Relationships between Skin Structure and Skin Function of Pregnant Women and Their Infants: A Prospective Cohort Study Open Access

There are periods in life when skin health is challenged [1]; for women, these include pregnancy and postpartum periods, and for infants the first months of life. Pregnant women often face a wide range of skin changes such as pigmentary, vascular, nail, and connective tissue alterations [2‒6]. Some of these effects are also assumed to continue after childbirth [7]. However, evidence about changes in the structure and function of the skin during this period is largely lacking.

On the contrary, the skin barrier maturation in infants is well described [8]. Evidence indicates that water retaining and transport mechanisms during the first year of life are very different to adults [9]. The skin surface pH in infants gradually declines indicating skin barrier maturation [8, 10‒12]. Changes regarding transepidermal water loss (TEWL) are less comparable across studies in infants and partly contradicting [8, 13‒15]. Stratum corneum hydration (SCH) in infants seems to vary based on age and body location. In Yonezawa et al.’s [14] study, baseline measurements of infant’s SCH averaged 23.2 (standard deviation [SD]: 7.9) arbitrary units (AUs) in the first week of life, as measured on the upper inner arm and thigh. Median SCH on the forehead was 45.7 (IQR: 32.2–58.2) AU in infants at the age of 2 weeks and even higher at the cheek (58.8 [IQR: 47.1–66.6] AU) [16]. Höger and Enzmann [12] showed that SCH increased after birth at all skin areas and remained stable after 30 days, i.e., for the volar forearm on a level of 80 AU.

While epidermal thickness (ET) in older adults is assumed to decrease with age [17], ET measurements in infants are less frequently reported and show greater variability across reports [18‒20]. Skin roughness parameters in infants show decreasing values within the first 4 weeks of life and then remain stable for up to day 90, indicating a smoothing of the skin surface [12]. In addition, most of the results are based on cross-sectional studies, not considering the possible changes over time. The overall aims of this study were to measure skin characteristics of pregnant and postpartum women and their infants over time and to explore whether maternal health and skin characteristics are related to the skin characteristics of infants.

In a prospective cohort study, we followed women during pregnancy and postpartum and their infants from 4 weeks until 6 months of age. A detailed study protocol was previously published [21].

The study visits were conducted in Berlin (Germany) at the Clinical Research Center for Hair and Skin Science at the Department of Dermatology, Venereology and Allergology, Charité-Universitätsmedizin Berlin from March 2021 to September 2023.

Pregnant women aged 18–45 years, beyond their first trimester, residing in Berlin or Brandenburg, Germany, without any skin conditions were eligible for inclusion.

The wide range of demographic, health, and skin-related exposures and outcomes collected are fully described in the study protocol [21]. In brief, noninvasive, standardized skin measurements were performed on women during pregnancy (visit 1) and on both, mothers and their infants at 4 weeks (visit 2) and 6 months postpartum (visit 3). Skin structure was measured by means of skin surface topography (roughness RA, RZ), ET, skin stiffness (total extensibility Uf, mm), and skin elasticity (Ur/Uf in %). Skin function was measured by means of TEWL, SCH, and skin surface pH, which are established parameters to characterize the skin barrier function. The definitions of all variables are shown in the online supplementary Table A (for all online suppl. material, see https://doi.org/10.1159/000546770). Skin dryness was measured by the Overall Dry Skin Score [22].

To address potential sampling bias from recruiting in a single geographical area, we compared the included women’s characteristics with national data regarding the age of the participants. Furthermore, we compared the infant’s characteristics with national data regarding birth weight.

Due to the exploratory character of this study, a formal sample size calculation was not performed. We aimed to include 100 pregnant females and their infants to obtain means and proportions with acceptable ranges of uncertainty. A loss-to-follow-up of one-third was assumed resulting in a total of n = 150 mothers to be included.

