© S. Leclère et al., Published by EDP Sciences, 2025

1 Introduction

An increasing number of individuals are exposed to psychological stress in current “modern” day-to-day life. Psychological stress is a normal physical, mental or emotional response to perceived challenges, threats, or demands in one’s environment (“stressors”). However, when stress becomes chronic, it can have negative impacts on the body. Psychological stress is known to cause and exacerbate a number of skin diseases associated with skin barrier defects, such as pruritus, psoriasis or atopic dermatitis (Hall et al., 2012; Graubard et al., 2021; Marek-Jozefowicz et al., 2022).

There are two main stress response axes in the body, namely the hypothalamic–pituitary–adrenal (HPA) axis and the sympathetic nervous system. In the HPA axis, psychological stress triggers the production of Corticotropin-Releasing Hormone (CRH) in the hypothalamus. This leads to the release of adrenocorticotropic hormone (ACTH) from the hypophyseal gland and to cortisol release in the adrenal gland. The skin also has its own equivalent of the HPA axis. Keratinocytes are able to synthesize cortisol (Pondeljak and Liborija, 2020; Cirillo and Prime, 2011). Local synthesis enables the skin to react to stress and is finely regulated to maintain cutaneous homeostasis. Numerous studies have described the negative effects of psychological stress on cutaneous homeostasis, mediated by an excessive level of cortisol. This can lead to immune and inflammatory response impairment; damage to the integrity of the stratum corneum and to epidermal barrier function (Choe et al., 2018; Altemus et al., 2001; Choi et al., 2005); delayed healing (Ebrecht et al., 2004); increase in reactive oxygen species production and DNA damage, factors which contribute to cutaneous aging (Krutmann et al., 2017). Research conducted by Choe et al. in 2018 showed a higher level of cortisol in the stratum corneum of individuals experiencing psychological stress (students taking exams) versus individuals not experiencing stress (students not taking exams). They also observed an increase of trans-epidermal water loss in stressed students, indicating skin barrier impairment (Choe et al., 2018). Other factors able to induce cortisol production include UV light (Skobowiat et al., 2013; Boudon et al., 2017; Slominski et al., 2018), temperature, and surrounding humidity (Takei et al., 2013; Zhu et al., 2014). The sympathetic nervous system plays a role in high alert situations in which it is necessary to engage the “fight or flight response”. It is responsible for the release of catecholamine neurotransmitters, in particular adrenalin and noradrenalin. These catecholamines are secreted in response to stress, leading for example to an increased heart rate (Chen and Lyga, 2014). The sympathetic nervous system is also responsible for releasing neuropeptides such as substance P. One of the most abundant neuropeptides of the central nervous system (CNS), substance P is involved in a wide range of physiological and pathophysiological processes, including regulation of stress and behaviors relating to emotions and anxiety (Ebner and Singewald, 2006). It is significantly involved in the transmission of pain signals, in inflammatory processes and sensitive skin.

Previous work has confirmed an association between sensitive skin and common psychological concerns (Farage, 2022). Symptoms of sensitive skin such as itching, stinging, burning, and pain can lead to sleep disorders, fatigue, stress, and anxiety. Conversely, lack of sleep and stress from external sources can make the individual with sensitive skin more susceptible to symptoms. This becomes a vicious circle that impacts on quality of life and well-being (Fig. 1). The purpose of this work was to develop a new cosmetic active ingredient which can countering this vicious circle between stress and the condition of dry and/or sensitive skin, which is a major challenge in cosmetic domain. We evaluated the biological efficacy of this ingredient by in vitro models targeting cortisol and substance P. We confirmed these effects by two in vivo studies.

Fig. 1 Vicious circle of psychological stress and skin conditions.

2 Ecodesigned Camellia Oleifera seed oil concentrate: origin, production, and characteristics

2.1 Camellia Oleifera

The Camellia genus is native to East Asia and includes more than 200 evergreen woody species. Some species have high economic value such as C. sinensis, C. japonica, and C. oleifera. C. oleifera, also known as oil-tea camelia, can grow on sterile, unfertilized ground, begin to produce fruit 6 yr after initial plantation, and remain highly productive for 80 yr. This species is mainly grown in China for the production of food oils (camelia oil, tea seed oil) (Yang et al., 2016). Camellia oleifera is a shrub 5 to 7 m tall with straight elliptical leaves. Its flowers are small, white and mildly fragranced. The fruits are capsules with 3 cavities containing 1 to 2 seeds 2 cm in size. The seeds are comprised of a husk and a kernel (Fig. 2), which represents around 60% of the seed’s total weight. Oil can be extracted directly from the seed or kernel.

