Black Girls Thrive Daily

Hair Loss at the Edges in Black Women

By Last reviewed 8 September 2026
Cover card reading: hair loss at the edges, scarring or not scarring.
The short answer

Thinning edges are not one condition. Traction alopecia, central centrifugal cicatricial alopecia, thyroid disease and hormonal causes look similar in a mirror and behave nothing alike. The distinction that matters most is whether the loss scars. Scarring destroys the follicle permanently, so an unexplained thinning edge is a reason to book a dermatology assessment early, not to wait it out.

Most of what is published about thinning edges treats the hairline as one problem with one fix. It is not. Traction alopecia and central centrifugal cicatricial alopecia (CCCA) have different locations, different mechanisms and different outcomes, and the two statistics quoted about them most often are both wrong. The CCCA number that circulates is a multiplication of two figures from a 326-woman convenience sample in Cleveland, published in Archives of Dermatology. The traction alopecia number is South African, from Khumalo et al. in JAAD. Neither is a US population prevalence.

What actually causes thinning at the edges, and how do you tell the causes apart?

Four broad categories account for most of it: tension-driven loss at the margins, inflammatory scarring loss that starts centrally and spreads outward, thyroid disease, and hormonal or androgen-related patterns. They differ in where they appear first, whether the follicle survives, and what else is going on with the scalp. Only an examination separates them reliably.

The table below is a differential, not a diagnostic tool. Where this site could not source a US figure, the cell says so rather than borrowing one. That is deliberate, and the reasoning behind it is set out in how we research: a number that cannot be traced to a primary document is cut rather than estimated.

Differential for thinning at the hairline and edges. Figures, where they exist, are from Kyei et al., Archives of Dermatology, 2011, and Khumalo et al., JAAD, 2008.
Cause Where it typically shows up Does it scar? Usually confused with
Traction alopecia The margins: frontal hairline, temples, the edges. Follows where sustained pulling has been applied. Not at the outset. No sourced US figure exists for how often it progresses to permanent loss. CCCA, and a receding hairline assumed to be genetic
Central centrifugal cicatricial alopecia (CCCA) The crown and central scalp first, spreading outward. Kyei et al. graded central scalp changes, not the hairline. Yes. Cicatricial means scarring, and that is what the condition is named for. Traction alopecia, and breakage blamed on styling
Thyroid disease Diffuse rather than confined to the edges. Requires bloodwork, not a scalp photograph. No Everything, because it is the first thing people search. Thyroid autoimmunity is less common in Black Americans, not more: TPO antibodies were positive in 4.5 percent of Black versus 12.3 percent of White participants in NHANES III.
Hormonal and androgen-related causes Pattern loss, often at the part and crown. Diagnosed clinically and with labs. No CCCA. This site could not source a US prevalence for hormonal hair loss in Black women specifically.

Why does it matter whether the hair loss scars?

Because it decides whether anything can come back. This is the single most important distinction on this page, and it is the reason the guidance here ends in an appointment rather than a routine. Everything else is detail; this is the part that is not recoverable once it has happened.

Non-scarring hair loss can regrow, because the follicle is still there. Scarring hair loss destroys the follicle, and hair does not grow back from a follicle that no longer exists.

That sentence changes the calculation on waiting. With a non-scarring cause, time spent identifying the problem costs relatively little. With a scarring cause, every month the inflammation continues is follicles lost permanently, and no later intervention recovers them. CCCA is scarring by definition: the “cicatricial” in its name is the medical word for scar. That is why an edge that is thinning for reasons nobody has examined is a different kind of problem from an edge that a dermatologist has looked at and identified.

Nobody can tell from a photograph, and neither can you from a mirror. It takes an examination of the scalp, sometimes with a biopsy. That is the limit of what looking can establish.

Is it true that CCCA affects about 15 percent of Black women?

No, and the figure is not from any study. It is produced by multiplying two numbers from a single paper: Kyei et al. found that 28 percent of a convenience sample of 326 African American women in Cleveland had central scalp alopecia of grade 2 or higher, and 59 percent of that subgroup had grade 3 or higher changes read as consistent with CCCA. Twenty-eight percent of 59 percent is roughly 15 percent. Nobody published that number as a prevalence.

