Cases of trisomy 21 and trisomy 18 among

Genetic results

We screen shotgun sequencing data from 9855 prehistoric and historic individuals (see “Methods” for descriptions of laboratory protocols, quality control measures and data processing methodology). For each individual, we record the number and proportion of reads that mapped to each of the autosomal chromosomes.

We identify six new positive cases of Down syndrome, two genetically female and four genetically male. As expected, the positive cases for Down syndrome showed an approximately 1.5-fold higher proportion of reads mapping to chromosome 21 (range: 1.44–1.52) when compared to the mean proportion for negative cases (Fig. 1). All normalised posterior probabilities are one (to machine precision) for each positive case.

Fig. 1: Fold-increase in reads mapping to chromosome 18 (x-axis) and chromosome 21 (y-axis), corrected for library protocol.

Filled black circles show individuals for which there were at least 1000 reads, and for which at most one chromosome was significantly overrepresented. Labels indicate individual IDs. Source data is provided as a Source data file.

We also reanalyse and confirm a previously published case of trisomy 21, the infant PN07 who was interred in a large Neolithic portal tomb in Poulnabrone, Ireland (3629–3371 BCE)15, showing that our method can be applied to sequencing data produced in any laboratory.

Among the 9855 screened individuals, we also identify one case (CRU013) of the comparatively rarer Edwards syndrome. The proportion of reads that mapped to chromosome 18 is 1.47 times greater than for the median for the negative group (Fig. 1) and the normalised posterior probability of the individual carrying Edwards syndrome is one (to machine precision). The individual was found to be genetically female, in line with the fact that Edwards syndrome is observed in genetically female individuals more often than in genetically male individuals by a ratio of 3:218.

Archaeological contexts

Of the six newly identified cases of Down syndrome, five date to between 5000 and 2400 years ago and were intramural burials (burials within settlements and not in dedicated necropoleis). The remaining case dates to the eighteenth century CE and was found within a Finnish church cemetery. We present these results in chronological order (Table 1).

Table 1 Sample information for all genetically confirmed cases of trisomy 18 and 21

The earliest case in our new data, YUN039 (2898–2700 cal BCE), is a 6-month-old, genetically female individual, discovered in a ceramic-vessel buried under the floor of a dwelling from the Early Bronze Age layer of Tell Yunatsite, from the Pazardzhik Province of southern Bulgaria. YUN039 was not found with any burial items19,20. LAZ019 (1398–1221 cal BCE) is a 12–16-month-old, genetically female individual, discovered buried in a tiny cist grave in a yard-like area within the confines of a dwelling at the Mycenaean site of Lazarides on the island of Aegina, Greece21. LAZ019 was buried wearing a necklace made of 93 beads, glass paste, faience and carnelian (4 beads), of different colours and sizes22.

CRU001 (600-400 BCE), CRU024 (779-549 BCE) and ERE004 (801-764 BCE) were neonatal individuals, aged ~38, ~28 and ~26 weeks gestation at the time of death, respectively23,24,25,26,27. CRU001 and ERE004 are genetically male, and CRU024 is genetically female. These remains were excavated from burials beneath dwellings found at the site of Alto de la Cruz (CRU) and Las Eretas (ERE), two of the first proto-urban centres in Early Iron Age Navarra, Spain. CRU024 was buried with rich grave goods, including bronze rings, a Mediterranean seashell, and surrounded by the complete remains of three sheep and/or goats. While CRU001 and ERE004 were apparently buried in common dwellings, CRU024 was found buried in a structure with a large, decorated fireplace, which was likely a place of ritual.

Also found at the site of Alto de la Cruz in an intramural burial was individual CRU013 (600-400 BCE), a positive case of Edwards syndrome, who was estimated to have died at ~40 weeks of gestation.

Finally, HKI002 (1640–1790 CE) was found buried in a wooden coffin under present-day Senate Square in Helsinki, previously a church graveyard in Post-Medieval Finland. The funeral attire of HKI002 contained bronze pins and decorative bronze flowers, which were a common trend of the time.

Rates of prevalence

The six observed cases of Down syndrome in our data indicate a past rate of prevalence of ~1:1643, which is significantly lower than the overall modern rate of prevalence for Down syndrome of 1:7059. It is known that the age of the mother is strongly associated with an increased rate of prevalence9. When compared to the rate of prevalence for present-day mothers under the age of 20 (1:1282), we continue to find a lower rate of prevalence in past populations (Fig. 2), although this difference is not statistically significant. However, past rates of prevalence could be underestimated as infant burials are generally under-represented in the archaeological record28. Similarly, it is possible that the average age of mothers was lower due to an increased risk of death for the mother during childbirth in pre-modern times and assisted reproductive technologies allowing increased instances of motherhood at older ages in modern populations, although this lower mean age may be offset by an increased age of menarche in premodern populations9,29,30,31,32.

Fig. 2: Estimated rate of prevalence (log-scale) grouped by mother’s age (from Mai et al.9).

The red dashed line indicates the total rate of prevalence, ignoring mother’s age, and the blue dashed line indicates the observed rate of prevalence in the premodern individuals. Source data is provided as a source data file.

The single case of Edwards syndrome indicates a premodern rate of prevalence of 1:9855, which is not statistically significantly different from the modern prevalence of 1:3226 (p ≈ 1)9. Finally, observing no cases of Patau syndrome in 9855 individuals is not unexpected, given that the modern rate of prevalence is 1:7143 (p ≈ 1)9. However, we caution that Edwards and Patau syndrome almost always result in pre- or perinatal deaths. Given that perinatal remains will often be poorly preserved due to reasons such as reduced bone mineralisation and smaller bone size, thus making their bones more vulnerable to destruction and their discovery more difficult, our rates of prevalence are likely underestimated33.

