Many hair growth supplements pass right through your system unabsorbed. Your body absorbs nutrients from real food 10x better than from a pill. Here are my Top 5 vitamins and minerals for hair density, where to find them, and the when you should supplement. Enjoy the read 👇

Men taking a slow-release minoxidil pill twice a day gained 36.4 hairs per square centimetre over six months. Men on placebo gained 6.6. That's from a 536-man phase 3 trial Veradermics reported on 19 Oct, and the gap between those two numbers matters more than either. The gap is about 30 hairs per square centimetre on twice daily and 22 on once daily (8.5 mg a dose), in men with mild to moderate pattern hair loss. Placebo groups grew some hair in this trial and in the company's first big one, so the difference is the fair measure. The count skipped vellus hairs, the fine colourless fuzz, so these are thicker hairs. The men's own verdict points the same way. On twice daily, 68.7% rated their coverage improved or much improved, against 15.0% on placebo, and the drug pulled ahead by month 2, the first check. The earlier trial, Study 302 (519 men), landed in the same place: 30.3 and 33.0 on the two doses, against 7.3 on placebo. Safety looks clean so far, with a hole in the data. The release reports no treatment-related serious side effects and no heart events of special interest. Ankle or leg swelling hit 4.5% on twice daily (0.6% on placebo) and unwanted body or facial hair 9.0% (none on placebo). Heart rate and blood pressure aren't reported, which matters for a drug whose tablet label carries a boxed warning about the heart. The new trial was randomised, double-blind and placebo-controlled, and run by the company. Its results so far are in a press release; there's no peer-reviewed paper yet. So it's still to come. Twelve-month safety data are due later this year and a trial in women in the first half of 2027; no filing date or price has been announced. Generic minoxidil tablets are approved only for hard-to-treat high blood pressure, so using them for hair loss is off-label. Thirty extra hairs per square centimetre is a real effect. Whether it beats finasteride or the generic tablets is something a trial with only placebo to compare against can't say.

Every "I'm a Norwood 3" goes back to a 1975 paper in which Dr O'Tar Norwood graded 1,000 white men to give hair transplant surgery a standard scale (Southern Medical Journal). It's an observational chart, so it shows how balding looked in white men, not women or anyone else.

In a study that followed 36,032 American men for about 20 years, those with the most hair loss at 45 got squamous cell skin cancer on the scalp about seven times as often as men with none (Li, Int J Cancer 2016). Away from the scalp, there was no clear difference. They were US health professionals who, in 1992, picked one of five drawings of a crown to match their hair at 45, then were tracked for new skin cancers. Over the whole body, the baldest men had about a third more squamous cell cancer and a fifth more basal cell cancer. Melanoma overall showed no link. Split by site, the picture sharpened: - scalp squamous cell cancer: 2.5 times the rate with moderate loss, 7 times with the most - scalp melanoma: also about 7 times, but from 14 cases in all, so a hint - elsewhere on the head, neck and body: no clear difference The excess sat where the hair had gone, which fits sunlight on newly bare skin. The analysis allowed for sunburns, moles, hair colour, time outdoors and how easily they burned as kids. The weak spot: hair loss was recalled years later, not photographed. A 2023 genetic study agrees (Ong, Nature Communications). In about 227,000 UK Biobank men, balding-linked gene variants predicted more squamous cell cancer and head and neck melanoma, but not melanoma elsewhere, and testosterone genes didn't explain it. Some balding genes also lighten skin and hair, which explains part of the link. Protection is cheap and available now. In a trial in Nambour, Australia (Lancet 1999), 1,621 adults randomised to daily SPF 15+ on the head, neck, arms and hands for 4.5 years had about 40% fewer squamous cell cancers, though the number of people affected didn't differ significantly. The American Academy of Dermatology's line for thinning hair: sunscreen on the scalp or a wide-brimmed hat. The bald-scalp link is observational, and I couldn't find a trial of sunscreen on bald heads. The case rests on where the cancers turn up, which is exactly where the hair isn't.

Coconut oil has real hair science, just not for hair loss. The usual study (Rele & Mohile, J Cosmet Sci 2003, by an industry R&D team) tested cut hair: coconut oil reduced protein loss in washing. No follicles were involved, so the regrowth verdict is none yet.

Smokers had 1.46 times the odds of pattern hair loss across 11 observational studies in a 2026 meta-analysis (Li et al., BMC Public Health), and 1.6 times the odds of it being moderate to severe. Association, not proof, but unlike your genes, smoking is a factor you can change.

