Mapping the Fine-Scale Aging Landscape of Young Chinese Women: A LOESS-Based Analysis of 45 Biophysical Skin Parameters Authors: Yumei Fan, Jiahuan Hu, Shuo Shan Published: 2026 ABSTRACT AbstractBackgroundFacial skin aging is a complex, progressive process, but subtle and nonlinear changes during young adulthood remain insufficiently characterized. This study aimed to develop an early-aging atlas of facial skin in young Chinese women and to identify key age-related turning points among the various aging traits.MethodsA total of 500 Chinese female volunteers aged 18–33 years participated in this study. We assessed 45 physiological skin parameters associated with barrier function, color, texture, pigmentation, and wrinkles using various noninvasive skin assessment techniques. Beyond Pearson correlation analysis, we used locally weighted scatterplot smoothing (LOESS) to model how these skin parameters evolve with age, capturing any nonlinear patterns.ResultsAge-related facial changes were detectable even within this young cohort. Sebum secretion and skin elasticity declined with age, whereas pigmentation, roughness, and wrinkle-related parameters increased. LOESS analysis suggested several approximate age-related transition points. Around 22 years of age, skin dullness and latent UV damage appeared to increase. Around 25 years, periocular wrinkles and mid-face roughness became more evident. After 27 years, brown spots showed a more apparent upward trend. Around 29–30 years, skin elasticity showed a more marked decline, accompanied by a secondary decrease in sebum secretion. These LOESS-indicated transition periods should be interpreted as exploratory rather than mathematically confirmed thresholds.ConclusionEarly facial aging in young Chinese women follows asynchronous and nonlinear trajectories across different skin dimensions. The identified transition points before age 30 provide a quantitative basis for understanding the temporal progression of early facial aging and may support age-specific prevention and intervention strategies.Keywords: skin aging, young Chinese women, skin physiological parameters, LOESS regression Introduction Skin aging is a complex biological process driven by both innate factors, such as genetics, metabolism, and hormonal shifts, and external influences like UV exposure, pollution, and lifestyle habits.1 This gradual decline is manifested on the face through notable variability and multidimensional changes, including widespread deterioration of skin structure and function. Clinically, facial aging is characterized by the appearance and deepening of wrinkles, skin laxity, decreased elasticity, uneven pigmentation, dullness, and compromised barrier function.2,3 Facial skin is continuously exposed to the external environment, making its aging characteristics particularly prominent and complex. Consequently, relying on a single parameter for assessment often fails to capture the full, dynamic nature of facial aging. Therefore, modern research on the skin increasingly employs quantitative phenotyping methods that utilize multiparametric, multidimensional, and noninvasive measurement techniques. Recent advancements in noninvasive measurement technologies now allow researchers to accurately quantify facial skin biophysical parameters and examine their association with aging. A significant body of cross-sectional studies has demonstrated distinct age-related patterns in various facial skin metrics: skin elasticity, luminance, and sebum secretion generally decline with advancing age, whereas wrinkles and hyperpigmentation tend to increase.4–7 Earlier studies often used broad age groups, such as 20–60 years or even wider ranges, or relied on coarse divisions such as 10-year intervals for comparison. While these approaches can identify overall aging trends, they often mask subtle yet important physiological changes occurring within narrower age ranges—particularly during young adulthood. This limitation hampers the ability to accurately determine the timing of early skin aging onset and how it evolves. Recent studies indicate that for women in East Asia, the third decade of life is a crucial period when notable skin aging changes may begin. Park et al,8 using an AI-based diagnostic system in Korean women aged 10–60 years, identified their 20s as a key period for facial aging, with notable increases in wrinkle and pigmentation scores during this age range. These findings suggest that microscopic aging processes may accelerate before visible signs emerge. An 11-year longitudinal study in Japan further supported this observation, demonstrating significant changes in individuals from ages 20 to 30 years in facial wrinkles, skin texture, hyperpigmented spots, and skin color, particularly lightness (L*) and yellowness (b*).9 Similarly, Zhang et al10 observed that in Chinese women, the number of facial aging features demonstrating significant changes sharply increased between ages 24 and 30 years, indicating the onset of an early accelerated aging phase. Notably, these phenotypic changes appear to have a molecular basis. Li et al,11 through multi-omics analyses, provided molecular evidence suggesting that the age of 30 years marks a significant transitional phase in the aging process among Chinese women. Overall, these findings suggest that changes in skin parameters between the ages of 20 and 30 years are likely not random but are instead influenced by programmed biological transitions consistent with the natural aging timeline. Mapping the nonlinear changes in skin parameters within this age range is therefore critical for