For decades, biomedical research has relied on male-dominated preclinical models, overlooking fundamental sex differences that shape disease biology, drug response, and patient safety. This omission has undermined translational success, delayed regulatory approvals, and exposed patients — especially women — to unnecessary risks. As personalized medicine and AI-driven discovery gain momentum, accounting for sex is no longer optional; it is essential to the credibility, efficiency, and fairness of the drug development process. Bridging this gap begins with a new scientific imperative: to embed sex as a cornerstone of experimental design, not an afterthought.
A Missed Variable with Major Consequences
Despite the growing sophistication of biomedical research and the promise of precision medicine, a fundamental biological variable remains routinely overlooked in preclinical study design: sex. Advances in omics technologies, personalized diagnostics, and computational modeling have allowed researchers to stratify patient populations with increasing granularity — yet many preclinical models continue to default to male animals and cell lines, marginalizing sex-based biological differences from the earliest stages of drug development.
This omission has far-reaching implications. Drug efficacy and safety are shaped by numerous factors — metabolism, immune response, gene expression patterns — all of which can differ significantly between sexes. When these distinctions are ignored in preclinical research, it becomes more difficult to anticipate how drugs will perform in diverse patient populations. This contributes to reduced translational success, higher rates of adverse drug reactions, and unequal therapeutic outcomes across sexes. The risks are not theoretical; they are borne out in real-world clinical data and regulatory patterns.
Recognizing this issue, the U.S. National Institutes of Health (NIH) implemented a policy in 2016 requiring researchers to account for sex as a biological variable in all NIH-funded preclinical research, including in the design, analysis, and reporting of data.1,2 This policy shift marked a turning point in how sex differences are treated in biomedical science, moving from an optional consideration to a mandated standard.
Still, the downstream effects of long-standing neglect are visible in clinical development pipelines. Drugs that perform well in male-centric preclinical models can falter in trials involving female participants, or worse, cause unanticipated harm. This contributes not only to costly development failures and regulatory setbacks but also to the ongoing marginalization of women in healthcare innovation.3,4 As the industry moves toward more equitable and effective healthcare, addressing this oversight in the early stages of drug development is no longer optional — it is essential.
A Historical Bias: Why Male Animals and Cells Became the Default
The routine exclusion of females from preclinical research has deep historical roots, shaped by outdated scientific assumptions and institutional inertia. For decades, male animals and cell lines were viewed as the default biological model, leading to a vast body of preclinical literature based largely on male biology — mirroring a bias across all aspects of society, from workplaces through civic planning that has long treated men as the default form of humanity. The bias in animal models can be traced back to mid-20th-century research norms, when early biomedical studies often favored experimental simplicity and assumed generalizability across sexes, despite growing evidence of sex-specific physiology and disease mechanisms.5,6
A prevailing justification for excluding females has been the presumed complexity introduced by hormonal fluctuations across the estrous cycle. Researchers feared that these changes would increase data variability and confound experimental results, especially in behavioral or neurological studies. This belief became entrenched in preclinical practice, discouraging the routine inclusion of females and limiting statistical analyses by sex.2 However, empirical evidence has since challenged this assumption. Analyses of variability across hundreds of studies reveal that female animals are no more variable than males when estrous cycle staging is not controlled — disproving the long-held notion that females inherently complicate study design.2,7
Beyond these scientific myths, practical and economic concerns have also played a role in sustaining the male bias. Including both sexes can double the number of experimental groups, increase resource needs, and lengthen study timelines. For some researchers, this added complexity has been framed as a barrier to innovation rather than a necessary investment in rigor and relevance.8,9 In practice, however, the costs of excluding sex as a variable — measured in failed translation, adverse events, and inequitable health outcomes — far outweigh the marginal resource savings.
The persistence of this male-centric model reflects a broader cultural resistance to change within the biomedical research community. Even as awareness grows, the legacy of historical bias continues to influence study design, funding priorities, and the interpretation of data. Undoing that legacy will require not just new policies, but a fundamental shift in scientific norms and expectations.
What We Miss When We Ignore Sex
Neglecting sex as a biological variable in preclinical research does not simply limit scientific curiosity — it has tangible consequences across disease understanding, drug safety, and therapeutic efficacy. Fundamental differences between male and female physiology extend beyond reproductive organs to nearly every organ system, influencing disease mechanisms and treatment responses in ways that cannot be inferred from male-only models.
