Summarized & reviewed by The Peptide Dispatch Editorial Team · Last reviewed September 14, 2026
Every comprehensive metabolic panel you have ever had measured alkaline phosphatase. It sits a few lines below albumin, it carries a reference range of roughly 40 to 130 units per liter depending on the laboratory and the assay, and unless the number is frankly high it receives no attention at all. That habit has two costs. The first is that the association between alkaline phosphatase and…
This dispatch covers Your Alkaline Phosphatase Is Normal. That Range Was Built to Catch Liver Disease, Not Calcifying Arteries. in the research research category, authored by The Peptide Dispatch Editorial Team. Estimated reading time: 10 minutes. The Peptide Dispatch curates peer-reviewed peptide research for self-directed learners. All summaries are presented for Research Use Only and do not constitute medical advice.
Every comprehensive metabolic panel you have ever had measured alkaline phosphatase. It sits a few lines below albumin, it carries a reference range of roughly 40 to 130 units per liter depending on the laboratory and the assay, and unless the number is frankly high it receives no attention at all.
That habit has two costs.
The first is that the association between alkaline phosphatase and cardiovascular events does not begin above the reference range. It runs straight through the middle of it. The second is that the bottom of the range, where a value is technically flagged but almost never investigated, is one of the few routine blood findings that should change what gets prescribed to you.
Neither of those is exotic. Both come from large cohorts and from mechanistic work that has been sitting in the literature for over a decade. What is missing is the interpretive habit.
Alkaline phosphatase is a family of enzymes, not one. The version that dominates the number on your panel is tissue-nonspecific alkaline phosphatase, produced mostly by bone and liver, with smaller contributions from intestine and kidney. It is a zinc-dependent enzyme. Structural work published in 2024 describes two zinc ions bound at its M1 and M2 catalytic sites, and finds that at sites of arterial calcification, calcium progressively displaces that zinc as the enzyme concentrates in the calcifying matrix (Gomez and Millán, Journal of Molecular Histology, DOI 10.1007/s10735-024-10207-3).
The job it does matters more than where it comes from. One of the enzyme's core substrates is inorganic pyrophosphate. Pyrophosphate is the body's endogenous brake on calcium-phosphate deposition in soft tissue. It is generated outside the cell from ATP, and tissue-nonspecific alkaline phosphatase is the enzyme that hydrolyzes it into ordinary phosphate, which removes the brake. A review of pyrophosphate biology in kidney disease lays out the balance plainly: several enzymes govern the pyrophosphate to phosphate ratio, and when that balance shifts, calcification follows (Azpiazu and colleagues, Nefrologia, DOI 10.1016/j.nefro.2017.07.005).
So the number filed under "liver panel" is also a readout on an enzyme whose day job is dismantling the molecule that keeps calcium out of your arteries.
This is the part that should change how the number is read.
A prospective cohort of 26,389 adults followed for an average of 7.3 years recorded 7,015 incident cardiovascular events. Comparing the highest quartile of alkaline phosphatase to the lowest, and adjusting for age, body mass index, smoking, drinking, diabetes, hyperlipidemia, hypertension, physical activity, aspirin and anticoagulant use, family history and kidney function, the hazard ratio in men was 1.22 for cardiovascular disease overall, 1.43 for stroke and 1.75 for hemorrhagic stroke. In women the cardiovascular hazard ratio was 1.12. The authors then did the thing that makes this study useful: they restricted the analysis to people whose alkaline phosphatase sat inside the normal range of 40 to 150 units per liter, and the association held, in a dose-dependent way (Liu and colleagues, Journal of Atherosclerosis and Thrombosis, DOI 10.5551/jat.63646).
A separate analysis of 34,147 US adults from a national survey, followed for a mean of roughly 11.6 years, recorded 5,413 deaths from any cause and 1,820 cardiovascular deaths. The highest alkaline phosphatase group carried a 1.30 hazard ratio for all-cause mortality and 1.39 for cardiovascular mortality against the lowest (Yan and colleagues, Frontiers in Endocrinology, DOI 10.3389/fendo.2023.1217369).
Two effects worth naming. Neither is enormous. A hazard ratio of 1.2 to 1.4 is a nudge, not a verdict, and it is the kind of effect size that is easy to confound. But it is measured on a number you already own, on a panel you already paid for, that currently generates no information at all unless it is abnormal.
The second study did something the first did not, and it cuts against a simple reading. The authors ran a mediation analysis and found that gamma-glutamyl transferase accounted for roughly 13 to 16 percent of the effect, vitamin D for 7 to 8 percent, and C-reactive protein for 8 to 10 percent. In other words, a meaningful slice of what looks like an alkaline phosphatase signal is really a liver stress signal, an inflammation signal, and a vitamin D signal wearing a different label.
This is observational data. It cannot establish that alkaline phosphatase causes anything. The causal evidence is animal work, and it is genuinely suggestive rather than conclusive. In a mouse model of a heritable calcification disorder, halving the enzyme genetically cut mineralization in the target tissue by 52 percent, and an oral inhibitor of the enzyme produced a 58 percent reduction. The same inhibitor failed entirely in a different calcification model, which tells you the pathway is specific rather than universal (Li and colleagues, Journal of Investigative Dermatology, DOI 10.1016/j.jid.2018.07.030).
Treat the enzyme as a marker with a plausible mechanism behind it. Do not treat it as a lever.
For anyone tracking metabolic health, there is a link that makes the marker more interesting than a generic risk score.
