Abstract
Objective:
Both excessive and inadequate gestational weight gain are associated with adverse health outcomes for the woman and her child. Antidepressant use in pregnancy may affect gestational weight gain, based on evidence in non-pregnant women that some antidepressants may cause weight gain and others weight loss. Previous studies of antidepressant use and gestational weight gain were small with limited ability to account for confounding, including by maternal mental health status and severity. We assessed the association of antidepressant continuation in pregnancy with gestational weight gain among women using antidepressants before pregnancy.
Methods:
Our retrospective cohort study included singleton live births from 2001–2014 within Kaiser Permanente Washington, an integrated healthcare system. Data were obtained from electronic health records and linked Washington State birth records. Among women with ≥1 antidepressant fill within 6 months before pregnancy, women who filled an antidepressant during pregnancy were considered “continuers”; women without a fill were “discontinuers”. We calculated mean differences in gestational weight gain and relative risks (RR) of inadequate and excessive weight gain based on Institute of Medicine guidelines. Using inverse probability of treatment weighting with generalized estimating equations, we addressed differences in maternal characteristics, including mental health conditions.
Results:
Among the 2,887 births, 1,689 (59%) were to women who continued antidepressants in pregnancy and 1,198 (42%) were to discontinuers. After accounting for confounding, continuers had similar weight gain to those who discontinued (mean difference: 1.3 lbs, 95% confidence interval [CI]: −0.1, 2.8 lbs) and similar risks of inadequate and excessive gestational weight gain (RR: 0.95, 95% CI: 0.80, 1.14 and RR: 1.06, 95% CI: 0.98, 1.14, respectively). Findings were comparable for specific antidepressants and trimesters of exposure.
Conclusions:
We did not find evidence that continuation of antidepressants in pregnancy led to differences in gestational weight gain.
Keywords: antidepressants, pregnancy, gestational weight gain, inadequate gestational weight gain, excessive gestational weight gain, adequacy of gestational weight gain
INTRODUCTION
Antidepressants are some of the most commonly used medications during pregnancy, with approximately 7%–8% of pregnant women in the US exposed, or approximately 320,000 exposed pregnancies per year.1–3 Studies in non-pregnant populations suggest that certain antidepressants (e.g. mirtazapine, paroxetine, sertraline) are associated with weight gain and others (e.g. bupropion, fluoxetine) with weight loss.4–12 Antidepressants could plausibly lead to either weight gain or weight loss by changing appetite and metabolism13 through differing action on monoamines, which are known to affect appetite, satiety, and feeding behavior.4,10,14 Antidepressants have high affinity for the histamine receptor H1 and serotonin receptor 5-HT2C, both of which are thought to impact weight gain.10 The impact of antidepressant use during pregnancy on gestational weight gain (GWG) has not been rigorously evaluated. Previous studies of GWG and antidepressant use were limited by small numbers or could not sufficiently account for underlying maternal depression and anxiety and associated characteristics,15–18 potentially leading to confounding by indication.19,20
Excessive GWG is associated with greater risk of gestational diabetes, macrosomia, caesarean section, and birth injuries, as well as greater lifetime risk of obesity and type 2 diabetes for the woman and her child. Inadequate GWG is associated with increased risk of preterm birth, low birthweight, neonatal infection, and lifelong adverse health consequences for the child.21–25 Given the substantial number of antidepressant-exposed pregnancies and the importance of potential adverse consequences, there is a need for additional study of the association of antidepressant use during pregnancy with GWG.
Therefore, our objective was to investigate whether, during the unique time of pregnancy in which women experience dramatic changes in appetite, satiety, and metabolism and typically gain 30 pounds or more, taking an antidepressant affects GWG. Specifically, we assessed the association of continuing, versus discontinuing, antidepressant use during pregnancy with GWG by utilizing electronic health record (EHR) data and linked state birth records. We employed inverse probability of treatment weighted (IPTW) models to account for maternal characteristics, including presence of mental health conditions and indicators of their severity, and assessed associations for individual antidepressants and specific trimesters of exposure.
METHODS
Overview
This was a retrospective cohort study using data from women enrolled in Kaiser Permanente Washington (KPWA, previously Group Health Cooperative), an integrated healthcare delivery system in Washington State, linked with Washington State birth records.26 We required women to have enrollment during the 6 months before pregnancy onset, as well as an antidepressant prescription fill during that time. By building the cohort in this manner, our analyses both reduced potential confounding by indication and mirrored the real-life decision women using antidepressants before pregnancy must make: whether to continue antidepressant use in pregnancy. Given that the vast majority of women who use an antidepressant in pregnancy were using an antidepressant before pregnancy, studying women continuing antidepressants into pregnancy has more clinical relevance than studying women newly starting antidepressants in pregnancy. The KPWA Institutional Review Board and the Washington State Department of Health Institutional Review Board approved this study (both with waivers of consent).
Cohort selection
We included live births from January 1, 2001 through December 31, 2014 to women enrolled in KPWA. KPWA maintains extensive electronic data on patient enrollment, demographics, encounters, diagnoses, procedures, and prescription fills. Over 60% of members receive comprehensive care from KPWA healthcare providers through the Integrated Group Practice, allowing ascertainment of additional data on vital signs and responses to mental health questionnaires. For this study, KPWA births were linked to Washington State birth records where we obtained gestational age information used to determine timing of antidepressant fills related to the start of pregnancy. The unit of analysis for our study was “births” instead of “women”, given that 7% of women had multiple births in our cohort, but we use the terms synonymously.
