Hair Growth, Thinning and Why Hair Falls Out

Every hair on your head is a small factory that runs for a few years, shuts down for a few weeks, rests for a few months, drops its hair, and starts again. Eight follicles are drawn here in cross-section — skin, oil gland, stem-cell bulge, blood-fed papilla and all — and each one is running that cycle out of step with its neighbours. That single fact is why you lose 50 to 100 hairs a day and never notice. Break the asynchrony and you get a shedding crisis. Shrink the follicles instead and you get pattern baldness. You can do both here, and then treat them.

Try this: press Telogen effluvium and watch the timeline at the bottom — the stressor lands, nothing happens, and the hair does not start falling until roughly three months later. Then switch to Androgenetic, let it run twenty years, and turn on 💊 Finasteride to see what still-living follicles can and cannot get back.

The Science in Plain Language

Four phases, eight clocks, and why 50–100 hairs a day is normal

A hair follicle is not a permanent tube that quietly extrudes hair forever. It is an organ that rebuilds itself over and over, and it is one of the very few organs in the adult body that does. Each cycle has four acts.

Anagen is the growing phase, and on the scalp it runs for roughly two to seven years. During anagen the bulb sits deep — often down in the subcutaneous fat — and the matrix cells wrapped around the dermal papilla are among the fastest-dividing cells in the human body. They push out about a centimetre of hair a month. Because anagen is measured in years and scalp hair grows about a centimetre a month, the maximum length your hair can reach is set almost entirely by how long your anagen lasts. People whose hair "won't grow past the shoulders" usually have a short anagen, not a growth problem.

Catagen is the shutdown, and it takes about two to three weeks. Matrix division stops, the lower two-thirds of the follicle regresses upward by programmed cell death, and the dermal papilla is pulled up behind it along a thin epithelial strand until it rests just under the bulge. In the animation this is the phase where the whole bottom of the follicle visibly retracts — it is shown in slow motion because at true speed it would be a flicker.

Telogen is the rest, classically about three months. The hair is now a "club hair": keratinised at the base into a little white bulb, no longer growing, held in place mechanically. Exogen is the separate, independently regulated act of actually letting go — the club hair is released, and it is usually pushed out by the new anagen shaft already growing beneath it. That is why you often find a hair with a small white bulb on the end in the shower drain: white bulb means telogen, which means it finished its life normally. It is not a root, and pulling it out did not damage anything.

Now the important part. A scalp carries roughly 100,000 follicles, and at any moment about 85–90% of them are in anagen and about 10–15% are in telogen. Crucially, they are asynchronous — every follicle is at a different point in its own cycle, with no coordination between neighbours. Do the arithmetic on 100,000 follicles turning over on a roughly four-year cycle and you get somewhere in the region of 70–100 hairs finishing every single day. That is the whole origin of the famous "50 to 100 hairs a day" figure. It is not a tolerance or an allowance. It is the unavoidable output of an asynchronous population. Press play, watch the eight follicles, and notice how seldom two of them shed together.

One nuance the model shows deliberately: a club hair is often retained for a while after telogen proper is over, and some follicles then sit empty for a period before restarting — a phase called kenogen. This is why the telogen count on a plucked-hair test (10–15%) is higher than three months of a four-year cycle would predict on paper. Exogen has its own biology and is not simply "the end of telogen".

Miniaturization: pattern hair loss is shrinking, not dying

This is the single most useful idea on this page, and the one most people have never been told. In androgenetic alopecia — male- and female-pattern hair loss — the follicles are not dying. They are shrinking. Every time a susceptible follicle completes a cycle, it comes back a little smaller, with a shorter anagen, a shallower bulb and a thinner shaft. A terminal hair that was 70 or 80 micrometres across and grew for four years is gradually replaced, over cycle after cycle, by a hair 20 micrometres across that grows for a few months and barely clears the skin. That thin, short, usually unpigmented hair is a vellus hair — the same kind of hair that covers your forearm.

Turn on the Androgenetic scenario and watch the front three follicles. The bulb climbs higher in the dermis each cycle, the shaft gets thinner and shorter, and eventually the follicle spends long stretches empty. The mean shaft diameter readout falls from the low 80s of micrometres into the 40s and 50s. The terminal-to-vellus ratio falls with it, and this is exactly what a dermatopathologist measures on a scalp biopsy: a normal scalp runs around 7:1 terminal to vellus; a ratio below 4:1 is the accepted histological marker of pattern hair loss.

