Hair curl type, hair habits and scalp heath, part 1: Clinical studies

Abstract

Background

Very curly, textured hair is often associated with specific scalp conditions, such as traction alopecia and seborrheic dermatitis. However, to date, very few detailed clinical studies have been performed to correlate scalp conditions to curl type and hair habits.

Methods

Healthy female panelists, with hair not damaged by chemical treatment, were recruited in both the US and South Africa (SA) based on their natural curl type: wavy (Walker Type 2), curly (Walker Type 3) or kinky (Walker Type 4). Both scalp and hair assessments were performed at visit 1 (baseline) and at visit 2, 1 week afterwards. Between the visits, panelists were asked to wash their hair four times with a standard shampoo and conditioner selected and provided by the authors.

Results

A significant proportion of panelists with Type 4 hair washed their hair less frequently than panelists with the other hair types and were more inclined to use hair styling techniques (braiding, extension, knots, etc.), which induce tension on the scalp. They also self‐perceived having drier and less oily scalps than panelists with the other hair types. Dermatologists' assessments showed the incidence of seborrheic dermatitis was higher in Hair Type 4 in the US. Dermatologists observed the highest levels of traction alopecia in Hair Type 4 in both countries. Scalp image analysis showed that visible scalp scaling was more common in Type 4 hair versus other hair types in both the US and SA. Washing the hair every 2 days between the two visits improved self‐perceived scalp cleanliness and scalp condition across all curl types in both countries.

Conclusions

Our results confirm that more frequent use of specific styling practices in Type 4 hair is linked to higher levels of traction alopecia. Results also suggest that lower wash frequencies observed mainly with Type 4 hair were associated with a higher degree of scalp irritation and scaling and that washing hair more frequently has the potential to improve scalp health. Our data also suggest that scalp image analysis is a more consistent way of measuring scalp colour and scaling.

A significant proportion of individuals with Type 4 hair reported less frequent hair washing and a greater reliance on styling practices that may alter scalp conditions. Our study demonstrates that increasing hair‐washing frequency improves scalp health by reducing itchiness, dryness and dandruff, while enhancing overall cleanliness.

INTRODUCTION

Textured hair is more prone to different types of scalp conditions [1]. These conditions are related to ethnicity, cultural practices and hair habits, such as frequent braiding of the hair, which exerts significant traction forces on the scalp and leads to a reduction of wash frequencies in very curly hair types. Scalp diseases often associated with very curly and textured hair include specific types of alopecia [2], such as traction alopecia (TA), discoid lupus erythematosus (DLE) and central centrifugal cicatricial alopecia (CCCA) and types of scalp dermatosis, such as seborrheic dermatitis (SD).

An epidemiological study by Alexis et al. [7] on the dermatological conditions experienced by people of different ethnicities in the New York area showed that black, African‐American patients tended to present to dermatologists more often with scalp complaints than white patients [3]. The two top scalp complaints are alopecia (8.3% black patients) and SD (6.3% black patients). In good agreement with these data, a US‐wide study examining dermatologists’ records from 1993 to 2009 also showed that SD was in the top five dermatological diagnoses for African‐American patients [4]. Other US studies show that TA is almost exclusively seen in African‐American subjects [5] and is linked to the use of pleated (cornrow) hair styles [6]. Hair extensions and infrequent hair oil use by African‐Americans were also significantly related to SD [6].

A wider study on African women visiting hairdressers in Cape Town has shown that the most frequent scalp conditions, seen on examination by a dermatologist, were TA (31.7%), androgenic alopecia (6.94%), SD (3.4%), dandruff (3.1%) and CCCA (2.7%) [7]. Interestingly, 52% of these women self‐reported having dandruff, suggesting that sub‐clinical dermatoses, not detected by dermatologists, might also be common.

