Cannabis-Microbiome Interactions in Varied Clinical Contexts: A Comprehensive Systematic Review

Background: As cannabis legalization extends to numerous countries, encompassing both medicinal and recreational applications, understanding its impact on human body is crucial. The endocannabinoid (eCB) system, regulated by naturally occurring (endogenous) and externally derived (exogenous) cannabinoid compounds, plays a pivotal role in the host’s metabolism. This system, in conjunction with the host's immunomodulatory mechanisms, influences the composition of the gut microbiota, resulting in beneficial outcomes for the gastrointestinal (GI) and immunological systems.

This systematic review aims to evaluate the association between cannabis treatment and the gut and oral microbiome, supporting further clinical trials in this area.

A comprehensive literature search was conducted on online platforms such as PubMed, Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library . The search encompassed studies published until July 20, 2022, focusing on adult populations with clinical abnormalities. Only English language studies were included. Identified studies were analyzed, considering predetermined subgroups based on different disease conditions. A random-effects meta-analysis was employed to qualitatively and quantitatively combined and assess the data.

The study comprised 9 studies with 2526 participants, including 2 clinical trials and several observational studies. The research explored the impact of marijuana (MJ) use on the microbiota across various clinical categories, such as HIV infection, pain/inflammation, systemic aspergillosis, obesity, cognitive deficits, and oral diseases. Qualitative analysis of the included studies revealed diverse and condition-specific effects of MJ use on the microbiota, such as decreased microbial diversity and increased cannabinoid excretion.

These findings shed light on the complex effects of cannabis use on the human microbiota, underscoring the need for furture research on the therapeutic potential of cannabis. This review provides valuable insights to guide future investigations in this field. Funding: None Registration ID: PROSPERO 2022 CRD42022354331

Marijuana, Microbiome, Cannabis, THC, Endocannabinoids. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 5 of 20

Marijuana (MJ), or Cannabis sativa, has a long history of use both for therapeutic and recreational purposes. It can also be called cannabis, weed, pot, or dope; all refer to the dried flowers, leaves, stems, and seeds of the cannabis plant (1). The chemical constituents the cannabis consist of a complex mixture of natural cannabinoids containing approximately 500 bioactive detected compounds with 70 different cannabinoids (2). Cannabidiol (CBD), a non-psychoactive cannabinoid molecule, had its chemical structure first determined in 1963, and the psychoactive cannabinoid, δ 9-tetrahydrocannabinol (THC), was then discovered in 1964 (3). People use it in various ways, including via smoking, inhaling, and as cannabis extracts. After alcohol and cigarettes, cannabis is the drug that is used most frequently for psychoactive purposes globally (4). In 2020, the global cannabis user population was over 4%, and nearly 6% among ages (5). Even though there is little evidence that adult patients consume cannabis, its consumption in older populations aged ≥ 50 years has been elevated from 15.1% in 2014 to 23.6% in 2016 since legalization for medical use encouraged former non-users to start using it (6). In the United States, the utilization of cannabis is increasing for medical and recreational purposes while the potential risk is declining (7, 8, 9). Even though there is little knowledge of the dangers and advantages of it, they primarily utilize it to control the symptoms. The endocannabinoid system (ECS) is a ubiquitous modulatory neurotransmission system in the brain, and it consists of endogenous neurotransmitters (also known as endocannabinoids (eCBs)) derived from fatty acids, the enzymes responsible for their breakdown, and cannabinoid receptors (CBRs) to which endogenous and exogenous, plant-based cannabinoids bind (10). Nearly all the effects of cannabinoids are mediated by CB1 and CB2 receptors, which are predominantly expressed in the brain and immune cells, respectively (11, 12, 13). Cannabinoids have a well-established, proven record as anti-inflammatory drugs with many immunosuppressive characteristics (14). It was shown that CBD stimulated myeloid-derived suppressor cells (MDSCs), which inhibited T cell proliferation in vitro and in vivo (15). It has also been demonstrated that THC can stimulate MDSCs without using toll-like receptor 4 (TLR4) and then stimulates regulatory T cells (Tregs) that play a role in their differentiation and functions and secrete immunosuppressive cytokines such as interleukin-10 (IL-10) and transforming growth factor β (TGF-β ) (16, 17). While increasing the legalized countries for cannabis use, more research studies are being explored on its potential therapeutic effects and adverse outcomes. Cross-sectional research of primary care patients indicated that those who use medical cannabis have more possible advantages than those who don't, but there are still adverse effects (18). From a systematic review of randomized clinical trials (RCTs) of medical cannabinoids, the serious adverse events are categorized into respiratory, thoracic, and mediastinal disorders, GI disorders, nervous system disorders, cerebrovascular disorders, general disorders and administration-site conditions, renal and urinary disorders, neoplasm, psychiatric disorders and others (19). Furthermore, it is evident that the most frequent physical health reasons are to manage pain (53%), sleep (46%), headaches/migraines (35%), appetite (22%), and nausea/vomiting (21%), while the most prevalent mental health reasons are anxiety (52%), depression (40%), and PTSD/trauma (17%) (20). A mutually beneficial symbiotic interaction exists between the host and the diverse GI microbial population that can be found here (21). The gut microbiota plays a significant role in host metabolism and is associated with the regulation of the inflammatory status of the host in the gut but is not limited to other organs like the brain (22). The intestinal microbiota affects neurological, endocrine, and immunological networks . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 6 of 20 through the gut-brain axis and the bilateral communication between the central and enteric nerve systems (23, 24). A recent study reveals that the gut microbiome plays a role in anxiety and depression-like behavior (25), and clinical research indicates that these disease situations are characterized by an abundance of pro- inflammatory and short-chain fatty acid (SCFA) producing bacterial species (26). This pathophysiology illustrates that dysbiosis may have positive or negative consequences on the host's inflammatory condition. The positive effects of cannabinoids on the GI and immunological systems are lowering intestinal permeability, controlling intestinal bacteria, and reducing inflammation, according to earlier preclinical studies (27, 28). Numerous active studies on cannabinoids in the field of natural medicines are underway at the moment since there is evidence to support their potential efficacy in treating cardiovascular disease, cancer, and inflammation (29, 30, 31, 32). On the other hand, it was linked to detrimental impacts on both physical and mental health in a dose-response manner, with daily or nearly daily use being linked to worse results (33). The gut-brain axis and bacterial metabolites and products have recently come to light as a mechanism by which intestinal bacteria can influence the physiology and inflammation of the central nervous system (CNS) (34). These mechanisms are dysregulated and then linked to altered blood-brain barrier (BBB) permeability and neuroinflammation during dysbiosis (34). A. muciniphila , one of the numerous bacterial species in the gut, significantly regulates the gut barrier and processes roughly 3–5% of the gut microbiota in healthy humans (35, 36). The regulation of intestinal barrier integrity, effects on immunological modulation and the enteroendocrine system, and mediators from the microbiome entering the body could all be implicated in this process. The amount of eCBs linked to alterations in Peptostreptococcaceae, Veillonellaceae , and Akkermansiaceae has been found to elevate in response to dietary treatments using certain fatty acids (37). In addition to enhancing the quality of life (QoL), cannabis use can affect eCB tone and promote mucosal healing in people with ulcerative colitis (UC) (38). It has also been documented to favor immune suppression in vivo through the modulation of the eCB system using cannabinoids (39). In a mouse model of Staphylococcal enterotoxin B (SEB)-induced acute respiratory distress syndrome (ARDS), AEA treatment increased the abundance of beneficial bacteria producing SCFAs like butyrate as well as production of antimicrobial peptides (AMPs) and tight junction proteins (TJPs), which are essential molecules sustaining epithelial barrier integrity in lung epithelial cells and decreased the pathogenic Enterobacteriaceae and Pseudomonas (39). While examining the effectiveness of THC therapies, Ruminococcus gnavus, a good bacterium, was discovered to be more prevalent and pathogenic A. muciniphila in the gut and lungs was decreased along with the enrichment of propionic acid (40). Another mouse study found that combining THC and CBD reduced the signs of experimental autoimmune encephalomyelitis (EAE), which was characterized by an increase in anti-inflammatory cytokine production, a decline in pro-inflammatory cytokines, a reduction in mucin-degrading A. muciniphila , and a reversal of the high level of lipid polysaccharides (LPS) (41). Collectively, these findings indicate that cannabis affects the gut microbiome. There is growing evidence that cannabis may promote healthy gut flora, communication between the gut and the brain, and overall robust gut health (10, 42). Cannabinoids can interact with their receptors in the gut, and it has been used for hundreds of years to treat the symptoms of inflammation and GI diseases, such as abdominal discomfort, cramps, diarrhea, nausea, and vomiting (43). Hence, the systematic review would be the first to evaluate several observational studies in human and clinical trials using marijuana to alter the microbiota to treat various diseases, hoping to encourage additional thorough research studies. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 7 of 20

