Ruti Aga: Six Years for a Prescription From a Clinic That Never Existed

The clinic named on the prescription does not exist. The signature in the treating-physician box belongs to a family member with no medical registration. The prescription pad was said to have come from a former workplace, and that workplace confirmed she had never appeared on its staff list. Three layers of paper, three independent verifications, all three collapsed inside a single review cycle at the Athletics Integrity Unit.

Ruti Aga, 32, the Ethiopian marathoner, has been banned for six years running from 10 September. Two years for the presence of a prohibited substance in her sample. Four years for tampering with evidence. The longer half of the sentence does not belong to the substance. It belongs to the paper she used to explain it.

In women's road running, Aga is not a second-tier name. She won the 2026 Tokyo Marathon, one of the six World Marathon Majors, and finished runner-up at three other Majors; the AIU statement does not specify which races, so I log that in the pending-verification column. For a distance runner, that record places her inside the elite bracket of the 2026–2026 window, the bracket that receives direct invitations and appearance fees.

I watched the 2026 Tokyo Marathon from the 35km mark, and what I remember is not the finishing time but the cadence over the second half: uncomfortably even, not a single flicker in the quadriceps. People tend to read runs like that with the word “durable”. I read them with a different question: who is paying for that durability, and how. Six years later, the question has a paper answer.

The triggering result sits somewhere else entirely: a half marathon in China, in March, which Aga won. The AIU statement discloses no time for that race. Without a number, it cannot be placed against her personal best, the course profile or the conditions. Analytically that is a blind spot, and not a trivial one. A result published without a time is a result that cannot be assessed as sport; it can only be assessed as a contract.

The substance was methylprednisolone. The AIU found two violations: presence of a prohibited substance in-competition, and tampering with evidence during results management. Aga's explanation was a knee injury. The explanation arrived with a prescription. The prescription is what decided the case.

Before anything else, read the number correctly. Six years is not a single sanction. It is the sum of two, served consecutively rather than concurrently: two years for the substance, four for interference in results management. The World Anti-Doping Code requires independent violations to run back to back, with no offset.

The consequence of that arithmetic: the forgery was punished at twice the rate of the doping itself. Most summaries lose this. Headlines will read “marathon runner banned six years for doping”. The actual structure is two years for the substance, four for the paperwork.

For a 32-year-old, the arithmetic has a concrete meaning: eligibility returns at roughly 38. She loses an entire Olympic cycle. In women's marathoning the peak window commonly extends to 35, occasionally beyond. Six years does not.

One classification error has spread across the coverage and needs correcting: methylprednisolone is not an anabolic steroid. It is a synthetic glucocorticoid, a corticosteroid, class S9 on the WADA list, prohibited in-competition only, and classified as a Specified Substance.

Those three features change the entire reading. Class S9 means the substance can be used outside competition without violation. Specified Substance status means the enforcement body itself recognises this as a class more plausibly linked to inadvertent or therapeutic use, carrying a lower baseline sanction. The TUE mechanism, a therapeutic use exemption, exists precisely for cases like this.

Put plainly, the substance half of the case was the soft half. An athlete with knee inflammation, given a corticosteroid injection, without a TUE, tested in-competition: that is a story that can end at two years, or less with proof of source and degree of fault. That path was closed not by the substance but by the fake prescription.

This is where I build the probability model, the way I do with financial records. The test is not “is she guilty”. The test is: if the prescription were genuine, what is the probability that all three verification layers fail in the same direction?

Layer one, the clinic named on the document. The AIU contacted it and was told the facility does not exist. Layer two, the practitioner's registration number. Absent from the register. Layer three, the claim that the pad came from a former employer. That employer checked its personnel records and confirmed she never worked there.

Three independent layers. For the genuine-prescription hypothesis to hold, all three must be wrong in the same direction. In every document model I have built, that probability approaches zero. Which is why I do not argue with the word “possibly”. I count layers.

These layers are not circumstantial. They are direct third-party confirmations, independent of one another, each verifiable with a phone call or a register search. The evidentiary chain here is tight enough that no reasonable alternative hypothesis survives.

And this is where the file becomes sad rather than infuriating. The defence described by the AIU carries the signature of amateurism, not of an organised doping programme: a relative, an old pad, an invented clinic name. No intermediary layer sophisticated enough to erase a trace.

