- Email: [email protected]
Increasing use of antidepressants in pregnancy
Research
www. AJOG.org
OBSTETRICS
Increasing use of antidepressants in pregnancy William O. Cooper, MD, MPH; Mary E. Willy, PhD; Stephen J. Pont, MD; Wayne A. Ray, PhD OBJECTIVE: The purpose of this study was to quantify the rate of exposures to antidepressants during pregnancy in a large cohort of women. STUDY DESIGN: This was a retrospective cohort study of 105,335 pregnancies among women enrolled in Tennessee Medicaid from 1999-2003. Pregnancies were classified according to antidepressant exposures during pregnancy using previously validated computerized pharmacy records linked with birth certificates.
white race (P ⬍ .0001), and education ⬎ 12 years (P ⫽ .008) were significant predictors of antidepressant exposure. The proportion of pregnancies with antidepressant use increased from 5.7% of pregnancies in 1999 to 13.4% of pregnancies in 2003 (p⬍.0001). The increase was mostly accounted for by increases in selective serotonin reuptake inhibitor exposures. CONCLUSION: There is an urgent need for further studies that better
quantify the fetal consequences of exposure to antidepressants.
RESULTS: During the study period, 8.7% of women giving birth had
exposure to any antidepressant; 6.2% had exposure to a selective serotonin reuptake inhibitor. Maternal age ⬎ 25 years (P ⬍ .0001),
Key words: antidepressants, fetal effects, medication exposures, pregnancy
Cite this article as: Cooper WO, Willy ME, Pont SJ, et al. Increasing use of antidepressants in pregnancy. Am J Obstet Gynecol 2007;196;544.e1-544.e5.
I
n recent years, selective serotonin reuptake inhibitors (SSRIs) have been increasingly used to treat major depression.1 However, recent studies have reported an increased risk of adverse neonatal effects related to both early and late fetal exposure to SSRIs, which have in-
From the Departments of Pediatrics (Drs Cooper and Pont) and Preventive Medicine (Dr Ray), Vanderbilt University School of Medicine, Nashville, TN, and the Division of Drug Risk Evaluation, Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD (Dr Willy). Received Aug. 15, 2006; revised Nov. 28, 2006; accepted Jan. 24, 2007. Reprints: William O. Cooper, MD, MPH, Associate Professor of Pediatrics, Department of Pediatrics, Vanderbilt University School of Medicine, AA-0216 MCN, Nashville, TN 37232-2504; [email protected] Funding to conduct the study was provided by the Food and Drug Administration (contract #HHSF223200510010C) and the Centers for Education and Research on Therapeutics, Agency for Healthcare Research and Quality (HS-0384). 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2007.01.033
544.e1
cluded both neurologic and cardiovascular abnormalities.2-6 Given that the prevalence of major depressive disorders in women of reproductive age is thought to be 10-15%, it is possible that large numbers of pregnant women are taking these medications.7 To estimate the potential public health consequences of SSRI use in pregnancy, it is important to know how many fetuses are potentially exposed to SSRIs and other antidepressant medications. At present there is limited information on the use of antidepressants during pregnancy.8 The aims of this study were to quantify the rate of pregnancy related exposures to antidepressants according to drug type and trimester of pregnancy, and to assess temporal trends in their use.
M ATERIALS AND M ETHODS The study was conducted using data from TennCare, Tennessee’s expanded Medicaid program, for which computerized records of filled prescriptions have been shown to be an accurate source of medication data and have high concordance with patient self-report of medication use.9-11 Links to vital records (birth, death, and fetal death certificates) files and medical records permit identifica-
American Journal of Obstetrics & Gynecology JUNE 2007
tion of pregnant women, including estimated conception dates.9,12-14 Vital records and TennCare enrollment files provided information on maternal and infant factors, including maternal age, race (from birth certificate self-report), and county of residence. The date of the last menstrual period was taken from the birth certificate when available (89.3% of all births) or estimated from birth weight using the raceand calendar year-specific distributions of gestational age for births with known last menstrual period (LMP).15 In a previous study using the same databases, the birth certificate LMP was within two weeks of the LMP found in the medical record 94% of the time.14 The first trimester was defined as the last menstrual period through the following 90 days, the second trimester as the next 90 days, and the third trimester as the remainder of the pregnancy. Pregnancies were included in the study if the date of delivery was between January 1, 1999 and December 31, 2003 and the mother was a Tennessee resident with complete information on the birth certificate for maternal age and date of delivery. Pregnancies were further restricted to those in which the mother was enrolled in the TennCare program from
Obstetrics
www.AJOG.org
Research
TABLE 1
Factors predicting any antidepressant and SSRI antidepressant exposures during pregnancy, women enrolled in TennCare, 1999-2003 Any antidepressant use
SSRI use
Characteristic*
Odds ratio
95% confidence interval
Odds ratio
95% confidence interval
Maternal age ⬎ 25 years
2.04
1.94-2.14
1.95
1.84-2.06
White race
4.33
4.02-4.67
4.55
4.17-4.97
Education ⬎ 12 years
1.09
1.02-1.15
1.15
1.08-1.24
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
* Logistic regression models included birth year, maternal age, race, years of education, county of residence, neighborhood income, and parity. Reference groups for the characteristics shown included age ⬍ 25 years (maternal age), black race (race), and education ⬍12 years (education).
