Most research peptides have an evidence problem. They have an attractive mechanism, a stack of rodent papers, and almost no useful human data.
ARA-290 has a different problem.
It has randomized human trials. It has objective evidence of small-nerve-fiber growth. Multiple studies converged on the same short subcutaneous regimen, and the adverse-event record is more legible than it is for most compounds in the peptide conversation. But the result that makes ARA-290 scientifically interesting is not a result I can credibly measure at home.
That does not make it a poor choice. It changes what the experiment can tell me. A compound can be biologically plausible, supported by human evidence, and worth trying for a potentially useful effect even when my N=1 cannot verify that the effect occurred.
With ARA-290, I am accepting the possibility that the most important result could be real, beneficial, and completely invisible to me.
What ARA-290 is
ARA-290 is an 11-amino-acid peptide derived from the helix-B surface of erythropoietin, or EPO. It also appears in the literature as cibinetide and pyroglutamate helix B surface peptide, usually shortened to pHBSP.
EPO is best known for stimulating red-blood-cell production. Its biology is broader than that, however. EPO also has tissue-protective, anti-inflammatory, and repair-associated effects in experimental models. Using EPO itself to pursue those effects would bring hematologic activity with it: higher red-cell mass and the risks that follow.
ARA-290 was designed to separate those jobs. It reproduces a small surface region of EPO associated with tissue protection but does not bind the classical EPO receptor in the same way and has not shown an erythropoietic effect in the human studies.
The design thesis is elegant:
Preserve the repair signal. Leave the red-blood-cell signal behind.
That is the origin of the compound—not a naturally occurring peptide discovered intact in the body, but a deliberately engineered fragment inspired by one face of a native hormone.
What it is proposed to do
The familiar mechanism diagram places ARA-290 at an innate repair receptor, proposed to contain the EPO receptor and the beta-common receptor CD131. Activation is then linked to JAK2 and downstream pathways that reduce inflammatory signaling, limit cellular injury, and support repair.
Animal and laboratory studies suggest that ARA-290 may help calm inflammation by changing how certain immune cells respond to injury. For example, in an experimental model of inflammatory bowel disease, ARA-290 reduced immune activity and turned down NF-kB, one of the body’s major “inflammation switches.” Researchers found that this effect appeared to involve proteins called CD131 and JAK2.
The exact way ARA-290 produces these effects is still debated, however. One popular theory is that it works through a special receptor made from the EPO receptor and CD131 working together. But a later laboratory study was unable to show that these two receptor proteins actually bind directly to each other—even when ARA-290 was present.
So the simple takeaway is that ARA-290 appears to have real anti-inflammatory effects in preclinical experiments, but scientists still do not fully understand exactly which receptor or signaling pathway is responsible for them.
That does not erase the clinical or preclinical effects. It means I should separate two claims:
- ARA-290 has produced biological effects consistent with tissue protection and nerve repair.
- The exact receptor arrangement responsible for those effects remains contested.
For my purposes, the best way to describe ARA-290 is as a small peptide derived from EPO that does not stimulate red blood cell production. Instead, research suggests it may help reduce inflammation and support nerve repair or regeneration. Scientists are still working out exactly which receptors and biological pathways are responsible for those effects.
The human evidence is narrow—but real
The compelling human story is not general recovery, muscle repair, longevity, or athletic performance. It is small-fiber neuropathy.
Small nerve fibers carry pain and temperature information and participate in autonomic function. Damage can produce burning, tingling, altered temperature sensation, abnormal sweating, pain, and other symptoms that are difficult to treat and sometimes difficult to quantify.
The first sarcoidosis trial: symptom signal, no direct regeneration measure
The first randomized pilot enrolled 22 people with sarcoidosis-associated small-fiber neuropathy. Twelve received 2 mg intravenously three times per week for four weeks; ten received placebo. The active group improved more on a small-fiber-neuropathy symptom questionnaire. Pain and fatigue improved in both groups without a clear treatment advantage, and the study reported no safety concerns. This was an encouraging symptom result, but it did not yet show that nerves had regenerated.
NERVARA: the first objective corneal signal
A second randomized trial enrolled 38 people with sarcoidosis-associated small-fiber neuropathy. This time, participants used 4 mg subcutaneously once daily for 28 days, followed by observation. Corneal confocal microscopy showed a median 14.5% increase in corneal nerve-fiber area in the ARA-290 group versus a 5.3% decrease with placebo. The active group also improved more on the small-fiber-neuropathy symptom questionnaire and walked farther on the six-minute walk test. The important counterweight is pain. Brief pain-inventory intensity fell by about the same amount in both groups. Ordinary skin intraepidermal nerve-fiber density also did not significantly change.
This is where the ARA-290 story becomes more interesting than “people said they felt better.” A structure that can be imaged objectively—the corneal small-fiber network—changed in the expected direction.
