Mind the Gap: how overlooking sex compromises research 

What shapes the way each of us responds to the world around us? One influential biological variable is sex: the combination of chromosomes, hormones, and reproductive anatomy that governs both development and adult physiology. This is a core feature of human biology, with impacts on health and disease that are still being uncovered. 

Historically, most research on the human body, and analogous work in mouse models, was conducted exclusively on males. Clinical trials in the United States were not mandated to include women until 1993, and currently still frequently fail to integrate sex as a covariable in statistical analyses. We now recognize that X and Y chromosomes do not only cause differences in baseline biology, but in an individual’s experience of disease and their responses to therapies. For example, many diseases disproportionately affect females. Systemic Lupus Erythematosus (SLE), Multiple sclerosis (MS), Rheumatoid arthritis (RA), scleroderma, and Graves’ disease are just a few examples of autoimmune disorders that affect far more females than males, at ratios ranging from 3:1 in the case of MS to as high as 15:1 in SLE. Some drugs have been shown to have sex-based differences in their pharmacokinetics, clinical effects, and rates of side effects. 

Despite these striking differences, biomedical research still often oversimplifies, or even avoids altogether, the impact of sex hormones and menstrual (or in mice, estrous) cycles and their impact on disease. Mouse studies often only include one sex in the name of maintaining rigorous control, operating on the assumption that including both sexes introduces “hormonal noise” that may muddy conclusions. But taking this approach overlooks the fundamental reality that these hormones and cycles are not “noise”; rather, they are meaningful factors that are critical to consider in research. 

To understand why, it helps to look at what these hormones actually do. The ovaries and uterus are complex organs that produce the hormones driving the menstrual cycle in humans and the estrous cycle in mice. The sex hormones estrogen and progesterone fluctuate and spike in characteristic patterns, and their receptors are expressed throughout the body, meaning their effects extend well beyond the reproductive organs, producing profound systemic changes. 

The immune system sensitive to these hormones. Immune cells are dynamically regulated by estrogen and progesterone signals, and hormone receptors expressed on immune cells shifts across the cycle, making their sensitivity to changing hormone levels a dynamic factor. During the follicular phase of the menstrual cycle, when estrogen peaks, antibody levels rise and immune cells shift toward a pro-inflammatory state. Later, during the luteal phase, progesterone drives anti-inflammatory effects and mild immunosuppression. 

The mouse estrous cycle is shorter – phases occur in hours rather than days or weeks like humans. Estrogen spikes during the proestrus are associated with heightened inflammatory activity in both innate and adaptive immune cells, whereas progesterone peaks during diestrus promotes immunosuppression. Being able to model these aspects of the human menstrual cycle in laboratory mice is massively valuable, as it allows the effect of sex hormones on immune function to be studied in a controlled system, even though mice will never fully reflect human physiology (for example, mice do not have a “period”, the bleeding associated with uterine lining shedding). Regardless, these rapid hormonal oscillations are not a minor detail, can meaningfully influence experimental outcomes in mouse models, and still need to be treated as an important consideration in research.  

What we’re missing 

When hormonal cycles are not factored into study design, the regulation of immune cells by these hormones becomes an unmonitored variable. Conclusions from mouse studies that do not take this into consideration could potentially be incomplete and not fully capture biology that exists in half of the population, which not only exacerbates the already widespread data reproducibility crisis that preclinical studies face, but also increases the likelihood that preclinical findings will fail in human clinical trials. 

Autoimmune diseases provide a clear illustration of how profoundly hormone cycles can impact disease. In mouse models of MS, estrogen and progesterone have clear impacts on disease severity. It is well established that MS flares drop significantly during pregnancy, when estrogen levels are at their peak. Despite recognizing these patterns, however, many mouse studies exclusively used only male mice, and few studies take the estrous cycle into account when conducting studies in female mice. 

A similar pattern emerges in vaccine responses. Female mice generate a stronger antibody response to a vaccine when the vaccine is administered during the high estrogen phase of the estrous cycle as opposed to during the high progesterone phase. A recent MedRxiv preprint reports a similar pattern in humans: people who received a COVID-19 vaccination during their follicular phase reported more side effects but were also slightly better protected from later infection when compared to people who received the vaccination during the luteal phase. Although this study has yet to be peer reviewed, it underscores the need to treat hormone cycles as biologically relevant variables rather than confounders. 

Taken together, these examples show that hormonal cycles are not a source of noise that need to be controlled. Research designs that ignore them do not eliminate that variation; they simply leave it unexamined. 

Closing the gap 

A relatively recent shift, funding agencies and academic journals are recognizing the importance of considering both sexes and hormonal cycles in research. In 2016, the National Institutes of Health in the United States implemented the Sex as a Biological Variable (SABV) policy. The goal of the policy is not only to identify sex differences in physiology and disease, but to consider sex as a variable that is incorporated into experimental design and interpretation of the results. 

Ultimately, ignoring these variables means ignoring fundamental biology of half of the global population, an oversight that does and will have consequences for medical advancement. Studying how hormones can impact immunity transcends sex binaries, with far-reaching implications for people who need hormone therapy, whether for endocrine disorders, menopause, or gender-affirming care. As research moves forward, integrating sex-related differences in study design is essential for building biomedical insights that reflect the full diversity of human physiology.