30 days prior to the date of the last menstrual period through the date of delivery or fetal death with no gaps in enrollment lasting longer than 7 days. From a total of 235,099 pregnancies identified from birth certificates and 1243 fetal deaths among women enrolled in TennCare at delivery, 105,335 pregnancies met all requirements for inclusion in the cohort. Most exclusions were for mothers who enrolled in TennCare after the LMP. All study pregnancies were classified according to antidepressant use during pregnancy. Prescribed medications were identified from TennCare pharmacy files, which included the date that the prescription was filled and the prescribed number of days of medication supply. TennCare pharmacy files contain information on all outpatient prescriptions which are reimbursed by TennCare, including prescriptions filled in pharmacies, purchased over the internet, or filled in other states. Maternal medication exposure during a given trimester was defined as presence of 1 or more days of supply during the trimester. Thus, if the prescription was filled during 1 trimester (or before the LMP) and the days of supply extended into the next trimester, exposure was considered to have occurred in both trimesters. In the assessment of temporal trends, the calendar year to which all maternal medication use was assigned was that of the date of delivery. Antidepressant medications studied included 4 categories of antidepressants: SSRIs (fluoxetine, sertraline, paroxetine, fluvoxamine, citalopram, escitalopram), serotonin-norepinephrine reuptake inhibitors (venlafaxine,
duloxetine), tricyclic and tetracyclic antidepressants (amitriptyline, desipramine, doxepin, imipramine, nortriptyline, proptriptyline, trimispramine, amoxapine, maprotiline, clomipramine), and other antidepressants (trazodone, bupropion, mirtazapine, phenelzine, tranylcypromine). For each pregnancy, only 1 exposure per category of antidepressants during a given trimester was counted. Exposures to multiple antidepressant categories during a trimester were allowed. Logistic regression was used to identify maternal factors associated with antidepressant use, including maternal age, race, parity, education, residence in a standard metropolitan statistical area, and neighborhood income quintile. Data were analyzed at the individual pregnancy level. To account for correlation due to women with multiple pregnancies during the study period, the robust Huber-White variance estimator was used to provide correct standard errors and confidence intervals of the estimated odds ratios.16 All regressions were performed with SAS version 9.1 (SAS Institute, Cary, NC). The analysis of temporal trends included adjustment for potential changes in maternal characteristics during the study period. To calculate the annual adjusted proportions of antidepressant use during pregnancy, calendar year was added to the model, with 2003 as the reference category. The adjusted odds (O) for each of the other years were obtained by multiplying the unadjusted odds for 2003 by the adjusted odds ratio for each calendar year. This was then converted to a proportion (p) using the identity p ⫽ O/(1⫹O).
Permission to perform the study and waiver of informed consent was obtained from the Vanderbilt University Institutional Review Board, the State of Tennessee Health Department, and the TennCare Bureau.
R ESULTS During the study period, the number of mothers who qualified for the cohort each year ranged from 19,421 to 21,997. The mean age (range of mean age 23.2 ⫾ 5.3 years to 23.3 ⫾ 5.2 years) and proportions of mothers who were prima gravida (proportion for each year ranged from 24.3-25.5%), were black race (28.240.5%), had greater than 12 years of education (14.1-15.1%), resided in an urban county (43.0-45.5%), and who resided in neighbors with the lowest income (23.6-26.8%) did not change materially during the 5 year period. Among the women giving birth during the entire study period, 8.7% had exposure to any antidepressant during pregnancy and 6.2% to an SSRI antidepressant. Maternal age greater than or equal to 25 years (P ⬍ .0001), white race (P ⬍ .0001), and education greater than 12 years (P ⬍ .01) were significant predictors of both any antidepressant and SSRI antidepressant exposure (Table 1). White race was the strongest predictor of pregnancy-related antidepressant exposure with odds ratios of 4.33 (95% confidence interval 4.02-4.67) for any antidepressant exposure and 4.55 (95% confidence interval 4.17-4.97) for SSRI exposure. The most commonly prescribed antidepressants included sertraline, with ex-
JUNE 2007 American Journal of Obstetrics & Gynecology
544.e2
Research
Obstetrics
www.AJOG.org
FIGURE
Percentage of pregnancies with antidepressant exposure 25.00%
Non-SSRI SSRI
20.00%
15.00%
10.00%
5.00%
0.00% 1999
2000
2001
2002
2003
Proportion of pregnancies with antidepressant exposure among women giving birth between 1999-2003 who were enrolled throughout pregnancy in the TennCare program.
Use of non-SSRI antidepressants increased as well, but to a lesser extent than that for the SSRI antidepressants. For women giving birth in 2003, 10.0% had first trimester antidepressant exposure and 6.4% and 5.9% had second and third trimester exposure, respectively (Table 2). SSRI exposure occurred most commonly, with the highest rate of exposures occurring in the first trimester. There was less frequent use during pregnancy of serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants, and other antidepressants, but, as was the case for the other antidepressants, such exposure was most frequent during the first trimester (Table 2).
