Association of treatment for prenatal depression with cardiovascular disease within 24 months after delivery

Abstract

Introduction

While the strong association of prenatal depression with an increased risk of new cardiovascular disease (CVD) within 24 months postpartum has been demonstrated in two distinct cohorts, it is unclear if treatment of prenatal depression may mitigate this risk. We aimed to estimate the cumulative risk of new CVD and new chronic hypertension in the first 24 months postpartum among pregnant individuals with prenatal depression treated with antidepressant medications or psychotherapy compared with those with untreated prenatal depression. We hypothesized that treatment for prenatal depression may be associated with a lower risk of CVD postpartum among those with prenatal depression.

Methods

Our longitudinal population‐based study included pregnant individuals with deliveries in the Maine Health Data Organization's All Payer Claims Data during 2007–2020. We excluded those individuals with any diagnosis before pregnancy of the cardiovascular conditions being examined postpartum, multifetal gestations, or no continuous health insurance during pregnancy. Prenatal depression, psychotherapy, and antidepressant medications were identified with ICD9/10 diagnosis codes, CPT codes, and pharmacy dispensing data, respectively. Primary outcomes were a dichotomous CVD composite (heart failure, ischemic heart disease, cerebrovascular disease, arrhythmia/cardiac arrest, cardiomyopathy) and chronic hypertension within 24 months postpartum. Adjusted Cox models were used to estimate hazard ratios.

Results

Among the 17,100 pregnant individuals with depression included in our analysis (13.5% of 126,333 total eligible pregnancies), 50.0% (n = 8556) had treatment with antidepressants and/or psychotherapy visits. Among the 16,953 pregnant individuals with depression included in the composite CVD outcome analysis, treatment with antidepressant medication use and/or psychotherapy was associated with a 25% lower risk of CVD (adjusted hazard ratio [aHR],, 0.75; 95% confidence interval [CI], 0.57–0.98). However, among the 15,240 pregnant individuals with depression included in the new chronic hypertension outcome analysis, no association was observed (aHR, 1.05; 95% CI, 0.84–1.32).

Conclusion

The cumulative risk of new CVD within 24 months postpartum was significantly lower in patients with antidepressant and/or psychotherapy use during pregnancy compared to those with untreated prenatal depression. This suggests that treatment of prenatal depression may mitigate CVD risk; however, given the limitations of observational research, further study is warranted to elucidate the pathophysiologic mechanisms that underlie this association.

1. INTRODUCTION

Cardiovascular disease (CVD) is the leading cause of death among women in the United States [1]. Because many cardiovascular deaths are preventable, it is crucial to determine which pregnancy conditions are associated with future CVD in order to inform evidence‐based screening and primary prevention strategies [2]. Prenatal depression is associated with an increased risk of new CVD within 24 months postpartum, even in the absence of co‐occurring hypertensive disorders of pregnancy [3, 4]. The increased risk is highest and consistent across prospective cohort studies for several subtypes of CVD, including new chronic hypertension (32%–44% increased risk), ischemic heart disease (44%–83% increased risk), and heart failure (38%–40% increased risk) [3, 4].

This association between prenatal depression and new CVD postpartum aligns with the abundant literature in the general population demonstrating that depression is a strong independent risk factor for CVD and sudden cardiac death [5, 6, 7, 8, 9]. There is even a dose–response relationship, with more severe depression conferring an even higher risk of cardiac mortality [6]. This accumulating evidence has spurred efforts to treat depression in order to prevent lethal cardiac events in the general population [7]. For example, preoperative depression is an independent risk factor for mortality following cardiac valve surgery, and research is ongoing to determine if treatment of depression improves cardiac outcomes after valve replacement surgery [10]. In addition, research has noted that pharmacologic antidepressants confer a protective effect against myocardial infarctions in nonpregnant adults with depression; however, pregnant and lactating individuals were excluded from this prospective study [11].

The study of prenatal depression and CVD among pregnant and postpartum patients has been mostly neglected due to the pervasive historical neglect of women's health in medical research [12]. In addition, hesitancy in recent years to robustly assess the safety of psychotropic medications in pregnancy has contributed to inconsistencies in clinical guidance and undertreatment of psychiatric conditions in the perinatal period [13, 14]. Despite the support of the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal‐Fetal Medicine (SMFM) on pharmacologic treatment of prenatal depression, many patients are still wary of antidepressant use during pregnancy [14, 15]. Further, it is unknown if treatment of prenatal depression may mitigate postpartum CVD risk [12].

Due to the increasing incidence of both prenatal depression and cardiac severe maternal morbidity (SMM), and the significant knowledge gaps on CVD in women, it is critical to include pregnant patients in studies examining both the association of treatment of prenatal depression with CVD risk [16, 17]. Therefore, we sought to determine whether treatment of prenatal depression with antidepressant medications or psychotherapy may lower the risk of new CVD within 24 months postpartum.

