Sex-related differences in response to zinc pyrithione shampoo vs. non-anti-dandruff shampoo
Abstract
enObjective
Sex-related differences in skin properties may be expected to impact dandruff formation and treatment. A meta-analysis approach was undertaken to investigate potential differences between males and females in response to zinc pyrithione (ZnPT) treatment vs. non-anti-dandruff (AD) shampoo. A separate pooled statistical analysis of ceramide and total protein loss endpoints was also undertaken to assess potential sex-related differences in stratum corneum properties that might influence response to ZnPT vs. non-AD shampoo in subjects with dandruff.
Methods
The meta-analysis approach included data from 17 half-head, double-blind, randomized studies (N = 2088) undertaken in Asia to assess the effectiveness of 1% ZnPT shampoo and/or non-AD shampoo in reducing dandruff severity, as assessed by Total Weighted Head Score Adherent Flake (TWHS-AF) methodology. Treatment duration was 4 weeks, with TWHS-AF measured at weekly intervals. Data from an additional three studies (N = 143) conducted in Asia were included in the pooled analysis of ceramide levels and protein loss from scalp skin of subjects with dandruff.
Results
Response to 1% ZnPT vs. non-AD shampoo was greater in males than in females at all time points; after 4 weeks, the between-treatment difference in TWHS-AF was −17.5 (95% confidence interval [CI] −19.5, −15.5) in males and −11.1 (−13.2, −8.9) in females. Sex-related differences were observed between males and females in response to both 1% ZnPT and non-AD shampoos. Males had a stronger response than females to treatment with 1% ZnPT shampoo, while dandruff decreased to a greater extent in females than in males when using non-AD shampoo. Statistically significant sex-related differences in ceramide levels and total protein loss were observed (both P < 0.01). Ceramide levels were 0.76 times lower (95% CI 0.60, 0.97) in males than in females, while total protein loss was 1.4 times greater (95% CI 1.1, 1.9) in males than in females.
Conclusion
Males show a greater response than females to 1% ZnPT shampoo, while females show a greater response than males to non-AD shampoo. These findings may in part be explained by the sex-related differences observed in stratum corneum properties, which may make males more prone to dandruff than females.
Résumé
frOBJECTIF
Les différences liées au sexe concernant les propriétés de la peau sont susceptibles d'avoir un impact sur la formation des pellicules et leur traitement. Une méta-analytise a été adoptée afin d’étudier les différences potentielles entre les hommes et les femmes en réponse au traitement à la pyrithione de zinc (ZnPT) par rapport au shampoing non antipelliculaire. Une analyse statistique groupée distincte des critères d’évaluation des céramides et de la perte totale de protéines a également été réalisée pour évaluer les différences potentielles liées au sexe au niveau des propriétés du stratum corneum qui pourraient influencer la réponse à la ZnPT par rapport au shampooing non antipelliculaire chez les sujets présentant des pellicules.
MÉTHODES
La méta-analytise a inclus des données provenant de 17 études par demi-tête, en double aveugle et randomisées (N = 2 088) réalisées en Asie pour évaluer l'efficacité de 1% de shampooing à base de ZnPT et/ou de shampooing non antipelliculaire dans la réduction de la gravité des pellicules, telle qu’évaluée selon la méthodologie Total Weighted Head Score Adherent Flake [TWHS-AF]. La durée du traitement était de 4 semaines, et le TWHS-AF était mesuré toutes les semaines. Les données provenant de trois études supplémentaires (N = 143) réalisées en Asie ont été incluses dans l'analyse groupée des taux de céramides et de la perte de protéines du cuir chevelu chez les sujets présentant des pellicules.