The quantity of oil contained in the seeds and fatty acid composition can vary depending on the variety, soil, and climate (Hu and Yang, 2018). Wen et al. (2018), consider light intensity during fruit growth and ripening (July-October) to influence fruit yield and quality. As for the varietal diversity, Yang et al. (2016) compared 10 different cultivars and seeds from several wild specimens and concluded that the oil content and fatty acid profile of their extracts were not influenced by artificial culture and cultivar selection. In many Asian countries (China, Taiwan, Japan, India, Indonesia) Camellia oleifera oil is used in cooking and used in traditional medicine. It is also recommended by the Food and Agriculture Organization of the United Nations as healthy alternative due to its specific composition (Guo et al., 2023, Li et al., 2022).

Fig. 2 Camellia oleifera tree & seed.

2.2 Molecular distillation

Molecular distillation is a highly precise separation technique that operates under a vacuum environment and at high temperature, making it ideal for separating and purifying sensitive compounds in a gentle and eco-friendly manner. This solvent-free extraction method enables to concentrate the unsaponifiable fraction of oil by a factor of 5x or 10x, to obtain an enriched oil while preserving complex molecular integrity. The unsaponifiable fraction of vegetable oil is the most biologically active fraction, despite constituting a minor part (between 0.1% to 2.0% and up to 15% in rare cases). The main constituents of interest are phytosterols, tocopherols, and tocotrienols well known for their anti-inflammatory, anti-oxidative, and wound healing properties (Fontanel, 2013).

2.3 Phytochemical profile

The analytical profile of the Tea Oil Concentrate (TOC) obtained by gas chromatography technology is described below (Tab. 1).

We obtained a product with a specific unsaponifiable fraction containing a large amount of β-amyrin ((3β)-Olean-12-en-3-ol). β-amyrin is described as the major pentacyclic triterpenoid compound found in medicinal plants. Previous studies have demonstrated the biological role of β-amyrin including anti-inflammatory, antioxidant, antinociceptive, antimicrobial, immune-boosting, anxiolytic, and antidepressant properties (Abdel-Raouf et al., 2015; Nogueira et al., 2019; Viet et al., 2021).

The safety of the extract had been demonstrated by several safety in vitro assays such as skin and ocular irritation, skin sensitization and mutagenicity.

3 Materials and methods

3.1 In vitro evaluation

3.1.1 Cortisol release in a skin explant model

Skin explants were obtained with the informed consent from abdominal surgery of a 55 yr old female Caucasian donor (Biopredic). Skin explants were cultivated into standard 12-well plates in contact in DMEM (ThermoFisher) medium at 37°C in 5% CO₂ humidified air. They were treated in systemic with two products: TOC at 0.001% (v/v) or a positive reference metyrapone at 1 mM (Sigma-Aldrich). The explants were incubated for 24h then stressed with ACTH 0.6 µM (Sigma-Aldrich) and kept in survival conditions in an incubator at 42°C to imitate a dry environment. The explants were treated again with products. After 24h incubation, the culture media were collected, and an enzyme-linked immunosorbent assay (ELISA) of cortisol (EIAHCOR, Invitrogen) was performed. All experimental conditions were performed at least in triplicate. For each condition, means and standard deviations were calculated and compared with untreated control and stress control (ACTH + 42°C) without tested products. Statistical significance between conditions was assessed using a two-tailed, unpaired Student t-test, with p < 0.05 being considered significant.

3.1.2 Skin barrier protective effect in a skin explant cortisol topical stress model

Skin explants were obtained with the informed consent from abdominal surgery of a 41 yr old female Caucasian donor (Biopredic). Skin explants were kept alive by culturing on metal grids into standard 12-well plates in contact in OxiProteomics® medium at 37°C in 5% CO₂ humidified air. Six hours after the explant reception and equilibration in culturing optimal conditions (D0), skin explant surfaces were damaged by tape stripping followed by 6h topical stress with an aqueous solution of cortisol 0.001% (v/v) incubation. The explants were then treated with two products: a formulation (cream) containing 1% TOC or a placebo. One day after the first application (D1), the explants were treated with cortisol again for 6h before applying the products. These treatments (6h cortisol + product) were repeated a total of 5 times (D0 to D4). The day after the last treatment (D5), the explants were collected to perform hematoxylin (Merck) and eosin (Invitrogen) staining to analyze stratum corneum thickness and Lucifer Yellow (Merck) staining to analyze skin barrier integrity. Light microscope images were collected with an epi-fluorescent microscope (EVOS M5000 Imaging System) and analyzed with ImageJ software (Schneider et al., 2012). All experimental conditions were performed at least in triplicate. For each condition, means and standard deviations were calculated and compared with untreated control and stress condition (cortisol) without tested products. Statistical significance between conditions was assessed using One Way ANOVA followed by Tukey’s or Dunnett’s test with p < 0.05 being considered significant.

