Impact of the Human Microbiome in Forensic Sciences: a Systematic Review

Numerous studies relate differences in microbial communities to human health and disease; however, little is known about microbial changes that occur postmortem or the possible applications of microbiome analysis in the field of forensic science. The aim of this review was to study the microbiome and its applications in forensic sciences and to determine the main lines of investigation that are emerging, as well as its possible contributions to the forensic field. A systematic review of the human microbiome in relation to forensic science was carried out by following PRISMA guidelines.

ABSTRACT

Numerous studies relate differences in microbial communities to human health and disease; however, little is known about microbial changes that occur postmortem or the possible applications of microbiome analysis in the field of forensic science. The aim of this review was to study the microbiome and its applications in forensic sciences and to determine the main lines of investigation that are emerging, as well as its possible contributions to the forensic field. A systematic review of the human microbiome in relation to forensic science was carried out by following PRISMA guidelines. This study sheds light on the role of microbiome research in the postmortem interval during the process of decomposition, identifying death caused by drowning or sudden death, locating the geographical location of death, establishing a connection between the human microbiome and personal items, sexual contact, and the identification of individuals. Actinomycetaceae, Bacteroidaceae, Alcaligenaceae, and Bacilli play an important role in determining the postmortem interval. Aeromonas can be used to determine the cause of death, and Corynebacterium or Helicobacter pylori can be used to ascertain personal identity or geographical location. Several studies point to a promising future for microbiome analysis in the different fields of forensic science, opening up an important new area of research.

INTRODUCTION

Numerous studies state that there is a relationship between a microbiome dysbiosis and the development of various pathologies (1–5). Studies on the microbiome point to its great potential in a clinical setting, enabling high-precision personalized medicine to be developed in the near future and offering preventive, diagnostic, and therapeutic measures (6–8). Among studies that have pointed to the great diversity of the microbiome, the Metagenomics of the Human Intestinal Tract project (9) reported the presence of 3.3 million nonredundant genes in the human intestinal microbiome alone. Due to the functional redundancy of different microbial systems, that is, the ability of different microbiomes to perform similar actions in different ways, some authors state that there is no single healthy microbiota composition, since microbial communities that involve a health condition could differ from person to person (10).

The human microbiota is a highly dynamic system that can be affected by a multitude of factors, including the spatial and temporal components, which are critical because they are associated with factors such as age, sex, life habits, geographical location, occupation, or interaction with other people (11, 12). From the forensic point of view, microorganisms are important for their role in the process of cadaveric decomposition (13–15). During the agonal period, for example, microorganisms may enter the body and subsequently be useful for diagnosis of the cause of death (16, 17). However, on many occasions, the microorganisms that cause fatal infections are not identified at the time of death (18).

On the other hand, there is increasing evidence that humans have an extremely diverse microbiome that can be useful in determining ethnicity, country of origin, and even personal identity (19, 20). Similarly, the composition of the microbiome present in the environment can be a useful indicator of geographical origin or as a means to link people, animals, or objects to each other or to a specific location (20, 21). Therefore, microorganisms can provide evidence in many different forensic scenarios, including investigations into sexual assault when there is no other type of evidence available (22).

Given the enormous forensic potential presented by microbial analysis, there is a need to develop standardized operating procedures for the collection, analysis, and interpretation of microbial evidence, as well as to create solid and complete databases for full implementation in the forensic context, thereby allowing the use of microorganisms as auxiliary evidence in criminal cases to clarify the causes of death, to provide identification and geolocation information, or to estimate the postmortem interval (PMI), among other uses (14, 17, 19, 20).

This contribution offers a systematic review of the literature on the microbiome in order to identify the main lines of research that are emerging and the possible contributions or limitations of such studies in the forensic sciences field.

SYSTEMATIC REVIEW

The methods used for this systematic review (covering 2009 to June 2020) were developed by reference to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (23) for studies published in accordance with the methods detailed in the Cochrane Handbook for Systematic Reviews of Interventions (77), such as reference 24. The protocol for this systematic review was registered with the International Prospective Register of Systematic Reviews (PROSPERO) prior to commencement.

Inclusion criteria.

All studies exploring the human microbiota in human forensic science in subjects aged 0 to 89 years old were included. The articles were chosen according to two main inclusion criteria: (i) application of the microbiome in forensic sciences and (ii) microbiome of human origin.

Search strategy.

Literature search strategies were developed in collaboration with a health sciences librarian using two scientific electronic databases (Medline and Google Scholar) and keywords.

For the articles included in the review, the key characteristics of the studies were identified: topic discussed, first author, and year. The following keywords and subject heading terms were used: postmortem and/or microbiology, forensic and/or microbiology, postmortem and/or microbiome, forensic and/or microbiome, thanatomicrobiome, sudden death and/or microbiome, and drowning and/or microbiome. The search in the two scientific electronic databases (Medline and Google Scholar) was limited to articles published in English and studies conducted in humans. Two independent reviewers revised titles and abstracts and then full-text publications with reference to the inclusion criteria. Study selection interrater agreement between the two reviewers was calculated as the proportion of positive agreement (25).

Data extraction.

Two independent testers retrieved duplicate data using Microsoft Excel. We checked and compared multiple reports from the same study and extracted them where specific data existed. For all studies that met the inclusion criteria, the following data were extracted: authors, year of publication, geographic location, study population, study design, sample size, age range, gender, ethnicity, method of microbiota analysis, type of bacteria detected at each anatomical site, provenance of the microbiome studied, and main microorganisms found.

Risk of bias assessment.

The risk of bias was assessed for each sample by comparison with the Cohort Research Checklist of the Critical Assessment Skills Program (CASP) (26). The following confounding variables within the CASP checklist were evaluated: sample size, age, gender, population analyzed, and location of the analyzed microbiome. Based on the CASP checklist, study output was graded as “bad,” “fair,” or “good.” The overall quality of the proof was rated as high, moderate, weak, or extremely low (27).

DESCRIPTIVE STUDIES

A total of 4,150 studies were identified in the two scientific electronic databases, PubMed (2,454) and Google Scholar (1,696) (Fig. 1). A total of 3,780 duplicates and nonrelevant studies were eliminated, and 370 studies were reviewed to assess their relevance. A total of 337 studies were excluded by these criteria: (i) reviews (n = 44); (ii) based on nonhuman samples (n = 176); (iii) based on clinical research (n = 88); and (iv) unspecific (n = 29).

FIG 1.

Flow diagram of the systematic review.

Finally, this search strategy identified 33 descriptive studies of microbiome and postmortem interval (n = 8), drowning (n = 4) and sudden death (n = 4), geolocation (n = 4), skin and surrounding microbiome (n = 4), sexual contact (n = 2), and identification (n = 7) that were included in this systematic review.

