A quantitative cross‐sectional analysis of the melanin index in the skin of preterm newborns and its association with gestational age at birth

Abstract Background Estimation of gestational age (GA) is important to make timely decisions and provide appropriate neonatal care. Clinical maturity scales to estimate GA have used skin texture and color to assess maturity at birth facing situations of the uncertainty of pregnancy dating. The size and darkness of the areola around the nipple to grade skin characteristics are based on visual appearance. The melanin index (M‐Index) is an optical skin parameter related to the melanin content in the tissue. This study is aimed to associate the M‐Index of the skin with the GA. Methods A cross‐sectional study evaluated 80 newborns at birth. A photometer device quantified the skin pigmentation on the areolae, forearms, and soles. Paired average differences of M‐Index were compared among the three body sites. The skin M‐Indexes were compared between subgroups of newborns until 34 weeks or with 34 and more. Results The skin over the areola had the highest values of M‐Index compared with the forearm or sole areas (P < .001 for both). Infants with a GA between 34 and <37 weeks had higher M‐Index values over the areola than the group with a GA with 24 to <34 weeks: 41.7 (8.9) and 38.3 (10.5) median (IQR), P = .005. Conclusions The measurable M‐Index values have the potential to improve physical evaluation in assessing GA at birth.

under visual inspection to the color assessment of skin is subjective, sometimes clinically inadequate, and discordant between observers. 5 In such evaluation, the color of the skin plays a major role despite being a subjective perception.
Skin color is the result of the existence of two main cutaneous chromophores, melanin and hemoglobin located in the epidermis and dermis, whose optical properties are well studied. 6 Photometric methods to measure skin color evaluate the interaction of light with chromophores taking into account the optical properties of these compounds, their concentrations, and specific wavelengths of the illuminating source. 5 The melanin index (M-Index) is an optical skin parameter related to the melanin content in the tissue. 7 In clinical practice, the phototype scale of Fitzpatrick et al is a semi-quantitative measure of the skin color that combines the intensity of the melanization and the erythemal response to sun exposure. 8 During the neonatal period, skin color is the result of pigments and skin thickness. 9 Analyzing skin reflectance at birth, Post et al reported that, in utero life, black and white infants already show skin reflectance differences due to the sex and site of the body after 32 weeks of gestation. 9 In fact, another report assessing the M-Index in 447 healthy neonates from 35 to 42 weeks of gestation showed that the skin color of neonates is mainly determined genetically. 10 Therefore, knowledge about physiological variations in skin pigmentation has clinical relevance because of associations with jaundice, anemia, plethora, and hormonal dysfunctions. 11 Studies on the skin pigmentation of newborn infants are scarce or focused on the measurements of transcutaneous bilirubin, 12 namely the influence of melanin in noninvasive measurements. 13 Meanwhile, assessment of GA at birth supported by noninvasive photometry by skin contact is a new approach, which is under clinical validation. There is a potential association between skin reflectance and GA at birth. 14 Post-natal approaches to pregnancy dating have limitations regarding accuracy and precision. 15 Preterm infants require special health care, mostly when they are born within <32 weeks of gestation, demanding inpatient care to survive. Innovations are still necessary to measure gestational age more efficiently in low-income settings, considering low-cost solutions. 16 This study aimed to assess the M-index of the skin in premature newborns and associate it with the GA at birth. We hypothesize that the darkness of the areola, assessed with a handheld photometer device, has the potential to contribute to pregnancy dating after birth in preterm neonates.

| Study design and setting
The design of the study was an observational cross-sectional gathering of obstetric data, a clinical observation of the skin color in women, and an assessment of the M-Index in the skin of newborns using a noninvasive optical device. Two referral perinatal centers in Brazil participated in this evaluation: Hospital das Clínicas, Universidade A sequential enrollment process selected women and their infants with the following inclusion criteria: age newborn up to 24 hours of life with a GA between 24 and 36 weeks + 6 days at birth, confirmed with an obstetric ultrasound assessment before 14 weeks of pregnancy. Exclusion criteria were malformation with structural skin alterations or skin modifiers as follows: anhydramnios, hydrops, congenital skin diseases, or chorioamnionitis.
Inpatient medical records and a quick interview with the women were the sources of clinical data. The first-trimester ultrasound was the reference for GA calculated at birth. Ranges of age: 24 to <34 weeks, 34 to <37 weeks bound two groups of analysis, according to the priorities of the newborn. 17 The color of the mother's skin was measured from the inner forearm using the Fitzpatrick reference. 8 This visual scale for skin phototypes ranged from 1 (pale light white) to 6 (very dark brown to black).
Neonates had their skin reflectance assessed at three body sites, the anterior distal forearm, the sole of the foot, and the areola.
Choosing such areas, the examiner avoided any interference with vital signal sensors in use, ensuring minimum handling and stable clinical conditions in preterm infants, mainly inside incubators. The areola was assessed due to this importance as a clinical maturity marker. 1,2,4

