Problematic Cannabis Use Is Associated with Reduced Rectal Microbial Species Richness and Diversity Among a Pilot Sample of Young Sexual and Gender Minorities

Abstract

Compared to young heterosexual men, young sexual and gender minorities (YSGM) have elevated systemic inflammation and unique intestinal microbial profiles, influenced by HIV infection and substance use. However, links between cannabis use and microbial dysbiosis in this population have not been well described. In this pilot study, we aimed to characterize the complex interrelationships between cannabis use and microbial community structure in YSGM in relationship to HIV status. Cannabis use was assessed by self-administered Cannabis Use Disorder Identification Test (CUDIT) questionnaires and rectal microbial community alpha-diversity metrics were assessed via 16S ribosomal ribonucleic acid (rRNA) sequencing in a subset of YSGM (n = 42) in the RADAR cohort (aged 16–29) in Chicago. Multivariable regression models were used to assess the relationship between cannabis use and microbiome alpha-diversity metrics, adjusting for HIV status and other risk characteristics, including inflammation, which was evaluated by plasma levels of C-reactive protein (CRP). Problematic cannabis use, but not general use, was significantly inversely associated with microbial community richness (Adj. Beta = −8.13; 95% confidence interval [CI]: −15.68 to −0.59) and Shannon diversity (Adj. Beta = −0.04; 95% CI: −0.07 to 0.009). No significant association was observed between CUDIT score and community evenness, nor was any significant moderation observed by HIV status. We observed that problematic cannabis use was associated with reduced microbial community richness and Shannon diversity, adjusting for within population differences in inflammation and HIV status. Future research should aim to assess how cannabis use contributes to microbiome-related health factors among YSGM and if decreasing cannabis use can restore gut microbial community structure.

Introduction

Microbiome dysbiosis has been implicated in several autoimmune and inflammatory diseases¹ as well as many cardiovascular-related pathologies, including obesity² and hypertension.³ Men who have sex with men (MSM), compared to men who have sex with women, have unique patterns of gastrointestinal microbiome composition with significantly richer and more diverse microbiota, differences only marginally associated with variations in diet.⁴ These unique patterns are further differentiated between those who engage in condomless receptive anal intercourse (CRAI) compared to those who do not; individuals that engage in CRAI have been shown to have unique rectal mucosal profiles, including greater levels of proinflammatory cytokines and distinctive microbial community structure, including a lower Shannon Index, a measure of sample richness and evenness, compared to individuals who do not engage in CRAI.^5
–7

In addition, the use of lubricant during receptive anal sex by MSM has been shown to significantly reduce the abundance of microbes in rectal tissue.⁸ HIV infection itself has also consistently been associated with reduced bacterial richness and intestinal dysbiosis,^4,9 elevated microbial translocation, and increased systemic inflammation.¹⁰ Substance use, meanwhile, has been shown to increase rectal mucosal inflammatory cytokine production regardless of HIV status.¹¹ In particular, cannabis use has been identified as a potential driver of microbial variation among people with HIV (PWH).¹¹

Together, these past studies suggest that MSM have a unique microbial community structure, which associates with elevated systemic inflammation among this population. Little work, however, has been done to examine the effect of the use of common substances, such as cannabis, on microbiome composition among MSM and whether this relationship is moderated by HIV status. To begin to fill this gap in the literature this end, we collected a pilot sample to examine microbiome diversity and develop a better understanding of within population variation in intestinal microbial composition, as well as the links between substance use, inflammation, and alterations in microbial community structure. In addition, we focused on young sexual and gender minorities (YSGM) given this population's high risk of HIV infection¹² and elevated use of cannabis.¹³ This a crucial step in developing biomedical interventions aimed at reducing HIV incidence and averting poor HIV-associated health outcomes.

Methods

Study participants and approval

Pilot data and samples were collected as a substudy of a larger cohort study. The parent study, RADAR, is an ongoing longitudinal cohort study of Chicago metropolitan area YSGM assigned a male sex at birth. The primary objective of this cohort study is to apply a multilevel perspective to a syndemic of health issues associated with HIV among YSGM. At the time of enrollment into RADAR, all participants were between 16 and 20 years of age, assigned male at birth, spoke English, and had a sexual encounter with a man in the previous year or identified as gay, bisexual, or transgender. All cohort members provided written informed consent and all study procedures and protocols were approved by the Northwestern University Institutional Review Board (STU00087614) and were carried out following relevant guidelines and regulations.

