In two new documents, FDA offers developers guidance on assessing sex-based differences

In early January 2025, the FDA issued two new draft guidance documents focused on gender representation in medical product research. Both documents provide recommendations on assessments of sex-based differences, primarily emphasizing inclusion of female participants. One document covers all medical products, proposing updates to policy from 1993, while the second is specific to devices and would update 2014 guidance.

Diversity in clinical research for medical products

• Historically, clinical research has not been the most representative. Although the FDA’s mandate directs the agency to ensure data submitted to support marketing of medical products represents the actual intended use population, certain groups have not been well represented. There has been a lack of representation of racial and ethnic groups in research populations, particularly populations who identify as Black/African-American, Hispanic/Latino, Indigenous and Native American, Asian, Native Hawaiian and Other Pacific Islanders, with white participants over-represented. Other demographic groups are also more likely to be underrepresented, including older individuals, pediatric populations, people with disabilities, and females. [See AgencyIQ’s historical primer on FDA policy related to diversity over the years.]
• After an April 2022 draft guidance, and new statutory authority in 2023, the FDA issued a new draft guidance on Diversity Action Plans (DAPs) in June 2024. DAPs are plans, developed by product sponsors, that include specific enrollment goals, the rationale by which these goals were developed, and “an explanation of how the sponsor intends to meet such goals.” The statutory provision granting FDA authority to require these plans as part of certain submissions also allows for certain research programs to be exempted from the DAP requirement through a waiver system, which could be initiated by the FDA or “at the request of a sponsor.” The 2024 draft guidance describes the various elements of a DAP (enrollment targets, rationale to justify those targets), the basics of their formatting, and how to submit a DAP to the FDA – including both the information they expect and when/how those documents should be sent in. [Read AgencyIQ’s analysis of this guidance here.]
• That guidance focused on enrollment of underrepresented racial and ethnic groups, defining underrepresented populations as: “Generally, for race and ethnicity, underrepresented populations may typically include participants who are Black or African Americans, Hispanic/Latinos, Indigenous and Native American, Asian, Native Hawaiian and Other Pacific Islanders, and other persons of color.”
• However, it also discussed underrepresentation in terms of other demographic categories: “Generally, for sex and age, underrepresented populations may typically include participants who are females and in the older adult and pediatric age groups, respectively.”
• Sex- and gender-based representation: According to a 2021 commentary published by the FDA (Office of Women’s Health (OWH), Office of the Commissioner), clinical researchers have increased representation of women in clinical trials: “72% of clinical trial participants were female” in an analysis of the Center for Drug Evaluation and Research (CDER) Drug Trial Snapshots from 2019. However, female participant inclusion varies by therapeutic area; that report from OWH found that female participants are still underrepresented in key areas including research into infectious disease and cardiovascular products.

The agency has now released a new draft guidance document on the inclusion of female research participants

