What Recent DNA Research Actually Shows About Hair
- Beth Thompson
- Aug 19
- 4 min read
What Recent DNA Research Actually Shows About Hair
I've spent much of my career with human hair, sorting it, assessing it, handling single strands closely enough to notice things most people never look for. I therefore read the recent forensic-genetics literature with a particular interest, because it revises an assumption I had long taken as settled: that a hair without its root tells you very little.
The assumption that stood for two decades
For most of the last two decades, the working position in forensic science seems straightforward. A hair root can yield a full nuclear DNA profile that can identify an individual. A shaft without a root was treated as mitochondrial-only territory: useful, but a lineage marker rather than an identifier, because mitochondrial DNA is shared down a maternal line and cannot separate a mother from her children or her siblings. Mitochondrial DNA exists in hundreds to thousands of copies per cell against nuclear DNA's two, so in a degraded shaft there is simply more of it to find. That copy-number difference is why mitochondrial analysis became the default method for rootless shafts.
What newer sequencing methods found
Massively parallel sequencing, next-generation sequencing, changed what is recoverable, because it can capture and read very short, highly fragmented pieces of DNA in bulk. That is exactly the state nuclear DNA is in once it has degraded inside a shaft.
In 2018, Brandhagen, Loreille and Irwin, working at the FBI Laboratory, applied shotgun sequencing to shed hairs and reported that nuclear DNA made up at least 88 percent of the recoverable human DNA in any given sample, and generally more than 95 percent, and that this held even in the most distal portions of the shaft, farthest from the root. The finding did not show that shafts hide a little nuclear DNA. It showed that nuclear DNA is the majority of what is there: abundant, but broken into fragments too small for older methods to read.
A study published in Genome Biology in February 2026 extended this directly. Applying genome-sequencing methods to rootless head and pubic hair shafts, the authors found that, on average, 96.94 percent of the recoverable human DNA was nuclear and 3.06 percent mitochondrial. For the 77 hairs that reached sufficient sequencing depth, greater than one-fold average coverage, genotype calls matched known reference profiles with greater than 99.4 percent concordance. The authors describe this as statistically compelling evidence for identity or non-identity, not merely a supporting lineage marker. The concordance figure belongs to the hairs that sequenced well, not to every hair tested.
The length of the hair is now a variable
A separate 2026 study, by Li and colleagues at Sun Yat-sen University, tested short-amplicon sequencing panels on shafts and roots from ten individuals and found that identification accuracy scaled with shaft length. From a 15-centimeter shaft, they reported likelihood-ratio effectiveness of 0.95 for individual identification, 0.93 for paternity testing and 0.74 for grandparent–grandchild kinship. Those figures fell to 0.63, 0.41 and 0.32 respectively from a 2 centimeter shaft.
The authors did not see this as a final result. Roughly 30 percent of the genotypes they called from shafts were inconsistent with the matching hair roots; several shafts showed contamination; and their stated conclusion is that contamination remains a substantial challenge when working with DNA of this quality. Longer hair yields more information, that much is documented, but it does so alongside error and contamination rates that are not yet trivial.
Does this displace mitochondrial DNA testing?
None of this replaces mitochondrial analysis. Mitochondrial DNA is still recovered in a high proportion of naturally shed hair cases, one long-cited figure puts it at 92.5 percent, and mitogenome sequencing has produced reportable results from shafts as old as 27 years. What mitochondrial DNA cannot do, by its biological nature, is separate individuals who share a maternal line. The newer work does not overturn it. It adds a second, individualizing layer that a shaft was previously assumed not to carry.
What changed, and what did not
The biology did not change. The nuclear DNA was always in the shaft. What changed is the sequencing technology's ability to read material that was always present but too fragmented and too sparse to interpret.
Where this becomes misleading
This research is about forensic identification, establishing whether a shaft came from a particular person, or from a relative of that person. It is not a method for establishing where hair came from, which population it belongs to, or whether a supplier's label is accurate. Those are different questions, and this body of work does not answer them. It is a genuine advance in what a single strand can reveal about identity. It is not a shortcut to provenance, and I would be cautious of anyone who presented it as one.
Closing point
A strand of hair holds more than we assumed. It still does not hold everything we might wish it did. Both of those are worth saying clearly.
Beth Thompson is the founder of Lux Symbolica, a Paris-based B2B rare hair curation and sourcing service operating at the intersection of professional expertise, and the luxury supply chain.
Professional enquiries by authorization only.
© 2026 Lux Symbolica SASU
Citations
Brandhagen, M.D., Loreille, O. & Irwin, J.A. (2018). Fragmented Nuclear DNA Is the Predominant Genetic Material in Human Hair Shafts. Genes, 9(12):640. https://doi.org/10.3390/genes9120640
Rootless hair as a reliable source of forensic genetic information (2026). Genome Biology. https://doi.org/10.1186/s13059-026-03981-8
Li, R., Wang, N., Dai, S. et al. (2026). Individual identification and kinship testing from hair shaft nuclear DNA: leveraging short amplicon strategy and bioinformatics models. International Journal of Legal Medicine, 140(2):713–723. https://doi.org/10.1007/s00414-025-03651-7
Routine Mitogenome MPS Analysis from 1 and 5 mm of Rootless Human Hair,
DOI: 10.3390/genes13112144
Comments