Taurine and aging: what the cohorts show

Taurine is the rare ageing compound whose famous paper contains no numbers at all. The Science 2023 paper that made it famous states its findings in prose and prints no sample size, no concentration and no effect size. That fact should sit in the first paragraph, not in a footnote, because every quantitative statement about taurine and ageing that a reader will meet comes from somewhere else.

What the Science paper actually says

Here is the substance of the abstract, and it is worth reading closely because the verbs do the work that numbers usually do.

Concentrations of circulating taurine decline with aging in mice, monkeys and humans. A reversal of that decline through taurine supplementation increased the health span and life span in mice and the health span in monkeys. Mechanistically, taurine reduced cellular senescence, protected against telomerase deficiency, suppressed mitochondrial dysfunction, decreased DNA damage and attenuated inflammaging. In humans, lower taurine concentrations correlated with several age-related diseases, and taurine concentrations increased after acute endurance exercise.

Then the authors’ own conclusion, which is more careful than the headlines built on it: taurine deficiency may be a driver of aging because its reversal increases health span in worms, rodents and primates and life span in worms and rodents, and clinical trials in humans seem warranted to test whether taurine deficiency might drive aging in humans.

Three things follow. The lifespan and health-span results are animals. The human statement is a correlation with age-related disease. And the authors themselves end by calling for the human trial that does not yet exist. Everything quantitative below is therefore from other papers, and that is not a criticism of this one.

The strongest human number, and what it is

The largest cohort figure available is from the Framingham Offspring Study, which measured 217 plasma metabolites in 2,067 participants free of dementia, mean age 55.9 years with a standard deviation of 9.7, of whom 52.4 % were women. Ninety-three developed incident dementia over a mean follow-up of 15.6 years with a standard deviation of 5.2.

Taurine came out at a hazard ratio of 0.74 with a 95 % confidence interval of 0.60 to 0.92. The authors describe it as a suggestive association and write that the taurine and hypoxanthine associations strengthen evidence that uric acid and taurine may be neuroprotective.

Read that interval carefully. It is 2,067 people followed for fifteen and a half years, which is exactly the kind of dataset a cohort purist wants, and it is an association computed in Cox models alongside 216 other metabolites. It is not a trial, taurine was not assigned, and the confidence interval excludes 1.0 without telling you whether the ninth percentile of the cohort is where the signal lives or the eightieth.

One detail in the source is worth flagging rather than smoothing: the abstract prints an identical confidence interval, 0.60 to 0.92, for taurine and for hypoxanthine. This page reports taurine’s figure as printed and does not attempt to resolve the duplication.

Where taurine shows up in the ageing metabolome, and where it does not

A second human cohort adds a detail that a broad summary usually flattens. In 92 cognitively unimpaired Taiwanese adults aged 20 to 87, with no obesity and no diabetes, fasting cerebrospinal fluid was analysed by mass spectrometry. Thirty-seven metabolites correlated positively with ageing and two negatively, including cysteine, pantothenic acid, 5-hydroxyindoleacetic acid, aspartic acid and glutamate.

Taurine is not on the ageing-correlation list. It appears in the sex-difference finding: levels were higher in women by propensity-matched comparison, alongside 5-HIAA.

So the honest statement is that this cohort identified taurine as differing between sexes in cerebrospinal fluid, and did not list it among the metabolites that track age. The page does not upgrade that into a finding about decline, because the abstract does not say it.

The one human trial, and why it cannot answer the question

There is a randomised controlled trial of a taurine-containing product in older adults, and its design is the reason this section exists.

One hundred and seven participants with mild cognitive impairment, mean age 62.94 years, were randomised 1:1 to 15 g a day of a whey protein powder containing milk fat globule membrane, taurine and B vitamins, or to placebo, for twelve months. On the Montreal Cognitive Assessment the mean change at twelve months was 3.23 (95 % CI 2.17 to 4.30) on the active arm against 1.42 (0.36 to 2.48) on placebo, giving a group-by-time difference of 1.81 (0.32 to 3.30). The Digit Symbol Substitution Test difference was 2.72 (0.20 to 5.23). Side effects did not differ significantly, P = 0.74.

That is a real trial with a real result. It is a trial of a three-ingredient powder. There is no taurine-alone arm, no taurine-alone placebo, and no way to attribute the 1.81-point MoCA difference to any single component. The authors’ conclusion names the whole powder, and this page names the whole powder too. A trial of taurine alone in older adults is precisely what the Science paper says is warranted, and it is not what this trial was.

The largest human trial with taurine in its name that is not about taurine

One more study needs naming because it is easy to cite wrongly. The CENTAUR trial screened 177 people with amyotrophic lateral sclerosis and randomised 137 of them, 89 to sodium phenylbutyrate plus taurursodiol and 48 to placebo. The active regimen was 3 g of sodium phenylbutyrate with 1 g of taurursodiol once a day for three weeks, then twice a day.

The primary outcome was the rate of decline in the ALSFRS-R score over 24 weeks: -1.24 points per month on the active drug against -1.66 on placebo, a difference of 0.42 points per month with a 95 % confidence interval of 0.03 to 0.81 and P = 0.03. The authors conclude that functional decline was slower.

Two things follow for this page. The compound is taurursodiol, a taurine conjugate, tested in combination with a separate drug in a specific neurodegenerative disease; it is not a taurine trial and it is not an ageing trial. And the secondary outcomes did not differ significantly between the groups, including isometric muscle strength, neurofilament H and slow vital capacity. A primary result of 0.42 points a month with a confidence interval touching 0.03 is a real signal that the authors themselves say needs a longer and larger trial.

The genetic-instrument approach

One recent analysis sidesteps the trial question entirely. A mendelian randomisation study used taurine-level data from the Ieu Open GWAS Project and immunocyte traits from a large European cohort, analysed with MR-Egger, weighted median and inverse variance weighting. Sensitivity analysis revealed no significant heterogeneity or pleiotropy, which is the standard check against the two commonest ways these analyses mislead.

The reported result is a causal relationship between taurine levels and 10 immunocyte phenotypes, three negative and seven positive, framed as a possible immunological route to skin anti-ageing effects.

What the abstract does not contain matters as much as what it does. There is no effect size, no sample size, and no skin outcome was measured. Mendelian randomisation estimates whether a lifelong genetically higher level is associated with a trait; it does not tell you what happens when someone starts taking taurine tomorrow, and it cannot.

Where this sits on the shelf

There is no taurine product in this catalogue and this page carries no product link, for the reason in the note at the top. The honest neighbours in the ageing section of this site are the NAD and NMN page and the Epithalon page. Both are more explicit about their human evidence than most pages on that shelf, and both are worth reading alongside this one to see how little a compound’s reputation tells you about its trials.

What the data do not show

They do not show a taurine concentration falling over time inside the same people, because no longitudinal human measurement is printed here. They do not show that taking taurine changes any clinical outcome in a human being, because no such trial exists; the one RCT tested a three- ingredient powder. They do not show an effect attributable to taurine in that RCT, because there is no taurine-alone arm. They do not show that the Framingham association is causal, because it is a cohort analysed in Cox models. And they do not show a lifespan or health-span effect in humans, because that result exists only in mice and monkeys.

What we supply

Nothing on this page is a product of ours and this page recommends nothing. It exists so a reader chasing an anti-ageing claim about taurine can see which numbers came from the famous paper and which came from somewhere else.

Research use only. This page summarises published studies for research reference. It is not medical advice, not a protocol, and not a suggestion for human use. Nothing we supply is for human or veterinary use.