Protocol and registration The systematic literature review was registered on PROSPERO ID 2022 CRD42022354331. Literature search According to the PRISMA declaration guidelines, a thorough literature review and meta-analysis were undertaken (44). The study's inclusion and exclusion criteria were developed using the PICO/PECOs methodology (45, 46). Four authors (MT, TO, SH, and AJ) independently assessed each study for eligibility in the systematic review by examining PubMed, Embase, and Cochrane databases. Any discrepancies were resolved through group discussions at each stage. The search was explicitly focused on cannabinoids and microbiome-related studies conducted in the English language. Studies that exclusively involved animals or lacked peer review were excluded. These databases encompassed epidemiological and intervention research, explicitly emphasizing the effects of cannabis treatment and its impact on the microbiome. The data extracted for analysis included studies published until July 20, 2022. Study selection Before conducting full-text reviews, two reviewers (MT, TO, SH, and AJ) independently assessed papers to determine their suitability based on predetermined inclusion and exclusion criteria. Full-text versions of all relevant documents were obtained for further data extraction. The inclusion criteria for the meta-analysis encompassed interventions investigating the effects of any cannabis treatment and its impact on microbiome modification, with or without active or placebo controls. The following publications were excluded: animal studies, in vitro research, review articles, protocols, letters, editorials, comments, suggestions, and guidelines. Any disagreements among the authors were resolved through consensus. Data extraction Independent reviewers (MT, TO, SH, and AJ) conducted the data extraction process and collected information on various variables. These variables included 1) study characteristics such as author names, publication year, study period, study type, country, sample size, and age range; 2) baseline characteristics of the included studies, encompassing participant information, study region, and clinical conditions reported by both patients and controls; 3) subgroup evidence categorized by specific diagnostic health problems; and 4) adverse reactions associated with the consumption of MJ—the data extraction involved thoroughly examining relevant text, tables, and figures. In the case of any discrepancies between the reviewers, they were resolved through

or reaching a mutual agreement. Risk of bias The evaluation of the risk of bias (ROB) in the extracted intervention study (47) was carried out by two independent authors (MT, TO, SH, and AJ). In the case of non-randomized clinical trials (48), the ROB was assessed using the ROBINS-I (Risk Of Bias In Non-randomized Studies - of Interventions) tool (49). For cohort and case-control studies that were included (50, 51, 52, 53, 54, 55, 56), the two authors independently assessed the ROB using the Newcastle-Ottawa Quality Assessment Scale (NOS) (57). Statistical analysis The prevalence of the baseline characteristics was described by total number and percentage by group. The overall mean age of included studies was calculated using combined mean and standard deviation (SD) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 8 of 20 techniques. For outcomes related to microbiome diversity in intervention trials, mean differences (MD) and a 95% confidence interval (95% CI) between groups were provided. The clinical and methodological variability of participant characteristics, study time, study type, and study site were also evaluated qualitatively.

