Background: Bloodstains are among the most frequently encountered biological traces found at crime scenes. They provide information for both human identification (HID) and investigative leads. A key application is the use of DNA methylation (DNAm) to estimate the chronological age of unknown donors. However, most DNAm models for age estimation have been developed and validated on fresh biological samples. Consequently, their applicability to realistic forensic traces, such as aged bloodstains, remains insufficiently explored. Beyond donor traits, estimation of the time since deposition (TsD) is essential for crime reconstruction. However, the use of DNAm to estimate TsD remains largely unexplored.

Aim: This study aimed to evaluate the stability of DNAm in bloodstains over time and under different environmental conditions with the purpose of estimating the age of the donor and the TsD of the stain.

Materials and Methods: A total of 92 samples were collected. The dataset included indoor and outdoor bloodstains from three donors, with TsD ranging from zero to 365 days. To capture seasonal variation, an additional outdoor subset with TsD of zero to nine days was analyzed for each season. DNAm profiles were generated using the Illumina Infinium MethylationEPIC v2 BeadChip Array. Different age prediction clocks (BLUP, skinHorvath, Levine, EN, Horvath) were evaluated by comparing predicted and chronological age using mean absolute error (MAE) and root mean squared error (RMSE). In parallel, TsD methylation variation was explored by comparing DNAm levels across time points and conditions. 

Results: Chronological age estimation remained accurate in bloodstains aged up to one year. Among the tested models, BLUP showed the best performance, with a MAE of 2.21 years, followed by skinHorvath (MAE = 3.92 years). In the indoor/outdoor setup, BLUP MAE increased from 1.38 years at deposition (fresh samples) to 2.71 years at 365 days. Absolute prediction error did not show significant differences across TsD levels, supporting the preservation of age-associated DNAm signatures over time and under different conditions. By contrast, signals associated with TsD were limited and influenced by inter-individual variability.

Conclusion: Age-associated DNAm signatures remained stable in bloodstains for up to one year under both indoor and outdoor conditions, supporting the use of DNAm-based chronological age prediction as a reliable approach in forensic casework. At the same time, the observed overall stability of DNAm across time points suggests that TsD-related changes are small and challenging to generalize across samples.

Authors

• Noemi Argirò (Institute of Legal Medicine, University of Brescia, Italy)
• Brando Poggiali (Section of Forensic Genetics, Department of Forensic Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark)
• Andrea Verzeletti (Institute of Legal Medicine, University of Brescia, Italy)
• Claus Børsting (Section of Forensic Genetics, Department of Forensic Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark)
• Marie-Louise Kampmann (Section of Forensic Genetics, Department of Forensic Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark)
• Jeppe Dyrberg Andersen (Section of Forensic Genetics, Department of Forensic Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark)