Risks and Benefits of Pre-delivery Taper In Pregnant Women Taking Antidepressants

Abstract

Background:

Serotonin reuptake inhibitors are commonly used for treatment of mental health problems in pregnancy but may cause neonatal adaptation syndrome. It is unknown whether reduction or discontinuation of medication prior to delivery may mitigate this effect.

Methods:

We present a case series of 38 women who either tapered their medication prior to delivery or maintained or increased their dose.

Results:

Greater reductions in maternal antidepressant dose just prior to delivery were associated with fewer admissions to the neonatal intensive care unit (NICU) for infants. There was a slightly greater increase in depressive symptoms across delivery for women who tapered, which was not statistically significant.

Conclusions:

NICU admissions may be less frequent among neonates whose mothers tapered their medication prior to delivery. Large prospective randomized trials are needed to further study this practice.

Significant Outcomes:

Pre-delivery taper of SRI medication in pregnant women is associated with reduced rates of NICU admission in neonates. An increase in maternal postpartum depressive symptoms after pre-delivery taper is possible, but did not reach significance in this cohort.

Limitations:

This is a small clinical case series without randomization or blinding. Residual confounding cannot be excluded.

Background:

Serotonin reuptake inhibitors (SRIs) are the most commonly used class of medication for depression, anxiety, and obsessive-compulsive disorder^Kern. They are considered acceptable for use in pregnancy^Yonkers. However, exposure to antidepressants in late pregnancy is associated with poor neonatal adaptation (PNA)¹, including irritability, abnormal crying, tremor, jitteriness, lethargy, respiratory distress, hypoglycemia, poor muscle tone, and, rarely, convulsions. PNA is estimated to occur in about 30% of infants antenatally exposed to SRIs². Rarely, in approximately one per thousand deliveries, a more serious syndrome called persistent pulmonary hypertension of the newborn (PPHN) may result^3,4. A recent study calculated the number needed to harm to be 417 to 5,000 women treated with antidepressants in late pregnancy⁵. The risks of both PPHN and PNA are dose-dependent and pertain to exposure in late, rather than early, pregnancy.

Guidelines for adult patients recommend that antidepressants be tapered gradually in order to avoid discontinuation syndrome⁶. Gradual reduction might also be preferable to abrupt discontinuation in neonates. Given that risk of PNA is related to SRI exposures later in pregnancy, it seems possible that reduction in dose or temporary discontinuation prior to delivery could ease the burden of discontinuation on the neonate⁷. However, this practice is controversial, with some practitioners warning that the risk for maternal relapse might be too great. The postpartum period is a time of uniquely high psychiatric risk¹⁰. Furthermore, the mother’s responsibility to care for her vulnerable infant raises the stakes for interventions that may interfere with optimal maternal functioning. At present, no published data on the practice of pre-delivery taper exist.

Two register-based studies^8,9 have attempted to compare risks for PNA based on the timing of exposure. A Canadian study⁸ identified significant differences in risks for both neonatal respiratory distress and convulsions based on exposure in the final two weeks of pregnancy (n=1605) compared to gestational exposure but not in the last two weeks of pregnancy (n=2122). These analyses were adjusted for many variables including total antidepressant exposure, number of episodes and psychiatric visits. Admission to the neonatal intensive care unit (NICU) was not examined. The authors then examined a smaller subset (n=239 pregnancies) using propensity score matching to pair each exposed case with an unexposed case that would be similar across many maternal variables. In this subset, no difference in respiratory distress was found. The propensity-matched subset contained no cases of convulsions in either the exposure or the comparison group.

The second⁹ study was a retrospective cohort study in Indiana of 2741 pregnant women prescribed 5 different antidepressant(s) before or during pregnancy. The authors reported reduced rates of respiratory distress and NICU admission when third trimester exposure was avoided, reaching statistical significance for escitalopram and buproprion, but not for fluoxetine, sertraline or citalopram.¹⁰.