RÉSULTATS
La réponse à 1 % de ZnPT par rapport au shampooing non antipelliculaire était plus importante chez les hommes que chez les femmes, à toutes les échéances; après 4 semaines, la différence de TWHS-AF entre les traitements a été de _17,5 (intervalle de confiance [IC] à 95 % −19,5, −15,5) chez les hommes et −11,1 (−13,2, −8,9) chez les femmes. Des différences liées au sexe ont été observées entre les hommes et les femmes en réaction à des shampoings à 1% de ZnPT et non antipelliculaires. Les hommes ont présenté une meilleure réponse que les femmes au traitement avec 1% de shampooing au ZnPT, tandis que les pellicules ont diminué dans une plus large mesure chez les femmes que chez les hommes avec l'utilisation du shampooing non antipelliculaire. Des différences statistiquement significatives liées au sexe ont été observées en ce qui concerne les taux de céramides et la perte totale de protéines (P < 0,01 pour les deux). Les taux de céramides étaient 0,76 fois plus faibles (IC à 95 %, 0,60, 0,97) chez les hommes que chez les femmes, tandis que la perte totale de protéines était 1,4 fois plus élevée (IC à 95 %, 1,1, 1,9) chez les hommes que chez les femmes.
CONCLUSION
Les hommes ont mieux réagi au shampooing à 1 % de ZnPT que les femmes, alors que celles-ci ont mieux réagi au shampooing non antipelliculaire que les hommes. Ces résultats pourraient en partie s'expliquer par les différences liées au sexe observées au niveau des propriétés du stratum corneum, ce qui peut rendre les hommes plus exposés aux pellicules que les femmes.
Introduction
Dandruff is a common scalp disorder that affects at least half of the adult population worldwide at some time 1. Dandruff is generally characterized by stratum corneum flaking and scaling, and is often accompanied by itch, skin tightness and erythema 2. While the precise aetiology of dandruff has still to be confirmed, the current scientific consensus is that dandruff development is influenced by three major factors: Malassezia colonization, sebum production and individual predisposition. However, the relative contributions of these factors as a trigger for dandruff formation remain the subject of debate 3.
It is well documented that the commensal yeast Malassezia is a strong contributory factor to dandruff development. In particular, the incidence of dandruff has been associated with an increased abundance of Malassezia restricta 4. However, the presence of Malassezia species on healthy scalps indicates that this alone is not a sufficient cause for dandruff formation; more recent evidence suggests that disruption of the stratum corneum may facilitate dandruff generation by affecting susceptibility to metabolites from Malassezia species. A healthy stratum corneum forms a protective barrier to prevent water loss, maintain hydration of the scalp and protect against external insults 3. Severe or chronic barrier damage can impair proper hydration, leading to atypical epidermal proliferation, keratinocyte differentiation and stratum corneum maturation, which may underlie some dandruff symptoms 5. Measurement of total protein, recovered from the scalp surface, can be used as marker of stratum corneum cohesion 6. The depleted and disorganized structural lipids of the stratum corneum in individuals with dandruff are consistent with a weakened barrier, as indicated by elevated trans-epidermal water loss 7 and measurement of diminished levels of total ceramides 6, 8, 9. Further evidence of weakening of the protective barrier in individuals with dandruff includes subclinical inflammation and greater susceptibility to topical irritants 3. The response to topically applied histamine has been shown to be greater in subjects with dandruff than controls, also suggesting compromised barrier integrity 9.
Dandruff is usually treated with anti-fungal agents such as zinc pyrithione (ZnPT), applied topically through shampoo. Studies have demonstrated sex-related differences in a number of skin properties that might be expected to impact dandruff formation and treatment, such as stratum corneum hydration and skin barrier function as measured by trans-epidermal water loss 10. In addition, there may be differences in hair washing and styling methods between males and females that could influence dandruff development 11. A previous meta-analysis found no sex-related difference in the efficacy of ZnPT treatment 12; however, the authors did not consider the relative contribution of detergency from non-anti-dandruff (AD) shampoo vs. ZnPT shampoo. Furthermore, therapeutic efficacy was only evaluated at a single time point compared with baseline.
This paper presents the results of a more recent meta-analysis undertaken to further investigate potential differences between males and females in response to ZnPT treatment, taking into account possible cleansing effects and examining treatment effects over several time points. The results of a pooled analysis from an additional three clinical studies undertaken to assess potential sex-related differences in levels of ceramide and total protein loss (as a marker of corneocyte loss) are also presented, in order to explore possible underlying mechanisms that might influence response to ZnPT.