3.1.3 Substance P quantification in a reinnervated epidermis model

This model was inspired by previously described models (Lebonvallet et al., 2012; Lebonvallet et al., 2013; Sakka et al., 2018; Bocciarelli et al., 2023). Human induced pluripotent stem cells (AXOL Bioscience) were differentiated into sensory neurons in culture inserts (Costar® Transwell® in a 24-well plate − Corning) according to manufacturer’s instruction. A collagen gel (Corning) was then applied to the sensory neurons, to enable the axons to develop in three dimensions. After 3 weeks of culture, an explanted epidermis (from abdominal surgery of an adult female Caucasian donor with the informed consent), dissociated from the dermis with dispase treatment, was deposited on the collagen gel. The epidermis model was incubated for 5 days, for the epidermis to become reinnervated by the sensory neurons. The epidermises were then treated topically with TOC formulated at 0.5% and 1% in paraffin (Sigma-Aldrich). After 24h incubation, the epidermises were treated topically with lactic acid 10% (Sigma-Aldrich) for 15 min. The culture media were then collected (DMEM − Lonza) and an ELISA test was carried out on substance P (Cayman Chemical, Advion Interchim Scientific®). All experimental conditions were performed in at least four replicates. For each condition, means and standard deviations were calculated and compared with untreated control and stress condition (lactic acid) without tested products. Statistical significance between conditions was assessed using the Mann-Whitney U test with p < 0.05 being considered significant.

3.2 In vivo evaluation

The efficacy of Camellia Oleifera (TOC) was evaluated by 2 in vivo studies.

3.2.1 “Well-being” in vivo study

The objective of this randomized double-blind comparative study was to evaluate the efficacy of TOC vs. placebo in subjects whose well-being was impacted by dry or sensitive skin conditions.

3.2.1.1 Subjects and products

Measurements were performed on 44 Caucasian women over 18 yr old with skin phototype I to IV, free from pathological findings on anatomical area studied, with even distribution of dry and/or sensitive skin on the face (with functional and/or clinical signs) and expressing psychological discomfort/stress/low mood due to their dry and/or sensitive skin. All volunteers gave their informed consent. The subjects were randomized (1:1) on inclusion and applied a cream containing the active ingredient (1% TOC) or a placebo for 28 days, twice a day (morning and evening) to their face and body. The evaluations were carried out on the face.

3.2.1.2 Physical self-perception profile (PSPP) questionnaire: self-esteem assessment.

Self-esteem was assessed using the physical self-perception questionnaire (Ninot et al., 2000) corresponding to the approved French version of the Physical Self-Perception Profile (PSPP) developed by Fox and Corbin (1989). This questionnaire comprises 6 scales (general level, physical level, physical condition, athletic performance, appearance, strength). The questionnaire’s hierarchical structure is used to differentiate levels of self-esteem, perceived physical value, and physical appearance.

3.2.1.3 Sensitive Scale questionnaire: sensitive skin assessment.

The Sensitive Scale (Misery, 2014) was used to evaluate the degree of overall cutaneous irritation and the severity of the skin condition. Based on 10 items corresponding to perceived signs (8 items) or obvious signs (2 items) in relation to cutaneous discomfort, the subject evaluated the intensity of each parameter from 0 to 10. A total score corresponding to the sum of the 10 items is calculated.

3.2.1.4 Scrapbooking Implicit Task (SIT) test: “Self-confidence / fulfilment” assessment.

The SIT test is based on a projective approach aiming to translate pre-verbal assumptions about the self into analog representations (e.g., pictures, etc.). During the SIT, the volunteer selects as fast as possible an image among a sample of three (8 presentations), in line with what he is currently feeling. All images are calibrated in terms of semantic categories and selected according to the targeted dimension: “Self-confidence / fulfilment”. A “SIT score” is calculated based on subject choices. The higher the score, the greater the match with the targeted dimension. Examples of 3 pictures used during the SIT test are illustrated Figure 3.

Fig. 3 Examples of pictures used during the SIT test.

3.2.1.5 Clinical scoring

Clinical scoring of scaling, redness, roughness and overall dryness were performed on the face by a validated technician expert under controlled and reproducible illuminations conditions (numeric scales from 0 (best score) to 9 (worst score)).

