Premature Epigenetic Aging, Deviations from Normative Brain Development, and Cognitive Performance in Young Adulthood

Premature Epigenetic Aging, Deviations from Normative Brain Development, and Cognitive Performance in Young Adulthood | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Premature Epigenetic Aging, Deviations from Normative Brain Development, and Cognitive Performance in Young Adulthood Klara Mareckova, Lukas Pelant, Radek Marecek, Anna Pacinkova, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9400608/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 10 You are reading this latest preprint version Abstract Premature epigenetic aging has been linked to poorer cognitive performance, but the neuroanatomical pathways underlying this association remain unclear. Therefore, we tested whether premature epigenetic aging predicts deviations from normative brain development and cognition in young adulthood. We followed 254 young adults from the ELSPAC prenatal birth cohort. Epigenetic aging was estimated using Horvath’s DNA methylation clock, CheekAge, and AltumAge. Structural MRI was processed with FreeSurfer to derive cortical thickness and surface area in 68 Desikan–Killiany regions and subcortical volumes in 14 Aseg regions. CentileBrain normative models trained on 37,407 MRI scans from individuals aged 3–90 years were used to quantify age- and sex-specific deviations from normative brain development. Cognitive performance was assessed using the WAIS-IV. Premature epigenetic aging estimated using Horvath’s clock was associated with more negative deviations from normative volume of the putamen and caudate, but not with other subcortical volumes, cortical thickness, or surface area. In women, this premature epigenetic aging was also associated with lower full-scale IQ and performance IQ, and moderated mediation analyses indicated that the deviations in putamen volume mediated the relationship between premature epigenetic aging and cognitive performance. In men, the relationship between premature epigenetic aging and cognitive performance was quadratic (U-shaped). No similar effects were observed when using CheekAge or AltumAge.Premature epigenetic aging estimated using Horvath’s clock and the associated deviations in development of the dorsal striatum in young adulthood may represent an early marker of poorer cognitive performance long before clinical symptoms might emerge. Health sciences/Pathogenesis Biological sciences/Neuroscience Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION The global population is undergoing profound demographic change. Although some regions continue to experience rapid population growth, population aging represents the dominant global trend. Declining fertility rates, alongside reduced mortality, have led to a steadily increasing proportion of older individuals worldwide (1, 2). In 2020, the global older-age dependency ratio (ODR) was 17, corresponding to 17 adults aged 65 years and older per 100 working-age individuals, and projections indicate an increase to 21 by 2030 and 26 by 2040 (2). Given the increasing percentage of adults over 65, there is also a higher prevalence of dementia as well as other health-related issues (3, 4, 5, 6) A growing body of evidence indicates that many age-related diseases originate in young adulthood or even early life (7, 8, 9), underscoring the need for a lifespan perspective on aging-related pathology. Identifying early biological markers and characterizing longitudinal neurodevelopmental trajectories are critical for advancing prevention strategies. Since chronological age alone fails to capture interindividual variability in biological aging, epigenetic clocks and normative models of brain development, allowing us to quantify deviations from age- and sex-specific development, have emerged as more informative and reliable estimates of biological aging. Epigenetic clocks have been increasingly recognized as predictors of health, cognition, and mortality (10, 11). Epigenetic clocks differ substantially in their biological targets, tissue specificity, and intended phenotypic scope. The first-generation epigenetic clocks (e.g., Horvath’s DNA methylation clock, Hannum’s DNA methylation clock) were designed primarily to predict chronological age based on DNA methylation patterns (11, 12). For example, the Horvath’s clock, based on DNA methylation profiles of 353 CpG sites, accurately predicts the age of different tissues, including the brain tissue (11). Horvath also demonstrated that the heritability of age acceleration is 100% in the newborns, but only 39% in older subjects, suggesting non-genetic factors become more relevant in later life (11). The second-generation epigenetic clocks (e.g., GrimAge (13), AltumAge (14), CheekAge (15)) have been explicitly optimized to predict health-related outcomes beyond chronological age, including disease risk, functional decline, and mortality. These clocks incorporate a substantially larger number of CpG sites and leverage machine learning approaches to capture methylation patterns associated with metabolic status, inflammation, lifestyle factors, and cumulative disease burden (14, 15). For example, the AltumAge, a deep-learning–based pan-tissue clock, utilizes more than 20,000 CpG sites and predicts chronological age and mortality (14). The CheekAge, developed specifically from buccal epithelial tissue, utilizes approximately 200,000 CpG sites to reflect lifestyle and health and shows strong associations with BMI and substance use (15). Epigenetic aging captures processes relevant to neurodevelopment (11, 15, 16), metabolic load (15, 17), and cellular aging (11, 18, 19). Research from our group showed that premature epigenetic aging, calculated using the Horvath’s clock, is associated with worse cognitive skills in women already in young adulthood (8) and that premature epigenetic aging relates to altered functional brain dynamics and cognition primarily in women (20). Previous structural neuroimaging research has consistently demonstrated that individual differences in cognitive performance and intelligence correlate with variations in regional brain macrostructure, including cortical thickness and subcortical volumes (21, 22, 23). However, the underlying brain structure reflecting the altered brain dynamics and cognitive performance is not well understood. Normative modeling of brain development, which allows quantification of individual-level deviations from age- and sex-specific normative brain development, might offer a unique insight into the relationships between altered epigenetic aging and altered cognitive performance. Large-scale normative models of brain development, such as the CentileBrain (24), trained on 37 407 MRI scans from individuals aged 3–90 years, allow us to characterize these structural processes with precision. Therefore, the current study aims to combine the estimates of premature epigenetic aging with deviations from normative brain development and cognitive performance and offer a unique analytical framework allowing for the detection of early signs of aging-related pathology. Specifically, we aim to test whether the relationship between premature epigenetic aging and worse cognitive outcomes in young adulthood might be mediated by deviations from normative development of brain structure. Further, we aim to determine whether these relationships between premature epigenetic aging, deviations from normative brain development, and cognitive outcomes might be sex-specific. MATERIALS AND METHODS Participants A total of 254 young adults (51.6% male) aged 28–30 years (M=28.97, SD=0.68) from the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) prenatal birth cohort (25) participated also in its neuroimaging – epigenetic follow-up in young adulthood, the Health Brain Age (HBA) study, and had complete data on brain structure, DNA methylation, and cognition. Demographic information regarding these participants is available in Supplementary Table 1 . For AltumAge, two extreme outliers, defined a priori as absolute z-scores > 3, were excluded before descriptive statistics and sex-comparison analyses (independent-samples t-test). All other analyses were conducted on the full available sample. All participants provided written informed consent to participate in the HBA study, including the agreement to merge data from HBA with their historical data from ELSPAC. Ethical approval for the HBA study was obtained from the ELSPAC ethics committee (ELSPAC/EK/2/2020), and all methods were performed in accordance with the relevant guidelines and regulations. Assessment of DNA Methylation and Epigenetic Aging DNA was isolated from buccal swabs collected in the late 20s, and DNA methylation was assessed using the Illumina EPIC Platform. Epigenetic aging was estimated using three different epigenetic clocks: Horvath’s clock (11), which we previously applied in Mareckova et al. (8, 25, 27), as well as the AltumAge clock (14) and the CheekAge clock (15), both of which were implemented and described in detail in Mareckova et al. (27), including the full analysis code. As described in our previous work (8, 26, 27), we used the R package ChAMP (28) to process the raw Illumina microarray data, and the beta mixture quantile normalization (BMIQ) (29) method to adjust the beta-values of type II design probes into a statistical distribution characteristic of type I probes. DNA methylation age was calculated using Horvath’s clock (11), based on 353 CpG sites, using the AltumAge clock (14), based on 20 318 CpG sites, and using the CheekAge clock (15), based on 115 533 CpG sites. See Supplementary Methods inMareckova et al. (27)for details and the link to the GitHub Repository. For each of these clocks, the epigenetic age gap estimate (EpiAGE) was calculated as the DNA methylation age residualized for batch, chronological age, and the proportion of epithelial cells. Positive values of EpiAGE reflect accelerated aging, and negative values decelerated aging. Structural Magnetic Resonance Imaging Structural magnetic resonance imaging (MRI) of the brain was performed on a 3T Siemens Prisma MRI scanner with a 64-channel head coil. T1-weighted (T1w) whole-brain MPRAGE images were acquired using a 64-channel head/neck coil with acquisition parameters: voxel size 1 mm 3 , repetition time (TR) 2,300 ms, echo time (TE) 2.34 ms, inversion time (TI) 900 ms, flip angle 8°. MRI data were processed in FreeSurfer v7.1.1, and cortical thickness and surface area in 68 regions from the Desikan-Killiany (DK) atlas, and volumes of 14 subcortical regions from the Aseg atlas were calculated. For quality control, all the registration and segmentation outputs from FreeSurfer were visually checked and passed the quality control. Normative Modeling and Calculation of the Deviations from Normative Brain Development Pretrained models from the CentileBrain Initiative (24) were employed to compute the age- and sex- specific normative deviation (z) scores for all measures (68 surface area regions, 68 cortical thickness regions, and 14 subcortical regions). These models were trained on 37 407 MRI scans from individuals aged 3–90 years (24). The resulting z-scores reflect whether each measure is larger or smaller than expected for an individual of the same age and sex. No further adjustment for intracranial volume (ICV) was performed, as correction is inherently part of the CentileBrain model calculation. Assessment of Cognitive Performance Cognition was assessed using the following subtests from the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV) (30): Picture completion, Matrix Reasoning, Arithmetic, Digit span, Information, Similarities, and Digit-symbol coding. These subtests allowed us to assess participants‘ perceptual reasoning (Picture completion, Matrix reasoning), working memory (Arithmetic, Digit span), verbal comprehension (Information, Similarities), and processing speed (Digit-symbol coding). Full-scale IQ was calculated based on the scores for these 7 WAIS-IV subtests according to Tam (31). Performance IQ (PIQ) was estimated from subtests assessing perceptual reasoning (Picture Completion, Matrix Reasoning) and processing speed (Digit-Symbol Coding). Verbal IQ (VIQ) was derived from subtests evaluating