Why this matters so much practically: a shrunken follicle can be coaxed back, but a follicle that has finally scarred over and closed cannot. Treatment that starts while the follicles are miniaturized but alive is working with living tissue. Treatment that starts on a scalp that has been bare and shiny for fifteen years is trying to resurrect something that no longer exists. This is why dermatologists sound impatient about early treatment — it is not salesmanship, it is the geometry of the follicle. Turn 💊 Finasteride on early in the androgenetic run and then late, and compare what comes back.

DHT, 5α-reductase, and why the pattern is patterned

Testosterone is converted inside the follicle to a much more potent androgen, dihydrotestosterone (DHT), by the enzyme 5α-reductase (type 2 predominates in the hair follicle; type 1 is more a sebaceous-gland enzyme). DHT binds the androgen receptor in the dermal papilla, and in susceptible follicles that signal shortens anagen and shrinks the papilla, which shrinks everything the papilla instructs.

The strange, revealing part is that the same hormone does the opposite elsewhere. The same DHT that thins the crown thickens the beard. Androgens are not "bad for hair" — the response is written into each follicle's own dermal papilla, which carries a positional identity it acquired in the womb and never forgets.

That positional memory is why baldness has a shape. Frontal and vertex follicles are androgen-sensitive. The occipital and lateral fringe — the horseshoe at the back and sides — largely is not. In the animation the back three follicles keep cycling at full size no matter how long you run the androgenetic timeline, and that is not an artistic choice; it is the reason hair transplantation works at all. Move a fringe follicle to the bald crown and it keeps behaving like a fringe follicle: it stays terminal in its new home. This is the principle of donor dominance, described by Norman Orentreich in 1959, and it is why a transplant is a redistribution of a fixed resource rather than a cure.

Two natural experiments confirm the DHT link without any drug involved. Men with an inherited 5α-reductase type 2 deficiency — the population studied in the Dominican Republic by Imperato-McGinley and colleagues — do not develop male-pattern baldness, and they also have small prostates. And men castrated before puberty historically did not go bald. The hormone is genuinely necessary; the genes decide which follicles listen.

Telogen effluvium and the three-month delay that makes people blame the wrong month

Something happens to the whole body — a high fever, major surgery, a general anaesthetic, childbirth, a crash diet, a new severe illness, iron deficiency, a thyroid swing, sometimes a drug. A large block of follicles that were happily in anagen all get pushed into catagen at the same time. The asynchrony that normally protects you is broken. Those follicles then do exactly what follicles do: about three weeks of catagen, then their telogen rest, and only then do they let go.

So the shedding does not begin when the stressor happens. It begins roughly two to four months later — classically taught as about three — and it can go on for a couple of months after that. Press Telogen effluvium and watch the bottom timeline: the red stressor line lands, the trace stays flat inside the normal band for a long stretch, and only then does the shed rate climb to two, three or four hundred hairs a day. The gap is drawn and measured for you.

That delay is why telogen effluvium is so often blamed on the wrong thing. Someone starts shedding handfuls in March and blames the new shampoo they bought in February, when the real culprit was the influenza and 40°C fever in December. If you are working out what caused a shed, look three months back, not three days.

The other thing worth knowing is that acute telogen effluvium is self-limiting and the follicles are entirely undamaged — they went through a normal telogen, just all at once. Once the trigger is gone, they re-enter anagen together, and a few months later people notice a fringe of short new hairs standing up around the hairline. That fringe is the recovery, not a new problem. The visible density lags behind the biology by many months, because the new hair has to grow a centimetre a month from zero.

Alopecia areata: an immune attack on the bulb, not a dead follicle

Alopecia areata is a different mechanism altogether. The anagen hair bulb is normally an immune-privileged site — it deliberately keeps a low profile so the immune system leaves it alone. In alopecia areata that privilege collapses, and cytotoxic T cells (CD8+, NKG2D+) swarm around the bulb in the pattern classically described as a "swarm of bees". Anagen is aborted abruptly, and the shaft breaks off close to the skin, leaving the diagnostic exclamation-mark hair: a short stub that is narrow and pale at the scalp and wider and darker at the tip. Select Alopecia areata and you will see immune cells converge on the middle follicles and those hairs snap off into stubs.

