Impact of doxycycline post-exposure prophylaxis for sexually transmitted infections on the gut microbiome and antimicrobial resistome.
Chu VT, Glascock A, Donnell D, Grabow C, Brown CE, Ward R, Love C, Kalantar KL, Cohen SE, Cannon C, Woodworth MH, Kelley CF, Celum C, Luetkemeyer AF, Langelier CR
- DOI
- 10.1038/s41591-024-03274-2
- Record issued
- 2026-08-16
- Engine
- 7.39.0
- Exported
- 2026-09-21
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/eed4ffd5-82b4-4025-90de-987814d27b6b is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted secondary analysis of a randomized trial, with strong reporting of study design, ethics, key resources, data/code availability, and transparency. The main weaknesses are minor: lack of explicit blinding and power analysis for the secondary analysis, and a few copyedit issues including an internal inconsistency in participant count (501 vs 510).
This is an observational secondary analysis of a randomized trial; both reviewers agreed on this classification. The statistics verification covered only a subset of tests (3 tests, 2 consistent, 0 inconsistent) and cannot confirm the correctness of all reported statistics. The citation check found no retracted or unresolved references. The copyedit pass flagged minor issues, including a participant-count inconsistency.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 2 tests: 2 consistent, 0 inconsistent; 2 via agent-written checks.
- CONSISTENTreported p = .009 · recomputed p = .009Reviewer 1Check the p-value for the Spearman correlation between doxycycline doses and tetracycline ARG proportion in the metagenome (rho=0.23, P=9.0e-3).
“Exposure to a higher number of doxycycline doses correlated with proportional enrichment of tetracycline ARGs in the metagenome (Spearman’s ρ = 0.23, P = 9.0 × 10 −3 )”
Taken as given: The sample size n=127 is used for the correlation test.; The test is two-tailed.Method: Two-tailed Spearman rank correlation p-value computed from rho and n.How we recomputed it: pR(0.23, 127) - UNCOMPUTABLEreported p = .023 · recomputed p = .009Reviewer 2Check p-value for increase in tetracycline ARG proportion in metagenome (46% to 51%) using Wilcoxon rank-sum test.
“The proportional mass of tetracycline ARGs in the resistome increased between day-0 and month-6 in DP participants from 46% to 51% in the metagenome ( P = 2.3 × 10 −2 )”
Taken as given: The reported p-value corresponds to a two-sided Wilcoxon rank-sum test comparing day-0 and month-6 within the DP arm.; The sample size for the test is 127 (DNA-seq samples).; The reported Spearman's rho of 0.23 is used as a proxy for the test statistic, which is not exact but provides a rough check.Method: Approximated using pR(r, n) with r=0.23 and n=127, which gives a two-sided p-value for Spearman correlation; this is not the exact test used but serves as a sanity check.How we recomputed it: pR(0.23, 127) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 2Check p-value for increase in tetracycline ARG proportion in metatranscriptome (4% to 15%) using Wilcoxon rank-sum test.
“and from 4% to 15% in the metatranscriptome ( P = 4.5 × 10 −6 )”
Taken as given: The reported p-value corresponds to a two-sided Wilcoxon rank-sum test comparing day-0 and month-6 within the DP arm.; The sample size for the test is 86 (RNA-seq samples).; The reported Spearman's rho of 0.55 is used as a proxy for the test statistic, which is not exact but provides a rough check.Method: Approximated using pR(r, n) with r=0.55 and n=86, which gives a two-sided p-value for Spearman correlation; this is not the exact test used but serves as a sanity check.How we recomputed it: pR(0.55, 86)
- lowinternal contradictionThe Methods section states '510 DoxyPEP trial participants' while the Abstract and other sections state 501 participants. This is likely a typo but should be corrected.
“For this analysis, a subset of 150 participants from the 510 DoxyPEP trial participants was selected for metagenomic sequencing of self-collected rectal swab samples.”
MethodsFind in source
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
6 major claims checked against the paper's own evidence: 1 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewers 1, 2The increase in tetracycline ARGs may be associated with potential bacterial pathogens.The paper shows correlation analyses suggesting associations with pathogens, but these are exploratory and not definitive.Evidence: Correlation analyses show increasing correlations with Bacteroides, Escherichia, and Staphylococcus.
