As representatives of northeast Asia, 3 east cities (Guangzhou, Nanjing, and Shijiazhuang) in China and Suwon in Republic of Korea were selected (Fig. 1). Guangzhou and Nanjing were selected as warm cities, and Shijiazhuang and Suwon as cold cities. In each temperature group, Nanjing and Suwon were selected as humid cities, Guangzhou and Shijiazhuang as dry cities. The data were obtained between December 21, 2015 and January 15, 2016. Geographic data of the 4 cities and climatic information during the study period are summarized in Table 1. The study aimed to recruit more than 80 healthy women from each city. In each city, at least 20 women were recruited by each age group: 18~21 years, 22~25 years, 26~37 years, and 38~49 years. Since the study was designed for the purpose of developing skin care products for certain social stages of life, age group was divided into college years, early employment years, mid-twenties to late-thirties, and late-thirties to just before 50 years of age. All subjects had lived in each city for at least 5 years. Subjects who were pregnant, lactating, or planning a pregnancy were excluded. None of the subjects had significant skin disorder, had taken any medications that could interfere with the study results, or had undergone any skin treatments (i.e., chemical peel, or laser treatments) within 6 months prior to the study. Finally, a total of 361 healthy Chinese and Korean women between 18 and 49 years of age in 4 cities were enrolled. This study was approved by the Institutional Review Board at IEC Korea (No. IECK (1)-IRB-15-K254443) and was conducted with strict adherence to the Declaration of Helsinki Principles.

Guangzhou and Nanjing (grouped as warm cities) showed higher mean temperature than Shijiazhuang and Suwon (grouped as cold cities) (p=0.000). In each group, Nanjing was more humid than Guangzhou (p=0.000), and Suwon was more humid than Shijiazhuang (p=0.000).

On the visit day, volunteers were required not to apply any cosmetics on their face. Once volunteers visited, skin surface temperature was taken from the cheek and forehead without acclimatization. Then they were instructed to read the research protocol that explained the study procedure before giving informed consent. They were required to complete a questionnaire to assess demographic profiles and use of cosmetics. Then, they washed their face with warm water without any cleansers or soap, and measurements were performed by multiple non-invasive devices for the following parameters: hydration, transepidermal water loss (TEWL), sebum, elasticity, skin pore, wrinkle, and skin tone (brightness). All study protocols were implemented in an environment-controlled room (temperature, 22℃±2℃ and relative humidity, 50%±10%).

Skin surface temperature was measured using a non-contact infrared thermometer. Skin electrical capacitance was measured using a Corneometer CM825 (Courage+Khazaka Electronic, Cologne, Germany), shown in system-specific arbitrary units (AU). This parameter was used to estimate the hydration level of stratum corneum. Higher values mean higher hydration level. TEWL was measured using Vapometer (Delfin Technologies, Kuopio, Finland), expressed in g/m². Higher values mean more water loss through the epidermis. Measurements of hydration and TEWL were performed 3 times from the cheek and forehead, and the average was calculated for each region. Sebum output was measured using a Sebumeter SM810 (Courage+Khazaka Electronic GmbH), expressed in µg/cm². Higher values mean more sebum secretion. The collection was performed once, each from the cheek and forehead. Skin elasticity of the cheek was measured by Cutometer MPA580 (Courage+Khazaka Electronic GmbH), a suction chamber device. R2, Ua/Uf, is considered as the main parameter that represents the skin's ability to regain its initial position after deformation and was most often used in previous studies1. Values close to 1 mean high elasticity. Skin pore area, wrinkle, and skin tone (brightness) were measured using Janus (PSI, Seoul, Korea), expressed in a 100-percentage scale17181920. Photographs of faces were taken at 3 different angles (frontal, left lateral, and right lateral) with a 10 megapixel ultra-high resolution camera, and were analyzed using built-in image analysis software. The extent of pore widening was assessed from forehead, nose, and both cheeks, and the degree of skin wrinkles was assessed from the crow's feet. The average of red, green, and blue values of the cheeks, nose, eyes, and forehead was calculated and converted to a 100-point skin tone grading. Higher values mean more pore widening, severe wrinkles, and brighter skin.

Variations of biophysical parameters according to age are shown in Fig. 2 and Supplementary Table 1. Skin surface temperature, hydration, TEWL, and sebum excretion showed values of determination coefficient (R-squared, R²) less than 0.1 (Fig. 2A~D). Skin elasticity and tone decreased, and skin pore and wrinkle increased with age (Fig. 2E~H); among the 4 parameters, wrinkle showed the largest determination coefficient (R²=0.455) (Fig. 2F~H).

Variations of biophysical parameters according to region are shown in Fig. 3 and Table 2. Skin surface temperature of the cheek and TEWL were lower in Suwon than the 3 cities in China (Fig. 3A, C). Nanjing and Suwon, which are more humid in each temperature group, showed much higher hydration level (Fig. 3B).

Sebum output was higher in warm cities than cold cities (Fig. 3D). Skin elasticity and pore size were not significantly different among 4 cities (Fig. 3E, F). The severity of wrinkle was the mildest and skin tone the brightest in Suwon among the 4 cities (Fig. 3G, H).

All assessed cosmetic habits were significantly different between Chinese (Guangzhou, Nanjing and Shijiazhuang) and Korean (Suwon) women (Fig. 4, Table 3). Korean women significantly used sunscreen more frequently than Chinese women (Fig. 4A). Furthermore, Korean women used sunscreen with higher sun protection factor (SPF) and protection grade of ultraviolet A (PA) index compared to Chinese women (Fig. 4B, C). A majority of subjects were wearing base makeup daily, with a higher percentage in Korean women (base makeup means wearing a skin cover material, such as foundation, blemish balm cream, or compact) (Fig. 4D). The type of moisturizer was classified into 3 categories: 1) no use or toner only; 2) light moisturizers, lotion only or a combined use of toner and lotion; 3) heavy moisturizers, cream only, or a combined use of toner and cream, lotion and cream, or toner, lotion, and cream. Korean women used heavier moisturizers than Chinese women (Fig. 4E).

By performing multiple regression, the association between cosmetic habits and biophysical parameters (values measured in cheek were chosen) was assessed after adjusting for age and geographical region (Supplementary Table 2). Marginal mean was calculated and is shown in Fig. 5. Either subjects who used sunscreen sometimes or everyday showed significantly lower sebum excretion than subjects who never used sunscreen (p=0.028 and p=0.010, respectively) (Fig. 5A). Skin elasticity was significantly higher in subjects who used sunscreen everyday than subjects who never used sunscreen (p=0.037) (Fig. 5A). Subjects who wore base makeup daily showed higher hydration level and lower TEWL than subjects who did not (p=0.004 and p=0.005, respectively) (Fig. 5B). Either the subjects who used light moisturizers or those who used heavy moisturizers showed more hydration and lower TEWL than subjects who never used moisturizers or used toner only (p=0.001 and p=0.036 for hydration and TEWL, respectively, between light moisturizer group and no use or toner only group; and p=0.001 and p=0.001 for hydration and TEWL, respectively, between heavy moisturizer group and no use or toner only group) (Fig. 5C). However, the difference of hydration and TEWL between light and heavy moisturizers was not statistically significant (p=0.693 and p=0.05, respectively). Skin elasticity was significantly higher in subjects who used heavy moisturizers than either the subjects who never used moisturizers or used toner only, or those who used light moisturizers (p=0.002 and p=0.036, respectively) (Fig. 5D). Other biophysical parameters had no statistically significant correlation with cosmetic habits.