Shedder status—An analysis over time and assessment of various contributing factors

The shedder status of a person is an important consideration when evaluating probabilities of DNA transfer during activity‐level assessments. As an extension of our previously published study, the shedder statuses of 38 individuals were reassessed 1 year later. The study found that shedder status may change over time for some individuals and was associated with one's gender, number of items touched, and mobile phone usage. In 29% of touch events, no DNA allele was detected and in 99% of touch events, the amount of DNA deposited was <2 ng. The study also found that in 0.6% of touch events, the participant could be excluded as a contributor of the observed DNA profile, with another person being included. Additionally, our investigations suggest that the current three‐category system for shedder status classification may require further refinement to better represent the individuals' shedder status in a population.


| INTRODUC TI ON
The concept of shedder status was first proposed by van Oorschot and Jones [1], who demonstrated that the dominant DNA profile recovered from an object handled by multiple persons may not always originate from the last person who handled it; rather, it was dependent on the individual. The term "shedder status" refers to the tendency of a person to leave behind DNA upon physical contact.
A person can be classified as a good shedder, a poor shedder, or an intermediate shedder depending on the amount or quality of DNA left behind. Over the years since the concept of shedder status was proposed, numerous studies have been conducted to categorize individuals by their shedder status, identify factors that influence shedder status, and examine the outcomes from the interactions of individuals of different shedder statuses with the environment [2][3][4][5][6][7][8][9][10][11][12][13]. Several studies with different approaches to determining an individual's shedder status are summarized and compared in a recent publication, which also looked at the impact of moisturizer application on one's shedder status [14].
The shedder status of an individual is important as it can address questions such as the probability of recovering the individual's DNA from a surface, whether the DNA recovered was deposited directly or through transfer, and whether an individual was the last person to handle an object, among others. Despite its importance, there is currently no standard method to determine an individual's shedder status or to classify an individual into a shedder group. In a study by Samie et al. [15], the authors noted that the amount of DNA deposited by an individual can fluctuate greatly, and consequently, recommended to assign an individual to a shedder group based on the overall results from multiple depositions. Another study by Johannessen et al. [16] attempted to use a fluorescent cell count method to determine an individual's shedder status. This method had the potential advantages of being quicker and simpler, doing away with the laborious DNA profiling processes and at the same time, eliminating variations caused by different laboratory protocols. Unfortunately, the method had limited success due to poor association between the number of detected cells and the actual quantity of DNA recovered.
In recent years, with the Courts' appreciation of DNA evidence expanding beyond the source and sub-source to activity level, the importance of shedder status in the context of Bayesian networks, to ascribe probabilities of transfer, persistence, and recovery of DNA has become more crucial than ever. While strong inferences can be made regarding the source of a DNA evidence using population data and software, inferences regarding the activities that produced the DNA evidence remain tentative. The Bayesian network is a powerful tool in the analysis of activity-level DNA evidence as it is able to integrate the probabilistic relationship of multiple variables.
However, there are challenges in accounting for every variable since the exact activities that occurred during a crime incident is often unknown. There are also other factors of importance in the analysis of activity-level DNA evidence such as background DNA and mechanisms of DNA transfer that are yet to be fully understood. Therefore, some practitioners are of the view that evaluations regarding activities are considered not robust and should not be allowed. However, others have opined that not pursuing the topic leaves recipients of the information without guidance, which may lead to erroneous conclusions [17]. A survey conducted on American forensic DNA practitioners found moderate support for activity-level reporting, with participants expressing major concerns in several key areas such as insufficient number of realistic activity-level studies and the need for standardized guidelines and approaches [18].
In this study, we examined the shedder status of 38 selected individuals from our previous study [19] after 1 year to investigate whether the shedder status assigned to an individual remained consistent over time. We also attempted to identify factors such as number and type of items handled prior to the touch event, which may be associated with the amount of DNA and number of alleles deposited upon physical contact. Finally, from the data obtained in our previous study, along with the new data from this extended study, we derived some observations regarding evidence that is obtained from touch DNA deposition which may be of interest to the forensic community. Though not exhaustive, the information in our study can assist in expanding the knowledge base in the field of DNA transfers and to support the ongoing advancements in interpretation of activity-based DNA evidence.

| Experimental design
There are three time points in this study: Phase 1 AM, phase 2 AM, and phase 2 PM.