verbal comprehension (Information, Similarities) and working memory (Arithmetic, Digit Span). Statistical Analyses Associations between EpiAGE, deviations from normative brain development, and cognitive performance were evaluated using linear regressions, and potential interactions with sex were considered. We also evaluated potential non-linear (quadratic) relationships between epigenetic aging and cognitive performance to capture potential complex dynamics not revealed by linear models. Multiple comparisons were corrected using the False Discovery Rate (FDR) method within each family of analyses. Finally, a moderated mediation analysis assessed whether the relationship between premature epigenetic aging and worse cognitive performance might be mediated by the epigenetic aging-related changes in normative brain development and whether these relationships differ by sex. The indirect effect was assessed using Model 14 of the PROCESS macro (32) for SPSS, version 5.0 (IBM SPSS Statistics, IBM Corp, Armonk, NY). Conditional indirect effects were obtained for significance using bootstrapped (10 000 iterations) bias-corrected 95% confidence intervals constructed around indirect effect estimates. RESULTS Premature epigenetic aging and its relationship with normative brain development Premature epigenetic aging calculated using the Horvath’s clock was associated with greater negative deviations from age- and sex-specific normative development of putamen (left: R 2 =0.02, FDRp=0.02, Figure 1A ; right: R 2 =0.02, FDRp=0.03, Figure 1B ) and caudate (left: R 2 =0.03, FDRp=0.01, Figure 1C ); right: R 2 =0.04, FDRp=0.01, Figure 1D ; see statistical details in Table 1 ), but no deviations from normative development of the other subcortical volumes (p>0.14, FDRp>0.38), cortical thickness in the 68 regions (p>0.05; FDRp>0.91), or surface area in the 68 regions (p>0.07, FDRp>0.94). No similar results were found using the two epigenetic clocks developed to reflect not only aging as such but also disease and mortality: CheekAge (cortical thickness in the 68 regions p>0.02, FDRp>0.49; surface area in the 68 regions p>0.04, FDRp>0.83; volume of the 14 subcortical regions p>0.07, FDRp>0.67) and AltumAGE (cortical thickness in the 68 regions p>0.003, FDRp>0.18; surface area in the 68 regions p>0.009, FDRp>0.54; volume of the 14 subcortical regions p>0.10, FDRp>0.63). All statistical details are provided in Supplementary Table 2 . Table 1 - Premature epigenetic aging and deviations from normative development of caudate and putamen. Standardized regression coefficients, nominal p values, and FDR-corrected p values are shown for associations of Horvath’s clock, CheekAge, and AltumAge. Significant associations after FDR correction are shown in bold. Horvath's clock CheekAge AltumAge Hem Region Std Beta p FDR p Std Beta p FDR p Std Beta p FDR p L caudate -0.198 0.002 0.012 -0.065 0.299 0.697 -0.079 0.212 0.631 R caudate -0.212 0.001 0.011 -0.074 0.240 0.671 -0.103 0.102 0.631 L putamen -0.182 0.004 0.017 -0.075 0.230 0.671 -0.062 0.319 0.631 R putamen -0.159 0.013 0.044 -0.101 0.109 0.671 -0.081 0.197 0.631 Premature epigenetic aging and cognitive performance Sex moderated the relationship between premature epigenetic aging calculated using Horvath’s clock and cognitive performance (interaction statistics for FSIQ: β=−0.192, p=0.029; interaction statistics for PIQ: β=−0.183, p=0.034). Post-hoc sex-stratified linear regressions indicated that premature epigenetic aging (Horvath’s clock) was associated with lower cognitive performance in women but not in men (FSIQ: men β=0.076, p=0.390; Figure 2A ; women β=−0.200, R²=0.06, p=0.027; Figure 2B ; PIQ: men β=0.011, p=0.905; Figure 2C ; women β=−0.257, R²=0.08, p=0.004; Figure 2D ). However, quadratic models revealed significant U-shaped relationships between premature epigenetic aging and cognitive performance in men (FSIQ: β=0.211, p=0.016, R²=0.05; Figure 2A ; PIQ: β=0.273, p=0.002, R²=0.07; Figure 2C ) but not in women (FSIQ: β=0.130, p=0.146; Figure 2B ; PIQ: β=0.127, p=0.148; Figure 2D ). In women, the linear term remained significant in the quadratic models (FSIQ: β=−0.204, p=0.023; PIQ: β=−0.261, p=0.003), consistent with a predominantly linear decline. No comparable associations were observed for CheekAGE or AltumAGE in analogous models (all p > 0.07). Neither of these clocks also showed the male-specific quadratic curvature seen for Horvath’s clock (all quadratic-term p’s ≥ .057). Normative brain development and cognitive performance Further analyses revealed women-specific relationships between greater negative deviations from normative development of putamen and lower full-scale IQ (left: sex × putamen interaction t = 2.62, p = 0.009; women: β = 0.28, p = 0.002, R² = 0.08; men: β = −0.03, p = 0.702, Figure 3A ; right: sex × putamen interaction t = 2.30, p = 0.022; women: β = 0.23, p = 0.012, R² = 0.05; men: β = −0.06, p = 0.531, Figure 3B , and particularly lower performance IQ (left: sex × putamen interaction t = 2.51, p = 0.013; women: β = 0.28, p = 0.002, R² = 0.08; men: β = −0.02, p = 0.835, Figure 3C ; right: sex × putamen interaction t = 2.16, p = 0.031; women: β = 0.23, p = 0.012, R² = 0.05; men: β = −0.04, p = 0.652, Figure 3D ). Quadratic terms were not significant in any sex-stratified post-hoc models (all p ≥ 0.171). Do deviations from normative brain development mediate the relationship between premature epigenetic aging and worse cognitive performance? Moderated mediation analyses indicated that the association between Horvath’s EpiAGE and cognitive performance was mediated by normative development of the putamen in women but not men. For full-scale IQ, the indirect effects were significant in women (left: ab = −0.186, 95% CI [−0.457, −0.036]; right: ab = −0.134, 95% CI [−0.368, −0.016]) but not in men (left: ab = 0.026, 95% CI [−0.075, 0.211]; right: ab = 0.035, 95% CI [−0.048, 0.202]). The index of moderated mediation was significant for both the left (index = −0.212, 95% CI [−0.567, −0.026]) and right putamen (index = −0.169, 95% CI [−0.483, −0.015]). Similarly, for performance IQ, the indirect effects were significant in women (left: ab = −0.218, 95% CI [−0.523, −0.040]; right: ab = −0.152, 95% CI [−0.411, −0.014]) but not in men (left: ab = 0.030, 95% CI [−0.086, 0.246]; right: ab = 0.037, 95% CI [−0.065, 0.223]). The index of moderated mediation was significant for both the left (index = −0.248, 95% CI [−0.667, −0.030]) and right putamen (index = −0.189, 95% CI [−0.545, −0.009]). See Figure 4 for statistical details. DISCUSSION We performed a neuroimaging–epigenetic follow-up of a prenatal birth cohort and showed that premature epigenetic aging in young adulthood, calculated using Horvath’s epigenetic clock, is associated with altered normative development of the dorsal striatum and that in women, the deviations from normative development of the putamen mediate the relationship between premature epigenetic aging and worse cognition. These findings substantially extend our previous research in the same cohort regarding premature epigenetic aging, brain dynamics, and cognition, and reveal the role of altered structural brain development and specifically the importance of the dorsal striatum in these relationships. Moreover, given that dopaminergic neurons in the substantia nigra release dopamine to the dorsal striatum, our findings align with previous PET research on age-related dopaminergic decline in the dorsal striatum (33, 34), as well as further research on age-related dopamine loss and cognitive deficits (10). The dorsal striatum, consisting of the caudate and putamen, is particularly relevant due to its involvement in dopaminergic circuits that support motor learning, cognitive speed, working memory, and reinforcement learning (35, 36). These domains exhibit early decline in dopaminergic aging models (10, 33, 34). The specifically larger associations between putamen and performance IQ, as compared to verbal IQ, described in the current study, are consistent with further evidence linking basal ganglia volume with figural and numeric, but not verbal, aspects of intelligence (23). While premature epigenetic aging estimated using Horvath’s clock was associated with deviations in both caudate and putamen, the mediatory effect on cognition was specific to the putamen. This may reflect the putamen’s prominent role in sensorimotor circuits and habitual control, which are closely linked to processing speed components of Performance IQ (35). In contrast, the caudate is deeply integrated into frontoparietal executive loops (37, 38), which might be more resilient or functionally compensated for in this healthy young cohort. A striking finding of our study is that the relationship between epigenetic aging and cognition differs by sex, not only in magnitude but also functionally. While women exhibited a direct, predominantly linear “dose–response” pattern, in which premature epigenetic aging was consistently associated with more negative deviations from normative striatal development and lower cognitive performance, men displayed a significant quadratic (U-shaped) relationship. This non-monotonic profile indicates that higher cognitive scores were observed at both the lower and upper extremes of the epigenetic aging spectrum, whereas intermediate EpiAGE values corresponded to lower performance. Although we confirmed that these results were not driven by statistical outliers (all standardized residuals |z| < 3; Cook’s distance < 1), the pattern suggests that, in young men, the response to epigenetic variance may be more complex than a simple linear decline. One plausible (but speculative) interpretation is the presence of sex-specific non-linear compensatory mechanisms. While classic biphasic or hormetic stress responses typically involve low-dose adaptation followed by high-dose toxicity or functional decline (39), the U-shaped pattern in young men implies a different dynamic. It suggests that more advanced epigenetic aging might act as a systemic challenge that triggers robust compensatory neural processes. Importantly, cognitive reserve is not merely a passive buffer against late-life pathology, but a dynamic capacity that can be actively mobilized to cope with demanding conditions already in young adulthood (40, 41). At the molecular level, this aligns with emerging neuroepigenetic evidence demonstrating that male and female brains engage distinct chromatin and transcriptional networks in response to systemic or biological challenges, providing a fundamental mechanism for sex-specific adaptation and resilience (42). The EpiAGE–putamen–cognition mediation pathway observed in women replicates and extends previous sex-specific fMRI findings from the same cohort (20) and points to a specific vulnerability of the female striatal system to premature epigenetic aging. Rather than being an isolated phenomenon, this susceptibility aligns with distinct sex differences observed in the structural maturation of subcortical volumes, including the putamen, throughout adolescence and early adulthood (44). Furthermore, this vulnerability may be rooted in the profound sensitivity of female striatal and dopaminergic circuits to the interplay between ovarian hormones and epigenetic programming (42, 45). Consequently, premature epigenetic aging may disproportionately disrupt these sensitive structural-functional couplings in young women. In contrast, the U-shaped curve in men implies a more complex interaction. It suggests that in young men, the impact of epigenetic deviations is not monotonic; moderate deviations might be buffered by compensatory neural mechanisms or higher physiological resilience thresholds that are not present in women (10). This aligns with the hypothesis of differential reserve capacities in early aging, where men may maintain cognitive function despite structural loads up to a certain tipping point, whereas women show earlier structural-functional coupling (10, 23). Furthermore, the absence of a significant mediation effect via the putamen in men reinforces that their cognitive variance related to aging is likely driven by different neural substrates or non-linear network reconfigurations that simple linear mediation models cannot capture. Potential explanations for this