Study selection In the initial literature search, 4,022 articles were identified across various databases. After removing 1,222 duplicates, the titles and abstracts of the remaining 2,800 studies were analyzed. Based on the predetermined inclusion and exclusion criteria, 2,766 publications were excluded. The remaining 30 articles underwent full- text screening to determine their eligibility for inclusion in the systematic review. Of these, 21 studies were disqualified for the following reasons: one editorial, one letter to the editor, one protocol, six studies with incorrect study designs, and twelve non-peer-reviewed studies. Ultimately, the systematic literature review included 9 studies that met the eligibility criteria (Figure 1). Figure 1: Flow diagram for identifying studies in the systematic literature review. Across different databases, 4,022 articles were identified, but 1,222 duplicates were removed at the initial identification step. After the

and full-text screening, only 9 out of the initial pool of 2800 studies were included in the systematic review. Records identified from: PubMed (n = 1681) Embase (n = 2225) CENTRAL (n = 116) Registers (n = 0) Records removed before screening: Duplicate records removed (n = 1222) Records marked as ineligible by automation tools (n = 0) Records removed for other reasons (n = 0) Records screened (n = 2800) Records excluded (n = 2766) Reports sought for retrieval (n = 30) Reports not retrieved (n = 0) Reports assessed for eligibility (n = 30) Reports excluded: Non-peer reviewed (n = 12) Wrong study design (n = 6) Editorial (n = 1) Letter to editor (n = 1) Protocol (n = 1) Studies included in review (n = 9) Reports of included studies (n = 9) Identification of studies via databases and registers Identification Screening Included . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 9 of 20 Study characteristics Among the included studies were one case-control study (54), six cohort studies, and two clinical trials (47, 48). These studies involved 2,526 participants and covered a research period ranging from 1983 to 2021, as indicated in Table 1. The data for these studies were collected from four different countries: the United States of America (USA), the Islamic Republic of Iran (Iran), the State of Israel, and the United Kingdom (UK). The study durations varied, ranging from 1 to 3 years. The age range of the study population spanned from 17 to 101 years. Table 1. Baseline characteristics of the included studies # Author Published year Study period Study type Country Sample size Age range Ref. 1 Payahoo 2019 2016 - 2018 Randomized double- blind clinical trial Iran 56 18 - 59 (47) 2 Habib 2021 2019 - 2020 Non-randomized clinical trial Israel 16 27-78 (48) 3 Kagen 1983 1982 Case-control study USA 38 17 - 36 (54) 4 Panee 2018 - Cohort study USA 39 21 - 36 (51) 5 Fulcher 2018 2014 - 2016 Cohort study USA 37 28-39 (52) 6 Newman 2019 - Cohort study USA 39 18 - 58 (56) 7 Vijay 2021 2018 - 2020 Cohort study UK 78 >45 (50) 8 Minichino 2021 - Cohort study UK 786 18 -101 (53) 9 Vallejo 2021 2019 Cohort study USA 1380 20-34 (55) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 10 of 20 Subject characteristics We discovered that 43% of participants played a role as patients, whose ages ranged from 17 to 101, as opposed to control groups, whose ages ranged from 18 to 87. The controls had a lower mean age of 28.4 years, whereas the patients' average age was 56.5 years. Women comprise more than 85% of the participants, with the USA contributing the most. Patients are more likely to suffer cognitive impairment (76.6%) than HIV infection cases (11.1%) and others (Table 2). Table 2. Baseline characteristics of the included clinical trials Characteristics Patients Controls Participants, n (%) 1051 (43) 1418 (57) Age range 17-101 18-87 Age, mean (SD) 56.5 (14.3) 28.4 (11.3) Gender, n (%) Male 154 (14.7) 47 (3.3) Female 897 (85.3) 1361 (96) NA 0 (0) 10 (0.7) Region, n (%) USA 201 (18.8) 1389 (96.6) Iran 28 (2.6) 10 (0.7) Israel 37 (3.5) 20 (1.4) UK 806 (75.2) 19 (1.3) Clinical consideration, n (%) HIV infection 117 (11.1) - Pain/Inflammation 54 (5.1) - Systemic aspergillosis 28 (2.7) - Obesity 27 (2.6) - Cognitive deficits 805 (76.6) - Oral disease 20 (1.9) - Risk of bias The review comprised 7 observational studies (50, 51, 52, 53, 54, 55, 56) assessed for the ROB using the NOS assessment technique (Figure 2). Among these studies, only one conducted by Minichino et al. displayed a very high risk of biases. This was due to the absence of a description of the non-exposed cohort, the inability to blind the outcome assessors, and a lack of follow-up information. Similarly, previous cohort studies by Newman et al., and Vallejo et al. also exhibited biases due to challenges in outcome assessment blinding and the absence of a follow-up timeline description. In the case-control study by Kagen et al., a significant bias was identified due to a lack of statements regarding case and control selection and exposure. Furthermore, in the reported results of the RCT conducted by Payahoo et al., there was evidence of selection bias as they did not include the specified lipid profile analysis mentioned in their protocol (Figure 3). On the . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 11 of 20 other hand, the non-RCT study conducted by Habib et al. demonstrated a well-performed risk assessment (Figure 4). Figure 2: Assessment of risk of bias for cohort and case-control studies using the NOS assessment tool Figure 3: Assessment of risk of bias for randomized controlled trials Figure 4: Assessment of risk of bias for non-randomized controlled trials . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 12 of 20 Microbiota alteration by different cannabis usage We conducted a qualitative analysis of the alterations in microbiota observed in patients with various clinical conditions when using MJ/medical cannabis and its compounds. This analysis also encompassed the examination of microbial diversity and factors associated with these changes, as presented in Table 3. Table 3. Qualitative analysis of microbial changes on the use of cannabis Author (Year) Study design Clinical