Exclusion criteria and counts of subjects excluded are presented in Figure 1.1. Briefly, in addition to limiting the population to women with an antidepressant prescription fill in the 6 months before pregnancy onset, we excluded those with their first antidepressant fill in pregnancy after the exposure window, defined below. As 80–90% of women using antidepressants have depression or anxiety, we aimed to restrict our population to women taking antidepressants for those indications.1,3 Women may not come in for care once every 6 months, and those who do not would not have an opportunity to have an anxiety or depression diagnosis code (Supplementary Table 1.1) recorded in this time window. Therefore, we did not require a depression or anxiety diagnostic code. However, we excluded women without a depression or anxiety diagnostic code prior to pregnancy if they had a diagnostic code for another antidepressant indication,1 e.g., migraine or fibromyalgia (Figure 1.1, Supplementary Tables 1.1 and 1.2).
Figure 1.1.
Identification of women with singleton, livebirths and a fill for an antidepressant in the six months prior to pregnancy for analyses of magnitude and adequacy of gestational weight gain.
Analyses examining adequacy of GWG per Institute of Medicine guidelines were restricted to births during 2003–2014, when data on body mass index (BMI) were available (Figure 1.1).
Exposure
Pharmacy dispensing data were used to ascertain information on medication exposure, including antidepressant name, date of prescription fill, and number of days supplied. Women with any antidepressant prescription fill during the exposure window were considered “continuers” and women without a fill were considered “discontinuers” (Figure 1.2).
Figure 1.2.
Study design
The relevant time window of exposure to antidepressants was defined as the first day of pregnancy through four weeks prior to delivery (Figure 1.2). This allowed time for a potential effect of exposure on outcome.
Exposure included selective serotonin reuptake inhibitors (SSRIs; sertraline, fluoxetine, citalopram, paroxetine, escitalopram, fluvoxamine), serotonin-norepinephrine reuptake inhibitors (SNRIs; venlafaxine, desvenlafaxine), and other antidepressants (mirtazapine, bupropion). We did not include tricyclic antidepressants or trazodone as part of our exposure since they are primarily used to treat pain and sleep disorders, respectively.27 If a woman had an antidepressant fill that overlapped into pregnancy but did not have a fill during pregnancy, she was categorized as a discontinuer; we addressed this assumption using a sensitivity analysis. We did not restrict continuers to women who filled the same antidepressant in the period before pregnancy and during the pregnancy exposure window.
Outcomes
We received GWG data from birth records supplied by staff at the Washington State Department of Health, who calculated it by subtracting pre-pregnancy weight from delivery weight per the birth certificate. Pre-pregnancy weight on the birth certificate is self-reported by the woman. Delivery weight on the birth certificate is typically ascertained from the last weight measured at a prenatal visit, which the delivery providers can access through the EHR. Pre-pregnancy BMI was calculated from self-reported measures on the birth certificate ([weight in kg]/[height in m]2). Adequacy of GWG was defined according to the Institute of Medicine guidelines,28 which define recommended weight gain ranges (inadequate, adequate, excessive) specific to pre-pregnancy BMI category.
We investigated the validity of birth certificate values for GWG, as compared with data from the KPWA EHR, among women with relevant data from both sources. GWG was defined as delivery weight from the EHR (the last measured weight at a prenatal visit in the three weeks prior to delivery) minus pre-pregnancy weight from the EHR (weight recorded in the EHR during the 6 months before pregnancy, or if missing, weight in the first trimester). For the 1012 women with GWG gain available from both the birth certificate and EHR, these measures had a correlation of 0.81, with a mean difference subtracting the EHR value from the birth certificate value of −0.6 lbs (95% CI −1.1, 0.03 lbs). Findings were similar when separately assessing continuers (correlation 0.82, mean difference −0.5 lbs, 95% CI −1.2, 0.3 lbs) and discontinuers (correlation 0.79, mean difference −0.7 lbs, 95% CI −1.6, 0.2 lbs).
Covariates
Maternal education, race/ethnicity, parity, and pre-pregnancy weight were ascertained from the birth certificate data because this information was not available in the EHR for members outside of the Integrated Group Practice.
Covariates obtained from KPWA electronic health databases included: (1) prior to pregnancy: Medicaid insurance coverage, membership in the Integrated Group Practice, smoking and other substance use disorder diagnoses, chronic disease diagnoses, mental health condition diagnoses, utilization of psychotherapy, psychiatry, and inpatient psychiatric hospitalization, psychotropic prescription medication fills, prescription fills of medications that may be associated with weight gain, and (2) from the time of delivery: birth year, maternal age, and baby’s sex (Supplementary Tables 1.1, 1.3, and 1.4). We allowed collection of covariate data to go back in time as long as women were continuously enrolled.29
The Patient Health Questionnaire-9 (PHQ-9) is a commonly-used, 9-item, self-reported depression screening questionnaire.30 PHQ-9 records were available for women in the Integrated Group Practice from approximately 2008 onwards. During this time period, KPWA did not practice universal depression screening, but rather the PHQ-9 was typically used when the patient initially presented with symptoms of depression and periodically afterward to monitor treatment response. PHQ-9 scores were available only for some women in our cohort, so we described the PHQ-9 scores during the two years before pregnancy but did not use them in statistical models.
Statistical analyses
We described baseline characteristics separately for continuers and discontinuers in our cohort using sample means and proportions. Associations between the exposure and outcomes were estimated using regression models with inverse probability of treatment weighting (IPTW). Covariates in the exposure models were chosen a priori based on our knowledge and review of current evidence. IPTW models were utilized in order to account for more potential confounders than possible within a multivariable adjusted model, described in detail in Methods Appendix 1.1.31 In brief, IPTW uses propensity scores to weight observations in the model by likelihood of exposure in order to improve balance in baseline covariates between exposed and unexposed. Quality of the balancing is assessed by standardized mean differences (SMDs)25 and is presented in Supplementary Table 1.6. For overall analyses for each of our outcomes, we weighted the IPTW models by all variables in Variable Set 4 (Supplementary Table 1.5), including variables describing demographic factors, chronic health conditions, pre-pregnancy weight, parity, substance use disorders, mental health conditions, fills for psychotropic and other medications, and mental health care utilization.