accurately identifying the onset and mechanisms of early aging in Chinese women. This study aimed to systematically characterize early facial skin aging features in young Chinese women using noninvasive assessment methods. We recruited 500 female volunteers aged 18–33 years and measured 45 biophysical skin parameters, including indices of wrinkles, skin tone, elasticity, melanin, and hemoglobin. Instead of traditional linear regression, we employed locally weighted scatterplot smoothing (LOESS), which is better suited for detecting localized nonlinear changes in the age-related skin patterns. By developing detailed models of parameter changes, we sought to map the dynamic progression of facial skin features from individuals aged 18 to 33 years, identify the ages at which specific signs of aging first emerge, and provide a quantitative basis for early anti-aging strategies specifically tailored to Chinese women. Materials and Methods Participants A total of 500 Chinese participants aged 18–33 years were recruited from a Beijing-based internal volunteer database established and maintained over several years by the Chinese Academy of Inspection and Quarantine according to predefined inclusion and exclusion criteria (Box 1). Participants who met all eligibility criteria and provided written informed consent were enrolled until the target sample size was reached. Therefore, the sampling strategy was random sampling within a Beijing-based internal volunteer database, rather than consecutive or convenience-based sampling. The research complied with the guidelines of the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of the Cosmetic Technology Center of the Chinese Academy of Inspection and Quarantine. The study was conducted from March to April 2024. Box 1Inclusion and Exclusion Criteria for Participant Enrollment Were as FollowsInclusion CriteriaFemale subjects aged 18–33 years;Participants with self-reported facial perceived age consistent with their biological age;Individuals with routine daily lifestyles without special occupational/environmental exposures (for example, exposure to sunlight, night shift work, and others), or specialized cosmetic intervention;Participants who were in good health at the time of the test;Participants who voluntarily agreed to participate and signed the informed consent form;Participants with the ability to complete the specified procedures according to the requirements of the test protocol.Exclusion CriteriaWomen who were planning to become pregnant, or those who were pregnant, breastfeeding, or within 6 months postpartum;Participants with severe systemic diseases, immunodeficiency, or autoimmune diseases;Participants with a history of allergic diseases or who had experienced cosmetic-related allergic reactions within the past 1–2 years;Participants currently participating in other clinical or cosmetic tests or who have participated in such tests within the previous month;Participants with dermatological conditions in the test area or currently receiving dermatologic medication;Participants who had acute medical conditions (for example, a cold and flu) that could interfere with the test;Participants who were considered unsuitable for participation based on the clinical judgement of the dermatologist. Inclusion and Exclusion Criteria for Participant Enrollment Were as Follows Female subjects aged 18–33 years; Female subjects aged 18–33 years; Participants with self-reported facial perceived age consistent with their biological age; Participants with self-reported facial perceived age consistent with their biological age; Individuals with routine daily lifestyles without special occupational/environmental exposures (for example, exposure to sunlight, night shift work, and others), or specialized cosmetic intervention; Individuals with routine daily lifestyles without special occupational/environmental exposures (for example, exposure to sunlight, night shift work, and others), or specialized cosmetic intervention; Participants who were in good health at the time of the test; Participants who were in good health at the time of the test; Participants who voluntarily agreed to participate and signed the informed consent form; Participants who voluntarily agreed to participate and signed the informed consent form; Participants with the ability to complete the specified procedures according to the requirements of the test protocol. Participants with the ability to complete the specified procedures according to the requirements of the test protocol. Women who were planning to become pregnant, or those who were pregnant, breastfeeding, or within 6 months postpartum; Women who were planning to become pregnant, or those who were pregnant, breastfeeding, or within 6 months postpartum; Participants with severe systemic diseases, immunodeficiency, or autoimmune diseases; Participants with severe systemic diseases, immunodeficiency, or autoimmune diseases; Participants with a history of allergic diseases or who had experienced cosmetic-related allergic reactions within the past 1–2 years; Participants with a history of allergic diseases or who had experienced cosmetic-related allergic reactions within the past 1–2 years; Participants currently participating in other clinical or cosmetic tests or who have participated in such tests within the previous month; Participants currently participating in other clinical or cosmetic tests or who have participated in such tests within the previous month; Participants with dermatological conditions in the test area or currently receiving dermatologic medication; Participants