• Up front: What’s the scope of this document? The new draft guidance is from all the FDA’s medical product centers –CDER, the Center for Biologics Evaluation and Research (CBER) and the Center for Devices and Radiologic Health (CDRH) – and applies to clinical research programs for all of these product types. This new draft guidance would, if/when finalized, replace this policy from 1993. However, the guidance primarily refers to Investigational New Drug (IND) investigations, with limited recommendations for devices – but the agency has also issued, in conjunction, a device-specific guidance on this same topic (see below for an analysis of that guidance).
• Some quick definitions: Differences between sex and gender. The new draft guidance opens with definitions of these terms, adapted using definitions from the National Institutes of Health (NIH) and Academies of Science, Engineering and Medicine (NASEM). Sex is a “biological construct based on anatomical, physiological, hormonal, and genetic (chromosomal) traits. Sex is generally assigned based on anatomy at birth and is usually categorized as female or male, but variations occur. Variations of sex refers to differences in sex development or intersex traits.” Relatedly, the FDA is defining gender as “a multidimensional construct that encompasses how an individual self identifies. Gender may be described across a continuum, may be nonbinary, and may change over the course of a lifetime. Gender may or may not correspond to a person’s sex assigned at birth.” For most people, “sex and gender are concordant, but FDA recognizes that sex and gender are not always concordant.”
• For the purposes of FDA’s guidance documents, and this analysis, “female” and “male” refer to sex, while “women” and “men” refer to gender.
• The guidance is focused on sex differences, as “biological differences… can impact outcomes in clinical trials and non-interventional studies.” Still, the agency also states that it “encourages sponsors to consider whether gender differences are relevant” to their studies. As females have been historically underrepresented in clinical research, the recommendations in the guidance focus on representation of females – notably, this also aligns with the DAP guidance document’s assertion that female participants should be considered in the definition of “underrepresented” populations.
• The guidance covers some historical precedent, citing the FDA’s work on the Demographics Rule in the late 1990s and then updates to the Investigational New Drug (IND) regulations in 2000 that established a system whereby the FDA can place a clinical hold on an investigation if people of reproductive potential (males or females) are being excluded from eligibility – although there are some carve-outs to that policy. Still, the agency argues that despite better inclusion of female participants in research overall, “female participants remain underrepresented in clinical trials for some therapeutic areas where the disease or condition affect both males and females.” The draft guidance also makes a (well-established) case as to why sex differences should be considered in research, including the potential for differential efficacy or safety considerations that could inform labeling, differences in pharmacokinetics/pharmacodynamics (PK/PD) because of physiological, anatomical or genetic factors, and the hormonal dynamics that differ with sex.
• The agency offers some high-level thinking on clinical trial design and conduct for enhancing representation of female participants. The agency recommends, in general, that sponsors ensure their trial population reflects the population for the disease or condition for which the product is being investigated, with a focus on “the underlying biology of the disease or condition to anticipate sex differences in PK, PD, and safety and effectiveness.” While participants should self-report sex information, “generally based on their sex assigned at birth,” the trial data should use the data standard definitions from the Clinical Data Interchange Standards Consortium (CDISC), as is standard practice. While the FDA does not require gender data, and CDISC does not offer this data field, sponsors may still include this information “if gender may influence the outcome of interest.”
• The policy recommendations in this guidance are broken into three categories: Clinical trial design and conduct, statistical concepts, and non-clinical considerations. There is also a section on “other general considerations” as a catch-all.
• For clinical trial design and conduct, the agency offers recommendations on recruitment, enrollment and retention, trial design, and on the enrollment of people who are pregnant or lactating. On recruitment, enrollment and retention, the agency recommends that sponsors evaluate their demographic distribution of potential participants at key time points, such as screening, inclusion/exclusion criteria and after consent, and assess how the population looks at these points to see if adjustments need to be made. As in other guidance documents focused on representation, the agency urges sponsors to engage with institutions or organizations that reflect the demographic groups of interest, in this case women’s groups, and also urges flexibility for participants (e.g., mobile health, digital health technologies, scheduling, support services for participants) and a focus on specific population needs and contexts, such as enrollment of women of different ages, races, hormonal statuses (e.g., menopausal) and people with reproductive potential “with appropriate risk mitigation efforts.”
