Venlafaxine in early pregnancy and risk of birth defects

This large population-based study from five Nordic countries examined whether taking selective serotonin reuptake inhibitors (SSRIs) or venlafaxine during early pregnancy increases the risk of birth defects. Analyzing over 2.3 million births, researchers found a small increase in the overall rate of major birth defects (3.7% vs 3.1%) and heart defects (1.5% vs 1.2%) among infants exposed to these medications. However, when the researchers compared siblings from the same family — where one baby was exposed to the drug and another was not — these associations disappeared, suggesting that the small observed increase may be explained by inherited family factors or lifestyle influences rather than a direct drug effect.

Table of Contents

• Key Points
• Why This Research Matters
• What This Study Adds to What We Already Know
• How the Research Was Conducted
• Key Findings: What the Researchers Discovered
• Heart Defects in Detail
• Other Birth Defects Examined
• The Sibling Comparison: A Powerful Approach
• How These Results Compare with Previous Research
• Study Limitations: What This Research Couldn't Prove
• What This Means for Patients
• Recommendations for Women and Their Doctors
• Frequently Asked Questions
• Source Information

Key Points

• In a study of over 2.3 million Nordic births, SSRI or venlafaxine exposure in early pregnancy was linked to 3.7% versus 3.1% major birth defects.
• Sibling comparisons showed no significant increased risk of birth defects, suggesting family or lifestyle factors may explain the small difference.
• Heart defects occurred in 1.5% of exposed infants versus 1.2% unexposed, but sibling analysis did not confirm a direct drug effect.
• Absolute risk increase was about 6 additional major birth defects per 1,000 exposed infants; most exposed infants had no birth defect.
• Women should not abruptly stop antidepressants due to depression risks; discuss continuing, switching, or stopping with a healthcare provider.

Why This Research Matters

Depression is a common condition during pregnancy, affecting a significant number of expectant mothers worldwide. Selective serotonin reuptake inhibitors (SSRIs) — a class of antidepressants that includes well-known medications like fluoxetine (Prozac), citalopram (Celexa), and sertraline (Zoloft) — are the most frequently prescribed antidepressants for pregnant women. In Europe, up to 3% of pregnant women take these medications, while in North America, the figure ranges from 4% to 10%.

Venlafaxine (brand name Effexor) belongs to a different drug class called serotonin-norepinephrine reuptake inhibitors (SNRIs). It works on two brain chemicals rather than one and is increasingly being used as an alternative to SSRIs. Because more women of childbearing age are being treated for depression, understanding whether these medications affect fetal development is critically important.

Previous studies on this topic have produced conflicting results. Some studies reported an increased risk of cardiovascular (heart) birth defects, while others found no increased risk. One American case-control study, based on just 77 exposed cases and 14 exposed controls, reported that venlafaxine use in early pregnancy was associated with several types of birth defects, with odds ratios ranging dramatically from 1.5 to 6.3. However, many earlier studies had important weaknesses: uncertain information on drug use, recall bias from self-reported data, limited ability to account for other risk factors (confounders), and low statistical power, meaning they were too small to detect rare outcomes reliably.

What This Study Adds to What We Already Know

This new research overcomes many of those limitations. The Nordic countries — Denmark, Finland, Iceland, Norway, and Sweden — maintain comprehensive nationwide registries of births, prescribed medications, and birth defects. These registries cover a combined population of about 26 million people and collect data prospectively (forward in time), eliminating recall bias. Every resident is assigned a unique personal identification number at birth, allowing researchers to link data across different registries accurately.

This study is one of the largest ever conducted on this topic, with a full study cohort of 2,303,647 singleton live births. It also employed a unique sibling-controlled design, comparing outcomes between brothers and sisters from the same family where one was exposed to the medication in the womb and the other was not. This powerful approach helps separate the effects of the drug itself from inherited genetic factors and shared family environment.

How the Research Was Conducted

Study Population and Data Sources

The study included women from all five Nordic countries who gave birth to a live singleton infant (a single baby, not twins or multiples) between 1996 and 2010. Each country contributed data for different time periods, depending on when their prescription databases became available: Denmark from 1997 to 2010, Finland from 1996 to 2006, Iceland from 2003 to 2007, Norway from 2005 to 2010, and Sweden from 2006 to 2010.

