Sign in

The Perplexing Topic of Hair 'Type': How Do We Classify Hair?

Contact Author Trefor A. Evans, Ph.D., TRI-Princeton, Princeton, NJ USA
Fill out my online form.

We have an intimate and personal relationship with our hair. When asked to describe it, many will offer detailed accounts of texture, thickness, condition, manageability and other descriptors. Not surprisingly, then, a popular and powerful marketing proposition involves positioning hair care products as being custom-designed to meet the specific needs of an individual’s hair “type.” Yet, this begs the question, “How do we classify hair?”

From a marketing perspective, it begins with the consideration of consumer language. Here, we are familiar with terms such as dry, oily, fine, limp, thick, coarse, damaged and others. But as the recurring theme in this series has explained, consumer language is nebulous and imprecise—and it has been shown how such self-diagnoses frequently do not equate to technical knowledge. A more scientific discourse is required.

So just how different is hair from the heads of various individuals? Is all hair essentially the same stuff extruded out onto the scalp in different sizes, shapes and colors? Or alternatively, are less-obvious or hidden factors present?

Basic Hair Architecture

Excerpt Only This is a shortened version or summary of the article you requested. To view the complete article, please log in or create an account. Registration is Free!

Discussions on hair’s basic architecture begin with its three major structural regions. The internal bulk is termed the cortex, although there may also be a hollow-like center, called the medulla. The delicate internal cortex is protected by a hard, resistant, scaly outer layer collectively termed the cuticle.

Thick hair fibers possess a pronounced medulla, while this component can be discontinuous or even absent in fine hair (see Figure 1). So, thin and thick hair types are certainly not vastly different entities; outside of the presence (or absence) of a medulla, they can be considered structurally and chemically akin.

The inherent curliness or straightness of individual hair strands is thought to be programmed by the hair follicle. Scalp biopsies have shown how curly hair sprouts from curved follicles, while straighter hair emerges from more vertical geometry.1 Furthermore, straight hairs grow from a symmetrical bulb at the base of the follicle, while curly hair has a distinctly asymmetrical bulb.1 These findings similarly suggest straight and curly hairs are made of the same stuff but shaped differently during growth.

Hair color is dictated by the presence of melanin pigment granules located in the cortex region. Higher concentrations produce darker hair color yet even in the darkest hair, it is still a minor component of the overall fiber structure. Therefore, despite a radically different appearance, light and dark hairs are not fundamentally different in makeup.

The scientific literature contains a number of studies where researchers have compared and contrasted compositional building blocks of different hair types. While there is some discrepancy in results, the general consensus suggests there are no meaningful differences in amino acid profiles. Thus, a strong argument seemingly can be made that hair is hair, is hair.

The Impact of Fiber Size, Shape and Color

It is positively anticlimactic to suggest the bewildering variations in human hair appearance and behavior can be rationalized merely based on size, shape and color. These properties seem too mundane—surely something more cryptic must be involved. Yet it is worth contemplating the myriad of ways the size and shape of fibers can impact the appearance, properties and maintenance of hair on a head—factors which then become the foundation of an individual’s daily hair care habits and practices.

First, thicker fibers weigh more. This greatly impacts the manner in which hair hangs on the head and moves. Fiber curvature also impacts hair motion; more fiber-to-fiber contact points occur with increasing curliness. In concert, these two characteristics may create manageability issues, where grooming forces increase, style options become restricted and frizz can become a problem.

The extremely curly/kinky nature of African hair creates notably trying conditions, too. This hair type is especially fragile, likely due in part to the higher grooming forces it requires, which contribute to an increased incidence of breakage. It also has been suggested this hair type is weakened by the presence of flaws that arise due to points of high stress along the especially kinky fibers.2

I suspect the highly elliptical cross-section of kinky/curly hair type to be a contributing factor. During manipulation, fibers must bend and twist to slide past each other but the highly elliptical cross-section of African hair limits fibers such that they are not free to move equally in all directions. A sizable bending differential exists about the major and minor fiber axes this could induce internal stresses within the fiber, in turn initiating structural flaws.

At the other extreme in fiber dimension, fine hair may dangle limply from the head, producing its own set of problems. Most notably, those “cursed” with such hair struggle to create styles with “body” and “volume.”