• For trial design, “for most drugs and devices,” the participant population should be powered to “allow for reliable benefit-risk assessments and to understand any potential sex-related differences in medical product response.” The agency cites the diversity action plan (DAP) policy here and says that sponsors working on their DAP should consider additional recommendations in this draft guidance specifically for sex differences. Similar to the DAP guidance, if there is a “plausible biological reason to expect a different response in females and males,” then the target enrollment of the program should reflect that possibility. Notably, the DAP guidance includes female participants as those in the “underrepresented” category (see footnote 10 of that guidance).
• Specifically for pregnant and lactating individuals, the agency offers some high-level thinking, urging developers to consider the benefits and risks of inclusion of these patient populations at different stages of the research program. When there is a scientific justification for exclusion of pregnant participants, the agency recommends sponsors “consider including PK sampling to inform drug dosing in participants who become pregnant during a trial and can safely remain in the trial,” although it does acknowledge that this is not always possible – in which case “it can be informative to collect relevant PK data even when the investigational medical product is discontinued.”
• Statistical concepts: An overview. The assessment of potential sex differences in safety or efficacy of a medical product informs the product’s labeling, including ensuring appropriate characterization of treatment effects or differential efficacy and determining risks and benefits for patient groups. Still, “the optimal analysis approach will depend on the type of inference that is sought,” the agency states, going over some examples based on demographic characteristics, and asks sponsors to assess treatment effects by sex for both effectiveness and safety endpoints.
• Analyses for differences should include information about uncertainty. The agency states that “statistical tests for detecting plausible magnitudes of differences in treatment effects by sex (i.e., tests of a treatment-by-sex interaction) tend to be underpowered,” and that a lack of statistical significance in an underpowered study “is not absence of a clinically meaningful difference.” When the population is sufficiently powered, the agency also cautions sponsors about incorrect conclusions based on the evaluation of differences beyond sex along (e.g., race, age, comorbidities). The guidance also recommends that sponsors use “data from all participants when estimating the treatment effect within a given sex,” rather than the data from a single sex. Per the agency, “Estimators of sex-specific treatment effects have greater precision than estimators based solely on the data for a given sex,” and subgroup analysis beyond sex (age, race) might lead to “random highs and lows.”
• What if there are differences expected? In the design stage, sponsors “should prespecify statistical analyses for evaluating and reporting differences in treatment effects between females and males,” and if there’s an anticipated difference between effects for the different sexes, “an early-phase trial (e.g., phase 1/phase 2) should ideally be performed to obtain information on differences by sex.” Pivotal trials can also be designed to account for these differences if evidence from early-stage trials points to “benefit for a particular sex and uncertainty around benefit for another sex” – but the agency urges sponsors to consider designing studies to mitigate risks of Type I error.
• Reporting results: The agency recommends including providing information about the demographics, including sex, of research participants and in marketing submissions – as is current practice.
• For nonclinical testing, the agency recommends that animal studies “use adequate numbers of male and female animals to permit the identification of any sex-based differences in toxicity or other safety assessments” as applicable.
• And finally, “Other General Considerations.” This section includes various general recommendations of what the agency would be looking for – or what sponsors should account for – when considering sex-based differences in their investigational products. Some of these bullets are still high-priority policy, such as the idea that FDA “can require a postmarketing study when applicable criteria are met,” including, potentially, for the identification of “an unexpected serious risk” for certain patients. The agency is also requesting that sponsors discuss how to address any sex-based differences in their applications.

