For most of human history grey hair has been treated as a one-way change. People notice a few silver strands, then more, and almost never see them turn back to their original colour. In everyday experience, going grey is something that happens and then stays.

Biology offered a clear explanation for this. Inside each hair follicle there is a small reservoir of melanocyte stem cells. These cells act like a pigment seed bank. They replenish the pigment-making cells that load melanin into the growing hair shaft and give it colour.

With age and repeated stress, this seed bank is thought to shrink. Fewer stem cells are left to replace worn-out pigment cells. At some point the reservoir is so depleted that the follicle can no longer supply enough melanin. The hair that grows out from that follicle emerges pale and then fully grey or white.

In this classic model, once the stem-cell bank is effectively empty, the process is considered irreversible. That follicle has crossed a line. It is expected to keep producing grey hair for the rest of its life, which is why greying has been framed as a one-way road rather than something the body can meaningfully reverse.

In the last few years, however, new work has started to challenge this simple picture. In particular, a 2021 study led by Ayelet Rosenberg at Columbia suggests a more nuanced scheme of hair greying, where at least some follicles retain the capacity to switch pigment production off and on again under certain conditions.

The Columbia team approached grey hair with a simple but powerful idea: treat each strand as a physical timeline of recent life [Ref 1]. Scalp hair grows at a fairly steady pace, roughly 1–1.3 centimetres per month, so every point along a single hair corresponds to a particular week in the past. The tip reflects months ago, the root reflects what is happening right now. Like growth rings in a tree trunk, length becomes a physical record of time.

Instead of just calling a hair “grey” or “not grey,” the researchers scanned the shaft under high-resolution imaging and converted its colour into numbers along the entire length. This produced detailed “hair pigmentation patterns” that show where the strand is darker, lighter or fully depigmented, millimetre by millimetre.

With this method, a single hair becomes a high-resolution record of how pigmentation in that follicle changes over time. The team can see when colour starts to fade, when it stabilises and, unexpectedly, when it comes back, all mapped onto an actual time axis rather than just a before-and-after snapshot.

This level of detail matters because greying is not a centralised process. Each follicle behaves autonomously, and transitions from one phase to the next happen independently in each follicle [Ref 2]. High-resolution mapping along single hairs is what makes it possible to see those individual follicle decisions over time.

When the researchers applied this new mapping, they expected to document the usual one-way march from dark to grey. Instead they found something far stranger. In several volunteers, some strands were “bi-colour” or even “multi-phase,” with sharp transitions in pigmentation along the same hair.

In a few cases the pattern was simple: a dark segment near the tip faded into a pale or fully grey segment closer to the root and then stayed grey, which fits the classic story. But other hairs showed the opposite twist. They started dark, turned grey for a stretch, and then became darker again toward the root.

To understand how life events might link to these colour shifts, the researchers asked volunteers to reconstruct their stress levels over the previous months. People went through calendars and noted periods of intense pressure at work, relationship problems or major life changes. This produced a personal “stress timeline” for each participant.

Because hair length maps onto time, the team could then line up these stress timelines with the pigmentation patterns along each strand. In several striking cases the match was very tight. A hair segment turned grey during a clearly stressful period, then returned to darker pigment in the weeks after the stress eased. In other words, the same hair recorded both a stress-linked loss of colour and a recovery when conditions improved.

The group also looked inside dark and grey hairs using proteomics. If the classic story of simple depletion were fully correct, one might expect grey hairs to be empty or inert. Instead they were packed with proteins, and not at random. Grey hairs showed higher levels of proteins linked to mitochondria, energy metabolism and stress responses, along with changes in structural and pigment handling proteins.

This points to a different picture. Many grey hairs are not dead, pigmentless leftovers. They are remodelled tissues that have shifted into a high stress, altered-metabolism state [Ref 1]. In that state pigment production is dialled down or shut off. Under the right conditions some follicles can move out of that state again and restart pigment, which fits with the rare but real reversals seen in the hair timelines.

To make sense of these patterns, the authors propose a simple threshold model for each follicle. Every hair is imagined as carrying an internal “load” that reflects both underlying ageing processes and short-term stress. This load is not fixed. It creeps up over the years as damage and wear accumulate, then jumps during stressful periods, and can ease back a little when conditions improve.

In this picture, visible greying happens when that combined ageing–stress load crosses a critical threshold. Once a follicle is pushed over the line, pigment production drops and the hair segment that grows during that time comes out grey or white.

The key nuance is what happens to follicles that sit close to the threshold rather than far beyond it. If a hair is just over the line, reducing stress or improving the local environment can pull the load back down. In those borderline follicles the pigment machinery can restart, which shows up as a grey segment followed by a darker segment on the same strand.

This framework explains several observations at once. Reversals are rare because many follicles are either comfortably below the threshold and still pigmented, or far above it and stably grey. They are more likely in younger or early-greying individuals, where many follicles hover near the tipping point, and they usually affect only a small fraction of hairs. However, recent work suggests that greying remains reversible while melanocyte stem cells and follicular melanocytes are still present [Ref 3]. Follicles that have not completely lost these cells may regain pigment if their microenvironment and energy balance recover.