“We found significant and increasing correlations between tetracycline efflux pump ARGs and the abundance of several genera encompassing clinically relevant human pathogens, including Bacteroides , Escherichia and, to a smaller extent, Staphylococcus , in DP participants after 6 months of doxycycline use.”
Discussion ¶6Find in source - supportedReviewers 1, 2Doxy-PEP use over 6 months increased the proportion and expression of tetracycline ARGs in the gut resistome.The paper provides direct evidence from DNA-seq and RNA-seq showing significant increases in tetracycline ARG proportion and expression in the DP arm.Evidence: Results show increases from 46% to 51% (P=2.3e-2) in metagenome and 4% to 15% (P=4.5e-6) in metatranscriptome.
“The proportional mass of tetracycline ARGs in the resistome increased between day-0 and month-6 in DP participants from 46% to 51% in the metagenome ( P = 2.3 × 10 −2 ) and from 4% to 15% in the metatranscriptome ( P = 4.5 × 10 −6 )”
AbstractFind in source - supportedReviewers 1, 2Doxy-PEP use minimally affected the gut microbiome's taxonomic composition.The paper reports no significant differences in alpha diversity, beta diversity, or bacterial mass, and no differentially abundant genera.Evidence: Results show no differences in diversity or mass, and differential abundance analysis found no consistent changes.
“Bacterial microbiome alpha diversity, beta diversity and total bacterial mass did not differ between day-0 and month-6 samples from DP participants when assessed by either DNA-seq or RNA-seq.”
AbstractFind in source - supportedReviewers 1, 2The impact of doxy-PEP was restricted to tetracycline ARGs without co-selection for other antibiotic classes.The paper reports no significant increases in other ARG classes and a sensitivity analysis showing decreases in MLS ARGs.Evidence: Results state 'No proportional increases were noted in other non-tetracycline ARG classes' and sensitivity analysis showed decreases in MLS ARGs.
“No proportional increases were noted in other non-tetracycline ARG classes (Fig. ), suggesting specificity of doxy-PEP use for tetracycline ARGs.”
ResultsFind in source - supportedReviewer 1Doxy-PEP use led to a dose-dependent increase in tetracycline ARG expression.The paper shows a strong positive correlation between doxycycline doses and tetracycline ARG expression (rho=0.55, P=3.7e-8).Evidence: Results show Spearman's rho=0.55, P=3.7e-8 for metatranscriptome.
“Exposure to a higher number of doxycycline doses correlated with proportional enrichment of tetracycline ARGs in the metagenome (Spearman’s ρ = 0.23, P = 9.0 × 10 −3 ) and metatranscriptome (Spearman’s ρ = 0.55, P = 3.7 × 10 −8 ).”
AbstractFind in source - supportedReviewer 2Doxy-PEP use increased the expression of tetracycline ARGs.The metatranscriptomic data show a significant increase in tetracycline ARG expression, supported by dose-response correlation.Evidence: RNA-seq: increase in expressed tetracycline ARGs (Padj=1.5e-2) and dose-dependent correlation (Spearman's rho=0.55, P=3.7e-8).
“Furthermore, in the metatranscriptome, the number of doxycycline doses was strongly positively correlated with both tetracycline ARG richness (Spearman’s ρ = 0.39, P = 2.2 × 10 −4 ) and the relative proportion of expressed tetracycline ARGs in the resistome (Spearman’s ρ = 0.55, P = 3.7 × 10 −8 ; Fig. ).”
ResultsFind in source
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites randomized trials of doxy-PEP and notes the lack of data on its impact on the gut microbiome and resistome. It acknowledges limitations of prior studies (culture-based or 16S rRNA) and justifies the use of metagenomics and metatranscriptomics. The hypothesis follows logically from the cited evidence.
“To address the outstanding question of whether doxy-PEP impacts the ecology of the gut microbiome and resistome, we studied longitudinally collected rectal swabs from DoxyPEP clinical trial participants using a combination of DNA-seq and RNA-seq approaches.”
