Claims About Risks Of Caffeine Consumption In Pregnancy Misleading

You may have seen claims in the news over the last week that pregnant women…

Claims About Risks Of Caffeine Consumption In Pregnancy Misleading

You may have seen claims in the news over the last week that pregnant women should avoid all caffeine, as it puts their unborn baby at risk, but are these claims really true?

The news coverage is based on a recently published academic ‘review’ article. The author — Jack James, a professor of psychology at Reykjavík University in Iceland — concludes in his article that “maternal caffeine consumption is reliably associated with major negative pregnancy outcomes” and suggests pregnant women should avoid all caffeine during pregnancy.

The review, which was published in the medical journal BMJ Evidence-Based Medicine, has been widely covered by international media such as The Guardian and The New York Times, among many other publications.

However, an analysis of its contents by Gideon Meyerowitz-Katz, an epidemiologist from the University of Wollongong in Australia, suggests its conclusions are misleading. Meyerowitz-Katz, who has written about this topic before, called out some of the article’s claims in a Twitter thread he published this week.

While it can’t be denied that high levels of caffeine intake by pregnant women can be dangerous for unborn babies, it is widely accepted, and indeed advised by health advisory bodies such as the World Health Organization, that drinking less than 300 mg caffeine per day (approximately 2-3 cups coffee) is associated with minimal health risks to mother and baby.

James’ article suggests that, contrary to this advice, all caffeine should be avoided in pregnancy. But, as highlighted by Meyerowitz-Katz, there are several problems with the review.

First, it is not based on new findings from a recent clinical study. It is a review of the already published medical literature by one person.

“This is a narrative review, which basically means that it is subject to the opinions of the author. To a great extent, narrative reviews are the academic version of opinion pieces,” Meyerowitz-Katz explained.

“What this means is that calling the paper new evidence is wrong. This review represents the opinions and expertise of Prof James, and while he seems to have read a lot that still doesn’t make it new evidence.”

Second, when Meyerowitz-Katz looked more closely at the 40 plus studies included in the review the conclusions drawn by James did not seem to align with the findings of all the original study authors.

For example, the biggest study included in the review that looked at links between pre-pregnancy caffeine intake and miscarriage was stated by James as evidence supporting his theory that all levels of caffeine consumption are harmful.

But, the study actually stated that only higher levels of coffee consumption (400 mg per day or more) could increase miscarriage risk. It should also be noted that the researchers in this study found that women who drank decaffeinated coffee had the same risk for miscarriage as those drinking caffeinated coffee and there were no links found between miscarriage and caffeinated tea or soda consumption.

Similarly, Meyerowitz-Katz pointed out two other studies — included in the analysis to back up claims that very small amounts of caffeine are harmful in pregnancy — that actually concluded that sticking to the WHO recommendation of less than 300 mg per day should be fine for pregnant women.

The third, and arguably most concerning, issue with the review article is that a table (table 2) included in the article to summarize its findings contains significant errors.

There are several mistakes, and you can see the details clearly outlined in Meyerowitz-Katz’s Twitter thread. But a couple of examples include:

  • The first column is titled ‘Risk per 100mg’ caffeine consumed – which should actually read relative risk increase per additional 100mg consumed. This may not mean much to a non-statistician, but Meyerowitz-Katz explains that the result is that all the risk estimates in the next column ‘Risk at 200mg’ caffeine consumption are wrong and overestimated.
  • The author then uses these incorrect estimates of risk to calculate the number of ‘negative pregnancy events’ associated with caffeine intake in the last column – again leading to a massive overestimate of the number of miscarriages linked to caffeine.

Meyerowitz-Katz and other researchers in the field agree that evidence suggests that high amounts of caffeine intake in pregnancy are a bad idea. It’s hard to predict what the results of future research might show, but it does seem that small amounts of caffeine intake during pregnancy may not be as damaging as implied in James’ review.

“My main take-home for anyone who is pregnant is to try not to worry about headlines, because they are usually misleading, and talk to your doctor if you’re worried,” said Meyerowitz-Katz.