Materials and methods
Meta-analysis of clinical trials assessing sex-related differences in response to ZnPT and non-AD shampoos
Study design
Data from 17 Unilever-sponsored, half-head, double-blind, randomized studies conducted in Asia since 2010 were included in four separate meta-analyses (Table 1). Studies were included if they assessed the effectiveness of 1% ZnPT shampoo (CLEAR, Unilever) and/or a non-AD shampoo in reducing dandruff severity.
| Study | Year | Age (years),amean (SD) | Test phase duration (weeks) | Treatment | Sample size, n (female/male)a | Meta-analysis groupsb |
|---|---|---|---|---|---|---|
| 1 | 2010 | 33.0 (8.1) | 4 | Four products, including 1% ZnPT and non-AD shampoo | 174 (102/72) | I, II, III, IV |
| 2 | 2013 | 31.5 (9.3) | 6 | Two products: 1% ZnPT and non-AD shampoo | 86 (86/0) | IV |
| 3 | 2013 | 34.0 (8.8) | 6 | Two products, including 1% ZnPT | 200 (100/100) | I |
| 4 | 2013 | 30.6 (10.0) | 4 | Two products, including 1% ZnPT | 208 (104/104) | I |
| 5 | 2014 | 46.1 (11.0) | 4 | Two products, including non-AD shampoo | 57 (33/24) | II |
| 6 | 2014 | 31.3 (9.5) | 4 | Two products, including 1% ZnPT | 139 (55/84) | I |
| 7 | 2014 | 33.3 (10.7) | 4 | Two products, including 1% ZnPT | 119 (48/71) | I |
| 8 | 2014 | 43.0 (12.0) | 4 | Two products, including non-AD shampoo | 48 (29/19) | II |
| 9 | 2014 | 35.7 (10.0) | 4 | Three products, including 1% ZnPT | 132 (78/54) | I |
| 10 | 2015 | 31.6 (9.8) | 4 | Five products, including 1% ZnPT | 228 (114/114) | I |
| 11 | 2015 | 32.2 (10.4) | 4 | Four products, including 1% ZnPT | 186 (108/78) | I |
| 12 | 2015 | 33.3 (11.0) | 4 | Two products, including 1% ZnPT | 128 (64/64) | I |
| 13 | 2015 | 33.3 (10.9) | 4 | Two products: 1% ZnPT and non-AD shampoo | 100 (51/49) | I, II, III, IV |
| 14 | 2016 | 31.2 (10.5) | 4 | Two products, including non-AD shampoo | 99 (51/48) | II |
| 15 | 2016 | 35.0 (11.9) | 4 | Two products: 1% ZnPT and non-AD shampoo | 100 (52/48) | I, II, III, IV |
| 16 | 2016 | 46.9 (10.5) | 4 | Two products, including non-AD shampoo | 35 (15/20) | II |
| 17 | 2016 | 33.9 (9.0) | 4 | Two products, including non-AD shampoo | 49 (27/22) | II |
- AD, anti-dandruff; ZnPT, zinc pyrithione.
- a At baseline.
- b See text for explanation of groups.
To be eligible for study inclusion, subjects were required to be 18–60 years of age and in good general health, and to have qualifying levels of dandruff 7 not currently treated with an AD shampoo. Exclusion criteria included: use of any medicine that could affect the study outcome; use of ketoconazole-containing treatments within the last 6 months; use of treatments containing selenium sulphide within the last 3 months; use of AD products within the last 2 months; hair colouring or perming within the last 2 weeks; other medical skin conditions; any known allergies or sensitivities to the ingredients in study shampoos; and pregnancy.
In all studies, dandruff severity was assessed using the Total Weighted Head Score Adherent Flake (TWHS-AF) methodology; this is a well-defined approach for assessing the efficacy of AD treatment 7, 9, 13. All studies involved a 2- to 3-week pre-treatment washout period, followed by a test phase of at least 4 weeks and a 0- to 4-week regression phase, with dandruff scored at weekly intervals throughout. During the pre-treatment washout period, there were two dandruff assessments (TWHS-AF): screening and rescreen. If the subject's screening score and other conditions met the inclusion/exclusion criteria, they were given a bottle of non-AD shampoo and instructed to wash their hair every other day for 2–3 weeks at home. At the end of the pre-treatment washout period, subjects were rescreened to evaluate if they should progress onto the test phase. Qualifying subjects were allocated two different test products, one for each side of the scalp (random allocation to side of scalp). In studies involving two products, the rescreen score was treated as baseline; in studies involving more than two products, there was a separate baseline visit following the washout period to ensure balance of dandruff score and sex across the product groups. During the test phase, subjects had their hair washed three times per week using the assigned product at the study site.