Diabetic neuropathy, symptom benefit, weaker structural confirmation
A trial in type 2 diabetes enrolled 49 people with painful neuropathy and used the same 4 mg subcutaneous daily regimen for 28 days. PainDetect scores improved more with ARA-290 at both day 28 and day 56. The overall change in corneal nerve-fiber density was not statistically significant. A subgroup with abnormal baseline corneal density showed a within-group increase, but that exploratory result is weaker than a clear randomized between-group effect.
This study supports a neuropathy signal, but it should not be upgraded into definitive proof of nerve regrowth in diabetic neuropathy.
Phase 2b: the strongest regeneration result
The most informative trial randomized 64 people with sarcoidosis-associated small-fiber neuropathy to placebo or 1 mg, 4 mg, or 8 mg of subcutaneous cibinetide daily for 28 days.
The 4 mg group produced the clearest result. Placebo-corrected corneal nerve-fiber area increased by 697 square micrometers per field, with a 95% confidence interval of 159 to 1,236 and a p-value of .012. Regenerating GAP-43-positive fibers in skin also increased significantly at 4 mg.
The dose response was not orderly. Neither 1 mg nor 8 mg reproduced the statistically significant corneal result seen at 4 mg. Pain improved across groups, and the placebo-corrected pain result in the prespecified moderate-to-severe subgroup did not reach statistical significance.
This is the core of the case for ARA-290. A randomized trial detected change in two objective nerve-regeneration measures. It is also the reason I resist describing the peptide as a painkiller. Structural regeneration was the clearer signal; subjective analgesia was not.
Negative studies define the boundary
Evidence becomes more useful when it tells me where the compound did not work.
In diabetic macular edema, nine people entered a 12-week study of 4 mg subcutaneous ARA-290 daily and eight completed it. The study found no meaningful improvement in visual acuity, central retinal thickness, retinal sensitivity, or tear production. In 36 healthy participants, a single 2 mg dose altered some aspects of emotional processing but did not produce a convincing antidepressant or mood-improving profile. Those results matter because they keep the thesis narrow. Human exposure does not show ARA-290 acting as a general-purpose restorative peptide. Its best evidence remains concentrated in people who already had small-fiber pathology.
What the newer animal literature adds
A 2025 rat sciatic-nerve-crush study gives the peripheral-nerve thesis a newer mechanistic layer. The investigators reported that ARA-290 reduced Schwann-cell NLRP3 inflammasome activation, NF-kB phosphorylation, and reactive oxygen species, alongside improved functional recovery after injury.
That is useful support for a nerve-injury model. It is not human evidence, and it does not justify the larger internet claim that ARA-290 should improve any kind of tissue recovery.
The literature also needs active maintenance. A frequently cited 2012 paper reporting protection from renal ischemia-reperfusion injury was retracted in February 2026 after the publisher identified highly similar blot lanes and concluded that selected bands were not representative of the experimental results.
One retracted kidney paper does not invalidate the independent neuropathy trials. It does mean that paper should be removed from any evidence stack used to argue for renal protection.
The human regimens that have actually been studied
ARA-290 discussions often drift from published exposure into invented “cycles.” The human record is much simpler:
- 2 mg intravenously three times per week for four weeks in the first sarcoidosis pilot.
- 4 mg subcutaneously once daily for 28 days in the NERVARA sarcoidosis trial and the diabetic neuropathy trial.
- 1 mg, 4 mg, or 8 mg subcutaneously once daily for 28 days in the dose-ranging Phase 2b sarcoidosis trial.
- 4 mg subcutaneously once daily for 12 weeks in the very small diabetic macular edema study.
- A single 2 mg dose in the healthy-volunteer emotional-processing study.
The pharmacokinetics are short. The diabetes study reported a terminal half-life of roughly 20 minutes after a 4 mg subcutaneous dose. Yet the clinical regimens were once daily, not continuous infusions or repeated doses throughout the day. That suggests the research model depends on triggering a biological program rather than maintaining a steady circulating level.
It does not establish an optimal regimen for a healthy person. No relevant trial has done that.
My regimen: 4 mg daily for 28 days
For my ARA-290 experiment, I will use 4 mg subcutaneously once daily for 28 days.
That selection is not based on peptide-forum convention. It is the exposure used in two randomized neuropathy trials, and 4 mg was the only dose with a statistically significant corneal nerve-fiber result in the Phase 2b dose-ranging study. Going to 8 mg would not be “more evidence based”; the 8 mg group failed to reproduce the 4 mg result.
Four weeks is also where the largest body of human tolerability data sits. The 12-week exposure comes from only eight completers in a study that missed its efficacy outcomes.
This makes 4 mg for 28 days the best-supported personal choice, not a personalized medical recommendation for anyone else and not evidence that the same exposure benefits a healthy, trained person.
Safety: reassuring at four weeks, incomplete beyond it
The short human trials are more reassuring than the usual research-peptide anecdotes. They did not show the rise in hemoglobin or hematocrit expected from erythropoietic EPO signaling, and routine laboratory findings did not reveal a consistent hematologic toxicity pattern.