C OMMENT posure occurring in 2.1% of pregnancies, paroxetine, with exposure occurring in 1.9% of pregnancies, and fluoxetine, with exposure occurring in 1.9% of pregnancies. Other commonly used antidepressants included amitriptyline (1.0% of pregnancies), trazodone (0.9% of pregnancies), and citalopram (0.8% of pregnancies). After adjusting for maternal age, race, parity, education, residence in a standard metropolitan statistical area, and neighborhood income quintile, the proportion of pregnancies with antidepressant use increased from 5.7% of pregnancies in 1999 to 13.4% of pregnancies by 2003, a more than 2-fold increase (Figure). This was largely due to increased use of SSRIs, with proportions increasing from 2.9% of pregnancies in 1999 to 10.2% of pregnancies in 2003.
In this study of antidepressant use during pregnancy in a large cohort of women enrolled in the TennCare program, we found a marked temporal trend of increasing use during the 5-year period 1999 through 2003. By 2003, more than 13% of pregnancies had an antidepressant exposure. SSRIs accounted for most of the increase in antidepressant exposures, with use of these drugs by more than 10% of pregnant women by 2003. Use of other antidepressants during pregnancy increased at a more moderate rate during the study period. Mothers most likely to have pregnancy-related antidepressant exposures included older women, white women, and women with more than 12 years of education. In other studies of medication use, white race has also been associated with greater use of antidepres-
sants.17 Interestingly, some populationbased studies have suggested that maternal depressive symptoms occur more frequently among racial and ethnic minorities.18 While younger maternal age has been associated with more frequent depressive symptoms among pregnant women, antidepressant use in the present study was more common among older mothers.18 Further research to understand the relationship between these factors and use of antidepressants is needed. The prevalence of antidepressant exposure in our cohort was higher than that described previously. Most of the prior population-based studies of antidepressant use in pregnancy include antidepressants in analyses of all other medications or of psychotropic medications as a group.2,19-21 Studies in European countries using data through 2001 identified use of antidepressants in ⬍ 1-1.4% of pregnancies.2,19 Reports from the United States, including a study from a large HMO research network by Andrade et al, noted slightly higher frequency of use—from 2.8-3.6%.20,21 The most probable explanation for the higher proportions of pregnancies with antidepressant use in the current study is that the data are from more recent years. Data for the first 2 years of our study were comparable to that described by Andrade et al. In addition, the current study included large numbers of Medicaid enrollees, who are known to have higher rates of chronic health conditions, including mental illness.22 Utilization of Medicaid pharmacy files minimizes concerns of selection bias, re-
TABLE 2
Use of antidepressants during pregnancy for 2003, according to drug class and trimester % of Pregnancies Medication exposure
First trimester
Second trimester
3rd trimester
Any exposure
Any antidepressant
10.0
6.4
5.9
13.3
SSRI antidepressants
7.2
5.0
4.8
10.2
SNRI antidepressants
1.3
0.5
0.4
1.4
Tricyclic antidepressants
1.1
0.4
0.3
1.3
Other antidepressants
2.4
1.2
0.9
3.1
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
SSRI, selective serotonin reuptake inhibitors; SNRI, serotonin norepinephrine reuptake inhibitors.
544.e3
American Journal of Obstetrics & Gynecology JUNE 2007
Obstetrics
www.AJOG.org call bias, and differential misclassification that may be associated with retrospective assessment of medication exposures. However, use of these files imposes limitations of their own. The prescription files recorded that a woman filled an antidepressant prescription during pregnancy, but did not indicate whether or not the prescription was taken as prescribed. It is possible that women may have stopped medication use after pregnancy was diagnosed, but in many cases, women filled antidepressant prescriptions during the second and third trimesters of pregnancy. Additionally, Medicaid pharmacy files were limited to outpatient prescription medications. Thus, for the small number of women who obtained prescriptions outside of the Medicaid program, exposures would not be captured. Because of the income requirements for participation in the Medicaid program, it is not anticipated that many persons paid for prescriptions out-of-pocket. Our study included only pregnant women enrolled in TennCare; thus, the results may not be generalizable to other populations. The study population included disproportionate numbers of mothers ⬍ 25 years of age, teenage mothers, mothers with ⬍ 12 years of education, and mothers living in urban areas.23 Nevertheless, Medicaid enrollees include a large proportion of all pregnant women and are thus important in their own right. In recent years in Tennessee, approximately 50% of women giving birth were enrolled in TennCare at the time of delivery; nationally, over one third of all births were covered by Medicaid.23,24 Utilizing vital statistics from the year 2003 and applying the 13.3% proportion of pregnancies with antidepressant exposure from that year of our study suggests that among Medicaid enrollees in the US, there were approximately 180,000 fetuses exposed to antidepressants each year.25 Exposure to antidepressants during both early and late pregnancy has been associated with adverse fetal outcomes, including cardiac malformations and pulmonary hypertension.5,6 There have been 2 Food and Drug Administration public health advisories noting these po-
tential risks.5,26 Prescribing of medications to women during pregnancy requires balancing the benefits to the mother against the risks to the developing fetus. This cannot be done unless there is accurate information on the medication’s fetal effects. Affective disorders are common among women of childbearing age and it is recommended that all pregnant women should be screened and treated for depression.27 Given that untreated depression may have serious clinical and social perinatal consequences, it would be difficult to avoid all use of antidepressant medications during pregnancy.26,28 Although the available data suggest there are adverse effects associated with fetal exposure to some antidepressants, these data are limited, particularly with regard to the effects of individual drugs.5,24 Thus, there is an urgent need for further studies that better quantify the fetal consequences of exposure to antidepressants. f ACKNOWLEDGMENTS We gratefully acknowledge the Tennessee Department of Health and the TennCare Bureau, who provided the data for the study.
