Ridiculous? When I was asked to comment on the local television news about a study that linked pre-conception diet with the sex of the baby, I figured I would be able to poke holes in the data in no time. The article was called, "You are what your mother eats: evidence for maternal preconception diet influencing foetal sex in humans," published in the Proceedings of the Royal Society of Biological Sciences.

The science seemed good, although hard to believe. They studied 740 British women who didn't know if they were carrying girls or boys, and had them track their diets before and during pregnancy. The researchers looked at elements in the diet and calories consumed and found that calories seemed to play a role in whether the mothers gave birth to boys or girls.

When the researchers grouped the mothers by calorie intake, the third of moms in the highest group had a 56 percent male birth rate, and the third with the lowest calorie diet, 45 percent. Eating breakfast cereal was one item that went along with having a boy. Only the diet at the time of conception made a difference; as you'd expect, diet later in pregnancy had no effect.

Here are some principles to understand before the study will make any sense.

• In some animal species, more males are born in times of plenty, and more females in times of want.
• It takes more energy to raise a male than a female. (You may doubt this if you have a 12-year-old daughter, but they mean available calories, not maternal aggravation).
• Sex is determined as soon as the sperm meets the egg. If sex is going to be influenced by environment, it must be in how many of each sort of sperm (X- or Y-bearing) reach the egg, or which sperm gets in, or which fertilized egg survives. We know that in most non-pregnant cycles, fertilization occurs but the conceptus is lost before the woman even misses her period. Selective survival of the earliest embryos is a potential way to affect male to female ratio.

One possible explanation based on evolution and natural selection:

If girls were more likely than boys to survive when they were born into hard times, the mothers who gave birth to girls would be more likely to pass on their genes to the next reproducing generation. And in times of plenty, perhaps boys had a selective advantage, and their moms were more likely to pass on their genes to grandchildren.

If genes could affect whether you had a boy or a girl based on nutrient availability, the genes that favored boys in times of plenty and girls in times of want would be more likely to be passed on to surviving offspring, who would then give birth selectively to boys or girls depending on nutrient availability.

In today's developed world, there is probably no difference in survival of boys and girls, but humans lived for a long time before farms and trucks, where food availability was more variable, and might have selected for saving the boys for when they would be more likely to make it.

This doesn't mean that mothers-to-be who want boys should pig out, and those who want girls should starve themselves. For one, as you can tell, diet only shifted the male-female odds a bit; it was not close to 100 percent effective. For two, I would never recommend eating an unhealthy or restrictive diet in early pregnancy.

But these effects are interesting, and surprising. As one "woman on the street" said during the television segment, "That's ridiculous! Everyone knows sex is determined by chromosomes!" And so it is. But the process may be more complex than we thought.