CHICAGO — Snips and snails and puppydog tails ... and cereal and bananas?

That could be what little boys are made of, according to surprising new research suggesting that what a woman eats before pregnancy influences the gender of her baby.

Having a hearty appetite, eating potassium-rich foods including bananas, and not skipping breakfast all seemed to raise the odds of having a boy.

The British research is billed as the first in humans to show a link between a woman's diet and whether she has a boy or girl.

It is not proof, but it fits with evidence from test tube fertilization that male embryos thrive best with longer exposure to nutrient-rich lab cultures, said Dr. Tarun Jain. He is a fertility specialist at University of Illinois at Chicago who wasn't involved in the study.

It just might be that it takes more nutrients to build boys than girls, he said.

University of Exeter researcher Fiona Mathews, the study's lead author, said the findings also fit with fertility research showing that male embryos aren't likely to survive in lab cultures with low sugar levels. Skipping meals can result in low blood sugar levels.

Jain said he was skeptical when he first heard about the research. But he said the study was well-done and merits follow-up research to see if the theory proves true.

It's not necessarily as far-fetched as it sounds. While men's sperm determine a baby's gender, it could be that certain nutrients or eating patterns make women's bodies more hospitable to sperm carrying the male chromosome, Jain said. "It's an interesting question. I'm not aware of anyone else looking at it in this manner," he said.

The study was published Wednesday in the Proceedings of the Royal Society B, a British medical journal.

The research involved about 700 first-time pregnant women in the United Kingdom who didn't know the sex of their fetuses. They were asked about their eating habits in the year before getting pregnant.

Among women with the highest calorie intake before pregnancy (but still within a normal, healthy range), 56 percent had boys, vs. 45 percent of the women with the lowest calorie intake.

Women who ate at least one bowl of breakfast cereal daily were 87 percent more likely to have boys than those who ate no more than one bowlful per week. Cereal is a typical breakfast in Britain, and in the study, eating very little cereal was considered a possible sign of skipping breakfast, Mathews said.

Compared with the women who had girls, those who had boys ate an additional 300 milligrams of potassium daily on average, "which links quite nicely with the old wives' tale that if you eat bananas, you'll have a boy," Mathews said.

Women who had boys also ate about 400 calories more daily than those who had girls, on average, she said.