REFERENCES 1. Pirraglia PA, Stafford RS, Singer DE. Trends in prescribing of selective serotonin reuptake inhibitors and other newer antidepressant agents in adult primary care. Prim Care Companion. J Clin Psychiatry 2003;5:153-7. 2. Kallen B. Fluoxetine use in early pregnancy. Birth Defects Res B Dev Reprod Toxicol. 2004;71:395-6. 3. Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1996;335:1010-5. 4. Moses-Kolko EL, Bogen D, Perel J et al. Neonatal signs after late in utero exposure to serotonin reuptake inhibitors: literature review and implications for clinical applications. JAMA 2005;293:2372-83. 5. Food and Drug Administration. FDA Public health advisory: paroxetine. Electronic citation. Available at: http://www.fda.gov/cder/drug/ advisory/paroxetine200512.htm. Accessed March 2, 2007. 6. Chambers CD, Hernandez-Diaz S, Van Marter LJ, et al. Selective serotonin-reuptake inhibitors and risk of persistent pulmonary hypertension of the newborn. N Engl J Med 2006;354:579-87. 7. Bennett HA, Einarson A, Taddio A, Koren G, Einarson TR. Prevalence of depression during
Research
pregnancy: systematic review. Obstet Gynecol. 2004;103:698-709. 8. Mills JL. Depressing observations on the use of selective serotonin-reuptake inhibitors during pregnancy. N Engl J Med 2006;354: 636-8. 9. Ray WA, Griffin MR. Use of Medicaid data for pharmacoepidemiology. Am J Epidemiol 1989;129:837-49. 10. West SL, Savitz DA, Koch G, Strom BL, Guess HA, Hartzema A. Recall accuracy for prescription medications: self-report compared with database information. Am J Epidemiol 1995;142:1103-12. 11. Johnson RE, Vollmer WM. Comparing sources of drug data about the elderly. J Am Geriatr Soc 1991;39:1079-84. 12. Piper JM, Mitchel EF Jr, Snowden M, Hall C, Adams M, Taylor P. Validation of 1989 Tennessee birth certificates using maternal and newborn hospital records. Am J Epidemiol 1993;137:758-68. 13. Cooper WO, Ray WA, Griffin MR. Prenatal prescription of macrolide antibiotics and infantile hypertrophic pyloric stenosis. Obstet Gynecol 2002;100:101-6. 14. Cooper WO, Hernandez-Diaz S, Arbogast PG, et al. Major congenital malformations after first-trimester exposure to ACE inhibitors. N Engl J Med 2006;354:2443-51. 15. Piper JM, Ray WA, Griffin MR, Fought R, Daughtery JR, Mitchel E Jr. Methodological issues in evaluating expanded Medicaid coverage for pregnant women. Am J Epidemiol 1990;132:561-71. 16. Dupont W. Statistical modeling for biomedical researchers: a simple introduction to the analysis of complex data. Cambridge, UK: Cambridge University Press; 2002. 17. Sleath BL, Jackson E, Thomas KC, et al. Racial differences in the use of antidepressants and counseling for depression among homeless women. Community Ment Health J 2006;42:77-85. 18. Rich-Edwards JW, Kleinman K, Abrams A, et al. Sociodemographic predictors of antenatal and postpartum depressive symptoms among women in a medical group practice. J Epidemiol Community Health 2006;60:221-7. 19. Malm H, Klaukka T, Neuvonen PJ. Risks associated with selective serotonin reuptake inhibitors in pregnancy. Obstet Gynecol 2005;106:1289-96. 20. Andrade SE, Raebel MA, Morse AN, et al. Use of prescription medications with a potential for fetal harm among pregnant women. Pharmacoepidemiol Drug Saf 2006;15:546-54. 21. Reefhuis J, Rasmussen SA, Friedman JM. Selective serotonin-reuptake inhibitors and persistent pulmonary hypertension of the newborn. N Engl J Med 2006;354:2188-90. 22. Thomas MR, Waxmonsky JA, Gabow PA, Flanders-McGinnis G, Socherman R, Rost K. Prevalence of psychiatric disorders and costs of care among adult enrollees in a Medicaid HMO. Psychiatr Serv 2005;56:1394-401.
JUNE 2007 American Journal of Obstetrics & Gynecology
544.e4
Research
Obstetrics
23. Ray WA, Gigante J, Mitchel EF Jr, Hickson GB. Perinatal outcomes following implementation of TennCare. JAMA 1998; 279:314-6. 24. US Department of Health and Human Services. A profile of Medicaid, Chartbook 2000. Electronic citation. Available at: http://www. cms.hhs.gov/TheChartSeries/downloads/ 2Tchartbk.pdf.