Fiber dimensions have a strong impact on many of hair’s physical properties. Fine hair possesses lower tensile strength3 and exhibits a higher tendency for breakage.4 The bending stiffness of fibers is also strongly dimension-dependent.5

Another banal but considerable contributor to overall hair perception is fiber density on the scalp. An individual may perceive their hair as being “thick” in texture; this perception is created by high fiber concentration rather than the dimensions of individual strands.

The vast spectrum of appearance, properties and behavior in hair may instinctively suggest complex and intricate contributors are involved, but as the above discussion shows, this is not necessarily the case.

Straight hairs grow from a symmetrical bulb at the base of the follicle, while curly hair has a distinctly asymmetrical bulb.

Hair Reactivity: Beyond Size and Shape

Stylists recognize how hair on the heads of different individuals can respond very differently to so-called chemical treatments. For example, the application of a permanent wave to one person’s hair may produce the desired result, while the exact same product and procedure may yield a wholly ineffectual outcome on the next person.

In the lab, scientists frequently use chemically damaged hair as a test substrate to improve the sensitivity of the testing techniques employed.6 Yet, a standard treatment protocol will not reproducibly yield the exact same outcome each time due to lot-to-lot variability in the reactivity of supplied hair. Why?

Attempts to understand this unknown begin with measurements to provide characterization. A previous article in this series described the Single Fiber Tensile Kinetic (SFTK) approach to measure the reaction rates of hair with perm solutions.7 Figure 2 shows results from such experiments, where a common perm solution was used in conjunction with hair from nine different donors.

Sizable differences in reactivity are observed. Poor perm performance may be a consequence of slow reaction rates, leading to insufficient transformation in the allotted time. That said, what is the reason for differing reaction rates? Perhaps something more mysterious lurks within the topic of hair type.

Secrets from the Wool Industry

The genesis of much of our knowledge about hair is derived from the wool industry. For example, the application of basic dyes to kinky Marino wool is known to produce differential staining of the fibers;8 i.e., the dye is adsorbed by one side of the fiber but not the other. In cross-section, this produces an appearance somewhat reminiscent of the yinyang symbol.

This finding has been interpreted as the asymmetrical distribution of two different types of cortical cells within the wool, traditionally termed ortho and para. The un-dyed para portion always lies on the inside of the fiber curl. This observation has produced an alternate hypothesis for the innate shape of straight and curly keratin fibers.

Electron microscopy studies show how these cells most notably differ in the packing of their microfibrils (intermediate filaments) within the cortical cells.9 Yet, chemical differences also apparently exist. Ortho cells readily adsorb basic dyes while para cells adsorb acidic dyes. Ortho cells swell substantially in basic solutions while para cells do not. Para cells have a higher cystine content—seemingly due to relatively higher matrix (keratin associated protein, KIF) content.

Comparable microscopy studies on hair were first performed by Swift10 in the mid 1970s. A similar asymmetrical arrangement of ortho and para cells was observed in highly curly African hair, while only para cells were seen in straight Japanese hair. Researchers at L’Oréal’s Paris laboratories confirmed these results in African hair but suggested a symmetrical annular arrangement of cells in straight hair.11 Bryson and coworkers12 have made similar observations but suggest the presence of four different cell types.

This is a captivating topic that is only now beginning to receive significant attention in the hair field. The concept provides a foundation from which hypotheses about differing internal structure and properties in hair may be rationalized. However, these secrets are not easily given up. Electron microscopy work in this area is hindered by tricky sample preparation and subjective interpretation. A more straightforward measurement approach for probing the internal structure of hair is needed.

Insights from Tensile Studies

The tensile properties of hair are a consequence of its internal structure, thus mechanical testing serves as a probe into this inner organization. For example, TRI-Princeton performed tensile experiments to evaluate the Young’s modulus of hair as a function of relative humidity. This approach was a modification of the conventional stress-strain experiment described in a previous article3 in that fibers were only stretched to 5% extension, which allows multiple experiments to be performed at different relative humidities using the exact same fibers. The initial hair properties were then “reset” after each experiment by exposing swatches to 90% RH for 4 hr.