All studies were conducted according to the principles of good clinical practice and in compliance with local government regulations, guidelines and standards applicable to such studies. All studies were reviewed and approved by an independent ethics committee before initiation. All study participants provided written informed consent.
Statistical analysis
To compare the effects of the test product (1% ZnPT shampoo) and non-AD shampoo over time and by sex, analysis of covariance was conducted for each study using PROC MIXED in SAS (SAS Institute Inc., Cary, NC, U.S.A.). The baseline value was included as a covariate, with both a side-level and a subject-level average to avoid any cross-level bias in product comparisons 14. Additional factors were included as fixed effects to adjust for any potential effects of the product, namely time point, sex and their interactions, and subject as a random effect. The protocol-recommended UN@CS covariance structure and the ddfm = KR option were used. Four separate meta-analyses were conducted, using R 15 with the metafor package 16, using random-effects models. Results are the combined estimate of the difference between and within groups at each time point. If the 95% confidence interval (CI) did not enclose zero, then the difference was considered to be statistically significant. The four meta-analyses were conducted to estimate:
- Difference between males and females when treated with 1% ZnPT shampoo
- Difference between males and females when treated with non-AD shampoo
- Difference between 1% ZnPT and non-AD shampoo treatment in males
- Difference between 1% ZnPT and non-AD shampoo treatment in females
Pooled analysis assessing sex-related differences in ceramides and protein loss
Study design
Data from three Unilever-sponsored studies conducted in Asia were included in the pooled analysis. Inclusion and exclusion criteria were as described above. In all studies, subjects used a non-AD shampoo at home for 3–4 weeks, after which the dandruff score was re-assessed (2 days after the last home wash). In subjects with TWHS-AF remaining at or above the study entry level after the washout period, buffer scrubs were collected for measurement of ceramides and protein loss; two buffer samples were collected from each subject (one from each side of the scalp). A section of scalp skin was exposed by making a straight parting in the hair and securing the hair with standard salon clips. To obtain a buffer scrub sample, a sterile plastic ring of 18 mm internal diameter and 60 mm height was placed on the parting and held steady by both the person carrying out the sampling and a second operator. Using a sterile plastic disposable pipette, 2.0 mL of collection medium (sterile phosphate-buffered saline, pH 7.9, containing 0.1% Triton X-100) was applied to the sample site. The scalp was then gently massaged with a round-ended Teflon rod for 1 min. This was repeated with a further aliquot of 2.0 mL of collection medium for a further 1 min.
All studies were conducted following the principles of good clinical practice and in compliance with local government regulations, guidelines and standards applicable to such studies. All studies were reviewed and approved by an independent ethics committee before initiation. All study participants provided written informed consent.
Buffer scrub extractions
Buffer scrub samples were centrifuged and the corneocyte pellet retained. Lipid content from the corneocyte pellet was extracted with 1 mL chloroform/methanol (2:1, high-performance thin-layer chromatography grade) for 2 h at room temperature (referred to as ‘lipid extract’). The lipid extract was dried at 37°C under nitrogen. Excess chloroform/methanol was allowed to evaporate from the remaining lipid-depleted corneocytes.
Total protein loss assay
Lipid-depleted corneocytes were extracted in 10 mmol/L phosphate buffer, pH 7.8, containing 1% (w/v) sodium dodecyl sulphate (SDS) and 20 mmol/L β-mercaptoethanol at 60°C for 1 h. The samples were centrifuged for 10 min at 16,249 ×g, and the SDS/mercaptoethanol protein extract was removed and analysed, using the BCA™ Protein Assay Kit (Fisher Scientific, Loughborough, U.K.).