We used generalized estimating equations with an exchangeable correlation matrix for all regression analyses, to account for the correlation among multiple births to the same woman.32 We calculated mean differences and 95% confidence intervals (CIs) via robust standard error estimates for the association of continuing antidepressant use in pregnancy, as compared with discontinuing use, with GWG using an identity link function. For the analyses of inadequate and excessive GWG with antidepressant continuation in pregnancy, we used a Poisson (log) link function to calculate relative risks (RRs) and 95% CIs.
Sub-analyses were undertaken to compare women who continued specific antidepressants with discontinuers (the group of women who discontinued any antidepressant, as in the overall analysis). Additionally, we compared outcomes in women who continued antidepressant use in specific trimesters, regardless of use in other trimesters, with the same group of discontinuers used in the overall analysis. We used Variable Set 3 for these sub-analyses due to smaller numbers of exposed women (Supplementary Table 1.5). For analyses of specific antidepressants with the fewest continuers, paroxetine and venlafaxine, we chose parsimonious models based on which characteristics had SMDs comparing continuers and discontinuers of ≥0.10, and from these variables, we included those associated with a >10% change in the mean difference in GWG after their addition to the crude model (paroxetine: Variable Set 1, venlafaxine: Variable Set 2, Supplementary Table 1.5). For consistency, we used these variable sets for analyses of all outcomes for venlafaxine and paroxetine.
Less than 1% of observations were missing prenatal weight, race/ethnicity, parity, or education. All other characteristics included in primary models did not have missing values. Analyses were restricted to subjects with known information for the variables of interest in each model.
Sensitivity analyses
We ran sensitivity analyses to investigate assumptions in our main analyses. One analysis defined exposure as receiving ≥2 antidepressant fills during pregnancy to address exposure misclassification, which could arise if women who only filled once in pregnancy never actually took the medication. We also conducted an analysis that included gestational age in the model, because studies have found that prenatal antidepressant use is associated with shorter gestational age, even after accounting for confounders.33,34 We did not adjust for gestational age in our primary analysis as it is in the causal pathway, and adjustment could lead to collider-stratification bias.35 Another sensitivity analysis restricted to women with a pre-pregnancy anxiety or depression diagnosis. To address potential lingering effects of medication in the body and the assumption that women whose last fill overlapped pregnancy discontinued before pregnancy, we conducted a sensitivity analysis that re-categorized them as continuers. In case the extent of antidepressant use before pregnancy affected GWG, we conducted sensitivity analyses stratifying by and, separately, adjusting for, proportion of days covered with antidepressant fills in the 6 months prior to pregnancy. We ran analyses separately for continuers with less than 50% of days covered with antidepressant fills and continuers with greater than or equal to 50% of days covered, to address a potential duration-response relationship. We assessed the association separately for male and female infants. For the analysis of magnitude of GWG, we adjusted for pre-pregnancy BMI instead of pre-pregnancy weight to address residual confounding, restricting to years when BMI was available. Sensitivity analyses used IPTW models weighted for characteristics in Variable Set 3 (Supplementary Table 1.5), except for the analysis in which we instead adjusted for those characteristics to assess whether the associations were similar to those found in IPTW analyses.
We used R version 3.4.2 to conduct statistical analyses (R Core Team [https://www.r-project.org]).
RESULTS
The source population included 57,743 births. After applying inclusion and exclusion criteria, the eligible population for analyses of mean GWG consisted of 2,887 births (Figure 1.1), including 1,689 (59%) to women who continued antidepressants in pregnancy and 1,198 (41%) to women who discontinued. The number eligible for the analyses of adequacy of GWG was 2,635 births (continuers: 1,537, discontinuers: 1,098).
Characteristics of continuers and discontinuers
Pre-pregnancy characteristics between continuers and discontinuers of antidepressant medication during pregnancy were similar, with a few exceptions: continuers were somewhat older, less likely to be nulliparous, more educated, and more likely to be non-Hispanic white than discontinuers (Table 1.1). Additionally, women who continued antidepressants were more likely to have had a psychiatry visit or to have filled an antidepressant prescription more than one year before pregnancy, as compared with discontinuers (Supplementary Table 1.6). Before pregnancy, the average length of enrollment in KPWHRI, during which we allowed collection of covariate data, was 4.2 years (standard deviation [SD]: 4.1 years) for continuers and 4.5 years (SD: 4.3 years) for discontinuers. SMDs in baseline characteristics were not meaningfully different between women who continued and discontinued antidepressants in pregnancy after weighting (Supplementary Figures 1.1 and 1.2).
Table 1.1.