“In contrast, metagenomic DNA sequencing (DNA-seq) allows for comprehensive assessment of bacterial genomes and the genetic potential for antimicrobial resistance, and metatranscriptomic RNA sequencing (RNA-seq) provides a functional profile of actively transcribed genes, including ARGs.”
The parent trial randomized participants 2:1 to DP or SOC, and the analysis subset was selected based on pre-specified criteria (arm, HIV status, sample availability). The paper describes the selection of 150 participants from the 501 trial participants. Power analysis is not reported for this secondary analysis, but the study is exploratory. Inclusion/exclusion criteria are described. Blinding is not mentioned, but this is a secondary analysis of an open-label trial; the outcome is objective sequencing data.
“Participants were randomized in a 2:1 ratio to the DP arm or the SOC arm.”
“Individuals were eligible for enrollment if they were at least 18 years of age; had male sex assigned at birth; had received a diagnosis of HIV or were on HIV PrEP; and had received a bacterial STI diagnosis of gonorrhea, chlamydia or early syphilis in the previous 12 months.”
“One outlier from the DP arm was omitted from the RNA-seq beta diversity plot for graphical purposes (coordinates NMDS1: 9.9; NMDS2: −0.3) but was included in the calculations.”
“Participants were randomized in a 2:1 ratio to the DP arm or the SOC arm.”
“Individuals were eligible for enrollment if they were at least 18 years of age; had male sex assigned at birth; had received a diagnosis of HIV or were on HIV PrEP; and had received a bacterial STI diagnosis of gonorrhea, chlamydia or early syphilis in the previous 12 months.”
“One outlier from the DP arm was omitted from the RNA-seq beta diversity plot for graphical purposes (coordinates NMDS1: 9.9; NMDS2: −0.3) but was included in the calculations.”
The paper reports that all participants were assigned male sex at birth, and provides age, race/ethnicity, HIV status, and CD4 counts in Table 1. It does not report sex justification because the study population is exclusively male by design (MSM and transgender women). Age and health status are reported. Demographics are adequately reported.
“No sex or gender analysis was carried out given that the study population enrolled only those who were assigned male sex at birth.”
“Age (years): median (IQR) | 39 (31–49) | 38 (32–50) | 38 (32–50) | 0.86”
“Living with HIV infection | 12 (40%) | 23 (39%) | 35 (39%) | 0.93”
“No sex or gender analysis was carried out given that the study population enrolled only those who were assigned male sex at birth.”
“Age (years): median (IQR) | 39 (31–49) | 38 (32–50) | 38 (32–50) | 0.86”
The methods state that the study protocol was approved by the University of California, San Francisco institutional review board, and all participants provided written informed consent. The parent trial is registered at ClinicalTrials.gov (NCT03980223). Regulatory compliance is implied by IRB approval and consent.
“The study protocol was approved by the University of California, San Francisco institutional review board, which served as the primary institutional review board.”
“All participants provided written informed consent, and no monetary compensation was provided for participation.”
“The study protocol was approved by the University of California, San Francisco institutional review board, which served as the primary institutional review board.”
“All participants provided written informed consent, and no monetary compensation was provided for participation.”
The paper identifies DNA/RNA Shield tubes, NEBNext kits, FastSelect, ERCC standards, and the CZ ID pipeline with versions. Software versions for R packages are provided. No antibodies or cell lines are used, so those criteria are not applicable.
“Swabs were self-collected into DNA/RNA Shield collection tubes (Zymo Research, R1107-E) and stored at −80 °C within 2 weeks of collection.”
“All analyses were conducted in RStudio (version 2023.09.1+494) using R (version 4.2.1) and performed for both DNA-seq and RNA-seq data.”
“Participants in the DP arm were counseled to take a 200-mg doxycycline hyclate dose within 72 h after condomless anogenital, vaginal or oral sex and no more than one dose every 24 h.”
“DNA-seq was carried out using a NEBNext Ultra II DNA Kit (New England Biolabs, E7645L).”