Pilot analytic data and samples included in the current study come from a subset of participants (n = 42) in the RADAR cohort. PWH were demographically matched (based on age, race, and ethnicity) to people without HIV (PWOH). Care was taken to ensure that no significant differences existed between those selected or not selected for this subset of data, including no differences based on biologic measures (e.g., CRP). All selected participants self-identified as black or African American to limit further confounding in this pilot data analysis and to reserve degrees of freedom for other key variables. All data came from the baseline study visit only (Accession no. PRJNA949231).

Demographic and substance use variables

Participants were asked to provide demographic information, including age, race/ethnicity, sexual identity, and gender identity at the baseline study visit. Body mass index (BMI) was collected and calculated objectively based on CDC recommendations.¹⁴ Number of instances of receptive anal intercourse were assessed as: (1) protected intercourse only, (2) unprotected intercourse only, and (3) the total number of any receptive anal intercourse.

Cannabis use was operationalized and assessed in three ways. First, use was as a binary variable, any use in the past 6 months or no use in the past 6 months based on urine test results of tetrahydrocannabinol level (THC; THC positive or THC negative). Urine samples were tested for metabolites using the Multi-Drug Screen Test Panel (DOA-264) dip cards from Innovacon, Inc., (San Diego, CA). Second, it was utilized as a categorical variable¹⁵ based on self-reported frequency of use in the past 30 days: (1) never, (2) intermittent use, ≤5 times in the past 30 days, and (3) frequent, ≥6 times in the past 30 days. Third, cannabis use and associated problems were assessed using an eight-item screen instrument, the Cannabis Use Disorder Identification Test (CUDIT).^16,17 CUDIT scores ranged from 0 to 32 and were operationalized as a continuous variable with higher scores indicating more problematic use.

The use of all other substances (methamphetamines, cocaine/crack, poppers, ecstasy, stimulants, depressants, prescription pain killers, ketamine, gamma hydroxybutyrate, inhalants, psychedelics, and heroin) were combined into a single binary variable where a value of 1 indicates use of any other substance and 0 indicates no other substance use.

Microbiome and multiplex biomarker analysis

Rectal swabs were self-collected by study participants as previously described.¹⁸ DNA was extracted from all rectal swabs included in this analysis at the same time. To do so, rectal swab tips were placed in 200 μL of lysis buffer (30 mM Tris-HCl, 10 mM EDTA, 200 mM sucrose, pH 8.2) for 10 min at 65°C. 10 mg/mL lysozyme solution (final concentration) was then added to each sample and incubated for 1 h at 37°C. Five percent sodium dodecyl sulfate was then added to a final concentration of 1% w/v and incubated at 56°C for 10 min. Extractions were then performed using Qiagen's DNeasy Blood and Tissue Kit.

Rectal microbial communities were assessed in extracted DNA via sequencing of the V4 region of the 16S ribosomal ribonucleic acid (rRNA) small subunit rRNA (SSU rRNA) using previously described 515 F (5′-GTGYCAGCMGCCGCGGTAA-3′)—806 R (5′-GGACTACNVGGGTWTCTAAT-3′) primers.¹⁹ Processing protocols were based on Earth Microbiome project protocols.²⁰ Triplicate amplicon concentrations were normalized and pooled to create a sequencing library. The library was cleaned and quantified by KAPA quantification (KAPA Biosystems, Wilmington, MA). A negative (water) control was extracted simultaneously with study specimens to monitor potential contamination at each step in the extraction and library preparation process.

The library was sequenced with an Illumina MiSeq (San Diego, CA) and a 2 × 150 bp cycle run utilizing 20% PhiX phage as a control. Processing of demultiplexed and trimmed FastQ files was done using the DADA2 R package (version 1.14.0).²¹ Taxonomic determination was conducted using the Silva 138 training set classifier.^22,23 The phangorn package (version 2.5.5) was used to create a phylogenetic tree.²⁴ Resulting files were then compiled and used for downstream analysis in R. The Phyloseq package (version 1.30) was used to analyze amplicon sequencing variants taxa count table.²⁵ The microbiome package (version 2.5–7) was used to calculate alpha-diversity metrics of richness, evenness (Pielou), and Shannon diversity.²⁶

Plasma levels of CRP were assessed using samples collected at the same study visit as rectal swabs collection, via a multiplex MESO QuickPlex SQ 120 electrochemiluminescence immunoassay platform assay (Meso Scale Discovery, MSD). CRP had a dynamic range of 0.001–49.6 mg/L and was operationalized as a continuous, logarithmic variable to account for non-normal distribution of the data.

