Keywords
anxiety - depression - psychotropic medication - maternal outcome - neonatal outcome
- neonatal adaptation syndrome
Identifying risk factors for mental disorders that may affect pregnancy is an important
prerequisite in developing interventions to lower adverse maternal and neonatal outcomes.
The estimated prevalence of perinatal anxiety and depression varies in published studies
and is estimated to be between 7 and 20%.[1]
[2] Previous studies have reported associations of perinatal anxiety and depression
with obstetrical and neonatal factors such as substance abuse, socioeconomic factors,
and medical complications of pregnancy, small for gestational age (SGA),[1]
[2]
[3]
[4]
[5] and postpartum depression. However, lacking from the literature are comprehensive
reports from the same cohort simultaneously studying prenatal, intrapartum, and neonatal
outcomes to assess associations of untreated and treated anxiety and depression in
comparison to unaffected pregnancies.
An important consideration is the medical management of these mental disorders in
pregnancy and the potential consequences.[6] Hanley et al estimated that 10.3% of all pregnancies may have some exposure to psychotropic
medications.[6] Currently, psychotropic medications require more extensive risk–benefit analyses
to determine efficacy and safety in pregnancy. For example, evidence from a large
study of publicly insured pregnant women[7] and a recent meta-analysis[8] are consistent with a potential increased risk of persistent pulmonary hypertension
associated with maternal use of selective serotonin reuptake inhibitors (SSRIs) in
late pregnancy. However, the absolute risk was noted to be small, and the risk increase
appears more modest than suggested in previous studies. Other studies have shown a
significantly increased risk of spontaneous abortion, preterm birth, and low birth
weight. All of the observed risks were of a very low magnitude, and the clinical significance
of these results is unknown.[9] However, benefits of antidepressant medication in pregnancy include prevention of
possible relapse of illness, with resultant hospital admission and suicidal ideations.[9] Consequently, studies analyzing the interactive effect of psychotropic treatment
on mental disorders with obstetrical and neonatal outcomes are required for counseling
and clinical management.
Therefore, the objectives of this study were (1) to determine the independent prenatal,
intrapartum, and neonatal correlates of pregnancies affected by depression and anxiety
and (2) to evaluate the differences in outcomes between cases of depression and anxiety
treated and not treated with psychotropic medications when compared with unaffected
pregnancies.
Materials and Methods
This was a retrospective cohort study of all women with singleton pregnancies and
their newborns who delivered from January 1, 2013, through December 31, 2014, in the
Beaumont Health System. An obstetrical outcomes database was developed using a combination
of International Classification of Diseases, ninth edition (ICD-9), coding and direct
querying of the electronic medical record. The extracted data were audited and verified
by manual chart review as indicated. The database included a maximum of 18 months
hospital follow-up data.
Beaumont Health is a not-for-profit health system located in suburban southeastern
Michigan and services a diverse population of approximately 5 million, comprising
Non-Hispanic Whites (68.5%), Blacks (21.6%), Asians (3.6%), and Hispanics (3.9%).
Cases of anxiety and depression were retrieved using the ICD-9 codes 293.84, 300.00,
300.01, 300.02, and 300.09 for anxiety and 296.2, 296.3, and 311 for depression (please
refer to the [Appendix A] for more details). A comprehensive manual chart review was completed on all women
identified as having depression or anxiety to determine treatment with psychotropic
medications during the antepartum period. This was accomplished by review of the admission
history, and physical activity as well as medication prescription during the hospital
admission. Those who had a positive history and were prescribed psychotropic medication
during the hospital admission were classified as “treated,” whereas those who were
not prescribed medication during the hospital admission were classified as “untreated.”
Cases of treated and untreated anxiety, depression, and comorbid anxiety and depression
were compared with unaffected pregnancies regarding various prenatal, intrapartum,
and maternal and neonatal factors and outcomes that were selected based on literature
reviews and biological plausibility.
