Cases of trisomy 21 and trisomy 18 among historic and prehistoric individuals discovered from ancient DNA

The study examines how past societies were affected by and responded to disease, focusing on chromosomal trisomies in historic and prehistoric populations. Osteological approaches alone often cannot diagnose conditions that do not reliably manifest in skeletal tissue, and rare diseases are under-represented due to taphonomic and methodological factors. Individuals with chromosomal trisomies carry three copies of a chromosome following nondisjunction during meiosis. Full autosomal trisomies are typically nonviable except for trisomy 21 (Down syndrome), trisomy 18 (Edwards syndrome), and trisomy 13 (Patau syndrome), with survival of trisomy 13 and 18 into childhood being rare without modern care. Documented ancient cases are few, with only a handful of Down syndrome cases suggested osteologically and one genetically identified Neolithic case from Ireland. The scarcity likely stems from low prevalence rates and limitations of ancient DNA data for copy-number analyses. This study aims to systematically screen a large aDNA dataset using methods suited to low-coverage data to identify trisomy cases and integrate genetic findings with osteological and archaeological evidence to infer social responses in past communities.

The paper contextualizes current knowledge: modern prevalence rates for live births, stillbirths, and terminated pregnancies are approximately 1:705 for trisomy 21, 1:3226 for trisomy 18, and 1:7143 for trisomy 13. Historical osteological reports have suggested a few Down syndrome cases, and a recent genetic identification of a Neolithic Irish infant with trisomy 21 exists; however, no confirmed prehistoric or historic cases of trisomy 13 or 18 were previously established. The authors note that conventional copy-number methods require longer reads and higher coverage than typical aDNA affords, explaining the lack of ancient detections. They also summarize known skeletal manifestations associated with Down syndrome (e.g., shortened long bones, cranial features, porosity, reduced bone mineral density) and the severe physical symptoms in Edwards syndrome, acknowledging variability and the lack of pathognomonic osteological markers.

• Samples and data: Shotgun sequencing data from 9,855 prehistoric and historic individuals were analyzed. Reads were adapter-trimmed and aligned to GRCh37 using BWA. Read counts mapping to each autosomal chromosome were tabulated after filtering for minimum read length ≥35 bp and mapping quality ≥25.
• Trisomy detection: For each individual and chromosome, the proportion of reads mapping to that chromosome was modeled with a beta-binomial framework to account for overdispersion from library protocols, sequencing runs, and DNA preservation. A binomial likelihood with a beta prior produced posterior probabilities for trisomy states. Parameters of the beta prior were estimated via maximum likelihood from a subset of 7,103 samples with ≥10,000 reads, stratified by library protocol. Screening required as few as 1,000 mapped reads per individual. Z-scores were computed as standard deviations of observed proportions relative to the beta-binomial expectation.
• Statistical analysis: Beta-binomial parameters estimated using Rfast::beta.mle; densities via extraDist::dbbinom; prevalence comparisons via stats::binom.test; visualization via ggplot2.
• Genetic sexing: Based on the fractions of reads mapping to X and Y (p_x and p_y), clustering individuals into XX vs XY as per Roca-Rada et al.
• Osteology: No re-examination of remains; osteological observations compiled from prior publications, photographs, and reports. Multiple age-at-death estimators were used for perinatal individuals, primarily long-bone lengths when appropriate. Observed skeletal markers were compared to features consistent with trisomy 21 or 18.
• Validation: Reanalysis of a previously published Irish Neolithic case (PN07) demonstrated method portability across laboratories.