Cannabis type Sample and detection Microbiome changes Summary Fulcher et al. (2018) Retrospective cohort study HIV-1 infection Marijuana (MJ) 16s rDNA sequencing using rectal swab - Positive association: Clostridium_IV, Solobacterium, Fusobacterium, Ruminococcus. - Negative association: Acidaminococcus, Prevotella, Dialister, Anaerostipes, Dorea - MJ use was the critical driver of microbiome variation [R2=0.01, p=0.14]. Vallejo et al. (2021) Retrospective cohort study HIV patients with vaginal discharge Marijuana (MJ) Affirm Vaginal Pathogens DNA Direct Probe using vaginal discharge - Lack of Lactobacilli - Overgrowth of facultative anaerobic organisms (such as Gardnerella vaginalis, Prevotella, Bacteroides, and Peptostreptococcus). - MJ use in reproductive-age women increases the odds of developing recurrent BV by two-fold (aOR=2.05), adjusting for confounders like age, ethnicity, cannabis use, insurance type, and asthma. Kagen et al. (1983) Case-control study Systemic aspergillosis Marijuana (MJ) cigarettes Culture from sputum, skin pustules, urine, nasal secretion, and lung biopsy - Fungi found in MJ users: A. fumigutus, A. flavus, A. niger, Mucor, Penicillium, Thermoactinomyces candidus, and Thermoactinomyces vulgaris. - MJ smoking can enhance fungal sensitivity, which raises the risk of exposure to and illness from fungal sources. Vijay et al. (2021) Longitudinal cohort study knee arthritis 2-AG, AEA, OEA, PEA 16s rDNA sequencing using feces - At baseline, AEA and OEA were positively associated with α -diversity significantly and with SCFA-producing bacteria such as Bifidobacterium (2-AG, p<0.01; PEA, p<0.01), Coprococcus 3 and Faecalibacterium (PEA, p<0.01) and negatively associated with Collinsella (AEA, p=0.004). - Using an exercise intervention, an increase in SCFA-producing bacteria and a decrease in the proinflammatory genus Collinsella are correlated with increases in eCBs circulating levels. - Approximately one-third of the anti-inflammatory effects of SCFAs are statistically mediated by eCBs. Minichino et al (2021) Cohort study anhedonia/ motivation Palmitoyl- ethanol- amide (PEA) 16s rDNA sequencing using feces - Microbial α -diversity was associated with both fecal PEA levels ( β = -0.31; p<0.001) and severity of anhedonia/ amotiva- tion ( β = -0.10; p=0.02). - Some β -diversity indexes correlated with fecal PEA levels or anhedonia/ amotivation (p < 0.01). - Faecal PEA positively associates with anhedonia/motivation ( β = 0.13; p<0.01). - None with serum PEA. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 13 of 20 Panee et al. (2018) Cohort study Cognitive deficits THC 16s rDNA sequencing using feces - Found a negative correlation between Prevotella and Bacteroides (p=0.012). - The lower Prevotella: Bacteroides ratio was associated with higher lifetime MJ use (p=0.052). - The ratio was almost 13-fold more excellent in non-users (p=0.34). Newman et al. (2019) Cohort study Oral cancer/ disease Marijuana (MJ) smoke 16s rDNA sequencing using swab samples of tongue and oral pharynx - Tongue: Capnocytophaga, Fusobacterium, and Porphyromonas were low, and Rothia was prominent. - Oral pharynx: Selenomonas were greater, and Streptococcus were lower. - Frequent MJ smokers have different surface oral mucosal microbiomes. - No variations at the lateral tongue location. - Differences were more in line with the cancer level in the oral pharyngeal location. Habib et al. (2021) Non- randomized Clinical trial Patients with Musculoskele tal pain Medical cannabis Culture from saliva - S. mutan decreased in week 1 and elevated in week 4. - Lactobacilli was increased at both time points. - Medical cannabis use may be associated with levels of oral microbiota. Payahoo et al. (2019) Randomized double-blind clinical trial Obesity Oleoylethan olamide (OEA) qRT-PCR from feces - After 8 weeks, the abundance of A. muciniphila increased significantly for the OEA group compared to placebo (p<0.001). - In the OEA group, energy intake (fat, protein, CHO) was decreased significantly (p=0.035). Abbreviation: 16s rDNA = 16S ribosomal deoxyribonucleic acid; 2-AG = 2-arachidonoylglycerol; AEA = anandamide; aOR = adjusted odds ratio; CHO = carbohydrate; eCBs = endocannabinoids; qRT-PCR = real-time quantitative reverse transcription PCR; HIV = human immunodeficiency virus; OEA = oleoyl ethanolamide; PBMC = peripheral blood mononuclear cell; PEA = palmitoylethanolamide; SCFA = short-chain fatty acid; THC = tetrahydrocannabinol. Adverse events Six patients (21.4%) reported symptoms, including coughing and wheezing after using MJ cigarettes, while one (3.5%) suffered drowsiness, night sweats, systemic aspergillosis, and coughing bouts. Steven et al. reported these negative consequences, which went away quickly after quitting smoking MJ (54).