In 2012, bald scalp was found to hold about twice as much prostaglandin D2 (PGD2), a molecule that slows hair growth in lab tests, as haired scalp from the same men. A pill blocking its receptor was then tested in balding men. Hair counts rose 6.7 per cm² on it, 7.0 on placebo. The 2012 study (Garza and colleagues, Sci Transl Med) used scalp left over from hair transplants. In 17 men, bald patches held 53.5 nanograms of PGD2 per gram against 25.6 in haired scalp. PGD2 left hairs 62% shorter in human follicles kept in a dish, and in mice it worked through one receptor, GPR44. Nobody had been treated, but at least ten GPR44 blockers already existed, because of the receptor's suspected role in allergies, and the authors suggested trying them on hair. In 2015 a computer study (Fong and colleagues, J Ethnopharmacol) docked 389 plant compounds into models of the enzyme that makes PGD2. Ricinoleic acid, about 90% of castor oil's fatty acids, was one of five predicted blockers; lab and animal tests were left for later. From 2016 to 2018, an Allergan-sponsored randomised trial gave men aged 18 to 49 with thinning crowns setipiprant, a GPR44 blocker first developed for allergies, at 1,000 mg twice a day (83 men) or placebo (74) for 24 weeks (DuBois and colleagues, Clin Cosmet Investig Dermatol, 2021). It didn't beat placebo on hair counts or ratings. A 12-man finasteride group gained 33.9 hairs per cm², but was dropped partway through and never formally compared. Treatment-related side effects were twice as common on setipiprant, all mild or moderate. A 2026 dermatology review of castor oil (Girdler and colleagues, Cureus) cites the PGD2 link next to that trial, and says its hair claims haven't been studied for clinical efficacy. For regrowth, castor oil rates none yet. Nothing aimed at PGD2 is approved for hair loss. The trial's authors suggest the lab biology may not carry over to a living scalp, or that it needed a higher dose. I found no larger trial since.

Ten. That's how many men on the product were analysed in the one placebo-controlled exosome hair trial I could read in full, followed for 16 weeks. Clinics already use exosome injections on thinning scalps, and the FDA says no exosome product is approved. The trial (Amini and colleagues, Life, 2025) randomised 24 men with early pattern loss at a Malaysian university to a commercial exosome product with two plant extracts, or saline, injected into the scalp under local anaesthetic in four sessions about two weeks apart. Twenty were analysed. On the product, the median gain was 9.5 hairs in a 0.5 cm² patch against 1.5 on saline (p=0.006). The spread was wide, two men had mild scalp irritation, and the authors call their sample small and follow-up short. The paper doesn't say where the exosomes came from. The company that supported the research materials also prepared the coded group labels, and the trial was registered a week after the paper was accepted. That trial sits near the top of a thin pile. A 2025 review (Al Ameer and colleagues, Clin Cosmet Investig Dermatol) found 11 studies of exosomes for hair loss: - 2 randomised trials - 3 single-group studies - 3 look-backs at clinic records - 1 case series - 2 case reports All reported improvement and none reported serious side effects, but the nine without random allocation can't show what would have happened anyway. The review's authors call exosomes promising and still ask for well-standardised, high-quality randomised trials. The FDA adds that exosome products meant to treat a condition need its approval, and in 2019 it warned of serious adverse events in Nebraska patients given unapproved ones (that alert doesn't mention hair). That leaves the evidence for exosomes in pattern hair loss weak, though clinics offer them now. With exosomes drawn from six different stem-cell sources plus a plant-extract product, and no standard dose or schedule, a result for one product says little about the next.

Trying for a baby while on finasteride raises two questions: whether it affects his sperm, and whether any of it reaches her. The second has the clearer answer, so start there. Finasteride does reach semen, in tiny amounts. In US label tests on 35 men taking 1 mg a day, 21 had none detectable. The label's upper estimate of a partner's exposure, 7.6 ng a day, is 650 times below a 5 microgram dose that had no effect on men's DHT. The UK label puts it under 0.001% of the dose per ejaculation. For pregnancies exposed this way, the UK label lists eight live male births in post-marketing reports and one retrospective report of a mild genital defect, hypospadias, which it says a single report can't pin on the drug; trials added nine boys with no defects reported. The firm warning is about women who are or may be pregnant handling crushed or broken tablets. His side is murkier. On the US label, a 48-week placebo-controlled study found median semen volume fell 11% on 1 mg against 8% on placebo; at 5 mg, the prostate dose, the drop was about 25%, and reversible. A 2020 review of sperm counts at 1 mg (Alajlan and Alfaraj, J Pak Assoc Dermatol) found seven small studies with 113 men; the largest, 91 men, saw no change. Against that sit reports. Both labels list infertility or poor semen quality after launch, with semen improving after stopping in some cases. A 2026 analysis of US and European safety databases (Gong et al., Frontiers in Pharmacology) found 469 and 288 such reports for finasteride, a far bigger share of its reports than for other drugs (reporting odds ratios of about 74 and 60). The authors say such signals "cannot directly establish causality", and the UK label notes other risk factors may have played a part in some. Finasteride 1 mg is available now on prescription. On evidence this thin, nobody can put a rate on the risk to sperm, and the UK label says plainly that long-term fertility data in humans are lacking.

