Curly hair: measured differences and contributions to breakage


  • Conflicts of interest: The author declares no conflicts of interest.

Harold Bryant
L′Oréal Institute for Ethnic Hair and Skin Research, Chicago, IL, USA


A consumer internet survey conducted by our Institute in 2005, of over 1200 women that self-identified as African American, Caucasian, Chinese, or Mexican, determined that African-American women experience hair breakage at a statistically higher rate as compared with their Chinese, Mexican, and Caucasian counterparts. In fact, 96% said they experienced hair breakage while 23% agreed it was their biggest hair problem. To better understand this reported phenomenon of increased fragility, one must consider that mechanical fragility of hair may be related to innate differences in the structure, the result of various diseases, or related to grooming habits and practices. In this paper, we review some of the previously reported structural differences observed in curly hair, laboratory measurements used to evaluate fragility, and the potential impact of grooming practices.

Historically, human hair was classified along racial lines of Caucasian, Mongolian (Asian), and Negroid (African or of African descent). Early studies investigated hair diameter, geometric shape, strength, and chemical reactivity.1–4 Since these early studies, information about the fiber surface, mechanical properties, proteins, amino acids, and lipids were reported.2,5–8 Despite the reported differences in shape, reactivity, swelling, and mechanical strength, bulk property chemical analysis reports no significant differences in protein content or amino acids. Results from the analysis of lipid content of African and Caucasian hair are inconsistent as Kreplak et al. report lower values for African hair and Nicolaides and Rothman report higher values.7,8 In general, African hair is the most elliptical in shape, requires less force to break, and tends to be the curliest compared with Asian and Caucasian. To quantify the different degrees of curl in hair, L’Oreal developed a worldwide curl classification system.9 The system, which is based on the analysis of curl diameter, number of waves and twists on hair from around the world, yielded eight distinct classes. Curl classification provides a standardized way to study the overall impact of curl on other attributes, including the relationship between curliness and grooming. In fact, as the hair becomes curlier it typically becomes thinner, has more twists, and is more difficult to manage because of the increased effort required during combing or brushing.10 The kinks and twists have often been referred to as local stress concentrators and are believed to be the locations where fibers are most likely to fail.2 The physical and mechanical properties described above must have their origins in the structural components of hair and despite the lack of definitive evidence, some speculation on the contribution of the individual structures to fragility is warranted.

Beyond general properties, it is important to understand the structures of hair and their potential impact on fragility. Hair consists of three main substructures: cuticle, cortex, and medulla, the last of which has no reported significance to structural integrity. The two structures considered responsible for hair strength are the cuticle and ortex. The contribution and importance of each is a topic still under debate. Some suggest that failure of the protective cuticle layer must precede any major structural failure as can be seen with split ends, where erosion of the cuticle layer exposes the cortex, which then fractures.2 Others feel that flexion, torsion, and repeated elongation, perhaps related to grooming, leads to a weakening of components in the cortex promoting fiber failure.2,11 The inability to monitor internal structures in a dynamic environment where loads can be applied and varied, as is reported for the cuticle in environmental chambers for scanning electron microscopy,12 makes it difficult to determine which structure fails first.

Outside of the consumer-perceived increased fragility for curly hair, our work indicates a clear correlation between degree of curliness and break stress (Fig. 1). Pertaining to the protective cuticle, differences in cuticle spacing, and the number of layers was reported based on race.13,14 In order of increasing space between cuticle layers are Caucasian, Asian, and African hair.13 In a more recent work, Japanese and Caucasian hairs were compared and showed that not only was the cuticle spacing smaller for Japanese hair but there were more cuticle layers.14 One may speculate that the tighter the layers and the greater they are in number, the greater protection they provide. At the time of this publication, no such work was published on number of cuticle layers in African hair. There are several reports comparing the components that make up the cortex in hair of different origin and curl. The overall arrangement of proteins as measured by X-ray diffraction did not reveal any differences.12 However, differences in the distribution and type of cortical cells were reported for curly hair compared with straight hair and differences in extractable proteins related to the country of origin of the hair was also reported.15,16

Figure 1.

 Break stress versus curl classification. AA, African American; CA, Caucasian; JA, Jamaican.

The remainder of this paper will focus on the perception of hair breakage: what the consumers know, what grooming habits may contribute, and which laboratory measurements correlate with the perception.

Materials and methods

Quantitative and qualitative consumer tests

A total of 715 African-American women, 18 and over, participated in a quantitative survey via internet (n = 353) or telephone interview (n = 362). The average age was 42 for the internet respondents and 38 for the telephone respondents. In the qualitative test, 73 African-American women, of whom 39 reported a serious hair breakage problem and 34 claimed not to have a problem, visited our facility. All women consented to give hair samples for lab testing and some were asked to bring their products and tools so that they may be observed during their typical grooming processes. All women were interviewed on site.

Hair sample analysis

All hair was individually swatched and washed using a 10% ammonium lauryl sulfate solution for 1 min followed by 1 min of rinsing in 40 °C tap water. The hair was then equilibrated in an environmental chamber at 45% relative humidity for a minimum of 8 h. Geometrical measurements of hair from each subject were taken using a laser scan micrometer (Mitutoyo, Kanagawa, Japan) at 22 ± 2 °C and 45 ± 5% relative humidity. Dimensional values for a single fiber were obtained by taking the average of three cross-sectional areas along the length of the fiber that was 3 cm in length. Results from 50 fibers per subject were averaged. The tensile properties of dry hair from all the subjects were determined using a miniature tensile tester (MTT) 675 (Dia-Stron Ltd., Hampshire, UK) using a strain rate of 12 mm/min and a gauge force of 1 g. The environmental temperature and humidity was 22 ± 2 °C and 45 ± 5% relative humidity, respectively. All tensile properties were measured on the first 50 hair fibers that exhibited normal failure profiles. Young’s modulus and break stress were normalized using the cross-section of each fiber.