HIV testing

Fingerstick blood samples were collected as part of each participant's visit every 6 months. Each participant's HIV infection status was determined using the Alere Determine HIV1/2 Ab/Ag Combo 4th generation point-of-care (POC) test. Those who tested positive on the POC HIV tests received confirmatory HIV antigen and antibody immunoassay testing following current CDC HIV testing guidelines.²⁷

Statistical analyses

Participant characteristics were described using means, standard deviations (SDs), and proportions, as appropriate. To assess differences in alpha diversity metrics, including (1) evenness (Pielou); (2) richness (observed); and (3) Shannon diversity, we used a Wilcoxon rank sum test to compare urine THC screen positive and negative participants, as well as participants categorized as nonproblematic cannabis users (CUDIT score <8) and problematic cannabis users (CUDIT score ≥8). We further assessed differences in alpha diversity metrics between participants categorized as never, intermittent, and frequent cannabis use based on self-report using a Kruskal-Wallis test. To assess differences in microbial composition (beta diversity) between cannabis users operationalized as (1) urine THC screen positive and negative; (2) never, intermittent, and frequent cannabis use based on participant self-report; and (3) problematic and nonproblematic cannabis use based on CUDIT score, we conducted principal coordinates analysis and PERMANOVA tests.

Multivariable linear regression models were next used to assess the relationship between urine THC screen and self-reported cannabis use and microbial community alpha-diversity metrics. All models were adjusted for demographic characteristics and known confounders, including age, BMI, log CRP, HIV status, and other substance use. All known confounders, as well as all covariates that were identified as statistically significant at the p < 0.05 level using the Wald test statistic, were included in the multivariable regression models. Supplementary analyses included assessing moderation based on gender identity and examination of receptive anal intercourse as potentially key confounding variable. All analyses were performed in Stata version 16.1 or R version 4.21.

Results

Table 1 depicts the demographic characteristics of the pilot sample. Mean age of the sample was 22.69 years (SD = 2.29). A plurality of participants used cannabis within the past 30 days (n = 26, 61.90%) with 20 (48.8%) screening positive for urine THC. Among those self-reporting any marijuana use, 10 participants (23.81%) reported use five times or less and 16 participants (38.10%) reporting use six times or greater in the past 30 days. Mean CUDIT score was 7.74 (SD = 6.75), indicating overall nonproblematic use but approaching the cutoff of eight for potentially hazardous use. Eight (19.05%) participants reported use of any other substances other than cannabis. Regarding HIV status, 22 (52.38%) were laboratory-confirmed as PWH and 20 (47.62%) as PWOH. Mean BMI and log CRP were 25.64 (SD = 7.49) and 0.77 (SD = 1.46), respectively.

Table 1.

Baseline Demographic Characteristics of Participants in the Analytic Sample, RADAR Study, Chicago (N = 42)

Variable	Total
Variable	n	%	Mean (SD)
Age, years	—	—	22.69 (2.29)
Race
Black alone	42	100	—
Gender identity
Cisgender	38	90.5	—
Transgender	4	9.5	—
Sexual identity
Gay	29	69.1	—
Bisexual	8	19.1	—
Different identity	5	11.9	—
Substance use
Marijuana use^a			—
Never	16	38.10	—
≤5 times	10	23.81	—
≥6 times	16	38.10	—
CUDIT score^b	—	—	7.74 (6.75)
Other substance use^c	8	19.05	—
HIV status
PWH	20	47.62	—
PWOH	22	52.38	—
BMI	—	—	25.64 (7.49)
Log CRP	—	—	0.77 (1.46)

In the past 30 days.

Assessed using the CUDIT test and scoring method; higher score indicates higher risk marijuana use.

Other drug use includes methamphetamines, cocaine/crack, poppers, ecstasy, stimulants, depressants, prescription pain killers, ketamine, GHB, inhalants, psychedelics, and heroin.

CUDIT, Cannabis Use Disorder Identification Test; GHB, gamma hydroxybutyrate; SD, standard deviation; BMI, body mass index; CRP, C-reactive protein; HIV, human immunodeficiency virus; PWH, people with HIV; PWOH, people without HIV.

