The medication cannot reach the fetus; the pattern points elsewhere
Em meio ao debate público sobre saúde materna e desenvolvimento infantil, uma análise abrangente de mais de 25 milhões de gestações concluiu que o uso de antidepressivos durante a gravidez não apresenta ligação causal clara com autismo ou TDAH nas crianças. O estudo, publicado no The Lancet Psychiatry, chega como resposta científica às afirmações do secretário de Saúde dos Estados Unidos, Robert F. Kennedy Jr., que atribuiu riscos desenvolvimentais a esses medicamentos sem respaldo em evidências. A ciência, aqui, convida à cautela interpretativa: associação não é causalidade, e o sofrimento não tratado da mãe carrega seus próprios riscos para a vida que cresce dentro dela.
- Afirmações do secretário Kennedy sobre antidepressivos e autismo geraram alarme entre gestantes e profissionais de saúde ao redor do mundo.
- A análise de 37 estudos anteriores mostrou que a associação inicial entre antidepressivos e autismo/TDAH desapareceu ao se considerar saúde mental materna, genética e histórico familiar.
- Um dado revelador: filhos de pais que tomaram antidepressivos — sem qualquer contato direto com o feto — também apresentaram taxas elevadas, apontando para fatores genéticos, não farmacológicos.
- Doses mais altas de antidepressivos não aumentaram o risco, enfraquecendo ainda mais a hipótese de que o medicamento em si seja o agente causador.
- Pesquisadores alertam que interromper antidepressivos durante a gravidez pode expor mães e bebês a riscos reais e documentados, como parto prematuro e desfechos maternos graves.
Uma análise de dados de mais de 25 milhões de gestações não encontrou evidências de que antidepressivos tomados durante a gravidez aumentem o risco de autismo ou TDAH nas crianças. O estudo, publicado no The Lancet Psychiatry e liderado por Wing-Chung Chang, da Universidade de Hong Kong, examinou 37 pesquisas anteriores envolvendo cerca de 650 mil gestações com exposição ao medicamento. Embora uma análise superficial sugerisse risco levemente elevado, essa associação se desfez ao incorporar variáveis como condições psiquiátricas maternas, histórico familiar e fatores genéticos.
O professor James Walker, da Universidade de Leeds, destacou o problema interpretativo central: comparar grupos sem controlar as condições de saúde das mães pode criar a ilusão de causalidade onde existe apenas correlação. A própria depressão ou ansiedade materna, independentemente do tratamento, pode influenciar o desenvolvimento neurológico dos filhos.
Dados adicionais reforçam essa leitura. Filhos de pais que usaram antidepressivos durante a gestação da parceira — sem qualquer via de exposição fetal — também apresentaram taxas mais altas de autismo e TDAH. O mesmo ocorreu com filhos de mães que tomaram os medicamentos antes, mas não durante a gravidez. Doses mais elevadas tampouco aumentaram o risco, sinalizando que os fármacos não são os responsáveis.
Uma exceção pontual foi identificada: dois antidepressivos tricíclicos mais antigos, amitriptilina e nortriptilina, mostraram associação com maior risco em filhos de mulheres com doenças psiquiátricas preexistentes — justamente aquelas com quadros mais graves, o que pode explicar o achado de forma independente. Esses medicamentos não são os SSRIs criticados publicamente por Kennedy.
Os pesquisadores concluem que gestantes não devem abandonar antidepressivos com base em relatórios de risco inconsistentes. Para casos de depressão leve, abordagens não farmacológicas como psicoterapia são recomendadas. Mas os riscos reais da depressão materna não tratada — incluindo aborto espontâneo e parto prematuro — devem ser pesados contra hipóteses que este estudo, em larga escala, não conseguiu sustentar.
A sweeping analysis of pregnancy data spanning more than 25 million cases has found no clear evidence that antidepressants taken during pregnancy increase the risk of autism or attention deficit hyperactivity disorder in children. The finding arrives as a direct challenge to claims made by Robert F. Kennedy Jr., the U.S. Health Secretary, who has asserted without scientific support that certain antidepressants pose developmental risks to fetuses and has also promoted the long-debunked theory linking vaccines to autism.