Clearly, a major problem with the studies published to date on scalp problems associated with textured hair is that they usually focus on patients presenting to doctors and dermatologists. As a result, they do not accurately represent the levels and types of scalp problems encountered in the general population. Another problem with published studies is that they usually use ethnicity as an indicator of natural hair curl levels, which is an over‐simplification and makes the data harder to interpret, particularly in mixed‐race populations.

In the present study, we compared scalp condition in healthy women panelists with a range of defined hair curl types in two countries, the United States (US) and South Africa (SA), using three evaluation methods: (i) professional evaluation by dermatologists, (ii) self‐perception based on questionnaires and (iii) scientific measurements based on macro‐imaging pictures. Recruitment was deliberately not made on the grounds of ethnicity. We also focused on panelists with natural, non‐chemically damaged hair. This took out a major variable, since it is widely known that some chemical treatments, such as relaxers used to straighten textured hair, can cause scalp irritation. Other relevant scientific measurements were performed on the scalp, including in vivo ATR‐FTIR spectroscopic analysis and the analysis of inflammatory cytokines. These results will be presented separately in a second publication related to this clinical study.

METHODS

Clinical study

This study was performed in two locations, Piscataway, New Jersey (US) and Cape Town, South Africa (SA). Tests in the US were performed in July–August 2019 (US summer). Tests in SA were performed in November–December 2019 (South African late‐spring to summer).

This study recruited around 120 panelists in each country for a total of 243. Panelists were recruited based on their natural curl type: wavy (Walker Type 2a, 2b 2c), curly (Walker Type 3a, 3b, 3c), and kinky (Walker Type 4a, 4b, 4c). In the US, 40, 39 and 39 panelists for Walker Hair Types 2, 3 and 4, respectively, were recruited. In SA, 40, 39 and 46 panelists for each curl type were recruited.

Subjects were all female, aged between 18 and 55, and had hair that was at least 6 cm in length or longer. All subjects self‐classified themselves as premenopausal or perimenopausal. A key exclusion criterion was that subjects must not have coloured or chemically treated their hair for at least 6 months prior to the start of the study. All hair was therefore considered largely ‘natural’ and the scalp not exposed to chemical aggression. Panelists were not recruited on the basis of existing scalp conditions or on existing product usage (e.g., use of leave‐on hair treatments or oils).

Panelists were invited to visit the clinical centre on two occasions. A first baseline visit (Visit 1) and a second visit 1 week afterwards (Visit 2). At the baseline visit, the panelists' fit with the inclusion and exclusion criteria were checked, and their hair types were assigned by a trained assessor. Panelists also completed a baseline questionnaire covering hair and scalp properties, hair styling habits and scalp issues. A dermatological assessment was also performed at Visit 1, and scalp pictures were recorded with a macro‐imaging system. Two different dermatologists were used, one in the US and one in SA.

At the end of Visit 1, panelists were given standard shampoo and conditioner products. Suave products in the US and Revlon Flex products in South Africa. Panelists washed their hair with these products at the clinical centre for the first wash and were instructed to wash their hair every other day with the same products until visit 2. So, each subject washed their hair four times between Visit 1 and 2.

Subjects who braided their hair were required to un‐braid their hair prior to washing and were instructed that they should not re‐braid the hair for the duration of the study. Subjects were instructed not to wear scarves, wigs and/or heat‐style their hair for the duration of the study. They were also asked to refrain from using any products in the hair, except for the shampoo and conditioner provided for the duration of the study.

At Visit 2, the panelists completed another questionnaire regarding any perceived changes in hair and scalp condition, and scalp images were taken again.

Scalp imaging

Digital clinical photography of the scalp was performed using a macro‐imaging system, the Skincam™ (Newtone Inc., Princeton, New Jersey). Images were taken in three zones around the scalp for each panelist at each visit. At least one image was taken for each zone, but sometimes 2–3 more if there were interesting features.