The distinction matters for two reasons. First, a convenience sample is not a population sample, and this one was recruited at churches and health fairs in one city. Second, the paper’s prevalence claim was formally challenged in the same journal. The study measured what it measured, which is scalp grading in a specific recruited group. It did not establish how common CCCA is among Black women in the United States.

The honest statement is that CCCA prevalence is unestablished. That is less satisfying than a percentage, and it is what the evidence supports.

Do a third of Black women in the US have traction alopecia?

No. That figure is South African and it is being reported as if it were American. Khumalo et al., working in Cape Town, found traction alopecia in 31.7 percent of women aged 18 to 86 and 17.1 percent of schoolgirls aged 6 to 21. Those are real, carefully collected numbers about a South African population.

There is no equivalent US population figure. None was found for this page, and one is not quietly substituted from another study. Transplanting a prevalence figure from one country to another is not a small liberty. It produces an invented statistic with a real citation attached, which is harder to catch than a fabrication.

What can be said is that traction alopecia is well described in the dermatology literature and that its mechanism, sustained pulling on the follicle, is not in dispute. How common it is among Black women in the United States is simply not known from published population data.

What does the evidence on hair straighteners actually show?

One large NIH cohort found an association between straightener use and uterine cancer. It found no causal chemical, collected no brand or ingredient data, and found no difference between racial groups. The word to use is associated. The word never to use is caused.

The Sister Study, published in JNCI, followed 33,947 women aged 35 to 74 for about 10.9 years and recorded 378 uterine cancers. Ever-use in the prior 12 months carried a hazard ratio of 1.80 (95 percent CI 1.12 to 2.88). Frequent use, more than four times a year, carried a hazard ratio of 2.55 (95 percent CI 1.46 to 4.45, P-trend .002). In absolute terms, NIH reported that 1.64 percent of never-users would be expected to develop uterine cancer by age 70, against 4.05 percent of frequent users. The number needed to harm was 85 for ever-use and 42 for frequent use. Hair dyes, bleach, highlights and perms were not associated with uterine cancer in the same cohort.

Two limits carry as much weight as the hazard ratios. The study identified no chemical responsible, so there is nothing specific to avoid on a label. And a separate NIEHS occupational analysis found no association between hairdressing work and uterine cancer, with a hazard ratio of 1.04 (95 percent CI 0.60 to 1.77).

The straighteners-and-fibroids claim is a different matter and it is not supported. In a NIEHS analysis of 4,162 Black women, use at ages 10 to 13 gave an odds ratio of 1.15 (0.96 to 1.36), frequent use 1.18 (0.99 to 1.42), and incident fibroids a hazard ratio of 1.14 (0.81 to 1.63). Every confidence interval crosses 1, and the association was similar in White women despite straightener use of 5 percent or less. If you are working out what does raise fibroid risk, the evidence is set out in the guide to uterine fibroids in Black women.

Has the FDA banned formaldehyde in hair straighteners?

No. The FDA has not banned formaldehyde in hair-smoothing products, and anyone who tells you otherwise is wrong. The rulemaking exists on paper and has not produced a rule.

The action is RIN 0910-AI83 in the OMB Unified Agenda, and when this page was last checked it was still listed at Proposed Rule Stage. No proposed rule has ever been published. There is no Federal Register citation for it. The projected date for a notice of proposed rulemaking has slipped six times.

The accurate wording is that a formaldehyde restriction is proposed for future rulemaking and has not been issued. This matters beyond pedantry. Anyone who believes a ban is in force will assume products on shelves were reformulated to comply with something, and nothing has been enforced because nothing has been issued.

Could the flaking and itching be part of what is happening?

It can be a separate problem sitting on top of the same scalp. Seborrheic dermatitis causes flaking, redness and irritation, and it is common. Inflammation and hair loss are not the same thing, and treating one as evidence of the other sends people down the wrong path.

A JAMA Dermatology systematic review of 121 studies covering 1,260,163 people put pooled seborrheic dermatitis prevalence at 4.38 percent (95 percent CI 3.58 to 5.17), and 5.64 percent among adults. In US ambulatory data, analysis of CDC NAMCS visits found seborrheic dermatitis among the top five dermatology-visit diagnoses for African American patients, but not for White patients. Prevalence by race was not established by that analysis, and this site could not source one.

