Did The CDC Overestimate The Number Of Alcohol-Exposed Pregnancies?

By Kelly Burch 01/18/19

A new study suggests that CDC may have inflated the number of pregnancies affected by alcohol use. 

pregnant woman with alcohol beverage

In 2016, the Centers for Disease Control and Prevention caused a stir by saying that women of childbearing age who aren’t on birth control should abstain from drinking so that they didn’t unintentionally have pregnancies affected by alcohol.

The CDC said at the time that 3.3 million women were at risk of having alcohol-exposed pregnancies each month. 

However, a new study finds that the CDC vastly overestimated the number of pregnancies that might be affected by alcohol. 

“We estimate 731,000 U.S. women had [alcohol-exposed pregnancies] and 481,000 resulted in [alcohol-exposed births],” researchers wrote. “Under our assumptions, the estimated expected actual number of [alcohol-exposed pregnancies] is 2.5 million less than the CDC estimate of the number at risk of an [alcohol-exposed pregnancies].” 

The study, published in the January edition of the journal Women’s Health Issues, was led by researchers Sarah Roberts and Kristen Thompson from the University of California San Francisco’s Department of Gynecology. They used the same data set that the CDC did, but they adjusted how the number of alcohol-affected pregnancies were calculated. They also considered that because of miscarriage and abortion, not all pregnancies lead to births. 

In the original calculations, the CDC assumed that any woman who both had an alcoholic drink and who had unprotected sex in the previous month could have a pregnancy affected by alcohol. However, this inflated the number of pregnancies affected, Roberts told Quartz. 

“That’s not the right way to do this,” she said. 

The calculations gave the maximum number of alcohol-exposed pregnancies that could possibly occur, rather than a more realistic, likely estimate, Roberts explained. 

In the new calculations, researchers raised the number of drinks a woman had to have to be considered “at risk” from one to seven. Roberts said this reflects the CDC’s classification of what level of drinking is considered potentially harmful. They tempered the likelihood of pregnancy resulting from unprotected sex using national data that shows that just 38% of unprotected sex leads to pregnancy, and 64% of those pregnancies lead to births. As a result, the study estimated the amount of pregnancies actually affected by alcohol as less than 25% of the original CDC estimate.  

The high number the CDC put out “contributed to a sense of moral panic around the topic,” Roberts said. In the future, it’s important that the agency take a more wholistic and realistic picture of the risk of drinking and unprotected pregnancy. 

“It would be great if future efforts could take into account this method,” she said. 

Please read our comment policy. - The Fix
Kelly Burch Contrib.jpg

Kelly Burch writes about addiction and mental health issues, particularly as they affect families. Follow her on TwitterFacebook, and LinkedIn.