Knowledge of breakage

Of the 715 women surveyed, 64% said that breakage was a real problem as opposed to the rest that just experienced it. Respondents overwhelmingly felt that hair breakage was a sign of poor hair quality. However, breakage was closely associated with, and was in fact, indistinguishable from shedding and split ends. This fact is supported by responses to the question of “What part of the fiber do you notice the breakage?”. Thirty-four percent of responses were positive for fibers broken at the fiber tip and 35% reported breakage in the middle of the fiber. When their fibers break, they notice them mostly on the comb (73%), brush (58%), and sink (52%).

Grooming practices

As expected, the majority of women noticed breakage during combing (81%) and brushing (62%). Ninety percent of these women reported using chemical treatments, including relaxers (79%) and 22% used more than one type of treatment concurrently. Many of the women indicated experiencing breakage during use of heat appliances, including blow dryers and curling irons at a rate of 62% and 42%, respectively. Nearly all the women responding (87%) acted to correct the problem. The majority of them used conditioners, oils, and moisturizers with a 71% satisfaction rate.

Qualitative observations

A small portion of the women were invited to visit our facility with their products and tools for observation during their normal grooming routine. Women that did not experience breakage were recruited and observed to provide information as a control group. In total, 73 women participated in the qualitative study and hair collection. The two groups of clients were observed doing their normal grooming routine to include shampooing, conditioning, combing, and styling. Several qualitative observations were made that suggest a trend toward an overall higher level of hair involvement/manipulation for the breakage group. In addition, this group used more products, brushes, and finer-toothed combs. Table 1 lists some of the other general observations.

Table 1.   General grooming observations
Breakage groupNo-breakage group
More manipulationLess combing, gentler motion
Smaller tooth combsLarger tooth combs
Use of brushesAir dry hair
Combing with more forceLess frequent relaxer use

Mechanical analysis

Hair samples were cleaned and equilibrated before analysis using the MTT. Samples from 73 participants were grouped into six subsets, including breakage, no-breakage, relaxed, natural, root, and tip. The results of multiple fiber analysis for each subject were compared with the self-assessed problem of hair breakage. The data in Table 2 show two of the parameters accessed by MTT (i.e., break stress and premature failure). Break stress is a measure of the force required to break a hair fiber and premature failure rate is determined by the number of fibers that fail before completion of a normal stress–elongation curve (Fig. 2). As expected, as we compare root with tip, there is a decrease in the mechanical values reported for the samples furthest from the scalp. This difference is larger for treated samples, which may be indicative of overlapping treatments or increased environmental susceptibility of the treated hair fibers. In a comparison of treated versus untreated fibers as a whole, the treated fibers have lower break stress and higher premature failure rates. Comparing the mechanical parameters for the breakage and no-breakage groups reveal that the premature failure rates are higher and break stress values are lower for those who experience hair breakage. It is interesting to note the relative nature of the self-assessment of hair breakage. For women with treated hair, identical break stress values were obtained despite nearly half of the women indicating they do not experience hair breakage. The same can be seen in the comparison of two groups with very similar premature failure rates: women with natural hair that experience breakage and women with treated hair that do not experience breakage. Figure 3 shows the relationship between perceived hair breakage and premature failure rate, which appears to be a better predictor than break stress, particularly for women with treated hair.

Table 2.   Mechanical properties of collected hair fibers
Subjects (no. of fibers tested)13 (475 fibers)17 (425 fibers)21 (550 fibers)22 (600 fibers)
  1. Bs, Break stress, PMF, premature failure; Treated, any chemical treatment.

RootNo-breakage NaturalBreakage NaturalNo-breakage TreatedBreakage Treated
BS203.7 ± 29.6198.07 ± 32.1167.06 ± 32.5160.55 ± 29.35
PMF (%)581122
TipNo-breakage NaturalBreakage NaturalNo-breakage TreatedBreakage Treated
BS190.4 ± 29.7175.50 ± 32.2138.66 ± 29.0138.78 ± 29.57
PMF (%)172835.6046.60
Total (%)PMF19333949
Figure 2.

 Typical stress/elongation profile

Figure 3.

 Premature failure for hair breakage groups


Structural differences in hair based on curl type exist and are published. One may speculate for the case of the cuticle differences that the greater distances between and the fewer number of cuticle layers may correlate with a weaker fiber due to a decrease in the protective shielding of the cortex provided by more spatial distribution of the cuticle layers. Hairs from women that perceive they have a hair breakage problem have a lower break stress and a higher premature failure rate; however, it is the latter that appears to be more sensitive to the perception when the hair is treated. The correlation between lower break stress and increased degree of curl, association of increased curl and amplified grooming difficulty, and the fact that the majority of African-American women use permanent hair-straightening treatments all contribute to the breakage problem. Grooming practices play a role in determining the level of hair breakage experienced. As observed in the qualitative study, women that did not have a problem had a less involved grooming process with fewer steps, fewer products, fewer appliances, and longer intervals between straightening treatments. It is important to note that in this study there was a group of women with either natural hair or treated hair that did not perceive they had a breakage problem and the mechanical data supported the difference in their hair compared with those with a problem. Based on our qualitative observations, the difference, within group, may center on more involved grooming practices, which may in turn lead to the perception of damage and alternation of the hair properties as measured.