The Skincam™ had an acquisition area of 20 mm diameter, an acquisition time of less than 3 s and a resolution of 60 pixels mm⁻¹ (i.e. a pixel size of 16 μm). Images were taken in cross and parallel polarisation modes to reveal different scalp surface components. Controlled and enclosed lighting in the Skincam™ meant that the system could accurately and reproducibly measure skin colour.

Scalp image analysis

The Scalp and Dandruff Image Analysis System (SDIAS; Newtone Inc., Princeton, New Jersey) used the images collected in both the US and SA to characterize scalp scaling and skin colour. All scalp images were independently analysed without knowledge of the panelists' details.

The SDAIS was able to detect hair fibres (regardless of colour, from dark to white) and remove them from the region of interest. This just left the exposed scalp skin for analysis. Once isolated, the colour of the exposed skin area could then be measured using a variety of parameters, including luminance (L*), yellowness (a*) and redness (b*).

The SDAIS had an algorithm for detecting images with and without skin scales. Skin scales could be further characterized based on the number and size of the bright spots on exposed scalp skin. The parameters measured were (i) the number of detected scales, (ii) global surface occupied by scales and (iii) average scale size.

Data distributions from the analysis of hair scale number and size were often strongly positively skewed, with medians lower than the means. This skew was corrected by the Log10 transformation of the data.

RESULTS

Age, curl level and ethnicity

The mean ages of the panelists recruited in the US were well balanced between the three hair types (35, 37 and 37, for Walker Hair Types 3, 4 and 5, respectively). The mean ages in SA were 35, 34 and 30. In SA, the hair type 4 was significantly younger than the same group in the US. The distributions of Walker sub‐curl types (a to c), within Hair Types 2 and 3, were similar in the US and SA; whilst for Type 4 hair, the SA cohort had higher levels of subtypes 4b and 4c (more curly hair). In the US, panelists with Hair Type 2 were mainly Caucasian, Hair Type 4 were mainly African‐American, and Hair Type 3 were a mixture of ethnicities (mainly Caucasian, African‐American and Hispanic). In SA, Hair Types 2 and 3 were mainly of mixed ethnicity (termed Coloured in SA), and Hair Type 4 was mainly Black panelists.

Washing habits

The baseline questionnaire at Visit 1 asked whether panelists washed their hair daily, 2–4 times a week, every week, every 2 weeks, every 3 weeks or every month. The responses showed that hair curl type had a large effect on washing frequency in both the US and SA (Figure 1a,b). Hair Type 4 panelists washed their hair less frequently than those with less curly hair types. In the US, 36% of panelists with Hair Type 4 left 14+ days between washes, versus 0% of panelists with Hair Type 2. In SA, 59% of panelists with Hair Type 4 left 14+ days between washes, versus 18% of panelists with Hair Type 2.

FIGURE 1.

Consumer habits (questionnaire visit 1) – Wash frequency versus curl type for (a) US and (b) SA.

Styling habits

The baseline questionnaire at Visit 1 asked panelists to describe which types of hair styling practices they had used over the past 12 months. Ten options were available, namely: braids, twists, knots, weaves, scarves, extensions, cornrows/plaits, buns/ponytails, banding and puffs. Questionnaire responses showed that the styling practices were very diverse across all hair types. In the US, buns and ponytails were the most popular types of styles used across all curl types (Table 1a). In both countries, panelists with Type 4 hair were more likely to favour styles such as braids, twists, weaves, extensions, cornrows, plaits and puffs than people with less curly hair. Head scarves were also more popular with Type 4 hair. Panelists in SA also liked ponytails and buns; however, unlike the US, panelists with Type 4 were less likely to favour this style (Table 1b). Braids were the most popular style for Type 4 hair in SA, but, unlike the US, cornrows and plaits were not particularly popular. In SA, wearing scarves during the day was most popular with Type 2 hair.

TABLE 1.

Types of hair styles used within the last 12 months versus curl type (questionnaire visit 1), in (a) US and (b) SA. Data represent percentages of each hair type.

