175 Grams of Brain No One Can Explain

What We Know

Bruce Lee spent the morning of July 20, 1973, going over dialogue for Game of Death, the film he’d paused to make Enter the Dragon — the movie that, in about four weeks, would open in the US and turn him into the biggest star Asian cinema had ever produced. That afternoon he went to the Kowloon Tong apartment of actress Betty Ting Pei, where producer Raymond Chow later joined them. Around six, Lee said his head hurt. Ting Pei gave him Equagesic, an aspirin-and-meprobamate painkiller he’d taken before without incident, and he lay down to rest.

He never got up. When Ting Pei couldn’t wake him, her doctor was called, then an ambulance — sent, at Chow’s direction, to Queen Elizabeth Hospital twenty-five minutes away rather than a hospital minutes closer, buying time to reach Lee’s wife before the story broke. Lee was pronounced dead there that night. He was 32.

It wasn’t his first brush with the same failure. Ten weeks earlier, on May 10, he’d collapsed during a dubbing session at Golden Harvest’s studio, in a room with the air conditioning switched off. He convulsed, vomited, and lost consciousness; doctors diagnosed cerebral edema and brought the swelling down with the drug mannitol. A full neurological workup followed: physical exam, blood-flow study, EEG. It found nothing wrong. He was cleared to work.

The July 23 autopsy, performed by pathologist R.R. Lycette, found the same thing that had nearly killed him in May, except this time it hadn’t stopped: his brain weighed 1,575 grams, about 175 over the upper end of normal, tense against its own casing. No injury, no tumor, no blocked vessel. Nothing explained why the swelling started, either time. A Hong Kong inquest that September returned a verdict of death by misadventure, reasoning that the Equagesic had triggered a hypersensitivity reaction. Lycette himself is on record having floated a different cause first, cannabis, found in Lee’s stomach. The hearing settled on the pill instead.

The Roadblock

The physical evidence that could settle this left with Lee. Hong Kong in 1973 ran no imaging beyond a visual exam of brain tissue at autopsy. No scan of what his brain looked like in the hours before it swelled exists, and no blood or tissue sample was preserved for the sodium and toxicology panels modern medicine would want to run first. What survives is one autopsy report, inquest testimony summarized secondhand, and the memories of the people who watched him die. None of it can be revisited.

Even the doctor who did the autopsy didn’t land in one place. Lycette suggested cannabis intoxication as a candidate cause before the inquest, then the hearing concluded on Equagesic hypersensitivity instead. Marijuana was already a poor mechanistic fit for cerebral edema, and Lee had taken Equagesic before without any reaction. If the pathologist who examined the body in person couldn’t settle on a single answer within weeks of doing it, the case was never going to close on paper alone.

That’s why the field keeps reopening it. Every theory below has to explain both the May collapse and the fatal one in July, using a fifty-year-old lab report and other people’s testimony instead of a body.

Best Guesses

Allergic reaction to Equagesic

This is the actual finding of the actual 1973 inquest, reached by the doctors with direct access to the body and the case file within weeks of the death: pathologist R.R. Lycette’s autopsy, backed by forensic expert Donald Teare, concluded a hypersensitivity reaction to Equagesic’s aspirin-meprobamate combination caused the fatal cerebral edema. Its problem is timing. Lee had taken Equagesic before without incident, and hypersensitivity reactions are supposed to build across repeated exposures, not vanish and then strike without warning on one specific afternoon — a gap the inquest never explained. The panel returned its verdict anyway: “death by misadventure.”

Heatstroke from lost sweat glands

Biographer Matthew Polly, researching Bruce Lee: A Life in 2018, found that Lee had had sweat glands surgically removed from his armpits months before his death, and that both collapses happened under compromised cooling — the May episode in a room with the air conditioning off, the July one during a stretch of hot, humid Hong Kong weather. Heatstroke autopsies commonly show cerebral edema, which fits. But axillary glands are a small share of the body’s total, and Lee reportedly sweated heavily during the May collapse anyway. That’s evidence losing them hadn’t meaningfully impaired his cooling. The authors of the 2022 hyponatremia paper, arguing against a rival theory, called the mechanism too small, alone, to explain a fatal edema.

Hyponatremia from water overload

A 2022 paper by seven Spanish nephrologists reconstructed Lee’s final months in detail: a diet built almost entirely on carrot and apple juice, heavy cannabis use known to drive excess thirst, roughly 15% body-weight loss in two months, and bloodwork from the May collapse suggesting kidney strain. They matched his final hours to documented fatal water-intoxication cases: headache, dizziness, vomiting, rapid decline after drinking water. What the paper can’t produce is a number: no blood test in 1973 ever measured Lee’s sodium level, so the entire case is built by matching a symptom pattern to a modern clinical profile, decades after the fact.

Undiagnosed seizure disorder

Its weakness comes first: every test anyone actually ran on Lee came back clean. The full neurological workup after the May collapse found no abnormality, and the July autopsy found no bitten tongue, a classic physical marker of a seizure. But a 2020 paper in Epilepsy & Behavior argued the May and July collapses read as two seizures from the same undiagnosed disorder, pointing to sudden unexpected death in epilepsy — a mechanism not formally defined in the medical literature until the 1990s, more than two decades after Lee’s doctors examined him with no term for what they might have been looking at.

The Verdict

Lee’s blood and tissue weren’t preserved for the sodium panel or toxicology screen that would settle two of the four theories above outright — the inquest closed within weeks, before anyone had reason to ask those questions. What’s left is a fixed, small set of facts: one autopsy report, two collapses ten weeks apart, testimony recorded secondhand. Every new specialist re-reads it through their own instrument. A team of nephrologists publishing in 2022 saw a kidney case. Epilepsy researchers publishing in 2020 saw a seizure disorder. The doctors in the examining room in 1973 saw a pill.

Short of Lee’s inquest transcript surfacing in fuller form than the secondary record preserved, there’s no new document left to find, and no tissue left in storage to test. The field will likely keep doing what it’s done for years: a new paper every so often, each internally consistent, none able to rule the others out. The 175 grams stay unexplained either way.