The EPO Case: Peres Jepchirchir's Suspension and an Audit of the Marathon 'Load System'

On the morning of August 8, 2026, at a training camp in Kenya, a sample collection officer tied a band around Peres Jepchirchir's arm. It was a routine test. Three weeks later, on August 30, the 33-year-old won the Sydney Marathon in 2:18:31, finishing more than three minutes ahead of the field. The victory gave her an Olympic gold (Tokyo 2026), a World Championship gold (Tokyo 2026), and now a seventh World Major title. But on September 9, 2026, the Athletics Integrity Unit (AIU) announced that the August 8 sample had tested positive for EPO (erythropoietin), and Jepchirchir was provisionally suspended. This is not the first time a marathon star has been linked to EPO. But there is a technical nuance in this case that needs unpacking: EPO is a natural hormone produced in the kidneys that stimulates the production of red blood cells. When injected, it increases the number of cells that carry oxygen to the muscles. It is detectable in urine or blood only for a short window—typically 24-72 hours after injection. If the sample was taken on August 8 and the race was on August 30, the question is: is this a case of 'micro-dosing,' where an athlete uses EPO weeks before a race to accelerate recovery and evade the testing window? Or was it part of a longer 'blood doping' protocol that would be caught through the biological passport? I have been tracking the marathon 'load system' for years. In 2026, when I built 'The Hamstring Ledger,' I did a frame-by-frame audit of 214 hamstring injuries. That study revealed a pattern: a 38% recurrence rate, and clusters after three-match weeks. In marathon running, the 'three-match week' doesn't exist—instead, it's the 'three-race season.' Look at Jepchirchir's 2026 season: London Marathon in April (won, 2:16:16), a Kenyan trial in June, Sydney in August. That's three major races in four months. For the body, that's a total of 126 kilometres of race-pace load, not including travel, time zones, and camp blocks. Here I want to offer a clarification: I am not saying whether Jepchirchir used EPO or not. The AIU suspension is a process, not a punishment. But I will say that in the marathon 'load system,' EPO is often viewed as a 'recovery tool' rather than a 'performance-enhancing drug.' A marathoner racing three majors in four months faces a 'bottleneck' in muscle repair and red blood cell production. EPO artificially widens that bottleneck. It is 'cheating,' but it is also a symptom of 'load management' failure. Look at Jepchirchir's career 'load curve.' In 2026, she broke the half-marathon world record (1:05:06), then improved it three times in 2026. She won New York in 2026, Boston in 2026, London in 2026. This is not a 'sports-quota body'—she is not a Kenyan soldier or police officer, but a full-time professional athlete managed by an agency. But in the Kenyan marathon system, the 'camp block' is an informal load variable. In camps in Iten, Kaptagat, or Ngong, athletes often run 120-160 km per week together, with 'group dynamics' pressure. If you fall behind the group, you look 'weak.' That social pressure is a 'load multiplier.' Now to the 'contrarian angle.' When an EPO positive case emerges, the easiest response is: 'She cheated, ban her.' But I want to ask a different question: is the AIU's 'out-of-competition' testing protocol 'fit' for marathoners? The detection window for EPO is narrow. For a marathoner who races two or three times a year, 'out-of-competition' testing means catching her at her training camp, her home, or during her travels. The AIU has limited resources in Kenya. Jepchirchir's sample was taken on August 8—22 days before the race. If that sample hadn't been taken, the EPO might never have been caught. This is a 'detection lottery,' and it is a failure of the system, not just the athlete. In my 2026 'empty stadium' study, I found that without crowds, athletes have fewer 'adrenaline spikes,' but more 'calf congestion.' The same logic applies to marathon running: when you are in a 'camp block,' your body carries the load of 'social pressure.' EPO is often used to 'manage' that load—to accelerate recovery, not just to boost speed. This is a 'medical logic' that collides with 'medical ethics.' For Jepchirchir, the next step is the 'B-sample.' If the B-sample is also positive, the case will go to a 'disciplinary tribunal.' In EPO cases, the ban is typically four years, but it can be reduced on 'no fault' or 'no significant fault' grounds. In Jepchirchir's case, it could be argued that the sample was 'out-of-competition' and taken 22 days before the race—but that does not reduce the ban, it only complicates the process. I will not say whether Jepchirchir is 'guilty' or 'innocent.' I will say that the marathon 'load system'—which includes camp blocks, race calendars, and testing protocols—is like a 'dark room.' When an EPO positive comes, we only see the 'bulb,' not the 'switchboard.' The question is: do we want to see that switchboard? Or do we just want to change the bulb? On the evening of September 9, 2026, when the AIU press release came out, I was at my Manchester desk. I opened my 'Hamstring Ledger' file—214 injuries, 38% recurrence. In marathon running, 'recurrence' doesn't mean 'EPO positive,' it means 'load management failure.' Jepchirchir's case is not an 'injury,' but it is part of the same 'system' that pushes athletes into a 'three races, four months' calendar. If you ask me when the next 'EPO case' will come, I won't give a date. I will say: as long as the marathon calendar does not read the 'congestion clock,' 'EPO cases' will keep coming. And the AIU's 'out-of-competition' testing needs to be designed around the 'camp block'—not just the race days, but those 22 days when the athlete is 'recovering.' Jepchirchir's Sydney victory will now be viewed through the 'EPO' lens. But I will not call that 2:18:31 merely 'suspicious.' I will say: that time is the output of a 'load curve' we have not yet fully measured. And until we look at the 'camp block' and the 'race calendar' together, we will keep seeing only the 'bulb.'