Risk of bias assessment.

According to the CASP risk of bias assessment, most studies (63.6%) were judged as “good” due to the considered variables, while 36.4% were judged as “poor” or “moderate,” largely due to confounding variables not being considered (Table 1).

TABLE 1.

Risk of bias assessment^a

Study	Address a clearly focused issue	Acceptable cohort recruitment	Exposure accurately measured	Outcome accurately measured	Important confounding factors identified	Important confounding factors accounted for	Precise results	Believable results	Results fit with other available data	Overall quality score
Postmortem interval determination
Adserias-Garriga et al. (41)	+	−	−	−	−	−	+	+	+	Moderate
DeBruyn et al. (76)	+	−	−	−	−	−	+	+	+	Moderate
Bell et al. (45)	+	−	−	−	−	−	+	+	+	Moderate
Pechal et al. (43)	+	+	+	+	+	+	+	+	+	Good
Pechal et al. (44)	+	−	−	−	−	−	+	+	?	Moderate
Javan et al. (14)	+	+	+	+	+	+	+	+	+	Good
Hauther et al. (46)	+	+	+	+	+	+	+	+	+	Good
Can et al. (15)	+	−	−	−	−	−	+	+	+	Moderate
Hyde et al. (42)	+	−	+	+	+	+	+	+	+	Good
Death by drowning
Uchiyama et al. (17)	+	+	+	+	+	+	+	+	+	Good
Rutty et al. (16)	+	+	−	−	−	−	+	+	+	Moderate
Huys et al. (28)	+	+	+	+	+	+	+	+	+	Good
Kakizaki et al. (29)	+	+	+	+	+	+	+	+	+	Good
Cause of sudden death
Leong et al. (51)	+	−	−	−	−	−	+	+	−	Poor
Highet et al. (52)	+	+	−	−	−	−	+	+	+	Moderate
Praveen and Praveen (30)	+	−	−	−	−	−	+	+	+	Moderate
Prtak et al. (18)	+	+	+	+	+	+	+	+	+	Good
Geolocation
Walker et al. 2019 (20)	+	+	+	+	+	+	+	+	+	Good
Brinkac et al. (54)	+	+	+	+	+	+	+	+	+	Good
Nagasawa et al. (33)	+	+	+	+	+	+	+	+	+	Good
Escobar et al. (55)	+	+	−	−	−	−	+	+	+	Moderate
Determination of personal belongings
Neckovic et al. (57)	+	+	+	+	+	+	+	+	+	Good
Phan et al. (58)	+	+	+	+	+	+	+	+	+	Good
Kodama et al. (21)	+	+	−	−	−	−	+	+	+	Moderate
Lax et al. (59)	+	+	+	+	+	+	+	+	+	Good
Determination of sexual contact
Williams et al. (22)	+	+	+	+	+	+	+	+	+	Good
Williams et al. (60)	+	−	+	+	+	+	+	+	+	Good
Human identification
Richardson et al. (19)	+	+	+	+	+	+	+	+	+	Good
Schmedes et al. (34)	+	+	+	+	+	+	+	+	+	Good
Schmedes et al. (35)	+	+	+	+	+	+	+	+	+	Good
Wilkins et al. (61)	+	+	+	+	+	+	+	+	+	Good
Park et al. (31)	+	+	+	+	+	+	+	+	+	Good
Leake et al. (32)	+	−	−	−	−	−	+	+	+	Moderate

^aData based on CASP-based risk of bias assessment. ?, this variable was unable to be assessed.

Participants were recruited from few geographic regions, making it difficult to generalize beyond these regions. Overall, the quality of the literature was good.

Laboratory methods.

The methods used to evaluate the microbiome varied between studies (Table 2). Most studies (24/33) used 16S rRNA gene sequencing to detect a wider range of bacteria. Five studies (18, 28–31) used culture for detection of the microbiome. Seven studies used PCR (16–18, 28, 32–34), and two studies used only whole-metagenome sequencing (20, 35).

TABLE 2.

Analysis of the techniques used for microbiome analysis

Study	Analysis technique
	Culture	PCR	16S rRNA gene sequencing	Whole-metagenome sequencing
Postmortem interval determination
Adserias-Garriga et al. (41)			✓
DeBruyn et al. (76)			✓
Bell et al. (45)			✓
Pechal et al. (43)			✓
Pechal et al. (44)			✓
Javan et al. (14)			✓
Hauther et al. (46)			✓
Can et al. (15)			✓
Hyde et al. (42)			✓
Death by drowning
Uchiyama et al. (17)		✓
Rutty et al. (16)		✓
Huys et al. (28)	✓	✓
Kakizaki et al. (29)	✓		✓
Cause of SIDS
Leong et al. (51)			✓
Highet et al. (52)			✓
Praveen and Praveen (30)	✓
Prtak et al. (18)	✓	✓
Geolocation
Walker et al. (20)				✓
Brinkac et al. (54)			✓
Nagasawa et al. (33)		✓
Escobar et al. (55)			✓
Determination of personal belongings
Neckovic et al. (57)			✓
Phan et al. (58)			✓
Kodama et al. (21)			✓
Lax et al. (59)			✓
Determination of sexual contact
Williams et al. (22)			✓
Williams et al. (60)			✓
Human identification
Richardson et al. (19)			✓
Schmedes et al. (34)		✓
Schmedes et al. (35)				✓
Wilkins et al. (61)			✓
Park et al. (31)	✓		✓
Leake et al. (32)		✓	✓

MICROBIOME ANALYSIS IN POSTMORTEM FORENSIC STUDIES

Postmortem interval determination by microbiome analysis.

The disruption of the immune system and the deterioration of the physical barriers that occur after death allow microbes to proliferate throughout the body (36). Changes in the variability and quantity of the microbiome after death can be used to determine the PMI, which is the main objective of the studies. It is important to note that the bacterial succession that occurs at the various stages of decomposition is affected by the physiological changes that the organism undergoes after death (13, 37).

A total of nine descriptive studies on the determination of PMI from the microbiome have been reviewed (Table 3). As shown in Fig. 2, as decomposition progresses and samples enter the swelling phase, as a consequence of oxygen depletion and the accumulation of gases such as carbon dioxide, methane, or sulfuric acid, the predominant aerobic organisms in the fresh state are replaced by anaerobic organisms. The advanced decomposition stage is characterized by the presence of microorganisms representing the soil, because during the decomposition of the corpse, if it is in the soil and there is vegetation, there is an increase in carbon and nutrients in the soil, which facilitates its proliferation. The dry remains stage is characterized mainly by the presence of spore-forming microorganisms, because their spores allow the rapid colonization of the new ecological conditions.

