Antidepressants Use Among Pregnant Women in Denmark From 2001 to 2023: A Population‐Level Drug Utilization Study

ABSTRACT

We evaluated trends in antidepressant use among pregnant women in Denmark from 2001 to 2023, comparing them with a matched comparison group from the general population. Data from the Danish Health Data Authority were used to assess annual antidepressant prescription redemption rates per 1000 pregnancies, with a focus on total antidepressant use, selective serotonin reuptake inhibitors (SSRIs) and serotonin noradrenaline reuptake inhibitors (SNRIs). Antidepressant use among pregnant women increased fivefold in the study period. In the comparison group, use increased 2.3‐fold. After a decline following 2011, antidepressant use among pregnant women began to rise again in 2019, driven primarily by sertraline, which accounted for 76% of prescription redemptions in 2023. Other drugs such as escitalopram, duloxetine, mirtazapine and amitriptyline showed modest increases, while the use of most other antidepressants declined or stabilized. This upward trend likely reflects the influence of updated clinical guidelines and greater clinical acceptance of antidepressant use during pregnancy. Our findings emphasize how prescribing practices are sensitive to shifts in public discourse and guideline revisions, highlighting the importance of ongoing pharmacovigilance in managing maternal mental health.

Plain Language Summary

We evaluated antidepressant use in pregnancy in Denmark from 2001 to 2023. Antidepressant use in pregnancy increased fivefold, compared with a little more than twofold in the comparison group. After a decline around 2011, use rose again from 2019, mainly due to sertraline, which by 2023 was the most prescribed antidepressant. We observed minor increases for a few other antidepressants, but most remained stable or declined. These latter trends likely reflect updated treatment guidelines and growing acceptance of antidepressant use in pregnancy.

1. Introduction

Worldwide, over 10% of pregnant and postpartum women suffer from depression, a serious condition that poses significant risks to both mother and child [1]. In Denmark, sertraline is currently the first‐line treatment among selective serotonin reuptake inhibitors (SSRIs) when non‐pharmacological interventions are insufficient [2, 3, 4]. As safety data on the use of antidepressants in pregnancy has developed substantially over the last 20 years, serotonin and noradrenaline reuptake inhibitors (SNRIs) are emerging as an alternative [5]. Despite large amounts of reassuring safety data, the use of antidepressants during pregnancy remains somewhat controversial, as concerns about fetal safety continue to influence prescribing. Throughout the study period, the preferred antidepressant has shifted, reflecting clinical uncertainty and evolving perceptions of safety. This variation in prescribing underscores the need to better understand how and why patterns of use have changed over time period [6, 7]. In this study, we investigate population‐level trends in antidepressant use among pregnant women in Denmark from 2001 to 2023.

2. Materials and Methods

The study was conducted in accordance with the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies [8].

2.1. Data Sources

Data on the number of prescription drug users among pregnant women were obtained from the publicly accessible Danish health database ‘eSundhed’, administered by the Danish Health Authorities [9]. In ‘eSundhed’, drug utilization among pregnant women is assessed by linking pregnancy and maternal data from the Danish Medical Birth Register with individual‐level prescription data on antidepressant use from the Register of Medicinal Product Statistics (RMPS) [10, 11]. Data linkage is achieved through each woman's unique Danish civil registration number, ensuring that each woman is counted only once, regardless of the number of prescriptions she redeems during pregnancy [12]. The Danish Medical Birth Register holds data on all pregnancies resulting in live births or in stillbirths after gestational week 22, since 1973, while the RMPS constitutes a nationwide prescription database in Denmark. This register has been maintained since 1994 [10, 11]. For more details on ‘eSundhed’, see Supporting Information.

2.2. Study Drugs

All antidepressants available in Denmark during the period 2001–2003 were examined individually, including sertraline (N06AB06), escitalopram (N06AB10), citalopram (N06AB04), fluoxetine (N06AB03), paroxetine (N06AB05), duloxetine (N06AX21), venlafaxine (N06AX16), amitriptyline (N06AA09), nortriptyline (N06AA10), clomipramine (N06AA04), dosulepin (N06AA16), mirtazapine (N06AX11), mianserin (N06AX03), vortioxetine (N06AX26), agomelatine (N06AX22) and isocarboxazid (N06AF01).