Appendix A
ICD-9 codes
|
Diagnosis
|
Code
|
Description
|
|
Maternal
|
|
Abnormal fetal heart rate
|
659.71
|
Abnormality in fetal heart rate or rhythm, delivered
|
|
Anxiety
|
293.84
|
Anxiety disorder in conditions classified elsewhere
|
|
|
300.00
|
Anxiety state, unspecified
|
|
|
300.01
|
Panic disorder without agoraphobia
|
|
|
300.02
|
Generalized anxiety disorder
|
|
|
300.09
|
Other anxiety states
|
|
Chorioamnionitis
|
658.4x
|
Infection of amniotic cavity
|
|
Depression
|
296.2x
|
Major depressive disorder, single episode
|
|
|
296.3x
|
Major depressive disorder, recurrent episode
|
|
|
311
|
Depressive disorder, not otherwise classified
|
|
Diabetes, gestational
|
648.8x
|
Abnormal maternal glucose tolerance
|
|
Diabetes, pregestational
|
250.x
|
Diabetes mellitus
|
|
|
648.0x
|
Diabetes mellitus of the mother complicating pregnancy
|
|
Hypertension, chronic
|
401.x
|
Essential hypertension
|
|
|
402.x
|
Hypertensive heart disease
|
|
|
403.x
|
Hypertensive chronic kidney disease
|
|
|
404.x
|
Hypertensive heart and chronic kidney disease
|
|
|
405.x
|
Secondary hypertension
|
|
|
642.0x
|
Benign essential hypertension complicating pregnancy
|
|
|
642.1x
|
Hypertension secondary to renal disease complicating pregnancy
|
|
|
642.2x
|
Other preexisting hypertension complicating pregnancy
|
|
Hypertension, gestational
|
642.3x
|
Transient hypertension of pregnancy
|
|
|
642.4x
|
Mild or unspecified preeclampsia
|
|
|
642.5x
|
Severe preeclampsia
|
|
|
642.6x
|
Eclampsia
|
|
|
642.7x
|
Preeclampsia or eclampsia superimposed on preexisting hypertension
|
|
|
642.9x
|
Unspecified hypertension complicating pregnancy
|
|
Placental abruption
|
641.2x
|
Premature separation of the placenta
|
|
Postpartum hemorrhage
|
666.0x
|
Third-stage postpartum hemorrhage
|
|
|
666.1x
|
Other immediate postpartum hemorrhage
|
|
|
666.2x
|
Delayed and secondary postpartum hemorrhage
|
|
|
666.3x
|
Postpartum coagulation defects
|
|
Preterm rupture of membranes
|
658.11
|
Premature rupture of membranes, delivered, with or without mention of antepartum condition
|
|
Previous cesarean
|
654.21
|
Previous cesarean delivery, delivered
|
|
Severe laceration
|
664.2x
|
Third-degree perineal laceration during delivery
|
|
|
664.3x
|
Fourth-degree perineal laceration during delivery
|
|
Shoulder dystocia
|
660.4x
|
Shoulder (girdle) dystocia during labor and delivery
|
|
Neonatal
|
|
Abstinence syndrome
|
292.0
|
Drug withdrawal
|
|
|
779.5
|
Drug withdrawal syndrome in newborn
|
|
Birth defects
|
Numerous
|
http://www.mdch.state.mi.us/pha/osr/CHI/birthdefects/BXDefectCodeGroups.html
|
|
Encephalopathy
|
348.3x
|
Encephalopathy
|
|
|
768.7x
|
Hypoxic–ischemic encephalopathy
|
|
|
779.2
|
Cerebral depression, coma, and other abnormal cerebral signs in fetus or newborn
|
|
Hypoglycemia
|
251.2
|
Hypoglycemia, unspecified
|
|
|
775.6
|
Neonatal hypoglycemia
|
|
Hypoxia
|
768.7x
|
Hypoxic–ischemic encephalopathy
|
|
|
768.9
|
Unspecified birth asphyxia in liveborn infant
|
|
|
770.12
|
Meconium aspiration with respiratory symptoms
|
|
|
770.88
|
Hypoxemia of newborn
|
|
|
775.81
|
Other acidosis of newborn
|
|
Persistent pulmonary hypertension
|
747.83
|
Persistent fetal circulation
|
|
Respiratory failure
|
V46.11
|
Dependence on respirator, status
|
|
|
518.81
|
Acute respiratory failure
|
|
|
518.83
|
Chronic respiratory failure
|
|
|
518.84
|
Acute and chronic respiratory failure
|
|
|
770.84
|
Respiratory failure of newborn
|
|
|
770.87
|
Respiratory arrest of newborn
|
|
Seizures
|
345.x
|
Epilepsy and recurrent seizures
|
|
|
779.0
|
Convulsions in newborn
|
|
Sepsis
|
038.x
|
Septicemia
|
|
771.81
|
Septicemia of newborn
|
|
785.52
|
Septic shock
|
|
995.91
|
Sepsis
|
|
995.92
|
Severe sepsis
|
Abbreviation: ICD-9, International Classification of Diseases, ninth edition.
Prenatal factors that were analyzed included maternal age, body mass index, gestational
age, self-reported race or ethnicity, parity, marital status, medical insurance type,
tobacco use, drug use, diabetes (gestational and pregestational), hypertensive disorders
(chronic hypertension, gestational hypertension, severe preeclampsia, and superimposed
preeclampsia), and previous cesarean delivery.
Intrapartum and maternal factors included induction of labor, exposure to cervical
ripening agents (misoprostol, dinoprostone, oxytocin, and Foley catheter), abnormal
fetal heart rate (ICD-9 code 659.71), systolic and diastolic blood pressures, pulse
rate, temperature, exposure to intravenous antibiotics (gentamicin or clindamycin),
exposure to intravenous antihypertensive medications (labetalol and hydralazine),
exposure to magnesium, exposure to insulin, preterm birth less than 37 weeks' gestation,
type of delivery, severe perineal laceration, chorioamnionitis, placental abruption,
postpartum hemorrhage, blood transfusion, postpartum readmission, preterm rupture
of membranes, shoulder dystocia, and length of stay (LOS).
