Accident statistics are meant to alarm us - that is what they are created for. So my heart always sinks when I read a newspaper headline about toddlers and water, children and stairs, strangers and danger, because the source of information is usually a number dressed up with a percentage sign. It is like a wagging finger reminding me that I am an even worse parent than I imagined for, say, letting my offspring bathe, climb stairs or venture out of the front door.
And on my favourite topic - the vexed question of where a baby sleeps - I know that parents are meant to feel scared, very scared. "Peril of Babies in Parent's Bed" and "Adult Bed No Place for Babies" growled this week's headlines. "Babies who sleep with their parents are 40 times more likely to suffocate than those sleeping in their own cots ..." In the past 15 years of research on babies and bedsharing, I have seen a handful of equally terrible claims. Every time the statisticians have to revise or eat their conclusions. Personally, I would like to put 10 statisticians in a bed and see them all roll over.
On the face of it, this week's scare story is as plausible as any other. Two officials from the United States Consumer Product Safety Commission and a paediatrician reviewed 1,400 infant deaths in the 80s and 90s. They found the risk of suffocation was greater for babies who fell asleep with their parents on beds, chairs or sofas.
But how meaningful is the new data? Retrospective statistical studies are highly unreliable, and the size of the sample must also be considered: 1,400 babies over six years, or 233 deaths a year. In 1997, more than 28,000 babies in the US died before the age of one. Mechanical suffocation can happen to babies wherever they sleep and there are highly effective ways of eliminating the risk. But we are not witnessing an epidemic and it hardly warrants the conclusion of Dr James S Kemp of the St Louis University School of Medicine, that "a sleeping baby belongs in a crib or other approved bed".
The idea that co-sleeping is inherently dangerous for babies is an urban myth. Generations of parents have been made to feel guilty if they cuddle up to their children at night, even though they may get more sleep, breastfeeding is easier, crying rates are drastically reduced, babies breathe more steadily, sleep hormones are stimulated, baby's core temperature is regulated by skin-to-skin contact, parental confidence is boosted and it becomes possible to react swiftly in a crisis. We feel guilty even though cot-death is unheard of in cotless cultures from Africa to Asia and South America, while thousands of American and some 200 British babies still die every year from Sudden Infant Death Syndrome (Sids).
The classic tale of Sids scaremongering came from New Zealand in 1991. Thanks to an emotional documentary featuring Anne Diamond, British parents first learned of the New Zealand Cot Death Prevention Programme and Back to Sleep campaign. Back to Sleep was positive and well-researched and has saved many thousands of infant lives. But back then, it was accompanied by an unexpected warning: do not sleep with your baby.
I attended a press conference in London where the New Zealand researchers handed out papers claiming that co-sleeping could lead to cot death. I was sitting with health researcher Dr Michel Odent, who whispered to me: "I am going to Japan tomorrow. There, they sleep with their babies and have never heard of cot death. How can I go to them and say, 'Do not sleep with your babies.' They will think I am crazy."
In a fit of statistical zeal, the New Zealand researchers had compared practice in Maori families with those in white culture. Yes, there was a high Sids rate among Maori families with their traditional culture of co-sleeping. But some not-so-traditional factors - maternal drinking and smoking and the poverty of the indigenous people - were the real culprits.
And an even greater irony had been sitting right under the researchers' noses. In the 90s, around 95,000 Pacific Islanders formed a substantial minority of the New Zealand population. The Pacific Islanders slept with their babies and had never heard of cot death until they arrived in New Zealand. It was only when they followed white doctors' advice and put their babies down to sleep alone that they experienced cot deaths.
The New Zealand statisticians were forced to retract their initial conclusions and all over the world a flurry of co-sleeping research began, the most influential and important being that of the British Confidential Enquiry into Stillbirths and Deaths in Infancy. It found that while some babies do die in their parents' beds, the risk factors are smoking, sofa-sleeping, drug use and social isolation. Bedsharing itself does not seem to pose extra risks and room-sharing, including bedsharing, is "strongly protective" against cot death.
Co-sleeping has been around for millennia. It is practised in all continents and is by far the world's preferred method of night nurture. Only in modern western homes has it become a sign of poor parenting or failure. No wonder the latest US study has problems making sense of its own figures. They have assumed that only 13%-14% of families co-sleep, because this is the number of American parents who will admit to it. The truth is that some 75% of western families take their babies and small children into bed at some point, a statistic which suddenly makes the US study appear even less significant than it did at the top of the page. Bring me no more statistics, I'm going back to bed.
· Deborah Jackson is the author of Three in a Bed: The Benefits of Sleeping with Your Baby (Bloomsbury, £7.99)