Flaking on its own is a scalp question, not a hair-loss question. The detail on what separates ordinary flaking from seborrheic dermatitis on textured hair is in the guide to dandruff and seborrheic dermatitis on natural hair. Bring both up at the same appointment if both are happening.

What about keloids along the scalp and hairline?

Keloids come up whenever a scalp or skin procedure is discussed, including a biopsy, and the question deserves a straight answer rather than avoidance. Black patients are substantially over-represented among people with confirmed keloids, and the keloids recorded are more often severe and more often multiple.

In a US electronic health record cohort of 24,453 adults with confirmed keloids, 18.8 percent were Black, against 8.8 percent of the matched general comparator group and 4.5 percent of the dermatology comparator. Major severity was recorded in 56.4 percent of the cohort. Black patients more often had more than one keloid than White patients, 32.5 percent versus 20.5 percent.

What that cohort cannot tell you is how common keloids are in the US population overall. A true population prevalence could not be sourced for this page. The practical point is that keloid history is worth raising with a dermatologist before any procedure on the scalp, as a factor in the conversation rather than a reason to skip the appointment.

How long is too long to wait before seeing a dermatologist?

Here is the judgement call, stated plainly: about eight weeks. If an edge has been visibly thinner for roughly two months and is not filling back in, that is the point to book an assessment. This is an editorial position on urgency, not a guideline threshold and not a diagnosis, and no published figure sets this number.

The reasoning is the scarring distinction. Most people wait far longer, often a year or more, because hair loss reads as cosmetic and appointments are inconvenient. That instinct is calibrated for a non-scarring cause, where waiting mostly costs time. It is badly calibrated for a scarring cause, where waiting costs follicles that do not come back. Because you cannot tell which one you have by looking, the sensible default is the one that is safe if it turns out to be the scarring kind.

An assessment is not a commitment to treatment. It establishes which category the loss falls into, which is the one thing no amount of reading, product-switching or waiting will settle.

Worth an appointment sooner rather than later

Book a dermatology assessment rather than waiting if any of these apply:

  1. An edge or patch has been visibly thinner for about eight weeks with no sign of filling back in.
  2. The scalp is tender, burning, itching or sore in the area that is thinning.
  3. The affected skin looks smooth or shiny where you would expect to see follicle openings.
  4. The thinning started at the crown or central scalp rather than at the hairline.
  5. Hair is coming out diffusely across the whole head rather than in one region, which is a bloodwork question as much as a scalp one.

Ask the dermatologist directly whether what they are seeing is scarring or non-scarring, and what would establish it. That is the question the whole differential turns on. If you are already booking, it is a reasonable moment to have blood pressure checked too, given what the numbers on high blood pressure in Black women look like; the USPSTF considers annual screening reasonable for Black adults as an increased-risk group.

Common questions about thinning edges

Can traction alopecia become permanent?

Tension-driven loss is not scarring at the outset, which is why early assessment matters. No sourced US figure exists for how often it progresses to permanent loss, so anyone quoting a percentage for that is not working from published US population data.

How do I know if it is CCCA or traction alopecia?

Location is the first clue: CCCA typically starts at the crown and central scalp, traction alopecia at the margins where pulling occurs. That is not enough to decide. A dermatologist examines the scalp and may biopsy, because the two are routinely mistaken for each other.

Should I stop using relaxers because of the cancer study?

That is a decision to make with a clinician, not from a webpage. The Sister Study found an association, not a cause, identified no responsible chemical, and put absolute risk by age 70 at 1.64 percent for never-users versus 4.05 percent for frequent users.

Is thyroid disease a likely explanation for thinning edges?

It is worth ruling out with bloodwork, but thyroid loss is usually diffuse rather than confined to the edges. Thyroid autoimmunity is also less common in Black Americans: TPO antibodies were positive in 4.5 percent versus 12.3 percent of White participants in NHANES III.

Does a scalp biopsy risk a keloid?

Raise it with the dermatologist before the procedure. Keloids are over-represented among Black patients, 18.8 percent of one US cohort of 24,453 adults with confirmed keloids, and it is a factor in planning rather than a reason to avoid assessment.

Sources

Medical disclaimer. This article is health information, not medical advice. It cannot account for your history, your medications or your test results. Talk to a licensed clinician before you change anything about your treatment. Read the full disclaimer.