Neonatal variables included infant sex, birth weight, Apgar scores, meconium-stained
amniotic fluid, birth defects, admission to the neonatal intensive care unit (NICU),
sepsis, persistent pulmonary hypertension, respiratory failure, hypoxia, hypoglycemia,
seizure, encephalopathy, abstinence syndrome, umbilical arterial and venous blood
gases, LOS, and a composite neonatal adverse outcome consisting of 5-minute Apgar
score less than 4, umbilical artery pH less than 7, umbilical artery base excess greater
than –12, sepsis, seizures, encephalopathy, and respiratory failure. Please refer
to the Appendix for a complete list of ICD-9 codes used.
Univariate analyses of associations between maternal and neonatal factors and psychiatric
diagnoses were performed by Student's t-test and chi-square test for continuous and categorical data, respectively. The associations
that were significant on univariate analysis were further evaluated by multinomial
logistic regression comparing each group to the unaffected referent pregnancies to
calculate adjusted odds ratios (aORs) controlling for baseline differences. Because
of the large numbers of cases and to assure clinical relevance, we only reported odds
ratios greater than or equal to 1.5 or less than 0.9. All statistical analyses were
performed using SAS version 9.4 (SAS Institute, Cary, NC). This study was approved
by our hospital's institutional review board.
Results
A total of 19,072 unique maternal–neonatal pairs were assessed for eligibility. Of
these, 915 were excluded for multiple gestations and 2,665 for incomplete records,
the majority of which were because of missing cord gas results. The remaining 15,492
(81.2%) deliveries were analyzed for this study. Of these, depression or anxiety was
present in 978 (6.3%), including anxiety alone (n = 453; 2.9%), depression alone (n = 305; 2%), and comorbid depression and anxiety (n = 220; 1.4%). These cases were further subdivided based on treatment with psychotropic
medications as follows: treated anxiety alone (n = 101), treated depression alone (n = 123), treated comorbid depression and anxiety (n = 92), untreated anxiety alone (n = 351), untreated depression alone (n = 182), and untreated comorbid depression and anxiety (n = 129). These groups were compared with the 14,514 (93.7%) unaffected reference pregnancies.
Considering significant maternal characteristics, patients with anxiety and depression
were more likely to self-report as Non-Hispanic Whites, use tobacco and illegal substances,
and be unmarried. Additionally, both treated depression and untreated depression were
associated with public health insurance, multiparity, and hypertension. Both treated
and untreated comorbid anxiety and depression were associated with hypertension, whereas
treated comorbid anxiety and depression were associated with diabetes and untreated
with public insurance. Only untreated anxiety was associated with decreased odds of
previous cesarean delivery. Maternal age, body mass index, gestational age, and infant
gender were similar across groups ([Table 1]).
Table 1
Baseline characteristics of patients with anxiety and depression (n = 978) compared with unaffected reference pregnancies (n = 14,514)
|
Variable
|
Anxiety, untreated (n = 351)
|
Anxiety, treated (n = 101)
|
Depression, untreated (n = 182)
|
Depression, treated (n = 123)
|
Comorbid depression and anxiety, untreated (n = 129)
|
Comorbid depression and anxiety, treated (n = 92)
|
Unaffected (n = 14,514)
|
|
Maternal age (y)
|