Selected baseline characteristics of women with births eligible for the analysis of mean gestational weight gain.
| Covariates | No antidepressant fill in pregnancy (n=1,198) | Antidepressant fill in pregnancy (n=1,689) |
|---|---|---|
| Maternal age at delivery, mean (SD) | 29.7 (5.8) | 31.4 (5.4) |
| Parity, n (%) | ||
| Zero | 534 (45) | 623 (37) |
| One | 417 (35) | 598 (36) |
| Two or more | 241 (20) | 462 (28) |
| Maternal race/ethnicity, n (%) | ||
| Hispanic | 83 (7) | 98 (6) |
| Non-Hispanic Asian | 33 (3) | 26 (2) |
| Non-Hispanic black | 47 (4) | 36 (2) |
| Non-Hispanic Native American | 15 (1) | 25 (2) |
| Non-Hispanic Native Hawaiian or Other Pacific Islander | 20 (2) | 21 (1) |
| Non-Hispanic white | 993 (83) | 1476 (88) |
| Maternal education, n (%) | ||
| High school diploma or less | 255 (21) | 305 (18) |
| Some college | 471 (40) | 617 (37) |
| Bachelor’s degree or more | 464 (39) | 755 (45) |
| Medicaid, n (%) | 47 (4) | 54 (3) |
| Pre-pregnancy weight in lbs, mean (SD) | 166.2 (44.5) | 169.1 (43.5) |
| During the two years prior to pregnancy, n (%) | ||
| Alcohol use disorder | 52 (4) | 74 (4) |
| Tobacco use disorder | 159 (13) | 235 (14) |
| Other drug use disorder | 48 (4) | 60 (4) |
| Depression disorder | 868 (73) | 1235 (73) |
| Anxiety disorder | 463 (39) | 719 (43) |
| Obsessive compulsive disorder | 22 (2) | 56 (3) |
| Post-traumatic stress disorder | 40 (3) | 55 (3) |
| Bipolar disorder | 56 (5) | 76 (5) |
| During the year prior to pregnancy, n (%) | ||
| Antipsychotic fill | 21 (2) | 38 (2) |
| Benzodiazepine fill | 266 (22) | 411 (24) |
| Mood stabilizer fill | 59 (5) | 89 (5) |
| Medication associated with weight gain fill | 174 (15) | 259 (15) |
| Psychotherapy visit | 225 (19) | 285 (17) |
| Psychiatry visit | 173 (14) | 271 (16) |
| Inpatient psychiatric hospitalization | 9 (1) | 19 (1) |
SD: Standard deviation
Less than 1% of observations were missing prenatal weight, race/ethnicity, parity, or education. All other characteristics included in this table did not have missing values.
This table contains the study population for the analysis of magnitude of gestational weight gain. The study population for the analyses of adequacy of gestational weight gain excluded some births presented in this table (Figure 1.1).
This table includes a subset of all covariates that were used to create the inverse probability of treatment weights and included in the overall regression models (Variable Set 4 in Supplementary Table 1.5). The full version of this table (with all included characteristics) is Supplementary Table 1.6.
Diagnostic codes for conditions are listed in Supplementary Table 1.1. Specific medications in medication categories are listed in Supplementary Table 1.3. Definitions of mental health care utilization are described in Supplementary Table 1.4.
Among continuers, we had a record of a PHQ-9 screening prior to pregnancy for 361 (61% of women who could have a record, based on when the PHQ-9 was used and recorded in the EHR in KPWA). Among discontinuers, we had 284 records (67% of women who could have a record). For continuers, the mean pre-pregnancy score was 9.5 (SD: 6.6) and for discontinuers the mean score was 10.3 (SD: 6.7), where a score of 5–9 represents mild depression and a score of 10–14 represents moderate depression.
Counts of continuers filling specific antidepressants in pregnancy are included in Supplementary Table 1.7, and number of antidepressant fills and days covered before and during pregnancy are included in Supplementary Table 1.8. The antidepressants most commonly used in pregnancy were sertraline, fluoxetine, and citalopram. The median proportion of days covered by antidepressant fills in the pregnancy exposure window was 60% (interquartile range: 26–91%).
Gestational weight gain
Continuers had an unadjusted mean GWG of 30.5 lbs (SD: 16.8 lbs), similar to discontinuers (mean: 30.3, SD: 16.1 lbs; Table 1.2). After IPTW, women who continued antidepressants had a similar amount of weight gain as discontinuers (mean difference: 1.3 lbs, 95% CI: −0.1, 2.8 lbs; Table 1.2). There was no association with mean GWG for women who continued any SSRI as compared with discontinuers (mean difference: 1.0 lbs, 95% CI −0.4, 2.4 lbs; Table 1.2). There was no association of mean GWG with individual antidepressants or trimesters of exposure (Table 1.2).
Table 1.2.
Association of antidepressant continuation in pregnancy with mean maternal gestational weight gain.
| No. | Weight gain (lbs) Mean (SD) | Crude mean difference in lbs (95% CI) | Inverse probability of treatment weighted mean difference (95% CI) | |
|---|---|---|---|---|
| No antidepressant fill in pregnancy | 1,198 | 30.3 (16.1) | 0.00 (Reference) | 0.00 (Reference) |
| Any antidepressant fill in pregnancy | 1,689 | 30.5 (16.8) | 0.4 (−0.9, 1.7) | 1.3 (−0.1, 2.8)a |
| SSRIs | 1,457 | 30.5 (16.7) | 0.4 (−0.9, 1.8) | 1.0 (−0.4, 2.4)b |
| Citalopram | 362 | 31.7 (15.3) | 1.4 (−0.5, 3.2) | 1.2 (−1.0, 3.4)b |
| Fluoxetine | 482 | 30.3 (17.8) | 0.7 (−1.7, 3.1) | 1.2 (−1.9, 4.4)b |
| Paroxetine | 150 | 27.8 (15.0) | −2.5 (−5.1, 0.0) | −2.5 (−5.9, 1.0)c |
| Sertraline | 514 | 30.0 (17.1) | 0.0 (−1.8, 1.9) | 0.7 (−1.3, 2.8)b |
| Bupropion | 232 | 29.1 (16.9) | −0.9 (−3.4, 1.6) | 0.9 (−1.7, 3.5)b |
| Venlafaxine | 103 | 29.0 (18.9) | −1.3 (−5.2, 2.6) | −1.6 (−5.9, 2.7)d |
| Any first trimester fill | 1,461 | 30.8 (16.8) | 0.7 (−0.6, 2.0) | 1.1 (−0.2, 2.5)b |
| Any second trimester fill | 989 | 29.7 (15.6) | −0.6 (−1.9, 0.8) | 0.1 (−1.3, 1.5)b |
| Any third trimester fill | 887 | 29.8 (15.9) | −0.3 (−1.7, 1.1) | 0.4 (−1.1, 1.8)b |
| Fill in all trimesters | 626 | 30.2 (15.8) | 0.1 (−1.4, 1.7) | 0.6 (−1.0, 2.3)b |
lbs: pounds; SD: standard deviation; CI: confidence interval; SSRIs: selective serotonin reuptake inhibitors
Numbers for specific medications may add to >100% due to some women having used multiple antidepressants (89% of deliveries were to women who used one medication, 10% used 2, and 1% used 3 or 4), and categories are overlapping.