Demographic characteristics were described using absolute and relative frequencies, means, and SDs. Metric outcomes were described using means and SD per group and time point. Frequencies and distribution were calculated for ordinal scaled variables. Associations of skin parameters between women and their infants were described using Pearson’s correlation coefficients (r), whereas r values from ±0.3 to ±0.5 were regarded as medium, values higher than 0.5 or lower than −0.5 as large [23]. Skin characteristics of women at visits 1 and 2 and infants at visits 2 and 3 were correlated for analyzing possible associations between women’s and infants’ skin characteristics.

Out of 109 screened pregnant women, all were included and attended the first visit. Of those, 95 mothers and their infants attended the second visit and 100 couples attended the third study visit. Reasons for not attending the study visits were illness of the infant, hospital stay of the infant, very hot temperature during summertime, or moving away from Berlin. Participation at each stage of the study is shown in Figure 1.

Demographic characteristics of included women are shown in Table 1. Mean age was 33 (SD: 4.2) years. Most of the women lived in an urban setting (95.4%). Seventy women delivered by vaginal birth, 26 by cesarean section, and in 13 cases, no information on delivery method was provided. The study cohort comprised 39 female and 56 male infants, born at a mean gestational age of 39.9 (SD: 1.14) weeks (Table 1).

Skin functional and structural parameters of women are shown in Table 2. TEWL ranged from visit 1 to visit 3 from 9.5 to 13.8 g/m²/h. SCH ranged from visit 1 to visit 3 from 31.6 to 38.5 AU. Skin surface pH remained relatively stable, ranging from 5.0 to 5.3. ET ranged between 110.4 μm and 107.6 μm. RA changed from 9.6 μm to 11.0 μm from visit 1 to visit 3. RZ slightly increased from visit 1 from 39.8 μm to visit 3 to 44.7 μm. Skin total extensibility (Uf, mm) remained relatively stable from visit 1 to visit 3 (Uf: 0.31 to 0.28 mm), whereas skin elasticity (Ur/Uf in %) changed from 0.79 to 0.68.

Skin functional and structural parameters of infants are shown in Table 2. TEWL remained relatively stable from visit 2 to visit 3 (11.2 g/m²/h and 12.0 g/m²/h). SCH increased from 36.8 AU at visit 2 to 41.2 AU at visit 3. The pH value at visit 2 was 4.86 and 4.99 at visit 3. Skin structure parameters of infants remained stable over the study period. ET did not change; at visit 2, it showed 129.4 μm, and at visit 3 it showed 129.7 μm. Skin roughness and skin elasticity parameters remained unchanged (RA: 12.0 μm to 11.8 μm) (Uf: 0.29 mm to 0.27 mm; Ur/Uf: 0.56 to 0.62), respectively. The Eczema Area and Severity Index (EASI) showed 0.15 s at visit 2 and 0.21 at visit 3. At visit 2, 63% of the infants had dry skin at head/neck region, mostly assessed as mild dryness. Dry skin at the extremities and trunk area was also present in about half of the infants. At visit 3, the percentage of dry skin was lower, and 12.2% up to 40.8% of the infants had dry skin, with the highest percentage in the head/neck region and at the lower extremities. All ODS results are presented in online supplementary Table B, C; Figures 2 and 3.

Correlation coefficients between skin characteristics of women at visit 1 and infants at visit 2 ranged from −0.22 to 0.41 (online suppl. Table D). Medium positive correlations existed between women’s skin roughness parameters RA and RZ and infants’ skin stiffness and skin elasticity.

Correlation coefficients between skin characteristics between women at visit 1 and infants at visit 3 ranged from −0.35 to 0.35 as shown in online supplementary Table E. Medium positive correlations existed between women’s skin roughness parameters RA and RZ and infants’ skin elasticity. There was a medium correlation between women’s skin stiffness and skin elasticity and infants’ skin elasticity. There was a negative medium correlation between women’s skin stiffness and infants’ pH of r = −0.35.

Correlation coefficients between skin characteristics of women at visit 2 and infants at visit 3 are shown in the online supplementary Table F. Medium correlation between women’s skin stiffness and skin elasticity and infants’ skin stiffness and skin elasticity ranged from 0.31 to 0.45. A medium correlation of r = 0.37 was observed between ET and TEWL.

