While many people know that smoking is the leading cause of preventable death in the US, it’s less appreciated that some of those deaths are newborn babies.
Researchers don’t fully understand why cigarettes increase the risk of infant death, but they think it has something to do with nicotine’s effect on brain regions that interfere with a baby’s sleeping and breathing patterns. Smoking is also known to restrict the blood flow that carries vital oxygen and nutrients between mom and baby.
When smoking kills, it can happen quickly. Roughly 3,600 babies in the US die suddenly every year for unknown reasons. The blanket term for these unexplained deaths is SUID, or sudden unexpected infant deaths, of which SIDS is the most well-known type.
In a new study in Pediatrics, researchers estimated that if expectant moms would just quit smoking, we could prevent 800 of those deaths.
For the paper, a collaboration between Microsoft and the Seattle Children’s Research Institute, researchers analyzed national vital statistics data from the Centers for Disease Control and Prevention on the 20 million US births and more than 19,000 cases of sudden infant death that occurred between 2007 and 2011.
The Microsoft team built a computational model to look at how the deaths correlated with maternal smoking behavior, controlling for a variety of potential confounding factors — or other variables that might have explained the increases in sudden deaths — including race and mothers’ education level.
“We wanted to make it so that the differences we saw in the data couldn’t be chalked up to differences in race or differences in any of these other variables,” said the study’s lead author, Tatiana Anderson, a PhD fellow at the Seattle Children’s Research Institute, “and we still found this increased risk.”
To be more specific, the researchers discovered that any smoking during pregnancy — even a single cigarette — was associated with a doubling of risk of a newborn baby dying suddenly in his or her sleep.
And, said Anderson, “every additional cigarette increased [sudden infant death] risk. So smoking a pack a day triples your risk, and even if you quit by the first trimester, that still results in a 50 percent increase in the risk of sudden infant death.”
There were some limitations to the study. For one, the researchers can only describe correlations between the variables they looked at, but not whether one caused the other. In addition, they didn’t have data on mothers’ alcohol consumption, which might have had an impact along with smoking.
The overall risk of SUID is also small — less than 1 percent — and it’s difficult to know how many deaths are actually caused by smoking. SUID is an umbrella term for unexplained newborn deaths. It includes SIDS (which means an autopsy was performed but the cause of death is unknown); unknown (there was no autopsy and no known cause of death); and accidental death (a child was found in an unsafe sleeping position, an autopsy was done, but there’s no known cause of death).
“It is still a small percentage of smoking mothers who have infants that die of SIDS or SUID,” explained Anderson. “We know statistically a certain number of cases are caused by smoking, but not exactly which ones, so they are still labeled as SIDS or SUID.”
But there’s one thing we know for sure: A surprisingly large proportion of women in America continue to smoke during pregnancy, increasing the risk of sudden infant death.
Too many American women smoke while pregnant
According to the latest data from the CDC, in 2016, 7 percent of women who gave birth smoked while pregnant across the US. For women ages 20 to 24, the pregnancy smoking rate was 11 percent.
Some states have shockingly high rates of maternal smoking. In West Virginia, a quarter of moms smoke while expecting. That was followed by Kentucky, Montana, Vermont, and Missouri, where more than 15 percent of moms smoked.
For John Kahan, the chief data analytics officer at Microsoft who was a key player in the study behind the scenes, that’s too many.
In October 2003, Kahan and his wife welcomed their fourth child — and first son — Aaron Matthew. “We were blessed that we had a few hours that no one knew anything was wrong,” he recalls. “We have pictures of my parents and other children holding Aaron.”
The baby boy seemed completely healthy, just like Kahan’s three daughters. But Kahan’s son never left the hospital.
About six hours after his birth, Aaron stopped breathing. Within three days, following doctors’ attempts to revive him, Aaron was declared dead.
To get closure on Aaron’s death, Kahan gave to charity over the years, even raising money for SUID research by climbing Mount Kilimanjaro in 2016. “I wanted to make sure no other parent experiences this,” he told Vox.
That same year, his personal tragedy took on a new — and public — meaning. A data scientist colleague at Microsoft, Juan M. Lavista Ferres, saw a photo of Aaron in Kahan’s office and asked about the baby boy. Kahan choked up, explaining that Aaron had died.
While Kahan was climbing Kilimanjaro, Lavista Ferres decided to download publicly available CDC data on births and infant deaths in the US and use Microsoft’s vast array of machine learning tools to analyze the data for patterns.
One of the first studies from that data set is the Pediatrics paper, on which Lavista Ferres is a co-author. And it offers one of the most granular looks to date at smoking as a contributor to sudden unexplained infant death.
“Since Aaron died, roughly 60,000 children in the US alone died, and their parents and grandparents don’t know why they died,” said Kahan. All told, the SUID rate has been flat since the 1990s, after public health campaigns informing parents of the risks of putting their babies to sleep on their stomachs led to a 40 percent decrease in the US SIDS rate.
Kahan and his wife never smoked, so the paper won’t shed any light on their situation. But Kahan hopes it’ll raise awareness about the problem and help others avoid the tragedy his family had to live through: “My mission is to ensure no other parent experiences the sudden loss of a child.”