Every protocol we build, published in full.

The presenting case

Anonymized archetype

A 41-year-old endurance runner with three years of progressive, exercise-resistant fatigue. Normal sleep architecture on a previous in-lab study. Normal mood inventory. No notable change in diet. Training volume is high but steady — roughly 55–70 km per week, two interval sessions, one long run. Previous clinicians described the standard panel as "normal."

On the optimal-range read, the picture is different. Ferritin is low. Free T3 sits at the bottom of the lab reference range despite a TSH and free T4 that are technically within normal. Vitamin D is insufficient. The fasting glucose is unremarkable; the CGM trace, however, shows an unusual post-prandial pattern after the longest training day.

The clinical reasoning

Why we are choosing this protocol over its closest alternatives

The simplest story that fits these labs is iron-limited erythropoiesis (the body has enough red cells today, but not enough iron stores to make new ones tomorrow), occurring alongside a non-thyroidal low-T3 pattern — sometimes called "euthyroid sick syndrome" — in the setting of sustained training load.

The competing hypotheses we considered and demoted: primary hypothyroidism (TSH is not elevated; thyroid antibodies were negative), depression (mood inventory and history don't support it), and overtraining as a single cause (volume is high but stable, and recovery markers are otherwise reasonable). The CGM pattern is interesting but, for now, we treat it as a marker to re-check, not a target.

Confidence: moderate on the iron-limited piece (well-established mechanism, well-fitting markers), moderate on the low-T3 piece (mechanism well-described, individual variability is real), low on the CGM pattern (interesting but not yet actionable).

The intervention, sequenced

Order matters. We change one thing at a time, on a schedule, so we can tell what's actually working.

Weeks 1–4 — Foundation

• Sleep window: consistent 22:30–06:30 anchor, ±30 min. No alcohol within four hours of sleep.
• Training volume: reduce by ~25% for four weeks. Keep one quality session per week; remove the second.
• Nutrition: add ~400 kcal/day, weighted toward carbohydrate around training. Daily protein target 1.6 g/kg.
• Daylight: 15–20 minutes outdoor light within an hour of waking on at least five days per week.

Weeks 1–12 — Supplementation

• Ferrous bisglycinate, 25 mg elemental iron, every other morning, taken with 250 mg vitamin C and away from coffee, tea, calcium, or grain-based meals by at least one hour.
• Vitamin D3, 4,000 IU daily, with the main fat-containing meal. Co-administered with 100 mcg vitamin K2 (MK-7).
• Magnesium glycinate, 300 mg, 60–90 minutes before sleep.

No proprietary blends. Each agent is chosen for a documented role and a documented marker we can re-check.

Prescriptions, this round

• None. The thyroid picture is consistent with non-thyroidal illness in the setting of training load and energy availability; we do not prescribe thyroid hormone in this scenario until and unless the picture persists after the upstream causes have been addressed for at least 12 weeks.

What we are explicitly not doing

• Not adding adaptogens, peptides, or "thyroid support" blends.
• Not chasing the CGM pattern with a separate intervention this round. We will re-check it at month four; if it persists, it becomes the next protocol.
• Not increasing training intensity until ferritin and free T3 have moved.

Expected response & timeframe

What we are predicting we will see, and when

• Subjective energy: meaningful improvement by week 6, more decisive by week 10. Mornings improve before training.
• Ferritin: rising at month 1 (target trend: +10–20 ng/mL), into mid-optimal range by month 4.
• Free T3: moving into mid-range by month 4 as energy availability and training stress normalize.
• 25-OH vitamin D: at or above 40 ng/mL by month 4.
• CGM re-check: month 4. Re-evaluate whether the post-prandial pattern persists.

What would tell us this isn't working

The honest pre-commitment

Markers we'll re-check

The follow-up panels are pre-committed before this protocol begins.