Venkata Jonnakuti, MD, PhD

All explainers

Perfusion pressure, dialysis, and the optic nerve

The optic nerve lives on a delicate balance: the blood pressure pushing blood into the eye, minus the pressure inside the eye pushing back. That difference is the ocular perfusion pressure. Hemodialysis can squeeze it from both sides at once. Build the pressures, then run a dialysis session and watch what happens.

Ocular perfusion pressure (OPP) is the driving head for optic nerve head and retinal blood flow: roughly the arterial pressure minus intraocular pressure. Hemodialysis can lower the numerator (intradialytic hypotension) while raising the subtrahend (an osmotic, disequilibrium rise in IOP). Build the pressures from the Goldmann equation, then simulate a four-hour session and its dialysis-side modifiers.

One scene, two steps. First build the pressures: set the blood pressure and the eye, and watch what is left for the optic nerve. Then run a dialysis session on that same patient and see the machine squeeze perfusion from both sides. One scene, two steps. Build the pressures: MAP from the cuff, IOP from the Goldmann equation, IOP = (F - U) / C + Pv, where outflow facility C is the term that fails in glaucoma. Then run a four-hour session on the same eye: ultrafiltration lowers MAP while an osmotic disequilibrium raises IOP.

Interactive explainer by Venkata Jonnakuti, MD, PhD. Open the full page
050100150200 SBP top 150 DBP bottom 85 MAP BP 107 blood pressure MAP = DBP + ⅓ pulse pressuretop and bottom numbers HEMODIALYSIS UF↓ Na+ 8 h BP pump dialyzer 0.0 L fluid removed arm, AV fistula dialysate bag catheter peritoneal cavity 0.0 L removed slow exchange across the lining t = 0.0 h on dialysis heart MAP BP — optic nerve ophthalmic arteryartery to the eye IOPeye pressure — OPP to nerve — — solute stays in the eye, so water follows it inas the blood is cleaned, water moves into the eye drain clogging the eye's drain the eye makes fluid here more fluid in, or a clogged drain, means higher eye pressure drain clogging C outflow, the drain Pv back-pressure — F inflow, the faucet the ciliary body makes fluid, the angle drains it swipe to see more

Push in, push back, what is left one shared scale, mmHg

blood pushes in (⅔ of MAP) — eye pushes back (IOP) — 4050 what is left for the nerve OPP — mmHg for the nerve, —

AutoregulationHow much blood gets through illustrative curve

flow falls flow held steady 0OPP 90 mmHg100%0 blood flow ≈ —% of normal

Optic nerve consequenceWhat it can do over time illustrative

to macula (T) optic disc, right eye NT visual field, right eye a year at this pressure nerve fibre loss ≈ — a teaching model, not a validated estimate

Over the sessionOver the session drag the chart or use the scrub bar

Pressures during dialysis (mmHg) 04080120 Ocular perfusion pressure (mmHg) below the knee: flow itself falls danger zone (OPP < 50) 05080 0h1h2h3h4h
Scrub0.0 h

Blood pressure

Systolic BP (SBP)Top number (systolic)150 mmHg
Diastolic BP (DBP)Bottom number (diastolic)85 mmHg

The eye (Goldmann)

Eye pressure (normal is about 10 to 21)13 mmHg

Glaucoma usually means the eye's drain is clogged, so pressure builds up. Slide it up to see what that does.

Aqueous inflow F, µL/min2.4
Outflow facility C, µL/min/mmHg0.28
Episcleral venous pressure Pv, mmHg9
Blood pushing in — − eye pushing back — = — reaching the nerve

Doctors call what reaches the nerve the perfusion pressure. Above about 50 is comfortable. Below 40, the nerve can be starved.

MAP = DBP + ⅓(SBP − DBP) = — mmHg
IOP = (F − U) / C + Pv = — mmHg (U ≈ 1.2)
OPP = MAP − IOP = — mmHg
Mean OPP (eye) ≈ ⅔ · MAP − IOP = — mmHg · diastolic OPP —

The eye sits above the heart, so pressure at the ophthalmic artery is roughly two thirds of the arm's MAP. Clinicians use this mean ocular perfusion pressure for risk.

Happy with this patient? Run a dialysis session on them Happy with these numbers? See what a dialysis day does

Fluid removed (ultrafiltration, liters)Fluid taken off during the session3.0 L
Pre-dialysis serum osmolality305 mOsm/kg
At 0.0 h MAPBP — IOPeye — OPPto nerve —

Peritoneal dialysis is continuous: solute and fluid shift slowly across the peritoneum, so there is no intradialytic hypotension and little osmotic disequilibrium. The hemodialysis levers do not apply. The chart shows the same four hours for comparison, and the faint curve is the previous scenario.Peritoneal dialysis works slowly all the time through the lining of the belly, so there is no big pressure drop and little water shift. The hemodialysis levers do not apply. The faint line is the previous scenario, for comparison.