Work published in 2024 examined what elevated glucose does to pyrophosphate handling across three models: rat aortic smooth muscle cells, chemically induced diabetic rats, and aortic smooth muscle cells from humans with diabetes. High glucose reduced extracellular pyrophosphate in all three. It downregulated the enzyme that produces pyrophosphate from ATP, upregulated a competing enzyme that routes ATP to plain phosphate instead, and raised tissue-nonspecific alkaline phosphatase activity markedly across every model. The net result was a lower pyrophosphate to phosphate ratio and a measurably greater tendency to calcify (Flores-Roco and colleagues, Cardiovascular Diabetology, DOI 10.1186/s12933-024-02502-w).
That is a coherent story rather than a coincidence. Chronic glucose elevation pushes the enzyme up and the brake down at the same time. It is also a reminder that a "normal" alkaline phosphatase drifting upward year over year on serial panels is information that a single snapshot destroys.
There is a kidney thread too. In 2,157 never-treated hypertensive adults whose alkaline phosphatase was entirely within the normal range, the enzyme correlated with estimated glomerular filtration rate at r equals minus 0.43, and emerged as the second strongest independent predictor of kidney function in the model (Sciacqua and colleagues, Scientific Reports, DOI 10.1038/s41598-020-66911-z).
Now the other direction, which is the more actionable half.
A German laboratory reviewed 6,918,126 adult measurements collected between 2011 and 2016. Eight and a half percent came in below 30 units per liter. Where a follow-up pyridoxal-5-phosphate level had been run, about 6 percent of those low values were elevated, which is the biochemical fingerprint of genuinely deficient enzyme activity. The authors put the combined figure at roughly 0.52 percent of tested adults, about one in 194, showing laboratory signs of hypophosphatasia, a heritable condition of low alkaline phosphatase activity affecting bone and dental mineralization (Schmidt and colleagues, Orphanet Journal of Rare Diseases, DOI 10.1186/s13023-021-02084-w).
Read that estimate with care. The follow-up test was ordered in well under 1 percent of the low-alkaline-phosphatase group, and it was almost certainly ordered on people who had already raised a suspicion, so the true population rate is likely lower than one in 194. The authors' own recommendation is the useful part: when alkaline phosphatase comes back low, run the confirmatory test automatically and tell the ordering physician, because at present nothing happens.
In a smaller referral series of 42 adults with unexplained low alkaline phosphatase, half carried a mutation in the gene encoding the enzyme. Most were heterozygous missense variants. Symptoms were mild, mostly musculoskeletal aches, but tooth loss was reported in 48 percent of carriers against 12 percent of non-carriers (Riancho-Zarrabeitia and colleagues, European Journal of Internal Medicine, DOI 10.1016/j.ejim.2015.12.019). A referral population is not a population rate, and 50 percent is not the number to quote at a dinner party. The pattern is what matters.
The clinical consequence is specific and it is why this deserves a paragraph rather than a footnote. Drugs in the bisphosphonate class work by suppressing bone turnover, and they are chemically related to pyrophosphate. Giving them to someone whose enzyme activity is already deficient is the wrong direction. A published case describes a 63-year-old woman who sustained an atypical proximal tibial fracture after four years of intermittent bisphosphonate exposure, at which point a look back through her chart showed her alkaline phosphatase had been low the entire time (Malabu and colleagues, Journal of the Endocrine Society, DOI 10.1210/js.2019-00265). The recommendation that came out of it is simple: check for the condition in anyone with persistently low alkaline phosphatase before starting that class of drug.
On zinc, a note against the internet consensus. The enzyme is zinc-dependent, so severe zinc deficiency does lower it. That does not make a low-normal alkaline phosphatase a zinc test. It is neither sensitive nor specific for marginal zinc status in adults, and reading it that way invites self-supplementation for a problem that has not been demonstrated. Low alkaline phosphatase has a long ordinary differential: recent bisphosphonate exposure, malnutrition, celiac disease, hypothyroidism, severe anemia, magnesium or zinc deficiency, some medications. A single low value is usually one of those. A persistently low value across several draws is the one worth chasing.
Before reading anything into a shift, rule out the mundane. Alkaline phosphatase is legitimately higher in adolescents and during any period of active bone growth or fracture healing. It rises in pregnancy because the placenta makes its own isoenzyme. It runs higher on average in older adults and differs by sex. Intestinal alkaline phosphatase can rise after a fatty meal in some people depending on blood group and secretor status, which is one reason fasting matters for comparability. And assays differ between laboratories, so a jump that coincides with a change of lab is a change of lab until proven otherwise.
Three habits, none of which require a new test.
Pull the number from every panel you can find and put the values in date order. A single reading tells you almost nothing. A value that has climbed from the low 50s to the low 90s across four years, while still labeled normal every time, is the pattern the cohort data describes, and it is invisible in any one report.
Read it alongside gamma-glutamyl transferase, C-reactive protein, vitamin D and a marker of glucose control rather than alone. The mediation analysis says a real share of the signal travels with those, so an isolated interpretation is likely to be wrong in one direction or the other.
If it is persistently at or below roughly 30 units per liter across more than one draw, that is the finding that warrants a conversation with your clinician, particularly before any decision about bone-density medication, and particularly if there is a history of adult fractures or unexplained tooth loss.
The broader point is the same one this series keeps arriving at. The reference range answers the question the test was designed for, which here is whether your bile ducts are obstructed or your bones are dissolving. It was not built to answer whether your arteries are quietly calcifying, and it does not stop being useful for that second question just because nobody is asking it.
Research for this article was retrieved from PubMed. All source studies are linked by DOI above.
Educational content only. This is not medical advice, and nothing here is a recommendation to start, stop or change any medication or supplement. Laboratory values require interpretation in the context of your full clinical picture by a qualified clinician who knows your history.
Educational content — not medical advice. Effects described are drawn from cited research in study subjects.