sex specificity include innate sex differences in striatal dopamine D2 receptor density and affinity (46), estrogen-mediated neuroprotection of the nigrostriatal dopamine system (47), or a differential sensitivity of the female brain to cumulative metabolic stress and environmental insults (48). Our analyses revealed a distinct sensitivity of Horvath’s clock to striatal development and cognitive performance, while other health-related clocks (CheekAge, AltumAge) failed to show significant associations. This divergence is likely attributable to the specific biological properties of the Horvath clock in the context of young adulthood. Horvath’s clock is a pan-tissue estimator tightly correlated with chronological age and cellular developmental processes, making it particularly suitable for detecting deviations from normative trajectories (11, 49). Notably, the male-specific non-linear associations observed between Horvath’s clock and cognitive performance were not reproduced by CheekAge or AltumAge, supporting clock-specific sensitivity to the processes captured in young adulthood. The second-generation or tissue-specific clocks are often optimized to predict mortality, environmental exposures, or physiological dysregulation associated with advanced aging and disease burden (50, 51). Although CheekAge was developed specifically from buccal epithelial tissue and was thus expected to show high sensitivity in our samples (15), it failed to show significant associations with brain structure or cognition. Therefore, the observed associations suggest that the variance in striatal volume and fluid intelligence is driven by an intrinsic biological aging rate captured by the Horvath clock’s rather than by external health deficits or cumulative physiological dysregulation (52). Together, these findings suggest that Horvath’s clock captures variance related to neurodevelopmental timing rather than cumulative health burden, which may explain its unique sensitivity to brain development and cognition in young adulthood. The main strengths of our study are the unique combination of neuroimaging and epigenetic data on a prenatal birth cohort, the use of three different epigenetic clocks as well as a robust large scale model of normative brain development, trained on 37 407 MRI scans from individuals aged 3–90 years old, and the even distribution of men and women allowing us to assess the potential interactions with sex. We bring novel evidence showing that premature epigenetic aging in young adulthood is distinctly associated with deviations from normative development in parts of the basal ganglia, and that these structural alterations mediate the relationship between accelerated epigenetic aging and worse cognitive performance. Still, our findings are limited by the cross-sectional nature of the current study, which precludes causal inference. While we modeled deviations in brain structure as a mediator between epigenetic aging and cognition, it is plausible that the relationship is bidirectional; for example, lifestyle factors associated with higher intelligence could influence both brain integrity and epigenetic aging rates. Future research should follow these participants longitudinally to reveal whether the premature epigenetic aging and the associated changes in normative brain development and cognition might develop into deteriorating brain health and more severe cognitive outcomes in older adulthood. Moreover, since the quadratic effects can be sensitive to model specification, future work should test robustness using alternative non-linear parametrizations (e.g., splines) and, ideally, longitudinal data to determine whether the observed U-shaped pattern in men reflects developmental dynamics or subgroup heterogeneity. In addition, since our DNA methylation data were based on buccal swab samples, we were not able to calculate other second-generation epigenetic clocks, such as the PhenoAge (50), and assess their relationship with normative brain development and cognitive outcomes. Therefore, future research should aim to collect blood from the participants and extend our findings with these additional analyses. Overall, this study provides novel evidence linking premature epigenetic aging to deviations from normative development of the dorsal striatum in young adulthood and worse cognitive performance in women. These findings indicate sex-specific neurodevelopmental pathways linking epigenetic aging markers to cognition and underscore the value of normative modeling in uncovering subtle deviations in brain maturation. Moreover, our study reveals that premature epigenetic aging in young adulthood and the associated deviations in the development of the dorsal striatum might be an early marker of worse cognitive performance long before clinical symptoms may develop. Our findings also suggest that the dorsal striatum and its function, including cognitive flexibility, goal-directed behavior, planning, working memory, attention, and learning, might be potential targets to prevent premature epigenetic aging and cognitive decline in women. Acknowledgemnts This work has received funding from the Czech Science Foundation, project no. 24-12183M, Czech Health Research Council (No. NU20J-04-00022), the Czech Ministry of Education, Youth and Sports (MEYS CR) (CEITEC 2020, LQ1601), and by project no. LX22NPO5107 (MEYS): Funded by European Union – Next Generation EU. Support with obtaining scientific data presented in this paper came from the core facility Multimodal and Functional Imaging Laboratory of Central European Institute of Technology, Masaryk University, supported by the Czech-BioImaging large RI project (No. LM2018129, funded by MEYS CR). Authors also thank the RECETOX Research Infrastructure (No LM2023069), financed by MEYS CR for its supportive background. This work was also supported by the European Union’s Horizon 2020 research and innovation program under grant agreement No 857560 (CETOCOEN Excellence). This publication reflects only the author's view, and the European Commission is not responsible for any use that may be made of the information it contains. Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. Dr. Nikolova is supported by Koerner New Scientist Award from the CAMH Foundation and a Discovery Grant from the National Sciences and Engineering Research Council of Canada (NSERC). Declarations Acknowledgemnts This work has received funding from the Czech Science Foundation, project no. 24-12183M, Czech Health Research Council (No. NU20J-04-00022), the Czech Ministry of Education, Youth and Sports (MEYS CR) (CEITEC 2020, LQ1601), and by project no. LX22NPO5107 (MEYS): Funded by European Union – Next Generation EU . Support with obtaining scientific data presented in this paper came from the core facility Multimodal and Functional Imaging Laboratory of Central European Institute of Technology, Masaryk University, supported by the Czech-BioImaging large RI project (No. LM2018129, funded by MEYS CR). Authors also thank the RECETOX Research Infrastructure (No LM2023069), financed by MEYS CR for its supportive background. This work was also supported by the European Union’s Horizon 2020 research and innovation program under grant agreement No 857560 (CETOCOEN Excellence). This publication reflects only the author's view, and the European Commission is not responsible for any use that may be made of the information it contains. Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. Dr. Nikolova is supported by Koerner New Scientist Award from the CAMH Foundation and a Discovery Grant from the National Sciences and Engineering Research Council of Canada (NSERC). Conflicts of Interest All authors declare no financial or any other conflicts of interest. References He W, Goodkind D, Kowal P. An Aging World: 2015 . International Population Reports P95/16-1. U.S. Census Bureau; 2016. U.S. Census Bureau. Story Map - An Aging World: 2020 . U.S. Census Bureau. Published 2020. Accessed April 12, 2026. Chang AY, Skirbekk VF, Tyrovolas S, Kassebaum NJ, Dieleman JL. Measuring population ageing: an analysis of the Global Burden of Disease Study 2017. Lancet Public Health . 2019;4:e159-e167. Chen C et al. A global analysis of adaptation to societal aging across low-, middle- and high-income countries using the Global Aging Society Index. Nat Aging . 2025;5:113-121. Ferrari AJ et al. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet. 2024;403:2133-2161. GBD 2019 Ageing Collaborators. Global, regional, and national burden of diseases and injuries for adults 70 years and older: systematic analysis for the Global Burden of Disease 2019 Study. BMJ . 2022;376:e068208. Bateman RJ et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med . 2012;367:795-804. Mareckova K et al. Longitudinal study of epigenetic aging and its relationship with brain aging and cognitive skills in young adulthood. Front Aging Neurosci . 2023;15:1215957. Moffitt TE, Belsky DW, Danese A, Poulton R, Caspi A. The longitudinal study of aging in human young adults: knowledge gaps and research agenda . J Gerontol A Biol Sci Med Sci. 2017;72:210-215. Backman L, Nyberg L, Lindenberger U, Li SC, Farde L. The correlative triad among aging, dopamine, and cognition: current status and future prospects. Neurosci Biobehav Rev . 2006;30:791-807. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol . 2013;14:R115. Hannum G et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol Cell . 2013;49:359-367. Lu AT et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019;11:303-327. de Lima Camillo LP, Lapierre LR, Singh R. A pan-tissue DNA-methylation epigenetic clock based on deep learning. npj Aging . 2022;8:4. Shokhirev MN et al. CheekAge: a next-generation buccal epigenetic aging clock associated with lifestyle and health. GeroScience . 2024;46:3429-3443. Sanli E, Kabaran S. Maternal obesity, maternal overnutrition and fetal programming: effects of epigenetic mechanisms on the development of metabolic disorders. Curr Genomics . 2019;20:419-427. Vidal CM et al. Insulin resistance in women correlates with chromatin histone lysine acetylation, inflammatory signaling, and accelerated aging. Cancers (Basel). 2024;16:2735. Pereira B et al. Epigenetic reprogramming as a key to reverse ageing and increase longevity. Ageing Res Rev . 2024;95:102204. Warner B, Ratner ER, Datta A, Lendasse A. A systematic review of phenotypic and epigenetic clocks used for aging and mortality quantification in humans. Aging (Albany NY) . 2024;16:12414-12427. Jordanek T et al. Accelerated epigenetic aging and its role in brain dynamics and cognition in young adulthood. Hum Brain Mapp . 2025;46:e70261. Luders E, Narr KL, Thompson PM, Toga AW. Neuroanatomical correlates of intelligence. Intelligence . 2009;37:156-163. Narr KL et al. Relationships between IQ and regional cortical gray matter thickness in healthy adults. Cereb Cortex . 2007;17:2163-2171. Rhein C et al. Neuroanatomical correlates of intelligence in healthy young adults: the role of basal ganglia volume. PLoS One . 2014;9:e93623. Ge R et al. Normative modelling of brain morphometry across the lifespan with CentileBrain: algorithm benchmarking and model optimisation. Lancet Digit Health . 2024;6:e211-e221. Piler P et al. Cohort profile: the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) in the Czech Republic. Int J Epidemiol . 2017;46:1379-1379f. Mareckova K et al. Epigenetic clock as a correlate of anxiety. Neuroimage Clin . 2020;28:102458. Mareckova K et al. Functional impact score of mitochondrial variants and its relationship with functional connectivity of the brain: potential origins of premature aging in young adulthood. Hum Brain Mapp . 2026;47:e70447. Tian Y et al. ChAMP: updated methylation analysis pipeline for Illumina BeadChips. Bioinformatics . 2017;33:3982-3984. Teschendorff AE et al. A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. Bioinformatics . 2013;29:189-196. Wechsler D. WAIS-IV: Wechsler Adult Intelligence Scale . 4th edn. Pearson; 2008. Tam WCC. The utility of seven-subtest short forms of the Wechsler Adult Intelligence Scale-III in young adults. J Psychoeduc Assess . 2004;22:62-71. Hayes AF. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. 