3.2.1.6 Mirror Test™: vocal analysis (prosody), cardiac activity and semantic analysis (verbatim)

The Mirror Test^TM is a specific method developed by Spincontrol (Tours, France) (Vial et al., 2018). Its objective is to confront a subject with their own reflection and to measure emotional reactions. The subject sits in front of a mirror and answers specific questions during self-observation, he answers specific questions which are recorded to detect voice and physiological response (e.g., heart signal). Exposure to self-reflection is highly emotional and leads to an emotional burden related to self-esteem. Quantification of this emotional response is performed by 3 different analyses. A neutral question is first asked for baseline measurements of each parameter. Two questions about the skin condition of the subject and the impact on mood and feeling are then asked.

Heart Rate Variability (HRV) is calculated from the heart signal. Slight variations in HRV are commonly used in psychosociological studies of stress as it offers one of the most stable parameters of a subject’s emotional stress. Stress and negative emotions are well known for disturbing cardiac coherence and lead to an irregular HRV signal. A more positive state of mind should therefore increase HRV values (Mather and Thayer, 2018).

Vocal stress is calculated from the voice signal. The emotional voice content (prosody) can be assessed by studying the vocal spectrum revealing variations in a number of parameters. Using a specific computerized solution which combined vocal parameters linked to stress as vocal intensity or fundamental frequency (Giddens et al., 2013), the vocal stress load is quantified. Verbatim (semantic analysis) is analyzed from verbal responses given using a statistical approach (text mining method) (Talib et al., 2016). This type of approach is the traditional social and human sciences approach. The frequency of the various verbal occurrences is calculated, and statistical significance is determined.

Results for HRV and vocal stress were given as a percentage of the baseline values measured using the neutral question.

3.2.1.7 Self-assessment questionnaire

After 28 days of treatment, the subjects completed a questionnaire to assess their overall opinion of the efficacy of the active product and placebo tested (answers on 4 points Likert scale).

3.2.2 “Cutaneous comfort” in vivo study

The objective of the second randomized double-blind comparative was to evaluate the efficacy of TOC vs. placebo on the skin of subjects with dry or sensitive skin conditions.

3.2.2.1 Subjects and products

Measurements were performed on 48 Caucasian women over 18 yr old with skin phototype I to IV, free from pathological findings on anatomical area studied, with dry and/or sensitive skin on the face with stinging and/or sensations of heat, burning and/or itching, and a self-perceived cutaneous discomfort ≥4 on a scale of 0 to 10 on inclusion. All volunteers gave their informed consent. The subjects were randomized (1:1) on inclusion and applied a cream containing the active ingredient (1% TOC) or a placebo for 28 days, twice a day (morning and evening) to their face. The evaluations were carried out on the face.

3.2.2.2 Evaluation of skin hydration

Epidermal hydration was assessed using the skin capacitance Corneometer CM 825 (Courage & Khazaka). The device determines the water content of the superficial epidermal layers to a depth of about 0.1mm and expresses the values in arbitrary units (a.u.). With this method, the skin can be classified as very dry skin (<30 a.u.), dry skin (30–40 a.u.), and normal skin (>40 a.u.) (Heinrich et al., 2003).

3.2.2.3 Evaluation of the skin barrier function by measuring Transepidermal Water Loss (TEWL)

Transepidermal water loss was measured using a Tewameter TM 300®. A steam flow crosses a probe placed on the skin with two sensors. The partial pressure difference is measured between the two sensors. The value obtained corresponds to the rate of water evaporation.

3.2.2.4 Clinical scoring

Clinical scoring of desquamation, redness, roughness, and overall dryness was performed on the face by a dermatologist under controlled and reproducible illuminations (scales from 0 (best score) to 9 (worst score). In addition, skin discomfort expressed by the subject was recorded (scales from 0 (best score) to 9 (worst score)).

3.2.2.5 Evaluation of skin color by spectrophotometry

Skin color was measured using a MINOLTA CM700-d spectrophotometer with an 8-millimeter diameter head. The spectrophotometer works in the color space CIELAB defined by the International Commission of Illumination (CIE) and converts colors perceived by humans into three parameters L*, a* and b*, where L* is skin lightness (from the darkest to fairest), a* and b* are color channels (a*from green to red and b* from blue to yellow) as illustrated Figure 4. Lightness L* and redness a* were assessed for this study.

Fig. 4 L*a*b* space illustration.

3.2.2.6 Evaluation of Quality of life by questionnaire

Subjects completed a quality of life questionnaire on D0 and D28 specifically designed for the study. The questionnaire included 19 questions covering 5 general dimensions (emotion, pain, sleep, mental state, general). An analysis of each question and of the total score (summing of the 19 questions) was performed.