The comprehensive study supported the link between cannabis usage and microbiota in a variety of patients suffering from oral illness, obesity, systemic aspergillosis, pain/inflammation, and HIV infection. The utilization of MJ was performed by substance use, oral capsules, or cigarette smoking. Microbial alteration in HIV patients using MJ Approximately 77% of HIV patients were lifetime MJ users, while the proportion of uninfected counterparts was 44.5% in a nationally representative sample (58). Research in this area suggests that cannabinoids are anti-inflammatory in the setting of HIV through the effect of ECS in the gut and through stabilization of gut-blood barrier integrity (59). Fulcher et al. conducted a study focusing on HIV-positive men who have sex with men to examine the impact of MJ usage on gut microbiota (52). The study consisted of two visits with a 6-month interval, during . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 14 of 20 which 49% of participants reported MJ use on the first visit, and 30% reported MJ use on the second visit. Utilizing permutational multivariate analysis of variance, the study revealed a positive correlation between MJ use in HIV patients and an increase in Fusobacterium and Anaerotruncus, as well as a negative association with the abundance of Dorea organisms (52), as indicated in Table 3. Typically, MJ users in HIV patients engage in higher-risk sexual conduct or practices leading to bacterial vaginosis (BV), characterized by lacking Lactobacilli and the overgrowth of facultative anaerobic organisms (55). Furthermore, the prevalence of trichomonas infection in rectal samples decreased from 14% to 5% over the 6 months (p=0.08), while gonorrhea and syphilis infections increased from 8% to 11% (p=0.66) and 0% to 5% ( p=0.16), respectively (52). On the contrary, a prospective study showed that the patients using marijuana were more than six times as likely to test positive for T. vaginalis (aOR=6.2, p=0.0003) (60). Additional clinical studies are needed to address the ongoing controversy surrounding the association between marijuana use and changes in sexually transmitted diseases. Vallejo et al. conducted a cohort study focusing on females of reproductive age who experienced vaginal discharge (55). Out of the participants, 15% reported marijuana (MJ) usage. Among those with recurrent bacterial vaginosis (BV), 28.7% (23 out of 80 patients) were MJ users, while among those without recurrent BV, 14.2% (185 out of 1300 patients) reported MJ usage ( p<0.01). Logistic regression analysis indicated a significant association between MJ use and recurrent BV, with an adjusted odds ratio of 2.05, as presented in Table 3. Interestingly, Fulcher et al. also found that 28.4% of the MJ users have a history of asthma patients while 18.3% of the non-users ( p<0.01) (52). Since cannabis has a bronchodilator effect on the airway, it has beneficial effects for asthma patients, yet, there are some detrimental effects on the lungs (61). With reported improvement in asthma symptoms, it might be used for medicinal or recreational purposes. Impact of MJ on microbiota regarding pain or inflammation To investigate the functional interactions between the eCB system and the gut microbiome in regulating inflammatory markers, Vijay et al. conducted a 6-week exercise intervention (50). They discovered that changes in anandamide (AEA) were positively associated with butyrate, and increases in AEA and palmitoylethanolamide (PEA) were correlated with decreases in TNF- α and IL-6. These associations statistically mediated one-third of the effect of short-chain fatty acids (SCFAs) on these cytokines (50). The findings suggest that the eCB system plays a role in the anti-inflammatory actions of SCFAs, indicating the involvement of additional pathways in the regulation of the immune system by the gut microbiota. Therefore, improved EC tone induced by exercise may mediate the shift in the gut microbiota to increased SCFA producers, thereby increasing the SCFA production without a dietary change. Musculoskeletal pain is the most common cause of chronic non-cancer pain, and the perceptions of these patients are that cannabis can help to relieve the pain with only minor adverse effects and improve psychological well-being (62). In a clinical trial on patients with musculoskeletal pain using medical cannabis treatment, it found an elevation of S. mutan and Lactobacilli in 4 th week even though the first was low at the first week (48). It was unexpected to see levels of S. mutans or Lactobacilli rise afterward. It is possible that cannabis either had a favorable or unfavorable effect on these particular bacteria, promoting the growth of these two oral bacteria. Different expression of microbiota in oral diseases after the use of MJ . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 15 of 20 Newman et al. found that genera earlier shown to be enriched on head and neck squamous cell carcinoma (HNSCC) mucosa, such as Capnocytophaga, Fusobacterium, and Porphyromonas, were at low levels at the tongue site in MJ users, while Rothia, which is found at depressed levels on HNSCC mucosa, was high (56). At the oral pharynx site, differences in bacteria were distinct, with higher levels of Selenomonas and lower levels of Streptococcus, as seen in HNSCC. In samples taken from the lateral border of the tongue and the oral pharynx, which are both different when it comes to the MJ-linked microbiome, it was found that daily/almost daily inhalation of MJ over the previous month correlates with differently abundant taxa of the oral microbiome. The use of MJ is associated with changes in bacteria levels, but it has not been established that it is the cause of how normal tissue develops into disease and then SCC. Furthermore, these changes were not consistent with malignancy. Lateral tongue sites demonstrated microbiological changes with MJ usage. Gut microbiota in cognitive deficits with the use of MJ Moreover, using MJ is associated with alterations in gut microbiota and mitochondrial (mt) function, leading to further cognitive deficits (51). It was associated with lower fruit and vegetable consumption and greater animal-based food consumption in adults and adolescents. It also found that a more extensive lifetime MJ use was associated with a lower Prevotella: Bacteriodes ratio, as indicated in Table 3. The authors suggested that MJ use and associated dietary change contribute to microbiome alteration along with lower dietary intake of antioxidants and fibers. Effect of cannabinoids in obesity in terms of microbiome Obesity is an excessive buildup of fat that can be unhealthy for health and cause an inflammatory response (63). Currently, herbal remedies are gaining popularity in the treatment of obesity and its co- morbidities, and Cannabis sativa derivatives are receiving much attention. In the randomized double-blind controlled clinical trial, it was discovered that energy intake, fat, protein, and carbohydrate declined significantly (p<0.001) in the OEA group (47). OEA supplement use significantly decreased the energy and carbohydrate intake of obese participants, and A. muciniphila bacterium increased considerably in the OEA group compared to the placebo group, suggesting that OEA could be used as a supplement for obese people (62). Dysbiosis in cognitive deficits using MJ Both mitochondrial (mt) dysfunction and gut dysbiosis also affect cognition. From the preliminary findings from Panee et al. (51), mt function correlated positively with Fluid Cognition and Flanker Inhibitory Control and Attention scores in MJ users but not in non-users (interaction p=0.0018–0.08).