544.e5
www.AJOG.org 25. Hoyert DL, Mathews TJ, Menacker F, Strobino DM, Guyer B. Annual summary of vital statistics: 2004. Pediatrics 2006;117:168-83. 26. Food and Drug Administration. FDA public health advisory: treatment challenges of depression in pregnancy. Electronic citation. Available at: http://www.fda.gov/cder/drug/advisory/SSRI_ PPHN200607.htm. Accessed July 19, 2006, August 7, 2006.
American Journal of Obstetrics & Gynecology JUNE 2007
27. American College of Obstetricians and Gynecologists. Committee opinion no. 343. Psychosocial risk factors: perinatal screening and intervention. Obstet Gynecol 2006;108: 469-77. 28. Bonari L, Pinto N, Ahn E, Einarson A, Steiner M, Koren G. Perinatal risks of untreated depression during pregnancy. Can J Psychiatry 2004;49:726-35.
www. AJOG.org
OBSTETRICS
Increasing use of antidepressants in pregnancy William O. Cooper, MD, MPH; Mary E. Willy, PhD; Stephen J. Pont, MD; Wayne A. Ray, PhD OBJECTIVE: The purpose of this study was to quantify the rate of exposures to antidepressants during pregnancy in a large cohort of women. STUDY DESIGN: This was a retrospective cohort study of 105,335 pregnancies among women enrolled in Tennessee Medicaid from 1999-2003. Pregnancies were classified according to antidepressant exposures during pregnancy using previously validated computerized pharmacy records linked with birth certificates.
white race (P ⬍ .0001), and education ⬎ 12 years (P ⫽ .008) were significant predictors of antidepressant exposure. The proportion of pregnancies with antidepressant use increased from 5.7% of pregnancies in 1999 to 13.4% of pregnancies in 2003 (p⬍.0001). The increase was mostly accounted for by increases in selective serotonin reuptake inhibitor exposures. CONCLUSION: There is an urgent need for further studies that better
quantify the fetal consequences of exposure to antidepressants.
RESULTS: During the study period, 8.7% of women giving birth had
exposure to any antidepressant; 6.2% had exposure to a selective serotonin reuptake inhibitor. Maternal age ⬎ 25 years (P ⬍ .0001),
Key words: antidepressants, fetal effects, medication exposures, pregnancy
Cite this article as: Cooper WO, Willy ME, Pont SJ, et al. Increasing use of antidepressants in pregnancy. Am J Obstet Gynecol 2007;196;544.e1-544.e5.
I
n recent years, selective serotonin reuptake inhibitors (SSRIs) have been increasingly used to treat major depression.1 However, recent studies have reported an increased risk of adverse neonatal effects related to both early and late fetal exposure to SSRIs, which have in-
From the Departments of Pediatrics (Drs Cooper and Pont) and Preventive Medicine (Dr Ray), Vanderbilt University School of Medicine, Nashville, TN, and the Division of Drug Risk Evaluation, Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD (Dr Willy). Received Aug. 15, 2006; revised Nov. 28, 2006; accepted Jan. 24, 2007. Reprints: William O. Cooper, MD, MPH, Associate Professor of Pediatrics, Department of Pediatrics, Vanderbilt University School of Medicine, AA-0216 MCN, Nashville, TN 37232-2504; [email protected] Funding to conduct the study was provided by the Food and Drug Administration (contract #HHSF223200510010C) and the Centers for Education and Research on Therapeutics, Agency for Healthcare Research and Quality (HS-0384). 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2007.01.033
544.e1
cluded both neurologic and cardiovascular abnormalities.2-6 Given that the prevalence of major depressive disorders in women of reproductive age is thought to be 10-15%, it is possible that large numbers of pregnant women are taking these medications.7 To estimate the potential public health consequences of SSRI use in pregnancy, it is important to know how many fetuses are potentially exposed to SSRIs and other antidepressant medications. At present there is limited information on the use of antidepressants during pregnancy.8 The aims of this study were to quantify the rate of pregnancy related exposures to antidepressants according to drug type and trimester of pregnancy, and to assess temporal trends in their use.
M ATERIALS AND M ETHODS The study was conducted using data from TennCare, Tennessee’s expanded Medicaid program, for which computerized records of filled prescriptions have been shown to be an accurate source of medication data and have high concordance with patient self-report of medication use.9-11 Links to vital records (birth, death, and fetal death certificates) files and medical records permit identifica-
American Journal of Obstetrics & Gynecology JUNE 2007
tion of pregnant women, including estimated conception dates.9,12-14 Vital records and TennCare enrollment files provided information on maternal and infant factors, including maternal age, race (from birth certificate self-report), and county of residence. The date of the last menstrual period was taken from the birth certificate when available (89.3% of all births) or estimated from birth weight using the raceand calendar year-specific distributions of gestational age for births with known last menstrual period (LMP).15 In a previous study using the same databases, the birth certificate LMP was within two weeks of the LMP found in the medical record 94% of the time.14 The first trimester was defined as the last menstrual period through the following 90 days, the second trimester as the next 90 days, and the third trimester as the remainder of the pregnancy. Pregnancies were included in the study if the date of delivery was between January 1, 1999 and December 31, 2003 and the mother was a Tennessee resident with complete information on the birth certificate for maternal age and date of delivery. Pregnancies were further restricted to those in which the mother was enrolled in the TennCare program from
Obstetrics
www.AJOG.org
Research
TABLE 1
Factors predicting any antidepressant and SSRI antidepressant exposures during pregnancy, women enrolled in TennCare, 1999-2003 Any antidepressant use
SSRI use
Characteristic*
Odds ratio
95% confidence interval
Odds ratio
95% confidence interval
Maternal age ⬎ 25 years
2.04
1.94-2.14
1.95
1.84-2.06
White race
4.33
4.02-4.67
4.55
4.17-4.97
Education ⬎ 12 years
1.09
1.02-1.15
1.15
1.08-1.24
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
* Logistic regression models included birth year, maternal age, race, years of education, county of residence, neighborhood income, and parity. Reference groups for the characteristics shown included age ⬍ 25 years (maternal age), black race (race), and education ⬍12 years (education).
