47⬅ Donald Trump: Report Irregularities / AI Decoy

President Trump's May 2026 medical report contains many irregularities, organized here into familiar information-corrupting buckets that impede efforts to understand his health:

• MORE Omissions ... of multiple test results
• MORE Ignorings ... of multiple clinical problems
• MORE Alterations ... of normal test ranges
• SEE BELOW Decoy ... with irrelevant material
• MORE Ambiguities ... of dates and phrasing
• MORE Misdiagnoses ... and incomplete diagnoses
• MORE Fawning ... including previous reports

However, it must be granted that the presidential physician during Trump's second term has twice accomplished something that had not been previously done since January 2018: release in-depth preventive medical assessments of the president. After the January 2018 physical, all subsequent preventive medical assessments of Trump — which excludes his covid bout — were one-page summaries that left many key parameters unspecified.

Comment: The mere existence of these more detailed assessments gives confidence that the system is operating and suggests the president is receiving attentive and principled medical care, despite what the paper trail might indicate. And so, in several places, I do not accept the possibility that Trump has incompetent doctors.
 
Although the irregularities in the reports raise many questions — especially the changed normal ranges MORE — an outsider finds many obstacles to understanding their origin and significance. Too much of the relevant information is available only to insiders. Genuinely innocent reasons could have propelled some of the irregularities into the report, so full judgment must await access to the inside information.

A decoy seems to exist in the president's May 2026 medical report.

It is the "AI-enhanced electrocardiogram" (AI-EKG) — an utterly worthless test — that took the place of an extraordinarily useful test: the standard 12-lead EKG.

But if the AI-EKG was a decoy, it backfired, because it ultimately can be interpreted as substantial evidence that the Trump has amyloidosis.

AI-EKG chronology and results

• Oct. 1, 2025 — The president's physician reports: "His cardiac age — a validated measure of cardiovascular vitality via ECG — was found to be approximately 14 years younger than his chronological age" MORE.

• Jan. 1, 2026 — The Wall Street Journal reports: "The White House provided the Journal with a summary of a Mayo Clinic AI-assisted analysis of Trump's electrocardiogram that estimated the president's cardiac age to be that of a 65-year-old" 1.

• May 29, 2026 — The president's medical report states: An "AI-enhanced electrocardiogram (ECG) analysis estimated his cardiac age, to be approximately 14 years younger than his chronological age" MORE.

Why the AI-EKG might be a decoy

• Because of its name, the AI-EKG camouflaged the absence of the standard EKG.
  • The standard EKG is an omission from the May 2026 report MORE.

• The name "AI-enhanced EKG" is misleading. It suggests that it can do more than a standard EKG, but exactly the opposite is emphatically true. A giant information loss occurred when the AI-EKG displaced Trump's standard EKG.

• The name "AI-enhanced EKG" is incorrect. It is an umbrella term that encompasses at least six distinct subsystems 2 3. Trump's report included the results of only one of the subsystems, the AI-EKG-age.
  • The AI-EKG-age subsystem calculates a "cardiac age" number for the patient.
  • There is also an AI-EKG-amyloid subsystem that calculates a probability that the patient has amyloidosis.
  • It is likely that Trump had an AI-EKG-amyloid test, but its result was not publicly reported. I have therefore listed it as an omission in the May 2026 report MORE.

• The May 2026 report misrepresents the output of the AI-EKG-age in multiple ways:
  • It erroneously claimed the AI-EKG-age measures "cardiovascular vitality."
  • It erroneously claimed the AI-EKG-age assessment of vitality was "validated." (Doubling-down, it appears.)
  • The prognostic value of the cardiac age number has never been validated in octogenarians.
  • It failed to mention that the AI-EKG-age is not FDA-approved 4.
  • It failed to adequately name the test, not even saying it was from the Mayo Clinic.
  • It provided no pointers that would aid a reader in interpreting the result.
  • It overstated the positivity of Trump's result, letting the user make the obvious (and incorrect) assumption that the "cardiac age" result could be simply interpreted as a linear age, when in fact the relation to chronological age is not linear at older ages.

• The AI-EKG-age could be a decoy in time, too. The test was performed before October 2025, yet it is reported in an official document dated May 2026 without saying whether the printed result is from pre-October or some later time.
  • That is sloppy and confusing.
  • The undated AI-EKG-age result is therefore listed as an ambiguity in the May 2026 report MORE.

The information loss

The AI-EKG-age reported for the president does only one thing: calculate a "cardiac age" for the patient. One lousy number.

Compare that to the amount of information available in a standard 12-lead EKG... I have two EKG textbooks from the 1980s: one takes 496 pages to say what the standard EKG can show 5 and the other takes 621 pages 6.