“All analyses were conducted in RStudio (version 2023.09.1+494) using R (version 4.2.1) and performed for both DNA-seq and RNA-seq data.”
The paper names tests (Wilcoxon rank-sum, PERMANOVA, Spearman, inflated beta regression) and provides exact p-values. Effect sizes with confidence intervals are shown in Extended Data Fig. 4 for regression models. Software versions are given. Data presentation includes box plots with individual points and per-group n. Mathematical plausibility checks were not possible for most statistics due to lack of raw data, but no obvious errors were found.
“P values were calculated using the two-sided Wilcoxon rank-sum test and adjusted for multiple comparisons.”
“The proportional mass of tetracycline ARGs in the resistome increased between day-0 and month-6 in DP participants from 46% to 51% in the metagenome ( P = 2.3 × 10 −2 ) and from 4% to 15% in the metatranscriptome ( P = 4.5 × 10 −6 )”
“The regression model is shown as a line, and the gray shaded areas surrounding the regression lines show the 95% confidence intervals.”
“P values were calculated using the two-sided Wilcoxon rank-sum test and adjusted for multiple comparisons.”
“The proportional mass of tetracycline ARGs in the resistome increased between day-0 and month-6 in DP participants from 46% to 51% in the metagenome ( P = 2.3 × 10 −2 ) and from 4% to 15% in the metatranscriptome ( P = 4.5 × 10 −6 )”
“The regression model is shown as a line, and the gray shaded areas surrounding the regression lines show the 95% confidence intervals.”
FASTQ files are deposited in SRA under BioProject PRJNA1099775, and de-identified patient data and source data are available on GitHub. Code is also available on GitHub. The data availability statement is concrete.
“All code can be found at https://github.com/infectiousdisease-langelier-lab/doxy-PEP .”
“FASTQ files containing non-host reads identified by the CZ-ID pipeline, after subtraction of reads aligning to the human genome, are available from the National Center for Biotechnology Informationʼs Sequence Read Archive under BioProject ID PRJNA1099775 (https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA1099775) .”
“De-identified patient data and source data can be found at https://github.com/infectiousdisease-langelier-lab/doxy-PEP .”
“All code can be found at https://github.com/infectiousdisease-langelier-lab/doxy-PEP .”
The trial is registered (NCT03980223). A reporting summary is mentioned. All outcomes are reported, including negative results. Limitations are thoroughly discussed. Conclusions are proportional. Funding and competing interests are declared.
“We also acknowledge the limitations of this study. First, samples represented only a subset of the DoxyPEP clinical trial participants over the first 6 months of follow-up, and, in the DP arm, samples from participants with higher doxy-PEP use were preferentially selected, which may have biased findings away from the null compared to the average individual using doxy-PEP.”
“This work is supported by a grant (R01AI143439 to A.F.L. and C. Celum) from the Division of Microbiology and Infectious Diseases, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH); the grant funded the DoxyPEP study design and data collection.”
“ClinicalTrials.gov registration number: NCT03980223 (https://clinicaltrials.gov/study/NCT03980223) .”
“Further information on research design is available in the linked to this article.”
“We also acknowledge the limitations of this study. First, samples represented only a subset of the DoxyPEP clinical trial participants over the first 6 months of follow-up, and, in the DP arm, samples from participants with higher doxy-PEP use were preferentially selected, which may have biased findings away from the null compared to the average individual using doxy-PEP.”
Registered (1 ID: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 49 references by DOI: 43 verified — 6 no DOI (shown, not verified).
- NO DOIGuidelines for the Use of Doxycycline Post-Exposure Prophylaxis for Bacterial STI PreventionNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIWHO announces the development of the Consolidated Guidelines on STI prevention and care (2022–2023)No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOISexually T ransmitted Disease Surveillance 2022: Gonococcal Isolate Surveillance Project ProfileNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAntibiotic Resistance Threats in the United States: 2019No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIModern Applied Statistics with SNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIvegan: Community Ecology PackageNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
3 data/code links checked; 3 live.
- dataBioProjectLIVEHTTP 200https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA1099775Resolves to BioProject (data repository).