3.2.2.7 Evaluation of cutaneous inflammation and stress by quantification of biomarkers

Two skin samples (swab and D-Squame) were taken from each subject on D0 and D28. Swab sampling was performed for skin inflammation assessment by quantifying the Interleukin-1α (IL-1α), which is found in the epidermis and released in response to various stimuli (Akdis et al., 2016). After defining an area on the cheek, the swabs were applied strongly on skin for 45 s. The samples were stored at −20 °C before analysis. IL-1α was extracted from swabs using a cocktail buffer. After a 30 min sonication, the cytokines were quantified using specific enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems). A decrease of this biomarker points to reduced cutaneous inflammation.

D-squame sampling were performed for skin stress assessment by quantifying cortisol. D-squame are adhesive tapes allowing to collect superficial corneocytes. After defining an area on the cheek, 5 successive D-squames were applied on the same area with a pressure during 30 s. The last 3 D-squames were kept and stored at −20°c. Corneocytes were extracted from these 3 D-squames and cortisol was quantified by liquid chromatography and mass spectrometry detection. A decrease of this marker points to less severe stress. Cortisol was analyzed in 15 subjects in the 1% TOC active ingredient group and in 17 subjects in the placebo group (cortisol was undetected in 9 subjects in the 1% TOC group and 7 subjects in the placebo group).

3.2.2.8 Self-assessment questionnaire

After 28 days of treatment, the subjects completed a questionnaire to assess their overall opinion of the efficacy of the products tested.

3.2.3 Statistical analysis of clinical studies

For both in vivo studies, statistical analysis was performed using R software version 3.6.1 (R; 2024). For each statistical test performed between 2 timepoints or between active ingredient vs. placebo, the normality of the data was tested by a Shapiro-Wilk test with a threshold α = 0.01. The following was performed depending on the result of this test (Millot 2018):

• For before/after comparisons of each product, the Student’s t-test for paired data (following normal distribution) or the Wilcoxon signed-rank test (not following normal distribution).

• For comparisons between the active ingredient vs. placebo, the Student’s t-test for unpaired data (following normal distribution) or the Mann-Whitney test (not following normal distribution).

A p-value p < 0.05 was considered statistically significant and p < 0.1 indicative of a borderline significant difference.

Results obtained by self-assessment questionnaire are interpreted by comparison of the percentage of positive answers to a percentage of 50% (considered as a random answer percentage) using a binomial statistical test. Active ingredient vs. placebo results were compared using a chi-square test.

4 Results

4.1 In vitro results

4.1.1 TOC decreases cortisol release in skin explant

ACTH and temperature (42°C) stress induced a significant increase in cortisol release (+28% vs. untreated explant, p < 0.05) (Fig. 5). The positive reference, metyrapone, decreased the amount of cortisol release (−16%) and protected skin explants from stress by 70%.

TOC 0.001% also allowed a significant decrease in cortisol (−21%, p < 0.01) and protected skin explants from stress by 93% (Fig. 5). These data show that TOC was able to reduce cortisol release and that it could protect and improve the skin barrier.

Fig. 5 Assay of cortisol produced by skin explants stimulated by ACTH + 42°C. $ p < 0.05 vs untreated control; ** p < 0.01 vs stress; two-tailed, unpaired Student t-test

4.1.2 TOC shows protective effect in cortisol stress model impairing the skin barrier explant

Damaging the stratum corneum by tape stripping followed by repeated cortisol treatment over five days significantly decreased stratum corneum thickness (−38% vs. untreated explant, p < 0.001) and the integrity of the skin barrier (Fig. 6). Diffusion of Lucifer Yellow fluorescence increased strongly (+141% vs. untreated explant, p < 0.001) showing an alteration of skin barrier. The cream formulated with 1% TOC significantly protected skin explants from stress by increasing stratum corneum thickness (+55%, p < 0.001, corresponding to 88% protection vs. untreated explant) and by reducing the penetration of Lucifer Yellow stain (−49%, p < 0.001, corresponding to 83% protection vs. untreated explant).

Placebo treatment showed an efficacy 35% and 36% protection, respectively, but the active ingredient was superior and the difference between active ingredient and placebo was significant (p < 0.001) (Fig. 6). Overall, TOC improved the integrity of the skin barrier, promoted skin barrier repair and protected it from psychological stress.

