Case Study 1 — A Literature That Answers Everything Except Your Question

The situation

Someone hands you a claim: growth-hormone-releasing peptides are among the best-studied compounds in existence, because animal agriculture has been researching growth-axis interventions for decades and the research is excellent.

The second half of that sentence is true. The livestock growth literature is real. It is decades deep, it is funded by an industry with a genuine and quantifiable stake in the answers, and it is methodologically careful in ways that would embarrass a good deal of human nutrition research. Sample sizes run into the thousands. Randomization is standard. Diets and housing are controlled to a degree no human study could achieve. Statistical analysis is competent and conventional.

So this is not a case study about bad science. That would be easy and it would teach you nothing. It is a case study about excellent science pointed somewhere else, which is a far more common and far more difficult problem.

What the literature is for

Start with the economics, because they explain everything about how these studies are designed.

An animal production operation converts feed into salable mass. The efficiency of that conversion is the central variable in the enterprise. A small proportional improvement in the amount of feed required per unit of gain, applied across a large number of animals over many production cycles, compounds into a meaningful figure. So does a shift in carcass composition that moves product into a higher grade.

Those are the questions the research exists to answer. It answers them well. And because it answers them well, it has produced exactly the endpoints those questions require:

THE MEASUREMENT SET

  average daily gain ............. how fast does mass accumulate
  feed conversion ratio .......... how much feed per unit of that mass
  dry matter intake .............. how much does the animal eat
  dressing percentage ............ what fraction survives to the carcass
  backfat thickness .............. fat depth at a standard site
  loin eye / ribeye area ......... cross-sectional muscle area
  marbling score ................. intramuscular fat, a grading criterion
  nitrogen retention ............. a proxy for protein deposition
  circulating GH and IGF-1 ....... did the axis respond as designed
  residue depletion .............. how long until the tissue is safe to eat

Every entry on that list is a real measurement of a real thing. Several are measured with more precision than anything in a typical human trial.

What the literature is not for

Now set that measurement set beside the questions an actual person asks about these compounds.

Read down the "Gap" column. The gaps are not gradations of weak evidence. Four of them are total, one is categorical — meaning the measurement could not exist in that species regardless of how well the study were run — and one is inverted, meaning the study was designed to establish the opposite of the thing being claimed.

That last row is worth dwelling on. The safety arm of production-animal research asks: how long must we wait until essentially none of this compound remains in the edible tissue? Its entire purpose is to establish that a human being will encounter almost nothing. Citing it as evidence that a human can safely encounter a great deal is not a weak inference. It is a reversal of the finding.

Three things the literature genuinely does establish

It would be dishonest to run this case study without saying what the evidence is good for, because it is good for several things and a reader who concludes "livestock research is useless" has learned the wrong lesson.

It establishes that the growth axis is manipulable. These studies demonstrate, at high precision and across species, that intervening on the growth-hormone axis changes growth-related outcomes. That is a real biological finding and it is not in dispute.

It establishes dose-response relationships in the studied species. Within that species, at those exposures, on those endpoints, the relationships are characterized well.

It establishes residue and depletion kinetics. These are genuine pharmacokinetic data in a real mammal, and they are more rigorous than anything available for most gray-market compounds.

None of the three is the claim being made. All three are worth knowing. Holding both facts at once is the skill this case study is trying to build.

The rhetorical move to watch

The argument you will meet in the wild rarely states the inference out loud. It usually runs:

"These aren't some untested research chemical. There are decades of peer-reviewed work on this class of compound — thousands of animals, controlled studies, published data. It's one of the most studied things out there."

Every clause is true. The conclusion the listener draws — so we know it is safe and effective for what I want — is supplied entirely by the listener. Nothing in the speaker's sentence asserted it, which is what makes the move so durable in argument: there is nothing to contradict.

The counter is not to dispute the literature. It is to ask what it measured. Once the endpoints are on the table, the argument resolves itself and nobody has to call anybody a liar.

Discussion questions

1. The case study says four of the six gaps in the table are "total" rather than "weak." Defend that characterization against someone who argues that all evidence exists on a continuum and that calling something "no evidence" is an overstatement.

2. Identify which single row of the gap table is most likely to persuade a skeptical reader, and explain why. Then identify which row a motivated reader is most likely to talk their way around.

3. The "three things the literature genuinely does establish" section could be omitted without changing the case study's conclusion. Argue for keeping it, in terms of what the chapter is trying to teach rather than what it is trying to prove.

4. Construct the same table for a different body of evidence — cell-culture studies, or a cosmetic ingredient's supplier data, or a nutrition epidemiology dataset. Fill in at least four rows. Which of your gaps are total, which categorical, which inverted?

5. The rhetorical move described here consists entirely of true statements. Write a two-sentence reply to it that does not dispute any of the statements and still defeats the implied conclusion.

6. Suppose someone funded a livestock-style study — same rigor, same sample size, same statistical care — designed to answer a human performance question. Sketch what would have to change in the design: population, endpoints, duration, blinding, and outcome ascertainment. Then estimate which of those changes would be hardest, and say why that difficulty is itself informative about the state of the evidence.