31.8 ± 5.5
|
31.4 ± 5
|
29.7 ± 5.8
|
31.5 ± 6.1
|
29.9 ± 5.4
|
32.3 ± 5.5
|
31 ± 5.2
|
|
Body mass index (kg/m2)
|
31.6 ± 6.2
|
32 ± 6.8
|
33.2 ± 6.8
|
32.5 ± 5.3
|
33.1 ± 7.1
|
32.1 ± 6.9
|
31.5 ± 6.1
|
|
Gestational age (wk)
|
39 ± 2.2
|
38.6 ± 2.6
|
38.7 ± 2.7
|
38.3 ± 2.2
|
38.8 ± 2.7
|
38.3 ± 2.4
|
39.1 ± 1.7
|
|
White race
|
279 (79.5)
1.91 (1.47–2.48)
[a]
|
88 (87.1)
3.33 (1.86–5.97)
[a]
|
123 (67.6)
|
99 (80.5)
2.03 (1.30–3.18)
[b]
|
103 (79.8)
1.95 (1.30–3.18)
[b]
|
80 (87)
3.28 (1.79–6.02)
[b]
|
9,726 (67)
|
|
Nulliparous
|
174 (49.6)
|
49 (48.5)
|
60 (33)
0.67 (0.49–0.91)
[c]
|
40 (32.5)
0.66 (0.45–0.96)
[c]
|
62 (48.1)
|
40 (43.5)
|
6,154 (42.4)
|
|
Married
|
257 (73.2)
0.71 (0.56–0.90)
[b]
|
73 (72.3)
|
111 (61)
0.40 (0.30–0.55)
[a]
|
82 (66.7)
0.52 (0.36–0.75)
[b]
|
74 (57.4)
0.35 (0.25–0.49)
[a]
|
65 (70.7)
|
11,533 (79.5)
|
|
Public insurance
|
72 (20.5)
|
23 (22.8)
|
79 (43.4)
2.60 (1.93–3.49)
[a]
|
40 (32.5)
1.63 (1.12–2.38)
[c]
|
44 (34.1)
1.75 (1.21–2.53)
[b]
|
28 (30.4)
|
3,312 (22.8)
|
|
Tobacco use
|
34 (9.7)
2.30 (1.60–3.30)
[a]
|
13 (12.9)
3.17 (1.76–5.70)
[a]
|
21 (11.5)
2.80 (1.76–4.44)
[a]
|
18 (14.6)
3.68 (2.22–6.10)
[a]
|
20 (15.5)
3.85 (2.17–6.83)
[a]
|
14 (15.2)
3.93 (2.43–6.38)
[a]
|
647 (4.5)
|
|
Drug use
|
20 (5.7)
2.77 (1.74–4.41)
[a]
|
8 (7.9)
3.94 (1.90–8.19)
[a]
|
21 (11.5)
5.98 (3.74–9.55)
[a]
|
5 (4.1)
|
20 (15.5)
8.41 (5.15–13.7)
[a]
|
15 (16.3)
8.93 (5.08–15.7)
[a]
|
310 (2.1)
|
|
Diabetes
|
24 (6.8)
|
13 (12.9)
|
17 (9.3)
|
9 (7.3)
|
13 (10.1)
|
13 (14.1)
1.90 (1.05–3.42)
[c]
|
1,158 (8)
|
|
Hypertension
|
56 (16)
|
17 (16.8)
|
33 (18.1)
1.63 (1.12–2.39)
[c]
|
22 (17.9)
1.61 (1.01–2.55)
[c]
|
26 (20.2)
1.86 (1.21–2.87)
[b]
|
21 (22.8)
2.18 (1.34–3.56)
[b]
|
1,734 (12)
|
|
Previous cesarean
|
53 (15.1)
0.67 (0.50–0.90)
[a]
|
16 (15.8)
|
43 (23.6)
|
28 (22.8)
|
25 (19.4)
|
22 (23.9)
|
3,043 (21)
|
|
Male infant
|
171 (48.7)
|
41 (40.6)
|
98 (53.9)
|
62 (50.4)
|
60 (46.5)
|
45 (48.9)
|
7,626 (52.5)
|
Note: Data are presented as mean ± standard deviation or n (%) and odds ratio (95% confidence interval).
a
p < 0.0001.
b
p < 0.01.
c
p < 0.05.
Significant independently associated intrapartum and maternal factors are listed in
[Table 2]. This shows that untreated anxiety was associated with increased risks of cesarean
deliveries (aOR = 1.69), induction of labor (aOR = 2.02), and longer maternal LOS
(aOR = 1.85), risks not present among patients with treated anxiety. Those with untreated
depression had higher odds of magnesium exposure (aOR = 1.82) and readmission within
6 weeks of delivery (aOR = 2.40), whereas treated depression was associated with decreased
odds of cesarean delivery (aOR = 0.59) but increased odds of requiring a blood transfusion
(aOR = 4.81). Untreated comorbid anxiety and depression were associated with increased
risks of cesarean delivery (aOR = 1.59), fever > 38°C (aOR = 2.03), postpartum hemorrhage
(aOR = 2.57), and maternal LOS (aOR = 1.96), whereas treated comorbid anxiety and
depression were associated with increased risks of severe perineal lacerations (aOR = 2.93)
and postpartum hemorrhage (aOR = 3.85). There were no significant associations with
abnormal fetal heart rate ICD-9 code, systolic or diastolic blood pressure, pulse
rate, exposure to intravenous antibiotics, exposure to intravenous antihypertensive
medications, insulin therapy, preterm birth before 37 weeks' gestation, operative
vaginal delivery, chorioamnionitis, placental abruption, preterm rupture of membranes,
and shoulder dystocia.
