The U.S. government’s medical research agency is taking steps to erase sex bias in pivotal biomedical studies that pave the way for human clinical trials, saying scientists too often favor male over female laboratory animals and cells.

A new requirement announced Wednesday by the National Institutes of Health for researchers applying for NIH funding is likely to have a big influence because the agency is one of the world’s top financial backers of biomedical studies, spending about $30 billion annually.

Beginning Oct. 1, researchers seeking NIH grants must report their plans for balancing male and female cells and animals in preclinical studies, with only “rigorously defined exceptions.” The NIH also plans to train grant recipients and its own staff members on designing studies without sex bias.

“Our goal is to transform how science is done,” wrote NIH Director Francis Collins and Janine Clayton, director of the NIH Office of Research on Women’s Health, in the scientific journal Nature.

The current over-reliance on male lab animals such as mice and rats and male cells in lab dishes in such research obscures important sex differences that could guide future studies involving human subjects and can lead to flawed findings, the NIH said.

The agency said inadequate inclusion of female cells and animals and inadequate analysis of data by sex may contribute to a “troubling rise” in findings by researchers that other researchers are then unable to verify in separate studies.

Twenty-one years ago, the NIH began requiring inclusion of women in NIH-funded research using human subjects. Previously, some studies being used to determine whether a new medicine or treatment worked simply excluded women.

Before new medicines or treatments can be tried on human subjects, they are tested on animals or cells in a lab. These preclinical studies can lead to human clinical trials or can doom approaches that do not seem to work.

“I really think this is a blind spot. I don’t know that there’s intentional bias here. But there’s certainly a reliance on male-only animal models, which have become the convention in some fields,” Clayton said in a telephone interview.

Clayton said the sex imbalance stems in part from an obsolete notion that the female hormonal cycle would cause too much variability in lab animals and disrupt a study’s results.

She pointed to examples of treatments working differently in men and women, including how well aspirin therapy protects against heart attack and how well nicotine patches and gum help smokers quit.