nutrients Article Abnormal Maternal Body Mass Index and Customized Fetal Weight Charts: Improving the Identification of Small for Gestational Age Fetuses and Newborns Nieves Luisa González González 1, *, Enrique González Dávila 2 , Agustina González Martín 3 , Marina Armas 4 , Laura Tascón 1 , Alba Farras 5 , Teresa Higueras 5 , Manel Mendoza 5 , Elena Carreras 5 and María Goya 5 1 Department of Obstetrics and Gynecology, University of La Laguna, Hospital Universitario de Canarias, 38200 Tenerife, Spain 2 Department of Mathematics, Statistics and Operations Research, IMAULL, University of La Laguna, 38200 Tenerife, Spain 3 Department of Obstetrics and Gynecology, Hospital Universitario Ntra Sra de Candenlaria, 38200 Tenerife, Spain 4 Department of Pediatrics, Evangelisches Krakenhaus König Elisabeth Herzberge, 10365 Berlin, Germany 5 Maternal-Fetal Medicine Unit, Department of Obstetrics, Hospital Universitari Vall d’Hebron, Universitat Autónoma de Barcelona, Pg. de la Vall d'Hebron, 119, 08035 Barcelona, Spain * Correspondence: nlglez@ull.edu.es; Tel.: +34-922678335 Abstract: Background: Obesity and thinness are serious diseases, but cases with abnormal maternal weight have not been excluded from the calculations in the construction of customized fetal growth curves (CCs). Method: To determine if the new CCs, built excluding mothers with an abnormal weight, are better than standard CCs at identifying SGA. A total of 16,122 neonates were identified as SGA, LGA, or AGA, using the two models. Logistic regression and analysis of covariance were used to calculate the OR and CI for adverse outcomes by group. Gestational age was considered as a Citation: González González, N.L.; covariable. Results: The SGA rates by the new CCs and by the standard CCs were 11.8% and 9.7%, González Dávila, E.; González respectively. The SGA rate only by the new CCs was 18% and the SGA rate only by the standard CCs Martín, A.; Armas, M.; Tascón, L.; was 0.01%. Compared to AGA by both models, SGA by the new CCs had increased rates of cesarean Farras, A.; Higueras, T.; Mendoza, M.; section, (OR 1.53 (95% CI 1.19, 1.96)), prematurity (OR 2.84 (95% CI 2.09, 3.85)), NICU admission (OR Carreras, E.; Goya, M. Abnormal 5.41 (95% CI 3.47, 8.43), and adverse outcomes (OR 1.76 (95% CI 1.06, 2.60). The strength of these Maternal Body Mass Index and Customized Fetal Weight Charts: associations decreased with gestational age. Conclusion: The use of the new CCs allowed for a more Improving the Identification of Small accurate identification of SGA at risk of adverse perinatal outcomes as compared to the standard CCs. for Gestational Age Fetuses and Newborns. Nutrients 2023, 15, 587. Keywords: customized growth charts; fetal weight; newborn weight; maternal body mass index; https://doi.org/10.3390/nu15030587 obesity; thinness; small for gestational age; birthweight; newborn weight; perinatal outcomes Academic Editors: Shannon L. Kelleher and Ann Anderson Berry Received: 24 November 2022 1. Introduction Revised: 8 January 2023 Small for gestational age (SGA) is defined as a birth weight of less than the 10th Accepted: 19 January 2023 percentile for gestational age. SGA and neonates are at greater risk of perinatal morbidity Published: 22 January 2023 and mortality, as well as obesity, cardiovascular diseases, and neurological disorders during childhood and cardiovascular diseases in adult life [1–5]. The effectiveness of SGA screening relies on the accurate estimation of gestational Copyright: © 2023 by the authors. age and fetal weight and depends on the selected reference fetal growth curve. Different Licensee MDPI, Basel, Switzerland. standard population birth weight curves are used to differentiate fetuses and newborns This article is an open access article with abnormal growth, SGA, and large for gestational age (LGA), from those with normal distributed under the terms and growth (appropriate for gestational age (AGA)). Standard population fetal growth curves conditions of the Creative Commons only consider gestational age and sex to calculate weight percentiles [6]. Attribution (CC BY) license (https:// The use of customized fetal and neonatal weight charts allows to