The most commonly reported issues in the dose-ranging trial included injection-site pain, diarrhea, fatigue, headache, and nausea, without a clear dose relationship. One participant in the 8 mg group experienced suicidal ideation that investigators considered possibly treatment related. In the diabetes study, four serious adverse events occurred in active-treatment participants; the reports did not establish a consistent causal pattern, but one participant stopped after worsening borderline renal insufficiency in the setting of an increased diuretic dose. A later fatal myocardial infarction following cellulitis was judged unrelated.
That record is not a major safety alarm. It is also far too small to exclude uncommon harms, identify long-latency risks, or describe repeated use. Most of the meaningful dataset is four weeks long. There is no completed Phase 3 program.
ARA-290 is not an approved medicine. The FDA currently lists cibinetide in 503A Category 3, meaning the agency has insufficient supporting information to evaluate it for inclusion on the 503A bulks list. Category 3 is not the same designation as Category 2, which covers substances that raise significant safety risks, but it is not an endorsement of safety, effectiveness, or routine compounding either.
The N=1 measurement problem
The objective endpoints that make ARA-290 compelling are precisely the endpoints I cannot put into a normal biohacking dashboard.
The trials used tools such as:
- Corneal confocal microscopy, which images small nerve fibers in the cornea.
- Skin biopsy, including staining for intraepidermal fibers and the regeneration marker GAP-43.
- Quantitative sensory testing, which measures detection thresholds for temperature and other stimuli under standardized conditions.
- Validated neuropathy symptom instruments in people with an established symptomatic condition.
Contemporary small-fiber-neuropathy guidance relies on clinical signs plus specialized measures such as quantitative sensory testing and skin-biopsy nerve-fiber density; there is no simple standalone home gold standard.
A wearable cannot substitute for these tests. HRV, resting heart rate, sleep score, soreness, workout readiness, and training performance may be valuable measures for other questions, but they do not demonstrate small-fiber regeneration.
Pain and altered sensation could be tracked at home if I began with a stable, specific neuropathic symptom. I could repeatedly score burning, tingling, allodynia, temperature perception, or a defined painful area. Even then, the human trials warn me that structural nerve change and pain relief do not necessarily move together.
Without a real baseline deficit, the problem becomes much worse:
- There may be a floor effect: healthy sensation has little room to improve.
- General soreness or recovery is too noisy and too far from the validated clinical target.
- A “no noticeable effect” result cannot distinguish an inactive compound from unmeasured nerve repair.
- A “felt better” result cannot distinguish nerve repair from expectation, training variation, sleep, or dozens of other inputs.
That is not a reason to dismiss the experiment. It means the experiment cannot give me a clean yes-or-no answer about nerve regeneration.
There is a legitimate difference between testing a claim and making a biological bet. A measurement-first experiment asks whether I can demonstrate that an intervention worked. This experiment asks whether a short exposure with randomized human evidence, an objective nerve-regeneration signal, and a reasonably reassuring four-week safety record is worth running even if its central effect remains below the threshold of perception.
For me, the answer is yes.
What I would expect
My expectation is modest and specific.
In a person with documented small-fiber neuropathy, ARA-290 has a credible chance of influencing nerve-regeneration biology over four weeks. The evidence for symptom relief is less consistent than the evidence for objective nerve-fiber change. Whether those findings generalize to a healthy biohacker is unknown.
I would not expect obvious muscle growth, a stimulant-like sensation, immediate analgesia, or a dramatic improvement in generic workout recovery. Those are not the outcomes supported by the human trials.
Running 4 mg daily for 28 days while healthy and without a defined neuropathic symptom may produce a personal report of “nothing clearly noticeable.” I would not interpret that as failure. The human trials found structural nerve changes even when pain improvement was inconsistent, so an unnoticed effect is entirely compatible with the published evidence.
The experiment can still tell me whether the exposure is tolerable, whether any subjective changes emerge, and whether something unexpected happens. What it cannot tell me is whether an uneventful four weeks contained useful small-fiber repair beneath the level of perception.
So my current conclusion has two parts:
- I will use 4 mg subcutaneously once daily for 28 days, the strongest literature-based choice.
- I think the human nerve-regeneration signal makes the experiment worth running even if I notice nothing.
- I can document exposure, tolerability, and any subjective changes, but without specialist pre/post testing this at-home N=1 cannot establish whether nerve regeneration occurred.
ARA-290 survives a serious literature review better than most peptides. In my framework, its combination of human structural evidence, a defined four-week exposure, and the possibility of an unnoticed but useful effect is enough to justify the experiment.
This will not be a clean efficacy test. It will be a documented exposure to a biologically credible intervention whose most important outcome may remain unmeasured. That is a limitation on the conclusion, not a reason to assume there was no benefit.