with dermatological conditions in the test area or currently receiving dermatologic medication; Participants who had acute medical conditions (for example, a cold and flu) that could interfere with the test; Participants who had acute medical conditions (for example, a cold and flu) that could interfere with the test; Participants who were considered unsuitable for participation based on the clinical judgement of the dermatologist. Participants who were considered unsuitable for participation based on the clinical judgement of the dermatologist. Assessment of Skin Physiological Parameters To assess skin physiological properties, participants were allowed to acclimate for 30 min in a room maintained at 22 ± 1 °C with relative humidity of 50 ± 5% before the measurements were made. Various parameters—including skin moisture, transepidermal water loss (TEWL), sebum levels, melanin and hemoglobin indices, skin color (L*a*and b* values), and wrinkles—were recorded at seven anatomical sites for each participant (details are provided in Figure 1). Before facial imaging and skin assessments, participants were instructed to cleanse their faces and remove all makeup and excess sebum. Schematic diagram of test position. Blue area: skin sebum; Purple area: skin hydration, TEWL, and color; Red area: winkles; Orange area: skin gloss, melanin index, erythema index, and elasticity. The skin hydration level of the outer epidermis on the cheek was measured using a Corneometer (CM 825; Courage and Khazaka, Germany), which operates based on capacitance measurement. For each assessment, at least five readings were taken from different locations within the test area on the cheek. The TEWL, gloss index (including both the gloss value and gloss derived from diffuse spectral reflectance [DSC]), elasticity, and sebum index were measured using specific devices: Vapometer (SWL5201, Delfin, Finland), Glossymeter (GL200, Courage and Khazaka, Germany), Cutometer MPA 580 (Courage and Khazaka, Germany), and Sebumeter (SM815, Courage and Khazaka, Germany). Skin parameters were assessed at three different sites within the test area for each measurement, and the average of these readings was used for subsequent analysis. Two instruments, the CM2500d chromameter (CR-400, Konica Minolta, Japan) and the Mexameter (MX18, Courage and Khazaka, Germany), were utilized to assess skin color parameters, including L*×a*×b* values, the individual typology angle (ITA°), melanin index (MI), and erythema index. Wrinkle assessment was performed using Fast Optical in vivo Topometry of Human Skin (FOITS technique, AEVA-HE, IEO THCH, France). The FOITS test areas included multiple facial regions, focusing on under-eye wrinkles, dynamic lines at the eye corners, nasolabial folds, and cheek lines. To evaluate the dynamic patterns at the eye corners, subjects completed a standardized smiling test. The analysis measured parameters such as Ra and Rz, as well as metrics related to the number, area, volume, and length of wrinkles. Ra represents the average deviation from the skin’s centerline within each topographical profile, indicating overall smoothness. Rz reflects the average maximum peak-to-trough distance, demonstrating the depth of the deepest wrinkles. Facial imaging was performed with the VISIA complexion analysis system (Canfield Scientific, Inc., VISIA Generation 7). Standardized frontal photos were taken under three lighting conditions: white light, UV light, and cross-polarized light. Following the manufacturer’s instructions for positioning and calibration, images were captured to maintain consistent head alignment and lighting conditions across all subjects. The VISIA software then conducted quantitative analysis, generating objective scores for features such as spots, texture, pores, UV spots, brown spots, red areas, and porphyrins. Statistical Analysis All statistical analyses were performed using R software (version 3.6.0). Initially, the normality of continuous variables was assessed using the Shapiro–Wilk test. Data that followed a normal distribution are presented as mean ± standard deviation, whereas non-normal data are presented as the median with the interquartile range. Correlation analyses were conducted to examine the relationship between age and the 45 skin parameters measured. Pearson correlation coefficients were used for variables exhibiting a normal distribution, whereas Spearman’s rank correlation coefficients were applied to variables with a non-normal distribution. The correlation coefficient reflects both the strength and the direction of each observed relationship. To gain a clearer understanding of the dynamic trends and potential nonlinear changes in skin parameters between individuals aged between 18 and 33 years of age, we employed LOESS regression models.12–14 This non-parametric smoothing technique enabled visualization of the continuous variation of each skin parameter with age, thereby reducing outlier effects and highlighting subtle physiological shifts. The age-related transition points described in this study were identified by visual inspection of the LOESS-fitted trajectories, focusing on apparent changes in curve direction, slope, or acceleration across adjacent ages. All statistical tests were two-sided, and statistical significance was defined as p < 0.05. Results A total of 500 young Chinese female volunteers aged 18–33 years were included in this study, with a mean age of 25.54 ± 4.62 years. The baseline demographic and lifestyle characteristics of the study population are summarized in Supplementary Table 1, including age distribution, BMI, skin