Case Study 1 — Devi: The Test Nobody Ran
A clinical case. Devi Raghunathan is an illustrative composite; the physiology, the diagnostic gap, and the repletion protocol are not.
Setup
Devi Raghunathan, 21. Division I 5000 metre runner. Vegetarian since twelve, for ethical reasons. 168 cm, 55 kg, 44 kg fat-free mass. Sixty miles a week.
Presenting history, over two years:
| Stress fractures | Three — two metatarsal, one tibial |
| Amenorrhea | Nine months |
| Performance | Declining across two seasons despite increased training |
| The specific complaint | "At about eighteen minutes there's nothing there, and then I have to run the rest of it on nothing" |
| Blood tests | Three, all reported as normal |
What had been measured: full blood count, basic metabolic panel, thyroid function.
What had not been measured: ferritin.
The two-year gap
I want to walk through what those three "normal" results actually were, because the failure is instructive and it isn't anyone's incompetence.
Test one (following the first stress fracture, roughly two years before I saw her): full blood count. Hemoglobin 13.1. Normal. The clinical question was "is she anemic?" — a reasonable question after an injury — and the answer was no.
Test two (a general check the following year): FBC and thyroid. Hemoglobin 12.8. Normal. Thyroid normal.
Test three (after the third fracture, three months before I saw her): FBC, metabolic panel. Hemoglobin 12.6. Normal.
Look at the trend. 13.1 → 12.8 → 12.6. Every single value sits inside the reference range and therefore generated no comment. And the trajectory is unmistakable once you put the three numbers next to each other — which nobody had, because they were ordered by different clinicians in different contexts and each was read in isolation.
She was descending the staircase from §14.2, one step at a time, over two years, with the floor sensor correctly reporting that she hadn't hit the floor.
Her ferritin, when finally measured: 11 ng/mL. Below every threshold in use — WHO's 15, the clinical 30, and the higher cut-offs some sports physicians use.
Why she got there: six factors, stacked
No single one would have done it. Together they were overdetermined.
| Factor | Effect |
|---|---|
| Non-heme iron only (vegetarian) | 2–20% absorption instead of 15–35%; requirements often cited as ~1.8× higher |
| Phytate | Her main iron sources — lentils, wholegrains, tofu — are also her main phytate sources |
| Tea with dinner | Polyphenols substantially inhibit non-heme absorption, and she drank tea with her main iron-containing meal |
| Menstrual losses | Until they stopped — and their stopping was itself a symptom, not a solution |
| Athlete-specific losses | Foot-strike hemolysis, GI microbleeding, sweat — and exercise-induced inflammation raises hepcidin, suppressing absorption for hours after hard training |
| Low energy availability | EA 31.8 kcal/kg FFM. You cannot eat adequate iron inside 2,100 kcal while running sixty miles a week. |
💡 Aha moment. Read the last row again, because it reorders everything.
Her iron problem is partly an energy problem wearing a mineral costume. At an intake of 2,100 kcal with 700 kcal/day of exercise expenditure, every micronutrient is under-supplied — not because she chose badly, but because there isn't enough food passing through to carry them.
Which means that supplementing iron alone would have been treating the most visible deficiency in a system that was under-supplied across the board. It might even have worked, partially, for a while — and it would have left the actual problem in place and delayed the diagnosis further.
This is the general principle: when several micronutrients are marginal at once, suspect the denominator.
What her training partner said
One detail from the consultation that I think about often.
I asked Devi whether anyone on her team had been checked for iron. She said her training partner had — because her mother was a nurse and had insisted. That teammate had been found deficient two years earlier, treated, and had gone on to a good season.
The difference between the two athletes was not their physiology, their diet, or their training. It was that one of them had a parent who knew which test to ask for.
That is not a clinical failure so much as a structural one, and it is why Exercise C1 exists in this chapter and why I've written the exact sentence out for readers to use.
The intervention
Ordered deliberately, because sequence matters here.
1. Energy first. ⚠️ Referral to a sports physician and a sports dietitian for RED-S assessment. Target intake raised from ~2,100 to 2,750–2,900 kcal/day, bringing energy availability toward 45 kcal/kg FFM. Bone health assessment given three fractures and nine months of amenorrhea.
Nothing else on this list works properly until this one does.
2. Iron repletion, supervised. Oral iron, dosed and monitored by her physician. Two details that made a material difference:
- Taken with vitamin C, away from tea, coffee, calcium and dairy. She had been drinking tea with dinner — her main iron meal — for years.
- Taken away from hard training sessions, because exercise-induced hepcidin suppresses absorption for hours afterward. (Alternate-day dosing was considered, on the basis that daily dosing itself raises hepcidin and can reduce fractional absorption — an area where practice has been shifting.)