2nd edn. Guilford Press; 2018. van Dyck CH et al. Age-related decline in dopamine transporters: analysis of striatal subregions, nonlinear effects, and hemispheric asymmetries. Am J Geriatr Psychiatry . 2002;10:36-43. Kim JH et al. Effects of age on dopamine D2 receptor availability in striatal subdivisions: a high-resolution positron emission tomography study. Eur Neuropsychopharmacol . 2011;21:885-891. Balleine BW, Delgado MR, Hikosaka O. The role of the dorsal striatum in reward and decision-making. J Neurosci . 2007;27:8161-8165. Zheng Q et al. Functional differentiation of the dorsal striatum: a coordinate-based neuroimaging meta-analysis. Quant Imaging Med Surg . 2023;13:471-488. Rieckmann A, Johnson KA, Sperling RA, Buckner RL, Hedden T. Dedifferentiation of caudate functional connectivity and striatal dopamine transporter density predict memory change in normal aging. Proc Natl Acad Sci USA . 2018;115:10160-10165. Choi EY, Yeo BTT, Buckner RL. The organization of the human striatum estimated by intrinsic functional connectivity. J Neurophysiol . 2012;108:2242-2263. Calabrese EJ et al. Biological stress response terminology: integrating the concepts of adaptive response and preconditioning stress within a hormetic dose-response framework. Toxicol Appl Pharmacol . 2007;222:122-128. Zihl J, Fink T, Pargent F, Ziegler M, Buhner M. Cognitive reserve in young and old healthy subjects: differences and similarities in a testing-the-limits paradigm with DSST. PLoS One . 2014;9:e84590. Stern Y. What is cognitive reserve? Theory and research application of the reserve concept. J Int Neuropsychol Soc . 2002;8:448-460. Kundakovic M, Tickerhoof M. Epigenetic mechanisms underlying sex differences in the brain and behavior. Trends Neurosci. 2024;47:18-35. Mareckova K, Devenyi GA, Andryskova L, Chakravarty MM, Nikolova YS. Maturation of hippocampal subfields in young adulthood and its relationship with cognition. Hum Brain Mapp. 2025;46:e70296. Koolschijn PCMP, Crone EA. Sex differences and structural brain maturation from childhood to early adulthood. Dev Cogn Neurosci . 2013;5:106-118. Ngun TC, Ghahramani N, Sanchez FJ, Bocklandt S, Vilain E. The genetics of sex differences in brain and behavior. Front Neuroendocrinol. 2011;32:227-246. Pohjalainen T, Rinne JO, Nagren K, Syvalahti E, Hietala J. Sex differences in the striatal dopamine D2 receptor binding characteristics in vivo. Am J Psychiatry . 1998;155:768-773. Dluzen DE. Neuroprotective effects of estrogen upon the nigrostriatal dopaminergic system. J Neurocytol . 2000;29:387-399. McEwen BS. Stress, sex, and neural adaptation to a changing environment: mechanisms of neuronal remodeling. Ann NY Acad Sci . 2010;1204:E38-E59. Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nat Rev Genet . 2018;19:371-384. Levine ME et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY) . 2018;10:573-591. Bell CG et al. DNA methylation aging clocks: challenges and recommendations. Genome Biol . 2019;20:249. Belsky DW et al. Quantification of biological aging in young adults. Proc Natl Acad Sci USA . 2015;112:E4104-E4110. Additional Declarations The authors have declared there is NO conflict of interest to disclose Supplementary Files SupplementaryTable2.docx Suplementary Table 2 SupplementaryTable1.docx Supplementary Table 1 Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: revise 11 May, 2026 Review # 2 received at journal 07 May, 2026 Review # 1 received at journal 04 May, 2026 Reviewer # 2 agreed at journal 22 Apr, 2026 Reviewer # 1 agreed at journal 22 Apr, 2026 Reviewers invited by journal 21 Apr, 2026 Editor assigned by journal 14 Apr, 2026 Submission checks completed at journal 14 Apr, 2026 First submitted to journal 14 Apr, 2026 Unknown event 13 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9400608","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":627119074,"identity":"0810dba5-59a2-4c08-8871-70c56628ec04","order_by":0,"name":"Klara Mareckova","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIiWNgGAWjYBACCXYEmxlEyIGIAw/waWFG02IM1pJAipbEBhCJT4tkM/OzDz8Ytsmbtzc/NvjYZpc+P+zwQ6AtdnK6Ddi1SDOzGc/sYbhtOOfMMePEmW3JuRtvpxkAtSQbmx3ArkWOGeh6HobbjDMkcpgP85xhzt04OwGk5UDiNpxa2D8z/mG4bT9D/g1IS3264ez0D3i1SDPzGDMDbUmcIcHDnMxTcThBXjoHvy2SzTzFzDIGt5Nn8KQZG86oOG64QTqn4ECCAW6/SBxv38z4puK27Qz2w48lPhhUy8vPTt/84UOFnRwuLRBggMw+gC5CEMg3kKJ6FIyCUTAKRgIAAIM2WREdzxOrAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-9120-9939","institution":"Central European Institute of Technology, Masaryk University (CEITEC MU)","correspondingAuthor":true,"prefix":"","firstName":"Klara","middleName":"","lastName":"Mareckova","suffix":""},{"id":627119075,"identity":"4541c6c9-e8e8-4981-b91c-6e508952bca3","order_by":1,"name":"Lukas Pelant","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lukas","middleName":"","lastName":"Pelant","suffix":""},{"id":627119076,"identity":"8c254423-dd24-463f-b0df-4c44e11335a7","order_by":2,"name":"Radek Marecek","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Radek","middleName":"","lastName":"Marecek","suffix":""},{"id":627119077,"identity":"6a5e4e57-e307-45d6-8d26-2a42e398e056","order_by":3,"name":"Anna Pacinkova","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Pacinkova","suffix":""},{"id":627119078,"identity":"b57f74c7-e94f-401c-a301-760e12892499","order_by":4,"name":"Jana Klánová","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jana","middleName":"","lastName":"Klánová","suffix":""},{"id":627119079,"identity":"1ef3aeb4-a5bd-4792-a3c0-3005bdaf0bc2","order_by":5,"name":"Yuliya Nikolova","email":"","orcid":"https://orcid.org/0000-0001-5144-3723","institution":"Centre for Addiction and Mental Health","correspondingAuthor":false,"prefix":"","firstName":"Yuliya","middleName":"","lastName":"Nikolova","suffix":""}],"badges":[],"createdAt":"2026-04-13 08:00:55","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9400608/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9400608/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108389610,"identity":"e92fe4ca-a1fc-41ed-b190-419a9c2f3a14","added_by":"auto","created_at":"2026-05-04 06:50:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":11638785,"visible":true,"origin":"","legend":"\u003cp\u003ePremature epigenetic aging estimated using Horvath’s DNA methylation clock was associated with greater negative deviations from normative development of the putamen (left – A; right – B) and caudate (left – C; right – D), indicating smaller volumes than expected based on the normative model.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/4389e4e47e5c0a3e083e4a83.png"},{"id":108493308,"identity":"cc9976c8-057e-455d-a70d-9421857d3355","added_by":"auto","created_at":"2026-05-05 09:59:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":10766062,"visible":true,"origin":"","legend":"\u003cp\u003eWomen showed linear declines, whereas men showed quadratic (U-shaped) associations between Horvath’s EpiAGE and cognitive performance as indicated by FSIQ (men – A; women – B) and PIQ (men – C; women – D).\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/34acc786cc34c3d372477fa1.png"},{"id":108389613,"identity":"58333df6-8d96-456f-944e-48c3284f2d8b","added_by":"auto","created_at":"2026-05-04 06:50:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":12095984,"visible":true,"origin":"","legend":"\u003cp\u003eIn women, greater negative deviations from normative development of the putamen (indicating smaller-than-expected putamen) were associated with worse cognitive performance, as indicated by lower FSIQ (left putamen – A; right putamen – B) and lower PIQ (left putamen – C; right putamen – D).\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/becdaa5a8a57ef2b19439def.png"},{"id":108389615,"identity":"e1cb69d3-8e54-4349-9e98-1921f8979440","added_by":"auto","created_at":"2026-05-04 06:50:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":351742,"visible":true,"origin":"","legend":"\u003cp\u003eIn women, deviations from normative development of the putamen mediated the relationship between premature epigenetic aging estimated using the Horvath clock and worse cognitive performance indicated by lower FSIQ (left putamen – A; right putamen – B) and lower PIQ (left putamen – C; right putamen – D).\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/57464f1b0fd308c09b88bdc3.png"},{"id":108494628,"identity":"b6630ff1-bdfd-4a97-9851-9c08168e9667","added_by":"auto","created_at":"2026-05-05 10:06:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":37345651,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/f4ad5d84-231f-4867-a367-aff2028ba821.pdf"},{"id":108492397,"identity":"500b765c-7c52-41b7-81d7-05601a59a5ab","added_by":"auto","created_at":"2026-05-05 09:57:40","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":52383,"visible":true,"origin":"","legend":"Suplementary Table 2","description":"","filename":"SupplementaryTable2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/af3970cb125d18322d86016c.docx"},{"id":108492900,"identity":"79657c68-92ce-49e7-b950-514aa101ed07","added_by":"auto","created_at":"2026-05-05 09:58:56","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":29534,"visible":true,"origin":"","legend":"Supplementary Table 1","description":"","filename":"SupplementaryTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9400608/v1/2a2a7310328a209994530cac.docx"}],"financialInterests":"The authors have declared there is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose","formattedTitle":"Premature Epigenetic Aging, Deviations from Normative Brain Development, and Cognitive Performance in Young Adulthood","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe global population is undergoing profound demographic change. Although some regions continue to experience rapid population growth, population aging represents the dominant global trend. Declining fertility rates, alongside reduced mortality, have led to a steadily increasing proportion of older individuals worldwide (1, 2). In 2020, the global older-age dependency ratio (ODR) was 17, corresponding to 17 adults aged 65 years and older per 100 working-age individuals, and projections indicate an increase to 21 by 2030 and 26 by 2040 (2). Given the increasing percentage of adults over 65, there is also a higher prevalence of dementia as well as other health-related issues (3, 4, 5, 6)\u003c/p\u003e \u003cp\u003eA growing body of evidence indicates that many age-related diseases originate in young adulthood or even early life (7, 8, 9), underscoring the need for a lifespan perspective on aging-related pathology. Identifying early biological markers and characterizing longitudinal neurodevelopmental trajectories are critical for advancing prevention strategies. Since chronological age alone fails to capture interindividual variability in biological aging, epigenetic clocks and normative models of brain development, allowing us to quantify deviations from age- and sex-specific development, have emerged as more informative and reliable estimates of biological aging.\u003c/p\u003e \u003cp\u003eEpigenetic clocks have been increasingly recognized as predictors of health, cognition, and mortality (10, 11). Epigenetic clocks differ substantially in their biological targets, tissue specificity, and intended phenotypic scope. The first-generation epigenetic clocks (e.g., Horvath\u0026rsquo;s DNA methylation clock, Hannum\u0026rsquo;s DNA methylation clock) were designed primarily to predict chronological age based on DNA methylation patterns (11, 12). For example, the Horvath\u0026rsquo;s clock, based on DNA methylation profiles of 353 CpG sites, accurately predicts the age of different tissues, including the brain tissue (11). Horvath also demonstrated that the heritability of age acceleration is 100% in the newborns, but only 39% in older subjects, suggesting non-genetic factors become more relevant in later life (11). The second-generation epigenetic clocks (e.g., GrimAge (13), AltumAge (14), CheekAge (15)) have been explicitly optimized to predict health-related outcomes beyond chronological age, including disease risk, functional decline, and mortality. These clocks incorporate a substantially larger number of CpG sites and leverage machine learning approaches to capture methylation patterns associated with metabolic status, inflammation, lifestyle factors, and cumulative disease burden (14, 15). For example, the AltumAge, a deep-learning\u0026ndash;based pan-tissue clock, utilizes more than 20,000 CpG sites and predicts chronological age and mortality (14). The CheekAge, developed specifically from buccal epithelial tissue, utilizes approximately 200,000 CpG sites to reflect lifestyle and health and shows strong associations with BMI and substance use (15).