Nutrition in early life plays a crucial role in shaping lifelong health outcomes. The early life period, spanning from conception to early childhood, is a critical window of susceptibility during which environmental factors can influence developmental processes and establish a trajectory of health or disease(Reference Godfrey, Gluckman and Hanson1). Nutrition, including micronutrient status and body composition, is foremost among such environmental factors, which also include exposure to parental smoking, endocrine-disrupting chemicals, biological and psychosocial stress. The concept of developmental origins of health and disease (DOHaD) proposed that early life exposures can have lasting impacts on health and disease risk later in life(Reference Godfrey, Gluckman and Hanson1). The DOHaD concept emerged from epidemiological studies in the UK during the 1980s, which revealed a clear relationship between lower birth weight and an increased risk of death from non-communicable diseases (NCD) in adulthood, including CVD(Reference Barker and Osmond2) and chronic obstructive airways disease(Reference Barker, Godfrey and Fall3). Impaired fetal development has been linked to a higher risk of a range of adult NCD that increase premature mortality risk, including type 2 diabetes, metabolic syndrome, osteoporosis, sarcopenia and CHD(Reference Poston, Godfrey, Gluckman and Hanson4), and to negative outcomes in childhood, such as stunted growth and reduced cognitive function(Reference Poston, Godfrey, Gluckman and Hanson4,Reference Castanys-Muñoz, Kennedy and Castañeda-Gutiérrez5) . These associations have been consistently replicated and are recognised as being exacerbated by environmental risk factors encountered in the postnatal environment.
To date, fixed genomic variations such as SNP and copy number variations have been found to explain only a modest proportion of the risk associated with NCD, and it is increasingly believed that the developmental environment plays a significant role in shaping later-life phenotypes by modifying epigenetic regulation of genes(Reference Kong, Kowalik and Garratt6). Epigenetic processes including DNA methylation, histone modifications and non-coding RNA (ncRNA) regulate gene expression by modulating the packaging and expression of the DNA without a change in genomic sequence. These changes can be maintained over multiple divisions in somatic cells. Environmental and lifestyle influences such as nutrition and stress can induce epigenetic modifications, effectively making the epigenome a molecular record holding the ‘memory’ of past exposures. During development, epigenetic processes contribute to phenotypic plasticity, allowing the fetus to adapt to predicted postnatal environments(Reference Godfrey, Lillycrop and Burdge7). However, when the phenotype is mismatched to the later environment, for example, from inaccurate nutritional cues from the mother or placenta or rapid socioeconomic shifts, this ‘mismatch’ can increase the risk of NCD in adulthood(Reference Godfrey, Lillycrop and Burdge8).
The impact of developmental epigenetic changes may not become apparent until later in life. Epigenetic biomarkers may therefore serve as indicators of previously undocumented developmental exposures and as predictors of future disease risk, enabling early intervention strategies to improve both early development and later health. Discovery and validation of perinatal epigenetic biomarkers, with both replication in independent cohorts and in vitro validation, is therefore an important and growing field. Recent advances in enzymatic approaches and high-throughput sequencing have enabled epigenetic biomarker discovery on a genome-wide scale(Reference Füllgrabe, Gosal and Creed9). The expanded coverage provided by these platforms is likely to uncover many new disease-related epigenetic modifications located outside well-known candidate regions such as CpG islands and gene promoters, complementing traditional candidate gene and array-based approaches that only interrogate small regions of the genome.
Epigenetic changes can be tissue-specific and can have tissue-specific consequences. Skeletal muscle has a lower priority in nutrient partitioning in the developing fetus, compared with the brain and heart, thus making it particularly vulnerable to nutrient deficiency. Epidemiological studies have consistently linked negative early-life environmental exposures to reduced muscle mass and function in later life(Reference Patel, Jameson and Syddall10–Reference Sayer, Syddall and Gilbody12), with epigenetic changes proposed as mediators. Recent research measuring CpG methylation across the genome (the ‘methylome’) of cultured myoblasts isolated from older individuals has shown associations with birthweight, growth during infancy and childhood illnesses(Reference Garratt, Sharkh and Burton13), with some of the differentially mediated regions associated with later-life grip strength and sarcopenia. In this review, we will summarise the evidence from cohort and intervention studies on the relationships between nutrition in early life, epigenetics and lifelong health.