What the dialysis team can change

Lowest perfusionLowest supply to nerve

—

In danger zone

— min

Peak eye pressureHighest eye pressure

—

Patient: MAP —, IOP — before dialysis. Edit the patientYour numbers: blood pressure —, eye pressure — before dialysis. Change them

How this simulation works

Pressures. MAP = DBP + (SBP - DBP) / 3. IOP = (F - U) / C + Pv with uveoscleral outflow U held at 1.2 µL/min. MOPP = 2/3 · MAP - IOP. Risk bands follow the epidemiology: 50 mmHg and above well perfused, 40 to 50 a thin margin, below 40 under-perfused.

Hemodialysis session. Over four hours, MAP(t) = MAP0 - D · (1 - (1 - t/240)^2.6), front-loaded because most intradialytic blood pressure loss happens early, with D = (4 + 5 · (UF - 0.5)) × lever factors (gentler ultrafiltration 0.55, holding antihypertensives 0.7, longer schedule 0.75). IOP(t) = IOP0 + R · (t/240)^1.4, late-peaking, with R = (1 + 0.15 · max(0, osmolality - 300)) × vulnerability × lever factors, capped at 18 mmHg (bicarbonate dialysate with sodium modeling 0.4, longer schedule 0.6). Vulnerability is clamp(0.28 / C, 0.6, 2.2): a low outflow facility cannot buffer an osmotic inflow.

Calibration. The typical patient is tuned so a session reproduces the cohort means reported by Lee et al. (2026): IOP rising about 2.6 mmHg to a peak at four hours, MOPP falling about 10 mmHg with most of the fall in the first hour. Their measured means are the hollow circles on the chart. The high-risk eye reproduces their IOP-rise subgroup, which climbed about 7.5 mmHg.

Peritoneal dialysis. The same equations with linear, continuous exchange: the blood pressure fall is 35% and the osmotic rise 30% of the hemodialysis values, spread evenly across the window. Illustrative.

Autoregulation. Relative flow = 1 / (1 + e^(-(OPP - 32) / 6)): a logistic knee near 40 mmHg. Illustrative shape, not a measured curve.

Optic nerve consequence. Session dose is the mean of (1 - flow) across the session; a year of three sessions a week gives loss = 1 - e^(-dose · 156 · 0.006). In Build mode, chronic exposure at the baseline pressure gives loss = 1 - e^(-(1 - flow) · 1.2). Nerve-fibre bundles and 24-2 field points fail in glaucoma order (inferior rim and superior arcuate field first, with a nasal step; fixation and the temporal island last). The scaling is a teaching choice, not a fitted estimate.

Share it: QR code, tour video, embed code
QR code linking to this explainer

Link with the guided tour

Embed the interactive (no site chrome)

A 52-second recording of the tour (1600 px MP4, 900 px GIF)For email or slides: MP4 or GIF.

The QR code opens the full page. Add ?embed=1 to strip the header and prose for slides and wikis; the audience toggle stays. Add #demo=1 to start the tour on load. Scenario links copied from the panel above keep every slider and lever.

What perfusion pressure is

For the optic nerve to stay healthy, blood has to be pushed into the eye against the pressure already inside it. Perfusion pressure is roughly the blood pressure driving flow minus the pressure inside the eye. If blood pressure falls, or eye pressure rises, less blood reaches the nerve. When that happens again and again, the nerve can slowly be starved. That slow loss of the optic nerve is glaucoma.

Where dialysis fits in

During hemodialysis, extra fluid and waste are cleared from the blood over about four hours. Two things can move at the same time. As fluid is pulled off, blood pressure can drop. And as the blood is cleared, water can briefly shift into the eye, so eye pressure can drift up, usually most toward the end of the session.2 A lower blood pressure and a higher eye pressure both shrink perfusion pressure. Someone on dialysis for years meets this many times a week.

What can help

The good news is that the dialysis side has real levers. Removing fluid more gently, using modern dialysis fluid with attention to salt, choosing longer or overnight schedules, and being careful about when blood pressure medicines are taken can all keep perfusion steadier. If you have glaucoma or are at risk, it is worth having your kidney team and your eye doctor talk to each other.

The two equations

Mean ocular perfusion pressure is commonly approximated as MOPP ≈ ⅔ · MAP − IOP, the factor accounting for the height of the eye above the heart. Low OPP is an established association with open-angle glaucoma: a meta-analysis of 43 studies found OPP roughly 2.5 mmHg lower in glaucoma patients than controls, with the association strongest in eyes with higher baseline IOP.1 IOP itself follows the Goldmann equation, IOP = (F − U) / C + Pv, where F is aqueous formation, U uveoscleral outflow, C trabecular outflow facility, and Pv episcleral venous pressure. In glaucoma the failing term is C. A low outflow facility raises baseline IOP and, critically, blunts the eye's ability to buffer an acute inflow of fluid.

