Why this part exists
For seven parts I have walked through what this paper found, and most of it is genuinely interesting. Now I want to do the other half of the job. A study is only as useful as its weakest link, and this one has a few that anyone excited about SS-31 should know about.
None of this means the paper is worthless. It means it is a first look, not a final answer. The authors themselves put "possibly" in the title. That word is doing real work.
Problem one: how many rats?
The methods section says a total of 192 rats were randomly divided into the groups.
The figure legends say six animals per group. The statistics for the behaviour and swelling tests are reported in a form that fits about 36 animals across six groups. The later tests fit about 24 animals across four groups.
That leaves a large gap between 192 rats used and the roughly 60 that appear in the reported results. There are reasonable explanations. Rats may have died after surgery, some may have been used for pilot work, and different rats were needed for different tissue tests. But the paper does not say. It does not report how many died, how many were excluded, or why.
That matters. If a lot of rats died after the bleed, and the survivors were the toughest ones, the results could look better than reality. Without the numbers, a reader cannot check.
Problem two: killed at 24 hours, scored at 72
The methods say all the rats in each group were killed at 24 hours after the bleed. The results then report behaviour scores at both 24 and 72 hours.
Both cannot be true for the same animals. Most likely, a separate set of rats was kept alive for the 72 hour scoring. That would be normal. But it is not stated, and in a paper this short, a sentence like that is exactly what readers need.
Problem three: small groups
Six animals per group is common in this kind of early work, but it is small. With groups that size, a single unusual rat can swing an average. It also makes it harder to trust the subtler findings, like the 10 mg/kg dose not beating the 5 mg/kg dose in Part 4.
The paper reports averages with standard error. Standard error makes spread look smaller than standard deviation does, which is fine as long as readers know which they are looking at.
Problem four: one sex, one timing, one dose
Only male rats were used. Female animals can respond differently to brain injury and to drugs, partly because of hormonal effects on blood vessels and inflammation. We simply do not know if the result holds in females.
The drug was given once, 30 minutes after the bleed. As I noted in Part 3, that is far faster than most people with a real bleed would ever be treated. It tells us SS-31 can work when given almost immediately. It tells us nothing about two, six or twelve hours later.
And after the dose test, everything was done at one dose, 5 mg/kg, at one time point, 24 hours.
Problem five: the authors' own gaps
To their credit, the authors list two limitations themselves. They did not test SS-31 on brain cells in a dish to confirm the effect directly. And they did not look at mitochondrial shape or the electrical charge across the mitochondrial membrane, which would have confirmed that the mitochondria themselves were being protected.
That second point matters more than it sounds. The whole story of this paper is "SS-31 protects mitochondria". Yet the study never looks at the mitochondria directly. It measures signals that come from them and infers the rest.
Problem six: a different lab, a different answer
In Part 6 I asked you to remember the Bax result. Here is why.
A 2021 study from the University of Wisconsin, which cites this very paper, set out to test the claim that SS-31 stops Bax from moving onto mitochondria. Using live imaging of individual cells, they found SS-31 slowed Bax slightly but did not delay when it started moving, did not reduce how much moved in the end, and did not stop cytochrome c from escaping.
Those were cultured cells, not rat brains, so the two studies are not directly comparable. But it does mean the mechanism this paper proposes is contested, not settled. We cover that study in its own series.
Problem seven: the version I read
The copy of this paper I worked from is the "journal pre-proof", the accepted manuscript before final copyediting. Some of the small inconsistencies above, such as a stray duplicated reference number in the methods, may have been tidied up in the final version. The core design questions would not have been.
So what do you do with this paper?
You take it for what it is. A small, well-intentioned first study, from a single lab, showing that SS-31 given almost immediately after a brain bleed in male rats reduced swelling, barrier leak, neuron death and oxidative damage at 24 hours.
That is a real finding and a good reason for more research. It is not proof that SS-31 treats brain bleeds, and it is certainly not evidence for anyone using it after a head injury.
The honest summary is the one the authors chose themselves: possibly. Anyone who drops that word when they quote this paper is selling something.
This series discusses a peer-reviewed laboratory study in rats. The results have not been shown in people. Nothing here is medical advice, and SS-31 is discussed for research purposes only.