Table 2
Intrapartum and maternal factors independently associated with anxiety and depression
(n = 978) compared with unaffected reference pregnancies (n = 14,514)
|
Variable
|
Anxiety, untreated
(n = 351)
|
Anxiety, treated
(n = 101)
|
Depression, untreated
(n = 182)
|
Depression, treated
(n = 123)
|
Comorbid depression and anxiety, untreated
(n = 129)
|
Comorbid depression and anxiety, treated
(n = 92)
|
Unaffected
(n = 14,514)
|
|
Induction of labor
|
96 (27.4)
2.02 (1.56–2.61)
[a]
|
18 (17.8)
|
32 (17.6)
|
18 (14.6)
|
25 (19.4)
|
19 (20.7)
|
2,229 (15.4)
|
|
Fever > 38°C
|
16 (4.6)
|
3 (3)
|
10 (5.5)
|
4 (3.3)
|
11 (8.7)
2.03 (1.07–3.86)
[b]
|
4 (4.4)
|
618 (4.3)
|
|
Magnesium exposure
|
25 (7.2)
|
7 (6.9)
|
21 (11.5)
1.82 (1.04–3.20)
[b]
|
6 (4.9)
|
8 (6.3)
|
8 (8.8)
|
719 (5)
|
|
Cesarean delivery
|
167 (47.6)
1.69 (1.31–2.18)
[a]
|
42 (41.6)
|
70 (38.5)
|
41 (33.3)
0.59 (0.36–0.96)
[b]
|
64 (49.6)
1.59 (1.04–2.42)
[b]
|
35 (38)
|
5,862 (40.4)
|
|
Severe laceration
|
6 (1.7)
|
3 (2.9)
|
3 (1.7)
|
3 (2.4)
|
2 (1.5)
|
4 (4.5)
2.93 (1.02–8.38)
[b]
|
284 (2)
|
|
Postpartum hemorrhage
|
10 (2.9)
|
1 (1)
|
5 (2.8)
|
7 (5.7)
|
8 (6.2)
2.57 (1.23–5.35)
[b]
|
9 (9.8)
3.85 (1.90–7.82)
[c]
|
398 (2.7)
|
|
Blood transfusion
|
3 (0.9)
|
1 (0.9)
|
2 (1.1)
|
5 (4.1)
4.81 (1.86–12.4)
[c]
|
3 (2.3)
|
2 (2.2)
|
128 (0.9)
|
|
Readmission within 6 wk of delivery
|
6 (1.7)
|
4 (4)
|
7 (3.9)
2.40 (1.10–5.21)
[b]
|
1 (0.8)
|
4 (3.1)
|
2 (2.2)
|
229 (1.6)
|
|
Length of stay > 5 d
|
31 (8.8)
1.85 (1.21–2.84)
[b]
|
5 (5)
|
17 (9.3)
|
7 (5.7)
|
10 (7.8)
1.96 (1.07–3.61)
[b]
|
8 (8.7)
|
598 (4.1)
|
Note: Data are presented as n (%) and adjusted odds ratio (95% confidence interval). Adjusted for maternal age,
body mass index, gestational age, race/ethnicity, parity, marital status, smoking
status, and history of drug use. There were no significant associations for abnormal
fetal heart rate ICD-9 (International Classification of Diseases) code, systolic or
diastolic blood pressure, pulse rate, exposure to intravenous antibiotics, exposure
to intravenous antihypertensive medications, insulin therapy, preterm birth <37 weeks'
gestation, operative vaginal delivery, chorioamnionitis, placental abruption, preterm
rupture of membranes, and shoulder dystocia.
a
p < 0.0001.
b
p < 0.05.
c
p < 0.01.
Significant independent associations of neonatal outcomes are listed in [Table 3]. This showed that untreated anxiety was associated with 5-minute Apgar score < 4
(aOR = 3.82) whereas treated anxiety was associated with meconium-stained amniotic
fluid (aOR = 1.85), respiratory failure (aOR = 5.84), and neonatal abstinence syndrome
(aOR = 11). Neonates of mothers with untreated depression had increased odds of hypoxia
(aOR = 2.83), but treated depression was associated with neonatal seizures (aOR = 12.3)
and neonatal abstinence syndrome (aOR = 10.2). Untreated comorbid anxiety and depression
were associated with encephalopathy (aOR = 15), whereas treated comorbid disease was
associated with increased odds of being SGA (aOR = 3.04), meconium-stained amniotic
fluid (aOR = 2.61), arterial base excess greater than –12 (aOR = 3.42), and venous
base excess greater than –12 (aOR = 5.62). There were no significant differences in
admission to the NICU, birth defects, persistent pulmonary hypertension, sepsis, hypoglycemia,
arterial and venous pH < 7, LOS, and the neonatal composite adverse outcome.
Table 3
Neonatal outcomes independently associated with anxiety and depression (n = 978) compared with unaffected reference pregnancies (n = 14,514)
