Dihexa is 10 million times more potent than BDNF for promoting synaptogenesis — the most powerful nootropic ever developed.

The "10 million times more potent than BDNF" claim is one of the most-cited and least-defended numbers in the recreational peptide community. It propagated from a 2013 paper from the Wright / Harding laboratory at Washington State University and has appeared in nearly every Dihexa marketing context since. The numerator is real; the framing is wrong.

Where the number actually comes from

McCoy et al. 2013 characterized Dihexa (N-hexanoic-Tyr-Ile-(6) aminohexanoic amide) as a metabolically stabilized angiotensin IV analog with procognitive activity in rodent models. The paper included a cell-culture synaptogenesis assay in dissociated hippocampal neurons. The reported finding: Dihexa promoted synapse formation in this culture system at concentrations approximately 7 orders of magnitude lower than BDNF required to produce the same effect.

This is the source of the "10 million times more potent than BDNF" figure. It is technically accurate for what the paper measured: a specific synaptogenesis endpoint, in a specific cell-culture system, at a specific concentration range, using molar potency as the comparator.

Why the framing fails as a cognitive-potency claim

Several structural reasons.

1. Cell-culture synaptogenesis is not human cognition. Dissociated hippocampal neurons in culture form synapses in response to many growth factors and small molecules. The synapse-counting endpoint in the McCoy 2013 assay is a downstream readout of growth-cone activity, not a direct measure of cognitive function in an intact organism. The translation gap from "cells form more synapses in culture" to "humans think better" is substantial and not addressed by the molar-potency comparison.

2. There are no published human Dihexa trials. The molar-potency-vs-BDNF figure is from rodent and cell-culture data. No human RCT, no Phase 1, no first-in-human exposure studies of Dihexa have been published. The closest clinical descendant is fosgonimeton (ATH-1017), an Athira Pharma developmental compound based on similar HGF/c-Met mechanism. The Athira Hua 2022 fosgonimeton Phase 1 showed safety and qEEG effects in healthy adults and Alzheimer's patients. The fosgonimeton Phase III LIFT-AD trial failed its primary endpoint in 2025. Athira paused development.

3. The "10 million times" headline obscures dose-dependence and biphasic curves. Many neuropeptide and growth-factor signaling molecules have inverted-U or biphasic dose-response curves. Higher molar concentrations produce ceiling or negative effects in many cases. The "more potent" framing implies a linear "more is better" logic that the underlying pharmacology does not support.

4. BDNF is a 27-kDa protein; Dihexa is a small modified tripeptide. Comparing molar potency of a small molecule to a protein is a category that pharmacologists routinely make but interpret carefully. Proteins have low intrinsic potency per molecule (they're large; they have receptor dissociation kinetics that limit per-molecule effect) but they have high physiological relevance (they're what the body actually uses for endogenous signaling). Comparing molar potency without engaging this asymmetry produces misleading framings on both sides.

5. The paper now carries an Expression of Concern. McCoy 2013 has carried a Journal of Pharmacology and Experimental Therapeutics Expression of Concern since September 2021. Two related Wright / Harding lab papers — Benoist 2014 (the HGF/c-Met dependence work) and Kawas 2012 (AngIV analog development) — were retracted in April 2025 as part of the broader Athira / Leen Kawas image-manipulation investigation. The mechanistic foundation that the "10 million times BDNF" claim rests on has substantially less research-integrity confidence than the original propagation of the figure implied.

See jpet-2025-dihexa-retraction-landscape for the full chronology.

What the actual evidence base looks like in 2026

• McCoy 2013 (foundational mechanism paper): Expression of Concern since 2021, not yet retracted
• Benoist 2014 (HGF/c-Met dependence): retracted April 2025
• Kawas 2012 (AngIV analog development): retracted April 2025
• No human Dihexa trials, ever
• Fosgonimeton (Athira's clinical successor): Phase III failed 2025, development paused
• The c-Met mechanism Dihexa was claimed to work through is also implicated in oncogenic pathways, raising chronic-use concerns

Where the critic has a real point

The molar-potency comparison in McCoy 2013 was a real finding at the methodology used. Cell-culture synaptogenesis at low Dihexa concentrations was reproducible within the laboratory at the time, and the angiotensin-IV-receptor-analog research program generated several papers consistent with procognitive effects in rodent models. The honest 2013 framing would have been: "an interesting in vitro and rodent finding worth investigating further in clinical development." That framing would have aged better than the "10 million times more potent than BDNF" marketing.

Where the critic loses the thread

The leap from "in vitro synaptogenesis assay shows a large molar-potency advantage over BDNF" to "Dihexa is the most powerful nootropic ever developed" is exactly the kind of evidence-base inflation that the methodology page section on what gets called out exists to prevent.

The right framing in 2026: Dihexa is a research compound with deeply compromised foundational mechanism literature. The "10 million times more potent than BDNF" figure is real for what it measured (cell-culture synapse counting) and meaningless as a clinical-potency claim. Anyone using Dihexa recreationally in 2026 is running an unguided self-experiment with a molecule whose mechanism papers are under serious research-integrity pressure and whose clinical descendant just failed Phase III.

See the cognitive enhancement decision guide for the honest evidence audit of where Dihexa actually sits among cognitive-enhancement options. The decision guide does not recommend Dihexa.