Here is the hopeful part, and it is a real one: the follicle survives. The stem cells in the bulge are not the target, so the machinery for building a new hair remains. This is why regrowth is genuinely possible even after long patches, why regrown hair often comes back white before it re-pigments, and why the animation shows the patch recovering when the attack subsides. It is also the biological reason the JAK inhibitors (baricitinib, and later ritlecitinib) that have been approved for severe alopecia areata can work — they interrupt the interferon-gamma signalling that drives the attack, and the follicle is still there to respond. It is an autoimmune disease with a fluctuating course, so it can also relapse.

Nutrition that matters — and nutrition that is marketing

Hair is a non-essential tissue with an enormous metabolic appetite, so the body throttles it early when supplies are short. That makes a handful of nutritional problems genuinely worth checking — and makes almost everything else on a hair-supplement label irrelevant.

Iron. This is the one with real substance. Iron deficiency — even without anaemia — is associated with telogen effluvium, and it is common in menstruating women, in people with heavy periods, in blood donors, in coeliac disease and after bariatric surgery. Ferritin is the test that matters, not haemoglobin: haemoglobin is the last thing to fall. Many dermatologists want ferritin comfortably above 30 µg/L and some aim higher for hair specifically, though the evidence for the higher targets is genuinely mixed and honest clinicians say so. Ferritin is also an acute-phase reactant, so inflammation can make it look falsely reassuring.

Protein and total calories. Hair is essentially structural protein. Sustained very low protein intake or aggressive calorie restriction reliably triggers telogen effluvium, which is why rapid weight loss and crash diets show up in dermatology clinics about three months later.

Zinc. True zinc deficiency causes hair loss, and correcting it fixes it. Supplementing zinc when you are not deficient does nothing for hair — and sustained high-dose zinc induces copper deficiency, which causes its own problems including anaemia and neurological symptoms.

Thyroid disease. Both hypothyroidism and hyperthyroidism cause diffuse hair thinning, and it is easily missed because the other symptoms can be vague. A TSH is cheap and worth doing.

What is not supported: biotin for people who are not deficient (see the myths below), collagen powders as a hair treatment, "DHT-blocking" shampoos that sit on the scalp for ninety seconds, and the general category of multi-ingredient hair gummies. Saw palmetto is often marketed as a natural 5α-reductase inhibitor; there is some small-scale human evidence and a plausible mechanism, but it is a long way from the evidence base behind finasteride and should not be presented as equivalent. Topical rosemary oil has one reasonably-cited comparison against minoxidil that is frequently over-quoted online — it is interesting, not settled.

Minoxidil and finasteride: what they do, how long they take, and what stopping costs

Minoxidil was an oral blood-pressure drug that grew hair as a side effect. Topically it opens ATP-sensitive potassium channels and increases blood flow to the follicle, and its net effect is to prolong anagen and enlarge the follicle, so hairs grow for longer and come back thicker. Two details are worth knowing. First, minoxidil is a prodrug: it must be converted to minoxidil sulfate by a sulfotransferase (SULT1A1) in the follicle, and people with low follicular enzyme activity are the classic non-responders. Second, and this surprises everyone: minoxidil frequently causes a shed in the first two to eight weeks. That is not the drug failing. It is telogen hairs being evicted early because new anagen hairs are pushing up behind them — press 🥤 Minoxidil in the animation and you will see that spike on the timeline before the diameter readout starts to climb. Give it four to six months before judging it.

Finasteride inhibits 5α-reductase type 2 and lowers scalp DHT by roughly 60–70%. It attacks the cause rather than the symptom, and in randomised trials it both slows loss and produces modest regrowth in still-living follicles. It takes three to six months to show anything and up to a year for the full effect. Sexual side effects were reported by a low single-digit percentage of men in the pivotal trials, close to but above placebo; a minority of men report persistent symptoms after stopping (the disputed "post-finasteride syndrome"), and the honest position is that this is debated rather than settled. It must not be used by women who could become pregnant, because blocking DHT interferes with the development of the male fetal genitourinary tract. Dutasteride blocks both enzyme types and is more potent; it is used off-label for hair in many countries.

And the part nobody wants to hear: both drugs are maintenance, not cure. They do not change the follicle's genetic programme; they suppress the signal or extend the phase for as long as they are present. Stop, and within roughly three to twelve months the follicles resume the trajectory they were on, and the hair you kept is lost fairly quickly. In the animation, turn a treatment off and keep watching — the miniaturization resumes from where it paused. Deciding to treat pattern hair loss is deciding to treat it indefinitely, and that is a reasonable thing to decline. It is just better to decline it knowingly.