TABLE 3.

Microbiome analysis in postmortem forensic studies^a

Reference	n	Age	Gender	Population analyzed	Microbiome location
Postmortem interval determination
Adserias-Garriga et al. (41)	3	27–81	W/M	USA	Oral (palate, tongue, inner mucosa of cheek and tooth surfaces)
DeBruyn et al. (76)	4	62–67	W/M	USA	Proximal large intestine (cecum)
Bell et al. (45)	10	17–67	W/M	USA	Cardiac tissue
Pechal et al. (43)	188	18–88	W/M	USA	Ears, eyes, nose, mouth, rectum, and umbilicus
Pechal et al. (44)	2	9–13	W/M	USA	External auditory canal, eyes, nares, mouth, umbilicus, and rectum
Javan et al. (14)	27	17–82	W/M	USA	Brain, heart, liver, and spleen
Hauther et al. (46)	12	51–85	W/M	USA	Intestine
Can et al. (15)	n.i.	n.i.	n.i.	USA	Blood, brain, liver, and spleen
Hyde et al. (42)	2	n.i.	n.i.	USA	Intestine and oral cavity
Death of drowning
Uchiyama et al. (17)	43	<10–80	W/M	Japan	Lung, kidney, liver, and blood
Rutty et al. (16)	20	14–93	W/M	UK	Brain, lung, spleen, and kidney
Huys et al. (28)	93	n.i.	n.i.	USA	Blood and bone marrow
Kakizaki et al. (29)	25	<10–80	W/M	Japan	Blood
Cause of SIDS
Leong et al. (51)	88	0–1	W/M	Australia	Fecal
Highet et al. (52)	154	0–1	W/M	Australia	Intestine
Praveen and Praveen (30)			W/M	USA	Gut flora
Prtak et al. (18)	121	0–2	W/M	UK	Blood and cerebrospinal fluid

^an, number of individuals or samples; n.i., not indicated; W/M, women/men; SIDS, sudden infant death syndrome.

FIG 2.

Representative diagrams illustrating the relationship between the microbiome and the postmortem interval (PMI). (A) Representative diagram of the changes in microbiota during the different stages of human decomposition (13). (B) Representation of microbial communities present before and after the bloat stage in human decomposition (76).

One study analyzed the daily differences in the oral microbial composition (palate, tongue, internal mucosa of the cheek, and dental surfaces) in the different stages of human decomposition to estimate the PMI (Fig. 2A). Different bacterial communities are observable in fresh, bloated, active, and advanced decay and also in the dry remains (38). The entire fresh stage was characterized by indigenous oral microbiome representatives. The predominant families in the bloat stage were Peptostreptococcaceae and Bacteroidaceae, which are mostly oral indigenous representatives, and Enterococcaceae, which is a gut microbiome representative. The translocation and proliferation of Clostridium in postmortem human internal organs is observed in several studies (14, 39). Clostridium species are believed to advance decomposition by breaking down lipids and complex carbohydrates associated with human tissue (40). Clostridium lipases are believed to significantly aid in fat hydrolysis under hot and humid conditions, with oxygen depletion and low redox conditions, while hydrolytic enzymes convert carbohydrates into organic acids and alcohols (40).

In the advanced decomposition stage, the predominant microorganisms found were of the class Gammaproteobacteria and the families Pseudomonadaceae, Alcaligenaceae, and Planococcaceae, which are frequently represented in soil. Finally, the dry remains were characterized by the presence of Bacilli and Clostridia, whose spores allow a rapid colonization of the new ecological conditions (41).

Other authors analyzed the microbiome of the proximal large intestine, revealing that although there was considerable variation between individuals, changes followed a similar path with time in reference 76. They observed how the taxon richness of the bacterial communities increased while the diversity decreased significantly, and they also described how the microbial communities present in the bodies changed over time (Fig. 2B). The same authors observed that levels of Bacteroides and Parabacteroides decreased over time and were significantly and inversely correlated with PMI, with Clostridium being the strongest positive predictor of PMI.

Pechal et al. (43) presented a large-scale evaluation of the postmortem human microbiome to determine if the microbiome in the first hours of death can be correlated with the state of health of the host before death. The authors collected samples with postmortem intervals ranging from less than 24 h to more than 73 h. The results of this study show that there is a strong differentiation between the microbiome present in different anatomical regions, and a microbial sequence can be observed that corresponds to the estimated time after death. Finally, this study also suggests that antemortem microbial communities persist in the first hours after death and may be useful to indicate the state of human health, although the authors point out that the value of the past-postmortem microbiome at 48 h of death can become more limited, and this time range could be reduced if there are extreme temperatures that affect the proliferation of specific microbial taxa.

Another study was carried out with two bodies that were found in a freezer (44) (Fig. 3A). Samples from the external ear canal, eyes, nostrils, mouth, navel, and rectum were analyzed when the bodies were completely frozen, when they were partially frozen (at 24 h), and when they were completely thawed (48 h later). The most notable increase in microbial diversity during the thawing process was documented in the nostrils, eyes, and rectum. An increase in the richness and diversity of six families was observed for which an increase in relative abundance was determined as the bodies passed from the frozen to the thawed state. However, two families decreased in relative abundance during the thawing period.

FIG 3.

Representative scheme of microbiota found after drowning according to the type of water in the external auditory canal, eyes, nares, buccal cavity, umbilicus, and rectum (44). (A) Differences between the microbiomes of frozen and thawed cadavers. (B) Differences in the microbiome according to gender in brain, heart, liver, and spleen (14). (C) Differences in the microbiome according to type of water during drowning. The percentages detected in blood, lung, and other closed organs are shown (17).

In one study, a total of 66 samples from the brain, heart, liver, spleen, blood, and oral cavity were analyzed, and microbial changes were seen to be dependent on the PMI and sex of the corpse (14). In female cadavers Pseudomonas and Clostridiales predominated, while male cadavers had a high abundance of Clostridium, Clostridiales, and Streptococcus. The most abundant in women was Pseudomonas, while Rothia was identified only in men (Fig. 3B).

Similarly, Bell et al. (45) examined the postmortem microbiomes of the cardiac tissues of 10 cadavers with a postmortem interval of 6 to 58 h. The investigation revealed that the cardiac microbiomes of male and female cadavers are different. The genera Streptococcus and Lactobacillus were found exclusively in men. The study also revealed a higher prevalence of Pseudomonas and Clostridium in women. Thus, this study provides data demonstrating that the microbiome has a discriminatory power for sex differences in postmortem heart samples.

Another study of changes in postmortem intestinal microbial populations concluded that Bacteroides and Lactobacillus could be used as quantitative indicators of PMI (46).

