A clinic that did not exist. A prescription pad the clinic itself disowned. And a “doctor” with no medical registration, who described himself as a family member and explained the paperwork by saying he had used “an old prescription pad from a former workplace.” The three lines in the Athletics Integrity Unit (AIU) statement that carry this entire case come not from a finish line but from a document check.

Ruti Aga, the Ethiopian marathoner, has six years. In headlines this is a doping case. In the paperwork it is two separate violations stacked: two years for the presence of a prohibited substance, four years for tampering with the evidence manufactured to explain it.

In the summer of 2026 I hand-coded 380 football matches off a free event feed because the university terminal was booked until October. The habit that came out of it never left: the most trustworthy cell in any report is the empty one. A filled cell tells you someone wrote something. An empty cell tells you someone left something out. This story lives entirely in the second kind.

The career ledger, and its blank cells

Aga is 32. The source gives no personal best — that is the first blank cell and I will leave it blank. My own record has 2:18:34 for second in Berlin 2026 and 2:20:40 for the win in Tokyo 2026, but that is external recall pending verification and I will not smuggle it into the AIU statement. What the source states plainly: the 2026 Tokyo Marathon title, and three second-place finishes at World Marathon Majors.

I coded Tokyo 2026 off the world feed. By then I had a habit of keeping split timings in a separate sheet, so that a year later I would remember not just who won but the shape of the win. Aga ran the second half faster than the first that day — the thing people later call form and the thing I log as a negative split. This piece is not about that race. But it matters that the athlete under discussion was not a fringe runner.

The violation attaches to a half marathon in China this March, which she won. The race is unnamed in the source. No mark. No splits. No field depth. What exists is that a post-race sample returned methylprednisolone, and that she explained it with a knee injury supposedly treated by that same “doctor.”

She had been provisionally suspended from 10 September; the statement came on a Thursday. The year is not stated — the second blank cell. An empty stadium has a data signature, and this statement should be read the same way: what is missing (the race name, the mark, the scope of disqualification, a single line from the national federation) speaks as loudly as what is written.

The classification error that bent the story

Start with the thing that has done the most damage to public understanding. A major report called methylprednisolone a “prohibited steroid” and said WADA classes it as performance enhancing. That is imprecise to the point of misleading. Methylprednisolone is a glucocorticoid — WADA Prohibited List class S9 — not an anabolic steroid (S1). And the decisive condition on S9 is that it is prohibited in-competition only, by injection, oral or rectal routes.

That one sentence rebuilds the case. The question is not whether she took a steroid. The question is by which route, and how many days before the race. A therapeutic knee injection is entirely legal out of competition; it becomes a violation only if residues are present in an in-competition sample. A road race's testing window opens at 23:59 the night before, which means an injection given days earlier can still be caught. In marathon terms this is the release angle of a javelin throw: the variable that decides the outcome is timing, not the mark.

Honesty is required here. The route-specific washout figures that once sat in the guidance are figures I will not quote, because I could not re-derive them against the current List. What is settled: since 2026, injections of glucocorticoids are prohibited in-competition, full stop. That is enough, and an invented number is always worse than an honest blank.

What two plus four actually says

Two years plus four years equals six. The arithmetic reads as a finding: the panel accepted the presence violation but refused the no-significant-fault mitigation that would have taken the presence sanction below two years. Four is the Code's standard tampering sanction. So one decision ran two different scales — lenient on the substance, absolute on the paper.

Here is the quiet paradox. A forged prescription is itself circumstantial evidence of concealment intent, yet it did not aggravate the presence sanction; it simply added four years from a separate direction. Technically clean. Practically terminal.

Backdating to 10 September is not a reduction. Provisional suspension time counts against the ban; the athlete gains nothing in real terms. It is procedural accounting, not mercy — and any report that frames it as relief has misread the ledger.

Three verifications, three failures

The forgery was dismantled in three separate places. One: the clinic named on the prescription did not exist. Two: the person who issued it was a family member with no medical registration. Three: the defence claimed an “old prescription pad from a former workplace,” and the clinic disowned the pad outright.