To investigate microbial community and compositional diversity, we first compared three key measures of microbial alpha-diversity, as well as beta diversity metrics between study participants. No significant differences in richness, evenness, Shannon diversity, or beta diversity were observed between participants with binary urine THC screen, categorized cannabis use frequency, or problematic cannabis use based on CUDIT score. Multivariable linear regression models were next used to examine the association between multiple measures of cannabis use and microbial alpha-diversity. No significant association was observed between either binary urine THC screen or categorical definitions of cannabis use and evenness, richness, nor Shannon diversity.

CUDIT score, indicating problematic cannabis use, was significantly inversely associated with both community richness (Adj. Beta = −8.13; 95% confidence interval [CI]: −15.68 to −0.59) and Shannon index (Adj. Beta = −0.04; 95% CI: −0.07 to 0.009), adjusting for known confounders and other demographic factors (Table 2). No significant association was observed between CUDIT score and community evenness nor was any significant moderation observed by HIV status. Supplementary analyses suggested no additional moderation by gender identity nor did receptive anal intercourse (protected, unprotected, or total acts) confound the analyses or improve model fit via Likelihood-ratio test.

Table 2.

Multivariable Linear Regression Models Examining the Association Between Microbiome Health and Problematic Marijuana Use (CUDIT Score) Among Young Men Who Have Sex with Men and Transgender Women—Chicago, Illinois, 2016–2018 (N = 42)

	Evenness (Pielou) (R² = 0.226)		Richness (R² = 0.333)		Shannon diversity (R² = 0.323)
Variable	Coef.	95% CI	Coef.	95% CI	Coef.	95% CI
CUDIT score^a	−0.001	−0.003 to 0.001	−8.13^*	−15.68 to −0.59	−0.04^*	−0.07 to −0.009
Age	−0.001	−0.006 to 0.003	−19.60	−40.32 to 1.12	−0.06	−0.14 to 0.02
BMI	−0.000005	−0.002 to 0.001	−3.90	−10.39 to 2.60	−0.01	−0.04 to 0.01
Log CRP	0.002	−0.006 to 0.010	−28.03	−63.29 to 7.22	−0.09	−0.23 to 0.05
HIV test result
PWOH	ref	—	ref	—	ref	—
PWH	−0.02	−0.04 to 0.006	−27.37	−129.45 to 74.70	−0.14	−0.54 to 0.25
Other drug use^b
None	ref	—	ref	—	ref	—
≥1 other drug	−0.02	−0.05 to 0.006	−20.38	−166.54 to 85.79	−0.14	−0.23 to 0.05

p < 0.05.

Assessed using the CUDIT test and scoring method; higher score indicates higher risk marijuana use.

Other drug use includes methamphetamines, cocaine/crack, poppers, ecstasy, stimulants, depressants, prescription pain killers, ketamine, GHB, inhalants, psychedelics, and heroin.

aOR, adjusted odds ratio; CI, confidence interval.

Discussion

In this pilot sample of YSGM from Chicago, we observed significant associations between problematic cannabis use and decreased microbial community richness and Shannon diversity metrics. No significant association was observed between any cannabis use nor frequency of cannabis use and any alpha-diversity measures. We also did not observe any significant moderation of these relationships according to HIV status. Together, our results suggest that problematic, but not general cannabis use, may associate with decreased microbial richness and diversity among young MSM, irrespective of HIV status.

Past research evaluating the intestinal microbiome among MSM has demonstrated uniquely diverse patterns of microbial composition among this population compared to their heterosexual counterparts.⁴ Recent work has also observed significantly elevated levels of systemic inflammation among this population, independent of HIV status.²⁸ The results presented in this pilot analysis contribute to these past bodies of work by demonstrating that problematic levels of cannabis use are associated with decreased measures of community richness and diversity, even after adjusting for differences in systemic inflammation, as evaluated by plasma CRP levels, and HIV status. Identifying and reducing chronically elevated levels of cannabis use among this high-risk population may aid in maintenance or restoration of microbial community structure.

Future work should aim to replicate the findings of this small pilot sample and to ascertain whether interventions targeting this population may improve gut microbiome diversity and potentially reduce downstream negative consequences, such as risk of HIV-associated comorbidities.

While we observed associations of problematic cannabis use with reduced microbiome diversity measurements among YSGM, our findings should be considered in the context of their limitations. First, no appropriate comparison sample of young heterosexual men was available. Second, cannabis use was self-reported and is reliant on participant recall and willingness to accurately report levels of use, both of which may bias the findings in a way that subjective measures would not. Next, we did not collect self-reported information regarding other medication use, in which many affect gut microbiome health (e.g., NSAIDS). We do not anticipate, however, meaningful changes to our findings as our sample is of young adults who are unlikely to report substantial use of microbiome-altering medications.

