Intraocular Pressure During Inversion
by Loren Fishman
There has been controversy regarding whether yoga inversion postures cause an untoward increase in intraocular pressure (IOP) that might be harmful for people with glaucoma. In order to clarify matters, we asked a healthy 67 year-old without glaucoma to perform headstand, handstand and other inverted postures while we measured left and right IOP with a Tonopen tonometer.
Headstand essentially doubled the IOP from 14-15 mmHg to 31-34 mmHg. They rose to that level within 30 seconds, and stayed elevated, but they rose no further until 13 minutes, when they were 38 and 33 mmHg on right and left respectively. Inverted lotus was studied at 14 minutes, inverted twists, both Parsva Sirsasana and Parivrttaikapada Sirsasana at 17 minutes, Eka Pada Sirsasana at 18 minutes, legs horizontal at 20 minutes, and Valsalva maneuver (holding breath and creating pressure in abdomen) at 21 minutes.
None of these maneuvers seemed to have much effect on the IOP: it stayed in the 29-34 mmHg range throughout.
Other inverted postures were then tested. While handstand reflected nearly the same doubling of IOP that was seen in headstand, the shoulder stand yielded values that were only 40% higher than those seen at rest. The plow elevated IOP by only 3 mmHg. Lying supine with thighs vertical and calves supported on the seat of a chair caused no rise, and viparita karani, the restorative pose, actually produced an IOP drop of one mmHg in each eye.
More is Less
In headstand, handstand and shoulder stand, the greatest rises in IOP occurred within the first minute of inversion. After that, the pressure dropped 2-5 mmHg, possibly reflecting the body’s adjustment to the inverted condition. This decrease continued until minute 13, when there was a bilateral rise. This gives warning about “trying” headstand for a short time in patients at risk. That “trying” period may be more hazardous than the next 13 minutes. However sustaining inversion for longer than that appeared to raise IOP even further, though quite gradually.
The series of variations in headstand did not change the IOP substantially.
The Valsalva maneuver (forced attempted expiration against the closed glottis) produced some blurred vision in the subject, and should be avoided by anyone attempting to duplicate this study.
We found six research articles in the last 35 years that had similar orientation and goals to ours. Most of their conclusions agreed with ours. Cook, et. al. found that inversion with inversion boots (so-called “gravity” boots) raised IOP from an average of 16 mmHg to 27 mmHg. (1). Galin, et. al. compared normal people’s IOP with eyes of people with glaucoma, finding that glaucomatous eyes, though higher in their baseline intraocular pressure, actually rose in pressure very little more than the eyes of normal people. The normal eyes’ pressures came down more than 10% in 5 minutes, the duration of Galin’s test, but glaucomatous eyes did not.(2)
All eyes’ IOP returned to normal within 15 seconds of stopping the inversion. Galin found absolutely no correlation with age or height in either group, in a total of 31 patients.
Friberg (3) analyzed the Goldmann equation, which equates IOP to the rate of production of aqueous humor in the eye divided by the facility of outflow of aqueous humor from the eye, plus the external episcleral venous pressures. Since inversion seemed to leave production and outflow essentially unchanged, he reasoned that it was the increased pressure in episcleral veins that was responsible for the IOP increases commonly seen during inversions. A feature of his study of 16 patients was the gonioscopic identification of blood in Schlemm’s canal in half of them after inversion, thereby confirming the continuity of the lumina of valveless episcleral veins with Schlemm’s canal. This is a frequent finding in certain clinical conditions, such as arteriovenous fistulae, that cause increased episcleral venous pressure and, subsequently, increased IOP as one type of secondary glaucoma. Still, Friberg also found that IOP essentially doubled within a very short time after inversion, and returned to normal in less than 10 seconds following it.
Bronner (4) found a decrease in retinal sensitivity with increased IOP on clinical grounds, something later documented by Friberg and Sanborn (5) using electrophysiological means. These latter investigators measured the electrical impulses generated by equal amounts of light striking the retinas before and during inversion. They found a significant drop in the size of these potentials in inversion, and cautioned against it under any circumstances, suggesting that continued inversion could impair and even destroy vision.
None of these other studies used actual yoga positions, but they observed very similar rises in IOP during inversion. It is gratifying to find that the general observations of medical science do apply to yoga positions as well. None of these studies examined shoulder stand, plow, nor the other asana investigated in our single case either. Most of these inverted positions show much smaller, and therefore much safer elevations in IOP.
Both Sides of the Story
On the other hand, here are retinal images, representative of a fluorescein angiographic study of a 67 year old man who has stood on his head in the Iyengar method for 30 minutes once weekly for 35 years.
The blood vessels pictured here are perfectly normal.
1 Cook J, Friberg TR. “Effects of inverted body position on intraocular pressure.” American Journal of Ophthalmology; 98:784-87, 1984.
2 Galin MA, McIver JW, Magruder GB. “Influence of position on intraocular pressure.” American Journal of Ophthalmology, 55:720, 1963. Friberg TR. “Portable transducer for measurement of episcleral venous pressure.” American Journal of Ophthalmology/ Sept. 1986: 396-7.
3 Bronner A, Franck A, Margraft C. “Influence de la position du corps sur le tonus oculaire.” BSOF, 1976; 6: LXXVI, 657-661.
4 Friberg TR, Sanborn G. “Optic nerve dysfunction during gravity inversion. Pattern reversal visual evoked potentials.” Archives of Ophthalmology; 103-1687, 1985.