Fig. 6 Histological analysis of explant model (hematoxylin, eosin stain); (b) Quantification of stratum corneum thickness; (c) Analysis of Lucifer Yellow fluorescent dye diffusion through the explants (yellow: Lucifer Yellow dye, blue: DAPI; (d) Quantification of Lucifer Yellow dye penetration. $$$ p < 0.001 vs control; *** p < 0.001 vs stress or placebo; one Way ANOVA followed by Tukey’s or Dunnett’s test

4.1.3 TOC inhibits substance P release in a reinnervated epidermis model

Lactic acid 10% induced an increase of substance P release (+48% vs. untreated epidermis). TOC 1% and 0.5% significantly decreased substance P release (−56% and −48% vs. lactic acid respectively, p < 0.05) (Fig. 7).

These data show that TOC inhibited substance P release and has the potential to improve the comfort of sensitive skin.

Fig. 7 Assay of substance P released by reinnervated epidermis model. * p < 0.05 vs lactic acid; Mann-Whitney U test.

4.2 In vivo results

4.2.1 “Well-being” in vivo study results

The group receiving active ingredient included 22 subjects aged 51.3±14.7 yr (range: 19–70 yr), including 18 subjects with dry/dryness-prone skin and 19 subjects with sensitive skin. The group receiving placebo included 22 subjects aged 45.3±13.8 yr (range: 23–68 yr), including 19 subjects with dry/dryness-prone skin and 20 with sensitive skin.

Results for the PSPP, SIT score, Sensitive Scale, clinical scoring, and Mirror Test™ (HRV and vocal stress parameters) are in Table 2.

PSPP results showed an improvement in the overall self-esteem with the active ingredient (+5.0%, p = 0.218) whereas the placebo diminishes the overall self-esteem (−3.6%, borderline significant change, p = 0.060). The difference between the active ingredient and the placebo was significant in favor of the active ingredient (−163.8%, p = 0.040). The same observation was made in the PSPP strength scale, with a borderline significant difference between the active ingredient and the placebo in favor of the active ingredient (−171.2%, p = 0.093).

The SIT score showed a significant improvement for the active ingredient between D0 and D28 (+42%, p = 0.030) without significant difference with the placebo (−30.6%, p = 0.833).

In the Sensitive scale evaluation, results showed that for all 10 signs evaluated and for the total score, the improvement was significant between D0 and D28 with the active ingredient (Tingling: −92.2%, Burning: −89.7%, Warm Felling: −84.7%, Tautness: −77.2%, Itching: −87.8%, Pains: −68.8%, General discomfort: −83.3%, Hot flushes: −72.2%, Redness: −79%, Desquamation: −75.7%, Total Score: −77.2%, p < 0.05) compared with only 5 signs and the total score with placebo (Warm Felling: −56.4%, Tautness: −45.2%, General discomfort: −51%, Hot flushes: −71.2%, Redness: −58.8%, Total Score: −50.7%, p < 0.05). There was a significant difference between the active ingredient and the placebo in favor of the active ingredient for signs of tingling (−74.2%, p < 0.001), burning (−68.6%, p = 0.027), and itching (−60.2%, p = 0.027). Borderline significant difference was also observed between the active ingredient and placebo in favor of the active ingredient for signs of warm feeling (−54.8%, p = 0.076), tautness (−40.0%, p = 0.096), general discomfort (−43.3%, p = 0.081).

The clinical scoring of scaling, redness, roughness and overall dryness assessed by an expert did not show any improvement of the cutaneous condition for the active ingredient.

Results from the Mirror Test™ showed a significant increase of the HRV with the active ingredient (+11.6%, p < 0.001) and a significant difference with placebo in favor of the active ingredient (−50.3%, p = 0.005). There was also a significant decrease of vocal stress with the active ingredient (−11%, p < 0.001), with a significant difference vs. placebo in favor of the active ingredient (−88.2%, p < 0.001). Outcomes on HRV and vocal stress showed an improvement in the emotional state of the subjects.

The verbatims or words used on D0 and D28 to answer the questions during the Mirror Test™ are illustrated in Figure 8. Using a Factorial Analysis of Correspondences, the two main statistical dimensions represented 79.4% of the total variance, indicating an accurate account of the semantic space evoked by both groups. Dimension 1 clearly discriminated the points in time (D0 vs. D28). The semantics expressed at D0 by the two groups were very similar and were clearly discriminated in Dimension 2 at D28. Both treatment groups expressed generally positive assessments at D28. However, the specific verbatim produced by the group testing the active ingredient was related to the reduction in cutaneous discomfort compared to the D0 skin condition, whereas the placebo group spoke in less specific terms (i.e., hydration, softness).

In the self-assessment questionnaire at D28 (Tab. 3), the active ingredient was perceived as significantly effective on 10 items (p < 0.05). The difference between the active ingredient and the placebo was significant in favor of the active ingredient for 6 items out of 20 (p < 0.05).

