Statisticians rubbish cereals link to baby boys

By staff reporter

- Last updated on GMT

Related tags Energy intake

A new paper has challenged research that concluded a child’s gender could be determined by its mother’s diet at conception.

Research conducted at the Universities of Oxford and Exeter and published in the Proceedings of the Royal Society B last year​studied dietary data from 740 British women who were unaware of their foetus’ gender, and drew conclusions about the kinds of food they ate whether they had a boy or a girl baby.

Breakfast cereals, for instance, were implied to be more conducive to having a baby boy – as was eating ​a higher quantity and wider range of nutrients, including potassium, calcium and vitamins C, E and B12.

The original research team saw that 56 per cent of women with the highest energy intake bore a boy, compared to 45 per cent with the lowest intake. No correlation was seen between gender and energy intake during pregnancy.

Media interest in the findings generated some 50,000 Google hits; however the validity of the results are now being questioned by a new paper published in the publication. US researchers Stanley Young, Heejung Bang and Kutluk Oktay say that the original results can easily be explained by chance.

They reach this conclusion after looking at two data sets from the original study, and noted that 132 food items were tested for two time periods. This makes for 162 in total – and with such a high number, they believe, it is highly likely that statistical significance will come about by chance.

Given that there is a 5 per cent chance data will show an effect when there is none, unless adjustments are made to account for double-testing around 13 false positives will occur.

Young, Bang and Oktay call the original paper “well-intentioned”,​ but say it is hard to believe that a woman’s diet can really affect whether she has a boy or a girl baby.

With such an arguable implausible result, they say it is “essential” ​that multiple testing be taken into account with transparent methods”. Without this, claims are lacking in credibility.

Sources: Proceedings of the Royal Society B​. January 14, 2009 online issue"Cereal-Induced Gender Selection? Most Likely a Multiple Testing False Positive"​Authors: Stanley Young, Heejung Bang, Kutluk Oktay.

Proceedings of the Royal Society BVolume 275, Number 1643/Jully 22, 2008"You are What your Mother Eats"​Fiona Mathews, Paul Johnson, Andrew Neil

Related topics Markets