- codeGitHubLIVEHTTP 200https://github.com/infectiousdisease-langelier-lab/doxy-PEPResolves to GitHub (code repository).
- codeGitHubLIVEHTTP 200https://github.com/vegandevs/veganResolves to GitHub (code repository).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly typo, consistency, clarity.
- MINORtypoExtended Data Fig. 6 legend“Clostroidioides difficile”→ Clostridioides difficileMisspelling of the genus name.
- MINORconsistencyMethods, Study design“510 DoxyPEP trial participants”→ 501 DoxyPEP trial participantsThe abstract and other sections state 501 participants; this appears to be a typo.
- MINORtypoAbstract“doxy-PEP use over a 6-month period was associated with an increase in the proportion of tetracycline ARGs comprising the gut resistome and an increase in the expression of tetracycline ARGs.”→ Consider rephrasing to avoid repetition of 'increase'.Minor stylistic issue.
- MINORconsistencyMethods, Study design“a subset of 150 participants from the 510 DoxyPEP trial participants”→ The parent trial is described as 501 participants elsewhere; verify the correct total.Potential inconsistency in participant count (501 vs 510).
- MINORclarityMethods, Statistical analyses“We used inflated beta-regression models to examine the association between doxy-PEP use (independent variable) and the proportional ARG class mass within the resistome (dependent variable) using the function gamlss() (family = ’BEINF’)”→ Clarify that 'inflated beta-regression' refers to zero-and-one inflated beta regression.Technical term could be expanded for clarity.
The published work is robust and well-reported; an informed reader should weigh the minor reporting gaps (blinding, power analysis, assumptions verification) and the internal inconsistency in participant count (501 vs 510) as low-severity issues. No erratum is warranted for the core findings, but a correction for the participant-count typo and the Clostridioides/Clostridioides misspelling would be appropriate.
- 1.HIGHcopyeditCorrect the participant count in Methods, Study design from '510' to '501' to match the Abstract and other sections.The internal inconsistency (501 vs 510) is a factual error that could confuse readers and undermine trust in the reported numbers.
- 2.HIGHcopyeditFix the misspelling of 'Clostridioides difficile' in the Extended Data Fig. 6 legend.Correct genus spelling is essential for scientific accuracy and professionalism.
- 3.MEDIUMrigorAdd a statement in Methods, Study design clarifying whether analysts were blinded to study arm during sequencing and analysis, or state that blinding was not applicable.Blinding is a standard design element that reviewers expect to see reported, even for secondary analyses.
- 4.MEDIUMrigorAdd a power analysis or sample-size justification for the secondary analysis in Methods, Statistical analyses.Reporting power or a justification for the subset size strengthens the study design and helps readers assess the precision of estimates.
- 5.MEDIUMstatisticsExplicitly state in Methods, Statistical analyses that assumptions for statistical tests (e.g., normality, homoscedasticity) were assessed or that non-parametric tests were chosen a priori.The current 'reported_but_inadequate' rating for assumptions verification is a minor gap that can be closed with a single sentence.
- 6.MEDIUMstatisticsReport exact p-values in figure legends (e.g., Extended Data figures) instead of threshold symbols where possible.Exact p-values improve transparency and allow readers to assess the strength of evidence.
- 7.MEDIUMreportingAdd a note in the Discussion about the generalizability of findings to other populations (e.g., cisgender women) given the all-male cohort.The study population is exclusively male, and a generalizability caveat would help readers interpret the findings appropriately.
- 8.MEDIUMdata codeAdd a DOI or versioned release for the GitHub repository in Data availability to ensure permanent access.A versioned release or DOI protects against link rot and provides a citable, stable reference for the code and data.
- 9.LOWcopyeditRephrase the Abstract sentence to avoid repetition of 'increase' (e.g., 'was associated with an increase in the proportion of tetracycline ARGs... and elevated expression of tetracycline ARGs').Minor stylistic improvement for readability.
- 10.LOWcopyeditClarify in Methods, Statistical analyses that 'inflated beta-regression' refers to zero-and-one inflated beta regression.Expanding the technical term improves clarity for readers unfamiliar with the specific model.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.