A study of the postmortem microbiome analyzed different tissues (blood, brain, liver, and spleen) and blood. It concluded that facultative anaerobic bacteria predominate in corpses with a short PMI and obligately anaerobic bacteria predominate in corpses with a longer PMI (15). In another study carried out on the bacterial species associated with human decomposition in the intestine and oral cavity, but focusing on the initial and final time points of the swelling stage, the authors emphasized that no definitive conclusion could be reached regarding changes in the structure of the community over time with the data set they analyzed (42).

Death by drowning.

Drowning is the usual cause of death for most victims recovered from watery environments (47). Determination of this type of death is normally based on pathological findings but is sometimes complicated when the typical signs of drowning are not obvious (48).

Diatom analysis can provide useful information for estimating the type and amount of water aspirated, as long as the diatom density is high enough (48). The presence of diatoms in closed organs (or bone marrow) generally suggests that the victim had entered the water while still alive. However, many diatoms aspirated into the lungs cannot enter the bloodstream because they are larger than the diameter of the alveolar capillaries (49). For this reason, several studies have explored the possibility of using the smallest aquatic microbes that can easily enter the blood circulation and that are detectable even in putrefied victims as markers that allow the detection of death by drowning (Table 3).

In one study, a triple PCR method with TaqMan probes was used to simultaneously detect eight species of bacterioplankton, which are dominant in the blood of drowned bodies, with the aim of confirming or ruling out drowning as the cause of death (17). The authors compared corpses drowned in different types of water, and the genus Aeromonas (A. hydrophila and A. salmonicida) was mainly observed in victims who had drowned in freshwater. They were found in lung samples (100%), blood (100%), and closed organs (85%) (Fig. 3C). In all the lung samples taken from victims discovered near estuaries, both seawater (Vibrio and Photobacterium) and freshwater bacteria were detected. In victims drowned in saltwater, the genera Vibrio and Photobacterium were detected in all lung samples, 90% of blood samples, and in 50% of the organ samples taken.

Using the methodology developed by Uchiyama et al. (17), samples of brain, kidney, spleen, and lungs from 20 bodies found in freshwater, brackish water, and salt water were analyzed by Rutty et al. to confirm the diagnosis of death by drowning. In the same study, a water sample from each of the places where the bodies had been found was analyzed as a control sample. The authors concluded that the PCR method used provides a fast, high-performance (4 to 6 h) backup test for a drowning diagnosis that could be easily applied (16).

Other authors (28) combined culture in selective ADA (ampicillin dextrin agar) enrichment medium with a specific PCR for Aeromonas species and evaluated the practical benefits of bone marrow harvesting using aseptic postmortem puncture to determine death by drowning. They analyzed three cases of drowning and 90 control cases of samples whose death diagnosis was other than drowning. Aeromonas species were detected in the lung, blood, and bone marrow samples from the three drowned bodies, while in the 90 cases used as controls all the samples were negative. This study confirms how the presence of Aeromonas species in bone marrow samples can be used as a marker to help diagnose drowning deaths.

Finally, a study analyzed the species of bacteria present in the blood samples of 25 corpses, of which 5 had been recovered from seawater, 10 from freshwater, 6 from estuaries, and 4 from dry land as nondrowned controls (29).

In the two victims submerged in freshwater but whose autopsy and diatom test findings excluded drowning as the cause of death, the results of the bacteriological tests did not indicate that the examined species entered the bloodstream. The freshwater bacterioplankton (Aeromonas species) was identified in the blood of the 8 victims who had drowned in freshwater, while marine bacterioplankton (Vibrio, Photobacterium, and Listonella) was found in the blood of the 4 victims who had drowned in seawater. Bacterioplankton was not detected in the 4 victims found on land, whose cause of death was not drowning. This study suggests that bacteria indigenous to the discovery sites do not easily invade the blood of corpses, and as it would be difficult to contaminate blood during autopsy or sampling, the authors concluded that bacteriological tests can be useful in those cases in which the density of diatoms is low. The authors were also of the opinion that the detection of bacterioplankton in a blood sample may support the conclusion of death by drowning.

Cause of SIDS.

Sudden infant death syndrome (SIDS) is defined as the sudden and unexpected death of an infant under 1 year of age, with the onset of the fatal episode apparently during sleep and which remains unexplained after extensive investigation (50). The determination of the cause is important in both forensic medicine and pediatrics, because it is the main cause of death for babies in the first year of life, and only in 20% of cases is a specific cause of death identified (18).

Several articles that analyze the relationship of the microbiome with death from SIDS are analyzed below (Table 3). Leong et al. (51) observed the composition of the microbiome in 44 cases of SIDS and in 44 healthy infants, where age, sex, and mode of feeding did not differ significantly between the two study groups. The authors found no significant difference in microbial diversity between SIDS cases and the controls. They also carried out specific tests for the detection of pathogens that had previously been related to SIDS (Clostridium difficile, Escherichia coli, and Staphylococcus aureus) and also found no significant difference between SIDS and healthy cases. However, a positive correlation was observed between the species richness of the samples analyzed and age in both groups.

Highet et al. (52) analyzed the intestinal contents of 52 SIDS cases and 102 fecal control samples similar in both age and gender. In all cases, Clostridium innocuum, Clostridium perfringens, Clostridium difficile, Bacteroides thetaiotaomicron, and Staphylococcus aureus were analyzed. The authors described a statistically significant increase in Clostridium difficile, Clostridium innocuum, and Bacteroides thetaiotaomicron in samples with SIDS compared with the controls when both groups were analyzed.

Furthermore, they observed that the SIDS samples showed a significantly more frequent dual colonization by Clostridium perfringens and Clostridium difficile than the healthy cases (17% versus 5%). Triple colonization by Clostridium innocuum, Clostridium perfringens, and Clostridium difficile was also significantly more frequent in SIDS samples (15% versus 3%).

They observed that SIDS babies who usually slept in the prone position had a higher frequency of colonization by Staphylococcus aureus (82%) than babies who usually slept in the lateral position (9%) or in the supine position (9%). Furthermore, in babies found in the prone position, Staphylococcus aureus was isolated from sterile sites (58%). For all these reasons, the authors concluded that the differences between the microbiome of babies who suffered from SIDS and that of healthy babies, while it remains to be shown whether they are critical differences that can lead to death or not, should be taken into account, since they may increase the susceptibility to infection and, consequently, to SIDS.

Another study (30) proposed a new hypothesis that the infant gut microbiome plays an important role in SIDS, during the period that is critical to both gut flora development and vulnerability to SIDS, by modulating the brainstem serotonergic system through the bidirectional microbiome-gut-brain axis, thereby “tilting the balance in favor of successful autoresuscitation during a sleep-related adverse autonomic event.”