Maternal diet and fetal development
A mother’s preconception and pregnancy diet, micronutrient status, body composition, metabolism, mood and lifestyle are all implicated in maternal pregnancy outcomes and offspring body composition, cardiometabolic, neurobehavioural and allergic outcomes. Examples include maternal diet quality and conception rates(Reference Lim, Loy and Colega14), micronutrient intake and preterm birth(Reference Samuel, Sakwinska and Makinen15), vitamin D supplementation and infantile atopic eczema(Reference El-Heis, D’Angelo and Curtis16), iodine status and child IQ(Reference Robinson, Crozier and Miles17), maternal adiposity and offspring obesity(Reference Dalrymple, El-Heis and Godfrey18) and maternal stress and offspring atopic eczema(Reference El-Heis, Crozier and Healy19).
Maternal diet during pregnancy is a critical determinant of fetal development and has been linked to a range of health outcomes in offspring(Reference Abdollahi, Soltani and de Souza20–Reference Borge, Aase and Brantsæter22). A maternal diet rich in fruits, vegetables and whole grains has been associated with improved fetal growth and development, as well as a reduced risk of adverse birth outcomes, such as preterm birth and low birth weight(Reference Chia, Chen and Lai23). In contrast, maternal diets high in sugar, salt and unhealthy fats have been linked to an increased risk of adverse outcomes(Reference Chen, Aubert and Shivappa24), and maternal diets with a high inflammatory potential have been associated with increased offspring adiposity during childhood(Reference Soh, Chong and Kwek25).
The Growing Up in Singapore Towards healthy Outcomes (GUSTO) cohort study, a prospective mother–offspring cohort study, has made significant contributions to our understanding of the relationships between maternal diet, fetal development and lifelong health(Reference Soh, Chong and Kwek25). In this study, maternal diet was assessed using a 24-h recall and food diary at 26–28 weeks of gestation(Reference Chong, Godfrey and Gluckman26). Associations were identified between a high-quality dietary pattern during pregnancy and lower risks of preterm birth and excessive offspring adiposity during childhood(Reference Chia, de Seymour and Colega27,Reference Chen, Aris and Bernard28) . The findings of this study have highlighted the importance of a healthy maternal diet in promoting optimal fetal growth and development.
Epigenetics and early life nutrition
Epigenetics encompasses a variety of heritable modifications that regulate gene expression without altering the underlying DNA sequence. These modifications can be stably maintained through multiple somatic cell divisions, thereby contributing to long-term gene regulation. The most widely studied form of DNA methylation is where a methyl group is added to the cytosine base of a CpG dinucleotide, primarily to regulate gene expression. This methylation can alter gene expression by either directly blocking the binding of transcription factors to DNA or by recruiting repressive protein complexes that induce local chromatin remodelling(Reference Bird and Macleod29). Patterns of DNA methylation are usually studied within CpG-rich islands in gene promoter regions(Reference Deaton and Bird30), but gene body and intergenic region DNA methylation are also thought to influence cell physiology. Histone modifications such as acetylation, methylation, ubiquitination and phosphorylation can directly affect chromatin structure and therefore the accessibility of the underlying genomic sequence, while also providing binding sites for proteins involved in gene regulation. Other mechanisms include the ncRNA, which are functional RNA molecules that are not translated into proteins. They can mediate mRNA degradation or translational repression and, when targeted to the promoter region of a gene, induce both DNA methylation and repressive histone modifications(Reference Kaikkonen, Lam and Glass31).
Studies in animal models have shown that maternal diet can induce lasting metabolic changes in offspring by modifying the epigenetic regulation of key metabolic genes. For example, when pregnant rats were fed a protein-restricted diet, their offspring showed reduced DNA methylation of the glucocorticoid receptor (GR) and PPARα genes in the liver(Reference Lillycrop, Phillips and Jackson32). This epigenetic change was associated with increased expression of GR and PPARα and a persistent change in metabolic pathways, specifically enhanced gluconeogenesis and fatty acid β-oxidation, which are regulated by these nuclear receptors. With growing concern over the widespread consumption of energy-dense Western diets, many studies have turned their attention to the effects of maternal high-fat intake. In rats, maternal high-fat feeding during pregnancy has been shown to reduce expression of Fatty Acid Desaturase 2 (FADS2), the rate-limiting enzyme in PUFA synthesis in the liver of offspring, and this reduction is accompanied by altered DNA methylation at CpG sites within the gene’s promoter region(Reference Hoile, Irvine and Kelsall33). Similarly, in mice, maternal obesity and diabetes have been linked to widespread changes in DNA methylation in the liver of offspring(Reference Li, Young and Maloney34). Notably, the window of epigenetic plasticity may extend beyond the prenatal period into postnatal life. For example, overfeeding in rat pups led to hypermethylation at two CpG sites in the promoter of proopiomelanocortin (POMC), a gene critical for appetite regulation. This hypermethylation prevented the up-regulation of POMC expression in response to elevated plasma leptin and insulin levels(Reference Plagemann, Harder and Brunn35).
