TL;DR
All three of these popular hair actives have a plausible story and thin human proof. Topical caffeine has a real in-vitro mechanism and one noninferiority trial against minoxidil, but sparse rigorous data. Collagen peptides are marketed hard for hair while the actual hair evidence is nearly absent: most trials are about skin, and hair is keratin, not collagen. Pumpkin seed oil rests on a single small 2014 trial that has never been replicated. None of the three is in an Unfurl formula, which is precisely why we can grade them without flinching.
Every so often the hair aisle adopts a new darling. Right now three of them are everywhere: caffeine (in shampoos and serums), collagen (in every powder, gummy, and "beauty" latte), and pumpkin seed oil (the supplement corner's current favorite). Each has a story good enough to sell. The question this article answers is the one the labels won't: for each of these, is there a real mechanism, and is there real human evidence, and are those two the same thing? (They almost never are.)
We're running each through the same lens we built in the snake-oil detector: mechanism plausibility, evidence tier, sample size, funding, replication. And a disclosure up front, because it changes how you should read this: none of these three ingredients is in any Unfurl product. That means we have no bottle to defend here, which you should treat with its own skepticism, since a brand is always freest to be honest about the things it doesn't sell. Weigh the reasoning, not our neutrality.
Does topical caffeine actually grow hair?
Caffeine has the most legitimate mechanism of the three and still surprisingly little human proof, a plausible lab story that hasn't been closed out in people. The mechanism traces to Fischer and colleagues in 2007, who published an in-vitro study on caffeine and human hair follicles in the International Journal of Dermatology. Working with follicles from men with androgenetic alopecia, they found that testosterone suppressed follicle growth in the dish, and that caffeine at low concentrations counteracted that suppression, and even stimulated follicle growth on its own. It's a genuinely interesting result, and it's why caffeine got into shampoos in the first place.
But notice the tier. That's a follicle in a dish: no bloodstream, no scalp, no hormones cycling through a living body. In-vitro plausibility is a starting line, not a finish line. The one notable human study is Dhurat and colleagues, 2017, an open-label trial in Skin Pharmacology and Physiology that tested a caffeine-based 0.2% topical liquid against minoxidil 5% in 210 men and reported the caffeine formula was "noninferior" to minoxidil over six months. That sounds impressive until you read it against our own framework: it was open-label (nobody was blinded), it was a noninferiority design (which can pass even when both arms are modest), and the honest questions about funding and independent replication apply. There is no large, blinded, placebo-controlled trial putting topical caffeine to bed.
Grade: mechanism plausible, human evidence thin. Caffeine isn't snake oil; the mechanism is real and the one comparative trial is suggestive. But "real mechanism in a dish plus one unblinded trial" is a long way from "proven on your head," and any product implying otherwise is stretching. There's a shampoo-versus-serum wrinkle, too: caffeine in a shampoo is on your scalp for the length of a rinse, while the studied formulations were leave-on. Contact time matters, and most caffeine shampoos quietly ignore it.
Do collagen supplements do anything for hair?
This is the one where marketing has outrun the data most severely: the hair evidence is close to absent, and the biology argues against the premise. Start with the biology, because it's decisive: your hair shaft is made almost entirely of keratin, not collagen. Collagen is the structural protein of skin, tendon, and connective tissue. Swallowing collagen does not deliver collagen to a keratin structure; it's digested into amino acids like any other protein, and your body allocates those amino acids wherever it sees fit, which is not preferentially to your follicles.
Now the evidence. The clinical literature on ingestible collagen is overwhelmingly about skin (hydration, elasticity, wrinkles), where there are real, if industry-heavy, trials. Direct evidence for hair is a different and much emptier story. A recent narrative review of collagen supplementation in Nutrients makes the state of things plain: the hair-and-nail claims rest on limited, short, heterogeneous, frequently multi-ingredient and industry-funded studies, exactly the profile that our snake-oil test tells you to discount. When a hair benefit does show up, it's usually inside a blended "beauty" supplement that also contains biotin, zinc, and other actives, so you can't attribute anything to the collagen. (And on the biotin in those blends, see why that's mostly theater for non-deficient people.)
Grade: marketing far ahead of the data. Collagen has a strong skin case and a nearly empty hair case, sold as if the two were the same. If your protein intake is genuinely low, correcting that could help your hair, but that's an argument for eating enough protein, not for a collagen scoop specifically. Of the three actives here, collagen is where the gap between the ad and the evidence is widest.
Is pumpkin seed oil a real hair treatment?
Pumpkin seed oil has exactly one trial worth citing, and once you read how it was built, it becomes a case study in why single studies don't settle anything. The trial is Cho and colleagues, 2014, published in Evidence-Based Complementary and Alternative Medicine. Seventy-six men with mild-to-moderate androgenetic alopecia took 400 mg a day of a pumpkin-seed-oil capsule or placebo for 24 weeks, and the treatment group showed a roughly 40% increase in hair count versus about 10% in the placebo group. On its face, a real randomized, placebo-controlled result, better designed than most ingredient studies in this category.