Any antidepressant fill in pregnancy also included mirtazapine, desvenlafaxine, escitalopram, and fluvoxamine, and SSRI antidepressants also include escitalopram and fluvoxamine, but there were too few births exposed to these medications to present separately here.
Weighted by characteristics included in Variable Set 4 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 3 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 1 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 2 in Supplementary Table 1.5.
In analyses of adequacy of GWG, 17% had inadequate weight gain (continuers: 17%, discontinuers: 18%), and 53% had excessive weight gain (continuers: 54%, discontinuers: 51%; Table 1.3). There was no evidence of an association between antidepressant continuation and risk of inadequate weight gain after weighting by confounders for any antidepressant (RR 0.95, 95% CI 0.80, 1.14) or the SSRI class of antidepressants, specifically (RR 0.94, 95% CI 0.78, 1.13). Findings from other trimester- and medication-specific analyses of inadequate GWG were similar (Table 1.3). We also did not observe evidence of altered risk of excessive GWG associated with continuation for any antidepressant (RR 1.06, 95% CI 0.98, 1.14) or SSRI’s, specifically (RR 1.06, 95% CI 0.98, 1.15; Table 1.4). Results were similar for individual antidepressants and exposure during specific trimesters (Table 1.4).
Table 1.3.
Association of antidepressant continuation in pregnancy with inadequate maternal gestational weight gain.
| No. | No. (%) with inadequate weight gain | Crude RR (95% CI) |
Inverse probability of treatment weighted RR (95% CI) | |
|---|---|---|---|---|
| No antidepressant fill in pregnancy | 1098 | 197 (18) | 1.00 (Reference) | 1.00 (Reference) |
| Any antidepressant fill in pregnancy | 1537 | 258 (17) | 0.94 (0.80, 1.11) | 0.95 (0.80, 1.14)a |
| SSRIs | 1327 | 219 (17) | 0.92 (0.77, 1.09) | 0.94 (0.78, 1.13)b |
| Citalopram | 344 | 52 (15) | 0.85 (0.65, 1.13) | 0.95 (0.68, 1.31)b |
| Fluoxetine | 450 | 75 (17) | 0.92 (0.72, 1.17) | 1.00 (0.77, 1.29)b |
| Paroxetine | 107 | 21 (20) | 1.09 (0.71, 1.65) | 1.11 (0.62, 2.01)c |
| Sertraline | 470 | 84 (18) | 0.99 (0.79, 1.25) | 0.99 (0.76, 1.29)b |
| Bupropion | 213 | 42 (20) | 1.10 (0.82, 1.49) | 0.96 (0.66, 1.38)b |
| Venlafaxine | 90 | 19 (21) | 1.19 (0.79, 1.82) | 1.22 (0.74, 1.99)d |
| Any first trimester fill | 1316 | 216 (16) | 0.91 (0.77, 1.09) | 0.93 (0.78, 1.11)b |
| Any second trimester fill | 905 | 170 (19) | 1.05 (0.87, 1.26) | 1.07 (0.87, 1.30)b |
| Any third trimester fill | 802 | 136 (17) | 0.94 (0.77, 1.15) | 0.94 (0.76, 1.16)b |
| Fill in all trimesters | 556 | 94 (17) | 0.93 (0.74, 1.16) | 0.93 (0.72, 1.19)b |
RR: relative risk; CI: confidence interval; SSRIs: selective serotonin reuptake inhibitors
Numbers for specific medications may add to >100% due to some women having used multiple antidepressants (89% of deliveries were to women who used one medication, 10% used 2, and 1% used 3 or 4), and categories are overlapping.
Any antidepressant fill in pregnancy also included mirtazapine, desvenlafaxine, escitalopram, and fluvoxamine, and SSRI antidepressants also include escitalopram and fluvoxamine, but there were too few births exposed to these medications to present separately here.
Weighted by characteristics included in Variable Set 4 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 3 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 1 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 2 in Supplementary Table 1.5.
Table 1.4.