|
Variable
|
Anxiety, untreated
(n = 351)
|
Anxiety, treated
(n = 101)
|
Depression, untreated
(n = 182)
|
Depression, treated
(n = 123)
|
Comorbid depression and anxiety, untreated
(n = 129)
|
Comorbid depression and anxiety, treated
(n = 92)
|
Unaffected
(n = 14,514)
|
|
Birth weight (g)
|
3,340 ± 617
|
3,220 ± 643
|
3,228 ± 670
|
3,156 ± 608
|
3,351 ± 653
|
3,036 ± 601
0.89 (0.84–0.93)
[a]
|
3,383 ± 542
|
|
Small for gestational age
|
22 (6.3)
|
4 (4)
|
15 (8.2)
|
6 (4.9)
|
10 (7.8)
|
13 (14.1)
3.04 (1.63–5.70)
[b]
|
672 (4.6)
|
|
Five-minute Apgar score < 4
|
1 (0.3)
3.82 (3.23–4.52)
[a]
|
1 (1)
|
2 (1.1)
|
0 (0)
|
1 (0.8)
|
0 (0)
|
11 (0.1)
|
|
Meconium-stained amniotic fluid
|
47 (13.4)
|
18 (17.8)
1.85 (1.10–3.11)
[c]
|
25 (13.7)
|
18 (14.6)
|
16 (12.4)
|
21 (22.8)
2.61 (1.58–4.32)
[a]
|
1,705 (11.8)
|
|
Admission to neonatal intensive care unit
|
57 (16.2)
|
26 (25.7)
|
28 (15.4)
|
31 (25.2)
|
19 (14.7)
|
26 (28.3)
|
2,030 (14)
|
|
Birth defect
|
17 (4.8)
|
3 (2.9)
|
10 (5.5)
|
4 (3.3)
|
6 (4.7)
|
3 (3.3)
|
491 (3.4)
|
|
Persistent pulmonary hypertension
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
29 (0.2)
|
|
Sepsis
|
2 (0.6)
|
0 (0)
|
2 (1.1)
|
0 (0)
|
1 (0.8)
|
0 (0)
|
20 (0.1)
|
|
Respiratory failure
|
1 (0.3)
|
2 (2)
5.48 (1.07–28.1)
[c]
|
2 (1.1)
|
0 (0)
|
0 (0)
|
1 (1.1)
|
32 (0.2)
|
|
Hypoxia
|
5 (1.4)
|
2 (2)
|
7 (3.9)
2.83 (1.15–7.01)
[c]
|
2 (1.6)
|
3 (2.3)
|
3 (3.3)
|
127 (0.9)
|
|
Hypoglycemia
|
17 (4.8)
|
11 (10.9)
|
11 (6)
|
13 (10.6)
|
6 (5.7)
|
9 (9.8)
|
736 (5.1)
|
|
Seizures
|
0 (0)
|
0 (0)
|
0 (0)
|
1 (0.8)
12.3 (1.28–119)
[c]
|
0 (0)
|
0 (0)
|
9 (0.1)
|
|
Encephalopathy
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
1 (0.8)
18.3 (2.16–156)
[c]
|
0 (0)
|
7 (0.1)
|
|
Adaptation syndrome
|
3 (0.9)
|
5 (5)
11 (3.41–35.7)
[a]
|
1 (0.6)
|
4 (3.3)
10.2 (2.98–34.7)
[b]
|
0 (0)
|
2 (2.2)
|
31 (0.2)
|
|
Arterial base excess greater than –12
|
5 (1.4)
|
1 (1)
|
0 (0)
|
0 (0)
|
2 (1.6)
|
3 (3.3)
3.42 (1.05–11.1)
[c]
|
124 (0.9)
|
|
Arterial pH < 7
|
2 (0.6)
|
0 (0)
|
0 (0)
|
0 (0)
|
1 (0.8)
|
1 (1.1)
|
44 (0.3)
|
|
Venous base excess greater than –12
|
2 (0.6)
|
1 (1)
|
0 (0)
|
0 (0)
|
1 (0.8)
|
3 (3.4)
5.62 (1.63–19.4)
[b]
|
62 (0.4)
|
|
Venous pH < 7
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
0 (0)
|
1 (1.1)
|
19 (0.1)
|
|
Length of stay > 3 d
|
47 (13.4)
|
14 (13.7)
|
25 (13.7)
|
17 (13.8)
|
18 (13.7)
|
18 (20.2)
|
1,409 (9.7)
|
|
Neonatal composite adverse outcome
|
9 (2.6)
|
4 (4)
|
4 (2.2)
|
1 (0.8)
|
4 (3.1)
|
4 (4.4)
|
168 (1.2)
|
Note: Data are presented as mean ± standard deviation or n (%) and adjusted odds ratio (95% confidence interval). For birth weight, the odds
ratio refers to a 100-g increase. Adjusted for maternal age, body mass index, gestational
age, race/ethnicity, parity, marital status, smoking status, history of drug use,
and mode of delivery. Composite neonatal adverse outcome: 5-minute Apgar score < 4,
umbilical arterial pH < 7, umbilical arterial base excess greater than –12, sepsis,
seizures, encephalopathy, and respiratory failure.
a
p < 0.0001.
b
p < 0.01.
c
p < 0.05.
Psychotropic medications were used by 339 subjects (34.6%). The most common medications
were SSRIs (n = 271; 79.9%), benzodiazepines (n = 54; 15.9%), and other antidepressants (n = 35; 10.3%). Other medications included serotonin–norepinephrine reuptake inhibitors
(n = 18; 5.7%), antiepileptics (n = 13; 4.4%), and antipsychotics (n = 10; 3.2%). A single drug was used by 285 (84.1%) and multiple drugs by 54 (15.9%).