2. MATERIALS AND METHODS

We conducted a longitudinal population‐based study of pregnant individuals with prenatal depression and singleton live birth or stillbirth deliveries during 2007–2020 to compare those with psychotherapy, antidepressant use, or both during pregnancy to those without these treatments. Treatment for prenatal depression was considered as a composite exposure (only psychotherapy, only antidepressant use, or both) and also as two separate exposures (psychotherapy and antidepressant use). Data from the Maine Health Data Organization's (MHDO) All Payer Claims Data (APCD) were used. The MHDO APCD included paid medical and pharmacy claims, as well as insurance coverage by month, for over 90% of Maine residents [18]. The data include claims paid by public payers and private insurers for inpatient and outpatient encounters, private physician visits, clinic visits, and prescription claims, as required by state statute and rules [18].

Code lists for prenatal depression, psychotherapy, antidepressant medications, and other psychotropic medications were sourced from public resources available from the Mental Health Research Network [19]. Diagnoses, medication fills, and psychotherapy were examined between 6 and 34 weeks of gestation (“during pregnancy”) to avoid the bias that can be introduced when assessing exposure during the late third trimester, the length of which varies across pregnancies [20]. Prenatal depression was identified using International Classification of Diseases, Ninth Revision (ICD‐9) and Tenth Revision (ICD‐10) clinical modification diagnosis codes and excluded codes for depressive disorders in full remission (Appendix 1) [3, 19, 21, 22]. Psychotherapy was identified with Current Procedural Terminology (CPT) procedure codes (Appendix 2). All antidepressant classes were included. Other psychotropic medications included benzodiazepines, other hypnotics, and other anti‐anxiety medications, anticonvulsants, antipsychotics, lithium, stimulants, and other attention deficit disorder medications. Other psychotropic medication (i.e., non‐antidepressant medications) fill was considered a potential confounder given its association with antidepressant use and it being a possible risk factor for CVD. Patients were considered to have been treated with psychotherapy during pregnancy if they had at least one visit during pregnancy. They were considered to have been treated with medication if they filled the medication of interest at least once during pregnancy, according to pharmacy claims data.

Primary outcomes were a postpartum CVD composite outcome (heart failure, ischemic heart disease, cerebrovascular disease, arrhythmia/cardiac arrest, and cardiomyopathy) and new chronic hypertension, consistent with previous studies using this cohort [3, 23, 24, 25]. The postpartum CVD composite outcome has been used in prior studies examining SMM due to cardiac causes [3, 24, 25, 26]. Ischemic heart disease is an inadequate supply of blood to the myocardium due to obstruction of the coronary arteries, which often leads to a myocardial infarction [27]. Several of these CVD outcomes, including stroke (cerebrovascular disease), myocardial infarction, arrhythmia, and peripartum cardiomyopathy, are recognized as indicators of SMM by ACOG and the US Centers for Disease Control and Prevention [28, 29]. New chronic hypertension was defined from 43 days postpartum until 24 months postpartum because hypertension up to 42 days after delivery is considered a hypertensive disorder of pregnancy [30, 31]. In contrast, the CVD composite outcome was defined from delivery date until 24 months postpartum, consistent with the definition utilized in previously published papers [3, 25, 26].

Exclusion criteria included records with implausible gestational age for stillbirth or live birth (<20 weeks), non‐Maine residents, multifetal gestations, those with implausible time to next pregnancy (<60 days), deliveries before 2007 or after 2020 (with incomplete data), and those without continuous health insurance during pregnancy and the first 2 months postpartum [3]. In addition, we excluded persons who did not have a depression diagnosis during pregnancy. Exclusion criteria also included any diagnosis before pregnancy of the cardiovascular conditions being examined postpartum, for example, those with pre‐existing chronic hypertension were excluded from the analysis with chronic hypertension as the primary outcome. For the analysis with the CVD composite as the primary outcome, those with chronic hypertension were included. In this model, the estimated hazard ratio was adjusted for pre‐existing chronic hypertension because of its potential as a confounder.

Cox proportional hazard models were used to estimate hazard ratios, adjusting for potential confounders, which we defined as factors that could affect both treatment for prenatal depression and the outcomes being examined. These confounders included maternal age at time of delivery, prepregnancy depression, pre‐existing chronic hypertension (for CVD composite outcome models), prepregnancy diabetes, obesity, smoking, nulliparity, pregnancy number in dataset, year of delivery, Medicaid coverage during pregnancy, county‐level measures (the number of general practice physicians per capita; the percentage uninsured adults, adults who smoke, and adults with obesity), ZIP‐code level measures (median percentage of residents living below federal poverty level, of non‐White race, and adults with less than college educational attainment), hypertensive disorders of pregnancy, gestational diabetes, and other psychotropic medication use during pregnancy. County‐ and ZIP‐code level measures were based on the residential ZIP code in the eligibility record for the patient at the month of delivery. We used Cox models to handle right‐censored data, as not all pregnancies had complete follow‐up. Censoring events were 24 months postpartum, loss of health insurance, or start of the next pregnancy, whichever was earliest.