A viral reel this year told men that cold drinks speed up balding. Trace the research that fact-checkers and headlines point to and you reach one 2023 online survey of young Chinese men, which linked self-reported hair loss to sugary drinks, and to a lot else besides. The reel drew more than 1.3 million views, according to a June fact-check by The Week, which cited that survey, a 2025 review and others. The review (Gomes et al., Nutrition and Health) gathered 17 diet studies, with 97% of participants women, and its only sugary-drink study is the survey. That survey (Shi et al., Nutrients 2023; 1,028 men aged 18 to 45) linked self-reported hair loss to plenty besides sugary drinks: - smoking, now or in the past - drinking alcohol - shorter sleep and less exercise - more fried food, sweets and ice cream, and fewer vegetables - severe anxiety When nearly every habit tracks with the outcome, a survey can't say what drives what, and its authors say the design can't establish cause. They also found anxiety carried part of the sugary-drink link. Men already seeing their hair thin may sleep worse and eat differently, which a one-off questionnaire can't untangle. Hair loss was self-reported, not examined, and 57.6% of the men, average age 28, placed themselves in the hair loss group. The drink link itself shrank as the researchers adjusted for other factors, from odds of 3.36 to 1.57, which wasn't statistically significant. As a hair-saving tip, cutting sugary drinks rates weak: one self-reported survey and no trial. There's nothing to buy. The reel's villain was the cold. The survey counted sugar-sweetened drinks by type and amount, and how cold they were never came into it.

Two months into minoxidil, the scalp itches and flakes, and forums have a ready answer: swap the liquid for the foam, since the liquid's propylene glycol is to blame. Sometimes it is. In published allergy cases, though, minoxidil itself turns up more often. Itch on minoxidil has look-alike causes, listed in a 2002 Journal of the American Academy of Dermatology paper (Friedman et al.): - irritation, no allergy involved - a seborrhoeic dermatitis flare (the flaky, dandruff-type rash) - allergy to the solvent, usually propylene glycol - allergy to minoxidil itself Only patch testing (each ingredient tried on the skin separately) tells the allergies apart. In Friedman's patients, propylene glycol was "the contactant in a majority of cases, not the minoxidil itself". A 2025 review in Acta Dermato-Venereologica (Junge et al.) gathered 21 published reports covering 142 patients; among positive patch tests, 54 people reacted to minoxidil and 12 to the base. Case reports collect the unusual, so that's no rate, and one series in the review (Aleid et al.) found the reverse: 8.8% reacted to propylene glycol, 5.5% to minoxidil. On the US labels I checked, the 5% foam lists no propylene glycol, while the 5% liquid lists alcohol, propylene glycol and water. Both deliver the same drug, so switching can help if propylene glycol is the culprit and not if minoxidil is. Friedman's paper says people allergic to minoxidil can't use any topical form. The review also notes nine women allergic to topical minoxidil who took low-dose tablets without problems. Those are off-label for hair loss, and the authors say a whole-body reaction can't be ruled out. It's all case reports and patch-test clinics; I couldn't find a trial comparing foam with liquid for itch. As a fix, switching rates weak, since it deals with one cause of four. Both forms are sold over the counter in the US. Nobody knows how common a true minoxidil allergy is. The review says its prevalence "remains unknown".

Does anyone else’s internal monologue suffer from Tourette syndrome? Notably when walking past coworkers in the office