\u003c/p\u003e \u003cp\u003eEpigenetic aging captures processes relevant to neurodevelopment (11, 15, 16), metabolic load (15, 17), and cellular aging (11, 18, 19). Research from our group showed that premature epigenetic aging, calculated using the Horvath\u0026rsquo;s clock, is associated with worse cognitive skills in women already in young adulthood (8) and that premature epigenetic aging relates to altered functional brain dynamics and cognition primarily in women (20). Previous structural neuroimaging research has consistently demonstrated that individual differences in cognitive performance and intelligence correlate with variations in regional brain macrostructure, including cortical thickness and subcortical volumes (21, 22, 23). However, the underlying brain structure reflecting the altered brain dynamics and cognitive performance is not well understood.\u003c/p\u003e \u003cp\u003eNormative modeling of brain development, which allows quantification of individual-level deviations from age- and sex-specific normative brain development, might offer a unique insight into the relationships between altered epigenetic aging and altered cognitive performance. Large-scale normative models of brain development, such as the CentileBrain (24), trained on 37 407 MRI scans from individuals aged 3\u0026ndash;90 years, allow us to characterize these structural processes with precision. Therefore, the current study aims to combine the estimates of premature epigenetic aging with deviations from normative brain development and cognitive performance and offer a unique analytical framework allowing for the detection of early signs of aging-related pathology.\u003c/p\u003e \u003cp\u003eSpecifically, we aim to test whether the relationship between premature epigenetic aging and worse cognitive outcomes in young adulthood might be mediated by deviations from normative development of brain structure. Further, we aim to determine whether these relationships between premature epigenetic aging, deviations from normative brain development, and cognitive outcomes might be sex-specific.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cp\u003eParticipants\u003c/p\u003e\n\u003cp\u003eA total of 254 young adults (51.6% male) aged 28\u0026ndash;30 years (M=28.97, SD=0.68) from the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) prenatal birth cohort (25) participated also in its neuroimaging \u0026ndash; epigenetic follow-up in young adulthood, the Health Brain Age (HBA) study, and had complete data on brain structure, DNA methylation, and cognition. Demographic information regarding these participants is available in \u003cstrong\u003eSupplementary Table 1\u003c/strong\u003e. For AltumAge, two extreme outliers, defined a priori as absolute z-scores \u0026gt; 3, were excluded before descriptive statistics and sex-comparison analyses (independent-samples t-test). All other analyses were conducted on the full available sample. All participants provided written informed consent to participate in the HBA study, including the agreement to merge data from HBA with their historical data from ELSPAC. Ethical approval for the HBA study was obtained from the ELSPAC ethics committee (ELSPAC/EK/2/2020), and all methods were performed in accordance with the relevant guidelines and regulations.\u003c/p\u003e\n\u003cp\u003eAssessment of DNA Methylation and Epigenetic Aging\u003c/p\u003e\n\u003cp\u003eDNA was isolated from buccal swabs collected in the late 20s, and DNA methylation was assessed using the Illumina EPIC Platform. Epigenetic aging was estimated using three different epigenetic clocks: Horvath\u0026rsquo;s clock (11), which we previously applied in Mareckova et al. (8, 25, 27), as well as the AltumAge clock (14) and the CheekAge clock (15), both of which were implemented and described in detail in Mareckova et al. (27), including the full analysis code.\u003c/p\u003e\n\u003cp\u003eAs described in our previous work (8, 26, 27), we used the R package ChAMP (28) to process the raw Illumina microarray data, and the beta mixture quantile normalization (BMIQ) (29) method to adjust the beta-values of type II design probes into a statistical distribution characteristic of type I probes. DNA methylation age was calculated using Horvath\u0026rsquo;s clock (11), based on 353 CpG sites, using the AltumAge clock (14), based on 20 318 CpG sites, and using the CheekAge clock (15), based on 115 533 CpG sites. See Supplementary Methods inMareckova et al. (27)for details and the link to the GitHub Repository. For each of these clocks, the epigenetic age gap estimate (EpiAGE) was calculated as the DNA methylation age residualized for batch, chronological age, and the proportion of epithelial cells. Positive values of EpiAGE reflect accelerated aging, and negative values decelerated aging.\u003c/p\u003e\n\u003cp\u003eStructural Magnetic Resonance Imaging\u003c/p\u003e\n\u003cp\u003eStructural magnetic resonance imaging (MRI) of the brain was performed on a 3T Siemens Prisma MRI scanner with a 64-channel head coil. T1-weighted (T1w) whole-brain MPRAGE images were acquired using a 64-channel head/neck coil with acquisition parameters: voxel size 1 mm\u003csup\u003e3\u003c/sup\u003e, repetition time (TR) 2,300 ms, echo time (TE) 2.34 ms, inversion time (TI) 900 ms, flip angle 8\u0026deg;.\u003c/p\u003e\n\u003cp\u003eMRI data were processed in FreeSurfer v7.1.1, and cortical thickness and surface area in 68 regions from the Desikan-Killiany (DK) atlas, and volumes of 14 subcortical regions from the Aseg atlas were calculated. For quality control, all the registration and segmentation outputs from FreeSurfer were visually checked and passed the quality control.\u003c/p\u003e\n\u003cp\u003eNormative Modeling and Calculation of the Deviations from Normative Brain Development\u003c/p\u003e\n\u003cp\u003ePretrained models from the CentileBrain Initiative (24) were employed to compute the age- and sex- specific normative deviation (z) scores for all measures (68 surface area regions, 68 cortical thickness regions, and 14 subcortical regions). These models were trained on 37 407 MRI scans from individuals aged 3\u0026ndash;90 years (24). The resulting z-scores reflect whether each measure is larger or smaller than expected for an individual of the same age and sex. No further adjustment for intracranial volume (ICV) was performed, as correction is inherently part of the CentileBrain model calculation.\u003c/p\u003e\n\u003cp\u003eAssessment of Cognitive Performance\u003c/p\u003e\n\u003cp\u003eCognition was assessed using the following subtests from the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV) (30): Picture completion, Matrix Reasoning, Arithmetic, Digit span, Information, Similarities, and Digit-symbol coding. These subtests allowed us to assess participants\u0026lsquo; perceptual reasoning (Picture completion, Matrix reasoning), working memory (Arithmetic, Digit span), verbal comprehension (Information, Similarities), and processing speed (Digit-symbol coding). Full-scale IQ was calculated based on the scores for these 7 WAIS-IV subtests according to Tam (31). Performance IQ (PIQ) was estimated from subtests assessing perceptual reasoning (Picture Completion, Matrix Reasoning) and processing speed (Digit-Symbol Coding). Verbal IQ (VIQ) was derived from subtests evaluating verbal comprehension (Information, Similarities) and working memory (Arithmetic, Digit Span).\u003c/p\u003e\n\u003cp\u003eStatistical Analyses\u003c/p\u003e\n\u003cp\u003eAssociations between EpiAGE, deviations from normative brain development, and cognitive performance were evaluated using linear regressions, and potential interactions with sex were considered. We also evaluated potential non-linear (quadratic) relationships between epigenetic aging and cognitive performance to capture potential complex dynamics not revealed by linear models. Multiple comparisons were corrected using the False Discovery Rate (FDR) method within each family of analyses. Finally, a moderated mediation analysis assessed whether the relationship between premature epigenetic aging and worse cognitive performance might be mediated by the epigenetic aging-related changes in normative brain development and whether these relationships differ by sex. The indirect effect was assessed using Model 14 of the PROCESS macro (32) for SPSS, version 5.0 (IBM SPSS Statistics, IBM Corp, Armonk, NY). Conditional indirect effects were obtained for significance using bootstrapped (10 000 iterations) bias-corrected 95% confidence intervals constructed around indirect effect estimates.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cem\u003ePremature epigenetic aging and its relationship with normative brain development\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePremature epigenetic aging calculated using the Horvath\u0026rsquo;s clock was associated with greater negative deviations from age- and sex-specific normative development of putamen (left: R\u003csup\u003e2\u003c/sup\u003e=0.02, FDRp=0.02, \u003cstrong\u003eFigure 1A\u003c/strong\u003e; right: R\u003csup\u003e2\u003c/sup\u003e=0.02, FDRp=0.03, \u003cstrong\u003eFigure 1B\u003c/strong\u003e) and caudate (left: R\u003csup\u003e2\u003c/sup\u003e=0.03, FDRp=0.01, \u003cstrong\u003eFigure 1C\u003c/strong\u003e); right: R\u003csup\u003e2\u003c/sup\u003e=0.04, FDRp=0.01, \u003cstrong\u003eFigure 1D\u003c/strong\u003e; see statistical details in \u003cstrong\u003eTable 1\u003c/strong\u003e), but no deviations from normative development of the other subcortical volumes (p\u0026gt;0.14, FDRp\u0026gt;0.38), cortical thickness in the 68 regions (p\u0026gt;0.05; FDRp\u0026gt;0.91), or surface area in the 68 regions (p\u0026gt;0.07, FDRp\u0026gt;0.94). No similar results were found using the two epigenetic clocks developed to reflect not only aging as such but also disease and mortality: CheekAge (cortical thickness in the 68 regions p\u0026gt;0.02, FDRp\u0026gt;0.49; surface area in the 68 regions p\u0026gt;0.04, FDRp\u0026gt;0.83; volume of the 14 subcortical regions p\u0026gt;0.07, FDRp\u0026gt;0.67) and AltumAGE (cortical thickness in the 68 regions p\u0026gt;0.003, FDRp\u0026gt;0.18; surface area in the 68 regions p\u0026gt;0.009, FDRp\u0026gt;0.54; volume of the 14 subcortical regions p\u0026gt;0.10, FDRp\u0026gt;0.63). All statistical details are provided in \u003cstrong\u003eSupplementary Table 2\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1 -\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ePremature epigenetic aging and deviations from normative development of caudate and putamen.\u003c/strong\u003e Standardized regression coefficients, nominal p values, and FDR-corrected p values are shown for associations of Horvath\u0026rsquo;s clock, CheekAge, and AltumAge. Significant associations after FDR correction are shown in bold.