Every letter we have printed

Prespecification matters here more than usual. A composition endpoint added after the parent results were known is a different kind of evidence from one written into the protocol, and the papers do not always make clear which they are reporting.

The proportion of loss that is lean tissue is broadly consistent with what is seen in dietary weight loss of similar magnitude, which is the comparison readers need and rarely get. The question is not whether lean mass falls but whether it falls more than it would otherwise.

The soft-tissue artefact point in your bone section is underplayed. In a patient losing twenty per cent of body mass the change in overlying tissue is well outside the range the calibration was validated over, and the published analyses do not report a sensitivity analysis for it. That is not a caveat, it is a gap.

Strength can be maintained while cross-sectional area falls, because a good deal of early strength change is neural. Readers using performance as a proxy for tissue are using a proxy that moves for other reasons, in both directions.

Some analytes have intervals that vary with body composition, which makes them awkward to interpret in exactly the population most likely to be measuring them. That interaction is real and is rarely mentioned in general discussion.

Testing many analytes at once guarantees flagged results by arithmetic alone. With twenty independent tests, at least one out-of-interval value is more likely than not in a person with nothing wrong.

The gap I would highlight is bioburden. Between sterile and unexamined there is a measurable middle, and a total viable count on a powder is neither expensive nor slow. It is the determination that would tell a buyer the most about a supplier’s environment, and nobody offers it.

A publication that keeps naming what is missing is doing something that cannot be done by a certificate, however good. The list of absent tests is editorial work rather than analytical work.

Turnaround is the practical obstacle. A compendial sterility test takes a fortnight, which for material with a short commercial life is a genuine cost rather than an excuse. Rapid methods exist and validating one is not trivial.

You keep saying you are criticising documentation and not honesty. I think that distinction is doing more work than it can bear. If a company knows buyers read a purity certificate as a safety document and issues one anyway, the omission is doing something.

Adherence typically falls over the course of a long trial, and a falling adherence curve will produce a flattening outcome curve without any change in response. The two are rarely plotted together.

Attrition produces apparent plateaus mechanically. If people whose trajectory flattens leave the study at a different rate from those whose does not, the remaining group mean moves for reasons that have nothing to do with anybody’s response.

Haemoconcentration is the artefact I would put first. A modest change in plasma volume shifts every concentration on the panel in the same direction at once, and the pattern is unmistakable if you look at the whole report rather than at the flagged line.

Creatinine is a good example of a value that moves for reasons unrelated to the kidney. Muscle mass, recent exercise and dietary intake all shift it, and during rapid weight change at least two of those are moving at once.

Assay methods for several of these differ between laboratories more than for common chemistry, and the between-method variation can exceed the width of the interval. Comparing across providers is particularly unreliable here.

Ferritin is an acute phase reactant, so a value taken during inflammation says something about the inflammation. Pairing it with a marker that is not acute phase is the standard remedy and it is routinely omitted.

A short thank you for the reference desk, which I use more often than the articles. Somebody has clearly spent a great deal of time on material that will never be anybody’s headline.

Consider the incentive on the service. A badge is a recurring revenue line and an expiry rule reduces it. That is not corruption; it is ordinary commercial gravity, and reforms that run against gravity need somebody outside the arrangement to push them.

Nobody expires anything. A badge granted three years ago is still on the page, still linking to a report about a lot that ran out long ago, and there is no mechanism anywhere in the chain that removes it. Decay is the missing feature.

The price table is the most useful thing you have published this year and also the thing most likely to be quoted out of context by somebody selling a comparison service. You might consider a note.

The submitter-pays problem is real and I think you overstate it slightly. A laboratory that shades results towards its customers loses the only asset it has the moment a second laboratory disagrees in public. The incentive to please the payer is immediate; the incentive to be right is larger and slower.

Speaking as a former quality manager, the safeguard that matters is whether the analyst knows whose sample it is before the run. Blinding at the bench costs nothing and removes the whole class of problem your article describes. It is a question worth putting to each of the four.

Multi-receptor compounds are described as though adding a target necessarily adds effect. The published ratios differ enormously between molecules, and a compound with a weak second activity is closer to a single-target agent than the label suggests.

The tolerability profile of a multi-receptor compound is not the sum of the single-target profiles either, and it is discussed as though it were. Both directions of that assumption are unsupported.

You describe biased agonism as "legitimate and probably important" and then decline to say which molecules are biased in which direction. That is a strange place to stop.

Half-life is quoted as a single number for compounds whose elimination is not first-order across the whole range, and the number then propagates into schedule arithmetic that assumes it is. The assumption is stated in the source and lost thereafter.

The hindbrain and hypothalamic sites are where the appetite effect lives, and the peripheral vagal contribution is more contested than a general reader would guess from most summaries. Your account distinguishes the two, which matters because they predict different things about tolerance.

Your piece treats the source as a fixed component. It is a consumable with a maintenance history, and a fouled source changes response for some compounds far more than others. Ask when the source was last cleaned, and you will occasionally learn more than the spectrum tells you.

Two ionisation techniques on the same sample is the cheapest confidence a laboratory can buy. Electrospray and matrix-assisted desorption fail in different directions, and a species that survives both is far more likely to be real than one that appears only in the technique the laboratory happens to own.

A caution on the word high-resolution as it appears in marketing. It is not a threshold; it is a continuum, and instruments separated by a factor of twenty in resolving power are sold under the same adjective. Ask for the figure and the mass at which it was obtained, and the adjective becomes unnecessary.

Analytical imprecision is published for every assay and almost never quoted alongside a result. A change smaller than the assay’s own variation is not a change, and a great deal of anxiety in this area concerns movements inside the noise.

Some analytes have intervals that vary with body composition, which makes them awkward to interpret in exactly the population most likely to be measuring them. That interaction is real and is rarely mentioned in general discussion.

My ferritin fell from 118 to 34 across a year on treatment and I was told this was expected because inflammation had fallen. Six months later I was clearly iron deficient. I appreciate the section on this being ambiguous, but the ambiguity was resolved in one direction and nobody looked.

Where intake falls substantially, protein adequacy is a more useful thing to think about than any single micronutrient, and it is not measured by any of the panels discussed. Albumin is a poor proxy and is used as one constantly.

Thank you for this. It answered the question I came with and raised two I had not thought to ask.

What I would fund, if I could fund anything, is a rolling blind purchase of the same molecule from the same source every quarter for three years. Longitudinal data on one line would settle more arguments than a hundred one-off determinations across the market.

A methodological plea. Publish the purchase date, the arrival date and the determination date for every blind result. Without all three, a poor figure cannot be attributed between manufacture, freight and storage, and every party will attribute it wherever suits them.

I run analytical services and I object to the framing of your blind comparison. You bought our cheapest tier, published the number it produced alongside a competitor’s most thorough package, and called the result a spread. It is not a spread. It is three different products, priced accordingly, and your own table says so two columns to the right of the headline figure.

Treating the four services as institutions rather than as brands is the right frame and it exposes how thin the institutional layer is. Four organisations, no common reporting standard, no shared reference material and no mechanism by which one can be compared with another except by a buyer paying twice.

Hydration is the dominant short-term error term in most of these methods and the least controlled. A standardised fasting and fluid protocol before scanning would tighten these datasets more than any statistical technique.