We did a subanalysis and compared those with multiple drugs (polypharmacy) with those
who used single psychotropic drugs regarding neonatal outcomes. On univariate analysis,
neonates born of mothers with multiple prenatal psychotropic drug use were significantly
more likely to be SGA and have neonatal hypoglycemia, neonatal abstinence syndrome,
NICU admission, and longer neonatal LOS; however, on adjusted regression analysis,
NICU admission and neonatal LOS were no longer significant ([Table 4])
Table 4
Significant associations of prenatal single versus multiple psychotropic drug use
with neonatal outcomes (n = 339)
|
Neonatal factor
|
Single psychotropic drug, n = 285
|
Multiple psychotropic drugs, n = 54
|
Univariate p-value (OR)
|
Independent p-value (OR) [95%CI]
|
|
Male gender
|
134 (47%)
|
29 (53.7%)
|
ns
|
ns
|
|
Birth weight
|
3,058.86 (43.6)
|
2,877.17 (100.58)
|
ns
|
ns
|
|
Neonatal hypoglycemia
|
25 (8.8%)
|
11 (20.4%)
|
(2.66)[a]
|
(2.6) [ 1.04–6.55][a]
|
|
Neonatal abstinence syndrome
|
3 (1.1%)
|
8 (14.8%)
|
(16.348)[b]
|
(8.3) [1.72–39.9][c]
|
|
Admission to neonatal intensive care unit
|
80 (28.1%)
|
25 (46.3%)
|
(2.2)[a]
|
ns
|
|
SGA
|
18 (6.3%)
|
8 (14.8%)
|
(2.57)[a]
|
|
|
Length of stay > 3 d
|
50 (17.5%)
|
22 (40.7%)
|
(3.23)[b]
|
(3.2) [1.15–9.09][a]
|
Abbreviations: CI, confidence interval; ns, not significant; OR, odds ratio; SGA,
small for gestational age.
Note: There were no significant associations with the neonatal composite score, arterial
pH < 7, venous base excess greater than –12, venous pH < 7, 5-minute Apgar score < 4,
and meconium-stained amniotic fluid. The independent p-value was adjusted for maternal age, body mass index, gestational age, race/ethnicity,
parity, marital status, smoking status, history of drug use, and mode of delivery.
a
p < 0.05.
b
p < 0.0001.
c
p < 0.01.
Discussion
In this hospital system, the prevalence of anxiety and depression overall was 6.3%,
which is somewhat lower than observed in other studies.[6] This is probably because most previous reports used self-reported screening instruments
in the identification of cases—a process that may overestimate with a resultant higher
prevalence in comparison to this study, which used ICD-9 coding and detailed chart
review for identification of cases.[1]
[10] However, the low prevalence noted in this study could also be because of potential
misclassification of both “noncases” and “cases.” It is likely that there are several
missed cases of depression or anxiety in the reference group. The findings in this
study support other previous reports showing associations between anxiety/depression
and smoking, substance abuse, and unmarried status.[11]
[12]
[13]
[14]
[15]
[16]
[17] But in contrast to most other studies that have likewise shown associations with
minority ethnic groups, our study showed an association with Non-Hispanic white women.[11]
[12]
[13]
[14]
[15]
[16]
[17] The increased risk of perinatal anxiety and depression in minority groups has been
explained by stress-related discrimination in addition to financial and social disadvantages.[18] However, another report in support of our findings demonstrated higher levels of
depression among White mothers compared with black and Hispanic mothers as well as
women of other ethnicities.[19] As an explanation, other studies have demonstrated that prolonged and repeated exposure
to adversity might provide minority women with greater resilience and skills to manage
psychological distress in addition to having higher levels of social support, spirituality,
and self-esteem resources.[20] Correspondingly, our study highlights that Non-Hispanic Whites may also be at increased
risk of anxiety and depression during pregnancy and the need for universal screening
of all pregnant women.
In this study, untreated prenatal anxiety and depression were associated with more
adverse maternal outcomes than treated (7 vs. 3, respectively) when both groups were
compared with those without anxiety and depression. Both prenatal anxiety and comorbid
anxiety and depression when untreated were associated with an increased risk of cesarean
birth, whereas treated depression demonstrated a decreased risk. A past history of
cesarean delivery has been found to be associated with a high incidence of antenatal
anxiety and depression.[21]
[22] However, our study did show a decreased risk of previous cesarean delivery with
those with untreated anxiety. The association of anxiety with previous cesarean birth
may not be very reliable since its validity would depend on the following: (1) if
the patient did have a diagnosis of anxiety during the previous pregnancy and (2)
if she was treated in that pregnancy. Our study's finding of increased risk of cesarean
birth with untreated prenatal anxiety and depression with a corresponding decrease
when treated is intriguing. Other studies have found no association between mode of
delivery and antenatal depression,[23]
[24] whereas another previous study did show an association between untreated depression
and cesarean section.[25] A recent systemic review showed that those requesting elective cesarean sections
had higher antepartum depression and anxiety levels but no different postpartum depression
levels than women who delivered vaginally.[26] It is important to note that in a Finnish study on major depression in pregnancy,
after a history of depression, the second strongest associated factor for major depression
was fear of childbirth, with a 2.6-fold increased prevalence.[10] Furthermore, other studies have shown that patients with antenatal anxiety and depression
have a fear of childbirth and show a preference for an elective cesarean delivery.[1] These findings suggest a need for greater attention to continuous assessment of
psychological well-being among women, especially those requesting elective cesarean
birth, as a potential intervention for the current major public health concern with
the increasing prevalence of cesarean delivery. Further prospective studies are required
to test this hypothesis.
In this study, treated prenatal anxiety and depression were also associated with more
adverse neonatal outcomes than untreated cases (7 versus 3 respectively) when both
groups were compared with those without anxiety and depression. The increased association
of treatment of anxiety/depression with psychotropic drugs during pregnancy with neonatal
adaptation syndrome supports previous findings.[9] Furthermore, treatment of comorbid anxiety and depression was associated with a
threefold risk for SGA neonate, whereas there was no association with preterm birth.
