Haftorn et al. Clin Epigenet (2021) 13:82 https://doi.org/10.1186/s13148-021-01055-z RESEARCH Open Access An EPIC predictor of gestational age and its application to newborns conceived by assisted reproductive technologies Kristine L. Haftorn1,2,3* , Yunsung Lee1,2, William R. P. Denault1,2,4, Christian M. Page2,5, Haakon E. Nustad2,6, Robert Lyle2,7, Håkon K. Gjessing2,4, Anni Malmberg8, Maria C. Magnus2,9,10, Øyvind Næss3,11, Darina Czamara12, Katri Räikkönen8, Jari Lahti8, Per Magnus2, Siri E. Håberg2, Astanand Jugessur1,2,4† and Jon Bohlin2,13† Abstract Background: Gestational age is a useful proxy for assessing developmental maturity, but correct estimation of gestational age is difcult using clinical measures. DNA methylation at birth has proven to be an accurate predictor of gestational age. Previous predictors of epigenetic gestational age were based on DNA methylation data from the Illumina HumanMethylation 27 K or 450 K array, which have subsequently been replaced by the Illumina Methylatio- nEPIC 850 K array (EPIC). Our aims here were to build an epigenetic gestational age clock specifc for the EPIC array and to evaluate its precision and accuracy using the embryo transfer date of newborns from the largest EPIC-derived dataset to date on assisted reproductive technologies (ART). Methods: We built an epigenetic gestational age clock using Lasso regression trained on 755 randomly selected non-ART newborns from the Norwegian Study of Assisted Reproductive Technologies (START)—a substudy of the Norwegian Mother, Father, and Child Cohort Study (MoBa). For the ART-conceived newborns, the START dataset had detailed information on the embryo transfer date and the specifc ART procedure used for conception. The predicted gestational age was compared to clinically estimated gestational age in 200 non-ART and 838 ART newborns using MM-type robust regression. The performance of the clock was compared to previously published gestational age clocks in an independent replication sample of 148 newborns from the Prediction and Prevention of Preeclampsia and Intrauterine Growth Restrictions (PREDO) study—a prospective pregnancy cohort of Finnish women. Results: Our new epigenetic gestational age clock showed higher precision and accuracy in predicting gestational age than previous gestational age clocks (R2 0.724, median absolute deviation (MAD) 3.14 days). Restricting the analysis to CpGs shared between 450 K and EPIC= did not reduce the precision of the clock.= Furthermore, validating the clock on ART newborns with known embryo transfer date confrmed that DNA methylation is an accurate predictor of gestational age (R2 0.767, MAD 3.7 days). = = Conclusions: We present the frst EPIC-based predictor of gestational age and demonstrate its robustness and preci- sion in ART and non-ART newborns. As more datasets are being generated on the EPIC platform, this clock will be valuable in studies using gestational age to assess neonatal development. *Correspondence: [email protected] †Astanand Jugessur and Jon Bohlin: Joint senior authors 1 Department of Genetics and Bioinformatics, Norwegian Institute of Public Health, Oslo, Norway Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Haftorn et al. Clin Epigenet (2021) 13:82 Page 2 of 13 Keywords: DNA methylation, Epigenetics, Gestational age, Illumina MethylationEPIC BeadChip, Assisted reproductive technologies, IVF, ICSI, MoBa, MBRN, PREDO Background variability, in part due to varying length of the follicular Accurate determination of gestational age is important phase. Ultrasound is much more precise but still depends for assessing fetal development and maturity. Tis is on the size of the fetus at the time of ultrasound [1, 21, necessary for investigating the impact of prenatal fac- 22]. On the other hand, for children conceived by ART, tors on pregnancy outcomes and any deviation from nor- the exact time when the embryo is transferred back to mal fetal development [1, 2]. Although gestational age the uterus is known. Although there may be some difer- at birth exhibits some normal variation, both preterm ences in the days before fertilization and embryo transfer, and post-term births are associated with an increased and the developmental speed may difer in the in vitro risk of adverse perinatal outcomes and health outcomes setting, the embryo transfer date (ETD) provides a more later in life [3–7]. Te efects of gestational age at birth direct estimate of gestational age [23]. Terefore, DNAm on health outcomes may be linked to epigenetic patterns data from ART births is particularly advantageous for established in utero or early in the postnatal period [8, developing and validating gestational age clocks. To our 9]. Changes in these patterns may interfere with critical knowledge, no gestational age clock has yet been devel- developmental processes [10–12] and trigger phenotypic oped using ETD, although its use has been called for pre- changes that persist throughout life. Tis may be even viously [16]. more pertinent to children conceived by assisted repro- In addition to gestational age prediction, gestational ductive technologies (ART), because ART procedures age clocks can be used to estimate gestational age accel- coincide with the extensive epigenetic reprogramming in eration (GAA), which is defned as the discrepancy the early embryo [13, 14]. between gestational age predicted from DNAm data DNA methylation (DNAm) is the most studied epige- and gestational age derived from clinical measurements netic mark in humans. It has, in recent years, been used [16, 24]. Investigating GAA is important because of its to build gestational age clocks that can predict gesta- reported association with several measures related to tional age [15–18]. Earlier clocks were built using DNAm birth outcomes, such as the cerebroplacental ratio (a data from the Illumina HumanMethylation27 (27 K) or robust indicator of prenatal stress [25]), higher mater- the Illumina HumanMethylation450 (450 K) BeadChip nal body mass index, and larger birth size [26]. Although arrays, both of which have subsequently been replaced children conceived by ART have a higher risk of sponta- by the Illumina MethylationEPIC BeadChip (EPIC). neous preterm birth [27] and other adverse perinatal out- EPIC has nearly twice (865,859 CpGs) as many CpGs as comes [28–30], only one small study has explored GAA 450 K, and a stronger focus on regulatory elements [19]. in ART children [31]. Although EPIC includes over 90% of the probes on 450 K To address these knowledge gaps, we developed a new [19], six to eight of the CpGs included in existing gesta- gestational age clock based on EPIC-derived DNAm tional age clocks are not present on EPIC. Tis discrep- data from newborns in the Norwegian Study of Assisted ancy may afect the precision of the published clocks Reproductive Technologies (START), which is a sub- in predicting gestational age when applied to DNAm study within the Norwegian Mother, Father and Child data generated on EPIC [20]. Terefore, it is essential to Cohort Study (MoBa) [32]. We validated this clock in develop a new gestational age clock that is updated and test sets of ART and non-ART newborns in START, and optimized for EPIC. Equally important is to elucidate also in an external dataset from the Finnish Prediction whether the additional CpGs on EPIC enhance gesta- and Prevention of Preeclampsia and Intrauterine Growth tional age prediction. Restriction (PREDO) study [33], which was used as a rep- A challenge in developing accurate gestational age lication cohort. We also used the new EPIC-based clock clocks is the lack of information on the exact gestational to explore diferences in GAA between ART and non- age of the newborns. Te standard approaches for esti- ART newborns. mating gestational age, based on ultrasound measure- ments or the last menstrual period (LMP), have thus Results far been used for training and testing epigenetic clocks. The EPIC gestational age clock Ultrasound and LMP are widely used in clinical settings Table 1 and Fig. 1 provide overviews of the datasets used and have their individual advantages and limitations. in this study. We ft a least absolute shrinkage and selec- While LMP can be informative, it sufers from large tion operator (Lasso) regression on DNAm data from 755 Haftorn et al. Clin Epigenet (2021) 13:82 Page 3 of 13 Table 1 Characteristics of the datasets used to evaluate the