Oxalates bind to iron in the gut. Spinach is loaded with oxalates. The conclusion writes itself: spinach iron is locked away, absorbed so poorly it barely counts.
When this was measured directly — the exact amount of oxalic acid found in a serving of spinach, added to a controlled meal, iron absorption tracked with stable isotopes — the effect was zero.
Do Oxalates in Spinach Really Block Iron?
Oxalic acid has no measurable effect on iron absorption in humans. When the exact amount found in a serving of spinach was added to a meal and tracked with stable isotopes, absorption was identical with and without it. Spinach iron is poorly absorbed, but the actual inhibitors are polyphenols and calcium — not oxalates.
— Bonsmann et al. 2008 · European Journal of Clinical Nutrition · n=13
Not a subtle zero. Not a statistical maybe. Iron absorption was virtually identical whether the oxalic acid was present or absent — so close the measurement could not tell the two conditions apart. The molecule every nutrition source has blamed for decades did nothing measurable to iron inside a human body.
Spinach iron IS poorly absorbed. Every source got the outcome right. What collapsed was the explanation.
Spinach and kale, side by side, told the rest of the story. Spinach absorbed less iron — and the data showed why: spinach brought far more polyphenols and calcium to the table. Those two compounds, present alongside the oxalates the entire time, explained the gap no one had attributed to them.
What every source named: Oxalates — binding iron, making spinach iron unavailable
What the measurement found: Polyphenols and calcium — the actual inhibitors of spinach iron absorption
The real culprits have a known weakness. Vitamin C reverses the way polyphenols block iron — roughly 50 milligrams per meal is enough to flip the effect. A single bell pepper clears that threshold twice over. A spinach and bell pepper pasta puts the fix on the same plate as the problem.
Thirteen women. One isotope study. A result so unambiguous it was never replicated — not because the science is thin, but because a P-value of 0.86 leaves nothing left to find. A null this clean from a controlled crossover means the effect is not small and hard to detect. It is absent.
The oxalate claim survived on repetition, not replication. The molecule sounded guilty enough — oxalic acid, iron binding, gut chemistry — and nobody checked the lineup with actual measurements. The question that stays open is how many other nutrition facts have survived the same way: repeated so widely they feel settled, tested so rarely they have never been caught.
What eggs do to a vegetable salad landed in the same territory — a pairing everyone assumed worked one way, with hidden mechanics nobody measured until someone did. The plate is more complicated than any single molecule.