type, daily stress, sleep quality, daily outdoor activity status, and other relevant baseline variables. Chi-square analysis showed no significant differences among age groups in the distributions of BMI, daily stress, sleep quality, or daily outdoor activity status, with all p values > 0.05. These findings suggest that the cohort was generally balanced across age groups with respect to these baseline lifestyle-related factors. However, skin type distribution differed significantly among age groups (p = 0.005). Overall, combination-oily skin was the most common skin type, accounting for 36.80% of all participants, followed by combination-dry skin (23.20%) and dry skin (17.80%). With increasing age, the proportion of combination-dry skin tended to increase, whereas the proportions of oily and normal skin tended to decrease. Association Between Age and Facial Skin Characteristics Pearson correlation analysis was conducted to examine the relationship between age and the 45 facial biophysical skin parameters in 500 female subjects aged 18–33 years. The correlation coefficients and significance levels are demonstrated in Figure 2 and presented in detail in Supplementary Table 2. The results reveal that, even within this young age group, several skin parameters exhibit significant aging-related patterns. Heatmap of Pearson correlation coefficients between age and 45 facial skin biophysical parameters. The color scale represents the Pearson correlation coefficient (r), ranging from −1.0 (dark blue, strong negative correlation) to +1.0 (dark red, strong positive correlation). Asterisks (*) indicate statistical significance at the p < 0.05 level. The matrix illustrates the relationships between age and various skin dimensions, including barrier function (Hydration, TEWL, and Sebum), elasticity (R2, R5, and R7), skin color (L*, b*, and ITA°), and regional wrinkle parameters. Abbreviations: NF, Nasolabial Folds; CF, Crow’s Feet; CL, Cheek Lines; UE, Under-eye wrinkles. Barrier Function Concerning skin barrier parameters, facial sebum content demonstrated a significant negative correlation with age (r = –0.116, p < 0.01), indicating that sebum secretion tends to decrease between the ages of 18 and 33 years. In contrast, skin hydration (r = 0.041, p > 0.05) and TEWL (r = 0.060, p > 0.05) did not demonstrate any significant association with age. These findings suggest that the water retention capacity of the stratum corneum and barrier function remain relatively stable within this age range, with no clear evidence of age-related decline. Skin Texture Age-related skin changes were primarily characterized by a decline in elasticity and an increase in roughness. Elasticity measures such as R2 (gross elasticity), R5 (net elasticity), and R7 (biological elasticity) were all found to be significantly negatively correlated with age (p < 0.001), with correlation coefficients of –0.208, –0.173, and –0.206, respectively. This pattern indicates a gradual reduction in skin resilience and firmness as age increases. Conversely, R0 (distensibility under suction) demonstrated no significant association with age (p > 0.05). Furthermore, both pore size (r = 0.293, p < 0.001) and skin texture (r = 0.238, p < 0.001) increased significantly with advancing age, reflecting a progressive coarsening of the skin surface. Skin Color Analysis of skin color revealed complex age-associated alterations in chromophore distribution. The L* value (r = –0.124, p < 0.01) and ITA° (r = –0.180, p < 0.001) both demonstrated significant negative correlations with age, indicating a gradual shift toward darker skin tones as age increases. Conversely, the b* value, which measures yellowness (r = 0.122, p < 0.01), displayed a positively correlated, suggesting an increase in yellow pigmentation. Notably, the MI demonstrated a slight negative correlation with age (r = –0.115, p < 0.05), which contrasts with the trend observed in L*; this discrepancy may be explained by seasonal variations or differences in the contributions of hemoglobin and pigmentation relative to localized spots. Furthermore, gloss parameters (Gloss and Gloss-DSC) exhibited weak but statistically significant associations with age (p < 0.01). Pigmentation (Spots) Pigmentation buildup was identified as a key characteristic of the aging process. Notably, UV spots, indicating hidden pigmentation, demonstrated the strongest correlation with age (r = 0.254, p < 0.001), followed by brown spots (r = 0.150, p < 0.001) and overall visible spots (r = 0.131, p < 0.01). This pattern suggests that subclinical UV damage accumulates progressively within this age range, frequently preceding the appearance of noticeable hyperpigmentation. Wrinkles Wrinkle parameters generally increased progressively with age across all measured areas, including nasolabial folds [NF], crow’s feet [CF], cheek lines [CL], and under-eye wrinkles [UE]. For NF, aging was more strongly reflected by deepening rather than proliferation: severity indicators such as volume (NF_V, r = 0.326), roughness (NF_Rz, r = 0.397), and average roughness (NF_Ra, r = 0.464) demonstrated moderate-to-strong positive correlations with age (p < 0.001). In contrast, wrinkle count (NF_N) did not demonstrate a statistically significant correlation with age (p > 0.05). In the periorbital and cheek regions, all evaluated metrics, including count, volume, length, and area, demonstrated significant positive correlation with age (p < 0.01). Notably, cheek line volume (CL_V) and crow’s feet roughness (CF_Ra) displayed correlation coefficients exceeding 0.3, indicating that these areas