Evidence from human cohort studies
Evidence from human cohort studies suggests a similar important role for epigenetic processes in holding the ‘memory’ of developmental exposures, with long-term consequences for the risk of NCD. For example, these have linked nutritional exposures during pregnancy to epigenetic changes that increase offspring susceptibility to childhood obesity(Reference Godfrey, Sheppard and Gluckman36).
The prevalence of childhood obesity is rising rapidly, posing immediate health risks for children and increasing their likelihood of developing obesity and related metabolic disorders in adulthood(Reference Reilly and Kelly37). The National Child Measurement Programme (NCMP) has also shown the disparity gap in child obesity widening each year, mostly driven by rising obesity rates in the most deprived areas and a relatively stable prevalence among the least deprived children. The gap in obesity prevalence between the most and least deprived areas in 2022–23 has reduced compared to 2020–21 from 19.5 to 17.1 percentage points but is still much larger than that seen in pre-pandemic years(38).
Several early-life risk factors have been identified that significantly increase the likelihood of childhood obesity. In the prospective Southampton Women’s Survey (SWS) parent–offspring cohort, five key risk factors were defined: maternal obesity, excessive gestational weight gain, low maternal vitamin D levels, smoking during pregnancy and a short duration of breastfeeding. At both 4 and 6 years of age, there was a positive graded association between the number of these early-life risk factors and increased childhood adiposity and obesity. After adjusting for potential confounders, children exposed to four or five risk factors had a relative risk of being overweight or obese of 3.99 (95% CI 1.83, 8.67) at age 4 and 4.65 (95% CI 2.29, 9.43) at age 6, compared to children with no risk factors(Reference Robinson, Crozier and Harvey39). Similarly, in the GUSTO prospective cohort study, six key risk factors were examined: maternal pre-pregnancy overweight/obesity, paternal overweight/obesity at 24 months post-delivery, excessive gestational weight gain, raised maternal fasting glucose during pregnancy (≥5.1 mmol/L), breast-feeding duration <4 months and early introduction of solid foods (<4 months). The adjusted relative risk of overweight/obesity in children with four or more risk factors was 11.1 (95% CI 2.5, 49.1) at age 4, compared to children with no risk factors(Reference Aris, Bernard and Chen40). Early interventions to change these modifiable risk factors could therefore make a significant contribution to the prevention of childhood obesity. These findings have paved the way for a new series of systematic genome-wide epigenetic investigations to find epigenetic biomarkers associated with child adiposity. Higher methylation of the Retinoid X Receptor Alpha (RXRA) gene promoter at birth was associated with child’s later adiposity, and associations were also observed between levels of RXRA methylation and mothers’ carbohydrate intake(Reference Godfrey, Sheppard and Gluckman36). Further observational studies have characterised perinatal DNA methylation variations related to Antisense long Non-coding RNA in the INK4 Locus (ANRIL) that mark obesity risk, replicated across three populations and with relevant physiological effects of altering ANRIL methylation in vitro (Reference Lillycrop, Murray and Cheong41). Decreased methylation of the SLC6A4 promoter region, a transport protein responsible for reuptake of serotonin and which may play a role in appetite and energy balance, was associated with higher maternal gestational weight gain as well as increased adiposity in infancy, early childhood and adolescence(Reference Lillycrop, Garratt and Titcombe42). Additionally, obese adults had lower methylation levels and decreased gene expression in adipose tissue compared to lean individuals(Reference Lillycrop, Garratt and Titcombe42). These findings suggest that altered SLC6A4 promoter methylation may provide a consistent marker of adiposity throughout the life course.