Sex-specific patterns are well documented across multiple disease areas. In cardiovascular disease, women often present different symptoms, risk profiles, and disease progression than men, yet many of the current diagnostic and therapeutic strategies were developed based on male-centric data.10,11 Neurodegenerative diseases, such as Alzheimer’s, are more prevalent in women, yet preclinical models of these disorders continue to be male-dominated, limiting insights into sex-linked pathophysiology.12 In autoimmune diseases, which disproportionately affect women, differences in immune system architecture and gene regulation play critical roles that are not captured in male-only studies. Even in the domain of pain, male and female animals have been shown to rely on different immune cell types for pain processing, suggesting distinct mechanistic pathways that should inform the development of analgesics.10
These biological differences also extend to pharmacokinetics and pharmacodynamics — how drugs are absorbed, metabolized, distributed, and eliminated by the body. Women experience adverse drug reactions more frequently and more severely than men, in part due to differences in body composition, enzyme activity, renal clearance, and hormone levels.2,8,13 These discrepancies often go unrecognized until after a drug reaches market, sometimes with serious consequences.
A striking example is zolpidem (best known as the brand name Ambien), a widely prescribed sedative–hypnotic. Years after its approval, data revealed that women exhibited significantly higher blood concentrations of the drug the morning after use, increasing the risk of impaired alertness and accidents. This led the U.S. Food and Drug Administration (FDA) to issue a sex-specific dose adjustment in 2013 — long after millions of women had taken the drug at a level calibrated for male metabolism.2 Such late-stage corrections highlight the cost of omitting sex-based analysis from earlier phases of research and development.
Even in fields like oncology and immunology, where personalization is a growing priority, the role of sex is still underappreciated. Research has shown that women and men may respond differently to targeted therapies and immunotherapies due to differences in immune regulation, hormone signaling, and genetic architecture.14,15 Ignoring these variables can undermine both the effectiveness and safety of treatments, while reinforcing inequities in access to optimal care.
The failure to systematically include and analyze sex in preclinical models obscures important biological signals, distorts our understanding of disease, and jeopardizes patient safety. A truly evidence-based approach to medicine must begin with recognizing that male and female biology cannot be treated as interchangeable.
Progress in Policy and Practice — But Not Fast Enough
In recent years, there has been a growing institutional recognition of the need to address sex bias in biomedical research. Key policy changes have begun to shift the landscape, most notably the NIH’s 2016 mandate requiring researchers to account for sex as a biological variable (SABV) in the design, analysis, and reporting of preclinical studies funded by the agency. This directive, aimed at correcting longstanding imbalances in scientific inquiry, marked a significant departure from earlier norms in which male models were considered sufficient by default.1 The FDA has also issued guidance emphasizing the evaluation of sex differences in clinical investigations, reinforcing the principle that medical innovation must be grounded in biological realism.16
These policy interventions have led to measurable — if uneven — progress. The proportion of studies including both male and female subjects has increased across many biomedical fields. However, inclusion alone is not enough. Most of these studies still fail to conduct sex-disaggregated statistical analyses, instead combining data from both sexes and reporting only aggregate outcomes.2 This practice can mask sex-specific effects and perpetuate the very biases that SABV policies were designed to address.
A review of the literature over a 10-year period revealed that while nearly half of studies in some disciplines now include both sexes, fewer than half of those perform any sex-based statistical analysis. Pharmacology stands out as a rare exception, where rates of such analyses have risen meaningfully in response to editorial pressure and evolving standards.17 In many other areas, the absence of detailed sex-based analysis remains the norm, limiting the interpretability and impact of otherwise inclusive research designs.
Efforts to promote sex equity in research have also taken shape outside the United States. The European Medicines Agency (EMA) issued guidance in 2005 encouraging greater inclusion of women in clinical trials, while the European Parliament adopted a resolution in 2017 urging stronger sex-based considerations in mental health and biomedical research.2 Despite these moves, enforcement has been inconsistent, and implementation varies widely across member states and therapeutic areas.