differentiate large creativecommons.org/licenses/by/ and small constitutional fetuses from those with pathological intrauterine growth and help 4.0/). in our understanding and diagnosis of abnormal fetal growth [7,8]. The first customized Nutrients 2023, 15, 587. https://doi.org/10.3390/nu15030587 https://www.mdpi.com/journal/nutrients Nutrients 2023, 15, 587 2 of 11 fetal growth charts (CCs) developed by Gardosi et al. [9,10] and adjusted for maternal height, weight, parity, and ethnic origin represent a gold standard for assessing the growth potential of each individual fetus. Different studies have shown that infants classified as SGA according to CCs were at an increased risk of adverse outcomes as compared to infants classified as SGA according to non-customized charts [11–18]. As a result, the use of CCs has expanded worldwide [19]. Abnormal maternal weight (both obesity and underweight) during pregnancy in- creases oxidative stress, alters the hormonal environment and microbiome, and can modify the epigenetic of the mother–placenta–fetus axis with short- and long-term deleterious consequences for mothers and their offspring [20–24]. Maternal obesity and thinness are also associated with several adverse perinatal outcomes including excessive or restricted fetal growth, preterm delivery, cesarean delivery, gestational diabetes, hypertensive dis- orders of pregnancy, and infant morbidity and mortality [25,26]. However, to the best of our knowledge, maternal obesity and thinness have never been considered so far as an exclusion criterion in the construction of CCs. If normal fetal weight curves are calculated excluding the weights of infants born to mothers with obesity or thinness, the efficiency of the models to identify fetuses and newborns with restricted or excessive intrauterine growth will be greater than that of conventional curves, as well as the possibility of avoiding associated adverse perinatal outcomes. The aim of this study was to establish the effectiveness of a new, customized fetal growth chart (new CC), which was constructed excluding cases with an abnormal maternal pre-pregnancy body mass index (BMI) to identify SGA infants at an increased risk of adverse perinatal outcomes versus that of standard CC 2. Material and Method Maternal characteristics and perinatal outcomes from a sample of 16,122 singleton- pregnancy neonates born at 26–43 weeks of gestation between 2018 and 2020 were collected from the Register of Perinatal Data of the University Hospital Vall d’Hebrón, Barcelona, Spain. From the initial sample of 16,122 infants, 1382 were excluded due to incomplete or implausible data; therefore, the final sample consisted of 14,740 infants. Newborns were classified by percentile birth weight as large for gestational age (LGA), SGA, or adequate for gestational age (AGA) using a standard customized chart [26] and the new CCs [27], which was constructed excluding cases with abnormal pre-pregnancy maternal BMI (<18.5 or >25 kg/m2 ) [28], only including women with a normal weight. LGA was defined as a weight above the 90th percentile, SGA as a weight below the 10th percentile, and AGA as a weight between the 10th and 90th percentile. Factors included in the final stepwise models and their coefficients for standard customized curves [26] and new CCs [27] prediction models for optimal fetal weight are included in Table 1. 2.1. Outcomes The maternal characteristics analyzed were: age, weight, height, BMI, parity, smoking habit, diabetes mellitus, and assisted reproduction techniques required, and the following perinatal outcomes were studied: mode of delivery, gestational age (days), prematurity (less than 28 weeks, between 28 and 34 weeks, and between 34 and 37 weeks of gestational age), birth weight, Apgar score at 1 and 5 min, pH value in the umbilical artery at delivery, admission to the neonatal intensive care unit (NICU), and perinatal mortality (stillbirth and neonatal death). In addition, we considered two composite outcomes (at least one perinatal outcome), 1: cesarean section, shoulder dystocia, Apgar score < 5 at 1st and 5th min, NICU admission or perinatal mortality; and 2: Apgar score < 5 at 1st and 5th min, NICU