The better the study, the weaker the link between balding and heart disease. Studies that start with heart patients and look back find the biggest effects; a German study that photographed scalps first and waited ten years found a small one that could be chance. Start at the bottom. Case-control studies begin with heart patients and healthy comparisons and ask about past hair loss. A 2013 BMJ Open meta-analysis (Yamada et al.) pooled three and got about 70% higher relative risk with balding; the authors flag recall as a weak spot. One rung up, cohort studies record hair first and wait. Yamada's three cohorts pooled to 32% higher risk for severe balding. The crown detail is from the same paper: where studies scored crown and front separately, severe crown loss came with 48% higher risk and frontal loss with 11%, not significant. Higher again, the German team photographed the scalps of 1,673 men aged 50 to 80 rather than asking, then followed them for about ten years (Pechlivanis, PLoS ONE 2019). Any balding came with 1.2 times the risk of a heart event, a range that includes no effect, with crown and frontal subgroups much the same. Balding was weakly tied to diabetes and body mass index. Genetics tests another idea, shared genes. Yap's team (Nature Communications 2018) studied 205,327 men and found no genetic correlation between balding and coronary artery disease. Metabolic trouble keeps turning up beside balding. A 2022 meta-analysis of 19 case-control studies (Qiu et al., Acta Dermato-Venereologica) found about three times the odds of metabolic syndrome in men with pattern hair loss, and systolic pressure about 8 mmHg higher. It's the bottom-rung design again, without adjustment for smoking or family history. On the best observational evidence (nobody can randomise baldness), a thinning crown is a weak signal at most, and there's nothing to treat or buy. A blood pressure cuff and a cholesterol test make far better heart checks than a mirror.

A 2023 study worked out why hairy moles grow hair, and a company co-founded by the scientist who led it now counts Eli Lilly among its investors. But the first injection that company tested in people isn't the molecule from the moles. On 21 June 2023, Maksim Plikus's lab at UC Irvine published the mole study in Nature (Wang et al.). In mice bred to grow moles, ageing pigment cells pumped out a protein called osteopontin, which woke nearby hair stem cells; remove it, or its receptor CD44, and the extra growth stopped. On human scalp follicles grafted onto mice, it brought on the growth phase sooner (11 samples against 7). Lab work, in other words. Six days later, Amplifica, a company Plikus co-founded, dosed its first human volunteer with an injectable called AMP-303. Its release described a randomised, double-blind, placebo-controlled study in men, with safety as the main goal. In November 2024, UC Irvine's biology school reported "more than a 15% increase in non-vellus hair" (thicker hairs, not fuzz) against placebo at 60 days, lasting up to 150 days. That's a summary of a company trial rather than a paper, and it doesn't say how many men saw it. In May 2025, the trial's poster at a dermatology research meeting was titled a "feasibility study of a novel polysaccharide". Osteopontin is a protein. Whatever AMP-303 is, and the company hasn't named it anywhere I could find, it isn't the mole molecule. In August 2026, Amplifica closed a $26 million round with Eli Lilly taking part, according to UC Irvine's innovation office. So where does that leave someone thinning now? The mole signal is early lab work, and I couldn't find a human trial of osteopontin itself. AMP-303 has one feasibility trial, reported in a poster and press releases with no peer-reviewed paper yet. Neither is available or priced. The next thing worth watching is a journal paper with the full AMP-303 data, and a name for what's in the syringe.

"Rosemary oil works as well as minoxidil" rests on one 2015 trial, and on one number in it: 2%. That was the strength of minoxidil it was up against, the weaker of the two, and it changes what the result means. Panahi et al. (Skinmed, 2015) randomly gave 100 men with pattern hair loss either rosemary oil or 2% minoxidil for six months. There was no placebo group, so nothing showed what six months of doing nothing would look like. At three months, neither group's hair count had moved significantly. By six months both had risen, and the difference between them wasn't statistically significant. Scalp itching was more common on minoxidil, though rosemary carries its own risk: a 2014 review of case reports (Miroddi et al.) describes allergic contact dermatitis, often blamed on carnosol, one of its compounds. Now the number. In the US, 5% minoxidil is sold over the counter as a "hair regrowth treatment for men". A 2025 network meta-analysis (Gupta et al., Int J Mol Sci) pooled 15 trials of over-the-counter options for men and estimated 5% minoxidil grows about 20 more hairs per square centimetre than rosemary over 24 weeks. That's an indirect estimate, and its rosemary data all come from the 2015 trial. So rosemary oil's evidence is weak: one small, single-blind trial against the weaker minoxidil, with no placebo, and a non-significant gap in 100 men doesn't make two treatments equal. It's easy to buy and might do something over six months. The newest rosemary trial I found doesn't settle it. Patel et al. (Cureus, 2025) tested two oil blends containing rosemary against coconut oil in 90 adults with thinning hair for 90 days. No pattern hair loss diagnosis was required, and the blends' maker paid for the study. What I couldn't find is the fair test: rosemary oil against 5% minoxidil and a true placebo. Until someone runs it, the 2% is doing most of the work in that claim.