Finally, the study of Prtak et al. (18) analyzed autopsies of SIDS cases in infants under 2 years of age, looking at microbiological and virologic evidence. They found potential pathogens in 59% of cases, postmortem microbiota and microbes that were not potentially pathogenic in 73% of cases, and 10% negative cases. The results of this study suggest that infection plays a key role in SIDS and highlight the benefit of microbiological investigations.

MICROBIOME ANALYSIS IN HUMAN IN VIVO FORENSIC STUDIES

Geolocation.

Studies carried out on the human microbiome to date have revealed the variations that exist in the microbial ecology of different populations around our planet (53). These differences may be due to factors such as the level of industrialization of each geographic region and/or to the lifestyle habits of each population. These facts increase forensic interest in finding microbial signatures that characterize each geographic region. In this review, we found four recent articles describing how the microbiome is related to geolocation (Table 4).

TABLE 4.

Microbiome analysis in human in vivo forensic studies^a

Reference	n	Age (yr)	Gender	Population analyzed	Microbiome location
Determination of human geolocation
Walker et al. (20)	293	n.i.	n.i.	New Zealand, USA, Nigeria, Portugal, Chile, Japan, and Colombia	n.i.
Brinkac et al. (54)	21	n.i.	n.i.	USA	Scalp hair and pubic areas
Nagasawa et al. (33)	144	18–89	W/M	China, South Korea, Taiwan, Thailand, Afghanistan, and the Philippines	Intestine
Escobar et al. (55)	126	19–68	W/M	USA, Spain, France, Denmark, South Korea, and Japan	Intestine
Determination of personal belongings
Neckovic et al. (57)	6	n.i.	n.i.	Australia	Hand
Phan et al. (58)	45	21–70	W/M	Australia	Hand
Kodama et al. (21)	88	24–69	W/M	USA (Hawaii)	Medical devices, pipes, manipulated objects, book, drinking container, glasses, identification card, armrest, steering wheel, computer devices, remote controls, mobile phones, door handles, switches, water taps, purse, razors, lighters, keys, cosmetics, combs, dumbbells, harmonica, nail clippers, and watch
Lax et al. (59)	91	n.i.	n.i.	Canada and USA	Mobile phones, shoes, and floor in the area
Determination of sexual contact
Williams et al. (60)	43	21–70	W/M	USA	Pubic hair
Williams et al. (60)	6	21–70	W/M	USA	Pubic hair
Human identification
Richardson et al. (19)	37	n.i.	W/M	USA	Skin
Schmedes et al. (34)	72	n.i.	W/M	USA	Skin
Schmedes et al. (35)	12	n.i.	W/M	USA	Skin
Wilkins et al. (61)	19	n.i.	n.i.	China	Skin and surfaces of objects
Park et al. (31)	15	n.i.	W/M	South Korea	Hand
Leake et al. (32)	2	25–69	M	Switzerland	Saliva

^an, number of individuals or samples analyzed; n.i., not indicated; W/M, women/men.

In one study, the authors analyzed the relative abundance of bacterial species in corpses from 12 cities of 7 countries and concluded that there was a clear difference in the most common species in each of the studied cities (20).

Another study analyzed the microbiome present in both the scalp and pubic hair of adults who lived in Maryland, California, and Virginia (54), finding that the microbial communities differed in composition between the different geographical locations analyzed. Peptoniphilus and Staphylococcus differed in abundance when samples from Maryland and California were compared in the case of both hair samples. It was also observed how comparisons between scalp hair collected in different cities have a greater potential to predict geolocation than pubic hair, a finding of great importance in forensic applications.

Nagasawa et al. (33) developed a method to determine the geographical origin of unidentified corpses by determining the genotype of Helicobacter pylori, a bacterium that is latently present in half of the world's population. The authors did not observe significant differences in the detection rate of H. pylori between the different sampling points of the gastric mucosa, between the causes of death, or the ages of the subjects. Finally, the authors amplified and sequenced the vacA gene from H. pylori, finding how the different genotypes showed specificity for geographic origin. The authors concluded that their results suggest that the H. pylori genome could provide valuable additional information for tracing the geographic origin of unidentified bodies.

In another study (55), the intestinal microbiome of Colombian adults was compared with that of North Americans, Europeans, Japanese, and South Koreans, and the results confirmed that the composition of the intestinal microbiota differed significantly among different populations. For example, the phylum Actinobacteria was present in a higher proportion in Japan, Colombia, and Europe but was practically absent or not found at all in South Korea and the United States. The phyla Firmicutes, Bacteroidetes, and Proteobacteria predominated in the intestinal microbiome of people analyzed in Colombia, while in the other analyzed regions there was a higher proportion of Bacteroidetes and lower proportions of Firmicutes and Proteobacteria. Tenericutes was more frequent in Europe but absent from Japan and in a very low proportion in the other regions. Finally, the verrucomicrobia were not found in either Japan or South Korea but were present in Colombia, Europe, and, to a lesser extent, the United States. Based on these data, the authors concluded that the geographic origin in the studied populations had an impact on the composition of the intestinal microbiota.

Determination of personal belongings.

Humans have personalized skin microbiomes that are generally stable over time and are transferred to the objects with which we interact, generating a microbial signature on personal objects (56). Below, we will review two articles that analyze skin microbial communities as a screening test to associate individuals with locations and objects in their environment (Table 4).

Neckovic et al. (57) carried out a study with the aim of verifying whether the microbiome of one individual could be transferred to another individual, to surfaces, and vice versa. Cotton and glass paper surfaces were used in the study. The microbiomes of six participants placed in three pairs were analyzed by analyzing two modes of transfer. The first transfer mode involved the pair of people shaking hands and then rubbing a surface with their right hands. The second transfer mode involved individuals who rubbed a surface with their left hands, exchanged the surface they had rubbed with their partner, and then rubbed the exchanged surface with their left hands. The authors concluded that transfer of the human skin microbiome took place between all pairs of participants, regardless of substrate type or mode of transfer.

Phan et al. (58) investigated whether the microbiome could be used as an indicator of donor characteristics. They analyzed the microbiome of 45 subjects who were asked to touch DNA-free cards with their dominant and nondominant hands. The authors compared the diversity and abundance of bacteria with the characteristics of gender, age, ethnic origin, labor force, home location, sample location, occupation, type of diet, use of humectants, use of hand sanitizers, and use of public transport. Correlations were found between the bacterial profile with gender, ethnic origin, type of diet, and the use of hand sanitizer. Specifically, the absence of Lactococcus indicated a mainly Chinese diet, while the absence of Alloiococcus indicated female gender, Asian ethnicity, and use of hand sanitizer. Tests of the prediction models demonstrated the highest precision for gender estimation, while the prediction of other characteristics showed less success. With these results, the authors conclude that there is a correlation between the presence of certain bacterial species in the donor's hands and the personal characteristics of potential forensic relevance, which shows a new application of the microbiome in forensic science.