Ceramide levels
The lipid extracts were reconstituted in chloroform/methanol. Extracts were placed in duplicate wells of a deep-well plate, and the volume was made up to 800 μL with methanol. HCl (1 mol/L) was added to each well and the samples were dried overnight at 60°C. Samples were reconstituted in 80 mmol/L imidazole, 8 mmol/L triethylamine and 6.4 mmol/L acetic acid in 80% isopropanol, pH 8.3, for 30 min at room temperature. Fluorescamine (2 mg/mL) in isopropanol was added, and 150 μL was transferred to a microtitre plate. Fluorescence was read immediately at 360 nm (excitation) and 460 nm (emission), using a Millipore fluorescence plate reader. The ceramide signal generated after hydrolysis was expressed as μg/μg extracted protein. Standard curves were created using Ceramide 2 (0–20 μg/mL; Givaudan [formerly Quest International], Ashford, U.K.) and sphingosine (0–10 μg/mL; Sigma, Poole, U.K.) treated in the same manner as the samples.
Statistical analysis
Total protein loss and ceramide (normalized by extracted protein) data from the three studies were included in the statistical analysis. Regression analysis of log-transformed data was conducted to include data from both sides of the head; consequently, the effect of subject-level variables (e.g. sex and age) and study-level variables (e.g. study number) could be explored. A separate regression model was fitted for total protein loss and ceramide to assess which of the exploratory variables might have an effect. Subject ID was added as a random effect in all models. Only results for the effect of sex are presented in this paper. Differences were considered statistically significant if they had a P-value of less than 0.05; where a difference was found to be significant, the direction of this significance is also given.
Results
Meta-analysis of clinical trials assessing sex-related differences in response to ZnPT and non-AD shampoos
Key features of the 17 studies included in this meta-analysis are shown in Table 1. Data were available for a total of 2088 subjects, all of whom were of Asian race and 53% of whom were female. Mean age at baseline ranged among studies from 30.6 to 46.9 years. Three studies directly compared 1% ZnPT vs. non-AD shampoo in males and females (n = 374), and one study made the comparison in females alone (n = 86). The remaining 13 studies investigated either 1% ZnPT (eight studies) or non-AD shampoo (five studies) in both males and females (n = 1628).
Meta-analysis I included both males and females treated with 1% ZnPT (11 studies), whereas meta-analysis II included both males and females treated with non-AD shampoo (8 studies). Comparisons between meta-analyses I and II demonstrated a clear difference between female and male subjects in terms of response to both 1% ZnPT and non-AD shampoo (Fig. 1). The difference in response to 1% ZnPT (females minus males) was statistically significant at Week 2 (1.4 [95% CI 0.3, 2.6]) and Week 3 (2.0 [95% CI 0.6, 3.4]), indicating that males have a stronger response than females to treatment with 1% ZnPT shampoo. In contrast, dandruff decreased to a greater extent in females than in males when using non-AD shampoo; the difference (females minus males) was statistically significant at Week 3 (−2.6 [95% CI −4.4, −0.8]) and Week 4 (−1.8 [95% CI −3.4, −0.2]).
Results from meta-analyses III and IV showed the response to 1% ZnPT vs. non-AD shampoo to be greater in males than in females at all time points evaluated (Table 2). After 4 weeks, the between-treatment difference in TWHS-AF was −17.5 (95% CI −19.5, −15.5) in male subjects and −11.1 (−13.2, −8.9) in female subjects.
| Time (weeks) | Mean between treatment difference in TWHS-AF (95% confidence interval) | |
|---|---|---|
| Males | Females | |
| 1 | −7.3 (−8.7, −6.0) | −4.0 (−5.0, −2.9) |
| 2 | −13.1 (−15.0, −11.3) | −7.4 (−8.6, −6.1) |
| 3 | −17.3 (−19.2, −15.5) | −9.2 (−10.9, −7.5) |
| 4 | −17.5 (−19.5, −15.5) | −11.1 (−13.2, −8.9) |
- TWHS-AF, Total Weighted Head Score Adherent Flake.
Pooled analysis assessing sex-related differences in ceramides and protein loss
The pooled analysis included data from a total of 143 subjects (98 females and 45 males). All subjects were of Asian race; mean age at baseline ranged from 24.5 to 25.4 years among the studies.
Statistically significant sex-related differences in ceramide levels and total protein loss (as assessed in buffer scrub samples) were observed (both P < 0.01). Ceramide levels were 0.76 times lower (95% CI 0.60, 0.97) in males than in females (Fig. 2), while total protein loss was 1.4 times greater (95% CI 1.1, 1.9) in males than in females (Fig. 3).