Finally, this sample was a single time point community sample rather than a longitudinal probability sample and, as such, findings may not generalize to the larger population of YSGM, particularly give our limited pilot sample size.

Conclusion

Here, we observed an association between problematic cannabis use and decreased rectal microbial richness and Shannon diversity, even after adjusting for known confounders and within population differences in inflammation. We also did not observe any significant moderation according to HIV status. These results suggest that cannabis use is associated with lowered rectal microbial alpha-diversity measures in a complex way among YSGM, and that more integrated analyses are warranted.

Footnotes

Acknowledgments

The authors thank the entire RADAR research team, particularly Dr. Thomas Remble and Antonia Clifford for overseeing the project and Daniel T. Ryan for data management. We also thank the RADAR participants for sharing their experiences with us.

Authors' Contributions

E.M. and J.A.M. lead conceptualization of the article, writing of the original draft, and assisted with formal analysis. C.B., H.H., and N.R.K., lead the formal analysis and assisted with writing the original draft. R.D., A.W.C., and B.M. assisted with conceptualization, methodology, data curation and acquisition, and reviewed and edited the article.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse or the National Institutes of Health. The sponsors were not involved in the conduct of the research or the preparation of the article.

Author Disclosure Statement

B.M. has received consulting fees from Hologic, Inc.; the other authors have no conflicts of interest to declare.

Funding Information

This work was supported by grants from the National Institute on Drug Abuse at the National Institutes of Health (F32DA046313, PI: E.M.; U01DA036939, PI: B.M.), the Viral Pathogenesis Core of the Third Coast Center for AIDS Research (CFAR, P30AI117943), and a pilot award from the University of Miami CFAR Developmental Core (GR015190/GR015573, PI: J.A.M.; UM CFAR P30-AI073961; PI: Dr. Savita Pahwa). J.A.M. is supported, in part, by grants R21OD031435, R01HD108015, R01DA054553, UG3AI69652, and TNPRC NIH core grant P51OD011104.

References

Belkaid

, Hand

. Role of the microbiota in immunity and inflammation. Cell, 2014; 157(1):121–141; doi: 10.1016/j.cell.2014.03.011

Turnbaugh

, Ley

, Mahowald

, et al. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature, 2006; 444(7122):1027–1031; doi: https://doi.org/10.1038/nature05414

Yang

, Santisteban

, Rodriguez

, et al. Gut dysbiosis is linked to hypertension. Hypertension (Dallas, Tex: 1979), 2015; 65(6):1331–1340; doi: 10.1161/HYPERTENSIONAHA.115.05315

Noguera-Julian

, Rocafort

, Guillén

, et al. Gut microbiota linked to sexual preference and HIV infection. EBioMedicine, 2016; 5:135–146; doi: 10.1016/j.ebiom.2016.01.032

Kelley

, Kraft

, de Man

, et al. The rectal mucosa and condomless receptive anal intercourse in HIV-negative MSM: Implications for HIV transmission and prevention. Mucosal Immunol, 2017; 10(4):996–1007; doi: 10.1038/mi.2016.97

Armstrong

AJS

, Shaffer

, Nusbacher

, et al. An exploration of Prevotella-rich microbiomes in HIV and men who have sex with men. Microbiome, 2018; 6(1):198; doi: https://doi.org/10.1186/s40168-018-0580-7

Vujkovic-Cvijin

, Sortino

, Verheij

, et al. HIV-associated gut dysbiosis is independent of sexual practice and correlates with noncommunicable diseases. Nat Commun, 2020; 11(1):2448; doi: https://doi.org/10.1038/s41467-020-16222-8

Haaland

, Fountain

, Hu

, et al. Repeated rectal application of a hyperosmolar lubricant is associated with microbiota shifts but does not affect PrEP drug concentrations: Results from a randomized trial in men who have sex with men. J Int AIDS Soc, 2018; 21(10):e25199-e; doi: 10.1002/jia2.25199

Kehrmann

, Menzel

, Saeedghalati

, et al. Gut microbiota in human immunodeficiency virus–infected individuals linked to coronary heart disease. J Infect Dis, 2019; 219(3):497–508; doi: 10.1093/infdis/jiy524

10.