On the other hand, Kodama et al. (21) analyzed 88 samples of objects found at 16 different crime scenes to ascertain whether it was possible to associate postmortem skin microbiomes with objects found at the crime scene. The authors used only the microbiome present in the right palm of the decedent. With an average precision rate of 75%, their results confirmed that it was possible to associate the postmortem microbiomes of the corpses and the microbiome present on the objects. The precision also varied according to the objects analyzed, so that various objects could be associated with 100% precision (medical devices, bottles, bongs, manipulated objects, books, drinking containers, glasses, identification cards, automobile armrests, and steering wheels). However, other objects were associated with an accuracy equal to or less than 67% (computing devices, remote controls, telephones, door handles, and light switches). Results for four objects were less than 60% accurate (water taps, purses, razors, and lighters), while for some, such as keys, cosmetics, combs, dumbbells, harmonicas, nail clippers, and clocks, no association could be made.

Furthermore, when these authors studied the structure of the postmortem microbial community during the transit and storage of corpses in the morgue, they showed how the skin microbiomes remained stable in all cases. This microbial stability was also reflected in the similarity observed between the skin microbiomes, personal items, and plastic bags used to transport the body.

On the other hand, in one study (59) an analysis was made of microbiomes present in shoes and mobile phones of two people, and samples were collected every hour on consecutive days. The microbial communities associated with mobile phones were less stable and more variable over time than the communities associated with footwear. Finally, the authors studied the biogeographic influence on the microbiome of the mobile phones and shoes of 89 volunteers from different places in Vancouver, BC (n = 29), Washington, DC (n = 26), and California (n = 34). They observed how the microbial communities on the telephones and the shoes were significantly different for the different cities, so that the analysis of both made it possible to determine from which of the three geographic regions the samples had come.

Determination of sexual contact.

The human microbiome of different regions of the body (intestine, oral, skin, and urogenital) differs in composition, although these microbiome regions are more similar to each other than to the microbiome of other people (11). This potential individuality of the human microbiome suggests that there is some transfer during sexual contact that would allow the human microbiome to be used in investigations of sexual assault when there is no other type of evidence (60). However, before the microbiome can be used in such a forensic context, it is first necessary to address issues such as the stability of the microbiome both on the individual's body and in stored samples, as well as the degree of transfer between individuals. Of the articles reviewed, two analyzed the usefulness of the microbiome in determining the existence of sexual contact between individuals (Table 4).

Recently, Williams et al. (22) demonstrated the stability of the pubic mound microbiome for 6 months. Furthermore, they analyzed microbiome samples from the pubic area and pubic hair from 43 individuals at different times (up to 12 weeks). They observed that more than 77% were represented by Corynebacterium (29.2%), Staphylococcus (21.5%), Propionibacterium (15.4%), and Lactobacillus (11.5%), with Corynebacterium being more abundant in men and Lactobacillus more abundant in women. In addition, they observed that the increased frequency of sexual activity does not necessarily mean greater similarity of the microbiomes.

Furthermore, the authors assessed the forensic potential of microbiome analysis in sexual assault. When the proportion of the woman's microbiome that appeared to be derived from the aggressor was evaluated, this method was able to very accurately predict the expected proportion of the aggressor when only a single suspect was being investigated. When the aggressor was an unknown person other than the known defendants, there was no case of erroneous attribution. When the evaluation was carried out with up to four potential attackers, the method showed solid results. The authors concluded microbiome analysis was useful when there is a small group of potential assailants to confirm that sexual contact occurred or to exonerate a suspect.

Williams et al. (60) analyzed the influence of storage time and temperature on pubic hair kept at room temperature (20°C), refrigerated (4°C), or frozen (–20°C). They observed how the variations due to storage time and temperature were random and had no significant influence on the taxonomic profiles of the samples.

Furthermore, after analyzing pubic hair samples from men and women, they observed how there were significant differences according to gender, reporting the existence of 10 orders that were significantly different (Bacillales, Bacteroidales, Bifidobacteriales, Campylobacterales, Clostridiales, Coriobacteriales, Enterobacteriales, Fusobacteriales, Lactobacillales, and Streptophyta). Bacillales was the only order more abundant in men than in women, while the orders more abundant in women were Bifidobacteriales and Lactobacillales. The authors concluded that, despite the absence of specific microorganisms for one genus, some were more abundant in one genus than in the other, which could allow their distinction.

Human identification.

The personal microbiome as a specific and exclusive signature of an individual can be stable over time, making the characterization of microbiomes potentially applicable to human forensic identification (35, 57). Here, six studies that have demonstrated the potential of using the human microbiome footprint for forensic identification are reviewed (Table 4).

In one study (19), the effect of individual microbial communities in public and private spaces where several people live was analyzed. The study showed that microbial samples associated with skin are useful to link people with the inhabited spaces, and that these microbial signatures were largely stable over a 4-week period.

However, it was seen how the presence of a second individual in the same space can interfere with the classification by acting as a confounding factor. The classification error was linearly correlated with the number of individuals per shared space.

In another study (34), the authors used a targeted sequencing method based on skin microbiome markers developed for human identification. The sequencing panel consists of 286 specific markers for the detection of 22 species belonging to the genera Corynebacterium, Propionibacterium, and Rothia. In this study, 72 samples of skin microbiomes from three body sites were analyzed: foot, hand, and chest. All samples, regardless of body site, were correctly assigned to their host with 92% accuracy, leading the authors to propose that the skin microbiome could be used for human identification in future studies.

Schmedes et al. (35) describe a novel approach to assigning skin microbiomes to their donors by comparing two types of studies: an analysis of the presence or absence of Propionibacterium acnes and an analysis of the diversity of species-specific markers (Corynebacterium aurimucosum, Corynebacterium jeikeium, Corynebacterium pseudogenitalium, Corynebacterium tuberculostearicum, Micrococcus luteus, Propionibacterium acnes, Propionibacterium granulosum, Pseudomonas species, Rothia mucilaginosa, Staphylococcus epidermidis, Malassezia globosa, and Propionibacterium sp. strain P101A). The authors found that the diversity of species-specific markers was significantly better than that of the study of the presence or absence of Propionibacterium acnes.

This same study was able to accurately identify individuals from the stable characteristics associated with skin microbiomes for a period of up to almost 3 years. The features described in this study provide the preliminary basis for the future development of a robust and reproducible method for profiling skin microbiomes for human forensic identification.