Researchers obtained information from nationwide health registries covering births, dispensed drugs, birth defects, and potential confounding factors. Reporting to these registries is mandatory and regulated by national laws in each country, ensuring high-quality, complete data. The sibling cohort included 2,288 singleton live births from families where siblings were discordant — meaning one sibling was exposed to the drugs while the other was not, and one had a birth defect while the other did not.

Defining Drug Exposure

Infants were considered "exposed" if their mothers filled a prescription for an SSRI from 30 days before the first day of the last menstrual period until the end of the first trimester (defined as 97 days after the last menstrual period). This time window captures the period when the fetal organs are forming, particularly the heart, which develops in the first trimester.

The following medications were included in the analysis:

• Fluoxetine (Prozac) — 6,250 exposed infants (0.27% of all births)
• Citalopram (Celexa) — 11,193 exposed infants (0.48%)
• Paroxetine (Paxil) — 2,879 exposed infants (0.12%)
• Sertraline (Zoloft) — 7,245 exposed infants (0.31%)
• Fluvoxamine (Luvox) — 255 exposed infants (0.01%)
• Escitalopram (Lexapro) — 3,950 exposed infants (0.17%)
• Venlafaxine (Effexor) — 2,763 exposed infants (0.12%)

An additional 2,237 infants (0.10%) were exposed to more than one of these drugs (classified as "mixed"), and 7,188 infants (0.31%) were exposed to other types of antidepressants, who were excluded from the main analyses. In total, 36,772 infants (1.6%) were exposed to an SSRI or venlafaxine during the first trimester. Notably, the proportion of exposed infants increased over the study period: from 0.6% in 1996–2000 to 1.5% in 2001–05, and reaching 2.2% in 2006–10.

Identifying Birth Defects

Major birth defects were recorded from medical birth registries, patient registries, and malformation registries, with diagnoses made within 365 days after birth. During the study period, Denmark, Norway, Sweden, and Iceland used the International Classification of Diseases, 10th revision (ICD-10), while Finland used ICD-9-CM codes with the Atlanta modification for birth defects.

Cardiovascular (heart) birth defects were categorized into specific subtypes:

• Any cardiac defects (overall heart defects)
• Situs anomalies and looping defects (problems with the heart's position and basic structure)
• Conotruncal and major aortic arch anomalies (defects of the large blood vessels leaving the heart)
• Atrial or ventricular septal defects (holes in the walls between the heart's chambers)
• Atrioventricular septal defects (more complex holes involving multiple chambers)
• Right ventricular outflow tract obstruction defects (blockages on the right side of the heart)
• Left ventricular outflow tract obstruction defects (blockages on the left side of the heart)

Based on findings from previous research, the study also examined these specific non-cardiac birth defects: anal atresia, hypospadias, clubfoot, limb reduction defects, craniosynostosis (early fusion of skull bones), omphalocele (abdominal wall defect), gastroschisis (another abdominal wall defect), and cystic kidneys.

Accounting for Confounding Factors

Researchers adjusted for several potential confounders — factors that could independently affect the risk of birth defects and might be more common among women taking antidepressants. These included:

• Country of residence
• Maternal age at delivery
• Calendar year of delivery
• Birth order
• Maternal smoking during pregnancy
• Maternal diabetes (defined as dispensed antidiabetic drugs from 90 days before the last menstrual period to 97 days after)
• Concurrent use of other drugs, including antiepileptics, anxiolytics and hypnotics (anti-anxiety and sleep medications), and agents acting on the renin-angiotensin system (blood pressure medications)

Due to missing data on some confounders, the adjusted analyses included 2,145,050 births, slightly fewer than the 2,303,647 in the crude (unadjusted) analyses.

Characteristics of the Study Population

Women who took SSRIs or venlafaxine differed from those who did not in several ways. Among the 36,772 exposed women, 25.5% smoked during early pregnancy, compared with 14.0% of the 2,266,875 unexposed women. Exposed women also had higher rates of diabetes (1.4% vs 0.7%) and were much more likely to be taking other prescribed medications (14.6% vs 1.0%), particularly antiepileptics, anti-anxiety drugs, and blood pressure medications. In the sibling cohort (980 exposed vs 1,308 unexposed), exposed mothers were also more likely to take other medications (14.3% vs 4.2%), though smoking rates were similar between the two groups (29.4% vs 28.5%). Citalopram and sertraline were the most commonly used SSRIs in the study.