In these register-based studies, exposure was calculated based on prescriptions either written or dispensed. However, monthly or quarterly prescriptions are unlikely to be altered by prescribers who anticipate only a short period of abstinence prior to delivery. Furthermore, reduced dosing (as opposed to complete discontinuation) is not detectable by this method. Hence, there is a high likelihood of misclassification when attempting to study medication changes over a short timeframe using registry data, which would tend to minimize or eliminate any true intergroup difference.

Given the difficulty of accurately tracking medication use and correctly assigning study groups when using registry data, these studies do not fully address the question of whether temporary SRI tapers in late pregnancy can reduce risk for PNA in newborns¹¹. Additionally, the putative counter-risk of maternal relapse has not been investigated at all.

There is currently no published evidence that directly evaluates the efficacy of intentional pre-delivery taper for the purpose of mitigating NAS. While register-based studies have been critical to discovering the association of PNA and PPHN with antenatal SRI exposure in the first place, the identification of efficacious interventions is likely to be achieved only by targeted, prospective, clinical studies. In the absence of any such study so far, we describe a clinical case series of women who were offered pre-delivery dose reduction of SRIs. We present both mental health and neonatal outcomes of women who chose pre-delivery dose reduction or cessation and compare them to those who maintained their usual doses through delivery.

Method:

Retrospective case series comparing women who reduced their SRI dose in the final weeks prior to delivery with women who maintained or increased dose prior to delivery

Study population:

The study protocol was approved by the Institutional Review Board of the Icahn School of Medicine at Mount Sinai (Study ID 22–00139). The requirement for individual patient consent was waived for this chart review of existing medical records.

Participants (n=38) were patients of TKR treated with an SRI during pregnancy, who were offered pre-delivery taper and either accepted or declined. All women offered this option were deemed by the treating clinician to be sufficiently psychiatrically stable to tolerate a short period at reduced dose or off their medication. Women who initiated SRI treatment after 30 weeks’ gestation were excluded. 18 women reduced or discontinued their medication between 30 weeks’ gestation and delivery, and 20 women maintained or increased their dose. None of the women included reported any prior history of suicide attempts or psychiatric inpatient hospitalizations, and none were actively abusing substances in pregnancy. Two women in the taper group had past histories of substance use disorders. Two women in the no-taper group had been treated in partial hospitalization programs in the past.

Study outcomes:

Neonatal health outcomes included respiratory distress, hypoglycemia, NICU admission, and Apgar score.

Maternal psychiatric outcome was measured as the difference in Edinburgh Postnatal Depression Scale (EPDS) score from the minimum recorded within 10 weeks antepartum to the maximum recorded within 10 weeks postpartum.

Study procedures:

Tapers were initiated between eight and four weeks prior to the expected delivery date, depending on the mother’s dose and whether she wished to achieve discontinuation or merely dose reduction. Tapers were subject to clinical tolerability, such that women who reported resurgent depression or anxiety at lowered doses did not have their doses further reduced beyond the minimum tolerable. All women were advised to return to their pre-taper doses immediately after delivery.

Women were divided into two groups: those whose dose was maintained or increased between 30 weeks’ gestation and delivery (no-taper group), and those whose medication was reduced or discontinued between 30 weeks’ gestation and delivery (taper group).

After delivery, medical charts of mothers and neonates were reviewed individually for evidence of each of the outcomes listed above. In cases where neonatal hospital records were not available due to delivery at an outside institution (6 in no-taper and 7 in taper group), information on the specified neonatal outcomes was obtained by the treating psychiatrist via patient interview at the first postpartum visit.

Statistical analysis:

Neonatal outcomes: Chi square tests were used to examine whether frequencies of respiratory distress, hypoglycemia, and NICU admission differed between neonates in the taper and no-taper groups. T tests were used to determine whether group means differed for gestational age at delivery, birth weight, and 1-minute Apgar score. Finally, logistic regression was used to further examine the NICU admission result, employing stratification of dosing groups and including potential confounders in the model.

Maternal outcomes: T tests were also used to assess differences in maternal symptoms of anxiety and depression as measured by the EPDS. Mean EPDS in the 10 weeks prior to delivery, mean EPDS in the 10 weeks post delivery, and mean difference from the lowest recorded EPDS antepartum to the highest recorded EPDS postpartum were analyzed.