Autoregulation and the knee

Retinal and optic nerve head blood flow are autoregulated: over a range of perfusion pressures, vessel tone adjusts so that flow stays roughly constant. Below a knee, autoregulation is exhausted and flow falls with pressure. That is why a threshold near 40 to 50 mmHg keeps appearing in the epidemiology, and why the curve in the inset matters more than any single number. Autoregulatory capacity is itself impaired in glaucoma and in vascular disease, which shifts the knee to the right and narrows the safe range. The curve shown is illustrative, not a measured one.

Intradialytic physiology

Ultrafiltration reduces intravascular volume and can produce intradialytic hypotension, lowering MAP. At the same time, rapid clearance of urea and other solutes from plasma outpaces their clearance from the eye, creating a transient osmotic gradient that draws water into the globe and raises IOP, a form of ocular disequilibrium. A 2026 prospective study of 103 eyes found IOP rising through the session and peaking at four hours while MOPP declined, with a pre-dialysis serum osmolality above 312 mOsm/kg predicting an IOP rise greater than 5 mmHg.2 The net effect on OPP is a falling numerator and a rising subtrahend at once.

The honest nuance

On average across modern cohorts, the historic intradialytic IOP spike has largely disappeared: a meta-analysis of 53 studies found the rise was tied to the old acetate dialysate and that bicarbonate dialysate, standard since the mid 2000s, produces no net rise or a small decline.3 But the same analysis found glaucoma and impaired outflow to be an independent moderator of a clinically meaningful rise. A small cross-sectional series likewise found IOP rising after a session while OPP did not change significantly, a reminder of how heterogeneous the response is.4 The risk is concentrated in the vulnerable eye, not the average one, which is exactly why the outflow-facility slider above changes the story so much.

Dialysis-side levers

Because the drivers are hemodynamic and osmotic, several are modifiable without touching the eye: reduce the ultrafiltration rate and target euvolemia to limit hypotension; use bicarbonate dialysate with sodium modeling to blunt the osmotic gradient; consider longer or more frequent, including nocturnal, schedules for gentler solute shifts; review the timing of antihypertensives on dialysis days; and consider peritoneal dialysis in high-risk glaucoma patients for its continuous, gentler shifts. For known glaucoma, co-management with pre and post-session IOP checks is reasonable. Reading this optic nerve is the same careful, quantitative habit I bring to reading an OCT and to color-coded RNFL analysis, where I have written about the pitfalls of trusting the map over the underlying scan.

What we do not know yet

  • People on dialysis do get glaucoma more often, but no study has yet proved that changing the dialysis settings prevents it. The levers here are based on the physiology, not on a trial.
  • Many people on dialysis have no rise in eye pressure at all. The risk is concentrated in eyes that already have glaucoma or a clogged drain, and in people whose blood is very concentrated before a session.
  • The "how much blood gets through" curve and the "over time" panel are teaching pictures. Your own eye may be more or less resilient.
  • Peritoneal dialysis is gentler on paper. Nobody has compared the two directly for eye outcomes.

What we do not know yet

  • Causality. The association between hemodialysis and glaucoma is observational. No interventional study has shown that modifying ultrafiltration, dialysate, or schedule changes optic nerve outcomes; the levers here are physiologic inference.
  • Heterogeneity. Pooled modern cohorts show little or no mean IOP rise. The clinically relevant signal lives in subgroups: impaired outflow, phakic eyes, and high pre-dialysis osmolality.
  • Measurement. Intradialytic IOP is usually measured by rebound tonometry and post-session by applanation; MOPP is calculated from brachial pressure, not measured at the nerve head. Choroidal and laminar perfusion are not captured by either.
  • Autoregulation. The curve is illustrative. Individual knees vary and shift rightward in glaucoma, diabetes, and vascular disease, so the safe range is narrower than the drawing in exactly the patients at risk.
  • Peritoneal dialysis. The gentler profile follows from its continuous kinetics; there are no head-to-head ophthalmic outcome data.
  • The consequence panel. The dose-to-fibre-loss scaling is a teaching choice. The pattern of loss is anatomically faithful; the magnitude is not a prediction.

References

  1. Kim KE et al. Ocular Perfusion Pressure and the Risk of Open-Angle Glaucoma: Systematic Review and Meta-analysis. Sci Rep 2020. doi:10.1038/s41598-020-66914-w
  2. Lee K et al. Factors Influencing Intraocular Pressure Elevation During Hemodialysis. Transl Vis Sci Technol 2026. doi:10.1167/tvst.15.5.7
  3. Chen SH et al. Changes in Intraocular Pressure During Hemodialysis: A Meta-analysis. J Glaucoma 2021. doi:10.1097/IJG.0000000000001842
  4. Sarr IL et al. Effects of Hemodialysis on Intraocular Pressure and Ocular Perfusion Pressure. Cureus 2024. doi:10.7759/cureus.59138

This explainer is for education and is not medical advice. The simulation, the autoregulation curve, and the consequence panel are simplified teaching models, not validated clinical predictors.

Back to all explainers

Curious about the rest of my work?

I work where RNA bioinformatics meets the eye clinic.

See my research