30 days prior to the date of the last menstrual period through the date of delivery or fetal death with no gaps in enrollment lasting longer than 7 days. From a total of 235,099 pregnancies identified from birth certificates and 1243 fetal deaths among women enrolled in TennCare at delivery, 105,335 pregnancies met all requirements for inclusion in the cohort. Most exclusions were for mothers who enrolled in TennCare after the LMP. All study pregnancies were classified according to antidepressant use during pregnancy. Prescribed medications were identified from TennCare pharmacy files, which included the date that the prescription was filled and the prescribed number of days of medication supply. TennCare pharmacy files contain information on all outpatient prescriptions which are reimbursed by TennCare, including prescriptions filled in pharmacies, purchased over the internet, or filled in other states. Maternal medication exposure during a given trimester was defined as presence of 1 or more days of supply during the trimester. Thus, if the prescription was filled during 1 trimester (or before the LMP) and the days of supply extended into the next trimester, exposure was considered to have occurred in both trimesters. In the assessment of temporal trends, the calendar year to which all maternal medication use was assigned was that of the date of delivery. Antidepressant medications studied included 4 categories of antidepressants: SSRIs (fluoxetine, sertraline, paroxetine, fluvoxamine, citalopram, escitalopram), serotonin-norepinephrine reuptake inhibitors (venlafaxine,
duloxetine), tricyclic and tetracyclic antidepressants (amitriptyline, desipramine, doxepin, imipramine, nortriptyline, proptriptyline, trimispramine, amoxapine, maprotiline, clomipramine), and other antidepressants (trazodone, bupropion, mirtazapine, phenelzine, tranylcypromine). For each pregnancy, only 1 exposure per category of antidepressants during a given trimester was counted. Exposures to multiple antidepressant categories during a trimester were allowed. Logistic regression was used to identify maternal factors associated with antidepressant use, including maternal age, race, parity, education, residence in a standard metropolitan statistical area, and neighborhood income quintile. Data were analyzed at the individual pregnancy level. To account for correlation due to women with multiple pregnancies during the study period, the robust Huber-White variance estimator was used to provide correct standard errors and confidence intervals of the estimated odds ratios.16 All regressions were performed with SAS version 9.1 (SAS Institute, Cary, NC). The analysis of temporal trends included adjustment for potential changes in maternal characteristics during the study period. To calculate the annual adjusted proportions of antidepressant use during pregnancy, calendar year was added to the model, with 2003 as the reference category. The adjusted odds (O) for each of the other years were obtained by multiplying the unadjusted odds for 2003 by the adjusted odds ratio for each calendar year. This was then converted to a proportion (p) using the identity p ⫽ O/(1⫹O).
Permission to perform the study and waiver of informed consent was obtained from the Vanderbilt University Institutional Review Board, the State of Tennessee Health Department, and the TennCare Bureau.
R ESULTS During the study period, the number of mothers who qualified for the cohort each year ranged from 19,421 to 21,997. The mean age (range of mean age 23.2 ⫾ 5.3 years to 23.3 ⫾ 5.2 years) and proportions of mothers who were prima gravida (proportion for each year ranged from 24.3-25.5%), were black race (28.240.5%), had greater than 12 years of education (14.1-15.1%), resided in an urban county (43.0-45.5%), and who resided in neighbors with the lowest income (23.6-26.8%) did not change materially during the 5 year period. Among the women giving birth during the entire study period, 8.7% had exposure to any antidepressant during pregnancy and 6.2% to an SSRI antidepressant. Maternal age greater than or equal to 25 years (P ⬍ .0001), white race (P ⬍ .0001), and education greater than 12 years (P ⬍ .01) were significant predictors of both any antidepressant and SSRI antidepressant exposure (Table 1). White race was the strongest predictor of pregnancy-related antidepressant exposure with odds ratios of 4.33 (95% confidence interval 4.02-4.67) for any antidepressant exposure and 4.55 (95% confidence interval 4.17-4.97) for SSRI exposure. The most commonly prescribed antidepressants included sertraline, with ex-
JUNE 2007 American Journal of Obstetrics & Gynecology
544.e2
Research
Obstetrics
www.AJOG.org
FIGURE
Percentage of pregnancies with antidepressant exposure 25.00%
Non-SSRI SSRI
20.00%
15.00%
10.00%
5.00%
0.00% 1999
2000
2001
2002
2003
Proportion of pregnancies with antidepressant exposure among women giving birth between 1999-2003 who were enrolled throughout pregnancy in the TennCare program.