Women’s skin functional characteristics showed small changes over the study period. TEWL changed from 9.5 to 13.8 g/m²/h and was higher during pregnancy and postpartum. SCH values ranged from 31.6 to 38.5, and pH values ranged from 5.3 to 5.0. All in all, these values showed small alterations in terms of biological variation, so the changes were regarded as clinically not relevant. Skin structure characteristics were relatively stable during pregnancy and 6 months after birth. Skin became rougher during pregnancy and postpartum.

Skin characteristics in infants showed that skin function at 4 weeks of age revealed as fully intact skin barrier, and TEWL, SCH, and pH values changed slightly until sixth month of age. Skin structure characteristics remained stable over the study period. Also, comparable to the results of the women, these values show small alterations without clinical relevance. Out of six skin characteristics measured, only skin stiffness and skin elasticity in women were moderately associated infants’ skin stiffness and skin elasticity during the study period.

A key strength of our study was the strict use of standardized operating procedures for all skin measurements with extensively trained study personnel. These procedures, developed and validated over many years, ensured highly comparable reproducible results across all measurement timepoints. Compared to other studies, we achieved a sample size that allowed us to show robust results. Therefore, recruitment was extended to two and half years, longer than anticipated, due to COVID-19 pandemic restrictions affecting study participation. Our very low dropout rate suggests no potential for attrition bias. Temporal technical issues with Visioscan^® VC98 resulted in some missing data, but these occurred by chance and had no confounding effect on the results.

According to our data, children’s skin tends to be dry in the first few months of life, but the degree of dryness decreases by the sixth month after birth. This can certainly be explained by physiological skin changes, but also by other factors that we have not measured. These probably include the change of seasons. In future studies, seasonal changes can be taken into consideration, but in this first explorative cohort study, this aspect was not analyzed.

Gallagher et al. [24] observed pregnant women at three times during pregnancy and two times after giving birth and reported a steady rise of TEWL until the second month after birth. Maximum median TEWL estimates were 20 g/m²/h; only at month two after birth TEWL was higher, but “normalized” until month six after birth [24]. They concluded that “largely the skin barrier remained clinically stable over a 11-month period of high hormonal activity” (p 159 [24]). An Australian study showed comparable mean TEWL estimates in pregnant women also with a small rising trend from 12.6 (SD: 3.92) to 14.2 (SD: 7.63) during pregnancy from week 18 to week 38. They concluded that the observed changes in TEWL may not have a clinical relevance [25]. Comparing pooled TEWL estimates of healthy humans at the same body site (6.8; 95% CI: 6.5–7.0 g/m²/h) [26], TEWL estimates of pregnant women in our study seemed to be a little bit higher, but not “out of range.”

SCH results in women showed a moderate increase from visit 1 to visit 3. Compared to healthy adults, the SCH was comparable after pregnancy. In a group of twelve young adults (n = 11 female, n = 1 male) (mean age 32.9, SD: 7.2 years), SCH values ranged from 37.0 (SD: 8.1) to 41.5 (SD: 6.6), also measured on the inner volar forearm [27]. Pooled data of SCH in adults at the volar forearm showed a result of 36.62 (95% CI: 34.14–39.10) showing a high similarity to our study results [28]. Overall, the increasing SCH might be associated with the observed TEWL increase.

From visit 1 to visit 3, skin pH values in women only showed a small change. Stevens et al. [25] reported similar results of skin surface pH during pregnancy; they observed that at week 18–24 pH values were 5.26 (SD: 0.60), and at week 36–38 it was 5.35 (SD: 0.58). Compared to healthy adults, the pH values were highly comparable. In a group of twelve young adults (n = 11 female; n = 1 male) (mean age 32.9, SD: 7.2 years), pH values ranged from 4.7 (SD: 1.0) to 5.2 (SD: 1.0) measured also on the inner volar forearm [27]. Summarizing the skin function characteristics of women, we found small changes over time, but these were not regarded as clinically relevant due to their limited range.