may serve as potentially early indicators of structural skin aging. Overall, skin aging in individuals aged 18–33 years appears to involve a combination of multiple changes rather than just isolated features. Major trends include decreased sebum production, lower skin elasticity, increased skin roughness, and progressive skin yellowing and dullness. Importantly, the data demonstrate that deepening nasolabial folds and the accumulation of hidden UV damage are the most reliable indicators of aging in this young adult group. Nonlinear Aging Trends of Biophysical Parameters Trends in Skin Barrier Function with Age A LOESS regression model was used to analyze the nonlinear patterns of skin barrier parameters—sebum content, hydration, and TEWL—across the cohort (Figure 3). The results demonstrated that skin hydration remained relatively stable between the ages of 18 and 33 years, but demonstrated clear physiological inflection points (Figure 3A). Specifically, hydration levels followed a curve that peaked slightly at the age of 29 years before initiating a modest decline, suggesting that approximately 29 years of age may represent a LOESS-indicated transition period for stratum corneum moisture retention. TEWL remained quite stable, consistently demonstrating low levels (approximately 14–15 g/m2/h) (Figure 3B). In contrast, facial sebum levels tend to decline with age (details are in Figure 3C). There is a sharp decrease during early young adulthood, dropping from approximately 80 to 60 µg/cm2 between the ages of 18 and 22 years, probably due to sebaceous gland activity stabilizing after puberty. After this, sebum levels remain steady between the ages of 24 and 29 years, then decrease again after the age of 30 years. LOESS regression analysis of skin barrier function parameters versus age. The solid lines represent the local weighted regression smoothing (LOESS) trends for (A) skin hydration, (B) TEWL, and (C) sebum level across the group aged 18–33 years. The gray shaded areas indicate the 95% confidence intervals. Points represent the distribution of data. Trends in Color Level with Age LOESS analysis demonstrated that skin color parameters change along different paths, emphasizing the complexity of skin tone changes during early aging (Figure 4). The fitted curves for facial brightness (L*) and ITA° demonstrated strong consistency (Figure 4B and C). Both metrics remained relatively high from 18 to approximately 22 years of age and then began to decline gradually thereafter. This pattern indicates that the cumulative impacts of photoaging and decreased metabolic turnover may begin to manifest as early as the mid-20s. Conversely, the b* value demonstrated a steady, linear increase across the entire age range studied (Figure 4D). This trend suggests that “skin yellowing” begins early and is probably caused by internal oxidative stress and the gradual accumulation of advanced glycation end-products. The MI remained relatively stable across ages, fluctuating mainly between 130 and 150 arbitrary units, with no clear signs of increasing pigmentation with increasing age (Figure 4A). However, skin gloss parameters (Gloss and Gloss-DSC) exhibited a nonlinear rise-and-fall pattern: they increased significantly from the age of 18 years to peak around ages 27–28 years, then remained stable or slightly declined (Figure 4E and F). LOESS regression analysis of skin color and complexion parameters versus age. Nonlinear aging trends for (A) MI, (B) L*, (C) ITA°, (D) b*, (E) Gloss, and (F) Gloss with DSC. Points represent the distribution of data. Trends in Texture Level with Age Skin texture degradation was characterized by a specific temporal sequence: loss of elasticity followed by surface roughening (Figure 5). The LOESS curves for Cutometer parameters (R2, R5, and R7) revealed that mechanical aspects of skin aging begin early. R2 initially declined between 18 and 25 years, then stabilized briefly from 25 to 28 years of age, before sharply declining again after 29 years of age (Figure 5A). Similarly, R5 and R7 demonstrated clear downward inflection points beginning at 29 years of age (Figure 5B and C). These results indicate that 29–30 years of age may represent an approximate transition period during which skin elasticity begins to show a more pronounced decline. Facial pore visibility increased consistently with age (Figure 5D), suggesting a cumulative expansion likely worsened by the loss of perifollicular collagen support. Additionally, the texture parameter derived from images demonstrated a gradual and ongoing rise in surface irregularity (Figure 5E). LOESS regression analysis of skin texture and elasticity parameters versus age. Nonlinear aging trends for (A) R2, (B) R5, (C) R7, (D) pores, and (E) texture. Trends in Spot Level with Age The differentiation between latent and visible pigmentation revealed distinct patterns of accumulation (Figure 6). UV spots demonstrated the most significant age-related increase among pigmentation metrics (Figure 6B). Values rose rapidly from 18 to 22 years of age and nearly doubled over 15 years, from about 11 to 19. This pattern indicates that deep, sub-clinical photodamage accumulates rapidly during the early youth period, well before visible imperfections become apparent. In contrast, brown spots exhibited a “plateau-rise” pattern (Figure 6C). Their levels appeared relatively stable before approximately 27 years of age, after which the LOESS-fitted curve showed a more apparent upward trend. The total number of visible spots (Spots) increased steadily and linearly across the age range (Figure 6A). These findings