Making scientific research work for women

Scientific research has traditionally treated the male body as the default, resulting in health inequity and poor outcomes for women in a world not designed for them. The Sex and Gender Equity in Research (SAGER) guidelines are an attempt to make sure research is designed for everyone, and that the outcomes better serve everyone.

Editor-in-chief of The Lancet Haematology Lan-Lan Smith and editor-in-chief of The Lancet Regional Health Americas Taissa Vila join Gavin to discuss how the history of scientific research led us to this point, what the SAGER guidelines are, and how science is addressing the research gap.

Read all of our content at https://www.thelancet.com/?dgcid=buzzsprout_tlv_podcast_generic_lancet

Check out all the podcasts from The Lancet Group:
https://www.thelancet.com/multimedia/podcasts?dgcid=buzzsprout_tlv_podcast_generic_lancet

Continue this conversation on social!
Follow us today at...
https://thelancet.bsky.social/
https://instagram.com/thelancetgroup
https://facebook.com/thelancetmedicaljournal
https://linkedIn.com/company/the-lancet
https://youtube.com/thelancettv

Transcript

This transcript was automatically generated using speech recognition technology and may differ from the original audio. In citing or otherwise referring to the contents of this podcast, please ensure that you are quoting the recorded audio rather than this transcript.

Gavin: Hello and welcome to the Lancet Voice. It's September, 2025, and I'm your host Gavin Cleaver. The history of scientific research has failed women. It feels like a bold statement, but the data, or rather the lack of data backs it. The history of scientific research is one of treating men as the default body, meaning research has been used to create systems and treatments which aren't suited for women.

This leads to poorer health outcomes and means that a lot of research just isn't covering everyone. The sex and gender equity and research guidelines, which we'll refer to as the SAGE guidelines, are an attempt to make sure every relevant scientific study, better accounts for sex and gender. As well as being your producer and host here at The Lancet Voice.

I'm also the co-chair of the Lancets Gender and Diversity Task Force. I'm joined to discuss sex and gender in research and data, and the SAGE guidelines on the line from Rio de Janeiro by my fellow co-chair and editor in chief of the Lancet Regional Health, America's Taste of Eila, and by the supervisor of the Gender and Diversity task force advocate of the SAGE guidelines across Elsevier and the editor-in-chief of the Lancet Hematology Lan Smith.

We hope you enjoy the conversation

Tesa Land. Thank you both so much for joining me on the Lance of Voice. Probably it's best to kick off at the beginning. Why do these guidelines exist by how has scientific study design failed women in the past? What's the importance of having properly disaggregated data in scientific research? 

Lan-Lan: Yeah, I'm happy to kick that off if you want.

So this is Lon and I'm also involved in the Elsevier sex and gender based analysis work stream. This has been something I've been working on for quite a while, and it, I'm working on it because it's, I find it so important. It's really a topic that I'm passionate about, and. You say, why is it important?

Traditionally, women have been excluded from scientific research. We've all know the classic crash test, dummy example, right? Crash test dummies. When they started manufacturing cars with more safety features, the default design was the male body and therefore. Women who get in car accidents or other people who do not conform to the average male are at more risk from greater injuries.

And the whole cardiology women present in different ways when having a heart attack from men and therefore are underserved. And these are just two really common examples of. Areas in research and development and design where the male is considered the default and women are tacked on as a, as an afterthought.

And I think we're understanding more now how. Different biological and societal impacts have on health outcomes, and this is why it's really important to be able to take into account sex and gender reporting. And this is not a new concept. The S seizure guidelines that you're talking about at Gap, and these were initially written up in 2016.

Obviously there's a lot more discussion before that about this, and I'm sure we'll come across it. We'll talk about a few more interesting examples as we get on. I'm sure 

Gavin: I was gonna say that still feels relatively recent, right? The 2016, given the kind of, yeah, the span of scientific research is what we're talking about.

Lan-Lan: That is true. You can say if you. Back in, if you wanna go back a little bit further, in 1993, which is not that long ago for me, but I still think Jurassic Park just came out yesterday, but the US National Institutes of Health had a revitalization act where it said that women have to be included in any NIH funded clinical research trials and designed to be analyzed if the variables affect women differently.

We do have a bit more going back, but yeah, I agree. If you're going back to the very earliest days of scientific research, then this is a relatively new area. 

Taissa: Yeah, paa, you so Bri, to your point about how science has failed women, and another example that to add to landline's example is exclusion of women from clinical trials because of hormone levels or childbearing age, women excluding from trials as well.

So it's only in the nineties that women from tire bearing potential. Could be included in drug trials at the FDA and it. You can imagine how much this exclusions had led to us defining doses, outcomes or symptoms, risks, everything, excluding people that could be presenting completely different from the, what we call a defo.