Statistical Analysis

The researchers used logistic regression analysis to estimate odds ratios (ORs) with 95% confidence intervals for all major and specific types of birth defects. An odds ratio above 1.0 suggests an increased risk; below 1.0 suggests a decreased risk. The 95% confidence interval indicates the range within which the true value likely falls — if the interval includes 1.0, the result is not statistically significant. For the sibling-controlled analyses, they used conditional logistic regression, comparing birth defect rates between siblings from the same family who were discordant for both drug exposure and having a birth defect.

Predefined sensitivity analyses tested different exposure windows — from the first day of the last menstrual period to 97 days after, and from 90 days before the last menstrual period to 97 days after. Additional analyses stratified by use of anti-anxiety and sleep medications, and restricted to women without previous psychiatric hospital admissions to account for the severity of the underlying psychiatric condition.

Key Findings: What the Researchers Discovered

Overall Birth Defects

Among the 36,772 infants exposed to any SSRI or venlafaxine during the first trimester, 1,357 (3.7%) had a diagnosis of a major birth defect. In comparison, 71,374 (3.1%) of the 2,266,875 unexposed infants had a major birth defect. This translates to a covariate-adjusted odds ratio of 1.13 (95% confidence interval 1.06 to 1.20) — meaning exposed infants had a 13% higher relative risk of birth defects after adjusting for confounders. However, in the sibling-controlled analysis, the adjusted odds ratio decreased to 0.92 (0.72 to 1.17), and this result was not statistically significant.

When looking at individual medications, the adjusted odds ratios for any major birth defect were:

• Fluoxetine: OR 1.25 (1.10 to 1.42) — 255 birth defects among 6,250 exposed infants (40.8 per 1,000)
• Citalopram: OR 1.19 (1.07 to 1.31) — 446 birth defects among 11,193 exposed infants (39.9 per 1,000)
• Paroxetine: OR 1.16 (0.95 to 1.41) — 116 birth defects among 2,879 exposed infants (40.3 per 1,000) — not statistically significant
• Sertraline: OR 1.06 (0.93 to 1.22) — 248 birth defects among 7,245 exposed infants (34.2 per 1,000) — not statistically significant
• Fluvoxamine: OR 0.77 (0.38 to 1.56) — 9 birth defects among 255 exposed infants (35.3 per 1,000) — not statistically significant
• Escitalopram: OR 0.89 (0.73 to 1.10) — 107 birth defects among 3,950 exposed infants (27.1 per 1,000) — not statistically significant
• Venlafaxine: OR 1.06 (0.86 to 1.32) — 95 birth defects among 2,763 exposed infants (34.4 per 1,000) — not statistically significant

Heart Defects in Detail

Because heart defects were the most consistently reported association in previous research, this study examined them in extensive detail. The prevalence of overall cardiac (heart) birth defects was 1.5% among infants exposed to any SSRI or venlafaxine (564 affected infants out of the exposed group), compared with 1.2% among unexposed infants (26,745 affected). The adjusted odds ratio was 1.15 (1.05 to 1.26). In the sibling-controlled analysis, however, the odds ratio dropped to 1.06 (0.91 to 1.24) — a result that was not statistically significant.

With the exception of escitalopram, exposure to each specific SSRI and to venlafaxine was associated with an increased prevalence of cardiac birth defects, with adjusted odds ratios ranging from 1.13 to 1.34:

• Fluoxetine: 17.0 per 1,000 affected, OR 1.34 (1.10 to 1.63)
• Citalopram: 15.7 per 1,000, OR 1.15 (0.98 to 1.34) — not statistically significant
• Paroxetine: 18.4 per 1,000, OR 1.30 (0.96 to 1.75) — not statistically significant
• Sertraline: 15.0 per 1,000, OR 1.13 (0.93 to 1.38) — not statistically significant
• Venlafaxine: 14.5 per 1,000, OR 1.14 (0.82 to 1.57) — not statistically significant

Atrial and Ventricular Septal Defects

Septal defects — commonly described as "holes in the heart" — were the most frequently observed cardiac finding. Among exposed infants, 376 (10.2 per 1,000) had an atrial or ventricular septal defect, compared with 17,503 (7.7 per 1,000) in unexposed infants. The adjusted odds ratio was 1.17 (1.05 to 1.31). Specific drug results showed:

• Fluoxetine: 11.8 per 1,000, OR 1.45 (1.15 to 1.84) — statistically significant
• Citalopram: 10.6 per 1,000, OR 1.17 (0.97 to 1.42) — not statistically significant
• Paroxetine: 12.9 per 1,000, OR 1.37 (0.96 to 1.95) — not statistically significant
• Sertraline: 9.7 per 1,000, OR 1.05 (0.82 to 1.35) — not statistically significant
• Escitalopram: 7.1 per 1,000, OR 0.82 (0.54 to 1.25) — not statistically significant
• Venlafaxine: 9.1 per 1,000, OR 1.10 (0.73 to 1.64) — not statistically significant

Right Ventricular Outflow Tract Obstruction Defects

These defects — blockages that obstruct blood flow from the right side of the heart to the lungs — showed the strongest association with drug exposure. Among exposed infants, 73 (2.0 per 1,000) had this type of defect, compared with 2,705 (1.2 per 1,000) in unexposed infants. The adjusted odds ratio for any SSRI exposure was 1.48 (1.15 to 1.89), with estimates for specific drugs ranging from 1.40 to 2.54:

• Fluoxetine: 2.4 per 1,000, OR 1.95 (1.17 to 3.25) — statistically significant
• Citalopram: 2.3 per 1,000, OR 1.65 (1.10 to 2.48) — statistically significant
• Paroxetine: 3.5 per 1,000, OR 2.54 (1.31 to 4.90) — statistically significant
• Sertraline: 2.1 per 1,000, OR 1.40 (0.81 to 2.42) — not statistically significant

However, in the sibling-controlled analysis for this defect, the adjusted odds ratio decreased dramatically to 0.56 (0.21 to 1.49), indicating no significant association and suggesting the full-cohort finding may be explained by factors other than the medication itself.

Other Heart Defect Subtypes

For conotruncal and major aortic arch anomalies (defects of the large arteries leaving the heart), exposure to paroxetine was associated with a statistically significant increased prevalence, with an adjusted odds ratio of 2.27 (1.01 to 5.07). Associations with other SSRIs were less clear and did not reach statistical significance. No significant associations were found for situs anomalies and looping defects, atrioventricular septal defects, or left ventricular outflow tract obstruction defects, though citalopram showed a borderline association with left ventricular outflow tract obstructions (OR 1.56, 0.99 to 2.46).

Other Birth Defects Examined

The study also looked at several non-cardiac birth defects that had been previously linked to SSRI use. The prevalence of clubfoot (a condition where the foot is twisted out of shape) and omphalocele (a defect where abdominal organs protrude through the belly button area) was increased after exposure to any SSRI. Both citalopram and sertraline were individually associated with an increased prevalence of clubfoot, and sertraline was also associated with anal atresia (a malformation of the anal opening).

Encouragingly, the study found no increased prevalence of hypospadias (a genital defect in males), limb reduction defects, craniosynostosis (premature fusion of skull bones), gastroschisis (another abdominal wall defect), or cystic kidneys among exposed infants.

The Sibling Comparison: A Powerful Approach

The sibling-controlled analysis is one of the most innovative and valuable aspects of this study. Researchers identified 895 families with at least one pregnancy exposed to SSRIs or venlafaxine and at least one child born with a birth defect. By comparing outcomes between siblings — who share approximately 50% of their genes and were raised in the same family environment — this design effectively controls for familial factors that could otherwise distort the results.

The results were striking. In the sibling-controlled analyses, the adjusted odds ratios were attenuated (reduced) compared with the full cohort analyses:

• Any major birth defect: sibling-adjusted OR 0.92 (0.72 to 1.17) — no longer statistically significant
• Any cardiac defect: sibling-adjusted OR 1.06 (0.91 to 1.24) — no longer statistically significant
• Right ventricular outflow tract obstructions: sibling-adjusted OR 0.56 (0.21 to 1.49) — no longer statistically significant

The researchers also confirmed that the order of events between siblings — whether the exposed sibling was born first or second, or whether the first-born had the birth defect — did not change the results. The confidence intervals in the sibling analyses were wider, which is expected because the number of families included was smaller, but the direction of the findings was consistent: once familial factors were accounted for, the apparent risk from the medications largely disappeared.