In the GUSTO cohort study, associations of maternal nutrition with pregnancy outcomes, fetal growth and childhood outcomes were examined, and DNA methylation was assessed at birth. Dietary patterns rich in vegetables, fruits and white rice were associated with lower risk of preterm birth and larger birth size, suggesting beneficial effects on developmental and growth outcomes(Reference Chia, de Seymour and Colega27). Associations between low-quality maternal diet and night-eating behaviours with higher insulin in the offspring were demonstrated and interact synergistically, especially in boys(Reference Chen, Loy and Tint43). Additionally, in epigenome-wide association studies, higher maternal dietary glycaemic index and glycaemic load in pregnancy were associated with offspring cord blood DNA methylation at multiple CpG sites, with some relating to genes expressed in tissue relevant to metabolic health; associations were seen particularly in overweight and obese mothers(Reference Küpers, Fernández-Barrés and Mancano44). The findings of this study highlighted the importance of maternal nutrition in shaping the epigenetic landscape of the newborn and suggested that these early-life epigenetic modifications may have lasting impacts on health and disease risk later in life.
Emerging evidence supports a further influence of postnatal nutritional status on epigenetic processes. For example, a study in the ALSPAC cohort identified associations between rapid weight gain in infancy and small increases in childhood methylation at two CpG sites, one of which was replicated in the SWS and was also associated with subsequent overweight and obesity(Reference Robinson, Brown and Antoun45).
The importance of preconception health
Epidemiological, clinical and basic science research has identified the period around conception as being critical in the processes mediating parental influences on the next generation’s phenotype and health. During this time, from the maturation of gametes through to early embryonic development, the nutrition of mothers and fathers can adversely influence the offspring’s long-term risks of cardiovascular, metabolic, immune and neurological morbidities. Such ‘developmental programming’ has been demonstrated for exposures including maternal overnutrition and obesity, maternal undernutrition, related paternal factors and the use of assisted reproductive treatment(Reference Fleming, Watkins and Velazquez46). Human studies and animal models demonstrate the underlying biological mechanisms, including epigenetic, cellular, physiological and metabolic processes.
Paternal influences on epigenetic processes
While the main focus has been on mechanisms driven by maternal nutrition and other exposures, emerging evidence suggests that paternal nutrition and lifestyle also influence sperm epigenomics and transcriptomics, with consequences for the development of the offspring(Reference Fleming, Watkins and Velazquez46,Reference Vaz, Burton and Kermack47) . Meta-analysis of mouse paternal and maternal protein undernutrition indicates distinct parental periconceptional contributions to postnatal outcomes(Reference Fleming, Watkins and Velazquez46). A 6-week paternal dietary intervention modified the small non-coding RNA (sncRNA) profile of human sperm in a subset of participants from the PREPARE trial(Reference Vaz, Burton and Kermack47). sncRNA sequencing revealed that supplementation with olive oil, vitamin D and n-3 fatty acids altered the expression of 3 tRF, 15 miRNA and 112 piRNA, which target genes related to fatty acid metabolism and transposable elements in the sperm genome.
Evidence from human intervention studies
While nutritional-randomised controlled trials before and during pregnancy with offspring phenotyping are sparse, several such trials have reported findings of potential importance. For example, the UK Pregnancies Better Eating and Activity Trial (UPBEAT) of a nutritional intervention in women living with obesity(Reference Antoun, Kitaba and Titcombe48) showed that a low glycaemic behavioural intervention can indeed change epigenetic processes in the offspring, with potential effects on adiposity in infancy. A pregnancy low-GI diet reduced infant subscapular skinfold thickness at the age of 6 months by 0.26 SD, and this beneficial effect of a prenatal nutritional intervention was also dependent on breast-feeding for ≥3 months(Reference Patel, Godfrey and Pasupathy49). Maternal GDM, fasting and 1-h and 2-h glucose levels from the oral glucose tolerance test were associated with numerous differentially methylated CpG sites in the infant’s cord blood DNA. Notably, the methylation changes linked to GDM and 1-h glucose were attenuated by the lifestyle interventions during pregnancy(Reference Antoun, Kitaba and Titcombe48).
Maternal vitamin D insufficiency is highly prevalent in many populations worldwide, and this can alter fetal bone growth and have lasting effects on the child’s bone health. In vitamin D-insufficient mothers, splaying of the metaphysis (similar to childhood rickets) is present as early as 19 weeks of gestation(Reference Mahon, Harvey and Crozier50) and reduced concentration of 25(OH)-vitamin D in mothers during late pregnancy is associated with lower childhood bone-mineral content at age 9 years(Reference Javaid, Crozier and Harvey51).
