An Inherent Imbalance
The inclusion of female subjects in biomedical research has been officially encouraged only since 1986.1 Before then, male organisms and cells predominated in clinical and preclinical studies. Decades of overlooking sex differences in biology have resulted in critical missteps in medical treatments, most infamously demonstrated by the tragedy surrounding thalidomide. It was prescribed in the late 1950s for morning sickness in pregnant women, but resulted in congenital abnormalities.2 Additionally, it is well-documented that women experience adverse drug reactions more frequently than men, though the underlying reasons remain poorly understood.2 Even now, many preclinical studies continue to use predominantly male models, including in cases where the intended human recipients of a drug are predominantly women.
Gender Bias in In Vivo Studies
Consideration of sex as a biological variable in animal research has historically been minimal, though recently gaining traction. Rats and mice are the most commonly used preclinical models in biological research. Between 1990 and 2009, 80% of in vivo studies used exclusively male rodents, with research in neuroscience, pharmacology, and physiology having the most pronounced male biases.3 More recently, a meta-analysis revealed that among studies specifying the sex of animals, only 16% included female models, while male models make up a disproportionate 84%.1 Even in disorders predominantly affecting women, only 12% of studies used female or mixed-sex models.4 Clearly, the historical gender bias remains significant and troubling.
Researchers cite that the main reason for excluding female animal models is the hormonal fluctuations associated with the rodents’ estrous cycle.2 This cycle, characterized by extreme shifts in progesterone and estrogen levels over 4–5 days, introduces unpredictability into experiment outcomes.3 Due to the complexity and variability attributed to these hormonal fluctuations, female animals have been excluded from preclinical studies.
Nevertheless, researchers attempting to mitigate hormonal influences have adopted strategies such as conducting experiments specifically during the diestrus phase, when reproductive hormone levels are relatively stable in female rodents.2 Even so, to adequately account for hormonal variation, studies typically require up to four times as many female animals as males, which contradicts the widely accepted 3Rs principle (Replacement, Reduction, and Refinement) for ethical animal research. Alternative strategies include creating genetically modified menopausal mouse models (e.g., Foxl2–/– mice) or estrogen receptor-deficient mice (ERα knockout).3 These methods are emerging but not widely used. Therefore, these limitations reflect ongoing challenges in addressing sex-based variability without compromising ethical research principles.
Gender Bias in In Vitro Studies
Gender bias extends beyond animal models and persists notably in cell-based studies. Historically, cell cultures in vitro were treated as asexual, causing many researchers to overlook the influence of sex at the cellular level. A substantial proportion of cell lines used in research remain unspecified in terms of biological sex. Moreover, approximately 15.5% of commercially available human cell lines lack clear sex identification, and this is even more lacking among animal and stem cell lines.1 A systematic review further highlighted this disparity, revealing that 69% of cell-based studies used exclusively male cell lines.1 This raises an essential question: Can we continue to disregard cellular sex without compromising our research findings?
Consequences of Overlooking Sex Differences
Inconsistencies in Data Interpretation
Failure to consider sex differences critically undermines the validity of research data. Often, studies employing both sexes pool the data and assume no meaningful differences exist between male and female subjects. Statistical methods commonly used, such as Student’s t-tests, assume samples originate from a homogenous population.4 However, this practice can obscure significant biological differences between sexes, leading to misinterpretations.
For example, one simulation study explicitly demonstrated that a treatment produced significantly greater effects in female animals than males.4 This effect became insignificant when the data was combined across sexes. This reveals that data pooling might not only reflect an assumption of negligible sex differences but may also highlight researchers' unfamiliarity with handling sex as a biological variable in data interpretation. Given that statistical errors in research papers are not uncommon, developing methodological rigor around sex-specific analyses is essential to generating accurate results.
Moreover, inconsistencies persist in reporting sex-specific information. Twenty-two percent of publications failed to specify animal sex altogether and yet, rigorous documentation of sex in research is critical.3 By clearly reporting data separately by sex, researchers can avoid inadvertently masking valuable insights, enhancing the reliability of scientific findings.
Poor Translation of Preclinical Findings to Humans
Ignoring sex differences in preclinical research also contributes to poor translational outcomes in clinical medicine. Alarmingly, surveys indicate that 88% of preclinical studies targeting disorders predominantly affecting women still relied exclusively on male animals.2 Such oversight directly impacts human health: between 1997 and 2000, eight out of 10 drugs withdrawn from the U.S. market were removed because they posed significant risks to women, risks that were not adequately predicted by preclinical research.2
Additionally, inherent biological and physiological differences between males and females — such as body mass, metabolism, and behavioral traits — significantly influence disease presentations. Prominent examples include sex disparities in cardiovascular and autoimmune diseases, which disproportionately affect women.1 The persistent underrepresentation of female models hampers our understanding of women's biology, limiting our ability to predict therapeutic efficacy accurately. Given that women constitute approximately half of the global population, neglecting female representation in preclinical research will have significant consequences for clinical outcomes.
Moving Forward: Addressing the Bias
Addressing gender bias in preclinical research is more than just a methodological adjustment, it represents a necessary ethical and scientific step toward more accurate and clinically meaningful research.
To advance effectively, separating data analysis by sex should become a routine practice. Data from male and female models should be analyzed independently before pooling to prevent masking meaningful gender-specific results.4 Additionally, clearly reporting the sex of animals, cells, or organoids within experimental designs is crucial, especially when studying X- or Y-chromosome linked disorders. In 2016, the Gender Policy Committee of the European Association of Science Editors (EASE) released international guidelines advocating sex reporting as a biological variable.3 Researchers could consult these guidelines and determine their applicability to their work, thereby contributing to the reduction of gender biases in scientific research.
Arguments that incorporating female animal models increases costs or contradicts the 3Rs should be carefully reconsidered. There is substantial evidence of sex as a fundamental biological variable significantly impacting research outcomes.4 While it is true that including female models incur extra costs, the long-term financial and scientific implications of neglecting sex differences are far greater. Neglecting these differences often leads to costly failures in clinical trials and drug withdrawals from the market. Therefore, one can argue that adopting gender inclusivity will reduce costs, enhance reproducibility, and improve translational success in the long run.
Acknowledging and systematically incorporating sex differences in both in vivo and in vitro studies will greatly enhance the translational value of scientific outcomes. As researchers become more aware of this issue, establishing rigorous standards for sex-based analysis and reporting must become an industry-wide norm rather than an afterthought. Only through changes in research design and reporting standards can we adequately address the historical inequities in preclinical research and improve healthcare outcomes for everyone, irrespective of sex.
References
1. Yoon, D.Y. et al. “Sex bias exists in basic science and translational surgical research.” Surgery. 156: 508–516 (2014).
2. Justice, M.J. “Sex matters in preclinical research.” Disease Models & Mechanisms. 17. (2024)
3. Allegra, S. et al. “Evaluation of Sex Differences in Preclinical Pharmacology Research: How Far Is Left to Go?” Pharmaceuticals. 16: 786–786 (2023).
4. Lee, S.K. “Sex as an important biological variable in biomedical research.” BMB Reports. 51: 167–173 (2018).