reveal a time delay: deep UV damage occurs in the early 20s, whereas visible pigmentation increases faster after 27 years of age. This underscores the importance of photoprotection among young adults to prevent the accumulation of UV-induced spots. LOESS regression analysis of facial pigmentation types versus age. The solid lines represent the LOESS trends for (A) spots, (B) UV spots, and (C) brown spots. Trends in Wrinkle Level with Age The LOESS analysis of wrinkle parameters (N, L, A, V, Ra, and Rz) across four facial zones (CF, UE, CL, and NF) revealed significant regional differences in aging patterns (Figure 7). CF metrics, particularly N and L, remained low during a “silent period” (18–24 years) but demonstrated a notable change around 25 years, suggesting a transition from dynamic to static lines. The UE region displayed a delayed onset, with major structural alterations (volume and roughness) appearing later, around 28 years. Unlike the sudden onset in the eye area, the mid-face aged more gradually. The number and length of CL increased steadily and continuously from 18 years of age. Similarly, NF demonstrated a transition from shallow to deep forms. Although the fold count remained relatively stable, the volume (NF_V) and area (NF_A) increased progressively from age 18 years, with a sharper increase after 25 years due to soft-tissue sagging and gravity. LOESS regression analysis of wrinkle parameters across four facial regions. The matrix displays aging trends for wrinkle Count (N), Length (L), Area (A), Volume (V), and Roughness (Ra, Rz) in four key areas: Crow’s Feet (CF), Under-eye (UE), Cheek Lines (CL), and Nasolabial Folds (NF). Discussion This study conducts a comprehensive high-dimensional quantitative analysis of 45 biophysical skin parameters in a cohort of 500 young Chinese women (aged 18–33 years). It systematically reveals the asynchronous progression of early skin aging with unprecedented temporal resolution. Unlike earlier studies that primarily relied on global parametric models such as linear analysis or binomial fitting, this study applies a non-parametric LOESS model that does not rely on predefined functional forms. This flexibility enables the model to detect subtle transitions and abrupt inflection points in skin features at specific ages more effectively. This methodological refinement allowed us to precisely identify the tipping points of several key biological indicators, highlighting the complex nature of phenotypic changes during young adulthood. This construction of this detailed atlas addresses a long-standing gap in understanding subclinical aging in younger groups and provides an important temporal reference for designing targeted preventive strategies. Dynamic Evolution of Facial Barrier Function Between the Ages of 18 and 33 Years This study revealed a clear phased pattern in facial sebum levels: an initial rapid decline from 18 to 24 years of age, followed by a plateau, and a subsequent decrease after 30 years. This progression is consistent with the natural physiological decline of sebaceous gland activity as sex hormone levels stabilize following puberty. These findings are highly consistent with large-scale epidemiological data from Chinese cohorts,5,15,16 further emphasizing that the third decade represents a critical transition from peak sebaceous gland activity to functional decline. Notably, despite a pronounced decline in sebum levels, TEWL remained remarkably stable throughout the group and did not demonstrate a strong association with age. This observation is consistent with findings from Luebberding et al,17 who observed that barrier integrity generally remains preserved even when surface lipid levels decline. Despite overall macroscopic stability, aging processes continue beneath the surface. Recent lipidomic studies18 support this, demonstrating that even with normal TEWL levels, the skin’s lipid composition changes markedly around 28 years of age. Specifically, these changes include depletion of ceramides and triglycerides, accompanied by an increase in cholesteryl esters. These findings indicate that between ages 18 and 33 years, the skin barrier gradually transitions from a state of being “optimally intact” to one of “hidden fragility” which remains partially compensated. These insights highlight the importance of targeted actions, such as proactively adding intercellular lipids, to strengthen the barrier before clear clinical signs become apparent. Early Evolution of Skin Color and Pigmentation This use of LOESS fitting models in combination with correlation analysis demonstrates that changes in skin color and pigmentation are among the earliest signs of aging in young women. We observed a notable decline in skin brightness (L* value) beginning at around 22 years of age. This finding is consistent with the observations of Miyamoto et al9 and Zhang et al,10 who described skin aging in East Asian women as a gradual process characterized by darkening and yellowing. Mechanistically, the decrease in L* values serves as an important phenotypic indicator, demonstrating that the skin’s natural antioxidant defenses are being increasingly overwhelmed by accumulated oxidative stress. Previous photobiological studies have demonstrated that the loss of skin radiance is strongly associated with ROS-triggered melanogenesis and the buildup of carbonylated proteins. These processes become more prominent when environmental stressors exceed the skin’s innate neutralization ability.19–21 Simultaneously, skin yellowness, indicated by the b* value, begins to increase steadily as early as the