Sex-based difference guidance number two focuses on medical devices

• This guidance is about 14 pages longer than the all-products guidance and offers more granular recommendations for device developers on sex-based assessments. While the draft guidance discussed above would update guidance from 1993, this draft guidance for devices would, if/when finalized, update a 2014 device-specific guidance on the study of sex differences.
• How does this device draft guidance compare to the guidance discussed above? This device-specific guidance urges developers to think about gender in addition to sex. Further, as this document is an update of a more recent document, the guidance offers more nuanced interpretations of existing recommendations, rather than all-new policy. However, some of the additions to the new device guidance mirror those in the all-product draft guidance as the agency recommends collection of data “during the trial and also prior to dropout, if it occurs” for pregnant participants, noting that “Collecting this type of data improves the ability to inform the instructions for use.” Similarly, both guidance documents urge sponsors to account for the potential for Type I error rates.
• How is the new draft guidance different from the 2014 final version? There are a few differences, but the overall structure of the guidance and its contents have been maintained. One key difference is that this draft guidance focuses on both sex and gender, meaning both biological sex and gender identity, both of which are emphasized throughout the guidance document. The new draft guidance is about four pages longer than the 2014 final version, offering some expanded information in individual sections. However, besides the re-written up front sections (definitions, background have all been updated with newer information and examples from the past decade) there are no entirely new sections in the guidance.
• However, many of the sections have been enhanced, with the FDA adding more nuance throughout – although it largely added nuance to content that was already in the guidance (including changing references solely to “sex” throughout to “sex and/or gender”). For example, while the 2014 guidance states that sponsors should “consider alternative communication strategies… for study recruitment, informed consent documents, and patient materials,” the new draft from 2025 states “Consider expanded communication strategies, such as community presentations and alliance building with area women’s groups, for study recruitment.”
• Something that’s been taken out: The 2014 final guidance recommends that sponsors consider a “continued access study” to focus on female participants (referred to as “women” in that guidance, which uses “women” to refer to sex rather than gender; this is reversed in the new draft guidance). The new draft guidance has removed this recommendation from both places it appears in the 2014 guidance.
• Some sections have seen more updates than others, but the overall recommendations on what information the FDA would like to see, and what it recommends sponsors consider at different stages in development/applications/post-market, are generally all maintained – albeit with expansions on several concepts, as noted above. Again, these are largely updates that add granularity to existing concepts from the 2014 guidance. For example, for comparative studies, the FDA noted in 2014 that evidence of potential differences in effects between sexes should be discussed with the FDA, as “in some cases, the interaction effect could be statistically significant but not clinically meaningful, or clinically meaningful but not statistically significant.” The 2025 draft guidance instead states in this section that “Additional analyses may be requested by FDA to investigate possible explanations for these differences, including, but not limited to, adjusting variables and/or interactions between treatment and variables such as age, body mass index (BMI), bone density or concomitant illness (e.g., diabetes). Additional data may also be necessary to appropriately evaluate the effect of sex and/or gender on study endpoints. In these cases, discussions with FDA are advised.” While the section and this bullet generally offer the same information – that the FDA will want to see sponsors address, justify and potentially discuss differential treatment effects – the new 2025 draft guidance includes a note that reviewers may want to see more information.
• This theme is carried throughout. In the expanded section on “Interpretation of Sex Specific Data,” the agency again states that “FDA may request additional studies in one or both sexes, or one or multiple genders, to support a premarket submission, implement specific post-approval study conditions, and/or recommend modifications of the design of subsequent studies.”
• Does this mean that the FDA expects sponsors to be collecting data on both sex and gender identity? It doesn’t seem so. As the agency explains, “For premarket submissions, FDA recommends researchers analyze and report on data already generated, whether it is sex-based, gender-based, or both as appropriate for the scientific question being studied. However, FDA does not anticipate that researchers will collect both sex and gender data for each clinical study, unless indicated by the scientific question at hand.” (emphasis added)

Analysis and what’s next

• It’s a little unclear how the FDA expects the new documents to interact with the DAP guidance. The new draft guidance for drugs, biologics and devices cites the DAP authority, including a reference in the section on trial design that sponsors “should consider” the recommendations in the draft guidance on sex differences when designing their DAP. That said, industry has long questioned how, exactly, the DAP authority extends to non-race/ethnicity-related demographic factors.
• If finalized, the all-products draft guidance would update policy from 1993. This document largely reflects many of the existing recommendations that apply to device programs under the 2014 final guidance.
• For the device-specific guidance, the extension to gender identities is likely to raise operational questions for researchers, especially as there is not currently a standardized format for gender identity recognized by the federal government – in general, collection of these data reflect the experiences of participants, and may change over time. Best practices collection of these data are still evolving. It’s also worth noting that while the FDA doesn’t seem to indicate it will require gender-identity-related data collection, the guidance does state that the FDA might ask ßfor additional data collection on this subject.
• The publication of these documents suggests that the agency is looking to move ahead with its diversity-related initiatives, even as the incoming administration has indicated that it will seek to halt these policies at the federal level. Going forward, it’s not entirely clear what’s next for the DAP guidance, or these related policies.
• Comments on both documents are due April 7, 2025.

To contact the author of this item, please email Laura DiAngelo ( ldiangelo@agencyiq.com).
To contact the editor of this item, please email Karen Early ( kearly@agencyiq.com).

Key Documents and Dates

• Study of Sex Differences in the Clinical Evaluation of Medical Products
• Evaluation of Sex-Specific and Gender-Specific Data in Medical Device Clinical Studies