Cure
Overview
Genomic tools trained on decades of male-dominated data can read female biology as low risk. Startups are building their own datasets to fix it, but researchers say the deeper problem runs through the NIH and the journals.
Human biology is becoming increasingly granular, molecular, and individualized with the advent of precision tools, but this shift threatens to reinforce deeply embedded biases in clinical research: a male default that frequently overlooks female biology and leads to misdiagnoses.
Cheaper genomic sequencing combined with AI tools that can scan billions of data points with ease has enabled highly tailored care, diagnostics, and therapies. This shift has fueled a massive boom in precision medicine, a market now valued at roughly $135 billion. As biotech companies and startups rush to address these market demands, the foundation these tools rely on remains fundamentally distorted.
To put it into perspective, trying to uncover granular, genetic-level insights within an overwhelmingly biased male dataset is like using a multi-billion-dollar space telescope to look for a specific exoplanet, but pointing it through a smudged lens. Zooming in further doesn't clarify the image; it just gives you a highly magnified view of the smudge. When the baseline data is inherently clouded by bias, precision tools may just document the distortion with higher resolution, often leading to misdiagnosis.
Thalidomide Disaster
In the early 1960s, thalidomide was widely prescribed to pregnant women to treat morning sickness. But unfortunately, the drug resulted in thousands of children being born with severe birth defects. In 1977, FDA advised a complete blanket ban of any women of "childbearing potential" from early-phase (Phase 1 and Phase 2) drug trials. This ban “was avoiding any risk of harming a fetus. But that is not the same as protecting women or protecting fetuses,” said Chloe Bird, a women's health researcher at Tufts Medical Center. “We don't protect hearts by not studying heart disease. We do the research and what protects it is science!” she added.
While the ban on women of childbearing capacity to participate in early-phase clinical research got lifted in 1993, women continue to pay the price for those historical restrictions. Although women account for roughly half of all cases of heart disease, cancer, and psychiatric disorders, their participation in clinical trials stands at just 41 percent. As a result, women continue to bear a disproportionate share of the burden for chronic conditions, migraines, depression, and anxiety disorders.
Missing Heart Attack Patterns
Take, for instance, traditional cardiovascular risk-scoring tools. While they calculate a 10-year heart attack risk using standard biomarkers like age, total cholesterol, and blood pressure, precision medicine tools can still fail to diagnose heart disease in women. Many of these models look for hard, calcified blockages in major arteries, a feature typical of male biology. Women, however, often develop non-obstructive plaque that is smoother, thinner, and spread throughout smaller blood vessels. Tools overwhelmingly trained on male datasets register these patterns in women as "low risk," while ignoring female-specific risk factors like high blood pressure during pregnancy, post-menopausal estrogen drops, or gestational diabetes.
Startups in the Scene
While working on her PhD in human genetics and genomics, Elizabeth Ruzzo, now the founder of precision medicine startup Adyn, encountered persistent patterns that limited her research. She was mapping genes associated with neurological conditions like epilepsy and autism when she noticed a stark sex bias: males were four times more likely to be diagnosed with autism than females. Was this a genuine biological anomaly, or a misdiagnosis? Trying to understand and dive deeper into the biological basis of the sex difference, Ruzzo was surprised to realize a massive medical gender research gap. “Women were not even required to be included in clinical trials in the US until 1993,” Ruzzo said. She realized that many reproductive misdiagnoses were even more dominant. Adyn’s flagship product is a personalized birth control test which uses precision medicine tools to find the most safe and effective birth control measures of an individual's unique body type, helping people avoid side effects including depression or blood clots.
Adyn collects an individual's more than 150 data points relating with lifestyle, other birth control measures, symptoms and medical history. Once the test kit arrives, participants collect a blood sample on the third day of their menstrual cycle. This specific timing provides an accurate snapshot of baseline biology before hormone levels begin to fluctuate. Participants also provide a saliva sample, allowing Adyn researchers to screen over 600,000 genetic markers for predispositions to severe birth control side effects, such as blood clots and clinical depression. Most importantly, instead of relying on the legacy and faulty datasets, Adyn is building its own ‘endogenomic datasets’, essentially mixing the field of endocrinology and genomics together and creating a massive first-of-its kind digital library that tracks how a person’s genes and hormone levels interact with each other. These datasets are actively trying to fix scientific blind spots by gathering massive amounts of biological datasets focused on female reproductive health.
Other companies are taking a similar approach, such as Evvy, a startup focused on gynecological health and the vaginal microbiome. Because traditional tests screen for only a few specific bacteria, more than 50% of vaginal infections are misdiagnosed. Evvy uses metagenomic sequencing to build a vast dataset on the vaginal microbiome and analyze its impact on fertility, preterm birth, and recurring infections.
Bigger Reforms
It remains to be seen how far startups like Adyn and Evvy can go in addressing the misleading datasets leading to misdiagnosis of so many women’s health conditions. “We want to continue to build out our endogenomics dataset to help with conditions far beyond birth control, from menarche to menopause,” Ruzzo said. By pairing proprietary biological data with large-scale peer-reviewed research, the company aims to deliver faster, more accurate diagnoses for under-researched conditions like endometriosis, uterine fibroids, and polycystic ovary syndrome. “When we’re this far behind, we need all the accurate and usable data we can get. Startups have the advantage of moving fast, and often retain the ability to recontact participants, so the data gets more useful over time,” she added.
However, individual startups can only do so much. Overhauling deeply flawed, male-dominated datasets requires systemic reform. Large government institutions have the power, reach, and budgets to take on big problems and collect valuable data without needing to demonstrate immediate outcomes. Bird emphasizes that the inclusion of women should be non-negotiable for any medical condition. To drive real change, the NIH, pharmaceutical companies, and other funders could require clinical trials to enroll women at least in proportion to their share of the disease burden, withholding or cutting funding if those targets aren’t met. “As long as it's not taken seriously enough, we're going to continue to have this problem,” Bird said. Furthermore, scientific journals could mandate that sex-specific results be published in both the abstract and main text, ensuring that findings on women are treated as primary science rather than something optional or secondary.
“For a long time [many] women’s health centers focused more on having nice pink waiting rooms than on having care and expertise for women,” said Bird. “But it certainly takes more for the existing institutions to work to fill those gaps in a healthcare system that was designed more around men's health.”