Fig. 8 “Well-being” in vivo study results: verbatims from the Mirror TestTM. Blue dots represent the verbal corpus for the active ingredient and placebo groups. Red triangles represent the main specific expressed terms by the subjects following the questions during the Mirror TestTM. The inserts list the specific terms significantly associated with a specific corpus (p < 0.05). Green arrows illustrate the change in the semantic spaces for each group.

4.2.2 “Cutaneous comfort” in vivo study results

The group receiving active ingredient included 24 subjects aged 50.0±16.7 yr (range: 24–71 yr). Nineteen subjects had dry/dryness-prone skin and 23 had sensitive skin in this group. The group receiving placebo included 24 subjects aged 54.2±16.8 yr (range: 24–74 yr). Of these, 22 subjects had dry/dryness-prone skin and 22 had sensitive skin.

The results for instrumental measurement (hydration and TEWL), clinical scoring, color (L* and a* parameters) and biomarkers (IL1-α and cortisol) are in Table 4.

Hydration significantly improved from D0 to D28 with the active ingredient (+25.1%, p < 0.001) but there was no significant difference vs. placebo (−10.4%, p = 0.787). For the skin barrier, a significant improvement was observed from D0 to D28 with the active ingredient (−12.5%, p < 0.001) along with a significant difference vs. placebo in favor of the active ingredient (−34.8%, p = 0.046).

The clinical scoring of desquamation (−57.1%, p = 0.016), redness (−32.9%, p < 0.001), roughness (−42.0%, p < 0.001) and overall dryness (−48.4%, p < 0.001) performed by a dermatologist showed a significant improvement with the active ingredient from D0 to D28 but no significant differences vs. placebo (p > 0.05).

Color measurements showed a significant improvement of lightness L* and redness a* with the active ingredient from D0 to D28 (+5.1%, p < 0.001) but no significant differences vs. Placebo (+15.4%, p > 0.05).

Results from biomarker quantification after 28 days of application showed that the active ingredient significantly decreased the quantity of IL1-α in skin (−42.2%, p < 0.05). The placebo also decreased levels of IL1-α with a borderline significant change (−4.1%, p = 0.087). The difference between the active ingredient and placebo was significant in favor of the active ingredient (−88.9%, p < 0.001). The active ingredient also significantly decreased cortisol (−25.2%, p = 0.002) but the difference between the active ingredient and placebo was not significant (−32.0%, p = 0.737).

The results for the quality of life questionnaire (Tab. 5) show a significant improvement with the active ingredient except for pain, sleep, and walking at night (improvement was borderline significant for these last two parameters). A significant improvement in favor of the active ingredient vs. placebo was observed on emotional (anger and irritability) and leisure activity aspects. A borderline significant improvement in favor of the active ingredient vs. placebo was observed on emotional (sadness), on sex life, and overall quality of life aspects.

Self-assessment questionnaire results showed that the active ingredient was perceived as significantly effective on 11 out of 14 items. No significant differences between the active ingredient and placebo were observed (Tab. 6).

5 Discussion and conclusion

Psychological stress is known to have an impact on skin and generates skin disorders of varying severity that affect well-being and quality of life (Hall et al., 2012; Graubard et al., 2021; Marek-Jozefowicz et al., 2022). Stress and skin conditions exacerbate each other in a vicious circle (Farage, 2022). The body has two major stress response systems, namely the HPA axis and the sympathetic nervous system, which allow the release of hormones and neurotransmitters such as cortisol and substance P. These responses are necessary, but when stress becomes chronic, they can have harmful consequences on the skin. Stress hormones, such as cortisol, can weaken the skin’s immune defenses, trigger allergic responses, delay healing, and disrupt the skin’s natural protective barrier (Zhang et al., 2024). Stress can increase overall inflammation in the body, which may have dermatological manifestations. Moreover, stress activates the nervous system, which can lead to increased sensitivity and reactivity in the skin (Chen and Lyga, 2014). These stress-induced changes can provoke various dermatological symptoms, such as eczema or psoriasis, that may in turn increase psychological distress (Hall et al., 2012). Visible symptoms of skin disorders cause embarrassment and affect social interaction, in addition to other symptoms such as physical discomfort due to itching or pain and sleep disruption caused by itching (Zhang et al., 2024).

During psychological stress, alteration of the epidermis and its barrier function are recognized targets (Choe et al., 2018). By reducing epidermal thickness and organization, psychological stress leads to reduced hydration and impaired protection against external agents. Studies have established the negative effects of psychological stress on the skin, which has been shown to impair the permeability barrier homeostasis (Garg et al., 2001) and stratum corneum integrity (Choi et al., 2005). As described in the literature (Zhang et al., 2024), we have observed that stress induces numerous negative impacts, including degradation of epidermal structure and barrier.