Participation-to-prevalence ratios (PPR) offer a quantitative lens on these disparities. Ideally, the proportion of women in clinical trials should match their representation among patients with the disease in question—a PPR between 0.8 and 1.2 is considered equitable. However, in areas like heart failure (PPR = 0.27), acute coronary syndrome (PPR = 0.51), and diabetes-related lipid trials (PPR = 0.74), women remain significantly underrepresented.2 Stroke trials also reveal persistent gaps, with a PPR of just 0.73 despite the high burden of disease in women.2
These data underscore a critical disconnect: while policies have begun to change the conversation, practice has not yet caught up. Inclusion without meaningful analysis is performative. Real progress requires structural incentives, cultural change, and the widespread adoption of methodological rigor that treats sex not as a nuisance variable but as an essential dimension of biomedical research.
The Translational Gap: From Bench to Bedside
The omission of sex as a biological variable in preclinical research does not remain confined to the lab — it follows drugs into the clinic, shaping trial outcomes, regulatory decisions, and ultimately, patient safety. A preclinical pipeline dominated by male models creates blind spots that may only be revealed during human trials, sometimes with damaging or even dangerous consequences. By that point, millions of dollars and years of effort may have been invested in compounds that are less effective — or more harmful — for half the population.3,18
One key example lies in the field of neuropsychiatry, where the efficacy of drugs such as antidepressants and antipsychotics has shown variation between sexes. Selective serotonin reuptake inhibitors often demonstrate stronger effects in women, while tricyclic antidepressants appear more efficacious in men.2 Similarly, men and women may respond differently to antipsychotic drugs, with varying side effect profiles and metabolic responses, yet most preclinical studies of these agents fail to model both sexes, and few clinical trials are powered to assess these distinctions.2
In oncology, sex-specific immune regulation and hormonal influences affect both disease progression and treatment response. Immune checkpoint inhibitors and targeted therapies show differential efficacy across sexes in some cancers, yet preclinical cancer models and early-stage studies have largely neglected these dynamics. Ignoring them not only impairs clinical trial design but risks deploying suboptimal therapies to large patient populations.2,19
The costs of this translational failure are steep. Drugs may fail in phase III trials due to unanticipated variation in response between male and female participants, wasting critical resources and delaying the availability of effective treatments. Even when approved, sex-blind development can result in higher rates of adverse events, post-market safety warnings, or withdrawal from the market altogether — undermining public trust in both regulators and manufacturers. These failures are not merely technical errors; they represent systemic inefficiencies rooted in outdated assumptions and avoidable omissions.
Conversely, when preclinical studies include and analyze both sexes from the outset, they lay the groundwork for more predictive models, better-calibrated dosing strategies, and more inclusive clinical trial designs. This approach not only improves the likelihood of success in later phases but also aligns drug development with the broader goals of personalized medicine and regulatory transparency.2,4 Bridging the bench-to-bedside gap begins with recognizing that sex differences are not confounding variables to be minimized, but fundamental aspects of biology that must be actively studied and understood.
A New Scientific Imperative: Designing Research for Real-World Relevance
If biomedical research is to meet the demands of modern medicine, it must reflect the full diversity of human biology — not just in theory but in design, execution, and analysis. The prevailing one-size-fits-all approach to preclinical experimentation has proven inadequate, yielding results that often fail to replicate in clinical settings. A shift toward more representative, sex-informed research practices is not just a matter of equity; it is a scientific necessity.
One of the most effective strategies for incorporating sex into preclinical studies is the use of factorial designs. These experimental models allow researchers to examine interactions between variables — such as sex and treatment — without requiring fully separate studies for each subgroup. In doing so, they maximize the information gained from each cohort and enable detection of sex-based differences in efficacy or toxicity that would otherwise go unnoticed. Far from being an inefficient burden, this approach can increase both the power and precision of research while improving translational success rates.2,9
Advances in data analysis and preclinical modeling now make it easier than ever to explore sex-based differences at a systems level. Platforms designed for managing and interpreting preclinical data increasingly allow for sex to be included as a standard parameter, facilitating more robust stratification and modeling.20 These tools not only support better study design but also promote a culture of accountability, making it harder to justify the omission of sex-based considerations in data reporting and regulatory submissions.