Myths, corrected

Hats, frequent washing and shampoo do not cause baldness. Hair grows from a follicle that is fed by a capillary loop from below; nothing that sits on the outside of the scalp starves it, and a hat does not "suffocate" a follicle. Washing does not cause hair loss either — it reveals it. Hairs that were released days ago stay tangled in the surrounding hair until the mechanical action of washing frees them, so people who wash less see more hair per wash and conclude that washing did it. Washing less does not reduce your shedding; it just batches it. (Extremely tight braids, buns and extensions genuinely can cause hair loss — traction alopecia — but that is sustained mechanical pull, not a hat.)

"You get it from your mother's father" is a folk simplification. There is a grain of truth in it: the strongest single genetic signal for male-pattern baldness sits at the androgen receptor locus on the X chromosome, which men inherit from their mother. But large genome-wide studies have found hundreds of associated regions scattered across the genome, and paternal baldness is also clearly predictive. The inheritance is polygenic and comes from both sides. Looking at your father's head is not a wasted glance.

Biotin does nothing for hair unless you are actually biotin-deficient — and it wrecks your lab tests. Biotin deficiency is genuinely rare (it happens in some inherited metabolic disorders, with prolonged raw-egg-white consumption, and occasionally with long-term anticonvulsants). If you are not deficient, supplementing does not thicken hair, and the marketing doses are wildly beyond any nutritional requirement. Meanwhile there is a real clinical hazard: high-dose biotin interferes with the streptavidin-biotin chemistry used in many immunoassays. It can produce falsely low troponin — which can mask a heart attack in an emergency department — and a falsely low TSH with a falsely high free T4 that mimics Graves' disease and has led to unnecessary treatment. The FDA has issued a safety communication on exactly this. If you take a hair supplement, tell whoever orders your blood tests, and stop it for a few days beforehand if you can.

Plucking one grey hair does not make three grow back. One follicle produces one hair. Plucking removes the shaft, the follicle enters a new cycle, and it produces one hair again — grey, because the melanocyte stem-cell pool that pigmented it is depleted and that does not reverse from being tweezered. What repeated plucking can do is inflame and eventually scar the follicle, so it is a bad habit for a reason that has nothing to do with multiplication.

Also worth retiring: shedding hair in the shower drain is not evidence of a problem by itself (count what you lose in a day, not what you see at once); "clogged follicles" are not the cause of pattern baldness; and scalp massage, while pleasant, has no convincing evidence behind it as a treatment.

When hair loss deserves a workup

Most shedding is either normal turnover being noticed for the first time, pattern loss, or a telogen effluvium that will resolve. A few situations genuinely need a doctor, and preferably a dermatologist:

1. Any sign of scarring. Smooth shiny skin where you can no longer see the tiny follicular openings, redness or scale around the follicles, burning, itching or tenderness, or a hairline that is receding as a clean band with loss of the eyebrows. Scarring alopecias — lichen planopilaris, frontal fibrosing alopecia, central centrifugal cicatricial alopecia — destroy the follicle permanently, and the window to save the remaining ones is short. This is the one to act on quickly.
2. Discrete round or oval bald patches appearing over weeks, especially with exclamation-mark hairs or pitted nails — that pattern suggests alopecia areata.
3. Sudden diffuse loss without an obvious trigger three months earlier, or a shed that keeps going beyond about six months (chronic telogen effluvium behaves differently from the acute kind).
4. Symptoms pointing at a systemic cause: heavy periods, fatigue, restless legs or pica (iron); cold intolerance, weight change, constipation or palpitations (thyroid); a new drug started a few months before the shed; recent rapid weight loss.
5. Hair loss in a woman with acne, irregular periods or new facial hair — worth investigating for an androgen excess such as PCOS rather than assuming pattern loss.

A sensible first round of tests is small: TSH, a full blood count and ferritin, with vitamin D and further hormonal tests if the story suggests them. A biopsy is reserved for suspected scarring alopecia or a confusing picture. And it is entirely reasonable to ask for a photograph and a part-width measurement at the first visit so that in six months there is something objective to compare against — because with a tissue that grows a centimetre a month, memory is a terrible instrument.

↑ Back to the animation

Connections

• All Interactive Visualizations
• Alopecia
• Hair Loss (symptom)
• Dermatology
• Iron
• Zinc
• Vitamin B7 (Biotin)
• Saw Palmetto
• Rosemary
• Testosterone & Male Puberty (animation)
• Iron Absorption & Hepcidin (animation)
• Thyroid Hormone Axis (animation)
• Autoimmunity (animation)
• Stem Cells & Differentiation (animation)