Association of antidepressant continuation in pregnancy with excessive maternal gestational weight gain.
| No. | No. (%) with excessive weight gain | Crude RR (95% CI) | Inverse probability of treatment weighted RR (95% CI) | |
|---|---|---|---|---|
| No antidepressant fill in pregnancy | 1098 | 564 (51) | 1.00 (Reference) | 1.00 (Reference) |
| Any antidepressant fill in pregnancy | 1537 | 828 (54) | 1.05 (0.98, 1.13) | 1.06 (0.98, 1.14)a |
| SSRIs | 1327 | 716 (54) | 1.05 (0.98, 1.14) | 1.06 (0.98, 1.15)b |
| Citalopram | 344 | 193 (56) | 1.09 (0.97, 1.22) | 1.07 (0.94, 1.21)b |
| Fluoxetine | 450 | 245 (54) | 1.07 (0.96, 1.18) | 1.04 (0.93, 1.17)b |
| Paroxetine | 107 | 51 (48) | 0.92 (0.74, 1.14) | 0.92 (0.70, 1.21)c |
| Sertraline | 470 | 245 (52) | 1.01 (0.91, 1.12) | 1.04 (0.93, 1.17)b |
| Bupropion | 213 | 110 (52) | 1.00 (0.86, 1.16) | 1.09 (0.94, 1.27)b |
| Venlafaxine | 90 | 49 (54) | 1.06 (0.87, 1.29) | 1.09 (0.85, 1.39)d |
| Any first trimester fill | 1316 | 713 (54) | 1.05 (0.98, 1.14) | 1.06 (0.98, 1.15)b |
| Any second trimester fill | 905 | 475 (52) | 1.02 (0.94, 1.11) | 1.02 (0.93, 1.12)b |
| Any third trimester fill | 802 | 422 (52) | 1.02 (0.94, 1.12) | 1.03 (0.94, 1.13)b |
| Fill in all trimesters | 556 | 299 (54) | 1.04 (0.95, 1.15) | 1.05 (0.94, 1.17)b |
RR: relative risk; CI: confidence interval; SSRIs: selective serotonin reuptake inhibitors
Numbers for specific medications may add to >100% due to some women having used multiple antidepressants (89% of deliveries were to women who used one medication, 10% used 2, and 1% used three or four), and categories are overlapping.
Any antidepressant fill in pregnancy also included mirtazapine, desvenlafaxine, escitalopram, and fluvoxamine, and SSRI antidepressants also include escitalopram and fluvoxamine, but there were too few births exposed to these medications to present separately here.
Weighted by characteristics included in Variable Set 4 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 3 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 1 in Supplementary Table 1.5.
Weighted by characteristics included in Variable Set 2 in Supplementary Table 1.5.
Sensitivity analyses
We did not find evidence of an association between continuing antidepressants in pregnancy and magnitude of GWG in any sensitivity analyses, including the analysis requiring ≥2 fills as the criterion for exposure (mean difference 0.7 lbs, 95% CI −0.7, 2.2 lbs) and the analysis that added gestational age in the model (mean difference 1.1 lbs, 95% CI −0.2, 2.5 lbs; Supplementary Table 1.9). Continuers and discontinuers also had equivalent risk of inadequate GWG after requiring ≥2 fills for exposure (RR 0.99, 95% 0.82, 1.20), including gestational age in the model (RR 0.93, 95% CI 0.79, 1.11), and testing the sensitivity of our findings to other modifications (Supplementary Table 1.10). Risk of excessive GWG was also similar between continuers and discontinuers in sensitivity analyses requiring ≥2 fills (RR 1.05, 95% CI 0.96, 1.14), accounting for gestational age (RR 1.07, 95% CI 0.99, 1.16), and in other sensitivity analyses (Supplementary Table 1.11).
DISCUSSION
In this retrospective cohort study using EHR data, we did not observe differences in GWG comparing women who continued antidepressant medications in pregnancy with those who discontinued these medications, after accounting for many potential confounders. We also did not find any associations when we analyzed individual antidepressants separately, or when we assessed adequacy of GWG according to Institute of Medicine categories. Results were robust to many sensitivity analyses.
Literature on the association between antidepressant use in pregnancy and GWG is sparse. A prospective cohort study by Wisner et al. found that GWG in women with depression was on average 10.9 lbs greater in women continuously using SSRI’s (n=23) compared with women with depression who did not use antidepressants (n=3), but this mean difference was not adjusted for confounders.15 This was notably different than our finding of only 1.0 lb greater weight gain in women using SSRI’s as compared with those not using antidepressants in pregnancy. After adjustment for age and race, however, Wisner et al. did not find evidence of a statistically significant difference in mean GWG nor risk of inadequate or excessive GWG associated with SSRI use, but they were limited by small sample size. Suri et al. reported essentially equivalent mean GWG in women with depression who took antidepressants (n=49) and did not take antidepressants (n=22), with 0.5 lbs greater mean weight gain in exposed, without adjustment for other confounders.16 Other studies have presented mean GWG stratified by prenatal antidepressant exposure, but because the unexposed group was not restricted to women with diagnosed depression or recent antidepressant use, it was not possible to disentangle the effects of the underlying mental health condition from the medication.17,18,36–38
Strengths of our study include the use of a large, well-defined study population, availability of medication fill data, extensive information about mental health utilization as an indicator of mental health condition severity, mental health screening test results for a portion of the population, and the ability to conduct sensitivity analyses, which yielded results consistent with the primary findings.
Our study was subject to limitations. Certain antidepressant medications were used by few women (e.g. mirtazapine, which is associated with weight gain in non-pregnant populations); this limited power to detect moderate associations for these medications. By studying a “prevalent user” cohort, we likely excluded any women whose weight was severely affected by an antidepressant, as they may have already discontinued that medication shortly after initiation due to this side effect. This could have attenuated our risk estimates. Women with more severe depression or anxiety may have been more likely to continue antidepressants than those with less severe depression or anxiety, which may have led to residual confounding. If women with only one antidepressant fill in pregnancy did not actually take the medication, there could be misclassification of exposure, which could weaken our risk estimates. Although the sensitivity analysis assessing validity of GWG from the birth certificate, as compared with the EHR, indicated strong correlation and very small mean differences in weight measures, there was likely some misclassification of our outcomes. Lastly, the majority of our study population was commercially insured and non-Hispanic white, and we only included live births, which may limit our findings’ generalizability.