admission or perinatal mortality. Perinatal outcomes were compared between infants classified as AGA and SGA according to the standard CCs or the new CCs. Nutrients 2023, 15, 587 3 of 11 Table 1. Coefficients for conventional customized curves [26], and new customized prediction models for optimal weight. Factors included in the final stepwise models [27]. Model Customized Curves New Customized Curves (18.5–25) Constant 3289.681 3304.579 Maternal height (MH, cm) (from 163) 9.392 6.987 Maternal weight (MW, kg) (from 65) Linear 4.856 7.510 Squared −0.098 - Cubed 0.001 - Parity (Birth ≥ 1) 113.889 113.530 Ethnic origin East-Asia 165.560 143.461 South America 161.550 134.161 Rest of Europe 67.927 68.934 North Africa 109.265 62.447 GA (weeks) (from 40) Linear 135.413 134.457 Squared −14.063 −13.435 Cubed −0.838 −0.803 Sex Male 67.188 67.552 Female −67.188 −67.552 Interactions Sex 6.890 8.501 Parity (Birth ≥ 1) 9.032 11.300 GA (linear) with MH (cubed) 0.006 0.008 MW (squared) 0.260 - MH (linear) −0.378 - GA (squared) with MH (squared) −0.008 −0.011 GA (cubed) with MH (linear) −0.032 - Mean square error 144,630.076 133,659.796 R2 0.454 0.451 Coefficient of variation (cv) 0.1156 0.1106 BMI: body mass index; GA: gestational age. 2.2. Subgroups and Comparisons In order to assess the effectiveness of the new CCs, the following groups of infants were considered and compared: - SGA according to the standard CCs, SGA according to the new CCs, SGA according to both the CCs and new CCs, SGA according only to the standard CCs, and SGA according only to the new CCs. - AGA according only to the new CCs, and AGA according to both the standard CCs and the new CCs. 2.3. Statistical Analysis The distribution of variables was investigated using histograms and the Kolmogorov– Smirnov’s test. Numerical data are reported as mean and standard deviation for parametric variables. Qualitative variables are reported as frequencies and percentages. Differences between groups were studied using the Student’s t-test. When three or more groups were compared, homogeneous subsets were indicated at a level of 5%, so that cases that are in the same subset did not differ significantly. Comparison between proportions was completed using the χ2 test and Fisher’s exact test when any of the expected values were <5. Logistic regression was used to calculate the odds ratio (OR) of adverse outcomes taking AGA according to both the CCs and the new CCs as a reference (OR = 1). The Nutrients 2023, 15, 587 4 of 11 confidence intervals for pH were calculated by analysis of covariance. Gestational age was considered as a covariable. Cohen’s kappa coefficient was used to assess agreement between the chart models and the scale proposed by Landis and Kock [29] was used to describe the level of agreement as follows: 0.21–0.40, “poor”; 0.41–0.60, “moderate”; 0.61–0.80, “good” and 0.81–1.00 “excellent”. For statistical analysis, SPSS 25.0 software (SPSS Inc., Chicago, IL, USA) was used. 3. Results The 14,740 infants included in the final sample were classified by weight as AGA, SGA, or LGA according to the standard CCs and the new CCs. Maternal characteristics and perinatal outcomes for these groups are shown in Table 2. Table 2. Maternal characteristics of small, adequate, and large for gestational age infants (SGA, AGA, and LGA) according to the use of new customized curves (new CCs) or conventional customized curves (CC). New Customized Curves (New CCs) Conventional Customized Curves (CC) Total SGA (S) AGA(A) LGA (L) p-Value (Ho- SGA AGA LGA (N = 14,740) (N = 1744) (N = 10,984) (N = 2012) mogeneous (N = 1431) (N = 11,197) (N = 2112) 11.8% 74.5% 13.7% Subset) 9.7% 76% 14.3% Maternal age, years 31.7 ± 5.9 32.1 ± 5.9 31.7 ± 5.9 31.3 ± 5.9 <0.001 (L; A; S) 32.2 ± 6.0 31.7 ± 5.9 31.4 ± 5.8 Weight, kg 64.4 ± 13.4 66.8 ± 15.8 64.2 ± 13.2 63.3 ± 12.0 <0.001 (L; A; S) 64.9 ± 14.2 64.2 ± 13.2 65.2 ± 13.8 Height, m 161.1 ± 6.6 161.4 ± 6.6 161.2 ± 6.6 160.4 ± 6.9 <0.001 (L; A-S) 161.5 ± 6.5 161.2 ± 6.6 160.4 ± 6.9 BMI, kg/m2 24.8 ± 4.9 25.6 ± 5.7 24.7 ± 4.8 24.6 ± 4.3 <0.001 (L-A; S) 24.9 ± 5.1 24.7 ± 4.8 25.3 ± 4.9 Nulliparous 7509 (50.9) 902 (51.7) 5595 (50.9) 1012 (50.3) 0.685 753 (52.6) 5721 (51.1) 