Vindicated by Ancient DNA, Forgotten by Academia: The Bittersweet Fate of My Polygenic Score Research
New preprint provides empirical validation for predicting complex traits in ancient populations
A new preprint by Cox et al. offers compelling evidence supporting the use of polygenic scores (PGS) to study complex traits in ancient populations. By directly comparing PGS predictions with skeletal measurements from 568 ancient Eurasians (38,000–600 BP), this work addresses critical questions about PGS reliability – with significant implications for my own research on cognitive traits and height across time.
Validations and Their Relevance
PGS Predicts Actual Ancient Phenotypes, even across ancestries!
The study demonstrates that height PGS significantly correlates with directly measured femur length in ancient individuals:
"The optimal PRS using European summary statistics explained R² = 0.099 of femur length variance(Fig. 1C).
Though reduced compared to modern populations (consistent with data limitations), this confirms PGS can quantify genetic effects on skeletal phenotypes – a methodology central to my work on ancient cognitive traits.
Trans-Ancestral PGS Validity
To address stratification concerns, the authors successfully applied an East Asian-derived PGS to ancient Europeans:
"The East Asian PRS (unconfounded by European ancestry) achieved R² = 0.053" (Fig. 1D).
This empirically supports the trans-ancestral portability of PGS – a cornerstone of my cross-population analyses of cognitive traits and height.
Genetic Contributions to Historical Height Differences
The study reframes the "Neolithic stature hypothesis" by revealing a genetic component to height declines:
"Neolithic populations showed significantly lower PGS than pre-agricultural groups (β = -0.38 SD, P=0.015)" (Fig. 2C).
"We find evidence for a genetic contribution to [height] differences, and limited evidence for environmental effects"
This aligns with our findings (Piffer & Kirkegaard, 2024) that Neolithic farmer ancestry contributed to reduced height in Europeans – not solely environmental factors: “The most pronounced decrease in Height PGS occurred from the Paleolithic to the Neolithic (Supplementary Figure S9). This fits the archaeological record of a reduction in stature in the bones of Neolithic farmers compared to Paleolithic foragers”.
This study supports our conclusion (Piffer & Kirkegaard, 2024): “Notably, a decline in height PGS was observed from the Paleolithic to the Neolithic periods, implying that the postagricultural revolution reduction in body size could be attributed not only to dietary limitations and heightened prevalence of diseases (Mummert et al., 2011), but also to genetic factors”.
This study also validates my finding (Piffer, 2025) of a negative effect of Anatolian farmer ancestry on height PGS in a sample of individuals from the North-Pontic Caspian (NPC) steppe:
For a summary of this study, check my previous blog post.
Gene-Environment Interactions in Deep Time
A striking finding: the lactase persistence allele (rs4988235) had strong effects on ancient height (β ≈ 0.35 SD) despite null effects today:
"If we were to run a GWAS for stature in Bronze/Iron Age Europe, rs4988235 would be the most significant SNP"
This highlights how PGS can recover historical gene-environment interactions.
Implications for Cognitive Genetics
This work provides three key validations for our research program:
Methodological robustness: PGS reliably quantifies genetic effects on skeletal proxies (height → femur length), supporting parallel work on cognitive traits.
Ancestry portability: East Asian PGS predicted ancient European height, affirming trans-ancestral PGS applications.
Genetic clines over time: Neolithic height declines were partially genetic – consistent with ancestry-based trait differences in my analyses.
The authors note:
"Our study represents the first systematic analysis of this scope in an ancient population, resulting in the first empirical evidence of gene-environment interaction in an ancient sample"
This independent replication of PGS efficacy across millennia and ancestries is encouraging validation for our approaches. It implies that:
Genetic effects on complex traits persist across temporal divides
PGS can disentangle genetic/environmental contributions in historical transitions
Skeletal proxies + ancient DNA enable direct tests of evolutionary hypotheses
A bittersweet victory
This study provides robust independent validation for our core methodologies, particularly the use of trans-ancestral PGS in ancient contexts and the genetic basis of historical height declines. As a researcher applying similar approaches to cognitive traits and height, I find these empirical confirmations deeply encouraging.