High-resolution molecular data has further expanded our understanding of how sex influences biology. Transcriptomic studies, which examine the complete set of RNA transcripts produced by the genome, have revealed extensive sex-linked differences in gene expression, particularly in tissues such as the liver, heart, and brain. In fields like pain and inflammation, these differences affect everything from immune signaling to receptor expression, with direct implications for drug responsiveness and side effect profiles.11,15 Incorporating sex-based transcriptomic data into preclinical assessments could help identify biomarkers, predict adverse events, and guide therapeutic development with greater precision.
Emerging best practices reflect a growing consensus within the scientific community. Researchers are increasingly expected to report the sex of animals and cells used, employ balanced cohorts that include both males and females, and power their studies to detect sex-based effects. While these practices are still far from universal, they are gaining traction in response to editorial standards, funding requirements, and institutional policy shifts.8,17
The tools, frameworks, and justifications for sex-inclusive research already exist. What is required now is a widespread commitment to using them — not as a box-checking exercise, but as a core element of scientific rigor. Real-world relevance demands that our models reflect real-world biology, and that begins with designing studies that account for the full spectrum of human difference.
Toward a Sex-Inclusive Biomedical Future
Creating a biomedical research ecosystem that truly accounts for sex differences requires more than updated policies or individual good intentions — it demands systemic accountability across every stakeholder group. Publishers, funding agencies, contract research organizations (CROs), contract development and manufacturing organizations (CDMOs), and pharmaceutical companies all play essential roles in shifting the paradigm from male-dominated research toward truly sex-inclusive science.
Journals can set higher standards by requiring transparent reporting of the sex of biological materials and rejecting manuscripts that omit sex-based data without justification. Funders can mandate sex as a biological variable not only in study design but also in analysis plans and statistical powering. CROs and CDMOs, as key players in outsourced preclinical work, can integrate sex balance into their default protocols, ensuring that sponsors are offered sex-informed study designs and encouraged to consider sex-specific endpoints. Pharmaceutical companies, which bear the costs of failed translation and post-market safety issues, have perhaps the strongest incentive to demand more representative models that better predict clinical outcomes.
But equity in biomedical research does not end with inclusion. Inclusion without analysis is insufficient and, at times, misleading. Simply enrolling female animals or patients means little if their data are not analyzed and interpreted separately. True equity requires that research be powered to detect sex-based differences, and that findings be disaggregated and transparently reported. This shift represents a move from superficial diversity toward meaningful, mechanistic understanding.
To that end, there is an urgent need for innovation in how we model sex-based disease mechanisms and treatment responses. Preclinical research must evolve to include models that better reflect the complexity of human sex differences — at the molecular, cellular, systemic, and behavioral levels. This includes not only using both sexes in animal models but also designing studies that account for hormonal cycles, reproductive status, and sex-specific gene expression. New in vitro platforms, such as organoids and tissue chips, offer promising avenues for studying sex differences in a controlled yet physiologically relevant manner.
As the life sciences industry increasingly embraces personalized medicine, the importance of sex-based research becomes even more pronounced. The use of artificial intelligence (AI) and machine learning to identify biomarkers, predict treatment responses, and optimize clinical trial design is only as good as the data on which these models are trained.21 If those data sets are skewed toward one sex, the resulting algorithms risk perpetuating existing disparities rather than mitigating them. Ensuring sex-balanced data inputs is thus essential for the validity of next-generation tools and the therapies they help shape.
Regulators have begun to recognize these challenges, and while guidance is evolving, the onus remains on industry and academia to lead. A sex-inclusive biomedical future is not just an ethical or political goal — it is a scientific imperative. Achieving it will require collaborative effort, sustained pressure, and a willingness to redesign the assumptions that have guided biomedical research for decades. The rewards, however, are substantial: safer, more effective therapies for everyone.
Conclusion: Precision Starts with Preclinical Rigor
Sex is not a confounding variable to be controlled or ignored — it is a fundamental dimension of biology that shapes health, disease, and therapeutic response. Decades of male-centric preclinical research have left critical blind spots in our understanding of human physiology, undermining the promise of precision medicine and contributing to avoidable failures in drug development. These oversights have real consequences, from regulatory setbacks to adverse events, and they disproportionately harm populations that have historically been underrepresented in biomedical research.
Correcting this imbalance begins with preclinical rigor. When researchers proactively account for sex differences in their models — designing studies that include both sexes, powering analyses accordingly, and interpreting results through a sex-informed lens — they lay a stronger foundation for clinical translation. These efforts not only improve scientific accuracy but also contribute to more equitable health outcomes by ensuring that therapies are safe and effective across the full spectrum of patients.