In addition, SSRIs and SNRIs were also analysed as therapeutic classes.

2.3. Study Population and Data Analysis

The annual use of antidepressants among pregnant women in Denmark was assessed relative to a comparison group from 2001 to 2023.

Within ‘eSundhed’ a matched comparison group was created. Ten women were randomly selected for each pregnant woman. These women had the same year and month of birth and lived in the same region. Women in the comparison group were required to have resided in Denmark for at least 1 year prior to selection to ensure completeness and accuracy of prescription data [12]. The larger comparison group (n = 10) is used to reduce random variance, and in subsequent data presentation, the number in the comparison group is divided by 10 to enable direct comparison with the pregnant cohort.

All data utilized in this study are fully anonymized. No statistical inferential analyses were performed.

3. Results

3.1. Overall Use of Antidepressants

The overall use of antidepressants among pregnant women increased nearly fivefold from 8 per 1000 in 2001 to 39 per 1000 pregnant women in 2023. The highest use was observed in 2011 (41 per 1000), followed by a sharp decline in the following year. Subsequently, consumption fluctuated until 2018, whereafter a renewed increase was observed. Antidepressant use within the comparison group also increased between 2001 and 2023, from 38 to 88 per 1000, and peaked in 2011 (93 per 1000) with a subsequent decline, albeit less sharp than that observed in the pregnant population. Over the 7‐year period from 2011 to 2018, the comparison group experienced a 27% decline in prescription redemption rates, while pregnant women showed an even greater reduction of 36%. As observed among the pregnant population, use of antidepressants increased in the comparison group. See Figure 1.

FIGURE 1.

Use of overall antidepressants and SSRI in Denmark from 2000 to 2023 per 1000 pregnancies. ¹SSRI, selective serotonin reuptake inhibitor.

3.2. SSRIs

SSRI use among pregnant women increased nearly fivefold from 6.9 in 2001 to 32 per 1000 in 2011. This was followed by a 32% decline in 2012. From 2012 to 2019, rates remained stable, but increased by 23% from 2020 to reach 30 per 1000 in 2023. See Figure 1. The time trend of individual SSRI use is illustrated in Video S1.

3.3. Sertraline

Between 2001 and 2023, sertraline use rose annually by an average of 1.0 per 1000 in pregnant women and 1.8 per 1000 in the comparison group. Sertraline consumption in the comparison group increased 6.5‐fold, while sertraline use in the pregnant population saw a 17‐fold increase from 1.4 to 24 per 1000. See Figure 2.

FIGURE 2.

SSRI and SNRI use in Denmark from 2000 to 2023 per 1000 pregnancies. ¹SSRI, selective serotonin reuptake inhibitor; ²SNRI, serotonin and noradrenaline reuptake inhibitor.

3.4. Other SSRIs

The use of citalopram among pregnant women increased from 2.0 in 2001 to a peak of 16.8 in 2010 and declined to 3.2 per 1000 in 2023. In the comparison group, usage declined from 11.3 to 6.4 per 1000.

Escitalopram use was very low at about or less than 1 per 1000 pregnant women from 2002 to 2005. From 2008, the use increased to 2 per 1000 pregnant women in 2023. In the comparison group, usage rose 24‐fold, from 0.2 in 2002 to 4.8 per 1000 in 2023.

In contrast, the use of fluoxetine among pregnant women declined from 2.5 in 2001 to 0.6 per 1000 in 2023. In the comparison group, usage declined from 4.3 to 2 per 1000.

Paroxetine use among pregnant women was very low in 2001 and declined even further from 1.4 in 2001 to 0.3 per 1000 in 2023. The comparison group exhibited a comparable pattern.

3.5. Serotonin and Noradrenaline Reuptake Inhibitors

There were no joint data for SNRIs as a group, but duloxetine and venlafaxine were examined separately. Results are illustrated in Figure 2.