Hair styling method	Percentage using this method in the last year or more frequently
	Hair type 2	Hair type 3	Hair type 4
(a)
Braids	35	54	62
Twists	28	23	46
Knots	5	10	18
Weaves	0	8	36
Scarves	5	33	64
Extensions	3	10	36
Cornrows/Plaits	5	33	59
Buns/Ponytails	83	95	67
Banding	8	10	15
Puffs	5	31	44
(b)
Braids	5	8	54
Twists	5	13	35
Knots	10	10	17
Weaves	15	13	30
Scarves	45	23	24
Extensions	45	23	24
Cornrows/Plaits	38	28	35
Buns/Ponytails	75	64	22
Banding	30	13	4
Puffs	8	15	7

Panelists' self‐perceived scalp problems

The baseline questionnaire at Visit 1 asked panelists whether they currently experienced any scalp issues, including itchy scalp, flaky scalp, dry scalp, oily or greasy scalp, dandruff, hair loss or sensitive/inflamed scalp. Survey responses in Table 2 showed that scalp issues were common across all hair types. Results showed that 70% of all panelists believed that they had at least one scalp issue. In the US, hair Types 4 had significantly more problems with self‐perceived dry scalp versus hair Type 2 and significantly fewer problems with oily/greasy scalp (p < 0.05, Chi‐squared test, Table 2). There was also a tendency for Types 3 and 4 to have more itchy and flaky scalps. Similar results were obtained in SA, although only the differences in oily/greasy scalp were significant (p < 0.05, Chi‐squared test). Interestingly, the incidence of dandruff was similar across all hair types in both countries. Approximately 33% of all panelists believed they currently experienced dandruff.

TABLE 2.

Self‐perceived scalp problems, versus curl type, in (a) US and (b) SA. The data represent the total number of subjects for each hair type. Rows are shaded to indicate which problems were significantly affected by Hair Type (p < 0.05, Chi‐squared test).

Scalp issues	Number of panelists			Chi‐squared test (p value)
	Hair type 2 (n = 40)	Hair type 3 (n = 39)	Hair type 4 (n = 39)
(a)
Itchy scalp	9	15	11	0.2908
Flaky scalp	4	9	11	0.1160
Dry scalp	5	13	21	0.0005
Oily/greasy scalp	14	8	3	0.0121
Dandruff	8	9	10	0.8364
Hair loss	10	6	9	0.5439
Sensitive/inflammed scalp	0	0	1	0.3601
None	16	11	11	0.4307

Scalp issues	Number of panelists			Chi‐squared test (p value)
	Hair type 2 (n = 40)	Hair type 3 (n = 39)	Hair type 4 (n = 46)
(b)
Itchy scalp	10	16	22	0.0873
Flaky scalp	5	5	4	0.7941
Dry scalp	9	11	19	0.1525
Oily/greasy scalp	6	2	0	0.0167
Dandruff	13	6	13	0.1913
Hair loss	11	6	16	0.1272
Sensitive/inflammed scalp	3	3	2	0.7738
None	12	13	10	0.4667

Dermatologist assessed dandruff and seborrheic dermatitis

Dandruff and SD were both assessed using a 10‐point scoring system (0 = no sign, 1 = mild+, 2 = mild++, 3 = mild+++, 4 = moderate+, 5 = moderate++, 6 = moderate+++, 7 = severe+, 8 = severe++, 9 = severe+++). Dandruff was not defined in words in the protocol. SD was defined as the scalp having a “reddish colour, swollen and greasy appearance and a white or yellowish crusty scale on the surface”.

Tables 3 and 4 show the relationships between hair type and dermatologist‐assessed dandruff and SD, respectively. The dandruff data show that the incidence of dermatologist‐assessed dandruff was similar across hair types in both countries, although the severity of the dandruff appeared to be higher in Type 4 hair in the US. Data also show that the overall incidence and severity of dermatologist assessed dandruff was much lower in South Africa versus the United States. Given that panelists' self‐assessment of dandruff was similar in both countries, the differences in dermatologists' scores probably indicate differences in grading methods between the two dermatologists.

