Boys were more likely to be born with genital birth defects if their fathers took the commonly prescribed diabetes drug metformin in the three months before conception, according to a new study by Stanford Medicine investigators and their collaborators in Denmark.
The finding adds to a limited but growing understanding of how paternal factors can affect the health of children.
For women, we know a lot about risk factors. There's a list of medications that women are not supposed to take when they're pregnant, for example. Given that men contribute half the DNA to a child, it makes sense that there could be some effects that travel through paternal pathways as well."
Michael Eisenberg, MD, professor of urology at Stanford and senior author of the study
In fact, the magnitude of the risk associated with metformin in the new study is on par with more well-known maternal contributors to birth defects, such as pregnancy after age 45 — a comparison that Eisenberg calls "sobering."
Using health-registry data from Denmark that tracked more than 1 million births, the researchers linked men's use of metformin during sperm development to higher rates of genital birth defects in their sons. It's notable that the genital birth defects occurred only in boys, and these fathers were less likely to have boys than girls compared with the general population.
These results suggest that metformin affects men's reproductive health in a way that can harm their sons, though the mechanisms are still unclear.
A paper describing the study was published March 28 in the Annals of Internal Medicine. The lead author is Maarten Wensink, MD, PhD, associate professor of public health at the University of Southern Denmark.
Disease or treatment?
While poorly managed diabetes in pregnant women has been associated with birth defects, little is known about how diabetes in men impacts their babies.
"Inherent in health is the question of whether it's the disease or the treatment that could cause risks," Eisenberg said.
Metformin, an oral medication, is an increasingly popular treatment option for Type 2 diabetes, but it is known to have effects on the reproductive system.
Eisenberg and colleagues set out to evaluate whether metformin use in men affects the rate of birth defects in their children.
Data from birth and patient registries in Denmark allowed the team to compare all live births from 1997 to 2016. They excluded babies born to mothers who had been diagnosed with diabetes or hypertension.
Because sperm takes about three months to fully mature, fathers who filled a prescription for a diabetes drug during the three months before conception were considered to have exposed their offspring to the drug.
Of the 1,116,779 babies included in the study, 3.3% had at least one major birth defect and 51.4% were boys. Among the 1,451 babies exposed to metformin, 49.4% were male and the rate of birth defects was 5.2%. In particular, genital birth defects were more common and occurred only in boys.
These rates mean that, in Denmark, where today approximately 120 babies per year are exposed through their fathers, metformin use may account for two additional babies born with birth defects every year.
Homing in on metformin
“One of the things that scientists like to do is try to disprove their own finding,” Eisenberg said. “We try to determine that what we found didn’t just occur by chance or from other factors.”
In their statistical analysis, the researchers controlled for variables that may have skewed the outcome, including birth year, paternal age, income, education, maternal age, tobacco use and education.
To further shore up their findings, the researchers compared exposure to metformin with exposure to two other common diabetes drugs, insulin and sulfonylureas. Exposure to insulin affected neither the rate of birth defects nor sex ratio. Among the 647 babies exposed to sulfonylureas, 5.1% had birth defects and 49.3% were boys — numbers that trend in the same direction as metformin exposure. But these numbers did not reach statistical significance, Eisenberg explained, and lacked a consistent pattern. The sulfonylureas trends were neither specific to the three-month sperm development window nor to genital defects, for example.
Men who filled metformin prescriptions before or after the three-month period of sperm development did not have offspring with a higher incidence of birth defects, the study found.
“It wasn’t just taking metformin at any time in a man’s life; it really has to do with taking it in that window when the sperm that is going to become the child is being developed,” Eisenberg said. “And it’s not just diabetes because the man likely had diabetes before and after that point.”
Finally, the researchers compared exposed babies to their unexposed siblings and found the siblings did not have increased rates of birth defects, showing that the effect didn’t track with individual fathers.
Although it’s not clear how metformin exposure leads to birth defects, Eisenberg noted that studies in mice have shown that the drug can cause reproductive harm. Pregnant mice given metformin, for example, have male offspring with smaller testicles. “We tried to tackle it in as many ways as possible to home in on the particular medication and then the particular time point when that exposure is critical,” Eisenberg said.
The decreased ratio of boys to girls could be due to severe cases of harm. “In some cases, the effect of exposure may be so potent that it even leads to fetal demise, specifically in male and not female offspring, and that could be why you see fewer male offspring being born,” Eisenberg said.
‘Don’t forget the father’
Should men aspiring to fatherhood avoid this particular drug? “I think that it’s a single study, so it’s hard to change clinical practice based on that,” Eisenberg said. “But for somebody considering fatherhood, this study emphasizes the importance of a father’s health on the health of a child.”
Eisenberg said it will be important to uncover the biological mechanisms at work and to expand the study beyond the relatively homogenous population of Denmark. “Ultimately, before you change guidelines, this should be studied and validated in other populations as well,” he said.
Eisenberg’s previous studies have shown that American fathers, like mothers, keep getting older. That likely means parents are also less healthy, and diseases like diabetes increasingly need to be considered in reproductive health.
“When we think about reproduction, we still mostly think about maternal factors, so this study brings further awareness that paternal factors can make a difference,” Eisenberg said. “Don’t forget the father.”
Other Stanford co-authors of the study are professor of biomedical data science Ying Lu, PhD; professor of biomedical data science Lu Tian, PhD; and professor of pediatrics Gary Shaw, Dr.PH.
Researchers from the University of Southern Denmark and Copenhagen University also contributed to the research.
The research was supported by the National Institutes of Health (grants HD096468 and 1UL1TR003142) and the Centers for Disease Control and Prevention.
Rare variant analysis reveals genetic spectrum of monogenic diabetes genes