Then you read the fine print, and it does the study in. First, the capsule wasn't pumpkin seed oil; it was a proprietary blend that also contained saw palmetto and other components, so the effect can't be cleanly attributed to pumpkin seed oil at all. That's the "proprietary complex" red flag firing exactly as designed. Second, the study was funded by a grant from the company that supplied the capsules, a textbook conflict of interest that calls for independent replication. Third, the authors themselves noted limitations: no histological confirmation and no measurement of DHT, so the mechanism is unestablished. And fourth, and most important, in the decade-plus since, this trial has not been replicated in humans. One small, industry-funded, multi-ingredient study is a hypothesis worth testing, not a treatment worth trusting.
Grade: single-trial, unreplicated, muddied by design. Pumpkin seed oil is not disproven; it's under-proven, and the one study people cite is weaker than its headline suggests. Interesting; not established.
The honest limitation, including on our own neutrality
Here's the counterweight, placed after the grading rather than before it: "thin evidence" is not the same as "disproven," and we don't want you to walk away thinking these three are frauds. Caffeine has a genuine mechanism. Pumpkin seed oil has a real, if flawed, positive trial. Collagen has a legitimate skin story and provides amino acids your body can use. Absence of strong evidence is a reason to keep your expectations modest and your money cautious, not a reason to sneer. Some of these will likely get better trials, and a few may hold up.
And turn the skepticism back on us while you're at it. We told you none of these is in an Unfurl formula, and that a brand is freest to be honest about what it doesn't sell, but that same freedom is a bias. It costs us nothing to grade caffeine harshly; it would cost us something to grade our own peptides that harshly. So don't take our neutrality as virtue. Take it as a reason to check our reasoning against the linked studies yourself, which is the whole point of citing them. This is not finasteride, and neither are any of these three; none of them will lower your DHT, and the honest ceiling for any cosmetic active is supporting the appearance of your hair while your own consistency does the slow work.
So what do we actually put in our formulas?
We build around actives whose evidence we're willing to show you in full, and we still label them "supporting," not "proven." Rather than caffeine, collagen, or pumpkin seed oil, our plant-first botanical serum leads with rosemary and a marine-algae blend, and our nine-oil scalp treatment is a twice-weekly night conditioner for the scalp environment, with neither claiming to regrow anything. Where an ingredient of ours does have human data, like saw palmetto, we lay out the trials and their limitations; where it's more preliminary, like our peptides, we say that too. You can see the whole lineup in the full collection.
Who should skip all of this: anyone whose thinning has a treatable cause not yet ruled out. Before you chase any trendy active, the two deficiencies genuinely worth testing are iron and vitamin D (see our ferritin and vitamin D pieces), and if you'd rather be routed than research, the 90-second hair quiz maps what you're seeing to a sensible next step, ours or otherwise.
Frequently asked questions
Does caffeine shampoo work for hair loss?
The mechanism is real in the lab: Fischer's 2007 in-vitro work showed caffeine counteracting testosterone-driven follicle suppression, but human proof is thin, resting mainly on one open-label trial. And shampoo is a weak delivery vehicle: caffeine rinses off in under a minute, while the studied formulas were leave-on. Modest expectations are the honest setting.
Is collagen good for hair growth?
The evidence is close to absent for hair specifically. Most collagen trials are about skin, and your hair is made of keratin, not collagen. Swallowed collagen is digested into amino acids your body uses wherever it wants. If your protein intake is low, fixing that may help, but a collagen scoop isn't specially proven for hair.
Does pumpkin seed oil regrow hair?
It rests on a single 2014 trial, Cho and colleagues, which showed a benefit, but the capsule was a proprietary blend containing saw palmetto (not pure pumpkin seed oil), the study was funded by the supplier, and it's never been replicated. Interesting, not established. "Regrow" is a drug claim no supplement can make.
Which of the three has the best evidence?
Topical caffeine, narrowly: it has both a mechanistic study and one comparative human trial, even if that trial was unblinded. Pumpkin seed oil has one flawed trial; collagen has essentially no direct hair evidence. None reaches the bar of a large, blinded, replicated, placebo-controlled result.
Why aren't these ingredients in Unfurl products?
Because we build around actives whose evidence we can show you and whose limitations we can state honestly. These three either have thin human data or a mechanism that doesn't fit hair biology well. That said, a brand is always freest to be skeptical about what it doesn't sell, so check our reasoning against the linked studies rather than taking our word.
Should I take any supplement for my hair?
Only to correct a genuine, tested deficiency, most usefully iron or vitamin D, which are both common and plausibly tied to hair. Trendy actives like collagen and pumpkin seed oil are far down the priority list behind ruling out a real, fixable cause with a clinician and a short lab panel.
Our editorial approach. Unfurl content is researched and written by named authors and cross-checked against peer-reviewed literature (PubMed, Cochrane, NIH ODS) before publishing. We cite every claim inline. Articles are editorially reviewed, not medically reviewed, and nothing on this site is medical advice. Talk to a clinician before starting any supplement. These statements have not been evaluated by the FDA.
A friendly note to close: we're a hair and scalp care company, not a medical one. Everything here is for general education, not medical advice, and it can't diagnose a condition or take the place of a visit with someone who can examine you. If your hair or scalp has you worried, please talk to a doctor or a licensed clinician you trust.