And it's totally define it. By a societal perspective of what the Defo is, which is male. 

Gavin: I think you made a really interesting point there, Tessa, which is well worth highlighting, which is the drugs, pharmaceuticals interact with male or female bodies in quite different ways, but the research hadn't really covered this, which is.

Huge loss in kind of healthcare outcomes for women having to, as you mentioned, land. Land, the kind of default model of a male in these trials. If every single piece of pharmaceutical dosing is set up to treat men, the women rolling the dice every single time. 

Lan-Lan: Yeah, and not just the primary outcomes for drug studies, either the adverse effects profile, so the side effects that people can have from different drugs can be very different as well.

And that kind of brings us round to the SAGE reporting guidelines, which is saying you've got to disaggregate your data by sex and gender, because otherwise you're not looking at the whole picture. 

Gavin: How often do you get a decent gender split in trials that are submitted to your journals, like a 50 50 gender split, for example, in the data?

Lan-Lan: For me, it really, it depends on the disease. Obviously for a disease like hemophilia, it's gonna be massively skewed towards males, but women and girls bleed too. That's another area where they can be a bit understood, served. In the hemophilia world or the more hematology oncology ones, there are different diseases that have slightly different sex skews.

So I think the important thing for us is that we. Get all that data reported and for trials, which where you have something manufactured like a clinical, like for CAR T-cell therapy, I see a lot of CAR T-cell papers. We actually see who were the people who weren't. It was aware it wasn't possible to manufacture the product.

Is there a sex bias or something like that? So when you make sure you report on the data, then you can start to tease out some interesting facets. 

Taissa: Yeah, I guess it's very disease and field oriented as well, because the guidelines, not only SE but specific founders and institutions have applied different scrutinizing or requests.

In some areas, the aggregation of data is way more advanced than in others. So for us as a general medicine journal and going across. Multiple areas. I can see the difference that, for example, it's much more common to see disaggregated data for specific fields compared to others where much more needs to be done on the importance of collecting and reporting those data Still.

Gavin: So it's more about having a kind of good rationale for it as you talk about with different subject areas. And that's partially what the SAGE guidelines are about, right? It's just at least flagging that these disaggregated data might be imbalanced in one way or another for this particular reason related to this particular study.

Lan-Lan: Yeah, I guess it's interesting to think about it that sage's kind of at the end of the line, it's the reporting guidelines. Whereas I think what's really important these days is that everyone actually bakes this stuff right into their study design. And I know a lot of funders are already asking for it, so it should be, so by the time it comes to us as publishers and editor, we should be seeing.

That was already supposed to be happening in the study is already, but unfortunately it's not always the case. 

Gavin: So I guess before we, we get into talking about that, it's probably best for us to outline what the Satre guidelines actually say. 

Lan-Lan: I think before we get into that, what might be even better is to actually just have a little chat about sex versus gender, because I think that's where people get confused a little bit and they think, what is relevant and what should we talk about?

Because for different studies we can make, definitely make an argument that sex and gender are relevant for everything. But when. From my perspective, working on a clinical journal where I see a lot of clinical trial data, then probably sex is going to be more relevant than gender for my journal. But people do use the terms interchangeably and they're not quite the same.

So sex generally refers to a set of biological attributes that are associated with. Physical and physiological features such as chromosome, genotype, hormonal levels, internal and external anatomy. So it's not even just x and y chromosomes. There's a lot more. It's all the. Biological processes involved with that, whereas gender refers to a social construct, gender's a social construct.

It's has to do with roles, behaviors, identities, which can definitely have an impact on health as well, for example. Women are more associated with caring responsibilities and therefore may have less time for them to care for themselves, and that definitely has a health impact. And that's one kind of example for how gender can influence outcomes as well.

Taissa: From my perspective, I do see a lot of papers with the intersection of medicine and social science in our journal. And for us, the desegregation of gender specifically is very important in many levels because it also helps us identify disparities in access to care and. Other levels of bias and stigma that define access to health and health outcomes on a conception, like in.

In a consequence of that so many countries, for example, men are less likely to seek care or a adhere to treatment for chronic diseases. This is very relevant. And on the other phase, as you mentioned, landline women have stigmas and caregiving burdens associated with identifying as a women and the gender level as well.

So there is the intersections of both of them. And with the wider society roles that they play are very important. And I think one of the things at the guideline level that we can is to. Try to change how we look at the design and what is important or not important to report and collect as data as well as we advance more and more to understand what a health outcome means.

The broader picture of it will identify that a lot of the data that we're not collecting actually very important to be segregated in the end. 

Gavin: You mentioned it there, taa, but it speaks to the intersectional nature of health outcomes. Doesn't that, I was reading up before this, and there's that classic study on the facial recognition.