1035 (49.0) Cigarette smoker 2118 (14.5) 459 (26.5) 1511 (13.9) 148 (7.4) <0.001 (L; A; S) 385 (27.1) 1578 (14.2) 155 (7.4) Diabetes mellitus 154 (1.0) 12 (0.7) 87 (0.8) 55 (2.7) <0.001 (S-A; L) 9 (0.6) 87 (0.8) 58 (2.7) Assisted reproduction 1150 (7.8) 140 (8.0) 863 (7.9) 147 (7.3) 0.333 119 (8.3) 879 (7.9) 152 (7.2) techniques Gestational age (days) 274 ± 14.2 267 ± 20.4 275 ± 12.6 273 ± 14.6 <0.001 (S; L; A) 267 ± 20.5 275 ± 12.6 273 ± 15.5 Cesarean section 3527 (23.9) 587 (33.7) 2358 (21.5) 582 (28.9) <0.001 (A; L; S) 488 (34.1) 2406 (21.5) 633 (30.0) Prematurity 1271 (8.6) 319 (18.3) 753 (6.9) 199 (9.9) <0.001 (A; L; S) 267 (18.7) 788 (7.0) 216 (10.2) <28 55 (0.4) 28 (1.6) 18 (0.2) 9 (0.4) 21 (1.5) 18 (0.2) 16 (0.8) (28–34) 286 (1.9) 105 (6.0) 148 (1.3) 33 (1.6) 89 (6.2) 160 (1.4) 37 (1.8) (34–37) weeks 930 (6.3) 186 (10.7) 587 (5.3) 157 (7.8) 157 (11.0) 610 (5.4) 163 (7.7) Umbilical artery pH 7.25 ± 0.094 7.23 ± 0.101 7.25 ± 0.093 7.25 ± 0.088 0.010 (S; A-L) 7.23 ± 0.104 7.25 ± 0.092 7.24 ± 0.094 Apgar < 7at 1st min. 541 (3.7) 151 (10.7) 313 (3.5) 77 (4.7) <0.001 (A-L; S) 129 (9.0) 325 (2.9) 87 (4.1) Apgar < 7 at 5th min. 100 (0.7) 36 (2.5) 56 (0.6) 8 (0.5) <0.001 (L-A; S) 30 (2.1) 57 (0.5) 13 (0.6) NICU admission 363 (2.5) 140 (8.0) 180 (1.6) 43 (2.1) <0.001 (L-A; S) 117 (8.2) 192 (1.7) 54 (2.6) Perinatal mortality 65 (0.4) 28 (1.6) 32 (0.3) 5 (0.2) <0.001 (L-A; S) 26 (1.8) 31 (0.3) 8 (0.4) Composite outcome 1 3878 (26.3) 665 (38.1) 2582 (23.5) 631 (31.4) <0.001 (A; L; S) 549 (38.4) 2642 (23.6) 687 (32.5) Composite outcome 2 832 (5.6) 245 (14.0) 473 (4.3) 114 (5.7) <0.001 (A-L; S) 208 (14.5) 495 (4.4) 129 (6.1) Results are shown as mean ± standard deviation or as frequency (%). Composite outcome 1: cesarean sec- tion, Apgar < 7 at 1st min, Apgar < 7 at 5th min, and NICU admission or mortality. Composite outcome 2: Apgar < 7 at 1st min, Apgar < 7 at 5th min, and NICU admission or mortality. S: SGA (small for gestational age); A: AGA (adequate for gestational age); L: LGA (large for gestational age). According to the conventional CC, the rates of SGA, AGA, and LGA were 9.7%, 76.0%, and 14.3%%, respectively. Using the new CC model, the rate of SGA was significantly higher 11.8%, and the rate of LGA and AGA were lower, 13.7% and 74.5%, respectively (p < 0.001). The rate of SGA by both the conventional CCs and the new CC was 81.9%. The SGA rates according only to the new CCs (SGA according to the new CCs, but AGE according to the standard CC) were 18%, and SGA rates according to only the standard CCs (SGA according to the standard CCs, but AGA according to the new CCs) were less than 0.01% (Figure 1). The Kappa agreement coefficient between the standard CCs and the new CCs was “excellent” (Kappa = 0.906; p < 0.001). Perinatal outcomes were significantly worse in infants classified as SGA according to only the new CCs as compared to infants classified as AGA according to the new CCs; he rate of cesarean section in SGA only according the new CCs was significantly higher than in AGA classified by the new CCs, (31.8% versus 21.4%, p < 0.001) and the rate of prematurity was more than double (16.9 % versus 6.9%, p < 0.01). The Apgar values at the first and fifth minutes were also significantly higher in SGA only classified by the new CCs (p < 0.001 and 0.007, respectively). The rate of NICU admission reached 7.6% in SGA only S: SGA (small for gestational age); A: AGA (adequate for gestational age); L: LGA 144 (large for gestational age). According to the conventional CC, the rates of SGA, AGA, and 145 LGA were 9.7%, 76.0%, and 14.3%%, respectively. Using the new CC model, the rate of 146 SGA was significantly higher 11.8%, and the rate of LGA and AGA were lower, 13.7% and 147 74.5%, respectively (p < 0.001). Table 2. 148 Nutrients 2023, 15, 587 The rate of SGA by both the conventional CCs and the new CC was 81.9%. The 5SGA of 11 149 rates according only to the new CCs (SGA according to the new CCs, but AGE according 150 to the standard CC) were 18%, and SGA rates according to only the standard CCs (SGA 151 according by the newtoCCsthe standard CCs, versus 0.6% in but AGA AGA according according to the to the newnew CCCCs) were group, (p