Skin properties measured by Boyer et al. [7] at the abdomen and at the thigh area in pregnant and nonpregnant women indicate that there were no differences regarding structural elasticity. While the total extensibility was lower in pregnant women at the abdomen side compared to nonpregnant women and in the postdelivery group, at the thigh area there were no differences between the three groups. In detail, the results of Boyer et al. [7] showed comparable data to our results: pregnant women in our study had a total extensibility of the skin of 0.31 (SD: 0.06) (Uf, mm), compared to 0.300 (SD: 0.063) (abdomen) and 0.337 (SD: 0.063) (thigh) [7]. Structural elasticity in our study remained stable and showed a very small decline from pregnancy to 6 months after birth similar to Boyer et al. [7] measured for abdomen area and thigh area [7].

During the study period, ET in women remained stable. Comparing our results to the pooled estimates of ET data measured by OCT of Lintzeri et al. [17], the ET estimates in our study were higher. Lintzeri et al. [17] showed an ET of 83.3 μm (71.5 μm–95.1 μm) on the volar forearm in healthy humans. There is potentially a little influence of pregnancy or postpartum period regarding ET, with rising values.

RA remained nearly stable from visit 1 to visit 3. RZ slightly increased from visit 1 to visit 3. Due to limited data in pregnant women, we compared results with a group of 11 healthy women and one man of similar age (mean age 32.9; SD: 7.2) [29]. This comparison group showed lower skin roughness results (RA) on the four measurement points of volar forearm between 6.7 (SD: 0.8) and 7.6 (SD: 1.1). Similarly, the RZ values in same sample group were lower (31.9 [SD: 4.1]; 36.4 [SD: 4.9]) [29]. Summarizing the skin structure characteristics of women, we only found small changes over time. RA, RZ, and ET were probably influenced by pregnancy and postpartum, but due to the low range, these changes were not regarded as clinically relevant.

Pregnancy causes significant physiological changes in women, particularly affecting the skin [2, 3]. These alterations may manifest as dermatoses or exacerbate preexisting conditions. Hormonal fluctuations during pregnancy play a crucial role in altering the skin barrier function and structure. While the effects of estrogen on skin have been extensively studied in the context of menstrual cycles and postmenopausal changes [30], research specifically focusing on pregnant women remains limited. Studies described an increase in skin surface lipids in postmenopausal women following hormone replacement therapy, proving the beneficial influence of estrogen [30]. This finding may indirectly suggest similar outcomes in pregnant women given that estrogen levels peak during pregnancy. The stable TEWL measurements observed throughout pregnancy and postpartum could be attributed to the protective character of estrogens, including enhanced water-binding capacity of the stratum corneum and the dermis, as well as improved skin elasticity and roughness.

From a study which evaluated dermal thickness across four groups: women with spontaneous ovulatory menstrual cycle (group I), those using one-phase contraceptives (group II), three-phase contraceptive users (group III), and pregnant women (group IV), the results showed that pregnant women exhibited increased skin thickness prepartum compared to 6 weeks postpartum [31]. The authors concluded that the hormone-induced water retention could explain these differences. The comparability of these results was limited due to different measurement techniques and due to different definitions of skin thickness. Eisenbeiss et al. [31] measured skin thickness including the dermis, whereas we only measured the ET. So, potentially, there is a certain degree of imprecision in this comparison.

TEWL values in infants depended strongly on anatomical site and time of measurement after birth [8, 13]. Compared to Stamatas et al. [15], who measured at the volar forearm and where TEWL values in the group of infants were even higher (about 15 g/m²/h), TEWL values in our cohort were about 11 (SD: 6.24) g/m²/h 4 weeks after birth and about 12 (SD: 6.58) g/m²/h at 6 months of age. In the cohort of Stamatas et al. [15], these values reached in the group of infants by the age of 7–10 years and in the adult group. In the longitudinal study of Chittok et al. [32], TEWL values for infants in week four were about 15–20 g/m²/h, a little bit higher as in our study cohort. But TEWL measurements were measured at the volar forearm, the right antecubital fossa, and the right thigh and then presented as mean. As the mean of data, three anatomical sites were reported, and there are chances of bias while comparing TEWL values with this study.