Discussion
The results from the four meta-analyses, using data from 17 clinical studies, confirm the effectiveness of 1% ZnPT-containing shampoo products in improving dandruff in both male and female Asian subjects. The results also showed sex-related differences for ZnPT effectiveness compared with non-AD shampoos in Asian consumers, driven by opposing sex-related responses to AD vs. non-AD shampoo. Our findings contrast with those of a previous meta-analysis of seven studies that found no sex-related differences in response to ZnPT-containing shampoo 12. However, this previous analysis did not take into account any effects of washing alone on dandruff over time beyond the pre-treatment washout period. The current meta-analysis approach also demonstrates a difference in the placebo effect between males and females, with dandruff in females improving over time to a greater extent than in males in the absence of ZnPT. The mechanism(s) for this difference are not clear, but may involve a long-term impact on barrier function of different behavioural aspects of hair washing between males and females, as well as possible differences in scalp skin physiology.
These findings are supported by the results of the pooled analysis undertaken to assess sex-related differences in scalp stratum corneum ceramide levels and total protein loss in subjects with dandruff. The results showed males to have significantly lower scalp ceramide levels and significantly greater protein loss (in response to gentle scrubbing) compared with females. The stratum corneum is an important element in scalp protection against external insults (e.g. microbes, pollution, ultraviolet light, dirt, chemical residues) and acts as the primary epidermal barrier to water loss, maintaining hydration, flexibility and integrity of the scalp 3. This outer protective layer consists of protein cells, lipids and water 17, 18. Total protein loss is a marker of flakes adhered to the scalp and of a weak barrier. Lipids, composed of ceramides, cholesterol and fatty acids, also play an important role in maintaining the integrity of the scalp stratum corneum. Ceramides are critical to the skin barrier structure and have been shown to be decreased in the scalps of subjects with dandruff 6, 8, 9. The finding that, compared with females, males have significantly lower ceramide levels per corneocyte and significantly more protein loss (i.e. corneocyte loss) in response to gentle abrasion implies an impaired skin barrier in males, which could make their scalps more prone to dandruff.
With a weaker scalp barrier, it is also possible that males do not benefit from the detergent effect when using non-AD shampoo. Detergents in the shampoo will help to physically remove external irritants, such as microbes, pollutants, dirt and chemical residues, and temporarily reduce their load. After washing, the external irritants reaccumulate. It is possible that males, with a weaker scalp barrier, are more sensitive to external irritant load than females, which might explain the lesser reduction in dandruff levels seen in males following thrice-weekly washing with non-AD shampoo.
Subjects’ perception of changes in their dandruff relative to the subjective TWHS-AF measurement was not evaluated in the studies included in these analyses. However, the observation of sex-related differences in response to shampoos containing 1% ZnPT is consistent with a consumer poll conducted in China (Unilever, data on file), which found that a higher percentage of men (68%) than women (55%) were satisfied with their scalp condition after using shampoos containing 1% ZnPT.
A key strength of the current analyses is the large number of subjects included; the largest pool of data was available for meta-analysis I (N = 1714) for assessing sex-related differences in response to ZnPT, with 143 subjects included in the pooled analysis assessing sex-related differences in ceramides and protein loss. Although only studies in Asian subjects were included in the current analyses, results of a previous meta-analysis found no differences in response to ZnPT between Asian and Caucasian populations 12, suggesting that the sex-related differences observed in the current analyses are widely applicable.
In summary, the results of these analyses suggest that males may be more prone to dandruff than females because of lower ceramide levels and a greater propensity for protein loss in the stratum corneum. These physiological differences may explain the finding that males experience less benefit than females from the detergency of non-AD shampoo and show a greater response than females to 1% ZnPT shampoo. These findings warrant further research.
Acknowledgements
All studies and analyses reported in this paper were performed and funded by Unilever Research & Development. Editorial assistance was provided by Jennifer Coward of Anthemis Consulting, funded by Unilever Research & Development. The authors thank Yajun Luo and Yingjie Li (Unilever R&D Shanghai, China) for conducting the studies. Luisa Collins, Fiona Baines, Jane Matheson, Graham Turner, Yuanyuan Diao, Yuanpei Li and Yingying Pi are employees of Unilever Research & Development.