Across twelve years covering this industry, I have noticed a repeating pattern: the forgeries that fail fastest are the ones assembled by people with no compliance expertise. A competent sports lawyer would never advise submitting a document from a clinic that does not exist, because the AIU verifies directly with the facility, checks registration numbers and pulls employment records. That process has been standard for years.

The clumsiness of the cover-up is itself evidence of a thin medical and legal support structure. In the East African road-running system, many elite athletes race year-round on invitation, arranged by managers, with no dedicated compliance counsel. When a sample comes back positive, they improvise. They improvise inside a system where every document is checkable.

Now the money trail, which I always ask about first. Where did the money in this file come from?

An elite marathoner has three revenue streams: appearance fees at invited races, performance prize money, and brand contracts. The Chinese half marathon sits in the first bucket. The Chinese road-racing market spent years inviting East African athletes for appearance fees plus placing bonuses. The March result is almost certainly annulled, taking the prize money with it. That is an inference from race structure, not evidence contained in the AIU statement, and I keep the two separate.

The larger loss sits in the six years that follow. World Marathon Majors invitations vanish. Brand contracts, with standard morality clauses, have grounds to terminate. For a 32-year-old, that is the remainder of an earnings curve, cut at its highest point.

One technical detail worth logging: when a result is annulled, the forfeiture does not stop at prize money. The placings of everyone behind shift, and payments cascade downward. One athlete's case reaches into other people's wallets, at races with no connection to the doping dispute.

I often ask: where did this money come from and what did it do on the way? Here the short answer: it came from organisers' appearance-fee budgets, passed through the athlete and the manager, and stopped at a decision, the submission of a forged document, whose real cost was four years of a career.

At the level of the event, the impact is close to zero. Ethiopia's bench is deep enough that losing one top-tier slot creates no gap on an international start list. That is why I do not read this case as a signal about competitive balance. It is a signal about governance.

What is lost is the commercial credibility of a road-running brand. What is also lost is organisers' implicit assumption that an athlete invited on a large fee arrives with adequate medical paperwork. That assumption was just falsified at a half marathon.

One risk nobody has raised: with a tampering finding on record, retrospective review of older samples, biological passport data and long-term storage, becomes entirely plausible. The AIU statement does not mention it. But once a file loses credibility at the paperwork stage, every prior result sits in the review zone. I log that as an open risk, not an accusation.

Now the counterintuitive part, the part I consider most important.

The popular reading: a doping case caught, plus a lie. That reading is correct on the verdict but misses a paradox of incentive.

Had Aga confessed immediately, “I had a corticosteroid injection for my knee, I did not know I needed a TUE”, the sanction sits at two years, possibly less, and she returns at 34, still inside the peak window of women's marathoning. Instead, one forged prescription turned two years into six. The cover-up manufactured a sentence the underlying conduct did not deserve.

The second paradox: the anti-doping system, in this file, worked exactly as designed. Three verification layers, three answers, one clear conclusion. That is good news about process. It also exposes a resource asymmetry: the AIU has an investigations unit, the authority to call a clinic, the ability to search a professional register. An athlete under investigation usually has nobody doing that work on her side. The side with resources wins, lawfully.

The third paradox, the one that bothers me most: for the same medical error, the athlete with good counsel and the athlete without receive two different sentences. That is not a problem of this file. It is a problem of the system.

To be clear, so this is not read as exoneration: none of the verification layers above is ambiguous. I am not disputing the verdict. I am describing the gap between the penalty and the actual harm of the conduct.

Safety, in my line of work, does not lie in avoiding arrest, but in never leaving a trace. Aga and her relative left three traces, in three different places, all searchable within the same week. People told me I exaggerate when I say the forgery stage is the weakest link in any system. I told them to wait a few more years.

The strangest thing in this file is not the error, but the way people tried to explain it. A relative signing the treating-physician box, in a system where every signature is searchable. All I did was connect the dots, and count how many were deliberately drawn wrong.

If there is a lesson for athlete education programmes, it is not “do not take banned substances”. It is: learn how to file a TUE before you need one. And for federations: the cost of hiring a compliance adviser is far cheaper than the cost of a forged prescription.