Study Links Smoking Grandmas to Autistic Children

The riddle of the autism spike remains unsolved, but this study is definitely worth a second look.

Originally Published: 
smoking granny
flickr / <a href=""> Moonfall Pix </a>

It’s entirely possible to believe in the importance and efficacy of vaccines while empathizing with anti-vaxxers. In the 1980s, roughly one in 2,000 children was diagnosed with an autism spectrum disorder. Today, one in 150 8-year-olds in the U.S. has a diagnosis. It’s not a small wonder that answering the question of why that has happened has become a growth industry. After all, experts haven’t found one way to account for the spike in diagnoses. However, researchers from the University of Bristol might have found part of what has driven the uptick: maternal smoking.

The new study, published in Scientific Reports, leans on data harvest from 14,500 children in the Avon Longitudinal Study of Parents and Children — aka Children of the ’90s. Pregnant mothers were recruited in 1991 and 1992 and their kids were followed for more than two decades. Researchers found that girls were 67 percent more likely to engage in repetitive behaviors and suffer from poor social communication skills if their maternal grandmother was a smoker. Overall, having a maternal grandmother who smoked increased the rate of autism diagnoses by 54 percent. This intergenerational effect was the strongest when the mothers themselves did not smoke during pregnancy. Experts suspect that this is due to some sort of epigenetic modification that the grandmother passes on to the mother when she’s pregnant, but that connection remains unconfirmed.

As easy as it is to get excited over the possibility of isolating a cause for the autism spike, there are a lot of caveats to consider. Researchers aren’t entirely sure why grandmaternal smoking disproportionately affects girls, especially given that previous studies have linked grandmaternal smoking to different growth patterns in both sexes. Of the 14,500 children track, there were also only 177 children diagnosed with ASD. Some 7,000 exhibited traits. Due to this significantly smaller sample size, researchers were unable to analyze grandsons and granddaughters separately.

Which is all to say that this study doesn’t offer a complete explanation. It highlights a strong correlation.

Still, historical timing makes the study really compelling. Congress didn’t require health warnings on cigarettes until 1965, right around the time many of the grandmothers in question may have been pregnant. The number of adults who smoke has declined from 42.4 percent in 1965 to 15.1 percent in 2015. These smokers would likely have grandkids after 1988, when autism rates started to rise. Likewise, the number of children ages 19 to 35 months receiving vaccines increased from 69 percent in 1994 to 83 percent in 2004. It makes sense how the effects of those unrelated trends might be confused. It is also almost certainly more complicated than just cause, epigenetics, and effect.

This article was originally published on