Our goal was to compile a comprehensive picture of the microbiome associated with MJ use. In the systematic review, it was reported that both the ECS and the GI microbiota could individually contribute to the manifestation of pain and others. Despite the limited literature dedicated to these interactions to date, this should be anticipated for further work in this exciting research field. Abbreviations CENTRAL: Cochrane Central Register of Controlled Trials; CIs: Confidence intervals; GRADE: Grading of Recommendation Assessment, Development and Evaluation; NOS: Newcastle - Ottawa Quality Assessment . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 16 of 20 Scale; PRISMA-P: Preferred Reporting Items for Systematic review and Meta-analysis Protocols; RCTs: Randomized controlled trials; ROBINS-I: Risk Of Bias In Non-randomized Studies - of Interventions; US: United States. Acknowledgments None Authors’ contributions MT and SR contributed to the analysis and writing of the manuscript. MT, TO, SR and KP contributed to the conception and design. MT, SR, and AJ conducted the data curation. MT, TN, NH, and KP contributed to the critical revision of the manuscript. All authors read and approved the final manuscript. KP is the guarantor of the review. Funding M.T. was supported by the Graduate Scholarship Programme for ASEAN or Non-ASEAN Countries, Chulalongkorn University. T.O. and K.P. were funded by the Second Century Fund (C2F), Chulalongkorn University. Availability of data and materials Data will be available as supplementary files. Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests.