We did not find evidence that continuation of antidepressants during pregnancy has a deleterious effect on GWG. Additional risks and benefits should be considered when deciding whether to continue antidepressants in pregnancy, including infant outcomes and the woman’s mental health, but our study does not raise concerns regarding risk of unhealthy GWG.
Supplementary Material
KEY POINTS.
Antidepressant use is associated with weight change in non-pregnant populations.
Prior evidence on whether antidepressant use in pregnancy affects gestational weight gain is sparse.
We accounted for confounding by characteristics such as mental health conditions and their severity.
We found no association between pregnancy antidepressant continuation and gestational weight gain.
ACKNOWLEDGMENTS
We thank Sharon Fuller and Eric Baldwin for pulling the information from the Kaiser Permanente Washington database to build our cohort, as well as programming our initial variables. We thank Gregory Simon and Christine Stewart for their advice about how to best extract and operationalize mental health indicators. We also acknowledge the Mental Health Research Network for their resources that we used to define mental health diagnoses, procedures, and medications. We thank James Fraser for his help with Institutional Review Board applications. We thank Susan Shortreed for her advice regarding the statistical analysis. We thank Sarah Randall for her help with manuscript formatting and submission.
FUNDING
Paige Wartko worked on this study while receiving a graduate student stipend from the Stroum Graduate Fellowship through the University of Washington Diabetes Research Center and the National Institute of Child Health and Human Development’s Reproductive, Perinatal, and Pediatric Epidemiology Training Grant (#T32 HD052462). Funds for data extraction came from a Group Health Foundation Partnership for Innovation grant.
Outside the work of this study, Paige Wartko receives funding from a research contract awarded to Kaiser Permanente Washington Health Research Institute from a consortium of pharmaceutical companies (Allergan, BioDelivery Sciences, Collegium, Daiichi Sankyo, Depomed, Egalet, Endo, Janssen, Mallinckrodt, Pernix, Pfizer, Purdue, and West-Ward) to conduct Food and Drug Administration-mandated studies on opioids. Also outside this study, Sascha Dublin is a co-investigator for a proposal that was submitted to GlaxoSmithKline by Harvard Department of Population Medicine, which is likely to be funded.
REFERENCES
- 1.Huybrechts KF, Palmsten K, Mogun H, et al. National trends in antidepressant medication treatment among publicly insured pregnant women. General hospital psychiatry 2013;35:265–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mitchell AA, Gilboa SM, Werler MM, et al. Medication use during pregnancy, with particular focus on prescription drugs: 1976–2008. American journal of obstetrics and gynecology 2011;205:51 e1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hayes RM, Wu P, Shelton RC, et al. Maternal antidepressant use and adverse outcomes: a cohort study of 228,876 pregnancies. American journal of obstetrics and gynecology 2012;207:49 e1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Serretti A, Mandelli L. Antidepressants and body weight: a comprehensive review and meta-analysis. The Journal of clinical psychiatry 2010;71:1259–72. [DOI] [PubMed] [Google Scholar]
- 5.Papakostas GI. The efficacy, tolerability, and safety of contemporary antidepressants. The Journal of clinical psychiatry 2010;71 Suppl E1:e03. [DOI] [PubMed] [Google Scholar]
- 6.Arterburn D, Sofer T, Boudreau DM, et al. Long-Term Weight Change after Initiating Second-Generation Antidepressants. J Clin Med 2016;5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wellman PJ, Davies BT, Morien A, McMahon L. Modulation of feeding by hypothalamic paraventricular nucleus alpha 1- and alpha 2-adrenergic receptors. Life sciences 1993;53:669–79. [DOI] [PubMed] [Google Scholar]
- 8.Terry P, Gilbert DB, Cooper SJ. Dopamine receptor subtype agonists and feeding behavior. Obes Res 1995;3 Suppl 4:515S–23S. [DOI] [PubMed] [Google Scholar]
- 9.Stahl SM, Pradko JF, Haight BR, Modell JG, Rockett CB, Learned-Coughlin S. A Review of the Neuropharmacology of Bupropion, a Dual Norepinephrine and Dopamine Reuptake Inhibitor. Prim Care Companion J Clin Psychiatry 2004;6:159–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Meister B Neurotransmitters in key neurons of the hypothalamus that regulate feeding behavior and body weight. Physiology & behavior 2007;92:263–71. [DOI] [PubMed] [Google Scholar]
- 11.Rezvanian H, Hashemipour M, Kelishadi R, Tavakoli N, Poursafa P. A randomized, triple masked, placebo-controlled clinical trial for controlling childhood obesity. World J Pediatr 2010;6:317–22. [DOI] [PubMed] [Google Scholar]
- 12.Ye Z, Chen L, Yang Z, et al. Metabolic effects of fluoxetine in adults with type 2 diabetes mellitus: a meta-analysis of randomized placebo-controlled trials. PloS one 2011;6:e21551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zimmermann U, Kraus T, Himmerich H, Schuld A, Pollmacher T. Epidemiology, implications and mechanisms underlying drug-induced weight gain in psychiatric patients. Journal of psychiatric research 2003;37:193–220. [DOI] [PubMed] [Google Scholar]
- 14.Papakostas GI. Tolerability of modern antidepressants. The Journal of clinical psychiatry 2008;69 Suppl E1:8–13. [PubMed] [Google Scholar]