Osteological results—trisomy 21

In this section, we identify osteological markers in our sample of seven individuals that could be indications of autosomal trisomies. We report all skeletal pathologies that are either known to be related to Down or Edwards syndrome or that are observed in more than one individual. However, we caution that osteological analyses alone are insufficient for such diagnoses, as no osteological symptom is pathognomonic for trisomy 21 or 18. Moreover, a large proportion of these features could be linked to environmental factors, such as malnutrition, and to other genetic disorders that can result in abnormal growth patterns or could even fit into normal growth variability2. However, the genetic diagnosis of these individuals offers an opportunity to suggest a relationship of the observed features to the identified chromosomal abnormality, directly or as a consequence. The osteological study of the individuals discussed in the present work are further complicated for two reasons: first, the skeletons are incomplete and often poorly preserved (see Supplementary Methods). Second, our cases are either late fetal losses, stillbirths or, at most, 16 months of age, and not all markers may be observable2.

To date, studies evaluating the skeletal manifestations of individuals with Down syndrome in past populations are few in number and often too statistically underpowered to answer questions about skeletal development, although studies based on modern individuals have shown that porosity and skeletal health issues, shortened femur length and reduced bone mineral density, are significantly more prevalent in cases of Down syndrome34. Several skeletal abnormalities have been identified in fetuses with Down syndrome, including: short stature with reduced length of long bones, microcephaly, brachycephaly, absence of nasal bone ossification and hypoplasia of the middle phalanx of the fifth digit35. Among the skeletal manifestations of Down syndrome that develop later in childhood are scoliosis, pes planus and Perthes disease35. Hence, we were motivated to look for abnormalities in the skeletons of the individuals that were genetically diagnosed as carrying trisomy 21.

Hand phalanges ossify at around 24 weeks of gestation, but they are so small that they are rarely recognised or correctly identified during excavation, thus making systematic recording impossible. Vitamin C, D, A, complex B and/or iron deficiency has also been associated with Down syndrome, individually or in combination, likely arising as a secondary complication of features of Down syndrome such as feeding difficulties36. Specifically, these deficiencies are commonly associated with cranial porosities including locations such as the sphenoid body, the external cranial vault, the superior aspect of the eye orbit (or cribra orbitalia), as well as on long bones, which can be anticipated in individuals with trisomy 2137.

Despite the different stages of completeness and preservation throughout the cases, there is overlap in the anthropological observations, with the most striking being the reported growth disorders on bones, followed by porosities in the cranial elements. Overall, we observe porosity in at least one of the bones of the cranium in 4/6 cases (Supplementary Figs. 2 and 3), visible in the orbital roofs (2/3), mandible (3/4), the upper palate (2/3), the frontal and parietal bones (3/4), and/or the occipital squama (4/6). We also observe additional bone formation on the jugular limb of the pars lateralis in both cases for which the occipital bone was sufficiently well preserved for observation. Less common were visible striae or pitting of either the long bones (1/3) or the ilium (1/2). HKI002 was the only individual for which we observed (Supplementary Fig. 4) inner bone (incus) malformation (1/1), and dental enamel hypoplasia38 (1/1) (Supplementary Fig. 5).

Other traits that can be associated with trisomy 21 overlap in three instances: individuals CRU001 and HKI002 present irregular bone growth on the pars lateralis segment of the occipital bone (e.g., flange, Supplementary Fig. 6), LAZ019 and YUN039 present morphological changes on the femora (Supplementary Fig. 5) (flattened and widened distal metaphysis, and antero-posterior flexion of the femur diaphysis, respectively) as well as cribra orbitalia and porotic hyperostosis. Additionally, HKI002 and YUN039 present signs of Vitamin C deficiency which often co-occurs with trisomy 21. CRU001 also displays a protrusion of the occipital squama, a cranial trait associated with trisomy 2139 (Supplementary Fig. 7). A compilation of the traits recorded for each individual can be seen in Supplementary Table 1, and we stress that some of these observed markers, such as dental hypoplasia and bone porosity, can be caused by many other diseases and even normal growth processes.

Osteological results—trisomy 18

Edwards syndrome is associated with very severe physical symptoms40,41, although very few clinical studies describe skeletal malformation42. In total, CRU013 presents numerous bone anomalies, some not being previously reported in the literature, such as: an abnormally shaped left scapula (Supplementary Figs. 8 and 9) which could be linked to describe chest malformations43, irregular morphology of the left hemiarch of the axis40 (Supplementary Fig. 10), extremely thin humeral diaphysis and pronounced curvature of the epiphyseal surfaces for both humeri (Supplementary Fig. 11), thinning of the diaphysis of the right femur (left missing) and the left tibia (right missing) (Supplementary Fig. 12). Furthermore, an incomplete bone fragment, likely the diaphysis of a fibula and not a clavicula, shows an unusual twist (Supplementary Fig. 13). The age of CRU013 was estimated at 40 weeks gestation when using the maximal length of both humeri (Supplementary Table 2). Of note, however, is the underdevelopment of the distal width of the right humerus, according to which the age estimate for the same individual would have been only 30 weeks of gestation. We thus caution that estimating the age of the individual using long bones, when these bones are shown to have undergone irregular development, and when short stature is a common symptom of trisomy 18, will likely produce downward-biased estimates.