How These Results Compare with Previous Research

The findings from this study both confirm and extend previous research. The observed increased prevalence of overall major birth defects associated with SSRI exposure (13%) was lower than the 21–30% increased risk reported in earlier Danish studies, and it extends the findings of a previous Finnish study that found no substantial increased risk of overall birth defects.

The 30% increase in cardiovascular defects seen with paroxetine or fluoxetine exposure in this study is of a similar size to effects reported from databases in the United Kingdom and the United States. However, the researchers noted that earlier studies reporting risks associated with combined use of SSRIs and benzodiazepines (anti-anxiety medications) were not confirmed — stratification by anti-anxiety and sleep medication use did not result in substantial variation in risk estimates between groups.

Importantly, the previously reported American case-control study — which found odds ratios between 1.5 and 6.3 for venlafaxine and several birth defects — was based on only 77 exposed cases and 14 exposed controls, far too small to draw reliable conclusions. This Nordic study, with 2,763 venlafaxine-exposed infants and 95 exposed infants with birth defects, found no significant association between venlafaxine and any birth defect (OR 1.06, 0.86 to 1.32 for overall birth defects).

The sensitivity analyses confirmed the robustness of the main findings. Using two different time windows for defining exposure produced effect estimates similar to the primary analyses, and restricting the analyses to women without previous psychiatric hospital admissions did not substantially change the results.

Study Limitations: What This Research Couldn't Prove

While this is one of the most rigorous studies ever conducted on this topic, it has limitations that should be acknowledged:

• Observational design: Because this was not a randomized controlled trial, it cannot prove cause and effect. It can only show associations.
• Prescription data vs. actual use: The study identified exposure by filled prescriptions, but researchers could not confirm that women actually took the medications as prescribed.
• Confounding by indication: Women taking antidepressants may have more severe depression, and depression itself may affect pregnancy outcomes. Although the researchers adjusted for many factors and used sibling comparisons, unmeasured factors could still influence results.
• Sibling analysis power: While the sibling design is powerful, it included fewer participants (2,288 births in 895 families), resulting in wider confidence intervals and less precision for rare outcomes.
• Different time periods by country: Each country contributed data from different years, which could introduce subtle variations in prescribing practices and diagnostic coding.
• No patient involvement: The study did not include patients in its design or planning.
• Exclusion of other antidepressants: The study focused only on SSRIs and venlafaxine; other antidepressant classes were not examined.

What This Means for Patients

For women who are pregnant, planning to become pregnant, or currently taking antidepressants, these findings are largely reassuring — but they require careful interpretation. The absolute increase in risk is small. Among 1,000 unexposed infants, approximately 31 will have a major birth defect. Among 1,000 infants exposed to SSRIs or venlafaxine, approximately 37 will have a major birth defect. The absolute difference is about 6 additional infants per 1,000 (0.6%).

Similarly, for heart defects, the difference is about 3 additional infants per 1,000 (1.5% vs 1.2%). Even the strongest association found — right ventricular outflow tract obstruction defects with paroxetine exposure — represented an absolute rate of 3.5 per 1,000 exposed infants, meaning more than 99.6% of exposed infants did not develop this condition.

The finding that the increased risks disappeared in the sibling-controlled analyses is particularly reassuring. It suggests that the small excess of birth defects seen among exposed infants in the general population may be related to underlying factors — such as genetics, family environment, or lifestyle — rather than to a directly harmful (teratogenic) effect of the medications themselves.

The study authors concluded: "In this large Nordic study no substantial increase was found in prevalence of overall cardiac birth defects among infants exposed to SSRIs or venlafaxine in utero. Although the prevalence of septal defects and right ventricular outflow tract defects was higher in exposed infants, the lack of an association in the sibling controlled analyses points against a teratogenic effect of these drugs."

Recommendations for Women and Their Doctors

These findings have practical implications for clinical care. Based on this research, women should not abruptly stop taking antidepressants when they discover they are pregnant. Untreated depression during pregnancy carries its own risks — to both the mother and the developing baby — and the evidence from this study suggests that the risk of birth defects from SSRIs or venlafaxine is small, if it exists at all.

