# The comparison register

> The other group | Really Research Review — Really Research Review’s comparison register: placebo responses, semaglutide against tirzepatide, mouse mechanisms and observational MOTS-c findings.

**06**

Some experiments add a treatment. Others remove a mechanism or measure a signal already present.

## Placebo responses

The [GHK-containing scalp mixture](https://pmc.ncbi.nlm.nih.gov/articles/PMC4969472/) improved hair counts more than placebo; counts increased in the placebo group too. The [STEP 1 semaglutide trial](https://pubmed.ncbi.nlm.nih.gov/33567185/) also recorded an average change in its placebo group, here a loss of weight.

Neither observation establishes that an inactive ingredient caused the change. A comparison group can capture other influences on a result. The controlled difference persuades me. Simply crediting treatment with every later change would claim too much.

The scalp trial has a further attribution problem because the active treatment combines GHK with another ingredient. Placebo helps establish that the mixture had an effect; separating the active ingredients would require more groups.

## A competing medicine

In [SURMOUNT-5](https://pubmed.ncbi.nlm.nih.gov/40353578/), semaglutide met tirzepatide within the same weight-loss trial. Tirzepatide produced the greater mean loss at 72 weeks. That is a particularly useful comparison when the question concerns the two medicines’ weight effects.

It leaves the heart and kidney questions with their own evidence. [SELECT](https://pubmed.ncbi.nlm.nih.gov/37952131/) and [FLOW](https://pubmed.ncbi.nlm.nih.gov/38785209/) compared semaglutide with placebo and studied different clinical populations. A universal winner across these outcomes would need quite different comparisons.

## Removing a mechanism

The NAD+ experiments offer a different sort of company. [Mice lacking CD38](https://pubmed.ncbi.nlm.nih.gov/27304511/) were protected against an age-related fall in NAD+. In the [mouse heart-failure model](https://pubmed.ncbi.nlm.nih.gov/40211954/), reducing HMGCS2 prevented NAD+ repletion from rescuing cardiac function.

One intervention removed an enzyme that consumes NAD+; the other reduced a proposed link between NAD+ and a heart benefit. These manipulations help explain why something happens inside the experimental model.

I value that capacity to trouble an explanation. The most interesting experiment may be the one that makes the hoped-for effect disappear. Its clinical relevance still requires treatment studies in people.

## The human observation

The [MOTS-c dialysis cohort](https://pubmed.ncbi.nlm.nih.gov/39111290/) compared people’s existing blood concentrations with later outcomes. Its participants received no MOTS-c intervention. Higher circulating MOTS-c was associated with the combined mortality and cardiovascular outcome, and its inclusion improved a risk model.

My question there is what the concentration signifies in illness. A treatment trial would ask what happens when investigators deliberately give MOTS-c. Changing the verb changes the experiment.

## Comparisons still missing

- A component-by-component test of the GHK scalp mixture would help assign the observed hair-count gain.
- Human MOTS-c treatment trials would have to measure benefit and safety directly, beyond mouse performance and human blood associations.
- NAD+ precursor research needs clearer connections between blood increases, tissue changes and useful human outcomes, as the [clinical review](https://pubmed.ncbi.nlm.nih.gov/41083806/) describes.

These gaps let me ask about ingredients, tissue and function, rather than merely attaching the word “early” to everything.

---

An independent field notebook on whether peptide findings keep their shape when the analytic lens shifts—not a clinic, a shop, or medical advice.