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"654\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 26.2997%;\" colspan=\"4\"\u003e\u003cstrong\u003eHorvath\u0026apos;s clock\u003c/strong\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 26.4526%;\" colspan=\"4\"\u003e\u003cstrong\u003eCheekAge\u003c/strong\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 24.9235%;\" colspan=\"4\"\u003e\u003cstrong\u003eAltumAge\u003c/strong\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\n \u003cp\u003eHem\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 85px;\"\u003e\n \u003cp\u003eRegion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003eStd Beta\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eFDR p\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003eStd Beta\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eFDR p\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003eStd Beta\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 55px;\"\u003e\n \u003cp\u003eFDR p\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ecaudate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.198\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.012\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.065\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.299\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.697\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.079\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.212\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 55px;\"\u003e\n \u003cp\u003e0.631\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ecaudate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.212\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.011\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.074\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.240\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.671\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.103\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.102\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 55px;\"\u003e\n \u003cp\u003e0.631\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eL\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eputamen\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.182\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.017\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.230\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.671\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.062\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.319\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 55px;\"\u003e\n \u003cp\u003e0.631\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 45px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eputamen\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e-0.159\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.044\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.101\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.109\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.671\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003e-0.081\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.197\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 55px;\"\u003e\n \u003cp\u003e0.631\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 6px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003ePremature epigenetic aging and cognitive performance\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSex moderated the relationship between premature epigenetic aging calculated using Horvath\u0026rsquo;s clock and cognitive performance (interaction statistics for FSIQ: \u0026beta;=\u0026minus;0.192, p=0.029; interaction statistics for PIQ: \u0026beta;=\u0026minus;0.183, p=0.034). Post-hoc sex-stratified linear regressions indicated that premature epigenetic aging (Horvath\u0026rsquo;s clock) was associated with lower cognitive performance in women but not in men (FSIQ: men \u0026beta;=0.076, p=0.390; \u003cstrong\u003eFigure 2A\u003c/strong\u003e; women \u0026beta;=\u0026minus;0.200, R\u0026sup2;=0.06, p=0.027; \u003cstrong\u003eFigure 2B\u003c/strong\u003e; PIQ: men \u0026beta;=0.011, p=0.905; \u003cstrong\u003eFigure 2C\u003c/strong\u003e; women \u0026beta;=\u0026minus;0.257, R\u0026sup2;=0.08, p=0.004; \u003cstrong\u003eFigure 2D\u003c/strong\u003e). However, quadratic models revealed significant U-shaped relationships between premature epigenetic aging and cognitive performance in men (FSIQ: \u0026beta;=0.211, p=0.016, R\u0026sup2;=0.05; \u003cstrong\u003eFigure 2A\u003c/strong\u003e; PIQ: \u0026beta;=0.273, p=0.002, R\u0026sup2;=0.07; \u003cstrong\u003eFigure 2C\u003c/strong\u003e) but not in women (FSIQ: \u0026beta;=0.130, p=0.146; \u003cstrong\u003eFigure 2B\u003c/strong\u003e; PIQ: \u0026beta;=0.127, p=0.148; \u003cstrong\u003eFigure 2D\u003c/strong\u003e). In women, the linear term remained significant in the quadratic models (FSIQ: \u0026beta;=\u0026minus;0.204, p=0.023; PIQ: \u0026beta;=\u0026minus;0.261, p=0.003), consistent with a predominantly linear decline.\u003c/p\u003e\n\u003cp\u003eNo comparable associations were observed for CheekAGE or AltumAGE in analogous models (all p \u0026gt; 0.07). Neither of these clocks also showed the male-specific quadratic curvature seen for Horvath\u0026rsquo;s clock (all quadratic-term p\u0026rsquo;s \u0026ge; .057).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNormative brain development and cognitive performance\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFurther analyses revealed women-specific relationships between greater negative deviations from normative development of putamen and lower full-scale IQ (left: sex \u0026times; putamen interaction t = 2.62, p = 0.009; women: \u0026beta; = 0.28, p = 0.002, R\u0026sup2; = 0.08; men: \u0026beta; = \u0026minus;0.03, p = 0.702, \u003cstrong\u003eFigure 3A\u003c/strong\u003e; right: sex \u0026times; putamen interaction t = 2.30, p = 0.022; women: \u0026beta; = 0.23, p = 0.012, R\u0026sup2; = 0.05; men: \u0026beta; = \u0026minus;0.06, p = 0.531, \u003cstrong\u003eFigure 3B\u003c/strong\u003e, and particularly lower performance IQ (left: sex \u0026times; putamen interaction t = 2.51, p = 0.013; women: \u0026beta; = 0.28, p = 0.002, R\u0026sup2; = 0.08; men: \u0026beta; = \u0026minus;0.02, p = 0.835, \u003cstrong\u003eFigure 3C\u003c/strong\u003e; right: sex \u0026times; putamen interaction t = 2.16, p = 0.031; women: \u0026beta; = 0.23, p = 0.012, R\u0026sup2; = 0.05; men: \u0026beta; = \u0026minus;0.04, p = 0.652, \u003cstrong\u003eFigure 3D\u003c/strong\u003e). Quadratic terms were not significant in any sex-stratified post-hoc models (all p \u0026ge; 0.171).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDo deviations from normative brain development mediate the relationship between premature epigenetic aging and worse cognitive performance?\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eModerated mediation analyses indicated that the association between Horvath\u0026rsquo;s EpiAGE and cognitive performance was mediated by normative development of the putamen in women but not men. For full-scale IQ, the indirect effects were significant in women (left: ab = \u0026minus;0.186, 95% CI [\u0026minus;0.457, \u0026minus;0.036]; right: ab = \u0026minus;0.134, 95% CI [\u0026minus;0.368, \u0026minus;0.016]) but not in men (left: ab = 0.026, 95% CI [\u0026minus;0.075, 0.211]; right: ab = 0.035, 95% CI [\u0026minus;0.048, 0.202]). The index of moderated mediation was significant for both the left (index = \u0026minus;0.212, 95% CI [\u0026minus;0.567, \u0026minus;0.026]) and right putamen (index = \u0026minus;0.169, 95% CI [\u0026minus;0.483, \u0026minus;0.015]). Similarly, for performance IQ, the indirect effects were significant in women (left: ab = \u0026minus;0.218, 95% CI [\u0026minus;0.523, \u0026minus;0.040]; right: ab = \u0026minus;0.152, 95% CI [\u0026minus;0.411, \u0026minus;0.014]) but not in men (left: ab = 0.030, 95% CI [\u0026minus;0.086, 0.246]; right: ab = 0.037, 95% CI [\u0026minus;0.065, 0.223]). The index of moderated mediation was significant for both the left (index = \u0026minus;0.248, 95% CI [\u0026minus;0.667, \u0026minus;0.030]) and right putamen (index = \u0026minus;0.189, 95% CI [\u0026minus;0.545, \u0026minus;0.009]). See \u003cstrong\u003eFigure 4\u003c/strong\u003e for statistical details.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eWe performed a neuroimaging\u0026ndash;epigenetic follow-up of a prenatal birth cohort and showed that premature epigenetic aging in young adulthood, calculated using Horvath\u0026rsquo;s epigenetic clock, is associated with altered normative development of the dorsal striatum and that in women, the deviations from normative development of the putamen mediate the relationship between premature epigenetic aging and worse cognition. These findings substantially extend our previous research in the same cohort regarding premature epigenetic aging, brain dynamics, and cognition, and reveal the role of altered structural brain development and specifically the importance of the dorsal striatum in these relationships. Moreover, given that dopaminergic neurons in the substantia nigra release dopamine to the dorsal striatum, our findings align with previous PET research on age-related dopaminergic decline in the dorsal striatum (33, 34), as well as further research on age-related dopamine loss and cognitive deficits (10).\u003c/p\u003e \u003cp\u003eThe dorsal striatum, consisting of the caudate and putamen, is particularly relevant due to its involvement in dopaminergic circuits that support motor learning, cognitive speed, working memory, and reinforcement learning (35, 36). These domains exhibit early decline in dopaminergic aging models (10, 33, 34). The specifically larger associations between putamen and performance IQ, as compared to verbal IQ, described in the current study, are consistent with further evidence linking basal ganglia volume with figural and numeric, but not verbal, aspects of intelligence (23).\u003c/p\u003e \u003cp\u003eWhile premature epigenetic aging estimated using Horvath\u0026rsquo;s clock was associated with deviations in both caudate and putamen, the mediatory effect on cognition was specific to the putamen. This may reflect the putamen\u0026rsquo;s prominent role in sensorimotor circuits and habitual control, which are closely linked to processing speed components of Performance IQ (35). In contrast, the caudate is deeply integrated into frontoparietal executive loops (37, 38), which might be more resilient or functionally compensated for in this healthy young cohort.\u003c/p\u003e \u003cp\u003eA striking finding of our study is that the relationship between epigenetic aging and cognition differs by sex, not only in magnitude but also functionally. While women exhibited a direct, predominantly linear \u0026ldquo;dose\u0026ndash;response\u0026rdquo; pattern, in which premature epigenetic aging was consistently associated with more negative deviations from normative striatal development and lower cognitive performance, men displayed a significant quadratic (U-shaped) relationship. This non-monotonic profile indicates that higher cognitive scores were observed at both the lower and upper extremes of the epigenetic aging spectrum, whereas intermediate EpiAGE values corresponded to lower performance. Although we confirmed that these results were not driven by statistical outliers (all standardized residuals |z| \u0026lt; 3; Cook\u0026rsquo;s distance\u0026thinsp;\u0026lt;\u0026thinsp;1), the pattern suggests that, in young men, the response to epigenetic variance may be more complex than a simple linear decline. One plausible (but speculative) interpretation is the presence of sex-specific non-linear compensatory mechanisms. While classic biphasic or hormetic stress responses typically involve low-dose adaptation followed by high-dose toxicity or functional decline (39), the U-shaped pattern in young men implies a different dynamic. It suggests that more advanced epigenetic aging might act as a systemic challenge that triggers robust compensatory neural processes. Importantly, cognitive reserve is not merely a passive buffer against late-life pathology, but a dynamic capacity that can be actively mobilized to cope with demanding conditions already in young adulthood (40, 41). At the molecular level, this aligns with emerging neuroepigenetic evidence demonstrating that male and female brains engage distinct chromatin and transcriptional networks in response to systemic or biological challenges, providing a fundamental mechanism for sex-specific adaptation and resilience (42).