1. NIDA. Cannabis (Marijuana) DrugFacts [Internet]. National Institutes of Health; 2019 [cited 2022. Available from: https://nida.nih.gov/publications/drugfacts/cannabis-marijuana . 2. Elsohly MA, Slade D. Chemical constituents of marijuana: the complex mixture of natural cannabinoids. Life Sci. 2005;78(5):539-48. 3. Pertwee RG. Cannabinoid pharmacology: the first 66 years. Br J Pharmacol. 2006;147 Suppl 1:S163- 71. 4. UN Office on Drugs and Crime. World drug report 2021: United Nations publication, Sales No. E.21.XI.8; 2021. 5. United Nations Office on Drugs and Crime. World drug report 2022 2022 [updated 2022; cited 2022. Available from: https://www.unodc.org/unodc/en/data-and-analysis/wdr2022_annex.html . 6. Subbaraman MS, Kerr WC. Cannabis use frequency, route of administration, and co-use with alcohol among older adults in Washington state. J Cannabis Res. 2021;3(1):17. 7. Compton WM, Han B, Jones CM, Blanco C, Hughes A. Marijuana use and use disorders in adults in the USA, 2002-14: analysis of annual cross-sectional surveys. Lancet Psychiatry. 2016;3(10):954-64. 8. Hasin DS. US Epidemiology of Cannabis Use and Associated Problems. Neuropsychopharmacology. 2018;43(1):195-212. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 17 of 20 9. Compton WM, Han B, Hughes A, Jones CM, Blanco C. Use of Marijuana for Medical Purposes Among Adults in the United States. JAMA. 2017;317(2):209-11. 10. Karoly HC, Mueller RL, Bidwell LC, Hutchison KE. Cannabinoids and the Microbiota-Gut-Brain Axis: Emerging Effects of Cannabidiol and Potential Applications to Alcohol Use Disorders. Alcohol Clin Exp Res. 2020;44(2):340-53. 11. Zou S, Kumar U. Cannabinoid Receptors and the Endocannabinoid System: Signaling and Function in the Central Nervous System. Int J Mol Sci. 2018;19(3). 12. Demuth DG, Molleman A. Cannabinoid signalling. Life Sci. 2006;78(6):549-63. 13. Di Marzo V. The endocannabinoid system: its general strategy of action, tools for its pharmacological manipulation and potential therapeutic exploitation. Pharmacol Res. 2009;60(2):77-84. 14. Giacobbe J, Marrocu A, Di Benedetto MG, Pariante CM, Borsini A. A systematic, integrative review of the effects of the endocannabinoid system on inflammation and neurogenesis in animal models of affective disorders. Brain Behav Immun. 2021;93:353-67. 15. Elliott DM, Singh N, Nagarkatti M, Nagarkatti PS. Cannabidiol Attenuates Experimental Autoimmune Encephalomyelitis Model of Multiple Sclerosis Through Induction of Myeloid-Derived Suppressor Cells. Front Immunol. 2018;9:1782. 16. Hegde VL, Nagarkatti M, Nagarkatti PS. Cannabinoid receptor activation leads to massive mobilization of myeloid-derived suppressor cells with potent immunosuppressive properties. Eur J Immunol. 2010;40(12):3358-71. 17. Berg BB, Soares JS, Paiva IR, Rezende BM, Rachid MA, Cau SBA, et al. Cannabidiol Enhances Intestinal Cannabinoid Receptor Type 2 Receptor Expression and Activation Increasing Regulatory T Cells and Reduces Murine Acute Graft-versus-Host Disease without Interfering with the Graft-versus-Leukemia Response. J Pharmacol Exp Ther. 2021;377(2):273-83. 18. Matson TE, Carrell DS, Bobb JF, Cronkite DJ, Oliver MM, Luce C, et al. Prevalence of Medical Cannabis Use and Associated Health Conditions Documented in Electronic Health Records Among Primary Care Patients in Washington State. JAMA Netw Open. 2021;4(5):e219375. 19. Wang T, Collet JP, Shapiro S, Ware MA. Adverse effects of medical cannabinoids: a systematic review. CMAJ. 2008;178(13):1669-78. 20. Leung J, Chan G, Stjepanovic D, Chung JYC, Hall W, Hammond D. Prevalence and self-reported reasons of cannabis use for medical purposes in USA and Canada. Psychopharmacology (Berl). 2022;239(5):1509-19. 21. Yan F. Mechanistic Understanding of the Symbiotic Relationship Between the Gut Microbiota and the Host: An Avenue Toward Therapeutic Applications. Cell Mol Gastroenterol Hepatol. 2020;10(4):853-4. 22. Dopkins N, Nagarkatti PS, Nagarkatti M. The role of gut microbiome and associated metabolome in the regulation of neuroinflammation in multiple sclerosis and its implications in attenuating chronic inflammation in other inflammatory and autoimmune disorders. Immunology. 2018;154(2):178-85. 23. Lu J, Hou W, Gao S, Zhang Y, Zong Y. The Role of Gut Microbiota-Gut-Brain Axis in Perioperative Neurocognitive Dysfunction. Front Pharmacol. 2022;13:879745. 24. Carabotti M, Scirocco A, Maselli MA, Severi C. The gut-brain axis: interactions between enteric microbiota, central and enteric nervous systems. Ann Gastroenterol. 2015;28(2):203-9. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 18 of 20 25. Simpson CA, Diaz-Arteche C, Eliby D, Schwartz OS, Simmons JG, Cowan CSM. The gut microbiota in anxiety and depression - A systematic review. Clin Psychol Rev. 2021;83:101943. 26. Tian P, Wang G, Zhao J, Zhang H, Chen W. Bifidobacterium with the role of 5-hydroxytryptophan synthesis regulation alleviates the symptom of depression and related microbiota dysbiosis. J Nutr Biochem. 2019;66:43-51. 27. Tanasescu R, Constantinescu CS. Cannabinoids and the immune system: an overview. Immunobiology. 2010;215(8):588-97. 28. Nasser Y, Woo M, Andrews CN. Cannabis in Gastroenterology: Watch Your Head! A Review of Use in Inflammatory Bowel Disease, Functional Gut Disorders, and Gut-Related Adverse Effects. Curr Treat Options Gastroenterol. 2020;18(4):519-30. 29. Ligresti A, Moriello AS, Starowicz K, Matias I, Pisanti S, De Petrocellis L, et al. Antitumor activity of plant cannabinoids with emphasis on the effect of cannabidiol on human breast carcinoma. J Pharmacol Exp Ther. 2006;318(3):1375-87. 30. Zuardi AW. Cannabidiol: from an inactive cannabinoid to a drug with wide spectrum of action. Braz J Psychiatry. 2008;30(3):271-80. 31. Wilkinson JD, Williamson EM. Cannabinoids inhibit human keratinocyte proliferation through a non- CB1/CB2 mechanism and have a potential therapeutic value in the treatment of psoriasis. J Dermatol Sci. 2007;45(2):87-92. 32. National Academies of Sciences Engineering and Medicine (U.S.). Committee on the Health Effects of Marijuana: an Evidence Review and Research Agenda. The health effects of cannabis and cannabinoids : the current state of evidence and recommendations for research. Washington, DC: The National Academies Press; 2017. xviii, 468 pages p. 33. Memedovich KA, Dowsett LE, Spackman E, Noseworthy T, Clement F. The adverse health effects and harms related to marijuana use: an overview review. CMAJ Open. 2018;6(3):E339-E46. 