Use of non-SSRI antidepressants increased as well, but to a lesser extent than that for the SSRI antidepressants. For women giving birth in 2003, 10.0% had first trimester antidepressant exposure and 6.4% and 5.9% had second and third trimester exposure, respectively (Table 2). SSRI exposure occurred most commonly, with the highest rate of exposures occurring in the first trimester. There was less frequent use during pregnancy of serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants, and other antidepressants, but, as was the case for the other antidepressants, such exposure was most frequent during the first trimester (Table 2).
C OMMENT posure occurring in 2.1% of pregnancies, paroxetine, with exposure occurring in 1.9% of pregnancies, and fluoxetine, with exposure occurring in 1.9% of pregnancies. Other commonly used antidepressants included amitriptyline (1.0% of pregnancies), trazodone (0.9% of pregnancies), and citalopram (0.8% of pregnancies). After adjusting for maternal age, race, parity, education, residence in a standard metropolitan statistical area, and neighborhood income quintile, the proportion of pregnancies with antidepressant use increased from 5.7% of pregnancies in 1999 to 13.4% of pregnancies by 2003, a more than 2-fold increase (Figure). This was largely due to increased use of SSRIs, with proportions increasing from 2.9% of pregnancies in 1999 to 10.2% of pregnancies in 2003.
In this study of antidepressant use during pregnancy in a large cohort of women enrolled in the TennCare program, we found a marked temporal trend of increasing use during the 5-year period 1999 through 2003. By 2003, more than 13% of pregnancies had an antidepressant exposure. SSRIs accounted for most of the increase in antidepressant exposures, with use of these drugs by more than 10% of pregnant women by 2003. Use of other antidepressants during pregnancy increased at a more moderate rate during the study period. Mothers most likely to have pregnancy-related antidepressant exposures included older women, white women, and women with more than 12 years of education. In other studies of medication use, white race has also been associated with greater use of antidepres-
sants.17 Interestingly, some populationbased studies have suggested that maternal depressive symptoms occur more frequently among racial and ethnic minorities.18 While younger maternal age has been associated with more frequent depressive symptoms among pregnant women, antidepressant use in the present study was more common among older mothers.18 Further research to understand the relationship between these factors and use of antidepressants is needed. The prevalence of antidepressant exposure in our cohort was higher than that described previously. Most of the prior population-based studies of antidepressant use in pregnancy include antidepressants in analyses of all other medications or of psychotropic medications as a group.2,19-21 Studies in European countries using data through 2001 identified use of antidepressants in ⬍ 1-1.4% of pregnancies.2,19 Reports from the United States, including a study from a large HMO research network by Andrade et al, noted slightly higher frequency of use—from 2.8-3.6%.20,21 The most probable explanation for the higher proportions of pregnancies with antidepressant use in the current study is that the data are from more recent years. Data for the first 2 years of our study were comparable to that described by Andrade et al. In addition, the current study included large numbers of Medicaid enrollees, who are known to have higher rates of chronic health conditions, including mental illness.22 Utilization of Medicaid pharmacy files minimizes concerns of selection bias, re-
TABLE 2
Use of antidepressants during pregnancy for 2003, according to drug class and trimester % of Pregnancies Medication exposure
First trimester
Second trimester
3rd trimester
Any exposure
Any antidepressant
10.0
6.4
5.9
13.3
SSRI antidepressants
7.2
5.0
4.8
10.2
SNRI antidepressants
1.3
0.5
0.4
1.4
Tricyclic antidepressants
1.1
0.4
0.3
1.3
Other antidepressants
2.4
1.2
0.9
3.1
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
SSRI, selective serotonin reuptake inhibitors; SNRI, serotonin norepinephrine reuptake inhibitors.
544.e3
American Journal of Obstetrics & Gynecology JUNE 2007
Obstetrics
www.AJOG.org call bias, and differential misclassification that may be associated with retrospective assessment of medication exposures. However, use of these files imposes limitations of their own. The prescription files recorded that a woman filled an antidepressant prescription during pregnancy, but did not indicate whether or not the prescription was taken as prescribed. It is possible that women may have stopped medication use after pregnancy was diagnosed, but in many cases, women filled antidepressant prescriptions during the second and third trimesters of pregnancy. Additionally, Medicaid pharmacy files were limited to outpatient prescription medications. Thus, for the small number of women who obtained prescriptions outside of the Medicaid program, exposures would not be captured. Because of the income requirements for participation in the Medicaid program, it is not anticipated that many persons paid for prescriptions out-of-pocket. Our study included only pregnant women enrolled in TennCare; thus, the results may not be generalizable to other populations. The study population included disproportionate numbers of mothers ⬍ 25 years of age, teenage mothers, mothers with ⬍ 12 years of education, and mothers living in urban areas.23 Nevertheless, Medicaid enrollees include a large proportion of all pregnant women and are thus important in their own right. In recent years in Tennessee, approximately 50% of women giving birth were enrolled in TennCare at the time of delivery; nationally, over one third of all births were covered by Medicaid.23,24 Utilizing vital statistics from the year 2003 and applying the 13.3% proportion of pregnancies with antidepressant exposure from that year of our study suggests that among Medicaid enrollees in the US, there were approximately 180,000 fetuses exposed to antidepressants each year.25 Exposure to antidepressants during both early and late pregnancy has been associated with adverse fetal outcomes, including cardiac malformations and pulmonary hypertension.5,6 There have been 2 Food and Drug Administration public health advisories noting these po-
tential risks.5,26 Prescribing of medications to women during pregnancy requires balancing the benefits to the mother against the risks to the developing fetus. This cannot be done unless there is accurate information on the medication’s fetal effects. Affective disorders are common among women of childbearing age and it is recommended that all pregnant women should be screened and treated for depression.27 Given that untreated depression may have serious clinical and social perinatal consequences, it would be difficult to avoid all use of antidepressant medications during pregnancy.26,28 Although the available data suggest there are adverse effects associated with fetal exposure to some antidepressants, these data are limited, particularly with regard to the effects of individual drugs.5,24 Thus, there is an urgent need for further studies that better quantify the fetal consequences of exposure to antidepressants. f ACKNOWLEDGMENTS We gratefully acknowledge the Tennessee Department of Health and the TennCare Bureau, who provided the data for the study.