- 15.Wisner KL, Sit DK, Hanusa BH, et al. Major depression and antidepressant treatment: impact on pregnancy and neonatal outcomes. The American journal of psychiatry 2009;166:557–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Suri R, Altshuler L, Hellemann G, Burt VK, Aquino A, Mintz J. Effects of antenatal depression and antidepressant treatment on gestational age at birth and risk of preterm birth. The American journal of psychiatry 2007;164:1206–13. [DOI] [PubMed] [Google Scholar]
- 17.Nulman I, Rovet J, Stewart DE, et al. Child development following exposure to tricyclic antidepressants or fluoxetine throughout fetal life: a prospective, controlled study. The American journal of psychiatry 2002;159:1889–95. [DOI] [PubMed] [Google Scholar]
- 18.Nulman I, Koren G, Rovet J, et al. Neurodevelopment of children following prenatal exposure to venlafaxine, selective serotonin reuptake inhibitors, or untreated maternal depression. The American journal of psychiatry 2012;169:1165–74. [DOI] [PubMed] [Google Scholar]
- 19.Hartley E, McPhie S, Skouteris H, Fuller-Tyszkiewicz M, Hill B. Psychosocial risk factors for excessive gestational weight gain: A systematic review. Women Birth 2015;28:e99–e109. [DOI] [PubMed] [Google Scholar]
- 20.Marcus SM. Depression during pregnancy: rates, risks and consequences--Motherisk Update 2008. Can J Clin Pharmacol 2009;16:e15–22. [PubMed] [Google Scholar]
- 21.Spaight C, Gross J, Horsch A, Puder JJ. Gestational Diabetes Mellitus. Endocr Dev 2016;31:163–78. [DOI] [PubMed] [Google Scholar]
- 22.Metzger BE. Summary and recommendations of the Third International Workshop-Conference on Gestational Diabetes Mellitus. Diabetes 1991;40 Suppl 2:197–201. [DOI] [PubMed] [Google Scholar]
- 23.Gilmore LA, Klempel-Donchenko M, Redman LM. Pregnancy as a window to future health: Excessive gestational weight gain and obesity. Seminars in perinatology 2015;39:296–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nehring I, Schmoll S, Beyerlein A, Hauner H, von Kries R. Gestational weight gain and long-term postpartum weight retention: a meta-analysis. The American journal of clinical nutrition 2011;94:1225–31. [DOI] [PubMed] [Google Scholar]
- 25.Rogozinska E, Zamora J, Marlin N, et al. Gestational weight gain outside the Institute of Medicine recommendations and adverse pregnancy outcomes: analysis using individual participant data from randomised trials. BMC pregnancy and childbirth 2019;19:322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Baldwin E, Johnson K, Berthoud H, Dublin S. Linking mothers and infants within electronic health records: a comparison of deterministic and probabilistic algorithms. Pharmacoepidemiology and drug safety 2015;24:45–51. [DOI] [PubMed] [Google Scholar]
- 27.Patten SB, Esposito E, Carter B. Reasons for antidepressant prescriptions in Canada. Pharmacoepidemiology and drug safety 2007;16:746–52. [DOI] [PubMed] [Google Scholar]
- 28.Institute of Medicine. Weight Gain During Pregnancy: Reexamining the Guidelines 2009. [PubMed] [Google Scholar]
- 29.Gilbertson DT, Bradbury BD, Wetmore JB, et al. Controlling confounding of treatment effects in administrative data in the presence of time-varying baseline confounders. Pharmacoepidemiology and drug safety 2016;25:269–77. [DOI] [PubMed] [Google Scholar]
- 30.Manea L, Gilbody S, McMillan D. A diagnostic meta-analysis of the Patient Health Questionnaire-9 (PHQ-9) algorithm scoring method as a screen for depression. General hospital psychiatry 2015;37:67–75. [DOI] [PubMed] [Google Scholar]
- 31.Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivariate Behav Res 2011;46:399–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Burton P, Gurrin L, Sly P. Extending the simple linear regression model to account for correlated responses: an introduction to generalized estimating equations and multi-level mixed modelling. Stat Med 1998;17:1261–91. [DOI] [PubMed] [Google Scholar]
- 33.Mitchell J, Goodman J. Comparative effects of antidepressant medications and untreated major depression on pregnancy outcomes: a systematic review. Archives of women’s mental health 2018. [DOI] [PubMed] [Google Scholar]
- 34.Sujan AC, Rickert ME, Oberg AS, et al. Associations of Maternal Antidepressant Use During the First Trimester of Pregnancy With Preterm Birth, Small for Gestational Age, Autism Spectrum Disorder, and Attention-Deficit/Hyperactivity Disorder in Offspring. Jama 2017;317:1553–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wilcox AJ, Weinberg CR, Basso O. On the pitfalls of adjusting for gestational age at birth. American journal of epidemiology 2011;174:1062–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Nulman I, Rovet J, Stewart DE, et al. Neurodevelopment of children exposed in utero to antidepressant drugs. The New England journal of medicine 1997;336:258–62. [DOI] [PubMed] [Google Scholar]
- 37.Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. Birth outcomes in pregnant women taking fluoxetine. The New England journal of medicine 1996;335:1010–5. [DOI] [PubMed] [Google Scholar]
- 38.Pastuszak A, Schick-Boschetto B, Zuber C, et al. Pregnancy outcome following first-trimester exposure to fluoxetine (Prozac). Jama 1993;269:2246–8. [PubMed] [Google Scholar]
- 39.Messerli FH, Bell DS, Fonseca V, et al. Body weight changes with beta-blocker use: results from GEMINI. Am J Med 2007;120:610–5. [DOI] [PubMed] [Google Scholar]
- 40.Domecq JP, Prutsky G, Leppin A, et al. Clinical review: Drugs commonly associated with weight change: a systematic review and meta-analysis. The Journal of clinical endocrinology and metabolism 2015;100:363–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Brookhart MA, Schneeweiss S, Rothman KJ, Glynn RJ, Avorn J, Sturmer T. Variable selection for propensity score models. American journal of epidemiology 2006;163:1149–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.