TABLE 3.

Dermatologist assessed incidence and severity of dandruff, versus curl type, in (a) US and (b) SA. The data represent percentages of each hair type. The overall incidence of dandruff was not significantly different between Hair Types in the US or SA (Chi‐squared test, p = 0.4517 and p = 0.0905, respectively). Boxes are shaded just to aid visualisation.

Severity	Dandruff
	Hair type 2	Hair type 3	Hair type 4
(a)
9 = Severe+++	0	0	0
8 = Severe++	0	0	0
7 = Severe+	0	0	0
6 = Moderate+++	0	0	2.6
5 = Moderate++	0	0	0
4 = Moderate+	0	2.6	5.1
3 = Mild+++	7.5	12.8	20.5
2 = Mild++	7.5	25.6	10.3
1 = Mild+	37.5	23.1	12.8
0 = No sign	47.5	35.9	48.7
(b)
9 = Severe+++	0.0	0.0	0.0
8 = Severe++	0.0	0.0	0.0
7 = Severe+	0.0	0.0	0.0
6 = Moderate+++	0.0	0.0	0.0
5 = Moderate++	0.0	0.0	0.0
4 = Moderate+	0.0	0.0	0.0
3 = Mild+++	0.0	0.0	0.0
2 = Mild++	0.0	0.0	2.2
1 = Mild+	12.5	7.5	21.7
0 = No sign	87.5	92.5	76.1

TABLE 4.

Dermatologist assessed incidence and severity of SD, versus curl type, in (a) US and (b) SA. The data represent percentages of each hair type. The overall incidence of SD was significantly different between Hair Types in the US, but not in SA (Chi‐squared test, p = 0.002 and p = 0.6188, respectively). Boxes are shaded just to aid visualisation.

Severity	Seborrheic dermatitis
	Hair type 2	Hair type 3	Hair type 4
(a)
9 = Severe+++	0	0	0
8 = Severe++	0	0	0
7 = Severe+	0	0	0
6 = Moderate+++	0	0	0
5 = Moderate++	0	0	2.6
4 = Moderate+	0	2.6	0
3 = Mild+++	5	5.1	12.8
2 = Mild++	5	0	15.4
1 = Mild+	10	5.1	15.4
0 = No sign	80	87.2	53.8
(b)
9 = Severe+++	0.0	0.0	0.0
8 = Severe++	0.0	0.0	0.0
7 = Severe+	0.0	0.0	0.0
6 = Moderate+++	0.0	0.0	0.0
5 = Moderate++	0.0	0.0	0.0
4 = Moderate+	0.0	0.0	0.0
3 = Mild+++	0.0	0.0	0.0
2 = Mild++	0.0	0.0	0.0
1 = Mild+	2.5	0.0	2.2
0 = No sign	97.5	100.0	97.8

The SD data show a significantly higher incidence (p < 0.05, Chi‐squared test) and severity of dermatologist‐assessed SD in Hair Type 4 in the US. However, in marked contrast to the US data, the overall incidence of dermatologist‐assessed SD in South Africa was very low, if not non‐existent. The very marked differences here between countries could be real, but they might also reflect, again, differences in scoring methods between the two dermatologists, highlighting the necessity to evaluate these scalp conditions with scientific methods that could be compared, whatever the country or the period in which the studies are carried out.

Dermatologist assessed traction alopecia

Dermatological assessments of marginal traction alopecia used the MTA‐S scoring system developed by Khumalo et al. (2008) [8]. Table 5 shows clearly that TA on the front of the scalp is a major issue for Hair Type 4 in both the US and South Africa, affecting 82% and 48% of women, respectively. No TA was observed for Hair Type 2 in both countries. As observed previously with dandruff and SD, the data show that the overall incidence and severity of TA were much lower in South Africa compared with the United States, indicating the variations that can be observed with these operator grading methods.