Wilkins et al. (61) carried out a study with the intention of associating the skin microbiome with the places of domestic residence. The taxonomic composition of the surface samples confirmed that most of the microbiota on domestic surfaces originated from the skin of the occupants. The most abundant family in all the samples was Moraxellaceae, dominated by the skin-colonizing genus Acinetobacter.

Among the 10 most abundant families on surfaces were the Staphylococcaceae, Micrococcaceae, Corynebacteriaceae, and Streptococcaceae, all associated with human skin. However, there were also abundant populations of families most probably derived from environmental sources, such as soil and vegetation, including Sphingomonadaceae, Methylobacteriaceae, Pseudomonadaceae, Rhodobacteraceae, and Xanthomonadaceae.

The authors also observed that most of the taxonomic units analyzed persisted on the skin or surfaces for a certain period, after which taxonomic units were indistinguishable. For all these reasons, the authors concluded that although microbiota traces have a potential forensic value, they are not static and therefore are degraded in a way that eliminates their useful characteristics for identifying people.

Park et al. (31) analyzed the diversity of the microbial communities that inhabit the palms of 15 individuals and evaluated their potential for human identification. The authors point out how the genus Staphylococcus was detected in all of the participants and Micrococcus and Enhydrobacter were detected in the majority of the participants (87% and 80% of the cases, respectively). The species with the highest proportion of Staphylococcus was Staphylococcus epidermidis (14 subjects), known as one of the most abundant skin bacteria. This species was followed by S. capitis subsp. capitis (11 subjects), S. warneri (9 subjects), S. hominis subsp. hominis, and S. hominis subsp. novobiosepticus (8 subjects). Furthermore, Micrococcus species, M. yunnanensis in particular, were commonly present (11 subjects).

The species that showed personal variations were Oceanobacillus caeni (1 subject), Paracoccus sanguinis (1 subject), Enterobacter aerogenes (1 subject), and Corynebacterium striatum (1 subject). With these results, the authors point out that some minor species were unique to specific individuals and, therefore, exhibited potential for personal identification. They also highlight that the main species can be applied as molecular biological markers at the subspecies level, especially Staphylococcus species that showed distribution in all of the participants. This is why the authors see a high potential of the cutaneous microbiome of the palm of the hand for personal identification.

Finally, in another study (32), the potential of the salivary microbiome to differentiate individuals was analyzed at different times. The samples were dominated by Firmicutes, Proteobacteria, Actinobacteria, Bacteroidetes, and Fusobacteria, and the authors concluded that it is possible to use the salivary microbiome to distinguish two people.

FUTURE CHALLENGES IN STUDYING HUMAN MICROBIOME IN FORENSIC SCIENCES

In this systematic review, the main results were obtained from recent studies that attempt to relate the microbiome to different aspects concerning forensic science, such as the determination of the postmortem interval, death by drowning or sudden death, geographical location, relationship with the environment, sexual contact, or human identification. As a whole, the studies serve to evaluate the potential and limitations of using the microbiome as a forensic tool.

Some specific investigational areas of the forensic application of the microbiome are still underdeveloped, but they are the beginning of a promising future due to their practical utility for the resolution of forensic cases. Systematic review is an essential tool to synthesize the available scientific information, increase the validity of the conclusions of individual studies, and identify areas of uncertainty where research is necessary.

Developments in DNA sequencing techniques have made it possible to identify human microbial communities, and the influence of the microbiome on the human body has been described by numerous authors (62–64). Much information is available on the microbiome in the clinical field, but there is now growing interest in its possible application in the field of forensic science (19, 41).

Failure of the immune system and physical barriers after death allow microorganisms to proliferate throughout the body and to colonize internal fluids and organs (36). Numerous studies have shown that during decomposition there is a predictable colonization pattern that allows the PMI to be determined (14, 15, 41, 42, 52). All of these studies develop models to estimate the PMI from the distribution of bacterial taxa, linked to the skin, oral cavity, abdomen, and fluids and internal organs.

It is important to consider that the bacterial succession that takes place in the different stages of decomposition is influenced by the physiological changes that the organism undergoes after death (13, 37). The decomposition of organic tissues begins with cellular autolysis by hydrolytic enzymes that result in the release of carbohydrates, proteins, minerals, and fats from cellular structures. At this point, the endogenous bacterial communities are, as would be expected, most abundant in the fresh stage of decomposition, and these organisms, usually aerobic, cause oxygen depletion (65, 66).

The availability of oxygen seems to be an important factor for the bacterial changes that can be observed during the different stages of decomposition (67, 68). In effect, oxygen depletion and the accumulation of gases, such as carbon dioxide, methane, and sulfuric acid, that occurs in the swelling phase favor the proliferation of anaerobic organisms that take advantage of these physiological changes (69). The accumulation of these gases occurs mainly in the abdomen, and it has been described that in the bloating stage, the endogenous communities of the intestinal microbiome colonize other areas of the body, such as the oral cavity (42).

When the pressure of the gases increases, the natural fluids escape through the natural orifices (nose, mouth, and anus) and can cause the skin to break, leading to the decomposition stage in which a large amount of mass is lost both by the release of these fluids and by the proliferation of the larvae in the putrescent tissues, which is why this stage is characterized by the presence of microorganisms related to myiasis (40). In the advanced decomposition stage, as a consequence of the loss of mass, the decomposition ceases, and if the corpse is in the soil and there is vegetation, there is an increase in carbon and nutrients in the soil, which is why this stage is characterized by the appearance of microorganisms representing the soil. Finally, the dry remains stage is mainly characterized by the presence of spore-forming microorganisms, because their spores allow a rapid colonization of the new ecological conditions (40, 42).

However, bacterial succession depends not only on the organs, tissues, or fluids but also on other variables, such as seasonal variations, temperature, or location of the body, while some studies demonstrate there are also variations in colonization patterns that depend on the sex of the individual (14). Some authors (44) also indicate the need to study how the conditions under which bodies are kept (frozen, burned, or embalmed) affect the microbiome.

The studies published on estimating the PMI based on the microbiome represent an important start when creating a catalog that includes the main microbial contributors and their evolution in the different stages of human decomposition. However, there is still a long way to go, since most studies are located exclusively in the United States and microbiomes vary depending on geographic location, even between regions within the same country (41, 42, 46). Therefore, it would be useful to create databases on bacterial evolution in the decomposition process in different geographic regions.

Future studies should study how the structure of the bacterial community changes as a function of time (41), from the fresh stage to the stage of skeletal remains and from the perspective of microbiology, entomology, and chemistry (42).

