Cancer ResearchResearch PaperOpen Access

Study Tracks Stress, Cortisol and Vaginal Microbiome After Abnormal Pap Results

A protocol for a 270-woman cohort tests whether stress and cortisol rhythms predict if abnormal cervical screens persist, with microbiome as a candidate mediator.

Sunday, November 22, 2026 0 views
Published in JMIR Res Protoc
a woman collecting a saliva sample into a labeled tube at a kitchen table beside a smartphone and Fitbit

Summary

This is a study protocol, not a results paper. Researchers are enrolling women aged 21 to 65 who have an abnormal cervical cancer screening result and are waiting for colposcopy. The goal is to learn whether psychological stress and disrupted daily cortisol rhythms predict whether the abnormality persists or resolves, and whether the vaginal microbiome helps explain that link. Participants complete stress surveys, twice-daily phone check-ins, home saliva cortisol sampling, Fitbit monitoring, and give blood, urine, saliva and vaginal swabs. The target is 270 women across Latina, Black and White groups. As of July 2026, only 3 women had enrolled and 2 had finished. No outcome data exist yet, so no conclusions about stress, cortisol or microbiome can be drawn.

Detailed Summary

Cervical cancer prevention does not end with screening. An abnormal result signals elevated risk but does not confirm a precancerous lesion, and many abnormalities resolve on their own. The waiting interval can bring uncertainty, repeat testing and sometimes unneeded procedures, and these burdens fall unevenly across racial and ethnic groups. This paper describes a study designed to probe the biology of that interval. It reports a plan and early recruitment numbers, not findings.

The hypothesis is mechanistic. Psychosocial stress and dysregulated cortisol may alter vaginal microbial ecology, immune response and local homeostasis, which in turn may influence HPV persistence and lesion behavior. The authors propose that higher stress and flatter or more disrupted diurnal cortisol patterns will be associated with persistence of abnormal findings at colposcopy, with vaginal microbiome features acting as candidate mediators or modifiers. This is a hypothesis drawn from prior literature, and it has not been tested here.

The design is a prospective cohort at an academic safety-net hospital. Target enrollment is 270 women aged 21 to 65 (90 Latina, 90 non-Latina Black, 90 non-Latina White). Time points are T0 (residual Pap material from the index screen), T1 (baseline visit with blood, urine, saliva, vaginal pH and a self-collected swab) and T2 (colposcopy with clinician-collected vaginal and cervical samples). Between T1 and T2, women complete 14 days of twice-daily ecological momentary assessment, twice-daily home saliva collection for cortisol, and passive Fitbit monitoring. Exclusions include prior cervical cancer, hysterectomy, pregnancy, HIV, and recent antibiotic or immunosuppressant use.

The two coprimary confirmatory analyses use Firth bias-reduced logistic regression. They estimate adjusted odds of persistence per standardized increase in baseline Perceived Stress Scale (PSS-10) score and in mean diurnal cortisol slope, adjusting for age, race or ethnicity and index screening risk stratum. Microbiome, metabolomic, HPV-related and wearable analyses are prespecified as secondary or exploratory. Feasibility and acceptability are also evaluated, including prescreen yield, completion rates and burden-related missingness.

Implementation has been slow. Recruitment began in September 2025. By July 15, 2026, 180 women had been prescreened, 25 met eligibility, 24 were contacted, 13 completed direct screening, 8 declined, 3 enrolled and 2 completed the study. The protocol went through six amendments, including expansion from 180 to 270 participants, adding Latina women, reducing EMA prompts from four to two daily, and extending the age limit to 65. Data collection is expected to continue through December 2028.

The main caveat is that nothing has been shown. There are no stress, cortisol or microbiome results, and the early enrollment is too small to interpret. The protocol is demanding, so selection bias toward women able to manage digital tasks and home freezer storage is a real risk. Any eventual finding would be observational and could show association, not causation. Still, the multimodal design could clarify a plausible mechanism linking stress biology to cervical outcomes.

Key Findings

  • Protocol study with no outcome results yet; no data on stress, cortisol or microbiome effects are reported
  • Target sample is 270 women aged 21 to 65: 90 Latina, 90 non-Latina Black, 90 non-Latina White
  • Coprimary exposures are PSS-10 score and mean diurnal salivary cortisol slope, analyzed by Firth logistic regression for persistence at colposcopy
  • Data streams include 14 days of twice-daily EMA, twice-daily home saliva cortisol, Fitbit data, and blood, urine, saliva and vaginal/cervical specimens at T1 and T2
  • Vaginal microbiome features are prespecified secondary mediator or effect-modifier candidates, not confirmatory endpoints
  • As of July 15, 2026: 180 prescreened, 25 eligible, 13 directly screened, 8 declined, 3 enrolled, 2 completed
  • Six protocol amendments, including EMA cut from 4 to 2 daily prompts and age range extended from 21 to 65; recruitment runs through December 2028

Methodology

Prospective, single-site cohort at an academic safety-net hospital, anchored to an abnormal cervical screen and followed to diagnostic colposcopy within 3 months. Exposures are measured at baseline and across a 14-day home monitoring window; the outcome is resolution or nonpersistence versus persistence at T2. Confirmatory analyses use Firth bias-reduced logistic regression adjusted for age, race or ethnicity and index risk stratum. Microbiome, metabolomic, HPV and wearable analyses are secondary or exploratory, and feasibility metrics are tracked alongside.

Study Limitations

Only 3 women had enrolled by July 2026, so feasibility is uncertain and no results exist. The demanding protocol and digital requirements may select for participants who differ from the broader screened population, and the single-site design limits generalizability. The cohort is observational, so any findings will show association rather than causation. The excerpt provided does not include the full measures section or any conflict-of-interest statement.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: