two techniques, one name. Worth keeping apart.

Fifteen-fold

Nitric oxide is made in the paranasal sinuses — the air-filled cavities in the bones around the nose — in large and continuous quantities. This was established the hard way, by measuring people whose cilia do not work: in Kartagener's syndrome the nasal NO is nearly gone, which located the source.

The gas matters downstream. It dilates the small vessels of the lung and improves the match between air and blood; it is antimicrobial in the sinus itself. Breathe through the nose and you carry a dose of it into the lungs on every inhale. Breathe through the mouth and you do not.

In 2002 two researchers at the Karolinska asked what would happen if the air in the sinuses were shaken. The sinuses connect to the nasal cavity through a narrow opening called the ostium; gas exchange across it is slow. Oscillating airflow, they reasoned, ought to speed it up.

Ten subjects. Nasal NO measured during quiet exhalation, and during humming, at a fixed flow rate.

Humming raised it fifteen-fold.

The part that gets left out

A follow-up from the same group, the next year, is the one that should be quoted alongside it, because it changes the practice.

The NO peak declines with repetition. Hum, and you get a large spike. Hum again immediately, and you get less. Again, less still — a stepwise fall. After a silent interval of three minutes it recovers completely. And measured immediately after repeated humming, the silent nasal NO is 5–50% below where it started.

The same paper found the ostium size to be the critical variable, with humming frequency mattering too — but both of those were established in the model, not in people, so treat the pitch question as open.

The verse, read against this, is oddly well-shaped. A quick inhale with sound, then a long slow hum on the exhale, done as a set — and the exhale is where the humming is, which is where the gas gets moved. Whether Svātmārāma knew anything about sinus ostia is not a question worth asking. He knew what the practice did to the person doing it.

Om, in a scanner

The other measurable thing about a sustained hum is what it does upstairs.

Twelve volunteers chanted om aloud inside a 3-tesla magnet, in alternating blocks, against a control of pronouncing "ssss" — a sound with the same effort and airflow and none of the vibration. During om there was significant deactivation in the right amygdala, and bilaterally in the anterior cingulate, hippocampus, insula, orbitofrontal cortex, parahippocampal gyrus, and thalamus. The "ssss" control produced neither activation nor deactivation in those regions.

The authors note the pattern resembles what is seen with vagus nerve stimulation — the implanted device used clinically for depression and epilepsy — and suggest the vibration may be reaching the vagus through the ear or the pharynx.

Two details worth holding onto, because they get inflated in retelling. The amygdala deactivation was on the right only, not both sides. And a deactivation in an fMRI is a drop in blood oxygenation in a voxel; it is not a measurement of calm.

What is actually established

That humming moves a great deal of nitric oxide out of the sinuses, in ten people, at one sitting, and that the effect exhausts within a few breaths and returns after three minutes of quiet. That a sustained vocal om quiets limbic blood flow relative to a hiss, in twelve people, once.

That is a thin evidentiary base for a practice that has been taught for six centuries, and it would be dishonest to dress it up. What is not thin is the direction of it. Every one of these findings points the same way the verse does: that the sound is not incidental to the technique, and that the exhale is where the work is.

The instruction stands on its own merits. Inhale with the note of the bee, and let the exhale be slow, slow.