Phosphate, FGF23, and Heart Disease: Does the Science Hold Up?

The Phosphate-FGF23 Hypothesis: A Compelling Story That Outran Its Evidence

A claim circulating in biohacking circles goes like this: inorganic phosphate additives — the kind found in processed meats, soft drinks, and packaged baked goods — silently trigger a hormonal cascade that drives up blood pressure, stiffens heart muscle, and accelerates cardiovascular aging. The villain is a hormone called FGF23. The mechanism is biologically plausible. The human evidence, as of now, does not support the conclusion being drawn from it.

We reviewed three studies cited in support of this claim. Here's what each one actually shows — and why the package falls apart when you read past the abstracts.

Study 1: The Review That Built the Case — and Flagged Its Own Limitations

The anchor of the phosphate-heart narrative is a review article published in International Journal of Molecular Sciences: High-Phosphate-Induced Hypertension: The Pathogenic Role of Fibroblast Growth Factor 23 (FGF23) Signaling in Sympathetic Nervous System Activation.

Review papers synthesize existing studies — they don't generate new data. The proposed mechanism is this: dietary phosphate elevates FGF23 → FGF23 activates the sympathetic nervous system → sympathetic activation raises blood pressure → chronic hypertension promotes cardiac fibrosis and eventual heart failure. Each link in that chain has some support in animal models, particularly rodent studies and in vitro work. But the authors themselves acknowledge that human epidemiological evidence is inconsistent, and the full mechanistic pathway has not been confirmed in healthy humans under controlled conditions.

When a review paper's own summary includes "proposed mechanism unconfirmed in humans," you're reading a scientific hypothesis, not a clinical finding. Rodent kidneys respond differently to phosphate loads than human kidneys do. Dietary context, microbiome composition, kidney filtration rates, and baseline phosphate metabolism all differ substantially between species. What happens to a mouse fed high-phosphate chow is not a reliable proxy for what happens to a person eating a processed diet.

The review is useful for identifying an interesting area of investigation. It is not evidence that dietary phosphate restriction improves cardiovascular outcomes in healthy adults.

Study 2: An Unrelated Cohort Study, Pressed Into Service

The second paper — Clinical Factors and Biomarkers During Pregnancy and Risk of Cardiovascular Disease, published in JAMA Cardiology — has nothing to do with dietary phosphate.

It's a cohort study examining whether complications and biomarkers measured during pregnancy predict a woman's long-term cardiovascular disease risk. Conditions like gestational hypertension, preeclampsia, and gestational diabetes, as well as inflammatory markers elevated during pregnancy, appear to forecast cardiovascular risk decades later. This is genuinely important research: pregnancy is emerging as a cardiovascular stress test that reveals underlying susceptibilities.

But there is no phosphate in this story. There is no FGF23. There is no processed food. Including this paper as evidence for a phosphate-heart narrative requires a conceptual leap the study was never designed to support. The connection is thematic — both topics involve cardiovascular risk — and that is where it ends. This is the kind of citation that looks persuasive in a listicle and falls apart under a five-minute read of the actual paper.

Study 3: A Drug Trial About Surrogate Endpoints

The third paper is methodologically the strongest of the three: Spironolactone and Fibrosis in Heart Failure Risk: Machine Learning Analysis of HOMAGE Trial Plasma Proteomics, published in MedComm.

This is a secondary analysis of the HOMAGE trial, an RCT examining spironolactone — a prescription aldosterone antagonist — in people at elevated risk for heart failure. Using machine learning on plasma proteomics, researchers found that spironolactone reduced levels of PICP, a biomarker associated with collagen synthesis in cardiac tissue. The implication: spironolactone may reduce cardiac fibrosis progression.

Three reasons this doesn't support the phosphate narrative:

• Surrogate endpoints are not outcomes. PICP is a biomarker — a signal that collagen production is occurring. The trial didn't measure whether patients lived longer, had fewer heart attacks, or experienced less heart failure. Biomarker reductions are hypotheses about future outcomes, not proof that outcomes improve.
• Spironolactone is a drug, not a dietary intervention. Its mechanism — blocking aldosterone receptors — has nothing to do with reducing dietary phosphate. You cannot use a drug RCT to argue for a dietary recommendation unless the trial actually tested that dietary change.
• This is a prescription medication requiring physician oversight. Spironolactone raises serum potassium, affects blood pressure, and carries real clinical risks. Positioning this study as supporting evidence for a general biohacking protocol misrepresents the scope of what the evidence addresses.

What Would Actually Change This Assessment

The phosphate-FGF23 pathway is a legitimate research area — primarily in patients with chronic kidney disease, where phosphate metabolism is already disrupted and FGF23 elevations are pronounced and clinically significant. In that population, the evidence is more developed. In healthy adults making dietary choices, the translational case is weak.

Evidence that would be convincing:

• A randomized controlled trial in healthy adults showing that reducing inorganic phosphate intake measurably lowers FGF23, then blood pressure, with hard cardiovascular endpoints tracked over a meaningful follow-up period.
• Large prospective cohort data carefully separating inorganic phosphate (from additives) from organic phosphate (from whole foods like meat, legumes, and dairy) and showing the additive form specifically predicts cardiovascular events after controlling for total calories, sodium, and overall diet quality.

Neither of these exists in the current evidence package.

The Practical Problem With the Whole Argument

Even if the FGF23 mechanism eventually proves out in human trials, the actionable recommendation it generates is "eat less processed food." This advice predates the entire FGF23 research program by decades. If a proposed cutting-edge molecular mechanism leads to the same conclusion as a 1985 nutrition brochure, you haven't generated a biohacking insight. You've generated a more complicated path to advice your grandmother already gave you.

That doesn't mean the research isn't worth tracking. The phosphate-FGF23 pathway could eventually yield genuinely useful guidance, particularly for individuals with early kidney dysfunction or markedly elevated cardiovascular risk profiles. But "worth watching" is not the same as "proven" — and the current evidence does not support treating this as an established causal mechanism in healthy people.

Verdict

Compelling villain. Missing crime scene. An animal-data-heavy review with acknowledged inconsistencies, an unrelated pregnancy cohort study, and a drug trial about surrogate endpoints do not collectively demonstrate that dietary phosphate damages the hearts of healthy people. Until human intervention data exists, this is a hypothesis in need of a trial — not a finding in need of a protocol.