(CNN)The data didn't make sense.
Five years ago, University of Maryland researcher Alisa Morss Clyne was studying pulmonary hypertension -- a type of high blood pressure that affects the arteries in the lungs -- in human cells she had cultured in her lab. But the results she was seeing just didn't stack up.
"We had these huge error bars. It didn't make any sense," she said. "And we said, OK, let's just graph it by male versus female, and what we found was really interesting."
The blood vessels in the lungs of people with pulmonary hypertension take up more glucose, and she found the female cells metabolized the glucose in way that changed a protein that was critical to blood vessel function.
In other words, the sex of the cells became an important variable that affected the outcome of the research.
"Because we were grouping our sexes together, we were missing the difference. We were getting the average with a big deviation," said Clyne, an associate professor at the University of Maryland's Fischell Department of Bioengineering and the director of the Vascular Kinetics Laboratory.
"I was shocked when we saw the cells themselves were different based on sex."
Concern has been growing in recent years that ignoring or downplaying differences in sex as a biological variable -- whether in cells under a microscope or in lab animals -- is undermining biomedical research at the earliest stages.
This matters because many diseases -- including Covid-19 -- affect men and women differently, and missing sex-based differences can make misdiagnosis and mistreatment more likely.
"When researchers don't consider sex as a biological variable, that means we have incomplete data. And if we have incomplete data, we run the risk of making judgments that are incorrect. And incomplete data and erroneous conclusions in preclinical research, ultimately, can have an impact on the health of all of us, the health of women and men," said Chyren Hunter, associate director for basic and translational research at the National Institutes of Health Office of Research on Women's Health, which is holding a conference on sex as a biological variable this week.
The default lab rat is male
While it's often unclear what sex the cells used in lab research are, and to what degree female cells are underrepresented, the default lab rat has long been male.
One study from 2011 found that in neuroscience research, male animals were used six times more often than females, and figures from a more recent analysis in 2017 suggest they have only improved marginally.
Women have been routinely included in clinical trials since the 1990s, but "integrating female animals into basic and preclinical research -- which is the key building block for studies in humans -- has been much slower," said Hunter.
Women are more likely to be misdiagnosed than men when it comes to a wide range of conditions, including heart attacks and ADHD. And women typically experience more -- and more intense -- side effects from pharmacological drugs.
Some drugs are more effective in men than women, including common over the counter ones such as ibuprofen and naproxen -- both forms of nonsteroidal anti-inflammatory drugs (NSAID).
And it goes both ways: Alosetron, a drug approved to treat severe irritable bowel syndrome, is only approved for treatment in women as it is largely ineffective in men.
"We're saying here that not all researchers need to study sex differences. But all researchers really should consider how biological sex can impact the questions that they're studying," Hunter said.
In the case of research into Covid-19, it's not clear that sex as a biological variable is consistently being taken into account even though men are three times more likely than women to be admitted to the ICU.
What's being done
Major funders of scientific research -- including the US National Institutes of Health, which handles 80,000 grants a year -- have required the research they fund to take account of sex as a biological variable.
Since 2016, NIH grant applications are required to include male and female animals or cells in both designing the research and in analyzing the results or provide strong justification for studying only one sex.
Hunter said that there have been signs of progress, but a five-year report card called for greater action.
The NIH doesn't have any mechanism to ensure the funded research adheres to what was set out in the grant application, and some feel the efforts made in grant applications are only lip service.
"It's been nearly six years, and we haven't seen the needle move in any considerable direction in terms of even understanding what the sex differences are at baseline level in between males and females," said Aditi Bhargava, a professor in the Department of Ob/Gyn and the Center for Reproductive Sciences at the University of California San Francisco.
Bhargava, who is the author of a new scientific statement on sex as a biological variable for the Endocrine Society, and others said that high-profile scientific journals -- like those published by Nature and Science -- need to do more to ensure the papers they publish take sex as a biological variable into account.
"It's up to the journal and the scientists who (peer) review for the journals to start holding people accountable and say this doesn't deserve to be a high profile paper because it only studied males. It can't be that impactful if we don't know the answer to the science question in females," said Rebecca Shansky, an associate professor in the Department of Psychology at Northeastern University in Boston.
Shansky wrote in Nature Neuroscience in March that considering sex as a biological variable "will require a global shift in science culture."
Many journals ask authors to adhere to SAGER guidelines on sex and gender equity in research, but that doesn't mean that researchers have to comply with them.
"We are not currently monitoring compliance with SAGER guidance, but we may in the future ask authors to provide a clear disclosure as to whether the study adhered to SAGER guidance," said Sowmya Swaminathan, head of Editorial Policy and Research Integrity, Nature Portfolio.
She said Nature does require researchers to fill in a detailed reporting summary that includes the sex of lab animals. An independent study published in 2019 in the journal BMJ Open Science, which evaluated the effectiveness of these reporting summaries, found 52% of a sample of Nature journal papers published after the reporting summary was implemented reported the sex of experimental animals, compared to 36% in a sample of articles on similar topics published by journals from other publishers.
Not just hormones
For a long time, female lab animals were thought too hormonal and messy for science, said Shansky, and one of the main reasons given for not using female mice was that they made data more "variable" as a result of the female reproductive cycle.
But this is a myth, she said. Hormones aren't a uniquely female problem. The testosterone levels of male mice can also vary widely.
The push for greater awareness of sex as a research variable is not just about using more female lab rats. It's also important not to take the behavior of male mice as the baseline for research, Shansky said. In her lab, she has been studying fear behavior in mice, and the typical fear response in mice has long been thought to be freezing. However, she found that female mice will often exhibit an escape response.
"They try to escape the boxes. They are afraid. They are just showing it differently. So say if we were looking for drugs that reduce fear, say to treat PTSD, and you were looking for a freezing response when the breathing goes down, you wouldn't be looking for the right response," she explained.
"If you based your interpretations