Is dutasteride just a stronger finasteride? In the pooled head-to-head trials, which ran six months, slightly yes. Beyond that the evidence goes quiet, and the two drugs differ in ways the forum shorthand skips. How much stronger? Zhou and colleagues (Clinical Interventions in Aging, 2019) pooled three randomised, double-blind trials in 576 men. After 24 weeks, dutasteride had grown about 29 more hairs in the counted area than finasteride (95% CI 19 to 38). Photo ratings favoured it too, though the expert panel's margins were small. Why would it be stronger? Finasteride blocks the type II form of 5-alpha reductase, the enzyme that turns testosterone into DHT. Dutasteride blocks types 1 and 2. Its label reports a median 90% drop in blood DHT after two weeks; finasteride's label reports 65% within a day. Different men in different studies, so that's a rough comparison at best. Are the side effects worse? In the pooled trials, libido, erection and ejaculation problems showed no significant difference, but trials this small and short can't rule a gap out. Dutasteride's own label data come from older men with enlarged prostates: in the first six months, impotence was 4.7% against 1.7% on placebo and decreased libido 3.0% against 1.4%, with both gaps closing after that. Is it licensed for hair loss? Not in the US. Its label covers an enlarged prostate, so for hair it's off-label. What changes day to day? Its half-life is about five weeks, so it lingers. The label asks men to wait at least six months after the last dose before giving blood, says women who are or may be pregnant shouldn't handle the capsules, and notes it roughly halves PSA readings, which matters for prostate checks later on. So it's available now, off-label, with decent evidence for a modest edge over six months. The pooled head-to-head trials stop at 24 weeks, a short window for a drug people take for years.

In 1989 the FDA went through the rub-on "hair growers" sold over the counter in the US (biotin, B vitamins, amino acids, jojoba oil, wheat germ oil, even estradiol) and ruled there wasn't adequate data to show any of them was safe and effective. That rule still stands. It's 21 CFR 310.527, published on 7 July 1989. The list runs to 17 entries, some of them odd company for a hair product, like tetracaine (a numbing agent) and sulfanilamide (an early sulfa drug). The sentence that matters comes after the list: there is "a lack of adequate data" for "these or any other ingredients" sold over the counter to grow hair or prevent hair loss. Read that as unproven rather than proven useless. It was a regulator weighing the evidence of the 1980s, and a review isn't a trial. But "any other" does a lot of work. Since January 1990, a non-prescription product promoted to grow hair or prevent hair loss has counted as a new drug, which needs an approved application before it can be sold for that purpose. Ingredients that missed the 1989 list get no pass. Minoxidil shows what the approved route looks like: a full drug application, approved on prescription in 1988 and sold over the counter since 1996. Cosmetics may change how hair looks, so a promise of thicker-looking hair stays on the cosmetic side; a claim that a product will "restore hair growth" is the FDA's own example of a drug claim. Pills and gummies come under supplement law, which allows claims about the body's structure or function if the label carries a disclaimer that starts "This statement has not been evaluated by the Food and Drug Administration." So in the US, a rub-on that promises regrowth needs an approval behind it, as minoxidil has. A supplement with that disclaimer is telling you, in its own small print, that nobody at the FDA has evaluated the claim. The FDA never said these ingredients can't work. It said the proof wasn't there, and the rule hasn't been amended since.

About 28 extra hairs per square centimetre over placebo. That's the average gain from PRP scalp injections across 13 randomised trials in a 2024 meta-analysis. The same paper rated the evidence low quality, and its reasons are worth knowing before anyone books a course. PRP (platelet-rich plasma) means drawing a little of your blood, spinning it in a machine and injecting the plasma into thinning areas, as the American Academy of Dermatology describes it. The number comes from Kieling and colleagues (Anais Brasileiros de Dermatologia, 2024), comparing PRP with placebo: 27.55 hairs per square centimetre, 95% confidence interval roughly 14 to 41. That average hides a mess. The trials disagreed so much that the heterogeneity score (I², how far results vary beyond chance) hit 96%. The authors saw signs of publication bias too, where positive trials get published and flat ones don't. Hair thickness didn't change significantly, most trials checked results within about six months, and at least three were company-funded. One striking result shows both sides. In Gentile and colleagues' half-head trial (Stem Cells Transl Med, 2015), 20 men had PRP on one side of the scalp and saline on the other, three sessions a month apart. Density rose 45.9 hairs per square centimetre on the PRP side and fell 3.8 on the saline side. About a year later, four of the 20 were losing hair again and had another round. That makes the verdict weak. The trials lean positive on average, but they're too inconsistent to predict individual results, and the reviewers called for better ones. PRP is available now at clinics, and it isn't a one-off. The AAD describes monthly sessions for three months, then top-ups every three to six months, and calls it "not a permanent solution". The commonest side effect in the trials was brief pain at the injection sites. The 28-hair average is real. It's also the average of trials that couldn't agree, which makes it a poor guide to any one scalp.