As a proxy measure of depression severity, models also controlled for any depression‐related emergency department (ED) visit or hospitalization during pregnancy using a previously published method applied to this cohort [32]. We identified all claims with evidence of ED visits (using procedure codes, revenue codes, and place of service codes) or inpatient hospitalizations (using bill type codes). We then sorted claims by unique de‐identified person identifiers and date of service, which allowed us to cluster encounters if they occurred on the same or adjacent days. If the last date of the prior encounter was at least 2 days before the first date of the next encounter, we considered it a distinct episode. Finally, among these encounters we identified, those with a depression diagnosis recorded as the primary diagnosis on the claim were considered depression‐related.

For models of antidepressant use as the exposure, the group with neither treatment (no antidepressant use and no psychotherapy use) was used as the referent group. Similarly, for models of psychotherapy use as the exposure, the group with neither (no antidepressant use and no psychotherapy use) was used as the referent group.

As a sensitivity analysis, we restricted our analysis to pregnancies with at least some insurance coverage in the 3 months before the start of pregnancy, so that we could better assess pre‐existing hypertension and CVD, which were excluded from their respective models. This improved our detection of these pre‐existing conditions, but reduced the sample size, limiting generalizability and statistical power.

SAS version 9.5 was used to perform the statistical analysis.

3. RESULTS

Among the 17,100 pregnancies with a depression diagnosis included in our analysis (13.6% of 126,333 total eligible pregnancies) (Figure 1), 50.0% (n = 8,556) had evidence of treatment with antidepressants or psychotherapy visits: 34.1% (n = 5826) used antidepressants and 24.0% (n = 4109) had psychotherapy, which were not mutually exclusive (Appendix 3). Baseline characteristics appeared similar between those with versus without treatment for prenatal depression, with the exceptions that older maternal age, private insurance, prepregnancy depression, any depression‐related emergency department/inpatient visit during pregnancy, and other psychotropic medication use during pregnancy appeared to be more common in the group with antidepressant or psychotherapy treatment (Table 1).

FIGURE 1.

Analytic sample identification. ^aPre‐existing chronic hypertension was defined as hypertension from 1 year prior to pregnancy start to 20 weeks of gestation.

TABLE 1.

Baseline characteristics, deliveries with depression during pregnancy ^* by treatment status in Maine 2007–2020.

			Either psychotherapy or antidepressant use
	TOTAL		No 50.0%		Yes 50.0%
	N	%	N	%	N	%
Total	17,100	100	8544	100	8556	100
Maternal age at delivery
Missing	9	0.1	6	0.1	3	0.0
15–19 years	1461	8.5	883	10.3	578	6.8
20–24 years	4704	27.5	2821	33.0	1883	22.0
25–29 years	4864	28.4	2477	29.0	2387	27.9
30–34 years	3644	21.3	1530	17.9	2114	24.7
35+ years	2418	14.1	827	9.7	1591	18.6
Cesarean section	5420	31.7	2646	31.0	2774	32.4
Gestational age at delivery
At least 37 weeks	15,285	89.4	7692	90.0	7593	88.7
20 to <37 weeks	1815	10.6	852	10.0	963	11.3
Delivery number in dataset
1	10,392	60.8	5132	60.1	5260	61.5
2 or more	6708	39.2	3412	39.9	3296	38.5
Health insurance coverage at time of delivery
Medicaid	13,047	76.3	7311	85.6	5736	67.0
Private	3938	23.0	1182	13.8	2756	32.2
Medicare	115	0.7	51	0.6	64	0.8
Last month of continuous insurance postpartum
<6 months	2471	14.5	1326	15.5	1145	13.4
6–11 months	1195	7.0	564	6.6	631	7.4
12–23 months	2068	12.1	1062	12.4	1006	11.8
24+ months	11,366	66.5	5592	65.5	5774	67.5
Any insurance coverage in 3 months before pregnancy	15,292	89.4	7466	87.4	7826	91.5
Hypertensive disorders of pregnancy	2532	14.8	1187	13.9	1345	15.7
Gestational diabetes	1759	10.3	805	9.4	954	11.2
Psychotropic medication use during pregnancy **	4072	23.8	1510	17.7	2562	29.9
Only psychotherapy during pregnancy *	2730	16.0	0	0.0	2730	31.9
Only antidepressant use during pregnancy *	4447	26.0	0	0.0	4447	52.0
Both psychotherapy and antidepressant use during pregnancy *	1379	8.1	0	0.0	1379	16.1
SSRI use during pregnancy *	4587	26.8	0	0.0	4587	53.6
SNRI use during pregnancy *	523	3.1	0	0.0	523	6.1
TCA use during pregnancy *	159	0.9	0	0.0	159	1.9
Bupropion use during pregnancy *	902	5.3	0	0.0	902	10.5
Other antidepressant use during pregnancy *	393	2.3	0	0.0	393	4.6
Any depression‐related emergency department/inpatient visits during pregnancy *	720	4.2	332	3.9	388	4.5
Prepregnancy cardiovascular disease ***	130	0.8	61	0.7	69	0.8
Prepregnancy hypertension	1845	10.8	866	10.1	979	11.4
Prepregnancy depression	8999	52.6	3763	44.0	5236	61.2

Note: Data source: Maine Health Data Organization's All Payer Claims Data.