The impact of antidepressants on fetal growth has been evaluated with inconclusive
results, with most studies reviewed lacking adequate control groups with heterogeneity
of outcomes.[9] In contrast to our findings, a recent study demonstrated an association of a positive
screen for depression using Edinburgh Postnatal Depression Scale (EPDS) with preterm
birth and SGA, but it noted that such risk was not apparent among women who were treated
with an antidepressant medication.[27] The difference in results may be partially explained by methodology since in our
study we identified patients with depression or anxiety based on ICD-9 coding and
we compared identified patients with depression or anxiety with those who were not
unaffected based on the use of psychotropic drugs. In the aforementioned study, patients
were identified by EPDS, and all patients who received antidepressants (either screen-positive
or screen-negative) were compared with all those who did not use antidepressants (both
screen-positive and screen-negative). Meconium-stained fluid, which is associated
with intrauterine hypoxia, was also increased in treated cases in our study. Our findings
and those of previous reports do support that psychotropic medications in pregnancy
are associated with some neonatal morbidity.
It should also be noted that untreated anxiety and depression were not without neonatal
morbidity and were associated with increased risks including low Apgar scores and
neonatal hypoxia. It conceivable that the increased maternal morbidity associated
with untreated anxiety and depression noted in this study and previous works can predispose
to neonatal morbidity.[1]
[10]
[27]
[28] For example, maternal fever, which is a predictor of neonatal sepsis and morbidity,
was only associated with untreated comorbid anxiety and depression. Furthermore, studies
in pregnant women have shown increased relapse rates in those who discontinued their
psychotropic medication during pregnancy, with increased risks of hospitalization
and suicide ideation.[29]
[30] Neonatal adaptation syndrome, which is the main neonatal morbidity associated with
antidepressant use, is known to be transient with the symptoms resolving within a
week with no apparent long-term adverse effects.[9] Correspondingly, the risks associated with pharmacological treatment must be balanced
with the effects of untreated antenatal maternal depression on the mother–fetus dyad,
which have the potential to be devastating.
Limitations
There are several limitations to this study. This is a retrospective cohort study
design, which has inherent limitations including missing data and accuracy of data.
Specifically, the used of ICD-9 codes for identifying cases has intrinsic limitations,
namely, accuracy of diagnostic code, potentially nonactive diagnoses, and noncoded
cases (i.e., unidentified cases). The latter is supported by the fact that the prevalence
of anxiety/depression in our study (6.3%) is lower than that reported in the literature.[31] Additionally, there were no data regarding the adequacy of prenatal care, duration
of anxiety and depression, and the level of control entering the pregnancy. However,
we did manual review and audits of the charts for verification and confirmation of
the data. We also performed multivariate logistic regression to eliminate potential
confounders, but it is possible that there are other factors that were not measured.
Such potential cofounders include quantity, timing, and duration of psychotropic medication
therapy. We also did not assess postpartum depression; however, multiple studies have
already confirmed the strong association of prenatal anxiety and depression with postpartum
depression.[30] We also did not directly compare treated cases versus untreated cases of anxiety
and depression, which would have provided differential effects of the exposure of
medications, but compared against unaffected pregnancies to determine the baseline
burden of disease and treatment. Moreover, there are always concerns with medication
compliance, and modest associations regarding medication treatment and outcomes can
be attributed to misclassification because nondifferential misclassification of the
exposure or the outcome will tend to bias results toward the null hypothesis.[32]
[33] Another limitation is that many of the neonatal adverse outcomes such as respiratory
failure, neonatal seizures, encephalopathy, low Apgar score, and abnormal blood gases
had very small numbers with correspondingly very wide 95% confidence intervals. Hence,
these findings, though significant, may not be clinically relevant. Furthermore, this
is a single-system study in southeastern Michigan; therefore, the findings may not
be applicable to different centers with different populations. Strengths of this study
include assessment of a large cohort of mother–fetus dyads using uniform methodology
and ascertainment of depression, anxiety, and maternal and neonatal outcomes by the
electronic health record with detailed review, thus avoiding the risk of recall bias
and overestimation seen in some earlier studies. Through detailed categorization of
these pregnant women, we were able to control for important confounding variables
associated with maternal and neonatal morbidity.
Conclusion
In contrast to the majority of previous studies, Non-Hispanic White mothers were at
a greater risk of prenatal anxiety and depression in addition to other social disadvantaged
factors, substance use, and medical disorders. Women with untreated anxiety and depression
in pregnancy had more adverse maternal outcomes. Untreated anxiety appeared to be
a risk factor for induction of labor and cesarean delivery. Treated patients with
anxiety and depression had more neonatal adverse outcomes; however, untreated patients
still had some neonatal morbidity presumably as a consequence of significant maternal
morbidity. A third of women used psychotropic medications during pregnancy, with an
increased risk of neonatal adaptation syndrome, which is known to be transient without
long-time sequelae. The universal identification of women at risk of anxiety and depression
during pregnancy may provide opportunities for therapy and promote well-being of mothers
and babies.