That was a really good example of how intersectionality works across data and health and there's a facial recognition system and a famous study. That found that the error rates on recognizing people's faces were 35% for darker skinned women, 12% for darker skinned men, 7% for lighter skinned women, and less than 1% for lighter skinned men.

So it shows how those different aspects of kind of xis of discrimination, as people refer to them, can interact on people's outcomes and how systems even interact with them as well. Yeah, it's really important for people to bear in mind that sex and gender are two. Scientifically two distinct aspects of a person, but they can both interact in different ways on a person's health outcomes, right?

Taissa: Yes, perfectly. Exactly. Where I think if you ask for what's the aim of the S guideline, I dunno if Len's gonna agree with me, but it's to change. Societal perception of what needs to be reported and why. So while we are acting at the end, the whole idea is that at some point we don't need to actually ask for this because this is the new default of how we do science and report it.

Lan-Lan: It's a great groundwork for this international set of guidelines that encourage really a systematic approach to reporting sex and gender across. A lot of disciplines, and then you can apply it as a comprehensive tool for researchers, authors, editors to make sure that we capture this really important metric.

Gavin: It just results in us publishing better research, right? The research is a more accurate knowledge base. The outcomes are better health for women better, better equity. Even the things we just talked about before, the usefulness of pharmaceuticals. All of this stuff depends on kind of this accurate reporting that we're talking about.

Lan-Lan: Yeah. And even if people's designs haven't been designed to disaggregate. Sex or gender. I often get a lot of pushback. Like we didn't design the trial, we would have to recruit twice as many people to get the power calculations, but you can still do that analysis, post-talk, publish it in the appendix, and then that data will be available for people to do a larger meta-analysis in the future, which can be really beneficial as well.

It doesn't need to be perfect. No 

Taissa: I would say that at this stage that we are now of knowledge about the aggregation. I think this is the most important message when I talk to researchers as well and that something clicked for them. When you. It's okay if when you desegregate, you'll have to make sure that you don't have the power for any conclusions, because that puts it in the pool of possibilities for future meta-analysis.

So it's still very important if you can please do it. And I think this is at this stage that we are, this is a really important message to send out. As for the researchers that are already recruited and done, that's still something that you can do. 

Gavin: And Lan you mentioned there about trial design. For our listeners who might not be all over the concept of trial design, what's the difference between designing your trial around capturing this sort of data and then versus running a post hoc analysis?

Lan-Lan: Well, a lot of it will probably come down to power calculations and having a little bit of thought. At the beginning about how to apply sex and gender, for example. Psychiatry study. Maybe they need to have a person's self-identified gender to be captured that information along with different biological variables, such as maybe they're measuring a certain hormone level and really thinking about how that needs to be incorporated.

How are we gonna collect that data? What are, how is it gonna be an individual will self identify as a certain gender or will we. Be looking at specific biological variables to define the sex and really considering at the beginning how these will be incorporated. And then saying if we want to do a proper disaggregated data, getting the biostatisticians to make the proper power calculations so that they can draw conclusions based on these areas.

Gavin: So this feels like a good time. Then given all of that to get to what the SAGE guidelines are. 

Taissa: Sure, I can give a brief outline. So the guideline is a set of recommendations of how to report sex and gender information in the study design, in the data analysis, in the results in the interpretation. And in your findings.

So it guides the alter through how to do a better reporting of the data that they already collected. So as Lenon was saying, at this stage you already have that. You can go back and be analyzed post hoc if you have enough data collected to do it. And what we're trying to achieve with that is change the overall mentality.

Of how we report it. And there the guideline specifically encourages the correct to use of the term sex and gender as we just referred to, all the way throughout study participants and the methods and requests that you think about how you're gonna collect the data, what methods are you gonna use to determine each variable that's there and report how you did that.

It advocates for separate reporting, as we mentioned, and also for discussion of the influence on the association of those variables with your findings. So it's not only about collecting, but if you can critically think about how they impact your data and your findings, and if you did not have them, then discuss.

How this limits your capacity to interpret your data in the end. So it also calls out for a full reflection on your findings using sex and gender interfaces and approaches as well. It provides you with a checklist, so this is really important. You can use the checklist to make sure that you are following everything, and importantly, I think the biggest aim is to improve transparency, reproducibility, and equity in publishing.

I guess this is the walkthrough. 

Gavin: I think it's all that. It's a really great overview of it. I think what really sticks with me is how. Not doing this by this point, is seen as a limitation on how it's important to, to discuss the lack of this sort of, not even the data reporting we've been talking about the lack of baking, the SAGE guidelines in from the start of a study design is.

Very much now recognized as a limitation in the scientific community. 