Abbreviations: SNRI, serotonin‐norepinephrine reuptake inhibitors; SSRI, Selective serotonin reuptake inhibitors; TCA, tricyclic antidepressants.

^*‘During pregnancy’ indicates 6 to 34 weeks' gestation.

^**Psychotropic medication use includes anticonvulsants, first‐generation antipsychotics, second‐generation antipsychotics, benzodiazepines, other hypnotics, injectable antipsychotics, lithium, stimulants, other medications used for attention deficit disorder, other anti‐anxiety medications.

^***Heart failure, heart disease, stroke, arrythmia, or cardiomyopathy.

Of the 17,100 pregnancies with a diagnosis of depression during pregnancy, 130 were excluded due to prepregnancy CVD and 1845 were excluded due to pre‐existing chronic hypertension (Figure 1). The cumulative risk of the composite CVD outcome (heart failure, ischemic heart disease, cerebrovascular disease, arrhythmia/cardiac arrest, and cardiomyopathy) within 24 months postpartum was 1.9% and the cumulative risk of new chronic hypertension within 24 months postpartum was 3.3% (Table 2); for each, there were more diagnoses occurring in the first few months postpartum than later on (Figure S1). In the final analytic models, 16,953 were included in the adjusted model for CVD and 15,420 were included in the adjusted model for chronic hypertension due to some deliveries missing covariate information (Figure 1).

TABLE 2.

Cumulative risk of new diagnosis in the first 24 months postpartum for cardiovascular disease and chronic hypertension for deliveries with depression, Maine 2007–2020.

	Prenatal depression Total	Number of outcome events	Cumulative risk by 24 months a among prenatal depression
Cardiovascular disease	16,970	250	1.9
Chronic Hypertension (43+ days after delivery) b	15,255	370	3.3

Note: Data source: Maine Health Data Organization's All Payer Claims Database.

^a Cumulative risk by 24 months. Censoring events were loss of health insurance coverage or start of the next pregnancy, whichever was earlier. Each model excluded records with any diagnosis before pregnancy of the cardiovascular conditions being examined postpartum, and records with implausible gestational age (<20 weeks), non‐Maine residence, multifetal gestation, those with implausible time to next pregnancy (< 60 days), and those without health insurance during pregnancy and the first 2 months postpartum and/or in the month of delivery.

^b Hypertension diagnoses in the first 42 days postpartum were not considered outcome events, as these were included in the definition of hypertensive disorders of pregnancy.

Among the 16,953 pregnancies included in the CVD outcome analysis, antidepressant medication use and/or psychotherapy visits were associated with a 25% lower risk of the composite CVD outcome (adjusted hazard ratio [aHR], 0.75; 95% confidence interval [CI], 0.57–0.98) (Figure 2A). Psychotherapy during pregnancy was associated with a 26% lower risk of the composite CVD outcome (aHR, 0.74; 95% CI, 0.52–1.06), though results were not statistically significant (Figure 3A). In addition, antidepressant use during pregnancy was associated with a 21% lower risk of the composite CVD outcome (aHR, 0.79; 95% CI, 0.60–1.05); results were not statistically significant (Figure 4A). For each exposure, adjusted hazard ratio (HRs) were lower than unadjusted HRs (Table S1).

FIGURE 2.

Weighted adjusted^a cumulative hazard curves and 95% confidence interval bands for new diagnosis of (A) cardiovascular disease and (B) chronic hypertension 43+ days after delivery in the first 24 months postpartum among people with and without antidepressant or psychotherapy use during pregnancy, deliveries with prenatal depression in Maine 2007–2020 (n = 17,100). ^a Adjusted for maternal age at time of delivery, prepregnancy depression, prepregnancy hypertension, prepregnancy diabetes, obesity, smoking, nulliparity, pregnancy number in dataset, year of delivery, Medicaid coverage during pregnancy, county‐level measures, ZIP‐code level measures, hypertensive disorders of pregnancy, gestational diabetes, psychotropic medication use during pregnancy, and any depression‐related emergency department visit or hospitalization during pregnancy.

FIGURE 3.

Weighted adjusted^a cumulative hazard curves and 95% confidence interval bands for new diagnosis of (A) cardiovascular disease and (B) chronic hypertension 43+ days after delivery in the first 24 months postpartum among people with and without psychotherapy during pregnancy, deliveries with prenatal depression in Maine 2007–2020 (n = 12653). ^aAdjusted for maternal age at time of delivery, prepregnancy depression, prepregnancy hypertension, prepregnancy diabetes, obesity, smoking, nulliparity, pregnancy number in dataset, year of delivery, Medicaid coverage during pregnancy, county‐level measures, ZIP‐code level measures, hypertensive disorders of pregnancy, gestational diabetes, psychotropic medication use during pregnancy, and any depression‐related emergency department visit or hospitalization during pregnancy.

FIGURE 4.