The active ingredient decreased cortisol and substance P release in skin explants and in reinnervated epidermis models, respectively. TOC also preserved the structure and improved the repair of skin barrier in a skin explant model stressed by cortisol. The in vitro studies presented here demonstrate that TOC has the potential to protect skin from the damaging effects of psychological stress.

The efficacy of the active ingredient demonstrated by in vitro studies was confirmed by two in vivo studies. The first demonstrated the positive impact of the active ingredient in subjects whose sensitive/dry skin had an impact on their well-being/emotions and their self-esteem by combining subjective questionnaires such as the PSPP and Sensitive Scale questionnaires, the subjective projection of skin conditions on image (SIT test) and objective approaches via physiological measurements carried out during the Mirror Test™ and verbatim analyses used to describe the perceived impact of the participant’s skin condition on mental state. The second in vivo study focused on the impact of the subjects’ sensitive skin/dry skin conditions on the skin physiological state itself. This analysis also demonstrated the efficacy of the active ingredient through instrumental measurement of hydration, skin barrier, color, but also through quantification of skin biomarkers of inflammation (Il1-α) and stress (cortisol) and through a quality of life questionnaire. These two in vivo studies have, therefore, highlighted the efficacy of the active ingredient that is both subjectively perceived via emotions, well-being, and quality of life and objectively measured by instrumental methods. The active ingredient has the potential to counter the vicious circle linked to the condition of dry and/or sensitive skin, thus providing effectiveness for the comfort and well-being of the skin and mind.

Acknowledgments

The authors thank the Expanscience laboratory team for their contributions: Clément Le Bras, Beatriz Soengas, Dorothée Piton, Marine Alluyn, Marion Kerdrain and Séverine Marchand. The authors thank Genel for testing the extract on the model of explant stressed by ACTH and temperature. The authors thank Oxiproteomics for testing the extract on the cortisol stress model impairing the skin barrier explant. The authors thank Université de Bretagne Occidentale, Laboratoire LIEN, for testing the extract on the reinnervated epidermis model. The authors thank Spincontrol and Dermscan for the clinical studies.

Funding

This research was funded by Laboratoires Expanscience.

Conflicts of interest

The authors are employees of Expanscience. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Author contribution statement

Conceptualization: S.L., M.L.R., and G.B.; formal analysis: M.L.R., G.B.; funding acquisition: C.B.; investigation: S.L., M.L.R., and G.B.; methodology: S.L., M.L.R., and G.B.; project administration: S.L., and C.B.; supervision: C.B; validation: S.L., M.L.R., and G.B.; visualization: S.L., M.L.R., and G.B.; writing—original draft preparation: S.L., M.L.R., and G.B.; writing—review and editing: S.L., M.L.R., G.B., and C.B. All authors have read and agreed to the published version of the manuscript.

References

All Tables

All Figures

	Fig. 1 Vicious circle of psychological stress and skin conditions.
In the text

	Fig. 2 Camellia oleifera tree & seed.
In the text

	Fig. 3 Examples of pictures used during the SIT test.
In the text

	Fig. 4 L*a*b* space illustration.
In the text

	Fig. 5 Assay of cortisol produced by skin explants stimulated by ACTH + 42°C. $ p < 0.05 vs untreated control; ** p < 0.01 vs stress; two-tailed, unpaired Student t-test
In the text

Fig. 6 Histological analysis of explant model (hematoxylin, eosin stain); (b) Quantification of stratum corneum thickness; (c) Analysis of Lucifer Yellow fluorescent dye diffusion through the explants (yellow: Lucifer Yellow dye, blue: DAPI; (d) Quantification of Lucifer Yellow dye penetration. $$$ p < 0.001 vs control; *** p < 0.001 vs stress or placebo; one Way ANOVA followed by Tukey’s or Dunnett’s test

In the text

	Fig. 7 Assay of substance P released by reinnervated epidermis model. * p < 0.05 vs lactic acid; Mann-Whitney U test.
In the text

Fig. 8 “Well-being” in vivo study results: verbatims from the Mirror TestTM. Blue dots represent the verbal corpus for the active ingredient and placebo groups. Red triangles represent the main specific expressed terms by the subjects following the questions during the Mirror TestTM. The inserts list the specific terms significantly associated with a specific corpus (p < 0.05). Green arrows illustrate the change in the semantic spaces for each group.

In the text