\u003c/p\u003e \u003cp\u003eThe EpiAGE\u0026ndash;putamen\u0026ndash;cognition mediation pathway observed in women replicates and extends previous sex-specific fMRI findings from the same cohort (20) and points to a specific vulnerability of the female striatal system to premature epigenetic aging. Rather than being an isolated phenomenon, this susceptibility aligns with distinct sex differences observed in the structural maturation of subcortical volumes, including the putamen, throughout adolescence and early adulthood (44). Furthermore, this vulnerability may be rooted in the profound sensitivity of female striatal and dopaminergic circuits to the interplay between ovarian hormones and epigenetic programming (42, 45). Consequently, premature epigenetic aging may disproportionately disrupt these sensitive structural-functional couplings in young women. In contrast, the U-shaped curve in men implies a more complex interaction. It suggests that in young men, the impact of epigenetic deviations is not monotonic; moderate deviations might be buffered by compensatory neural mechanisms or higher physiological resilience thresholds that are not present in women (10). This aligns with the hypothesis of differential reserve capacities in early aging, where men may maintain cognitive function despite structural loads up to a certain tipping point, whereas women show earlier structural-functional coupling (10, 23).\u003c/p\u003e \u003cp\u003eFurthermore, the absence of a significant mediation effect via the putamen in men reinforces that their cognitive variance related to aging is likely driven by different neural substrates or non-linear network reconfigurations that simple linear mediation models cannot capture. Potential explanations for this sex specificity include innate sex differences in striatal dopamine D2 receptor density and affinity (46), estrogen-mediated neuroprotection of the nigrostriatal dopamine system (47), or a differential sensitivity of the female brain to cumulative metabolic stress and environmental insults (48).\u003c/p\u003e \u003cp\u003eOur analyses revealed a distinct sensitivity of Horvath\u0026rsquo;s clock to striatal development and cognitive performance, while other health-related clocks (CheekAge, AltumAge) failed to show significant associations. This divergence is likely attributable to the specific biological properties of the Horvath clock in the context of young adulthood. Horvath\u0026rsquo;s clock is a pan-tissue estimator tightly correlated with chronological age and cellular developmental processes, making it particularly suitable for detecting deviations from normative trajectories (11, 49). Notably, the male-specific non-linear associations observed between Horvath\u0026rsquo;s clock and cognitive performance were not reproduced by CheekAge or AltumAge, supporting clock-specific sensitivity to the processes captured in young adulthood. The second-generation or tissue-specific clocks are often optimized to predict mortality, environmental exposures, or physiological dysregulation associated with advanced aging and disease burden (50, 51). Although CheekAge was developed specifically from buccal epithelial tissue and was thus expected to show high sensitivity in our samples (15), it failed to show significant associations with brain structure or cognition. Therefore, the observed associations suggest that the variance in striatal volume and fluid intelligence is driven by an intrinsic biological aging rate captured by the Horvath clock\u0026rsquo;s rather than by external health deficits or cumulative physiological dysregulation (52). Together, these findings suggest that Horvath\u0026rsquo;s clock captures variance related to neurodevelopmental timing rather than cumulative health burden, which may explain its unique sensitivity to brain development and cognition in young adulthood.\u003c/p\u003e \u003cp\u003eThe main strengths of our study are the unique combination of neuroimaging and epigenetic data on a prenatal birth cohort, the use of three different epigenetic clocks as well as a robust large scale model of normative brain development, trained on 37 407 MRI scans from individuals aged 3\u0026ndash;90 years old, and the even distribution of men and women allowing us to assess the potential interactions with sex. We bring novel evidence showing that premature epigenetic aging in young adulthood is distinctly associated with deviations from normative development in parts of the basal ganglia, and that these structural alterations mediate the relationship between accelerated epigenetic aging and worse cognitive performance. Still, our findings are limited by the cross-sectional nature of the current study, which precludes causal inference. While we modeled deviations in brain structure as a mediator between epigenetic aging and cognition, it is plausible that the relationship is bidirectional; for example, lifestyle factors associated with higher intelligence could influence both brain integrity and epigenetic aging rates. Future research should follow these participants longitudinally to reveal whether the premature epigenetic aging and the associated changes in normative brain development and cognition might develop into deteriorating brain health and more severe cognitive outcomes in older adulthood. Moreover, since the quadratic effects can be sensitive to model specification, future work should test robustness using alternative non-linear parametrizations (e.g., splines) and, ideally, longitudinal data to determine whether the observed U-shaped pattern in men reflects developmental dynamics or subgroup heterogeneity. In addition, since our DNA methylation data were based on buccal swab samples, we were not able to calculate other second-generation epigenetic clocks, such as the PhenoAge (50), and assess their relationship with normative brain development and cognitive outcomes. Therefore, future research should aim to collect blood from the participants and extend our findings with these additional analyses.\u003c/p\u003e \u003cp\u003eOverall, this study provides novel evidence linking premature epigenetic aging to deviations from normative development of the dorsal striatum in young adulthood and worse cognitive performance in women. These findings indicate sex-specific neurodevelopmental pathways linking epigenetic aging markers to cognition and underscore the value of normative modeling in uncovering subtle deviations in brain maturation. Moreover, our study reveals that premature epigenetic aging in young adulthood and the associated deviations in the development of the dorsal striatum might be an early marker of worse cognitive performance long before clinical symptoms may develop. Our findings also suggest that the dorsal striatum and its function, including cognitive flexibility, goal-directed behavior, planning, working memory, attention, and learning, might be potential targets to prevent premature epigenetic aging and cognitive decline in women.\u003c/p\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eAcknowledgemnts\u003c/h2\u003e \u003cp\u003eThis work has received funding from the Czech Science Foundation, project no. 24-12183M, Czech Health Research Council (No. NU20J-04-00022), the Czech Ministry of Education, Youth and Sports (MEYS CR) (CEITEC 2020, LQ1601), and by project no. LX22NPO5107 (MEYS): Funded by European Union \u0026ndash; Next Generation EU. Support with obtaining scientific data presented in this paper came from the core facility Multimodal and Functional Imaging Laboratory of Central European Institute of Technology, Masaryk University, supported by the Czech-BioImaging large RI project (No. LM2018129, funded by MEYS CR). Authors also thank the RECETOX Research Infrastructure (No LM2023069), financed by MEYS CR for its supportive background. This work was also supported by the European Union\u0026rsquo;s Horizon 2020 research and innovation program under grant agreement No 857560 (CETOCOEN Excellence). This publication reflects only the author's view, and the European Commission is not responsible for any use that may be made of the information it contains. Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. Dr. Nikolova is supported by Koerner New Scientist Award from the CAMH Foundation and a Discovery Grant from the National Sciences and Engineering Research Council of Canada (NSERC).\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgemnts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work has received funding from the Czech Science Foundation, project no. 24-12183M, Czech Health Research Council (No. NU20J-04-00022), the Czech Ministry of Education, Youth and Sports (MEYS CR) (CEITEC 2020, LQ1601), and \u003cem\u003eby project no. LX22NPO5107 (MEYS): Funded by European Union \u0026ndash; Next Generation EU\u003c/em\u003e. Support with obtaining scientific data presented in this paper came from the core facility Multimodal and Functional Imaging Laboratory of Central European Institute of Technology, Masaryk University, supported by the Czech-BioImaging large RI project (No. LM2018129, funded by MEYS CR). Authors also thank the RECETOX Research Infrastructure (No LM2023069), financed by MEYS CR for its supportive background. This work was also supported by the European Union\u0026rsquo;s Horizon 2020 research and innovation program under grant agreement No 857560 (CETOCOEN Excellence). This publication reflects only the author\u0026apos;s view, and the European Commission is not responsible for any use that may be made of the information it contains. Computational resources were provided by the e-INFRA CZ project (ID:90254), supported by the Ministry of Education, Youth and Sports of the Czech Republic. Dr. Nikolova is supported by Koerner New Scientist Award from the CAMH Foundation and a Discovery Grant from the National Sciences and Engineering Research Council of Canada (NSERC).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;All authors declare no financial or any other conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eHe W, Goodkind D, Kowal P. \u003cem\u003eAn Aging World: 2015\u003c/em\u003e. International Population Reports P95/16-1. U.S. Census Bureau; 2016.\u003c/li\u003e\n\u003cli\u003eU.S. Census Bureau. \u003cem\u003eStory Map - An Aging World: 2020\u003c/em\u003e. U.S. Census Bureau. Published 2020. Accessed April 12, 2026.\u003c/li\u003e\n\u003cli\u003eChang AY, Skirbekk VF, Tyrovolas S, Kassebaum NJ, Dieleman JL. Measuring population ageing: an analysis of the Global Burden of Disease Study 2017. \u003cem\u003eLancet Public Health\u003c/em\u003e. 2019;4:e159-e167.\u003c/li\u003e\n\u003cli\u003eChen C et al. A global analysis of adaptation to societal aging across low-, middle- and high-income countries using the Global Aging Society Index. \u003cem\u003eNat Aging\u003c/em\u003e. 2025;5:113-121.\u003c/li\u003e\n\u003cli\u003eFerrari AJ et al. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. \u003cem\u003eLancet.\u003c/em\u003e 2024;403:2133-2161.\u003c/li\u003e\n\u003cli\u003eGBD 2019 Ageing Collaborators. Global, regional, and national burden of diseases and injuries for adults 70 years and older: systematic analysis for the Global Burden of Disease 2019 Study. \u003cem\u003eBMJ\u003c/em\u003e. 2022;376:e068208.\u003c/li\u003e\n\u003cli\u003eBateman RJ et al. Clinical and biomarker changes in dominantly inherited Alzheimer\u0026apos;s disease. \u003cem\u003eN Engl J Med\u003c/em\u003e. 2012;367:795-804.\u003c/li\u003e\n\u003cli\u003eMareckova K et al. Longitudinal study of epigenetic aging and its relationship with brain aging and cognitive skills in young adulthood. \u003cem\u003eFront Aging Neurosci\u003c/em\u003e. 2023;15:1215957.\u003c/li\u003e\n\u003cli\u003eMoffitt TE, Belsky DW, Danese A, Poulton R, Caspi A. The longitudinal study of aging in human young adults: knowledge gaps and research agenda\u003cem\u003e. J Gerontol A Biol Sci Med Sci.\u003c/em\u003e 2017;72:210-215.\u003c/li\u003e\n\u003cli\u003eBackman L, Nyberg L, Lindenberger U, Li SC, Farde L. The correlative triad among aging, dopamine, and cognition: current status and future prospects. \u003cem\u003eNeurosci Biobehav Rev\u003c/em\u003e. 2006;30:791-807.\u003c/li\u003e\n\u003cli\u003eHorvath S. DNA methylation age of human tissues and cell types. \u003cem\u003eGenome Biol\u003c/em\u003e. 2013;14:R115.\u003c/li\u003e\n\u003cli\u003eHannum G et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. \u003cem\u003eMol Cell\u003c/em\u003e. 2013;49:359-367.\u003c/li\u003e\n\u003cli\u003eLu AT et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019;11:303-327.\u003c/li\u003e\n\u003cli\u003ede Lima Camillo LP, Lapierre LR, Singh R. A pan-tissue DNA-methylation epigenetic clock based on deep learning. \u003cem\u003enpj Aging\u003c/em\u003e. 2022;8:4.