3. Dietary changes, all additive rather than restrictive: - Vitamin C source at every iron-containing meal — peppers, citrus, tomatoes - Tea moved to between meals, which cost nothing - Legumes soaked before cooking, reducing phytate - Fortified cereals retained and eaten away from dairy
4. Protein raised from 1.1 g/kg (61 g) to 1.6 g/kg (88 g), per Chapter 8 — for tissue repair and lean mass, and because the energy increase made it achievable.
5. Monitoring: ferritin and hemoglobin at three months, then six.
What happened
| Baseline | 3 months | 8 months | |
|---|---|---|---|
| Ferritin | 11 | 24 | 48 |
| Hemoglobin | 12.6 | 13.0 | 13.6 |
| Energy intake | ~2,100 | ~2,650 | ~2,800 |
| Energy availability | 31.8 | ~40 | ~45 |
| Menses | Absent 9 months | Absent | Returned at ~7 months |
| Weight | 55 kg | 56.5 kg | 57 kg |
| Training | Reduced during assessment | Rebuilding | Full |
The eighteen-minute problem resolved gradually over the second and third months — before her ferritin was anywhere near normal, which is worth noting. Some of that is iron; a good deal of it is simply having eaten enough.
Her weight went up two kilograms, and this was the hardest part of the entire intervention for her. Chapter 6's Case Study 1 records what she said: "So everything I've been good at is the problem."
Three fractures, and none since.
Analysis
1. The trend was visible and nobody saw it. 13.1 → 12.8 → 12.6, across three tests ordered by three clinicians in three contexts. Each was correctly interpreted in isolation and the sequence was never assembled. This is a systems failure that better individual clinical reasoning wouldn't have fixed.
2. The right test costs almost nothing. Ferritin is inexpensive, widely available, and has clear indications — female, endurance athlete, vegetarian, fatigue. She met four. The gap was not knowledge or cost; it was that nobody thought to add one line to a form.
3. Sequence mattered more than content. Iron alone might have partially worked and would have left RED-S in place. When multiple micronutrients are marginal, suspect the denominator — and fix energy before optimizing anything downstream of it.
4. The free interventions were substantial. Moving tea away from meals, adding vitamin C, soaking legumes, timing supplements away from training. None cost anything and together they meaningfully changed how much iron a given diet delivered. These are almost never mentioned.
5. Symptoms improved before the numbers did. Which is a useful thing to tell a patient in advance, because a person waiting for a lab value to move will often quit before it does.
6. And the uncomfortable comparison. Her training partner got tested two years earlier because her mother was a nurse. ⚠️ Access to the right question is unevenly distributed, and a chapter that just says "get your ferritin checked" is speaking to people who can and will. The sentence in Exercise C1 exists to make that question portable.
Discussion Questions
-
Three clinicians each read a hemoglobin correctly and the trend was missed. What system change would surface a declining-but-normal trajectory? What would it cost in false alarms?
-
Devi met four indications for ferritin testing and it was never ordered. Is this a knowledge gap, a time gap, or a protocol gap? What would you change first?
-
The chapter argues energy had to be fixed before iron. Construct the case for treating them simultaneously. Is there a scenario where iron-first is right?
-
Her weight rose two kilograms and she found this the hardest part. How would you prepare an athlete for that in advance? Write the actual words.
-
The training partner was tested because her mother was a nurse. What does that tell you about how health knowledge is distributed, and what can a book realistically do about it?
Your Turn
This is the exercise with the highest expected value in Part III.
Tick honestly:
- [ ] I menstruate
- [ ] I have heavy periods
- [ ] I do regular endurance training
- [ ] I'm vegetarian or vegan
- [ ] I'm pregnant or planning pregnancy
- [ ] I donate blood regularly
- [ ] I have unexplained fatigue
- [ ] I drink tea or coffee with meals
Two or more of the first seven → ask for ferritin.
The sentence, verbatim, so you don't have to compose it in the room:
"Could I have my ferritin checked as well as my full blood count? I've read that a normal hemoglobin doesn't rule out iron deficiency."
Add "with a CRP" if you want to be thorough, since inflammation can mask a low result.
And if you ticked the last box, do the free thing regardless of testing: move tea and coffee away from meals, and put a vitamin C source alongside iron-containing food. It costs nothing, it takes no willpower, and it meaningfully changes how much iron your existing diet delivers.
One more thing. If you know a woman who trains seriously — a runner, a cyclist, a rower — ask whether her ferritin has ever been checked. The answer is usually no. Devi's teammate got tested because someone told her mother to ask.