Weighted adjusted^a cumulative hazard curves and 95% confidence interval bands for new diagnosis of (A) cardiovascular disease and (B) chronic hypertension 43+ days after delivery in the first 24 months postpartum among people with and without antidepressant use during pregnancy for depression during pregnancy, deliveries with prenatal depression in Maine 2007–2020 (n = 14370). ^aAdjusted for maternal age at time of delivery, prepregnancy depression, prepregnancy hypertension, prepregnancy diabetes, obesity, smoking, nulliparity, pregnancy number in dataset, year of delivery, Medicaid coverage during pregnancy, county‐level measures, ZIP‐code level measures, hypertensive disorders of pregnancy, gestational diabetes, psychotropic medication use during pregnancy, and any depression‐related emergency department visit or hospitalization during pregnancy.

Among the 15,240 pregnancies included in the chronic hypertension analysis, antidepressant medications and/or psychotherapy visits were not associated with new chronic hypertension (aHR, 1.05; 95% CI, 0.84–1.32) (Figure 2B). In addition, when analyzed separately, neither psychotherapy nor antidepressant use was associated with new chronic hypertension, (aHR, 0.97; 95% CI, 0.73–1.30) and (aHR, 1.09; 95% CI, 0.86–1.39), respectively (Figures 3B and 4B). Like the CVD outcome analysis, for each exposure, adjusted HRs were lower than unadjusted HRs (Table S2).

In our sensitivity analysis, where the analysis was restricted to pregnancies with at least some insurance coverage in the 3 months before the start of pregnancy (which reduced the sample size), the aHRs were similar to our primary results, but none of the findings were statistically significant (Model 5 in Tables S1 and S2).

4. DISCUSSION

Overall, only 50% of patients with prenatal depression were treated with antidepressant medications or psychotherapy. The cumulative risk of new CVD within 24 months postpartum was 25% lower in patients with prenatal depression with antidepressant and/or psychotherapy use during pregnancy compared to those without evidence of these treatments. Findings were of similar magnitude for psychotherapy and antidepressants, when examined separately; however, they were not statistically significant. In contrast, the cumulative risk of new chronic hypertension within 24 months postpartum was not different in patients with antidepressant and/or psychotherapy use during pregnancy compared to those who did not have these treatments. These findings were similar when psychotherapy and antidepressants were examined separately.

The association of treatment of prenatal depression with a lower risk of CVD postpartum requires further exploration in other populations, especially given the limitations of observational studies; however, our results appear to have biologic plausibility. The prevalence estimates of depression and antidepressant use during pregnancy among those with depression in our study (14% and 34%, respectively) were similar to those found for perinatal mood disorders among commercially insured women (19% and 33%, respectively) during 2008–2020 [33], supporting the validity of our methodology for identifying these factors. In addition, the suspected protective effect of depression treatment during pregnancy on CVD was similar to what has been noted in nonpregnant cohorts. Over two decades ago, researchers began studying whether successful treatment of depression is associated with a reduction in cardiac morbidity and mortality [34, 35]. While treatment of depression in the general population does not appear to significantly reduce cardiac mortality [36], some pharmacology treatments do appear to mitigate adverse cardiac outcomes. For example, the use of selective serotonin reuptake inhibitors (SSRI), the most commonly used antidepressants, confers a protective effect against myocardial infarction [11]. This is possibly attributable to the inhibitory effect SSRIs have on serotonin‐mediated platelet activation, which has been demonstrated in several distinct cohorts [11, 37, 38]. Pregnancy is a unique physiologic state characterized by major compensatory changes in every major organ system, including but not limited to circulatory changes and increased oxidative stress beginning in the first trimester to protect the fetus from rejection by the maternal immune system [39]. Specifically, the increased oxidative stress during pregnancy is due to a systemic inflammatory response that leads to an elevated presence of reactive oxygen species and reactive nitrogen species in the circulatory system [40]. This increased oxidative stress peaks in the second trimester of pregnancy [41]. Therefore, due to the unique physiology of pregnancy with increased inflammation and oxidative stress, known precursors to adverse cardiac events, further research focusing on depression treatment during pregnancy is crucial, especially given that pregnancy was an exclusion criteria for some prospective studies on the association of SSRIs and myocardial infarction [11, 39, 42].

In the general population, psychotherapy has been studied in patients with CVD and co‐morbid depression to determine both the efficacy for treatment of mood disorders and the potential impact on recurrent CVD. While cognitive behavioral therapy has been demonstrated to significantly reduce depression and anxiety in patients with CVD, the possible impact on cardiovascular events is mixed [43]. However, one large randomized control trial in Sweden found that a cognitive behavioral therapy intervention program significantly decreases the risk of recurrent CVD by 41%, specifically acute myocardial infarction [44]. In addition, they found a strong dose–response relationship with a higher attendance rate at therapy sessions conferring an even lower risk of recurrent CVD [44]. In this study, therapy was focused on stress management, suggesting a possible mechanism is decreased emotional reactivity [44]. Emotional reactivity is strongly linked to oxidative stress, which has deleterious effects on the cardiovascular system [45, 46].