\u003c/li\u003e\n\u003cli\u003eShokhirev MN et al. CheekAge: a next-generation buccal epigenetic aging clock associated with lifestyle and health. \u003cem\u003eGeroScience\u003c/em\u003e. 2024;46:3429-3443.\u003c/li\u003e\n\u003cli\u003eSanli E, Kabaran S. Maternal obesity, maternal overnutrition and fetal programming: effects of epigenetic mechanisms on the development of metabolic disorders. \u003cem\u003eCurr Genomics\u003c/em\u003e. 2019;20:419-427.\u003c/li\u003e\n\u003cli\u003eVidal CM et al. Insulin resistance in women correlates with chromatin histone lysine acetylation, inflammatory signaling, and accelerated aging. \u003cem\u003eCancers (Basel).\u003c/em\u003e 2024;16:2735.\u003c/li\u003e\n\u003cli\u003ePereira B et al. Epigenetic reprogramming as a key to reverse ageing and increase longevity. \u003cem\u003eAgeing Res Rev\u003c/em\u003e. 2024;95:102204.\u003c/li\u003e\n\u003cli\u003eWarner B, Ratner ER, Datta A, Lendasse A. A systematic review of phenotypic and epigenetic clocks used for aging and mortality quantification in humans. \u003cem\u003eAging (Albany NY)\u003c/em\u003e. 2024;16:12414-12427.\u003c/li\u003e\n\u003cli\u003eJordanek T et al. Accelerated epigenetic aging and its role in brain dynamics and cognition in young adulthood. \u003cem\u003eHum Brain Mapp\u003c/em\u003e. 2025;46:e70261.\u003c/li\u003e\n\u003cli\u003eLuders E, Narr KL, Thompson PM, Toga AW. Neuroanatomical correlates of intelligence. \u003cem\u003eIntelligence\u003c/em\u003e. 2009;37:156-163.\u003c/li\u003e\n\u003cli\u003eNarr KL et al. Relationships between IQ and regional cortical gray matter thickness in healthy adults. \u003cem\u003eCereb Cortex\u003c/em\u003e. 2007;17:2163-2171.\u003c/li\u003e\n\u003cli\u003eRhein C et al. Neuroanatomical correlates of intelligence in healthy young adults: the role of basal ganglia volume. \u003cem\u003ePLoS One\u003c/em\u003e. 2014;9:e93623.\u003c/li\u003e\n\u003cli\u003eGe R et al. Normative modelling of brain morphometry across the lifespan with CentileBrain: algorithm benchmarking and model optimisation. \u003cem\u003eLancet Digit Health\u003c/em\u003e. 2024;6:e211-e221.\u003c/li\u003e\n\u003cli\u003ePiler P et al. Cohort profile: the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) in the Czech Republic. \u003cem\u003eInt J Epidemiol\u003c/em\u003e. 2017;46:1379-1379f.\u003c/li\u003e\n\u003cli\u003eMareckova K et al. Epigenetic clock as a correlate of anxiety. \u003cem\u003eNeuroimage Clin\u003c/em\u003e. 2020;28:102458.\u003c/li\u003e\n\u003cli\u003eMareckova K et al. Functional impact score of mitochondrial variants and its relationship with functional connectivity of the brain: potential origins of premature aging in young adulthood. \u003cem\u003eHum Brain Mapp\u003c/em\u003e. 2026;47:e70447.\u003c/li\u003e\n\u003cli\u003eTian Y et al. ChAMP: updated methylation analysis pipeline for Illumina BeadChips. \u003cem\u003eBioinformatics\u003c/em\u003e. 2017;33:3982-3984.\u003c/li\u003e\n\u003cli\u003eTeschendorff AE et al. A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. \u003cem\u003eBioinformatics\u003c/em\u003e. 2013;29:189-196.\u003c/li\u003e\n\u003cli\u003eWechsler D. \u003cem\u003eWAIS-IV: Wechsler Adult Intelligence Scale\u003c/em\u003e. 4th edn. Pearson; 2008.\u003c/li\u003e\n\u003cli\u003eTam WCC. The utility of seven-subtest short forms of the Wechsler Adult Intelligence Scale-III in young adults. J Psychoeduc \u003cem\u003eAssess\u003c/em\u003e. 2004;22:62-71.\u003c/li\u003e\n\u003cli\u003eHayes AF. \u003cem\u003eIntroduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. 2nd edn.\u003c/em\u003e Guilford Press; 2018.\u003c/li\u003e\n\u003cli\u003evan Dyck CH et al. Age-related decline in dopamine transporters: analysis of striatal subregions, nonlinear effects, and hemispheric asymmetries. \u003cem\u003eAm J Geriatr Psychiatry\u003c/em\u003e. 2002;10:36-43.\u003c/li\u003e\n\u003cli\u003eKim JH et al. Effects of age on dopamine D2 receptor availability in striatal subdivisions: a high-resolution positron emission tomography study. \u003cem\u003eEur Neuropsychopharmacol\u003c/em\u003e. 2011;21:885-891.\u003c/li\u003e\n\u003cli\u003eBalleine BW, Delgado MR, Hikosaka O. The role of the dorsal striatum in reward and decision-making. \u003cem\u003eJ Neurosci\u003c/em\u003e. 2007;27:8161-8165.\u003c/li\u003e\n\u003cli\u003eZheng Q et al. Functional differentiation of the dorsal striatum: a coordinate-based neuroimaging meta-analysis. \u003cem\u003eQuant Imaging Med Surg\u003c/em\u003e. 2023;13:471-488.\u003c/li\u003e\n\u003cli\u003eRieckmann A, Johnson KA, Sperling RA, Buckner RL, Hedden T. Dedifferentiation of caudate functional connectivity and striatal dopamine transporter density predict memory change in normal aging. \u003cem\u003eProc Natl Acad Sci USA\u003c/em\u003e. 2018;115:10160-10165.\u003c/li\u003e\n\u003cli\u003eChoi EY, Yeo BTT, Buckner RL. The organization of the human striatum estimated by intrinsic functional connectivity. \u003cem\u003eJ Neurophysiol\u003c/em\u003e. 2012;108:2242-2263.\u003c/li\u003e\n\u003cli\u003eCalabrese EJ et al. Biological stress response terminology: integrating the concepts of adaptive response and preconditioning stress within a hormetic dose-response framework. \u003cem\u003eToxicol Appl Pharmacol\u003c/em\u003e. 2007;222:122-128.\u003c/li\u003e\n\u003cli\u003eZihl J, Fink T, Pargent F, Ziegler M, Buhner M. Cognitive reserve in young and old healthy subjects: differences and similarities in a testing-the-limits paradigm with DSST. \u003cem\u003ePLoS One\u003c/em\u003e. 2014;9:e84590.\u003c/li\u003e\n\u003cli\u003eStern Y. What is cognitive reserve? Theory and research application of the reserve concept. \u003cem\u003eJ Int Neuropsychol Soc\u003c/em\u003e. 2002;8:448-460.\u003c/li\u003e\n\u003cli\u003eKundakovic M, Tickerhoof M. Epigenetic mechanisms underlying sex differences in the brain and behavior. \u003cem\u003eTrends Neurosci.\u003c/em\u003e 2024;47:18-35.\u003c/li\u003e\n\u003cli\u003eMareckova K, Devenyi GA, Andryskova L, Chakravarty MM, Nikolova YS. Maturation of hippocampal subfields in young adulthood and its relationship with cognition. \u003cem\u003eHum Brain Mapp.\u003c/em\u003e 2025;46:e70296.\u003c/li\u003e\n\u003cli\u003eKoolschijn PCMP, Crone EA. Sex differences and structural brain maturation from childhood to early adulthood. \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e. 2013;5:106-118.\u003c/li\u003e\n\u003cli\u003eNgun TC, Ghahramani N, Sanchez FJ, Bocklandt S, Vilain E. The genetics of sex differences in brain and behavior. \u003cem\u003eFront Neuroendocrinol.\u003c/em\u003e 2011;32:227-246.\u003c/li\u003e\n\u003cli\u003ePohjalainen T, Rinne JO, Nagren K, Syvalahti E, Hietala J. Sex differences in the striatal dopamine D2 receptor binding characteristics in vivo. \u003cem\u003eAm J Psychiatry\u003c/em\u003e. 1998;155:768-773.\u003c/li\u003e\n\u003cli\u003eDluzen DE. Neuroprotective effects of estrogen upon the nigrostriatal dopaminergic system. \u003cem\u003eJ Neurocytol\u003c/em\u003e. 2000;29:387-399.\u003c/li\u003e\n\u003cli\u003eMcEwen BS. Stress, sex, and neural adaptation to a changing environment: mechanisms of neuronal remodeling. \u003cem\u003eAnn NY Acad Sci\u003c/em\u003e. 2010;1204:E38-E59.\u003c/li\u003e\n\u003cli\u003eHorvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. \u003cem\u003eNat Rev Genet\u003c/em\u003e. 2018;19:371-384.\u003c/li\u003e\n\u003cli\u003eLevine ME et al. An epigenetic biomarker of aging for lifespan and healthspan. \u003cem\u003eAging (Albany NY)\u003c/em\u003e. 2018;10:573-591.\u003c/li\u003e\n\u003cli\u003eBell CG et al. DNA methylation aging clocks: challenges and recommendations. \u003cem\u003eGenome Biol\u003c/em\u003e. 2019;20:249.\u003c/li\u003e\n\u003cli\u003eBelsky DW et al. Quantification of biological aging in young adults. \u003cem\u003eProc Natl Acad Sci USA\u003c/em\u003e. 2015;112:E4104-E4110.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"translational-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"tp","sideBox":"Learn more about [Translational Psychiatry](http://www.nature.com/tp/)","snPcode":"41398","submissionUrl":"https://mts-tp.nature.com/cgi-bin/main.plex","title":"Translational Psychiatry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9400608/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9400608/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePremature epigenetic aging has been linked to poorer cognitive performance, but the neuroanatomical pathways underlying this association remain unclear. Therefore, we tested whether premature epigenetic aging predicts deviations from normative brain development and cognition in young adulthood. We followed 254 young adults from the ELSPAC prenatal birth cohort. Epigenetic aging was estimated using Horvath\u0026rsquo;s DNA methylation clock, CheekAge, and AltumAge. Structural MRI was processed with FreeSurfer to derive cortical thickness and surface area in 68 Desikan\u0026ndash;Killiany regions and subcortical volumes in 14 Aseg regions. CentileBrain normative models trained on 37,407 MRI scans from individuals aged 3\u0026ndash;90 years were used to quantify age- and sex-specific deviations from normative brain development. Cognitive performance was assessed using the WAIS-IV. Premature epigenetic aging estimated using Horvath\u0026rsquo;s clock was associated with more negative deviations from normative volume of the putamen and caudate, but not with other subcortical volumes, cortical thickness, or surface area. In women, this premature epigenetic aging was also associated with lower full-scale IQ and performance IQ, and moderated mediation analyses indicated that the deviations in putamen volume mediated the relationship between premature epigenetic aging and cognitive performance. In men, the relationship between premature epigenetic aging and cognitive performance was quadratic (U-shaped). No similar effects were observed when using CheekAge or AltumAge.Premature epigenetic aging estimated using Horvath\u0026rsquo;s clock and the associated deviations in development of the dorsal striatum in young adulthood may represent an early marker of poorer cognitive performance long before clinical symptoms might emerge.\u003c/p\u003e","manuscriptTitle":"Premature Epigenetic Aging, Deviations from Normative Brain Development, and Cognitive Performance in Young Adulthood","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 06:50:22","doi":"10.21203/rs.3.rs-9400608/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2026-05-11T09:34:18+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-07T22:23:42+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-04T10:28:55+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-04-22T18:21:04+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-04-22T08:25:35+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2026-04-21T14:14:27+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-14T14:45:13+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-14T14:42:14+00:00","index":"","fulltext":""},{"type":"submitted","content":"Translational Psychiatry","date":"2026-04-14T06:50:23+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2026-04-13T14:48:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"translational-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"tp","sideBox":"Learn more about [Translational Psychiatry](http://www.nature.com/tp/)","snPcode":"41398","submissionUrl":"https://mts-tp.nature.com/cgi-bin/main.plex","title":"Translational Psychiatry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6fc953a4-b9c1-4074-9338-5672d9b2d201","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"revise","date":"2026-05-11T09:34:18+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-07T22:23:42+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-04T10:28:55+00:00","index":1,"fulltext":"This content is not available."}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":66745659,"name":"Health sciences/Pathogenesis"},{"id":66745660,"name":"Biological sciences/Neuroscience"}],"tags":[],"updatedAt":"2026-05-11T09:36:30+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 06:50:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9400608","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9400608","identity":"rs-9400608","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}