TABLE 5.

Dermatologist‐assessed incidence and severity of traction alopecia (total anterior score) versus curl type, in (a) US and (b) SA. Data represent percentages of each hair type. The overall incidence of traction alopecia was significantly different between Hair Types in the US and SA (Chi‐squared test, p < 0.001 and p < 0.0001, respectively). Boxes are shaded just to aid visualisation.

Severity	TA front
	Hair type 2	Hair type 3	Hair type 4
(a)
9 = Severe+++	0	0	0
8 = Severe++	0	0	0
7 = Severe+	0	0	0
6 = Moderate+++	0	7.7	28.2
5 = Moderate++	0	2.6	7.7
4 = Moderate+	0	2.6	7.7
3 = Mild+++	0	17.9	23.1
2 = Mild++	0	5.1	7.7
1 = Mild+	0	0	7.7
0 = No sign	100.0	64.1	17.9
(b)
9 = Severe+++	0	0	8.7
8 = Sever ++	0	0	2.17
7 = Severe+	0	0	0
6 = Moderate+++	0	0	8.7
5 = Moderate++	0	0	4.35
4 = Moderate+	0	0	0
3 = Mild+++	0	0	17.39
2 = Mild++	0	0	6.52
1 = Mild+	0	0	0
0 = No sign	100.0	100.0	52.2

Panelists' perception changes in scalp condition after 1 week of controlled hair washing between visits 1 and 2

At the end of Visit 1 panelists were given a standard shampoo and conditioner and were asked to wash their hair right away at the study centre under the supervision of trained operators and then once every 2 days until they returned, 1 week afterwards, for Visit 2. At Visit 2 panelists were asked to complete a follow‐up questionnaire.

The follow‐up questionnaire asked whether panelists perceived changes in scalp cleanliness, scalp itchiness, scalp dryness or dandruff. The questionnaire asked whether they had perceived an increase or decrease in each scalp issue or no change. Overall, panelists perceived significant changes in scalp condition after a week of controlled hair washing period between visits 1 and 2. Indeed, Figure 2 shows that all hair types in both countries perceived an increase in scalp cleanliness and a decrease in scalp itchiness, dryness and dandruff. Further data analysis showed that only improvements in scalp cleanliness were significantly altered between hair types in SA (p = 0.0328, Chi‐squared test). Otherwise, all the other attributes changed equally between hair types in both countries (p > 0.05). Pooling the data from both countries showed that improvements in scalp cleanliness and dandruff were better in panelists with Type 4 hair versus other curl types (p = 0.0075 and p = 0.0252, respectively. Data not presented in Figure 2).

FIGURE 2.

Consumer self‐perception. Changes in self‐perceived scalp condition between Visit 1 and Visit 2 versus curl type, in the (a) US and (b) SA.

Plotting the changes in self‐perceived scalp condition versus regular wash frequencies indicated by the panelists during the first visit, we observed that improvements in scalp cleanliness, itchiness, dryness and dandruff tended to be experienced most by panelists who normally washed the least frequently (Figure 3). Chi‐squared tests comparing the distribution of responses across different wash frequencies showed that improvements were significantly different for cleanliness (p = 0.0217), itchiness (p = 0.0356) and dandruff (p = 0.0186). Differences were not significant for dryness (p = 0.1534).

FIGURE 3.

Effect of regular wash frequency on the self‐perception improvement in scalp condition between Visit 1 and Visit 2.

Scalp scaling using macro‐imaging system

Some typical scalp images obtained by the Skincam™ are shown in Figure 4. Analysis of scalp images using SDIAS revealed that the number of panelists with visible scales was significantly different between the hair types at both Visit 1 and Visit 2, in both countries (p < 0.05, Chi Squared test) (Figure 5). Curly hair was more likely to have visible scales. Furthermore, the incidence of scaling appeared to decrease between Visit 1 and Visit 2 in Hair Types 2 and 3, but not in Hair Type 4. The overall number of panelists with visible scales was not different between NJ and SA, at Visit 1 or 2 (Chi‐squared, p > 0.05). This showed that the incidences of scalp scaling based on the scientific method were roughly similar between the two countries.

