Menopause remains one of the most significant physiological transitions in a woman’s life, with cascading effects on cardiovascular, metabolic, skeletal and neurological health. Yet according to new research from the University of Colorado Anschutz, the condition is astonishingly absent from the very records clinicians and researchers rely on most: the electronic health record. In a sweeping analysis of nearly 396,000 women enrolled in the National Institutes of Health’s All of Us Research Program, investigators found that menopause was documented in electronic health records for only about seven percent of the women in the dataset — a gap so large that it raises fundamental questions about how women’s health information is captured, stored and ultimately used in scientific research.
The study, published in the journal Menopause, compared what participants reported about themselves with what was actually written in their medical charts. The contrast was stark. Researchers identified approximately 193,000 menopause observations in survey data collected directly from participants, compared with roughly 28,000 menopause diagnoses recorded in EHR data. In other words, menopause appeared nearly seven times more often in participant surveys than in the structured clinical documentation that drives much of modern biomedical research. Among the women who themselves reported being menopausal in surveys, only about 14 percent had the status reflected in their health records.
For principal investigator Audrey Hendricks, an associate professor of bioinformatics at CU Anschutz, the discrepancy represents more than a documentation nuisance — it is a structural blind spot in the data infrastructure of women’s health. “Ultimately, we cannot study what we do not measure. We cannot treat what we do not know,” Hendricks said in the study’s announcement. Menopause, she noted, carries enormous implications for long-term health outcomes, and failing to consistently capture when it occurs limits the scientific community’s ability to understand how the transition shapes disease risk.
The scale of the All of Us Research Program is precisely what makes the finding both possible and important. Launched by the NIH as one of the most ambitious precision medicine initiatives ever undertaken, All of Us gathers data from hundreds of thousands of participants through multiple channels: detailed participant surveys, information extracted from electronic health records across a nationwide network of health care provider organizations, and genomic sequencing. This layered architecture is meant to allow researchers to cross-reference self-reported information with clinical documentation, giving a more complete picture of health and disease than either source alone.
That cross-referencing capability is exactly what the Colorado team exploited. When the researchers compared the two data streams, they found a reassuring pattern in one direction: nearly all participants who had a menopause diagnosis in their EHR also reported menopause in their survey responses. This alignment suggests that when clinicians do document menopause, the documentation generally matches women’s own accounts of their bodies. The problem is not accuracy — it is coverage. Far too few women ever receive the diagnosis code in the first place, and the resulting dataset is riddled with silent gaps rather than errors.
The consequences of those gaps extend well beyond counting. Electronic health records are built primarily for billing and clinical care, and reproductive transitions like menopause often go unrecorded unless they trigger a specific treatment decision or clinical event. But researchers increasingly recognize that menopause timing is not a trivial detail. The age at which a woman reaches menopause is thought to be intertwined with her risk of cardiovascular disease, osteoporosis, cognitive decline and other chronic conditions. Early menopause, whether spontaneous or surgically induced, may carry different risk profiles than later transitions, and hormone therapy decisions often hinge on precisely this kind of timing information.
Yet the study found that even the most basic details were frequently missing. Age at menopause — arguably the single most important piece of data for anyone studying the transition’s relationship to disease — was often unavailable not just in EHRs but in the survey data as well. Without that anchor, researchers attempting to model how reproductive history shapes health trajectories face immediate obstacles. They cannot reliably define study populations of women by menopausal status, cannot stratify analyses by timing or type of menopause, and cannot accurately calculate the number of participants needed to detect meaningful effects in future studies.
This last point — statistical power — is where the practical guidance value of the study becomes apparent. By quantifying just how sparse menopause documentation is within All of Us, the researchers have given the scientific community a realistic map of what is and is not possible with the current data. Anyone planning a menopause-related study using the program’s resources now has empirical benchmarks for how many women will have usable documentation, how many will need to be recruited through additional surveys, and how definitions of menopausal status should be constructed to avoid biased samples. In effect, the paper is both an alarm and a user’s manual.