Previous studies regarding pH values for infants provided more homogenous results; pH value of 4.86 (SD: 0.37) at week four was similar to the results of Hoeger and Enzmann [12] and Theunis et al. [10], who examined healthy infants. Regarding the infants’ results at month six, similar pH values (4.99) compared to women (5.02) were found. Also compared to Fluhr et al. [33], who evaluated the alteration of pH values in infants in different age groups, our study results suggested that alteration in pH values in infants might be restricted to the first weeks of life as mentioned by Ludriksone et al. [8] and Fluhr et al. [33].

TEWL and pH values have been shown to be age and body site dependent in infants, and SCH seemed to be equal on each body site [12]. Similar to pH changes in infants, SCH stabilized 30 days postpartum, albeit at a higher level compared to 3 days after birth [12]. Summarizing the skin function characteristics of infants, we found small changes over time. However, the results are fully comparable with other studies involving healthy infants.

Skin structure parameters of infants remained stable over the study period. ET did not change. It was measured 129.4 μm (SD: 13.67) at visit 2 and 129.7 μm (SD: 13.21) at visit 3. Compared to adults, infants’ epidermis seems to be thicker [18]. Vitral et al. [19] measured ET at the forearm of healthy infants at day 1 in their life to be 172.4 μm (SD: 19.6) (95% CI: 169.8–175.1). These results probably overestimate ET, because high-frequency ultrasonography was used [17]. Kakasheva-Mazenkovska et al. [20] measured ET of 109 μm by biopsies in a sample of children between 0 and 1 year. Due to the heterogeneity of time points and measurement methods, the comparison of these results is limited.

Skin roughness parameter RA and RZ of infants remained stable over the study period (RA: 12.03 μm to 11.84 μm; RZ: 43.72 to 41.50). Studies which evaluated directly after birth showed remarkable alterations in the first days of life. Hoeger and Enzmann [12] showed a decline in skin roughness parameters in the first weeks of life, indicating a smoothening of the skin surface. After 4 weeks, these parameters remained stable, like our results from 4 weeks to 6 months of age. Hoeger and Enzmann [12] measured the skin roughness by a mechanical sensor system, not by Visioscan, and combined data from four body sites in their calculations, therefore limiting the comparison. Trojahn et al. [34] calculated roughness estimates in a group of infants up to 2 years and found RA to be 6.1 μm (SD: 1.1) and RZ to be 29.6 μm (SD: 6.5) at the volar forearm. So, our results indicate higher roughness estimates at both visits.

Skin stiffness [35] and skin elasticity (Ur/Uf) estimates of the infants remained stable from visit 2 to visit 3. Compared to results of Visscher et al. [36], Ur/Uf estimates seem to be higher whereas Ur estimates seem to be lower. Visscher showed Ur/Uf estimates of 0.36 (SD: 0.04) measured at the arm and Uf of 1.26 (SD: 0.15) at the arm. Visscher et al. [36] used a Cutometer 575 (Courage+Khazaka electronic GmbH, Koln, Germany), but different from our settings, 6-mm probe aperture, 2 s 200 mbar negative pressure, and 2 s relaxation [36]. Therefore, the results can only be compared to a limited extent due to different reactions of the skin to the measurement settings of the Cutometer.

According to the ODS assessment of infants at visit 2, dry skin at head/neck region was present in 63% of the infants. Including the head/neck region, dryness at the lower and upper extremities and at the trunk (which occurred in less frequencies) was mainly in mild form. At visit 3, the percentage of infants with dry skin was in general lower compared to visit 2. Comparable to our results, Hoeger and Enzmann [12] showed in their study that neonatal skin was relatively dry and rough as compared to that of older infants. Summarizing the skin structure characteristics of infants, we found small changes during the study period, with elevated skin roughness and mild skin dryness.