| Phase 1-Shedder status determination
This phase was performed in Tan et al. [19]. Eighty-one participants, comprising of 53 female and 28 male laboratory employees, washed their hands thoroughly with soap and water, and then dabbed dry with clean paper towels. Participants put on a medical face mask and resumed normal activities for 15 min with instructions to avoid food consumption, wearing gloves or touching another person.
Thereafter, participants gripped a sterile 50 mL plastic tube in each hand for 10 s. DNA deposited on each tube was collected with wet and dry cotton swabs. All samples were collected between 8 AM and 10 AM in the morning. Negative controls of five new sterile 50 mL plastic tubes were swabbed and assessed to be free from any detectable DNA beforehand. Participants then completed a questionnaire by recalling from memory and listing down the items they had touched with each hand during the 15-min interval. Each item touched was counted once irrespective of the number and duration of the contact. The process was repeated twice on another 2 days within the same week, giving a total of six samples per participant which resulted in a total of 486 samples for phase 1.

After 1 year (phase 2 AM)
A subset of 38 participants from phase 1, comprising of 28 females and 10 males, repeated the experiment 1 year later. Experimental procedures were identical to phase 1, that is, the DNA deposited on the sterile 50 mL plastic tubes were collected after participants had

| Statistical analyses
Analyses were performed using SPSS Statistics for Windows, ver-

| Variations in shedder status of participants
The proportions of participants of the three shedder groups, for the 81 participants in phase 1, which was conducted in the morning, and the subsequent shedder status of a subset of 38 participants in phase number of alleles, and amount of DNA detected were found to be significantly different depending on participants' gender and mobile phone usage (Tables S1-S3). In contrast, no such differences were detected with respect to the parameter of hand dominance and selfcontact. These findings from the expanded dataset of combined phase 1 and phase 2 results were similarly identified in phase 1 of our study [19]. Two additional variables were examined in phase 2, (1) whether one had touched their own clothing and (2) whether the deposition occurred in the morning or afternoon; no significant differences in the number of self-alleles, total number of alleles, and amount of DNA were detected (Tables S1-S3). Additionally, a moderate Spearman's rank correlation coefficient of 0.520 (p < 0.001) was found between DNA quantity and number of self-alleles, suggesting that the amount of DNA recovered may not be the best predictor of a person's shedder status.
The Spearman's rank correlation test was also performed on all the data collected in phases 1 and 2 to examine for possible correlations between DNA quantity or quality (number of total alleles and self-alleles) with the number of items touched (total items, personal items, and shared items), and only a weak correlation was detected (Table S4). However, when the data were categorized based on shedder status and the Kruskal-Wallis test was performed followed by post hoc tests, it was observed that there were significant differences between the number of items touched by participants of the three different shedder statuses (Table S5) only. In general, the median number of items touched by good shedders (6) was higher compared to intermediate shedders (4), which was in turn higher than that touched by poor shedders (3; Figure S1).

| Characteristics of DNA profile obtained from a single touch event
This study involved the collection of a large number of 942 touch samples which provided for a rich dataset to study the types of DNA profile that could be obtained from a single touch event. A distribution of the number of self-alleles deposited in a single touch event is shown in Figure 4. At 15 min after handwashing, 29% of the touch events did not yield any detectable self-alleles. Only 20% of the touch events yielded 16 or more self-alleles, which is the minimum number of alleles needed for a profile to be reportable according to laboratory guidelines. In terms of secondary transfers, the major-

| Shedder status is associated with one's gender and activities
In agreement with the results of phase 1, mobile phone usage and gender demonstrated an association with one's DNA deposition when results from phase 2 were included (Tables S1-S3) [19]. One possible reason for the low number of good shedders in this study could thus be the presence of more female than male participants, since female participants left behind lower DNA quantity and significantly fewer alleles per touch event compared to male participants. This observation is also in agreement with studies from other research groups which had examined for gender differences [6][7][8][9]. Our results are in agreement with previous studies which found that the amount of DNA transferred through contact is affected by prior activities like touching one's hair, repeated contact with the same object, and physical activities [20][21][22], confirming the association of shedder status with one's activities and behavior. The data gathered across phase 1 and 2 also recognizes that by not limiting participants' activities during the 15-min interval between handwashing and DNA deposition, we would be categorizing the participants based on a combination of both transferred self-DNA (from items touched prior to deposition) and innate propensity of skin cell shedding, instead of innate propensity of skin cell shedding alone [19]. While it may be argued that this is not the true shedder status of an individual but rather is a measure of an individual's DNA deposition ability, it is nonetheless an accurate reflection of an individuals' DNA contribution to the environment in the real world.
In the study by Jansson et al. [11], the authors likewise noted that

| Shedder status may change over time
Our study showed that while some participants maintained their shedder status, others switched from one category to another. The Petricevic [3], despite the many differences in experimental design and methods used.