34. Rutsch A, Kantsjo JB, Ronchi F. The Gut-Brain Axis: How Microbiota and Host Inflammasome Influence Brain Physiology and Pathology. Front Immunol. 2020;11:604179. 35. Derrien M, Vaughan EE, Plugge CM, de Vos WM. Akkermansia muciniphila gen. nov., sp. nov., a human intestinal mucin-degrading bacterium. Int J Syst Evol Microbiol. 2004;54(Pt 5):1469-76. 36. Belzer C, de Vos WM. Microbes inside--from diversity to function: the case of Akkermansia. ISME J. 2012;6(8):1449-58. 37. Castonguay-Paradis S, Lacroix S, Rochefort G, Parent L, Perron J, Martin C, et al. Dietary fatty acid intake and gut microbiota determine circulating endocannabinoidome signaling beyond the effect of body fat. Sci Rep. 2020;10(1):15975. 38. Tartakover Matalon S, Azar S, Meiri D, Hadar R, Nemirovski A, Abu Jabal N, et al. Endocannabinoid Levels in Ulcerative Colitis Patients Correlate With Clinical Parameters and Are Affected by Cannabis Consumption. Front Endocrinol (Lausanne). 2021;12:685289. 39. Sultan M, Wilson K, Abdulla OA, Busbee PB, Hall A, Carter T, et al. Endocannabinoid Anandamide Attenuates Acute Respiratory Distress Syndrome through Modulation of Microbiome in the Gut-Lung Axis. Cells. 2021;10(12). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 19 of 20 40. Mohammed A, Alghetaa HK, Zhou J, Chatterjee S, Nagarkatti P, Nagarkatti M. Protective effects of Delta(9) -tetrahydrocannabinol against enterotoxin-induced acute respiratory distress syndrome are mediated by modulation of microbiota. Br J Pharmacol. 2020;177(22):5078-95. 41. Al-Ghezi ZZ, Busbee PB, Alghetaa H, Nagarkatti PS, Nagarkatti M. Combination of cannabinoids, delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD), mitigates experimental autoimmune encephalomyelitis (EAE) by altering the gut microbiome. Brain Behav Immun. 2019;82:25-35. 42. Kalaydina R-V, Qorri B, Szewczuk MR, editors. Preventing Negative Shifts in Gut Microbiota with Cannabis Therapy: Implications for Colorectal Cancer2017. 43. Ahmed W, Katz S. Therapeutic Use of Cannabis in Inflammatory Bowel Disease. Gastroenterol Hepatol (N Y). 2016;12(11):668-79. 44. Page MJ, Moher D, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ. 2021;372:n160. 45. Huang X, Lin J, Demner-Fushman D. Evaluation of PICO as a knowledge representation for clinical questions. AMIA Annu Symp Proc. 2006;2006:359-63. 46. Morgan RL, Whaley P, Thayer KA, Schunemann HJ. Identifying the PECO: A framework for formulating good questions to explore the association of environmental and other exposures with health outcomes. Environ Int. 2018;121(Pt 1):1027-31. 47. Payahoo L, Khajebishak Y, Alivand MR, Soleimanzade H, Alipour S, Barzegari A, et al. Investigation the effect of oleoylethanolamide supplementation on the abundance of Akkermansia muciniphila bacterium and the dietary intakes in people with obesity: A randomized clinical trial. Appetite. 2019;141((Payahoo L.; Khajebishak Y.) Assistant Professor of Nutrition Sciences, Maragheh University of Medical Sciences, Maragheh, Iran(Alivand M.R.) Department of Medical Genetics, Faculty of Medicine, Tabriz University of Medical Sciences, Tabriz, Iran(Soleiman). 48. Habib G, Steinberg D, Jabbour A. The impact of medical cannabis consumption on the oral flora and saliva. PLoS One. 2021;16(2):e0247044. 49. Sterne JA, Hernan MA, Reeves BC, Savovic J, Berkman ND, Viswanathan M, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919. 50. Vijay A, Kouraki A, Gohir S, Turnbull J, Kelly A, Chapman V, et al. The anti-inflammatory effect of bacterial short chain fatty acids is partially mediated by endocannabinoids. Gut Microbes. 2021;13(1):1997559. 51. Panee J, Gerschenson M, Chang L. Associations Between Microbiota, Mitochondrial Function, and Cognition in Chronic Marijuana Users. J Neuroimmune Pharmacol. 2018;13(1):113-22. 52. Fulcher JA, Hussain SK, Cook R, Li F, Tobin NH, Ragsdale A, et al. Effects of Substance Use and Sex Practices on the Intestinal Microbiome During HIV-1 Infection. J Infect Dis. 2018;218(10):1560-70. 53. Minichino A, Jackson MA, Francesconi M, Steves CJ, Menni C, Burnet PWJ, et al. Endocannabinoid system mediates the association between gut-microbial diversity and anhedonia/amotivation in a general population cohort. Mol Psychiatry. 2021;26(11):6269-76. 54. Kagen SL, Kurup VP, Sohnle PG, Fink JN. Marijuana smoking and fungal sensitization. J Allergy Clin Immunol. 1983;71(4):389-93. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint Page 20 of 20 55. Vallejo V, Shan W, Ilagan JG, Goldberg GL. Marijuana use in women of reproductive age and recurrent bacterial Vaginosis. Journal of Reproductive Medicine. 2021;66(9-10):256-62. 56. Newman T, Krishnan LP, Lee J, Adami GR. Microbiomic differences at cancer-prone oral mucosa sites with marijuana usage. Sci Rep. 2019;9(1):12697. 57. GA Wells, B Shea, D O'Connell, J Peterson, V Welch, M Losos, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses Ottawa Hospital Research Institute: Ottawa Hospital Research Institute; 2021 [cited 2021. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp . 58. Shiau S, Arpadi SM, Yin MT, Martins SS. Patterns of drug use and HIV infection among adults in a nationally representative sample. Addict Behav. 2017;68:39-44. 59. Ellis RJ, Wilson N, Peterson S. Cannabis and Inflammation in HIV: A Review of Human and Animal Studies. Viruses. 2021;13(8). 60. Crosby R, DiClemente RJ, Wingood GM, Harrington K, Davies SL, Hook EW, 3rd, et al. Predictors of infection with Trichomonas vaginalis: a prospective study of low income African-American adolescent females. Sex Transm Infect. 2002;78(5):360-4. 61. Jarjou'i A, Izbicki G. Medical Cannabis in Asthmatic Patients. Isr Med Assoc J. 2020;22(4):232-5. 62. Furrer D, Kroger E, Marcotte M, Jauvin N, Belanger R, Ware M, et al. Cannabis against chronic musculoskeletal pain: a scoping review on users and their perceptions. J Cannabis Res. 2021;3(1):41. 63. Cavalheiro E, Costa AB, Salla DH, Silva MRD, Mendes TF, Silva LED, et al. Cannabis sativa as a Treatment for Obesity: From Anti-Inflammatory Indirect Support to a Promising Metabolic Re-Establishment Target. Cannabis Cannabinoid Res. 2022;7(2):135-51. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 11, 2023. ; https://doi.org/10.1101/2022.12.31.22284080doi: medRxiv preprint