1. Have an open conversation with your healthcare provider. Discuss the risks and benefits of continuing, switching, or stopping antidepressant treatment during pregnancy, weighing both the potential medication risks and the risks of untreated depression.
2. Do not stop medications abruptly. Suddenly discontinuing antidepressants can cause withdrawal symptoms and a return of depression, which may be harmful to both mother and baby.
3. Consider the absolute risks, not just relative risks. While relative increases of 13–15% sound concerning, the actual chance of a birth defect remains approximately 97% or higher for most women taking these medications.
4. Address modifiable risk factors. The study found that women taking antidepressants were more likely to smoke and to have diabetes. Quitting smoking and managing blood sugar during pregnancy are things women can do that are proven to reduce birth defect risks.
5. Consult a specialist if needed. Women taking paroxetine specifically may want to discuss this study's findings with their doctor, as paroxetine showed the strongest associations with certain heart defects, although still not confirmed as a causal relationship.
6. Talk to a fetal medicine specialist if you have concerns. Women who have already taken these medications during early pregnancy can ask about additional ultrasound monitoring, including detailed fetal heart imaging.

Every pregnancy carries a baseline risk of birth defects of roughly 3%, regardless of medication exposure. These medications may add a very small amount to that baseline risk — or possibly no additional risk at all, as the sibling analysis suggests. For many women, the benefits of treating depression during pregnancy — improved mood, better self-care, healthier lifestyle choices, and stronger bonding with the baby — will outweigh the small potential risks identified in this study.

Frequently Asked Questions

What does this Nordic study say about taking SSRI antidepressants or venlafaxine during early pregnancy and birth defect risk?

In more than 2.3 million births, infants exposed to SSRIs or venlafaxine in the first trimester had a slightly higher rate of major birth defects (3.7% versus 3.1%). However, comparing siblings where one was exposed and one was not, this difference disappeared, suggesting family factors may explain the small increase rather than a direct drug effect.

Should I stop taking my antidepressant if I find out I am pregnant?

The study's authors advise against abruptly stopping antidepressants, because untreated depression during pregnancy carries its own risks. The evidence suggests the risk of birth defects from these medications is small, if it exists at all. Talk to your healthcare provider about the risks and benefits of continuing, switching, or stopping treatment.

Why did the sibling comparison change the findings?

The study compared siblings from the same family where one baby was exposed to the medication and another was not. These siblings share genes and family environment. In this analysis, the increased rates of birth defects disappeared, suggesting inherited family factors or lifestyle, not the drug itself, may explain the small increase seen in the general population.

Were heart defects increased in infants exposed to SSRIs or venlafaxine?

In the full group, heart defects occurred in 1.5% of exposed infants versus 1.2% of unexposed infants. But in the sibling comparison, the association was no longer statistically significant. Septal defects and right ventricular outflow tract defects appeared more often in exposed infants, yet sibling analysis did not confirm a direct drug effect.

Does venlafaxine increase the risk of birth defects?

This large Nordic study found no significant association between venlafaxine and any birth defect. The odds ratio for overall birth defects was 1.06, with a confidence interval that included 1.0. An earlier American study suggesting higher risks was very small, with only 77 exposed cases, and was not confirmed here.

What should I discuss with my doctor about antidepressant use in pregnancy?

Have an open conversation about risks and benefits of continuing, switching, or stopping treatment. Do not stop suddenly. Also address modifiable risks: women taking antidepressants were more likely to smoke and have diabetes, so quitting smoking and managing blood sugar can lower birth defect risk. If concerned, ask about fetal heart ultrasound monitoring.

Source Information

Original article title: venlafaxine in early pregnancy and risk of birth defects

Authors: Kari Furu, Helle Kieler, Bengt Haglund, Anders Engeland, Randi Selmer, Olof Stephansson, Unnur Anna Valdimarsdottir, Helga Zoega, Miia Artama, Mika Gissler, Heli Malm, Mette Nørgaard

Journal: The BMJ (British Medical Journal), 2015;350:h1798

Published: 17 April 2015 (accepted 10 March 2015)

DOI: 10.1136/bmj.h1798

This patient-friendly article is based on peer-reviewed research. The original study was funded and conducted using nationwide health registry data from Denmark, Finland, Iceland, Norway, and Sweden. As with all medical research, this study represents a snapshot of evidence at a point in time, and women should consult current healthcare guidance when making decisions about medication use during pregnancy.