Our findings that treatment of prenatal depression with antidepressants or psychotherapy is not associated with new chronic hypertension are difficult to explain, especially given prior literature in nonpregnant patients that demonstrated a possible reduction in hypertension with cognitive behavioral therapy [47]. This highlights the importance of further research to confirm our findings and explore the mechanism. Because of the complex pathophysiology of both depression and hypertension, there are many pathways by which depression may influence cardiovascular regulation. While further research into this null finding is critical, there may be some biologic plausibility to the results. The pathophysiology of chronic hypertension is partly due to changes in the renal–angiotensin–aldosterone system [48]. Therefore, SSRIs’ inhibitory effect on platelets may have less impact on the development of chronic hypertension within 24 months postpartum than on the development of CVD.

While the association of prenatal depression with new CVD postpartum has been previously demonstrated [3, 4], our finding that treatment of prenatal depression is associated with a lower risk of CVD postpartum is novel and highly clinically relevant. Prevalence of both prenatal depression and cardiac SMM among women has been increasing steadily in recent years [2, 49]. Despite the well‐established association of increased maternal morbidity and mortality with prenatal depression, adequate treatment of prenatal depression is severely lacking in current clinical practice [50]. Approximately 20% of pregnant and postpartum women will develop a new mental health condition; yet, these women are often not treated or are undertreated [51, 52]. Therefore, clinicians must prioritize treatment of prenatal depression for a multitude of reasons including the risks of untreated prenatal depression such as preeclampsia and preterm birth and the benefits of treating prenatal depression such as a potential lower risk of cardiac SMM [53, 54].

This stark reality of undertreatment of mental health conditions during pregnancy was evident in our cohort as well, where only 50% of the patients diagnosed with prenatal depression received antidepressant or psychotherapy treatment. This low uptake of treatment is consistent with prior cohorts where treatment for prenatal depression ranged from 47%–60% [21]. This is likely due to pervasive barriers to treatment including lack of access to skilled providers, social stigma, and lack of time for treatment [21, 55, 56]. Interestingly, this frequency is actually higher than a previous report in the general population where only 35.7% of patients with a new diagnosis received treatment [57]. This may be due to improved access to care during pregnancy, which is a well‐known entry point into the healthcare system [58]. Given the increasing prevalence of prenatal depression, policy changes are needed to increase engagement in depression treatment [55, 56]. Implementation of a depression care model by Obstetricians and Gynecologists (OBGYNs) [59] is now even more crucial to women's health in order to optimize not only their mental well‐being but also potentially to reduce the staggering burden of preventable cardiovascular morbidity and mortality among women. The key framework for implementation of a depression care model in OBGYN practice includes on site screening, patient engagement, education, treatment including assistance with referrals, and tracking follow‐up care and symptom severity [60]. In addition, providers must integrate care utilizing consultations with our colleagues in Psychiatry and Psychology [59]. Early integrated depression treatment programs in OBGYN practices are well accepted and extremely effective, likely ushering in a new wave for the future of mental health care for women [59, 61].

Our study had multiple strengths. Primarily, it addresses a novel, highly clinically relevant question regarding the possible cardiovascular risk mitigation of prenatal depression treatment. Given the risk of prenatal depression and the significant opportunity to optimize treatment of mental health during pregnancy, our results suggest an additional reason why all clinicians must prioritize screening and treatment of mental health disorders during pregnancy. In addition to the effects of antidepressants and psychotherapy on perinatal depressive symptoms, we demonstrate that treatment of prenatal depression may also lower the risk of life‐threatening CVD within 24 months postpartum. Furthermore, pregnant patients have long been excluded from research on the association of depression and CVD [12]. Therefore, one additional strength of our study is the focus on pregnant patients, an underrepresented cohort of patients in research, which is critical to promote equitable care for all [62]. Our cohort includes a large patient population with a robust, comprehensive database of inpatient, outpatient, office‐based, and clinic claims.

In addition, our findings are timely as misinformation is circulating regarding the treatment of depression in pregnancy with antidepressants. ACOG has issued a statement confirming the safety of antidepressants in pregnancy [50]; however, many patients are still wary of antidepressant use during pregnancy. Complications of untreated depression, including suicide, are a major cause of preventable maternal mortality [63]. Our study helps to quantify the potential benefits of depression treatment during pregnancy and advocates for continued screening and treatment of mental health disorders during pregnancy.

Our study has several limitations, including those common to observational, retrospective studies utilizing claims data. There is a lack of data on race, ethnicity, and other covariates including measures of depression severity and duration of depressive episode. While an association between prenatal depression treatment and lower risk of CVD postpartum was demonstrated, correlation does not imply causation. Therefore, further prospective studies to confirm our findings are critical. Despite using a standardized prenatal depression code list from peer‐reviewed articles, some instances of prenatal depression may not have been captured and therefore not included in the study cohort, especially among those without continuous health insurance and patients who did not disclose their symptoms of depression to their provider [21, 64]. Despite the large sample size, a composite outcome needed to be utilized given the rarity of CVD outcomes in the study population. Similarly, all antidepressants were included in one exposure. Therefore, the results lacked granularity and further studies with even larger population sizes are crucial to determine which adverse cardiovascular events are most impacted by treatment of prenatal depression. Specifically, larger studies are needed to explore the impact of treatment on CVD endpoints by different antidepressant classes given their distinct mechanisms of action. In addition, because pharmacy dispensing data was utilized, we could not confirm the actual use of antidepressants.