Subsequent to the above observational studies, in the MAVIDOS trial, supplementation of 1000 IU/day cholecalciferol during pregnancy did not affect the primary outcome of offspring neonatal bone mineral content (BMC), but did show a significant increase in infant bone mass for winter births(Reference Cooper, Harvey and Bishop52). In the same cohort, effects of gestational supplementation became more apparent over time; the intervention was positively associated with bone mineral density (BMD) at age 4 years(Reference Curtis, Moon and D’Angelo53) and BMC and BMD at age 6–7 years(Reference Moon, D’ Angelo and Curtis54), suggesting a sustained beneficial effect of supplemental vitamin D supplementation in pregnancy on offspring bone health. Furthermore, significantly reduced DNA methylation at several CpG sites near the RXRA gene, known to play a role in bone metabolism, was observed in umbilical cord DNA(Reference Curtis, Krstic and Cook55), suggesting a potential epigenetic mechanism by which maternal vitamin D supplementation may influence fetal bone development. Supplementation with 1000 IU/day cholecalciferol did not reduce the incidence of preterm birth but was associated with a greater likelihood of a spontaneous vaginal delivery(Reference Moon, D’Angelo and Crozier56). MAVIDOS also provided the first randomised controlled trial evidence for the role of antenatal vitamin D supplementation in reducing the risk of infantile atopic eczema. The protective effects were seen in infants who breastfed more than one month but not in those who breastfed less than one month(Reference El-Heis, D’Angelo and Curtis16).
In our multicentre NiPPeR randomised controlled trial, we have reported that a maternal nutritional supplement taken preconception and during pregnancy substantially decreased the incidence of preterm birth, particularly cases associated with preterm pre-labour rupture of membranes(Reference Godfrey, Barton and El-Heis57). Moreover, nutritional intervention before and during pregnancy halved the incidence of obesity in the offspring at age 2 years(Reference Lyons-Reid, Derraik and Kenealy58). At recruitment preconception, over 90% of the trial participants had marginal or low concentrations of one or more of folate, riboflavin, vitamin B12 or vitamin D during, and many developed markers of vitamin B6 deficiency in late pregnancy(Reference Godfrey, Titcombe and El-Heis59). Current work is examining the influence of the nutritional intervention on offspring epigenetics in the NiPPeR trial.
Evidence now points to the opportunity to reduce maternity disparities through intervention before and between pregnancies. In 2024 the NIHR launched its first ‘Challenge’ funding call, focused on new ways to tackle inequalities in maternity care. The resulting NIHR Maternity Disparities Consortium (2025–2030) will bring together a diverse range of organisations, including nine lead UK universities who will collaborate with local councils, NHS trusts, charities, industry and other health organisations. The Consortium will focus on inequalities before, during and after pregnancy(60).
Conclusion
In conclusion, the evidence from cohort and intervention studies highlights the critical importance of nutrition in early life for shaping lifelong health outcomes. Maternal diet during pregnancy is a key determinant of fetal development and has lasting impacts on health and disease risk later in life. Epigenetics is a key mechanism by which early life nutrition can influence lifelong health, and interventions during early life have been shown to improve health outcomes later in life. Further research is needed to fully understand the relationships between nutrition in early life, epigenetics and lifelong health and to develop effective interventions to promote optimal health outcomes.
Collectively, the evidence supports lasting effects of maternal and infant nutrition on offspring health and human potential, with epigenetic processes likely to be an important underpinning mechanism.
Acknowledgements
We thank the participants and their families for their enthusiastic involvement in the MAVIDOS and NiPPeR studies.
Author contributions
Keith M Godfrey: Conceptualisation and writing original draft
Paula Costello: Co-drafting review and editing
Sarah El-Heis: Writing review and editing
Financial support
KMG is supported by the UK Medical Research Council (MC_UU_12011/4), the National Institute for Health and Care Research (NIHR Senior Investigator (NF-SI-0515-10042) and NIHR Southampton Biomedical Research Centre (NIHR203319)), and the Wessex Medical Trust, the Gerald Kerkut Charitable Trust and the Rosetrees Trust. For the purpose of Open Access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.
Competing interests
KMG has received reimbursement for speaking at conferences sponsored by companies selling nutritional products and is part of an academic consortium that has received research funding from Bayer, Boehringer Ingelheim, Nestec, BenevolentAI Bio Ltd and Danone.
Open access