3.6. Other Antidepressants

Use of antidepressants other than SSRIs and SNRIs remained generally low, with typically fewer than 1 per 1000 pregnant women redeeming a prescription as illustrated in Figures S1–S3. Amitriptyline and mirtazapine were exceptions, showing an increase from 0.2 to 1.3 per 1000 and from 0.3 to 1.9 per 1000, respectively, between 2001 and 2023.

4. Discussion

Antidepressant use among women, both in the pregnant and the comparison population, increased from 2001 to 2023. Prescription redemptions among pregnant women rose nearly fivefold, while the comparison population saw a 2.3‐fold increase. By 2023, 3.9% of pregnant women in Denmark had redeemed antidepressants, making it a common clinical practice. Among them, 76% used sertraline, equating to more than 1 in 34 pregnant women redeeming a sertraline prescription.

Antidepressant consumption increased steadily until 2011, followed by a sharp decline. Only sertraline, duloxetine, mirtazapine and amitriptyline have surpassed their 2011 levels, while other antidepressants have not recovered.

SSRI consumption, primarily driven by sertraline, has risen steadily since 2001. The increase in sertraline use persists despite an overall decline in antidepressant consumption from 2011 to 2018. Sertraline has largely replaced other treatments, and this is aligning with updated guidelines recommending its use [13]. An observed increase in the use of duloxetine among pregnant women likely reflects growing clinical evidence and guideline recommendations supporting its use [14].

The 2011 decline in SSRI use may stem from multiple factors, including reduced depression prevalence or lower prescription rates due to guideline changes or prescriber behaviour influenced by patient preferences. However, a decrease in depression and anxiety prevalence is unlikely, as global rates are rising [15, 16]. Reviewing the guidelines from the Danish Medicines Agency (2011), the decision support platform www.pro.medicin.dk (2005–2022) and the Danish Psychiatric Association (2014 and 2022), no substantial revisions explaining this shift were found [2, 13, 17] (T. Nødvig, personal communication, 2024). A more plausible explanation is heightened media focus in 2011 on antidepressant use in pregnancy. Several major Danish media outlets featured stories of severe adverse pregnancy outcomes, e.g., major congenital malformations and neonatal death attributed to the use of SSRIs in pregnancy, which may have affected prescribing patterns at the time [18, 19, 20]. Since 2019, antidepressant use among Danish pregnant women has steadily increased. The latest data from 2023 show levels nearly equivalent to the 2011 peak. Sertraline accounts for most of this increase, while the use of other antidepressants, with the exception of escitalopram, mirtazapine and amitriptyline, has declined. Escitalopram, mirtazapine and amitriptyline have seen some increases of use but remain low compared with sertraline. Specifically for escitalopram, the increased use may be attributed to its formal approval for treating anxiety and the accompanying recommendations in Danish guidelines [21, 22, 23].

Trimester‐specific data show that antidepressant use is highest in the first trimester, lowest in the second and rises again in the third, see Figure S4. This pattern has remained stable from 2001 to 2023. However, interpretation requires caution, as prescriptions may reflect preconception use, delayed initiation due to teratogenic concerns or late pregnancy relapse. In addition, prescription redemption does not guarantee use, and early first‐trimester redemptions may reflect preconception decisions or unintentional exposure.

In a European context, antidepressant use during pregnancy has ranged from 1.6% (Europe‐wide average, 2019) to 13.4% in the United Kingdom (2018), with intermediate rates reported in Spain (5.8%), Iceland (7.0%) and Norway (1.8%) [24, 25, 26, 27]. Some of these estimates are based on prescriptions issued rather than redeemed, which may overestimate actual use. Most are based on averaged data across multiple years and do not allow time‐trend analyses, as made possible by the eSundhed data used in this study.