Yet it is disappointing to note that our prior work, which directly anticipated key findings here, remains uncited. Our 2024 study (Piffer & Kirkegaard) first demonstrated that Neolithic European height declines were partially genetic rather than solely environmental, using similar PGS approaches. My paper on the genetics of the Steppe Pastoralists (Piffer, 2025) also showed how the Yamnaya were genetically much taller than the Anatolian farmers, yet it also went uncited. Instead, Cox et al. (2025) cite Akbari et al., who themselves overlooked our work despite addressing similar questions about ancient DNA.
This reflects a concerning pattern in academia where foundational contributions are sometimes excluded from citation chains, even when they provide methodological precedence in applying PGS to ancient ancestry shifts, conceptual anticipation of genetic versus environmental effects in Neolithic transitions, and direct empirical alignment showing farmer ancestry negatively correlating with height. Such omissions undermine the scholarly record and fail to acknowledge the incremental nature of scientific progress.
References
Akbari, A., ..., and D. Reich. (2024). Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. Working paper, BioRxiv.
Cox, S. L., Kaymak-Loveless, K., Shin, C., Alihodžić, T., Alt, K. W., Atanassova, N., Binder, D., Čaušević-Bully, M., Chohadzhiev, A., Chokhadzhiev, S., Duday, H., Gaydarska, B., Khudaverdyan, A., Micó Perez, R., Nicklisch, N., Novak, M., Oliart Caravatti, C., Réveillas, H., Rivollat, M., Rottier, S., Tončinić, D., Zaüner, S., & Mathieson, I. (2025). Effects of ancestry, agriculture, and lactase persistence on the stature of prehistoric Europeans. bioRxiv. https://doi.org/10.1101/2025.07.11.664181
Piffer, D., and E. O. Kirkegaard (2024). Evolutionary trends of polygenic scores in European populations from the Paleolithic to modern times. Twin Research and Human Genetics, 27, 30-49.
Piffer, D. (2025). Why Were the Yamnaya so Successful? An Evolutionary Polygenic Approach. Mankind Quarterly, 65(4), 510-532
I wonder what it is that stops these researchers to cite your research. Is it cowardice - being afraid to be associated with people like you and Emil Kirkegaard, just by mentioning your research? Or is it that most of these people are leftist and woke to the core and thus hate people like you? Or maybe they are not aware of your research?
I think it is probably a mixture of the first two causes. I also think that this is a damning statement about the state of current science and how it is corrupted by ideology and the fear of being called 'racist'. Not considering other research because of ideology / fear is clearly hindering scientific progress.
By the way, a similar incident comes to my mind: Charles Murray's book Human Accomplishment appeared in 2003. It is an extremely well researched study of the topic until about 1950.
In 2023, Benoit de Courson et al. published an article with the title Quantifying the Scientific Revolution in the journal Evolutionary Human Sciences. Figure 4. of this article shows the 'geographic distribution of scientific production' in the 16th, 17th and 18th centuries. The maps - in particular the first two - are almost identical to the well-known map in Murray's book on page 297 which shows the 'European Core' of accomplishment. Nevertheless, the article ignores Murray's book completely.
his position, was arguing that polygenic scores for traits like height are accurate in Europeans but less reliable in non-European populations because selection pressures for height-related alleles differ across groups.
Overall, I think his point is not that we’ll never be able to identify these traits accurately, but rather that more research is needed for populations that are currently understudied.
I believe you countered by arguing that humans are highly similar as a species and that the traits in question have been under recent selection.
If I were to put myself in Kevin Bird’s shoes, I think he might argue that even the selection for these traits could be influenced by genetic drift and random mutations, which may not be consistent across populations and could represent noise in larger samples.
For populations like those in India and Africa, I think more research is needed, as they are heavily understudied. However, I believe we can make significant progress with further study.
I may have misrepresented Kevin Bird’s argument, but I think this is close to what he would say directly just based on how he writes in his essays obviously I can't do a psychological analysis but I think he could make an argument like this