Results:

Neonatal outcomes:

NICU admission rate in the taper group was 6%, and in the no-taper group 42% (chi-square statistic 6.708, p < 0.01, Figure 1a). In the taper group, 3 infants of 18 had respiratory distress and 4 hypoglycemia, while in the no-taper group, 7 out of 20 had respiratory distress and 5 hypoglycemia (both nonsignificant; Supplementary Table 1). Maternal age, gestational age at delivery, birth weight, and 1-minute Apgar did not differ between groups (Table 1).

Figure 1. Neonatal and maternal outcomes by pre-delivery taper status.

Figure 1a. NICU admissions among infant, by taper status of mothers. Difference significant by chi square test (chi-square statistic = 7.60, p-value = 0.006).

Figure 1b. Edinburgh Postpartum Depression Scale scores before and after delivery. A maximum pre/post difference was calculated by taking the lowest recorded EPDS in the ten weeks prior to delivery, and subtracting it from the highest recorded EPDS in the ten weeks following delivery. No significant differences between groups by T test.

Table 1:

Maternal and neonatal characteristics (mean, SD).

	Maternal age (y)	Weeks exposed to SRI	GA at delivery	Birth weight	1min Apgar	Minimum antenatal EPDS	Maximum postnatal EPDS	EPDS difference
Taper	34.4 (3.3)	34.4 (7.4)	38.8 (1.5)	3319 (430)	8.55 (0.69)	5.33 (3.6)	9.12 (5.7)	4.00 (4.6)
No taper	33.8 (4.0)	32.0 (12.0)	38.5 (2.0)	3043 (439)	8.50 (0.65)	6.30 (3.3)	8.00 (4.4)	1.58 (2.9)

No significant between-group differences on any measure by T test

The significant difference in NICU admissions was further explored using logistic regression to evaluate the differences in outcomes among women who increased, maintained, reduced, or discontinued their medication. Women who increased their medication between 30 weeks’ gestation and delivery had a 50% rate of infant admission to the NICU. Women who maintained their dose had a 42.9% NICU admission rate, and those who reduced their dose an 8.3% NICU admission rate. No NICU admissions were recorded among infants of women who discontinued their medication prior to delivery (Figure 2). This difference remained significant when duration of medication exposure and indicators of maternal psychiatric severity were included in the model (Supplementary Table 2). For women who discontinued, the mean period of time between discontinuation and delivery was 10.3 days (SD 8.7 days, range 3–26 days).

Figure 2.

NICU admissions among women who increased, maintained, reduced, or discontinued their SRI prior to delivery. Difference significant by logistic regression (p-value = 0.017).

Maternal psychiatric symptoms:

Among all women in the study, EPDS scores were recorded a mean of 3.8 times (range 1–6 administrations) between 10 weeks antepartum and 10 weeks postpartum. EPDS scores did not differ between groups (Figure 1b). There was a somewhat larger increase in EPDS in the taper (+4.0 EPDS points) than in the no-taper group (+1.58 EPDS points), but this difference did not reach statistical significance (p=0.06). (Table 1)

Conclusions:

This clinical case series compared 18 women who tapered their antidepressant dose prior to delivery to 20 women who did not. Groups were clinically comparable. Results suggest that predelivery taper may reduce risk for NICU admission in the infant. There was a trend toward increased postpartum EPDS scores in the mothers who tapered, but this difference did not reach statistical significance in this cohort. The current series suggests that reductions in dose might mitigate risk for PNA, and that even a short period of medication abstinence prior to delivery might provide benefit to the neonate. Our results should be interpreted with caution, given that residual confounding cannot be excluded, This practice should therefore be investigated in larger cohort studies and/or a randomized controlled trial, as the current case series was limited by small sample size and absence of randomization

Future Directions.

This is the first report to explore the question of whether pregnant women who are treated with SRIs should consider a temporary reduction in dose prior to delivery. It is of critical importance to conduct larger, targeted clinical studies to quantify risks and benefits for mother and child.

Data Availability:

The data that support the findings of this study are available on request from the corresponding author, TKR. The data are not publicly available due to their containing information that could compromise the privacy of research participants.