| Shedder status categories should be further refined
Considering the results from both phase 1 and phase 2 of our study,

| Profile quality and DNA quantities derived from touch DNA samples
The large number of touch samples provided a rich dataset which enabled a study into the characteristics of touch DNA. Analysis of alleles deposition showed that it is common for an individual to leave behind zero self-alleles even when contact has occurred which was the case in 29% of depositions ( Figure 4). Our results suggest that while individuals may make a similar number of contacts, the DNA of some individuals may not be detected as much, or at all, in the environment. While this difference could be a result of differing methodologies and DNA detection sensitivities across laboratories, this also highlights the importance of taking into consideration a person's shedder status and activities prior to the touch event when performing activity-level assessments. Foreign alleles were detected in 21% of samples, which contrasted with the relatively higher 79% in Goray et al. [7] and the 49% in Manoli et al. [6]. However, participants in those two studies had less restrictions placed on their activities while also having a longer time interval between handwashing and sample deposition, which would have likely increased the accumulation of foreign DNA on their hands.
This study also provided insights into the quantity of DNA deposited, with an observed range of 0-12.68 ng. It was rare (1.5%) for touch events to deposit more than 2 ng of DNA; the majority (92%) of touch events deposited <0.5 ng of DNA ( Figure 6). These results The majority (87%) of the DNA profiles generated in this study either had no detectable alleles or were uninterpretable mixtures.
Only 13% of the DNA profiles recovered from the touch events could be interpreted, and the majority (95%) of these could be matched to the person who had held the plastic tubes. This suggested that when a reportable DNA is recovered under similar circumstances, it would almost certainly be a result of primary transfer.
In the remaining instances (4.9% of interpretable profiles, 0.6% of all samples), the recovered DNA profile was of a different person while the participant who had held the tube was excluded as being a contributor-the result of secondary transfer. The study by Goray et al. [7] reported that 2.9% of their samples yielded profiles which were likely the consequence of secondary transfer with the depositor being excluded, much higher than the 0.6% in this study. The results were not unexpected since participants in this study deposited their DNA 15 min after handwashing, whereas participants from the 2016 study had 2-5 h between depositions, and had few restrictions on handwashing and activities, giving them more opportunities to accumulate foreign DNA on their hands [7]. Further analysis of our data showed that the foreign DNA profile in our study could mostly be matched to a coworker or family member (data not shown). The data from this study can thus be applied to provide estimates for touch events taking place under similar circumstances such as depositions occurring on a nonporous material held for approximately 10 s at approximately 15 min after handwashing.

| CON CLUS ION
This study confirms that while shedder status does exist, it is not necessarily static. While it may remain the same in a subset of individuals, it is not so for others and can gradually change over time.
The propensity of a person to deposit DNA on a surface is found to be associated with one's gender, the number of items touched, and the nature of activities. Uncovering more of such activities may enable us to better elucidate the mechanisms that relate to DNA deposition. The study showed that in 29% of touch events, no DNA from the participant is detected and the probability of depositing more than 2 ng of DNA in a touch event is low (1.5%). In 0.6% of touch events, secondary transfer could result in the participant being excluded as a contributor of the DNA profile obtained, and another person being included. Finally, the study provides support for the need to standardize shedder studies in the community and for refinements to the current three-category classification system for better assignment of shedder status.

ACK N OWLED G M ENTS
The authors thank Nurul Insyirah Binte Ishak, Marlene Abdul Mugni, Chew Mee Hui, Aw Zhen Qin, and Lee Jun Yu for their assistance in the execution of the experimental plans. We also express our appreciation to all volunteers for their participation in this study.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no competing interests, financial, or otherwise.