Researchers examined data from 37 prior studies involving nearly 650,000 pregnancies where mothers took antidepressants, compared against almost 25 million pregnancies with no such exposure. While the initial analysis showed children of mothers who used antidepressants had a slightly elevated likelihood of autism or ADHD diagnosis, this association collapsed when researchers accounted for the mothers' own mental health conditions, family history, genetic factors, and other variables known to influence neurodevelopmental risk. The work, published in The Lancet Psychiatry, was led by Wing-Chung Chang of the University of Hong Kong, who described the findings as reassuring evidence that commonly prescribed antidepressants do not raise the risk of these developmental disorders.
The distinction matters enormously. James Walker, an emeritus professor of obstetrics and gynecology at the University of Leeds, explained the crucial interpretive problem: a simple comparison between children whose mothers took antidepressants and those whose mothers did not may reveal a difference, but that difference does not prove the medication caused it. The underlying health conditions of the mothers themselves—depression, anxiety, other psychiatric illness—could account for any observed risk, independent of the drugs used to treat those conditions.
Several patterns in the data further undermine any direct causal link between medication and fetal harm. Children whose fathers took antidepressants while the mothers were pregnant showed elevated autism and ADHD rates, yet paternal medication cannot reach a developing fetus in the womb. Similarly, children whose mothers took antidepressants before pregnancy but not during it also showed increased risk. These findings point toward shared genetic and familial traits rather than direct drug exposure as the explanation. Higher doses of antidepressants did not increase risk either—another signal that the medications themselves are not the culprit.
One narrow exception emerged: in women with pre-existing mental illness, two older tricyclic antidepressants, amitriptilline and nortriptilline, showed associations with increased ADHD and autism risk in offspring. These drugs are typically reserved for patients whose depression has not responded to other treatments, suggesting the women prescribed them may have had more severe or complex psychiatric conditions that could independently influence their children's developmental outcomes. Notably, these older tricyclics are not the selective serotonin reuptake inhibitors, or SSRIs, that Kennedy has most vocally criticized.
The research team emphasized that pregnant women with depression should not stop taking antidepressants based on inconsistent or potentially confounded risk reports. For women with mild depression, non-pharmacological approaches such as psychotherapy may be preferable. But the evidence does not support abandoning medication during pregnancy out of fear of neurodevelopmental harm. The stakes of untreated maternal depression—increased risk of miscarriage, preterm birth, and poor maternal and infant outcomes—must be weighed against speculative risks that this large analysis has failed to substantiate.
Notable Quotes
Our study provides reassuring evidence that commonly used antidepressants do not increase the risk of neurodevelopmental disorders like autism and ADHD in children— Dr. Wing-Chung Chang, University of Hong Kong
If you simply compare children whose mothers took antidepressants with those whose mothers did not, you may find a difference. But that does not mean the medication caused the difference— Dr. James Walker, University of Leeds
The Hearth Conversation Another angle on the story
Why does it matter that fathers' antidepressant use also showed this pattern? That seems like the most important detail.
Because it breaks the causal chain. If the medication itself harmed the fetus, paternal medication shouldn't matter at all. The fact that it does suggests the real risk factor is something inherited—genetics, family psychiatric history, the environment the child grows up in.
So the mothers' depression itself could be the issue, not the treatment?
Exactly. Untreated depression during pregnancy carries real risks. The study is saying: don't confuse correlation with causation. Yes, children of mothers on antidepressants show slightly higher autism rates. But that's because mothers with depression are more likely to have children predisposed to these conditions anyway.
What about those two older drugs that did show a link?
They're prescribed to people with the most treatment-resistant depression. So even there, you're looking at women with more severe illness, not necessarily a drug effect. The researchers were careful not to overclaim.
Kennedy's been making a lot of noise about this. Does this study settle the question?
For anyone reading carefully, yes. But settling a question and changing public perception are different things. The evidence here is strong, but it requires understanding why confounding variables matter—which is not intuitive.