Given the limitations of our study, further research is critical to provide more granularity for both subtypes of CVD and subtypes of antidepressant medications. In addition, we need to explore if treatment of prenatal depression is also associated with a lower future CVD risk for those with adverse pregnancy outcomes, specifically for those with hypertensive disorders of pregnancy [65]. Because top causes of pregnancy‐related death in the United States include cardiovascular conditions and suicide in the setting of mental illness [63, 66, 67], further studies on treatment of prenatal depression and CVD risk, especially exploring a potential causal relationship, can lead to saved lives through improved cardiovascular and mental health. It is also critical to examine if there is a dose–response relationship with both longer duration of antidepressant use and more therapy sessions conferring an even lower risk of recurrent CVD, as was found in nonpregnant cohorts [44]. Lastly, future adequately powered studies are needed to understand the impact of treatment before and during pregnancy on CVD risk versus treatment only before pregnancy, as well as assessing the impact of treatment initiation during pregnancy versus non‐initiation on CVD risk among those with new episodes of depression during pregnancy.

5. CONCLUSION

In conclusion, the cumulative risk of new CVD within 24 months postpartum was significantly lower in patients who treated their depression with antidepressants or psychotherapy during pregnancy. The finding of potential benefit of prenatal depression treatment on the development of new CVD is novel and should be studied in other settings and pregnant populations. Our study is a first step in assessing the association of depression treatment during pregnancy on postpartum CVD risk. Further research is needed to confirm which subtypes of CVD are most impacted by treatment of prenatal depression, to identify the causal pathway, and to understand dose‐response relations.

CONFLICT OF INTEREST STATEMENT

K.P. has research contracts from AbbVie, GSK, Pfizer, and Sanofi. H.S.L. is on the external DSMB for Pfizer COVID‐19 vaccination. All other authors declare no conflicts of interest.

ETHICS STATEMENT

The study protocol was approved by the University of Southern Maine's Institutional Review Board.

Supporting information

Supporting Information

APPENDIX 1. ICD9/10 codes for prenatal depression and cardiovascular diagnoses.  ^a

1.1.