The implications reach beyond a single research program. Large-scale biomedical data resources have become the backbone of modern epidemiology, powering genome-wide association studies, drug safety surveillance and population health modeling. But these resources inherit the blind spots of the clinical systems that feed them. If a condition is routinely undocumented, it becomes invisible to algorithms, underrepresented in cohorts and systematically excluded from discovery science. The under-documentation of menopause fits a broader and well-documented pattern in which aspects of women’s health — from endometriosis to perimenopausal symptoms — receive less research attention and less systematic data collection than conditions affecting other populations.
Hendricks and her colleagues argue that the solution lies in making reproductive health information a routine, expected element of clinical documentation rather than an afterthought. The All of Us program, which continues to grow and refine its data collection instruments, offers a promising venue for such improvements: standardized survey questions about menopausal status and timing, paired with better incentives for clinicians to record these details in the EHR, could dramatically change the research landscape. More complete data would allow investigators to examine how factors such as age at menopause relate to disease risk across the program’s remarkably diverse participant population — one of the program’s defining strengths.
The stakes of getting this right are measured in the health of millions. Women spend roughly a third of their lives in the postmenopausal state, and the hormonal shifts of the transition influence some of the leading causes of illness and death. Understanding who is at elevated risk, and when, depends on datasets that faithfully record the transition itself. “We have an enormous opportunity to use large-scale datasets to understand women’s health across the menopause transition and to identify who may be at greater risk for disease,” Hendricks said. “But we need to make sure that the information researchers need is actually being collected. We must do a better job of capturing women’s health information, including reproductive health and measures related to menopause.”
For now, the message of the study is sobering but constructive. Nearly 400,000 women generously shared their health information with a landmark national research program, and their own accounts revealed a dimension of their health that their medical records had largely erased. The sevenfold gap between lived experience and clinical documentation is a call to action for health systems, electronic record vendors and researchers alike: close the data gap, and the science of women’s midlife health can finally move at the pace it deserves.
Subject of Research: Documentation of menopause and reproductive health information in electronic health records compared with participant-reported survey data in the NIH All of Us Research Program
Subject of Research: Medicine
Article Title: Menopause in the All of Us Research Program: A Descriptive Study
Article References: Staples, J. W., White, S. L., Giacalone, A., Pozdeyev, N., Sammel, M. D., Stranger, B. E., Valencia, C. I., Santoro, N., & Hendricks, A. E. (2026). Menopause in the All of Us Research Program: a descriptive summary of electronic health record and survey response across sociodemographic characteristics. Menopause. https://doi.org/10.1097/gme.0000000000002884
Image Credits: AI Generated
DOI: 10.1097/GME.0000000000002884
Keywords: menopause, electronic health records, All of Us Research Program, women’s health, reproductive health data, NIH, disease risk, health data documentation
Cite Scienmag News
APA
MLA
Chicago
Ophelia Keating. (September 8, 2026). Electronic health records largely omit menopause, new research reveals. Scienmag. https://scienmag.com/electronic-health-records-largely-omit-menopause-new-research-reveals/
Copy citation
Download RIS
Tags: analysis of health record accuracy for women’s health conditionsanalysis of self-reported versus clinical menopause datachallenges in capturing menopausal status in electronic health systemschallenges in capturing menopause information in healthcare systemscomparison of self-reported versus clinical menopause datadisparities in health record-keeping for menopausal womendisparities in women’s health data collectiongender disparities in electronicimpact of EHR omissions on menopause researchimpact of missing menopause information on biomedical researchimplications for cardiovascular and metabolic health researchimplications of EHR omissions for menopause-related health researchimportance of comprehensive health records for menopause-related researchinfluence of incomplete health records on women’s health studieslarge-scale analysis of menopause data inMenopause documentation in electronic health recordssignificance of accurate menopause recording in EHRssignificance of menopause in health outcome studiesunderreporting of menopause in clinical recordsunderrepresentation of menopause in clinical documentationwomen’s health data gaps