Three independent checks, three failures. This is not the signature of a filing error. It is the signature of an athlete operating without a credentialed team physician. A genuine knee injection, properly documented, is defended with records — a discharge summary, an injection note, a pharmacy bill. None of it was real. And that is the great irony: the knee-injury explanation was mechanistically plausible. Intra-articular glucocorticoid for knee synovitis is routine sports medicine in distance runners. That plausibility died the moment it was defended with paper instead of a doctor.

One more cell I will not fill: the TUE. An athlete with a chronic knee problem either holds or applies for a therapeutic use exemption, or observes the washout interval. Neither appears to have happened — I do not know that directly; I am inferring it from the fact that a forgery was attempted instead. Inference, labelled as inference.

The tampering charge rests on an interview, not a document — statements made under questioning. That detail matters, because language and translation conditions in AIU interviews with Ethiopian athletes are a recurring point of friction. Not an excuse. A fact about the room in which this decision was written.

Thirty-two to thirty-eight

Six years at 32 means a return at roughly 38. Against the depth of Ethiopian women's marathon running, that is a six-year ban on paper and a career ending in practice. The national pool holds a sub-2:16 cohort and world-record class at the top; no selector has a rational reason to pick a 38-year-old with a doping record, least of all in an ecosystem whose real contract is a federation quota rather than a sponsorship. One athlete's removal is invisible at national-team level.

Her record is now bifurcated. Tokyo 2026 and three Major second places are the record of a genuinely elite marathoner; the same record is commercially radioactive in a sport where distance-running sponsorship is already thin. The source does not state the disqualification scope. Standard AIU practice voids the result and prize money from the competition where the sample was collected; earlier results do not automatically fall unless prior use is established. So the March half marathon, its title and its cheque — my strongest reading, not a stated fact.

Where I part with the headlines

A six-year sanction has landed on a substance with no proven endurance benefit. A glucocorticoid reduces pain and swelling; it does not build muscle or carry oxygen. Over 42 kilometres its only legitimate advantage is that an injured runner can stand on the start line — an integrity question in itself, but not a steroid question.

Two readings exist and I want both on the table. On one hand this is AIU rigour: four years for tampering, no softening. On the other it is a consistency problem: six years for a drug whose ergogenic value is unproven. When the pattern repeats across East African road-racing cases, the question shifts from guilt to proportionality. I do not have an answer to that. I have the arithmetic and a note that correlation is not causation.

The real finding is not the drug. It is that nowhere in this file is there a registered team doctor. That is bigger than an individual lapse. Injured marathoners carry pressure to protect appearance-fee income; glucocorticoids are the cheapest, most available tool; the surrounding medical network is informal. The forged prescription sits downstream of that governance gap, not upstream of it.

And then the Chinese label-race circuit. For East African marathoners it is the income tier between Majors — few races, good fees, low risk. That same tier is now the AIU's highest-yield testing ground. The appearance-fee economy and the testing economy are colliding in the same race. This is not one athlete's story. It is the story of a tier no broadcaster puts a camera on.

What to watch next

First, the CAS window — typically around 21 days from the reasoned decision. A document-based forgery finding is hard to overturn, so the odds are low, but the window is open. Second, whether an agent or management is ever named, or the statement stops at one family member. Third, whether the Ethiopian federation says a single line; the source carries the AIU and nothing national.

Fourth, and most consequential: how many similarly profiled cases are already in the pipeline from the same training clusters. I did not trust a single table after 380 matches until I could rebuild it myself. Here I have six years, two violations and a clinic that never existed. One centimetre is a full argument — and this entire document was invented.

Method note: every fact attributed to the AIU statement is the AIU's — the six-year ban, the two-plus-four split, the provisional suspension from 10 September, the Thursday statement, the three-stage forgery check. Everything external — the Berlin 2026 and Tokyo 2026 marks, washout figures, disqualification scope — is my recall or standard practice, pending verification. Sample size: one case. This is a case, not a trend.