As a specific example, by producing just a "cardiac age" it is impossible for the AI-EKG-age to say anything at all about amyloidosis. By contrast, the standard 12-lead EKG can show multiple different signs of amyloidosis.

Cynically, therefore, substituting the AI-EKG-age for the standard EKG suppresses any and all specific abnormalities that the standard EKG might have disclosed.

The AI-EKG-age result is not "cardiovascular vitality"

My readings in the medical literature (see references below) failed to explain why the May 2026 report labeled the output of the AI-EKG-age subsystem as "cardiovascular vitality." That is a nice-sounding phrase, which carries a certain connotation, but nowhere is it mentioned in any of the Mayo Clinic's publications that I reviewed.

And that is just what one would expect. The notion of "cardiovascular vitality" does not exist in day-to-day clinical medicine. Cardiology excels at quantifying several aspects of how the heart performs, but there is no accepted way to combine them, and there is no combination so brazen to label itself as "vitality."

The Mayo researchers call it "a biomarker of age" 7.

The AI-EKG-age result is overstated

Trump's result sounds impressive, but as the image below shows 8, the AI system awards a 65 year old cardiac age to a substantial fraction of 79 year olds (indicated by the black dot). In fact, it looks like the average cardiac age given to 79 year olds is about 69 — 10 years younger than Trump's age at the time of the test.

Clearly, the AI system is less accurate in older patients, and this alone should give pause about attaching a meaning to Trump's AI-EKG-age result.

AI age predictions vs. chronological age. Straight line shows where the two ages would be identical. Each of the thousands of test patients contributes to the dots. Black dot indicates President Trump's position: his chronological age is 79 and his AI estimated cardiac age is 65. From 8.

Trump's AI-EKG-age result is 100% useless

The AI-estimated age for Trump's heart has no firm medical significance. In other words, it may sound significant to say that his heart has similarities to a 65 year old's, but there is no evidence to suggest that the similarities make any difference in how the heart will behave over time in the real world.

The Mayo AI-EKG-age subsystem was trained by feeding it hundreds of thousands of electrocardiograms where the age of the patient was known. That's all there was. The system got no data on how long the patients lived afterwards and it didn't know how sick the patients were at the time of the electrocardiogram.

So it's a pretty blind system, and that's the first reason it's useless.

The Mayo team performed a later study in much younger patients than Trump (average age 54) and found that an AI-estimated age younger than the chronological age conferred no survival benefit when compared against patients who had an AI-estimated-age close to their chronological age 7.

Survival according to the gap between AI predicted heart age and chronological age. Survival is better when the lines are closer to the top. Bottom two lines show poorer survival when the AI predicted heart age is higher than the chronological age (one line for somewhat higher and a worse line for much higher). Three lines cluster at the top and are statistically the same. These three lines are for patients whose AI estimate was the same, somewhat lower, or much lower than their chronological age, Trump included. Figure is from 7.

So these are the second and third reasons the system is useless: no prognostic testing in patients of Trump's age, and even if we extrapolate the testing in younger patients, there is no news for Trump except that he's average.

The fourth reason the result is useless is that it leads to no actionable steps. If a patient's heart reads out as excessively old, what can be done about that? The patient cannot be made younger! Perhaps in the normal world a cardiologist might run a slew of tests to try to find the reason(s) the heart is reading out as old, but Trump is already getting those tests, so doing the AI-EKG-age adds zero.

Speculation: For this last reason, especially, it is likely that the real point of getting Mayo's AI-EKG-age was not to get the age reading, but to get the readings from the other subsystems in the Mayo AI-EKG dashboard... primarily the AI-EKG-amyloid that detects early amyloidosis.

The AI-EKG-age suggests Trump has amyloidosis (Speculation #1)

With apologies, this is a rather complicated bit of speculative reasoning, written only secondarily for laypersons, most of whom will probably want to skip it.

1. There has been little published research on how readings from the standard electrocardiogram (the kind that was invented in 1911) change with age. It has been noted that the "R-wave height" increases with age 9.
   
   An idealized single heartbeat as it appears on an electrocardiogram. The tallest spike is called the "R wave" and its height has significance.

2. Longstanding hypertension increases the height of R-waves.

3. The builders of the Mayo AI-EKG-age subsystem identified hypertension as a factor causing their software to label a person's cardiac age as older than their chronological age 8.

4. From all the above, it is reasonable to conclude that taller R-waves is a significant factor in leading the AI-EKG-age subsystem to label a heart as old. Conversely, but equivalently, shorter R-waves will lead the software to label a heart as young.
   • The AI-EKG-age subsystem was trained on a group of 775,000 patients 7 in whom approximately half had hypertension 8a 10a. These patients may be expected to have a higher than normal R-wave height, and there will be more of them among older patients in the training group. Reasonably, this would lead the software to strongly associate tall R-waves with older hearts.