Our convenient sampling method may have introduced a selection bias compared to a randomized sampling method. When considering women’s mean age, the mean age of the included women was 33 years, which aligns well with German national statistics. While average age of first-time mothers in Germany was 30.3 years, the average age for second and subsequent children ranges between 32 and 35 years. Given that our study included both first-time mothers and women with previous children, our sample’s mean age is consistent with national demographic patterns [37]. Regarding the mean birth weight and gestational age at birth, also the infants were representative for the infant population in Germany [38]. According to this comparison, our study results might be generalized to pregnant women and infants from 4 weeks to 6 months of age in Germany.

We performed the first prospective cohort study in mother and infants on skin characteristics and function from pregnancy over 4 weeks postpartum to 6 months of age. Skin characteristics of pregnant and postpartum women revealed a stable skin status with a mild to moderate increase of TEWL and skin roughness. Thus, pregnancy and postpartum period affected those skin parameters, whereas SCH, pH, skin stiffness, and skin elasticity were not affected.

At 4 weeks postpartum, the infants’ skin barrier and function characteristics were consistent with previously reported values for healthy infant skin during the first 6 months of life. However, our cohort had increased skin roughness after 4 weeks and decreased skin roughness after 6 months compared to previous studies.

Since the correlation analysis between women’s and infants’ skin barrier characteristics and function revealed no or only a weak correlation between pairs, there is only a restricted association of the maternal skin characteristics to infant skin barrier status and development over time. It seems more important to take care of infant’s individual status independent from their mother’s skin characteristics.

Based on this observational study, no broad general recommendations can be drawn during pregnancy and after birth and infants. However, based on our findings it may be justified to consider using skin care for maintaining barrier quality and function.

• 1.

  In pregnant women with a positive effect on roughness and TEWL

• 2.

  In infants improving dry skin and roughness.

This first prospective cohort study provided evidence of the skin barrier status and function in mothers and their infants during pregnancy and up to 6 months postpartum, showing mild to moderate changes during pregnancy. In our cohort, healthy women with no clinical signs of skin disease or dermatitis were included. Our observational results provide a foundation for future interventional studies addressing the following research questions:

• 1.

  To evaluate skin care regimen on skin roughness and TEWL in women during pregnancy and postpartum

• 2.

  To evaluate skin care regimen on skin roughness and dry skin in infants in their first year of life.

This observational study is an investigator-initiated trial conducted at the Charité-Universitätsmedizin Berlin, Germany. The project was financially supported by the Dehaa Rossun Research Center of Lunaler Group, Kowloon, Hong Kong. We highly appreciate the participation of all women and their infants who participated in the study during the challenging time of the COVID-19 pandemic. We would like to thank Mrs Uhl (Courage+Khazaka electronic GmbH; Germany) for her enormous support and advice on technical and software issues.

The Ethics Committee of the Charité-Universitätsmedizin Berlin reviewed the synopsis, the informed consent form (ICF), and the study protocol for the application of the ethical approval. Approval has been granted on 14 December 2020. The committee’s reference number is EA2/184/20. Written informed consent was obtained from the included women. For the infants, written informed consent was obtained by both parents, or from the parent with sole custody.

Ulrike Blume-Peytavi is a scientific member of the Board of the Dehaa Rossun Research Center. Gavin Zhou is the head of Dehaa Rossun Research Center of Lunaler Group. All other authors state that they have no conflicts of interest.

This observational study is an investigator-initiated trial conducted at the Charité-Universitätsmedizin Berlin, Germany. The project was financially supported by the Dehaa Rossun Research Center of Lunaler Group, Kowloon, Hong Kong. This funding organization was included in the development of the study concept but was not included in the analysis, interpretation, and publication of the results of the study. The funder had no role in the design, data collection, data analysis, and reporting of this study.

U.B.‐P. contributed with conceptualization, funding acquisition, methodology, resources, supervision, and writing – review and editing. J.K. contributed with conceptualization, methodology, and writing – review and editing. D.W. contributed with project administration, investigation, and writing – original draft. K.H., T.T., A.C., R.A., and A.F. contributed with investigation and writing – review and editing. G.E. contributed with formal analysis and writing – review and editing. G.Z. contributed with conceptualization, review, and editing.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.