The systematic nature of this testing, in combination with improved reproducibility by with using the exact same fibers, gives high confidence in results. We began noticing differing modulus vs. relative humidity curves for hair from different sources; Figure 3 shows examples. These results are all from single-source hair procured from donors, as opposed the blended hair traditionally provided by suppliers. It should be emphasized that all hair used was virgin in nature, i.e., without chemical treatments; FT-IR was used to evaluate cysteic acid levels and confirm the hair’s condition via oxidative damage.

Significant, consistent and sometimes considerable differences were observed between hair samples. These differences manifest more prominently under low humidity conditions. Normally, we test on blended hair and tensile experiments are performed in a wet state or at medium humidity (50-60%). This illustrates how such observations have previously gone unnoticed. While we still do not understand the reason for this behavior, and complimentary testing is needed, the results imply differences in internal hair structure.

Most hair is modified from its natural state, considering 70-75% of women use chemical treatments, plus high-temperature styling.

Attempts at Hair Classification

Historically, attempts to classify hair began with race. This position has received some criticism yet it is understandable why such an approach was used. Prominent differences are present in terms of appearance and properties between, for example, Caucasian, Asian and African hair.

In general, Asian hair possesses a larger diameter, is heavily pigmented and is virtually spherical in cross-section. Caucasian hair is not as thick, is more variably pigmented and distinctly elliptical in cross-section. African hair can possess variability in thickness, but its appearance is dominated by an especially curly/kinky conformation. African hair is almost always heavily pigmented and can be highly elliptical in cross-section.

Today, sizable emerging markets in South America and India are drawing increased attention to such hair “types.” And while there is an understandable desire for a demarcation from race, unquestionably, these geographical descriptors immediately convey images (albeit stereotypical ones) of hair with very different appearances and properties.

Researchers at L’Oréal took up the task of trying to classify hair irrespective of geography.13 They proposed a segmentation system based on hair shape, specifically curvature and kinkiness. This system consists of eight groupings ranging from straight and mostly straight (Type 1) to extremely kinky (Type 8). The system provides a convenient means for visualizing and describing hair with different degrees of curl/curvature/kink, which is useful when dealing with curly/kinky hair. Yet the hair substrate is too complex to be effectively described by a single term alone. For example, thick, straight and heavily pigmented Asian hair; and fine, straight, blonde Caucasian hair would both be classified as Type 1—even though they are vastly different in appearance and properties.

I propose that additional descriptors be added to the L’Oréal groupings to provide further distinction. For example, a second term may be used to designate fiber thickness. Arbitrarily, labels could include thick (T) for a diameter greater than 80 µm; medium (M) for a diameter between 65 µm and 80 µm; and fine (F) for diameters less than 65 µm.

An indicator for pigmentation level would be useful, too. To avoid instituting new terms, I suggest using the lightness (L) value from the conventional L*a*b* color measurement.14 To recap, this scale spans from 0 to 100, with higher numbers indicating lighter coloration. For aesthetic purposes and to avoid numeric confusion, these values would be subscripted. Thus, by this proposed system, thick, heavily-pigmented Asian hair may be expressed by the descriptor 1T20; while fine, blonde Caucasian hair might be suitably distinguished as 1F70.


Hair is a remarkable bio-composite material with a highly complex structure. The bewildering variability in appearance, properties and behavior may instinctively suggest inherent differences in this structure as a function of hair type. Yet this article makes the case that a majority of these properties can be rationalized simply by the size, shape and color of fibers.

Despite such an anticlimactic positioning, this viewpoint nonetheless provides technical rationalization as to why different products are needed for different hair types. For example, those with especially thick, curly hair desire assistance during detangling, brushing and general grooming—and the lubrication provided by conventional conditioner products helps greatly.

This general benefit is desirable to most but not necessarily to the same degree. Fine, thin hair may be weighed down by heavy surface deposits, so a lighter treatment is more appropriate. Therefore, product lines within hair brands usually feature a range of conditioner variants consisting of similar base formulas, but differing in conditioning dosage.

Fiber size and shape may also lead to a variety of specific maintenance and style-related issues, which dictate subsequent habits and practices. On occasion, rather drastic approaches may be employed. For example, thick, curly hair may more frequently be subjected to heat straightening or even chemical relaxation. Or, fine hair may be treated chemically via perms in attempts to impart “volume” and “body.” Such treatments are harbingers of hair damage, which in turn produce their own maintenance issues. Therefore, appropriate product-based responsive actions are needed to retard and mask sensorial negatives.