age of 18 years. Notably, our LOESS model demonstrates that the inflection point for the b* value occurs much earlier than the age of 26 years reported by Huang et al.5 This difference implies that nonlinear modeling is more effective at detecting subtle early metabolic changes that linear methods might miss. Biologically, this early yellowing is mainly caused by the quick buildup of specific chromophores in the epidermis. Laughlin et al22 provided direct evidence from a study of young women (aged 20–29 years), demonstrating that individuals with duller complexions exhibited significantly higher levels of advanced glycation end products (AGEs), particularly carboxymethyl lysine (CML), in their epidermis. Additionally, the accumulation of novel yellow chromophores (AGEY)23 and stress-related endogenous bilirubin24 within the epidermis offers a mechanistic explanation for the persistent increase in b* values after the 18 years of age. Overall, this early-stage yellowing indicates active epidermal oxidative-glycation responses, which occur before the chronic glycation of dermal collagen typically seen after 26 years of age. This study further clarifies how pigmented spots evolve through two distinct biological phases. The first, the “Accumulation Phase” (ages 18–22 years), demonstrates a rapid rise in deep UV spots, while superficial brown spots stay relatively unchanged. This indicates that UV damage accumulates silently in the deeper skin layers before becoming visible. The second, the “Manifestation Phase” (after 27 years of age), involves a quick appearance of these latent damages on the surface, leading to a notable increase in visible brown spots. These results not only reinforce but also refine previous findings that the ages of 20 to 30 years mark a vital transition for pigmentation changes.5,8 This offers a scientific explanation for the widely observed sudden appearance of spots reported by women around age 30 years. Based on this timeline, our study highlights that the period between 18 and 22 years old is a prime golden window for photoprotective measures and antioxidant treatments. Applying focused interventions during this time to address hidden ultraviolet damage can effectively help prevent the formation of long-lasting pigmentary problems. Phased Evolution of Skin Elasticity and Texture This study integrates correlation analysis with LOESS fitting models to characterize the nonlinear patterns of skin elasticity and texture decline with age. Our findings demonstrate that skin elasticity parameters (R2, R5, and R7) are significantly negatively correlated with age (p < 0.001), strongly consistent with previous studies on the age-related reduction of skin biomechanical properties.1,25–27 From a microstructural standpoint, this notable reduction in elasticity has a clear histological foundation. Multiphoton scanning laser biomicroscopy studies have revealed that dermal elastic fibers gradually degenerate with age, demonstrating curling, fragmentation, and thinning.28 Additionally, previous research has indicated that the elastin score—which indicates fiber integrity and tortuosity—correlates positively with the R7 index. These results imply that the macroscopic decrease in elasticity directly reflects the structural deterioration of the deep dermal scaffold. More biologically significant is that the LOESS model accurately mapped a nonlinear pattern of skin elasticity: it stays nearly unchanged until 29 years of age, then drops sharply in a “cliff-like” manner between 29 and 30 years of age. This finding is strongly consistent with molecular insights from Li et al,11 who used multi-omics analysis to identify the age of 30 years as a crucial transition point in Chinese women’s aging process, characterized by significant changes in extracellular matrix-related gene expression. Our macroscopic phenotypic data provide concrete physical evidence for this molecular shift, confirming that the age of 30 years marks the pivotal moment when the dermal collagen and elastic fiber network switches from a “maintenance phase” to an “accelerated degradation phase”. Regarding skin texture, we identified distinct patterns of evolution in different regions. While wrinkle parameters in the periorbital and cheek areas increased notably with age—confirming that the third decade is a crucial period for wrinkle formation8—the NF region demonstrated a different trend. In this area, aging mainly involved an increase in wrinkle depth (volume/roughness) rather than the number of wrinkles. This indicates that early aging of the mid-face involves deeper tissue changes, such as fat pad displacement or the breakdown of retaining ligaments, rather than just superficial epidermal folding. Additionally, the LOESS model demonstrated that the acceleration in surface texture deterioration—such as the formation of crow’s feet and more roughness in the mid-face—begins at 25 years of age, consistent with findings by Zhang et al10 and Huang et al.5 This study introduces a “five-year lag” model of skin aging: superficial textural decline begins at the age of 25 years. Deep dermal elastic network remains intact until around the age of 30 years, when systemic collapse occurs. These insights offer a clear timeline for dermatological treatments. Early on, skincare should emphasize antioxidant protection and barrier health at the age of 25 years. When approaching 30 years of age, a shift to deep structural repairs—such as pro-collagen peptides or elastic fiber protectants—is crucial to address the rapid deterioration of the dermal matrix. Study Strengths and Limitations