Taurine and Biological Age: Part Three: The Reframing

Abstract

Part One reported an 80 per cent lifetime decline in taurine. Part Two reported no association with age at all. The usual way to settle such a disagreement is to decide which team was careless.

A paper published in npj Aging in January 2026 proposes a third option: that both measured correctly, and the question was framed on the wrong axis. Chronological age, they argue, averages across people who are ageing at very different rates.

Measuring 146 adults aged 20 to 97 and stratifying the older ones by frailty rather than by birthday, they found a relationship that a straight line would erase entirely.

1.The Objection to Age

Two people of 78 can differ enormously in physiological reserve. One walks four kilometres; the other cannot rise from a chair unaided. Treating them as a single data point at x equals 78 is standard practice and it discards most of what distinguishes them.

The frailty phenotype is the established alternative. Assessed by the Fried criteria, it sorts older adults into robust, prefrail and frail on the basis of measured characteristics rather than years elapsed.

The proposal in this paper is that frailty status captures heterogeneity that chronological age cannot, and may therefore reveal associations that age-based analysis obscures.

2.The Cohort

One hundred and forty-six community-dwelling adults in Baltimore, aged 20 to 97. Forty-five younger adults aged 20 to 50, with a mean age of 26. One hundred and one older adults aged 69 and over, phenotyped for frailty: 41 robust with a mean age of 76, 31 prefrail at 78, and 29 frail at 80.

Taurine and upstream metabolites of its biosynthesis pathway were measured from serum by metabolomics, sex-adjusted and normalised.

The first result replicates Part Two rather than Part One. Comparing younger with older adults, the difference in taurine was not statistically significant, at a p-value of 0.29.

The authors address the contradiction directly rather than ignoring it. Their older adults, aged 69 to 97, represent a later life stage than the population in the original study, and they note that high inter-individual variability in taurine can obscure age-related patterns in smaller or more heterogeneous samples.

3.The Shape That Emerged

Stratified by frailty, the pattern was not monotonic. Taurine was highest in robust individuals, lowest in the prefrail, and intermediate in the frail.

A curve of that shape, regressed linearly against age, returns approximately nothing. The decline from robust to prefrail is cancelled by the partial recovery from prefrail to frail, and the average across all older adults sits close to the average of the young. Two groups measuring carefully would report exactly what Parts One and Two reported, depending on where their sampling fell.

The pathway analysis gives the shape a possible mechanism. Robust individuals showed efficient flux through taurine biosynthesis. The prefrail showed the greatest metabolic disruption, described as bottlenecks in the pathway. The frail showed persistent disruption with partial restoration of taurine, which the authors interpret as compensation.

Inflammation tracked the same structure. Tumour necrosis factor alpha correlated negatively with taurine specifically in the prefrail group — not in the robust, not in the frail.

4.What This Would Mean If True

It would mean taurine is neither a driver of ageing nor irrelevant to it, but a marker of a transition — and one that stops being informative once the transition has occurred.

It would also mean that the population in which supplementation could plausibly matter is narrow and specific: people who are prefrail. Not healthy older adults, in whom taurine is already high, and not the frail, in whom it has partly returned by some other route.

That is a considerably smaller claim than the one on the label of any taurine product, and it is testable. It also cuts against the trial we describe in Part Four, which enrolled healthy people aged 55 to 75 and excluded chronic illness.

5.Why We Are Not Treating This as the Answer

This is a cross-sectional study of 146 people, and the frailty groups contain 41, 31 and 29 participants respectively. A three-point non-monotonic curve fitted across groups of that size is a hypothesis, not a finding.

The sex distribution is also badly unbalanced in a way that maps onto the result. The frail group was 86 per cent female; the prefrail group was 65 per cent male. Taurine concentrations were sex-adjusted, which is the correct handling, but adjustment is not the same as balance when the imbalance aligns with the grouping variable.

And a reconciling explanation is the most seductive kind. It resolves a contradiction, it flatters everyone involved, and it is exactly the sort of story that gets accepted before it has been tested. The honest position is that it is the best available account of why two competent groups disagreed, and that it now needs its own replication.

Editorial Comment

MindHeaven® makes no claim about taurine and ageing, and none is authorised in the European Union.

We are including this part because a series that ran only from hypothesis to refutation would be a satisfying story and an incomplete one. Scientific disagreements are more often resolved by discovering that the question had a hidden assumption than by discovering that someone was wrong.

That said, notice what has happened to the claim across three parts. It began as a driver of ageing in five species. It is now, at best, a marker of one transition in one subgroup of older adults, unreplicated. Part Four covers the trial that tested the original version of the claim, and finished while all of this was going on.