• Screening outcome: Among 9,855 individuals, six new cases of trisomy 21 (Down syndrome) and one case of trisomy 18 (Edwards syndrome) were identified; a previously published Neolithic case of trisomy 21 (PN07) was confirmed. All new cases were infant, perinatal, or stillborn.
• Genetic signal: Trisomy 21 cases showed ~1.5-fold higher mapping proportions to chromosome 21 (range 1.44–1.52) relative to negatives; the trisomy 18 case showed a 1.47-fold increase for chromosome 18. Normalized posterior probabilities of trisomy were ~1 (machine precision) for all positives. The trisomy 18 individual (CRU013) was genetically female, aligning with a known female bias (3:2).
• Archaeology and context: Five of the six new trisomy 21 cases date to ~5000–2400 BP and are intramural burials; one case is from a Finnish church cemetery (1667–1800 CE). Notable burials include: LAZ019 (Aegina, Greece; 12–16 months) with a necklace of 93 beads; CRU024 (Navarra, Spain) with rich grave goods including bronze rings and a seashell, plus three complete caprines; contemporaneous early Iron Age Spain sites (Alto de la Cruz and Las Eretas) yielded three trisomy 21 cases and the sole trisomy 18 case.
• Prevalence estimates: Observed Down syndrome prevalence 1:1643, lower than modern overall 1:705; also lower than mothers <20 years (1:1282), though not statistically significant when stratified by maternal age. Edwards syndrome observed prevalence 1:9855 vs modern 1:3226 (not significantly different). No trisomy 13 observed, which is consistent with modern prevalence 1:7143 given sample size.
• Osteology: No pathognomonic markers; overlapping features included cranial porosities (observed in at least one cranial bone in 4/6 Down cases), cribra orbitalia and porotic hyperostosis in some, irregular growths on occipital bone/pars lateralis where preserved, femoral morphological changes, inner ear (incus) malformation and enamel hypoplasia (HKI002), and signs of vitamin C deficiency in some. The Edwards case (CRU013) showed severe skeletal anomalies (e.g., scapula malformation, axial hemiarch irregularity, very thin and abnormally curved humeri, thin femoral/tibial diaphyses, twisted likely fibula fragment) with discordant age estimates when using long-bone dimensions.
• Social inferences: Mortuary treatments indicate care and community inclusion; individuals were buried according to local norms or with exceptional items, suggesting recognition as community members rather than stigmatization.

The study demonstrates that large-scale aDNA screening coupled with a beta-binomial Bayesian approach can identify rare chromosomal aneuploidies in ancient individuals even at low coverage. The lower observed prevalence of Down syndrome compared to modern rates likely reflects non-random sampling and preservation biases, particularly the under-representation and poorer preservation of perinatal and infant remains. Osteology alone is insufficient for diagnosis due to variability and non-specificity of lesions; nonetheless, recurring features (cranial porosity, abnormal occipital growth, femoral changes) broadly align with known manifestations. The clustering of three trisomy 21 cases and one trisomy 18 case at two nearby Early Iron Age Spanish sites may indicate a higher local frequency of burials of trisomy carriers or specific cultural practices, warranting further investigation. Mortuary practices across cases suggest that affected individuals were treated with care and integrated into community ritual norms. As ancient DNA datasets grow, improved detection and refined screening will enable broader insights into the prevalence and social treatment of rare genetic disorders in the past.

This work presents the first systematic genetic screening and osteological description of autosomal trisomies in premodern samples, identifying six new Down syndrome and one Edwards syndrome cases and confirming a prior Neolithic Down syndrome case. The novel beta-binomial method detects trisomies from low-coverage aDNA and integrates archaeological and osteological data to contextualize findings. The study underscores that osteology cannot reliably diagnose trisomies in isolation and that preservation and sampling biases likely depress observed prevalence. Future research should expand dataset size, refine detection methods for low-coverage data, conduct targeted re-examinations where preservation allows, and investigate cultural and demographic factors contributing to apparent local clusters of cases.

• Preservation and sampling bias: Perinatal and infant remains are under-represented in archaeological contexts and preserve poorly, likely underestimating prevalence.
• Osteological ambiguity: No pathognomonic skeletal markers for trisomy 21 or 18; many observed features (e.g., porosity, enamel hypoplasia) are non-specific and can result from multiple conditions or normal growth variability.
• Incomplete skeletons: Frequent incompleteness and poor preservation limit osteological assessment and comparability across cases.
• Age estimation uncertainty: Long-bone-based age estimates can be biased in individuals with growth disorders (e.g., in the trisomy 18 case).
• Methodological constraints: Reliance on low-coverage read-mapping proportions assumes consistent mapping biases across libraries; despite protocol-specific modeling, residual overdispersion or batch effects may remain. Non-random site selection and temporal/geographic sampling further limit generalizability.