4.1. Methodological Considerations in Drug Utilization Studies in Pregnancy

Estimating drug consumption based on redeemed prescriptions per 1000 pregnancies may lead to overestimation, as some patients obtain medication without using it. This limitation applies to both the pregnant study population and the comparison group. One approach to counteract this is restricting analysis to patients redeeming multiple prescriptions within a year, reducing bias from non‐adherent users. However, this poses new challenges, as it excludes patients who discontinue treatment after a single prescription due to recovery or side effects. Studies from Denmark and Norway have shown that discontinuation of antidepressants during pregnancy is common, often occurring in the first trimester, with frequent re‐initiation among women with severe symptoms. Initiation during pregnancy is less frequent and typically occurs early, with most new users discontinuing before delivery [28, 29].

4.2. Strengths and Limitations

The validity of data from the National Prescription Registry and the Danish Medical Birth Register, provided by the Danish Health Authority, is generally considered high. A 1996 study assessing its accuracy by cross‐referencing redeemed prescriptions of strong analgesics with the National Board of Health's surveillance system found an error margin of just 0.05% [30]. Because these medications are strictly prescription‐based, this suggests a high degree of reliability in the registry's data.

A key limitation of this study is that the database does not account for polypharmacy among antidepressants, potentially counting the same patient multiple times if switching medications or using multiple antidepressants simultaneously. Additionally, prescription records reflect redemptions rather than actual use, meaning medication acquired during pregnancy may be taken postpartum, or pre‐pregnancy redemptions may be misclassified. However, this effect is likely minimal, as antidepressants are seldom dispensed in quantities sufficient for an entire pregnancy, ensuring refills are recorded accurately. On the other hand, pregnant women may have redeemed prescriptions prior to the date of conception and subsequently used the medication during pregnancy, potentially during a period when they were unaware of being pregnant. Such drug use is not recorded as exposure during pregnancy. Another limitation is the potential random inclusion of pregnant patients in the comparison group. Because controls were selected from the general female population matched on birth month, birth year and region of residence, some of these individuals may themselves have been pregnant during the same period as the index pregnancy. This misclassification can reduce the contrast between the exposed and comparison groups, particularly if medication use differs systematically between pregnant and non‐pregnant women.

As data from eSundhed are limited to non‐individualized prescription and birth registry information, clinical outcomes such as congenital malformations, gestational or neonatal complications cannot be assessed. However, the marked increase in antidepressant use during pregnancy underlines the importance of maintaining continued surveillance of such outcomes in appropriate data sources.

5. Conclusion

In Denmark, the use of antidepressants in pregnancy is dominated by SSRIs, particularly sertraline. The pattern of use demonstrated a maximum of about 41 per 1000 pregnancies in 2011, followed by a slow decrease to about 27 per 1000 pregnancies in a stable pattern from 2015 to 2020. From 2021, a noticeable increase to 2011 levels was observed. The pattern largely reflects changes to guideline recommendations.

The widespread and increasing use of antidepressants in pregnancy substantiates the significance of continuous post‐marketing safety pharmacovigilance.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Use of TCAs in Denmark among pregnant women from 2000 to 2023 per 1000 pregnancies. ¹TCA, tricyclic antidepressant.

Figure S2: Use of NaSSAs in Denmark among pregnant women from 2000 to 2023 per 1000 pregnancies. ¹NaSSA, noradrenergic and specific serotonergic antidepressant.

Figure S3: Use of other antidepressants in Denmark among pregnant women from 2000 to 2023 per 1000 pregnancies.

Figure S4: Overall use of antidepressants by trimester in Denmark among pregnant women from 2000 to 2023 per 1000 pregnancies.

Video S1: Supporting information.

Andersen A.-J., Enevoldsen K., Enevoldsen A., Gram E., and Damkier P., “Antidepressants Use Among Pregnant Women in Denmark From 2001 to 2023: A Population‐Level Drug Utilization Study,” Basic & Clinical Pharmacology & Toxicology 137, no. 4 (2025): e70105, 10.1111/bcpt.70105.

Data Availability Statement

The data that support the findings of this study are available in eSundhed at https://www.esundhed.dk. These data were derived from the following resources available in the public domain: ‐ Brug af lægemidler under graviditet, https://www.esundhed.dk/Emner/Laegemidler/Brug‐af‐laegemidler‐under‐graviditet.