Mothers' Depression, Not SSRI Use, Best Explains Researchers' Results

Back in October, researchers published the findings from a study that suggested that mothers who take a common form of antidepressants (SSRIs like Prozac) while pregnant are at greater risk for producing offspring that will later have speech or language problems.

However, this month, the researchers got some push back in the journal where the original study was published. And in reviewing the results of the study, it appears the researchers overstated the association and import of the relationship they found.

There are a few classes of antidepressants prescribed to treat depression, but by far the most common class is known as selective serotonin reuptake inhibitors (SSRIs). SSRIs have been studied more extensively than virtually any other class of psychiatric drug, both because they are so commonly prescribed and because people take them over such long periods of time (years, usually).

The original study (Brown et al., 2016) examined the health records of the Finnish offspring of 15,596 whose mothers with depression used SSRIs during pregnancy; 9537 whose mothers had depression but did not use SSRIs in pregnancy; and 31,207 whose mothers did not have a psychiatric diagnosis (or a history of purchasing SSRIs).

As Medscape originally reported:

After adjustment for covariates, analysis showed a 37% greater risk for speech and language disorders among children of mothers who purchased SSRIs at least twice during their pregnancy compared with children whose mothers had depression but did not receive medication (hazard ratio [HR], 1.37; 95% confidence interval [CI], 1.11 – 1.70; P = .004).

The risk was 63% greater in comparison with children of mothers who did not have psychiatric illness (HR, 1.63; 95% CI, 1.37-2.01; P < .001).

When reporting on data, it’s fine to pose risk in terms of overall greater chances of something happening like this. But it also helps your audience to understand what this means in real terms.

So it was up to other researchers and professionals to help put this research into real-world context. Too bad it happened in the JAMA Letters section — a place where few media outlets look. Rasminsky & Burt (2017) skewered the original researchers’ findings:

Buried in the online Supplement of the article, we learn that the rate of speech/language delay was 1.62% in the SSRI-exposed group vs 1.85% in the depression-exposed group, compared with 1.04% of the unexposed group [with no psychiatric illness]. In other words, both depression and antidepressants appear to increase the risk for speech/language delay, but neither by very much.

There’s also much we don’t know: does a mother’s psychiatric status in the postpartum period impact the trajectory of the speech delay? When speech delay is diagnosed early, does it resolve with treatment?

You see how much more informative those statistics are in explaining this finding?

  • In mothers without psychiatric illness, we can expect speech/language delays in about 1 percent of their offspring
  • In mothers with untreated depression, we can expect speech/language delays in about 1.85 percent their offspring
  • In mothers with SSRI-treated depression, we can expect speech/language delays in about 1.65 percent of their offspring

As Miller et al. (2017) concluded, “The authors’ findings appear to have more to do with the mothers’ illness than their treatment. When the authors compared pregnant mothers with depression-related psychiatric disorders taking SSRIs with pregnant mothers with depression-related psychiatric disorders not taking SSRIs, they found no difference in the adjusted risk for speech/language, scholastic, or motor disorders” [emphasis added].

What the original study clearly shows isn’t that SSRIs are causing any kind of problem — it’s the underlying depression that is at fault. In fact, treatment with the SSRIs actually appears to be reducing the risk of speech/language delays in depressed moms’ children. This from a study with “inherent study design difficulties (Cohen & Nonacs, 2016) and small absolute differences” (Rasminsky & Burt, 2017).

It’s a shame this isn’t what the researchers focused on when discussing their results. Instead, they apparently decided to focus primarily on the comparison of SSRI use to moms without any psychiatric illness. While this results in flashier headlines, in my opinion, it completely skews the results.

Researchers have a responsibility to put their findings into proper context. Failure to do so could very well result in moms making ill-informed choices to quit their antidepressant medication — unnecessarily putting both themselves and their child at even greater risk.



Brown et al. (2016). Association of selective serotonin reuptake inhibitor exposure during pregnancy with speech, scholastic, and motor disorders in offspring. JAMA Psychiatry, 73, 1163-1170.

Cohen, L.S. & Nonacs, R. (2016). Neurodevelopmental implications of fetal exposure to selective serotonin reuptake inhibitors and untreated maternal depression: weighing relative risks. JAMA Psychiatry, 73, 1170-1172.

Miller et al. (2017). Serotonin Reuptake Inhibitor Use During Pregnancy. JAMA Psychiatry. doi:10.1001/jamapsychiatry.2016.4073

Rasminsky, S. & Burt, VK. (2017). Serotonin Reuptake Inhibitor Use During Pregnancy. JAMA Psychiatry. doi:10.1001/jamapsychiatry.2016.4068