FIGURE 4.

Examples of scalp images (cross‐polarized). (a) Mild+ dandruff (b) Moderate++ SD.

FIGURE 5.

Percentage of panelists with visible scales versus curl type at Visit 1 and Visit 2 based on image analysis, in the (a) US and (b) SA.

Whilst the number of panelists with visible scales was different between hair types and between visits (Figure 4), when we looked at just the panelists with scales and studied the number of visible scales, the differences between the hair types and visits were no longer visible. However, surprisingly, both the area covered by scales and the scale size were lower in Hair Type 4, in both the US and SA (ANOVA, p < 0.05). The reasons why this should be the case are not clear.

Scalp colour using macro‐imaging system

The individual typology angle (ITA) skin colour scale is widely used to measure skin colour [9] and classifies skin as very light, light, intermediate, tan, brown and dark, based on the L* and b* values. In this study, the colour of the skin areas between scales and between hair fibres was analysed. These data showed that hair Types 2 and 3 could be classified as brown (mean ITA = −3.88 and −18.56, respectively), and Type 4 was classified as dark (mean ITA = −49.32).

Clearly, the major differences in underlying skin colour related to different types of curl complicate investigations of skin redness and inflammation. However, we could analyse differences in skin colour in relation to scaling within each hair type. Figure 6 shows that, as expected, scaling was associated with a darker skin colour (between the scales), particularly in Hair Type 4 panelists. Presumably, this is related to inflammation of the underlying skin. Differences in skin lightness (L*) between skin with and without scales were significant for Hair Type 3 in SA and in Hair Type 4 in both NJ and SA (p < 0.05, two‐tailed t‐test, equal variances, Visit 1). No clear differences were seen in skin tone (a* or b*).

FIGURE 6.

Scalp skin lightness (L*) with and without scales versus curl type at Visit 1 in the (a) US and (b) SA (means +/−s.d.)

Further analysis investigated the changes in skin colour between Visit 1 and Visit 2. No significant changes were seen in L*, a* or b*, in any Hair Types (ANOVA, p > 0.05). Presumably, the short period between Visit 1 and 2 was not long enough to dramatically change skin colour.

DISCUSSION

Effects of washing habits

A key hypothesis, going into this study, was that lower wash frequencies, associated with Hair Type 4, would lead to significant build‐up of sebum on the scalp and to more skin inflammation. The breakdown of sebum triglycerides by lipases into fatty acids, such as oleic acid, is believed to trigger skin irritation [10].

Questionnaire data on wash habits confirmed that, as expected, wash frequencies in both the US and SA reduced as hair curl levels increased. It is likely that less manageable hair and more involved styling practices probably dissuade consumers with Type 4 hair from more frequent washing.

Scalp irritation and scaling increased with increased hair curl levels. Analysis of scalp images showed that the incidence of scalp scaling was higher in Hair Type 4 in both the US and SA. Self‐perceived scalp dryness was also higher in panelists with Hair Type 4 in the US. Dermatologist assessments showed that the incidence and severity of SD were worse in Hair Type 4 in the US, although not in SA.

In keeping with our hypothesis, washing the hair every other day for a week improved scalp condition. Scalp imaging showed reductions in the incidence of scalp scaling in Hair Types 2 and 3, in both the US and SA. Panelists also reported self‐perceived improvements in scalp cleanliness, itch, dryness and dandruff across all hair curl types in both countries. Furthermore, good correlations were observed between panelists' usual wash frequencies and the improvements in cleanliness, itch and dandruff delivered by more frequent washing. Those panelists who normally washed their hair the least frequently were more likely to see improvements.