| M-Index measurement
A handheld multiband reflectance photometric device, previously detailed, 18

| Data analysis
A descriptive analysis is expressed as the frequencies, the mean, and standard deviation (SD) presented baseline characteristics of the study group and skin M-Index, whereas the median and interquartile range (IQR) were preferred for non-normally distributed continuous variables. Independent mean or median tests, Student's t, Mann-Whitney, or Kruskal-Wallis, evaluate the relationship between the measurements and groups of interest. Paired t tests compared M-Index differences among the three body sites of the infant. The significance level for hypothesis tests will be 5%, together with 95% confidence interval (CI). A sample size of 80 newborns had the power of 82% for detecting a mean difference between the two groups, considering a medium size effect (0.6) and 95% of confidence, in onetail t test. 20  and ranged from 24.1 to 36.9 weeks. We then stratified the infants by GA and found 30 (37.5%) were between 24 and <34 weeks, and 50 (62.5%) were between 34 and <37 weeks. Birth weight ranged from 510 to 2545 g; with a mean (SD) of 1967 (SD = 628) grams.

| RE SULTS
Measurements were taken, 34 (42.5%) were female and 46 (57.5%) were male. The evaluation occurred inside an incubator in 36 (45.0%) infants and 37 (46.3%) infants were under intensive care. Sixteen newborns were siblings, and the number of mothers was 72 women.
Most mothers had a phototype classification of type-III (34/72 42.5%) or type-IV (22/72 27.5%). Table 1   monitoring attached over areolae. The areola area was the darkest site of the three areas evaluated on the body of the newborn. Paired comparisons revealed significant within-subject differences in the M-Index on the areola compared with the forearm (Line 2 P < .001), and the sole (Line 2 P < .001), see Table 2.  (Table 4 line 2).

| Key results and interpretation
The central contribution of this analysis was associating the color of the areola with the pregnancy chronology at birth using a quan- Skin color by inspection was, in the past, recognized as an independent predictive variable of GA at birth. 21 Regarding the skin color evaluation of newborns using optical approaches, there are scarce studies for comparisons. Previous results explored changing in the skin reflection due to pigment distribution and concentration, as well as tissue thickness according to GA. 9 In that article, ethnicity of the infant was based on the appearance of the parents.
There was a difference in the skin reflection between white and black neonates, after 32 weeks of gestation. 9 The present study failed in demonstrating a relationship between the M-Index as- The sample analyzed here consisted of a broad scale of prematurity from 24 weeks toward the near term, most of them inside incubators and receiving intensive neonatal care. In preterm newborns, the lack of the skin barrier is a cause of loss of water by evaporation, hypothermia, and less protection against external agents. 22 The fragility of the premature infant's skin needs to be promptly recognized at birth to face the important clinical challenges of newborn attention.
The concern of offering the best neonatal health care justifies the exploration of skin maturity based on the optical properties of this tissue.

| Limitations and perspectives
In situ measurements of the skin pigmentation involve mathematical bio-optical models to represent the interaction of light with the human skin. This study used the simplified principle in which the assessment of the intensity of the reflected red light is related to the optical skin density of melanin content, the M-Index. 5 The reasons for the three layers' optical model assumption were previous reported to compare our results since commercial devices adopted the same model. 10,23 Despite an experimental handheld device, the repeatability of acquisitions was excellent, as previously reported. 17 For Nonphysiologic or adaptive hyperbilirubinemia is the most common morbidity in the neonatal period, which is more severe in premature infants. 24 In our sample, there was no phototherapy ongoing. We be-

| CON CLUS IONS
We noticed that quantification of the areolar darkness objectively evaluated the nipple's physical differences between the newborns at a cutoff of 34 weeks. The measureable M-Index  values have the potential to improve physical evaluation to assessing GA at birth. in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CO N FLI C T O F I NTE R E S T S
Authors RNG and ZSNR declare a patent deposit on behalf of the