The microbiome has also been shown to be useful in determining drowning as a cause of death as well as for determining whether drowning occurred in seawater, freshwater, or brackish water (16, 17, 28, 29). The studies we have reviewed analyze the microbiome present in internal organs, blood, and bone marrow, since these media are generally sterile before drowning. Furthermore, the predominance of some bacterial species that reach these organs and fluids with the aspiration of water during drowning would interrupt the proliferation of other bacteria that might invade or contaminate the bodies after death (29).

Along the same lines, some authors affirm that the native bacteria of the discovery sites do not easily invade the blood of the corpses, and it is difficult for them to contaminate the blood during the autopsy or sampling, even when the corpses have suffered significant injuries (16, 29). This represents an important advantage of the microbiome compared to other conventional drowning diagnostic methods, where aseptic sampling techniques remain the most widely used.

On the other hand, and despite what might be expected for victims discovered after long periods of time, it has been seen how the microbial evidence present in blood does not disappear due to the lack of nutrients or the accumulation of waste substances, but, as described in reference 16, it is possible to confirm death by drowning in corpses in an advanced state of decomposition by detecting certain microorganisms.

Therefore, future studies with a greater number of submerged bodies, both drowned and nondrowned, should be carried out to standardize valid detection methods for the diagnosis of death by drowning based on the microbiome.

Regarding the determination of the cause of death in cases of SIDS, although the pathogenic mechanism underlying the condition is still unclear, among the most prominent hypotheses are those pointing to infections and sepsis, as some studies suggest based on bacteria found in blood samples and tissues from unexpectedly dead babies (18, 52, 70, 71). Another hypothesis is that death caused by SIDS is related to transient bacteremia with no detectable histological changes (18).

It has also been proposed that an altered physiology as a consequence of dysbiosis of the intestinal microbiome could contribute to SIDS, although the data provided by reference 51 contradict this theory, with the authors stating that SIDS is not associated with a substantial change in the intestinal microbiology. However, the high percentage of infections observed in the pediatric population that dies as a consequence of SIDS, and the presence of potentially pathogenic organisms in many of these cases, confirm the importance of conducting bacteriological, viral, and toxicological investigations in all SIDS cases in a multidisciplinary approach (18).

Furthermore, knowledge of the specific composition of host and environment microbiomes can help determine the geographic origin of samples, since microbial communities differ in composition and function according to geographic location and even between different cities in the same country (54). Different strains of Helicobacter pylori can be linked to specific geographic settings (33), and some studies have also shown that a microbial signature can be associated with geographic locations in the country of origin, as can the composition of taxa present on the 16S RNA gene (20, 54).

A possible limitation of using the microbiome to determine geographic provenance is that microbial indicators associated with location can vary by interacting with new environments or by sudden changes in a person's lifestyle, such as diet or disease. To evaluate whether the microbiomes are robust in the face of such changes, it would be convenient to carry out longitudinal studies that evaluate these variables.

Along the same lines, some studies have evaluated the potential of the microbiome to reveal whether a particular person has touched an object or has recently been in a specific space. In this sense, it has been shown how the transfer of the microbiome from the skin of the hand can occur between individuals who do not live together through contact with objects shared between both individuals, which opens the possibility that the study of the microbiome can be used to associate individuals with other individuals or surfaces with which they have interacted (57). In addition, some authors also affirm that the study of the cutaneous microbiome of the hand is not only useful as a tool to link a subject with a surface with which it has been in contact but also has a high potential to find out features of a subject useful in research forensics, such as sex or ethnicity (58).

It has been described how it is possible to associate, with a high degree of precision, the microbiome present on some objects with the cutaneous microbiome of the individuals who interacted with them (21, 59).

It has also been seen how the microorganisms present in a specific soil can often determine the microbiome present on the shoes of individuals who walk through this soil (59). This new technique for associating a person with an object or location represents an important advance in forensic science.

Some studies mention the potential of analyzing the microbiome of the pubic mound area as a tool for determining whether there had been previous sexual contact, since the microbiome is highly individualized and characteristic of gender (60). The microbiomes of couples with sexual contacts also tend to be more similar to each other than to those of unrelated people, although how long the contact must be for the transfer to occur has not been determined, which suggests that any transfer during a single sexual encounter, such as rape, is unlikely to be detected (22).

Therefore, it is necessary to undertake studies to directly evaluate the transfer probabilities associated with sexual assaults and to evaluate how long any resulting mixture of microbiomes is maintained. Even so, the potential to identify a specific individual from microbial fingerprints obtained from different parts of the body, even over long periods of time, has also been demonstrated in several studies (32, 34, 35).

Some studies describe how cutaneous microbiomes can be accurately grouped according to the host from which they come, although the precision decreases if the samples are collected at different times (34, 35). Although cutaneous microbiomes are not stable and can degrade, making it difficult for them to be used in human identification (61), this is not a problem when studying the salivary microbiome (32). It should be noted that the stability, reproducibility, and sensitivity of microbiome-based tests and other critical factors must be considered to accurately identify microbial DNA profiles for forensic application.

Furthermore, there are very few studies that have evaluated the sensitivity of current technologies to obtain microbiome profiles from samples of limited amounts of biomass, as is often the case in forensic investigations (72). In this respect, standardization proposals exist to optimize the yield of forensic and clinical postmortem microbiology by means of adequate sampling (73) and applying microbiome sampling protocols in postmortem sudden death studies (74). Some authors recently point to the need to include microbiome analysis in routine forensic investigation, including the necessary means as a further resource in the forensic toolkit (75). One final remark is that the introduction of novel techniques in forensic science may also require changes in legislation (Fig. 4).

FIG 4.

Recommendations for future microbiome forensic research.

The microbiome can play a fundamental role in various forensic fields, such as the determination of the postmortem interval, identification, geolocation, association with objects in the environment, sexual contact, and cause of death by drowning and sudden infant death syndrome. Concerning future prospects and actions, there is a need to standardize protocols for the collection of samples of a microbiological nature that allow the results obtained in different investigations to be standardized and optimized. This will allow clear and definite conclusions to be reached so that they can be treated as evidence to help elucidate the postmortem interval, cause of death, or the identity and geographic origin of a victim, among other applications. To consolidate the recent inclusion of the microbiome in forensic analyses, it is necessary to continue related research. For its broader application in forensic science, it would be helpful to develop reliable and robust databases of microbiomes that include metadata associated with humans, such as geographic origin, ethnicity, diet, or social information. In addition, it will be necessary to collect and analyze many more samples to establish reliable standards that are legally valid. It would also be convenient to increase standardized processes for the collection, storage, and analysis of samples to avoid contradictory results due to contamination or alteration of the microbiomes.