A future in which drug development consistently reflects real-world biology is within reach but only if sex differences are treated not as an afterthought, but as a central consideration from the outset. Achieving this will require alignment across the research ecosystem, from funders and publishers to CROs, CDMOs, regulators, and innovators. At stake is not only the efficiency of the drug development pipeline, but the integrity of biomedical science and the trust of the patients it aims to serve. Precision starts with the models we choose to build—and that means building with both sexes in mind.
References
1. Justice, Monica J. “Sex matters in preclinical research.” Dis. Model Mech. 17: dmm050759 (2024).
2. Barus, Romain, et al. “Sex differences: From preclinical pharmacology to clinical pharmacology.” Therapies. 78: 189–194 (2022).
3. Rosch, Anke. “Sex Matters… In Preclinical Drug Development?” Clinical Leader. 18 Jan. 2022.
4. Mauvais-Jarvis, Franck, Arthur P Arnold, and Karen Reue. “A Guide for the Design of Pre-clinical Studies on Sex Differences in Metabolism.” Cell Metabolism. 25: 1216–1230 (2017).
5. Lee, Heisook, et al. “It is time to integrate sex as a variable in preclinical and clinical studies.” Experimental & Molecular Medicine. 50: 1–2 (2018).
6. McGrath, Nicole. “The Impact of New NIH Requirements on the Preclinical Research The Impact of New NIH Requirements on the Preclinical Research Sex Disparity --- A Meta-Analysis.” Honors College. 501 (2019).
7. Mir, Franco Rafael and María Angélica Rivarola. “Sex Differences in Anxiety and Depression: What Can (and Cannot) Preclinical Studies Tell Us?” Sexes. 3: 141–163 (2022).
8. Zucker, Irving and Brian J Prendergast. “Sex differences in pharmacokinetics predict adverse drug reactions in women.” Biology of Sex Differences. 11: 32 (2020).
9. Buoncervello, Maria, et al. “Preclinical models in the study of sex differences.” Clinical Science. 131: 339–469 (2017).
10. Olivera, Sol and Delyth Graham. “Sex differences in preclinical models of hypertension.” Journal of Human Huypertension. 37: 619–625 (2023).
11. Ramirez, F Daniel, et al. “Sex Bias Is Increasingly Prevalent in Preclinical Cardiovascular Research: Implications for Translational Medicine and Health Equity for Women: A Systematic Assessment of Leading Cardiovascular Journals Over a 10-Year Period.” Circulation. 135: 6 (2017).
12. Voskuhl, Rhonda. “Preclinical studies of sex differences: A clinical perspective.” Biology of Sex Differences. Dec. 2016.
13. Hilmer, Sarah and Paula Rochon. “Sex and Age Differences in Geriatric Pharmacotherapy.” Public Policy & Aging Report. 33: 132–135 (2023).
14. Cattaneo, Annamaria, et al. “Recommendations for the Application of Sex and Gender Medicine in Preclinical, Epidemiological and Clinical Research.” Journal of Personalized Medicine. 14: 908 (2024).
15. Guma, Elisa, et al. “Comparative neuroimaging of sex differences in human and mouse brain anatomy.” eLife. 15 Mar. 2024.
16. Evaluation of Sex Differences in Clinical Investigations: Guidance for Industry. U.S. Food and Drug Administration. Jan. 2025.
17. Karp, Natasha A and Neil Reavey. “Sex bias in preclinical research and an exploration of how to change the status quo.” British Journal of Pharmacology. 12 Nov. 2018.
18. Tosi, Laura L, et al. “Sex and Gender Differences in Clinical and Preclinical Studies.” Musculoskeletal Key. Accessed 15 May 2025.
19. Allegra, Sarah, et al. “Evaluation of Sex Differences in Preclinical Pharmacology Research: How Far Is Left to Go?” Pharmaceuticals (Basel) 15: 786 (2023).
20. Jacobson, Moriah L. “The Issue of Sex vs. Gender in Preclinical Animal Model Studies.” Taconic Biosciecnes. 23 Mar. 2021.
21. Alvaro, David. “The Gender Bias Built Into AI — And Its Threat to Women’s Health.” Pharma’s Almanac. 14 Apr. 2025.