Lan-Lan: Yeah, that's, I think that's really important and hopefully journal groups like The Lancet Cell Press, other big journals asking for this information, we're flagging up that this is something that is of value and important and we'll help educate people for trial design and study design going forward.

I know a lot of funding bodies are already asking for it, so hopefully the message is getting out there. 

Gavin: As journal editors, how have you found this progressing over time? Do you find that most of your studies that are submitted to you have this contained? Do you find a higher percentage over the last few years?

Is it moving in the right direction? 

Lan-Lan: We implemented the SAGE guidelines into our information for authors in 2023, so we only ourselves as a Lancet group adopted them. Relatively recently, although when they were first published in 2016, we obviously said, ah, this is a great idea, but it's only been incorporated at in 2023.

And I have to say it's not just the Lancet Group. Elsevier itself has done an amazing job across this. In 2023, we got it added to All Lancet Group. Titles, all of cell press because we haven't even touched on periodic cells and animal models yet. But yes, Sage can apply to those as well. And over 2,300 Elsevier titles.

So that's, I have to say, that was a great accomplishment. I think at this time it feels very much like we're really. Doing more prompting. I've not had a huge amount of pushback from authors. There's obviously, you get ones which say, oh, we can't because wouldn't really be meaningful because the study's not powered.

But in general, we do get people who agree with our prompting to add disaggregated data, and hopefully in the future they'll just do it automatically. 

Gavin: Because I think the argument is that researchers already have a lot to do. They're stressed, they've gotta get all these research studies out. But in what we are saying we don't think is absolutely right is that this is a central part of putting out a good study.

Lan-Lan: If your funding body asked you to do it in the first place, then you should be reporting on it as well. Yeah. 

Taissa: To 

Lan-Lan: your 

Taissa: point about how we are receiving it, as LAN said, this is really recent, but earlier in the year, we conducted a internal alter survey. On the topic at The Lancet. So we did a pilot study to understand how our elders were receiving the implementation of the guideline.

It was a small survey based study that we did, but the overall acceptance of the alters of the guidelines is quite high. The pushbacks are similar to what Leland reported, but we have like over 50% of people responding that they encourage the initiative that a high percentage of people saying that after being prompted by this first submission, they will change their approach in their next submission.

So that is very good feedback that we received. We also collected feedback on how we as editors could help. Make the SAGE R guidelines more accessible and visible. So how can we serve outs from the editorial perspective and acting on that? That's one of the reasons why we are doing the webinar on SAGE R in the near future.

This is one of our actions prompted by that survey on things that we can do to encourage and promote this guidelines adherence more and more. 

Gavin: Yes, our Marcos colleagues would be quite annoyed if we got through this podcast without plugging the webinar that the three of us are doing October the 21st.

If you would like to join us for a webinar to discuss the sage o guidelines, you can find the link to sign up in the show note. I just wanted to finish up by asking you both what you think the next steps are. After the SAGE guidelines. Now, obviously there's no such thing as after the SAGE guidelines.

It's always gonna be something that is a constant back and forth between editors, authors, publishers, research funders, all that sort of thing. But in terms of intersectionality in kind of data and reporting, what would you like to see next? 

Lan-Lan: I think you hit the nail on the head when you said intersectionality and TASA touched on this earlier.

That is really. Important. There's so many things that affect health outcomes and just looking, I, we've got amazing group within the lens of the group for racial equity race, and we also have a race and ethnicity guideline within our information for authors as well. Being able to consider all of these factors together and pulling in socioeconomic issues.

These all have determinants. It gets more complicated the more factors you try to incorporate, which is why I think we're starting out with saying. Sex and gender. These are more and more basic things, but as once everyone's on board with that, then expanding outwards and really considering other important factors and bringing that all together really.

I guess it's personalized medicine again, isn't it? Where you need to look at trying to get things that are relevant towards individuals. 

Taissa: Yeah, I guess that's where we were going. And I think that if we wanna stick with the population level, the future would be not to look at separate well desegregated data.

Yes. But once you have the segregated data, then you can look at how they intersect to influence the outcome that you're looking at. So I think if I wanna see something in the future is that every single research question accounts for context. Of the individual, even if it's a population level research, the question needs to account for context.

And context involves everything from sex to gender, to socioeconomic aspects to political influence in health, and so on. I guess if we get there 

Gavin: onwards to fully personalized medicine. Alright. Thank you both so much for joining me on the podcast today, and I'll see you on the webinar in a month's time.

Maybe some of our listeners will join us there, but for now, thank you both for your time. 

Lan-Lan: Thank you. Thank you so much.

Gavin: Thanks so much for listening to this episode of The Lancet Voice. If you'd like to hear more from all of our journals, you can go to the lancet.com/podcasts to see offerings from across every journal. See you again soon and take care.