Mark Vonnegut on Bad Science in Anti-Cosleeping Recommendation

Bad science sets out to make a point, looks neither to the left nor to the right but only straight ahead for evidence that supports the point it sets out to make. When it finds evidence it likes, it gathers it tenderly and subjects it to little or no testing.

And that's exactly what the commission did in this, its first attempt to quantify the number of fatalities resulting from the practice of parents "co-sleeping" with their babies and toddlers. The study, published in the October issue of the Archives of Pediatrics and Adolescent Medicine, was based on the deaths of 515 children under age 2 from 1990-1997. All 515 died in adult beds. The commission concluded that all the deaths were caused by that fact.

Such a conclusion amounts to connecting dots that shouldn't necessarily be connected. If a young child dies in bed with his or her parents, there can be many reasons. Completely normal babies can go home after a "well baby" checkup (with or without immunizations) and die of Sudden Infant Death Syndrome, or SIDS. Those babies could be in a safety-certified crib, or in an adult bed. Babies can also have overwhelming infections, strokes, or congenital heart disease--all of which may or may not be correctly diagnosed on autopsy.

[...]

Small numbers also make for bad science. Given how many babies and toddlers there are in the United States, 515 is a very small number. And the number of those deaths due to a "co-sleeping" parent rolling over on a child, rather than the child becoming caught in bedding or rails, is even smaller: 121. The amount of reliable, testable information available in each case is virtually nonexistent. To really understand the cause of something that claims the lives of, roughly, 1 in 100,000 babies and toddlers, you need to follow at least 1 million control cases. Only then can you examine a wide array of variables.

[...]

What else makes for bad science? Investigators with a vested interest in their conclusion. A study that concluded there was no danger lurking in the "family bed" wouldn't have been covered by the press, wouldn't have made such a stunning video as the one that accompanied the commission's report (complete with dolls being strangled by the sheets) and wouldn't have advanced anyone's career, nor proved to Congress that the agency was worth its weight in taxes.

The truth is, almost all mammals (including humans) sleep with their babies. Indeed, most human babies in most cultures sleep with their parents, and always have. Sleeping apart from babies is a mostly British, Northern European experiment that we would do well to wait on until the data is in.

[...]

Parents choose to sleep with their babies and toddlers for intensely practical and personal reasons: less crying and more sleep. A crying, unhappy baby can shape the behavior of his or her parents better than any shock ever shaped the behavior of a lab rat. So if snuggling in bed together is what it takes to end that crying, a shared bed is what a parent will choose.

[...]

Not many generations ago, the practice of breast-feeding was just about killer off by bad science that "proved" the practice was unsanitary. Now science can't say enough nice things about breast-feeding; just this week brought news that breast-fed babies are less likely to develop leukemia. Maybe in a few generations, we'll see studies that indicate that babies who sleep with their parents have fewer ear infections, do better in school, and don't engage in pseudo-science when they grow up.

Highlights of bad science from the report.