Taken together, these data seem to confirm that infrequent washing may lead to skin irritation and that simply washing the scalp more frequently with a basic shampoo and conditioner can significantly improve scalp condition. In good agreement with our findings, studies on Japanese female volunteers have shown that not washing the hair allows more sebum triglycerides to be converted into fatty acids and that the build‐up of these fatty acids is correlated with skin itchiness. Interestingly, washing the hair and scalp removed the decayed sebum and immediately reduced scalp itch [11]. Our results also agree with studies on Asian volunteers that showed that more frequent washing was associated with improved scalp condition [12].

It should be noted that the increased scalp irritation associated with Hair Type 4 noted at Visit 1 may have been influenced by more factors than just wash frequency. Higher use of leave‐on hair and scalp products by consumers with this hair type has been shown to be associated with scalp irritation [6]. It is possible that use of such products, in combination with less frequent washing, could have further increased scalp irritation. The relative and combined effects of wash frequency and use of leave‐on products on scalp health require further investigation.

Effects of styling practices

A hair habit associated with Hair Type 4 is the highest use of hair styling methods, such as braids, that pull the hair from the scalp. This creates major scalp problems. The widest spread is the traction alopecia. Our study showed that anterior traction alopecia was present in 82% and 48% of women with Type 4 hair in the US and South Africa, respectively, but not in women with Type 2 hair. The clinical information collected during this large project, and previously reported studies, highlight the need for additional investigations to find ways of minimizing this form of hair loss.

Effects of geographic location

Running parallel legs of our study in the US and SA enabled us to compare differences in wash habits and styling practices between the two countries and the severity of scalp problems. Looking at the wash habits data, the overall differences in wash frequencies between hair curl types were similar in both countries, with Hair Type 4 washing least frequently in both countries. Styling practices were different in both countries, but the use of specific hair styles, such as braids, was still very popular in both countries for Hair Type 4. The scalp images showed that the incidence of scalp scaling was similar in both countries across each hair type, with Hair Type 4 having the highest incidence of scaling in both countries. Furthermore, the increases in skin darkening in panelists who had visible scales were similar across countries. Self‐perceived scalp health was overall quite similar across different hair types in the US and SA. Increased scalp dryness was significantly higher in the US for Hair Type 4 and directionally higher in SA. Increased oily/greasy scalp was significantly higher in both the US and SA for Hair Type 2. The largest differences between countries were observed in the dermatologists' data, which suggested that the incidence of SD was higher in the US, and that the severity of dandruff and SD was higher in the US. Given the similarities in the instrumental data and panelists' self‐perception data, we must assume that these differences are largely down to differences in the scoring methods of the dermatologist in the US and the dermatologist in SA.

CONCLUSION

Our results confirm that more frequent use of specific styling practices in Type 4 hair is linked to higher levels of traction alopecia. Results also suggest that lower wash frequencies observed mainly with Type 4 hair result in a higher degree of scalp irritation and scaling and that washing hair more frequently has the potential to improve scalp conditions. The different methods used in this study to evaluate the scalp condition (self‐perception, dermatologist assessment and scientific measurements) made it possible to highlight that scalp image analysis is a more consistent way to evaluate scalp conditions such as colour and scaling during clinical trials.

Further studies associated with this project investigated the effects of hair curl and wash frequency on scalp health using additional instrumental measures and biological markers, including FTIR spectroscopy of the scalp surface (detecting, for example, sebum lipids, sebum composition or scalp hydration), scalp surface pH measurements, and analysis of skin inflammatory markers (IL‐1α and IL‐1RA). These results will be reported in a second, associated paper.

AUTHOR CONTRIBUTIONS

SGA led the study. SGA, RS and PC conducted the study and analysed the data. MJ performed the image analysis work. SGA and PC authored the paper.

CONFLICTS OF INTEREST STATEMENT

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.