	ICD‐9 Any code starting with:	ICD‐10 Any code starting with:
Prenatal depression	6484, 296.20, 296.21, 296.22, 296.23, 296.24, 296.25, 296.30, 296.31, 296.32, 296.33, 296.34, 296.35, 296.82, 298.0, 300.4, 301.12, 309.0, 309.1, 309.28, 311	F06.31, F06.32, F32.0, F32.1, F32.2, F32.3, F32.4, F32.81, F32.89, F32.9, F33.0, F33.1, F33.2, F33.3, F33.41, F33.8, F33.9, F34.1, F43.21, F43.23, F53.0, O99.34
Heart failure	428.0, 428.1, 428.20, 428.21, 428.22, 428.23, 428.30, 428.31, 428.32,428.33, 428.40, 428.41, 428.42, 428.43, 428.9, 402.01, 402.11, 402.91, 404.01, 404.03, 404.11, 404.13, 404.91, 404.93, 402.01, 402.11, 402.91	I09.81, I11.0, I13.0, I13.2, I50.1, I50.20, I50.21, I50.22, I50.23, I50.30, I50.31, I50.32, I50.33, I50.40, I50.41, I50.42, I50.43, I50.9, I50.810, I50.811, I50.812, I50.813, I50.814, I50.82, I50.83, I50.84, I50.8
Ischemic heart disease	410.x, 411.1, 411.8	I20.0, I20.1, I20.8, I20.9, I21.01, I21.02, I21.09, I21.11, I21.19, I21.21, I21.29, I21.3, I21.4, I21.9, I21.A1, I21.A9, I22.0, I22.1, I22.2, I22.8, I22.9, I23.0, I23.1,I23.2, I23.3, I23.4, I23.5, I23.6, I23.7, I23.8, I24.0, I24.1, I24.8, I24.9, I25.10, I25.110, I25.111 I25.118, I25.119, I25.2, I253, I25.41, I25.42, I25.5, I25.6, I25.700, I25.701, I25.708, I25.709, I25.710, I25.711, I25.718, I25.719, I25.720, I25.721, I25.728, I25.729, I25.730, I25.731, I25.738, I25.739, I25.750, I25.751, I25.758, I25.759, I25.760, I25.761, I25.768, I25.769, I25.790, I25.791, I25.798, I25.799, I25.810, I25.811, I25.812, I25.82, I25.83, I25.84, I25.89, I25.9
Cerebrovascular disease/stroke	430, 431, 432.x, 433.x, 434.x, 435.x, 436, 437.x, 6715.x, 6740.x, 997.02	G45.0, G45.1, G45.2, G45.3, G45.8, G45.9, G46.0, G46.1, G46.2, G46.3, G46.4, G46.5, G46.6, G46.7, G46.8, G97.31, G97.32, I60.00, I60.01, I60.02, I60.10, I60.11, I60.12, I60.20, I60.21, I60.22, I60.30, I60.31, I60.32, I60.4, I60.50, I60.51, I60.52, I60.6, I60.7, I60.8, I60.9, I61.0 I61.1, I61.2, I61.3, I61.4, I61.5, I61.6, I61.8, I61.9, I63.00, I63.011, I63.012, I63.013, I63.019, I63.02, I63.031, I63.032, I63.039, I63.09, I63.10, I63.111, I63.112, I63.119, I63.12, I63.131, I63.132, I63.133, I63.139, I63.19, I63.20, I63.211, I63.212, I63.213, I63.219, I63.22, I63.231, I63.232, I63.233, I63.239, I63.29, I63.30, I63.311, I63.312, I63.313, I63.319, I63.321, I63.322, I63.323, I63.329, I63.331, I63.332, I63.333, I63.339, I63.341, I63.342, I63.343, I63.349, I63.39, I63.40, I63.411, I63.412, I63.413, I63.419, I63.421, I63.422, I63.423, I63.429, I63.431, I63.432, I63.433, I63.439, I63.441, I63.442, I63.443, I63.449, I63.49, I63.50, I63.511, I63.512, I63.513, I63.519, I63.521, I63.522, I63.523, I63.529, I63.531, I63.532, I63.533, I63.539, I63.541, I63.542, I63.543, I63.549, I63.59, I63.6, I63.8, I63.9, I66.01, I66.02, I66.03, I66.09, I66.11, I66.12, I66.13, I66.19, I66.21, I66.22, I66.23, I66.29, I66.3, I66.8, I66.9, I67.841, I67.848, I97.810, I97.811, I97.820, I97.821 I60.2, I63.033, I63.113, I62.9, I65.1, I65.29, I65.09, I6781, I67.2, I67.82, I67.89, I67.4, I67.1, I67.7, I67.5, I67.6, I67.9, I67.848, O87.3, O99.4
Arrhythmia/cardiac arrest	427.31, 427.32, 427.41, 427.42, 427.5	I49.01, I40.92, I46.9, I48.0, I48.1, I48.2, I48.91, I48.4, I48.92
Cardiomyopathy	674.5x, 425.x	O90.3, I42.1, I42.2, I42.4, I42.5, I42.6, I42.7, I42.8, 142.9, 143
Hypertension	401.x, 402.x, 403.x, 404.x, 405.x	I10, I11, I12, I13, I15, I16

^a Codes ending in  .x refer to the stem codes.

APPENDIX 2. Psychotherapy procedure codes.

2.1.

CPT	DESCRIPTION
90804	Outpatient psychotherapy 20–30 min
90805	Individual therapy with med mgt  20–30 min
90806	Outpatient psychotherapy 45–50 min
90807	Individual therapy with med mgt 45–50 min
90808	Outpatient psychotherapy 75–80 min
90809	Individual therapy with med mgt 75–80 min
90810	Interactive therapy 20–30 min
90811	Interactive therapy with med mgt 20–30 min
90812	Interactive therapy 45–50 min
90813	Interactive therapy with med mgt 45–50 min
90814	Interactive therapy 75–80 min
90815	Interactive therapy with med mgt 75–80 min
90832	Psychotherapy, 30 min with patient and/or family member
90833	Psychotherapy, 30 min with patient and/or family member with med mgt
90834	Psychotherapy, 45 min with patient and/or family member
90836	Psychotherapy, 45 min with patient and/or family member with med mgt
90837	Psychotherapy, 60 min with patient and/or family member
90838	Psychotherapy, 60 min with patient and/or family member with med mgt
90839	Psychotherapy for crisis, first 60 minutes
90840	Psychotherapy for crisis 30 min add‐on
90844	Outpatient psychotherapy 45–50 min (replaced by 90806)
90845	Psychoanalysis
90847	Family psychotherapy (conjoint psychotherapy, with patient present)
90849	Multiple‐family group psychotherapy
90853	Group psychotherapy (other than of a multiple‐family group)
90857	Interactive group psychotherapy
90875	Individual psychophysiological therapy incorporating biofeedback training by any modality, 20–30
90876	Individual psychophysiological therapy incorporating biofeedback training by any modality, 45
90880	Hypnotherapy

Abbreviation: med mgt, medication management.

APPENDIX 3. Cross‐tabulation of antidepressant use and psychotherapy during pregnancy for deliveries with depression.

3.1.

		Antidepressant use
		Yes	No	Total
Psychotherapy during pregnancy	Yes	1379 (8.1%)	2730 (16.0%)	4109 (24.0%)
	No	4447 (26.0%)	8544 (50.0%)	12,991 (76.0%)
	Total	5826 (34.1%)	11,274 (65.9%)	17,100 (100.0%)

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.