El-Far

, Tremblay

. Gut microbial diversity in HIV infection post combined antiretroviral therapy: A key target for prevention of cardiovascular disease. Curr Opin HIV AIDS, 2018; 13(1):38–44; doi: 10.1097/COH.0000000000000426

11.

Fulcher

, Hussain

, Cook

, et al. Effects of substance use and sex practices on the intestinal microbiome during HIV-1 infection. J Infect Dis, 2018; 218(10):1560–1570; doi: 10.1093/infdis/jiy349

12.

Centers for Disease Control and Prevention. HIV Surveillance in Adolescents and Young Adults. Atlanta, GA, USA; 2010.

13.

Marshal

, Friedman

, Stall

, Thompson

. Individual trajectories of substance use in lesbian, gay and bisexual youth and heterosexual youth. Addiction, 2009; 104(6):974–981; doi: 10.1111/j.1360-0443.2009.02531.x

14.

CDC. Defining Adult Overweight and Obesity; 2016. Available from: https://www.cdc.gov/obesity/adult/defining.html [Last accessed: August 15, 2018 ].

15.

Morgan

, Khanna

, Skaathun

, et al. Marijuana use among young Black men who have sex with men and the HIV care continuum: Findings from the uConnect cohort. Subst Use Misuse, 2016; 51(13):1751–1759; doi: 10.1080/10826084.2016.1197265

16.

Adamson

, Sellman

. A prototype screening instrument for cannabis use disorder: The Cannabis Use Disorders Identification Test (CUDIT) in an alcohol-dependent clinical sample. Drug Alcohol Rev, 2003; 22(3):309–315; doi: 10.1080/0959523031000154454

17.

Adamson

, Kay-Lambkin

, Baker

, et al. An improved brief measure of cannabis misuse: The Cannabis Use Disorders Identification Test-Revised (CUDIT-R). Drug Alcohol Depend, 2010; 110(1–2):137–143; doi: https://doi.org/10.1016/j.drugalcdep.2010.02.017

18.

Hudson

, D'Aquila

, Mustanski

, Morgan

. A proportion of self-collected rectal swabs yield human immunodeficiency virus sequences phylogenetically related to those from plasma human immunodeficiency virus RNA. AIDS Res Hum Retroviruses, 2020; 36(1):92–95; doi: 10.1089/AID.2019.0170

19.

Caporaso

, Lauber

, Walters

, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proc Natl Acad Sci U S A, 2011; 108(supplement_1):4516–4522; doi: https://doi.org/10.1073/pnas.1000080107

20.

Walters

, Hyde

, Berg-Lyons

, et al. Improved bacterial 16S rRNA gene (V4 and V4-5) and fungal internal transcribed spacer marker gene primers for microbial community surveys. mSystems, 2016; 1(1):e00009–e000015; doi: https://doi.org/10.1128/mSystems.00009-15

21.

Callahan

, McMurdie

, Rosen

, et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods, 2016; 13(7):581–583; doi: https://doi.org/10.1038/nmeth.3869

22.

Pruesse

, Quast

, Knittel

, et al. SILVA: A comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res, 2007; 35(21):7188–7196; doi: 10.1093/nar/gkm864

23.

Yarza

, Yilmaz

, Pruesse

, et al. Uniting the classification of cultured and uncultured bacteria and archaea using 16S rRNA gene sequences. Nat Rev Microbiol, 2014; 12(9):635–645; doi: 10.1038/nrmicro3330

24.

Schliep

. Phangorn: Phylogenetic analysis in R. Bioinformatics, 2011; 27(4):592–593; doi: 10.1093/bioinformatics/btq706

25.

McMurdie

, Holmes

. Phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One, 2013; 8(4):e61217; doi: https://doi.org/10.1371/journal.pone.0061217

26.

Lahti

, Sudarshan

. Tools for microbiome analysis in R. Version 2.5–7; 2017. Available from: http://microbiome.github.com/microbiome [Last accessed: February 09, 2023 ].

27.

Centers for Disease Control and Prevention. Guidelines and Recommendations. In: Services DoHaH, editor. CDC: Atlanta, GA; 2019.

28.

Morgan

, Taylor

, Ryan

, et al. Systemic inflammation is elevated among both HIV-uninfected and -infected young men who have sex with men. AIDS, 2019; 33(4):757–759; doi: 10.1097/QAD.0000000000002093