5. Lower-than-normal R-wave height is an electrocardiographic hallmark of cardiac amyloidosis 11, which tends to be a disease of older people.

• From 2000 to 2019 the Mayo Clinic saw only 2541 patients with cardiac amyloidosis 12. So even if every one of those cases has been used to train AI-EKG-age (an impossibility), that would only be 0.3% of the training group. This small percentage of short R-waves in old hearts in the training cases is going to be swamped by the very large percentage of training cases with tall R-waves in older hearts.

6. Thus, the sofware will probably read short R-waves as an indicator of young age, so that amyloidosis patients are consistently rated as having younger hearts than their chronological age.

Therefore, if this is all true, then the lower AI-estimated-age for Trump is exactly what would be expected if he had cardiac amyloidosis. The lower AI-estimated age is actually evidence for Trump having cardiac amyloidosis.

I wouldn't take this conclusion to the bank, but I think at this point the onus is on the Mayo people to prove that amyloidosis does not cause a systematic error in their AI-EKG-age subsystem.

The AI-EKG-age suggests Trump has amyloidosis (Speculation #2)

The AI-EKG is available only from the Mayo Clinic because it is still experimental. So why in the world would the White House go to the trouble of asking for the test?

A potential answer, as hypothesized earlier, is that the Mayo AI-EKG system has six (or more) subsystems, and that the president's physician was interested in other subsystems more than the AI-EKG-age subsystem.

This multiplicity of subsystems is easily seen in a snapshot of the Mayo "AI-EKG dashboard" that collects results from the multiple subsystems.

This screen grab from 2 shows the Mayo AI dashboard, which includes separate readouts for multiple different parameters. Amyloidosis risk is one of those parameters, although it is not shown in this example. (The amyloidosis readout is shown, however, in 3.) This dashboard shows five readouts for one patient, and shows how probabilities and calculations have changed for that patient over several years.

Performing the AI-EKG-age suggests Trump has amyloidosis if one accepts that his medical team was actually interested in the AI-EKG-amyloid, the subsystem that calculates the patient's probability of amyloidosis by analyzing the standard EKG.

The Mayo Clinic already has considerable experience with the AI-EKG-amyloid subsystem, having enrolled more than 20,000 subjects in a clinical trial before Trump's EKG was analyzed 13.

Why the AI-EKG-amyloid could be so helpful

The AI-EKG-amyloid system could be extremely useful in Trump's case. The White House has Trump's standard EKGs on file back to 2018 (if not earlier) so they could have fed them all at once to Mayo, which could then plot how the probability of amyloidosis changed over the years, as in the made-up example below.

This is not Trump's results, but a made-up example of how the Mayo AI-EKG-amyloid subsystem could be useful in a patient like him. It shows that the subsystem rated the patient's 2021 EKG to have a zero chance of indicating amyloidosis, whereas the patient's 2024 EKG is 80% likely to be indicating amyloidosis. Mayo researchers say that their AI-EKG-amyloid subsystem has, in some patients, provided indications of amyloidosis years before anything else made them suspicious 2.

If a physician saw this type of plot, it would more than justify aggressive testing for amyloidosis, which translates to a biopsy analysis (probably not, for the president, unless recent surgical samples are on file) or a TcPYP nuclear scan of the heart (takes hours).

Conclusion

From a public relations standpoint, the AI-EKG-age brought multiple benefits. It enabled suppression of whatever abnormalities were on the standard 12-lead EKG and it helped paint a rosy picture of Trump's cardiac health, regardless of its actual state.

If the White House came clean about the circumstances related to the AI-EKG, perhaps an innocent truth behind all this would be disclosed, but based on the information that is public now, and the information in the scientific literature (which I cannot claimed to have scrutinized to the maximum possible degree), I can think of no reason to include the AI-EKG-age in Trump's medical reports other than public relations.

Such a conclusion has several weaknesses. My reconstruction of these motivations is necessarily subjective. Various areas on this page are marked as speculation. And perhaps I have missed a reference or two from the literature or have been too narrow in clinical reasoning. If so, please let me know and I will correct and apologize. But there is much here that is on firm ground, enough to warrant a "decoy" hypothesis.

Finally, there is the co-existence of the AI-EKG-amyloid with the AI-EKG-age at the Mayo Clinic. The coincidence is absolutely arresting. It equals the famous coincidence in the movie Casablanca: "Of all the gin joints, in all the towns, in all the world, she walks into mine." But here, of all the experimental, unapproved diagnostic tests being developed in all of medical science, in all the world, Trump's EKG walks into the Mayo Clinic's AI-EKG joint, where testing for amyloidosis is the specialty of the house. The fundamental things apply.