In short, most hair-related properties and issues can be rationalized in terms of fiber size and shape, and/or the consequences of dealing with especially trying conditions that associate with size and shape. The Achilles Heel of this thesis, however, is the varying reactivity of different hair types with chemical treatments. This behavior implies potential differences in internal structure of hair—and attention has been drawn to ongoing research pertaining to the presence and distribution of different cortical cells.

Regarding hair classification, an idea is proposed here for expanding an earlier designation to provide additional detail. That said, a hugely important variable still missing from this descriptor is some indication of hair health.

Outside of size, shape and color, and their related properties, it is difficult to find any fundamental fiber parameter that changes significantly with hair type. Yet sizable differences become readily apparent after certain cosmetic treatments or practices that are well-known to compromise hair structure. Mechanical properties change; friction increases; light reflection decreases. Despite these consequences, the use of such treatments is highly prevalent, with the end result apparently justifying the means.

Industry folklore suggests around 70-75% of all women utilize some chemical treatment on their hair. Add to this the use of high-temperature straightening and curling irons and it becomes apparent that most hair will be in a considerably modified form from its freshly grown state.

The bottom line is, even if minor innate differences exist between hair types, they seemingly would be overwhelmed and swamped out by the variety of everyday habits and practices, which can considerably damage and greatly impact hair properties; in fact, the next article in this series will focus on this topic of hair damage.


  1. S Thibaut, O Gaillard, P Bouhanna, DW Cannell and BA Bernard, Human hair shape is programmed from the bulb, Br J Dermatol 152(4) 632-638 (2005)
  2. Y Kameth, S Hornby and HD Weigmann, Mechanical and fractographic behavior of Negroid hair, J Cosmet Sci 35 21-43 (1984)
  3. TA Evans, Measuring hair strength, part 1: Stress-strain curves, Cosm & Toil 128(8) 590-594 (Aug 2013)
  4. TA Evans, Hair breakage, ch 8 in TA Evans, R Wickett, eds, Practical Modern Hair Science, Allured Business Media, Carol Stream, IL USA (2012.
  5. FJ Wortmann, G Wortmann, HM Hake and W Eisfeld, Analysis of the torsional storage modulus of human hair and its relation to hair morphology and cosmetic processing, J Cosm Sci 65 59-68 (2014)
  6. www.cosmeticsandtoiletries.com/testing/efficacyclaims/Hair-as-a-Test-Substrate-premium-282462371.html (Accessed Oct 24, 2016)
  7. TA Evans, Modifying hair’s structure from the Inside, Cosm & Toil 131(2) 36-41 (Mar 2016)
  8. M Horio and T Kondo, Crimping of wool fibers, Tex Res J 23 373-386 (1953)
  9. GE Rogers, Electron microscopy of wool, J Ultrstruct Res 2 309-330 (1959)
  10. JA Swift, The histology of keratin fibers, in RS Asquith, ed, The Chemistry of Natural Protein Fibers, Plenum Press, New York (1977) pp 81-146
  11. S Thibaut, P Barbarat, F Leroy and BA Bernard, Human hair keratin network and curvature, Int J Dermatol 46 suppl 1, 7-10 (2007)
  12. WG Bryson et al, Cortical cell types and intermediate filament arrangements correlate with fiber curvature in Japanese human hair, J Struct Biol 166 46-58 (2009)
  13. G Loussouarn etal, World diversity of hair curliness: A new method of assessment, Int J Dermatol 46 suppl 1, 2-6 (2007)
  14. TA Evans, Quantifying hair color fading, Cosm & Toil 130(1) 30-35 (Jan/Feb 2015)

Related Content



Figure 1. Optical microscopy image of fine blonde hair

Figure 1. Optical microscopy image of fine blonde hair

Optical microscopy shows both the absence and discontinuous nature of the medulla.

Figure 2. Kinetic data of reaction rates

Figure 2. Kinetic data of reaction rates

Reaction rates for a common perm solution with hair from different sources

Figure 3. Young’s modulus versus relative humidity plots for different hair samples

Figure 3. Young’s modulus versus relative humidity plots for different hair samples

We began noticing differing modulus vs. relative humidity curves for hair from different sources. These results are all from single-source hair procured from donors, as opposed the blended hair traditionally provided by suppliers.

Next image >