The primary strength of this study lies in its focused investigation of individuals aged 18–33 years, a demographic often overlooked in pre-aging research. With a substantial sample size of 500 participants and comprehensive quantitative assessments across 45 biophysical skin parameters, this study addresses a significant gap in understanding the microscopic aging process of young skin. The use of LOESS nonlinear modeling provides an exploratory methodological advantage over traditional linear regression by allowing visualization of nonlinear age-related trajectories without imposing a predefined functional form. Based on visual inspection of the LOESS-fitted curves, this study suggested several approximate age-related transition periods, including around 22 years for skin color changes and latent UV damage, around 25 years for periorbital and mid-face texture changes, around 27 years for the more apparent increase in brown spots, and around 29–30 years for the more marked decline in skin elasticity. These findings provide a useful age-specific framework for developing targeted skincare and early anti-aging strategies. However, this study has certain limitations. First, as a cross-sectional design, it captures population trends but cannot definitively distinguish true longitudinal aging changes from cohort effects. Moreover, the proposed age-related transition points were based on visual interpretation of LOESS-fitted curves rather than formal mathematical change-point detection. Although LOESS modeling is useful for exploratory visualization of nonlinear patterns, it does not by itself provide statistically defined thresholds. Therefore, the ages identified in this study should be interpreted as approximate LOESS-indicated transition periods rather than definitive biological cut-off points. Future longitudinal studies should apply segmented regression, derivative-based approaches, or other formal change-point methods to validate the exact timing and biological significance of these transition points. Second, although the focus was on biophysical phenotypes, it did not account for skin aging exposome factors such as urban pollution and diet (glycemic load), factors known to accelerate aging in young adults. In addition, previous aesthetic procedures were not specifically excluded or systematically controlled for in this study. Such procedures, including laser or light-based treatments, radiofrequency procedures, injectable fillers, botulinum toxin injections, and chemical peels, may affect skin biophysical parameters such as pigmentation, wrinkles, elasticity, texture, and barrier-related variables. Therefore, their potential confounding effects cannot be fully ruled out. Future studies should collect detailed histories of aesthetic procedures and account for these factors in the study design or statistical analysis. Future studies should include these external variables to develop a more comprehensive genotype-phenotype-exposome aging model. Third, our results are based on macroscopic biophysical measurements. While aligning with Li et al’s transcriptomic data, the study did not perform concurrent lipidomic or proteomic analyses on the same subjects to validate the proposed molecular mechanisms. Lastly, participants were recruited solely from Beijing. Since climate significantly affects barrier function, these findings may not be fully applicable to populations in high-humidity areas like Southern China, highlighting the need for multi-center validation. Conclusion This study established a fine-scale map of early facial aging in young Chinese women aged 18–33 years using 45 biophysical skin parameters and LOESS regression analysis. The results showed that early facial aging is asynchronous and nonlinear across skin dimensions. Skin dullness and latent UV damage appeared around 22 years of age, periorbital wrinkles and mid-face roughness became more evident around 25 years, brown spots increased after 27 years, and marked declines in elasticity and sebum secretion occurred around 29–30 years. These findings provide a quantitative basis for understanding the temporal progression of early skin aging in young Chinese women. Acknowledgments The authors thank the Cosmetic Technology Center of the Chinese Academy of Inspection and Quarantine for their support in participant recruitment and noninvasive skin biophysical measurements. The authors also thank Home for Researchers editorial team (www.home-for-researchers.com) for language editing service. Data Sharing Statement The data used and/or analyzed during the current study are available from the corresponding author on reasonable request. Ethics Approval and Informed Consent This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Cosmetic Technology Center of the Chinese Academy of Inspection and Quarantine on February 26, 2024 (Protocol Reference Number: 2024-216-ZH-002). This ethics committee was used because the study was conducted as a collaborative project, and participant recruitment, informed consent procedures, and all noninvasive skin biophysical measurements were performed by the Cosmetic Technology Center of the Chinese Academy of Inspection and Quarantine. All procedures were performed in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments. Informed consent was obtained from all participants. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors declare no conflicts of interest in this work. References Associated Data Data Availability Statement The data used and/or analyzed during the current study are available from the corresponding author on reasonable request.