REFERENCES 1. Pirraglia PA, Stafford RS, Singer DE. Trends in prescribing of selective serotonin reuptake inhibitors and other newer antidepressant agents in adult primary care. Prim Care Companion. J Clin Psychiatry 2003;5:153-7. 2. Kallen B. Fluoxetine use in early pregnancy. Birth Defects Res B Dev Reprod Toxicol. 2004;71:395-6. 3. Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1996;335:1010-5. 4. Moses-Kolko EL, Bogen D, Perel J et al. Neonatal signs after late in utero exposure to serotonin reuptake inhibitors: literature review and implications for clinical applications. JAMA 2005;293:2372-83. 5. Food and Drug Administration. FDA Public health advisory: paroxetine. Electronic citation. Available at: http://www.fda.gov/cder/drug/ advisory/paroxetine200512.htm. Accessed March 2, 2007. 6. Chambers CD, Hernandez-Diaz S, Van Marter LJ, et al. Selective serotonin-reuptake inhibitors and risk of persistent pulmonary hypertension of the newborn. N Engl J Med 2006;354:579-87. 7. Bennett HA, Einarson A, Taddio A, Koren G, Einarson TR. Prevalence of depression during
Research
pregnancy: systematic review. Obstet Gynecol. 2004;103:698-709. 8. Mills JL. Depressing observations on the use of selective serotonin-reuptake inhibitors during pregnancy. N Engl J Med 2006;354: 636-8. 9. Ray WA, Griffin MR. Use of Medicaid data for pharmacoepidemiology. Am J Epidemiol 1989;129:837-49. 10. West SL, Savitz DA, Koch G, Strom BL, Guess HA, Hartzema A. Recall accuracy for prescription medications: self-report compared with database information. Am J Epidemiol 1995;142:1103-12. 11. Johnson RE, Vollmer WM. Comparing sources of drug data about the elderly. J Am Geriatr Soc 1991;39:1079-84. 12. Piper JM, Mitchel EF Jr, Snowden M, Hall C, Adams M, Taylor P. Validation of 1989 Tennessee birth certificates using maternal and newborn hospital records. Am J Epidemiol 1993;137:758-68. 13. Cooper WO, Ray WA, Griffin MR. Prenatal prescription of macrolide antibiotics and infantile hypertrophic pyloric stenosis. Obstet Gynecol 2002;100:101-6. 14. Cooper WO, Hernandez-Diaz S, Arbogast PG, et al. Major congenital malformations after first-trimester exposure to ACE inhibitors. N Engl J Med 2006;354:2443-51. 15. Piper JM, Ray WA, Griffin MR, Fought R, Daughtery JR, Mitchel E Jr. Methodological issues in evaluating expanded Medicaid coverage for pregnant women. Am J Epidemiol 1990;132:561-71. 16. Dupont W. Statistical modeling for biomedical researchers: a simple introduction to the analysis of complex data. Cambridge, UK: Cambridge University Press; 2002. 17. Sleath BL, Jackson E, Thomas KC, et al. Racial differences in the use of antidepressants and counseling for depression among homeless women. Community Ment Health J 2006;42:77-85. 18. Rich-Edwards JW, Kleinman K, Abrams A, et al. Sociodemographic predictors of antenatal and postpartum depressive symptoms among women in a medical group practice. J Epidemiol Community Health 2006;60:221-7. 19. Malm H, Klaukka T, Neuvonen PJ. Risks associated with selective serotonin reuptake inhibitors in pregnancy. Obstet Gynecol 2005;106:1289-96. 20. Andrade SE, Raebel MA, Morse AN, et al. Use of prescription medications with a potential for fetal harm among pregnant women. Pharmacoepidemiol Drug Saf 2006;15:546-54. 21. Reefhuis J, Rasmussen SA, Friedman JM. Selective serotonin-reuptake inhibitors and persistent pulmonary hypertension of the newborn. N Engl J Med 2006;354:2188-90. 22. Thomas MR, Waxmonsky JA, Gabow PA, Flanders-McGinnis G, Socherman R, Rost K. Prevalence of psychiatric disorders and costs of care among adult enrollees in a Medicaid HMO. Psychiatr Serv 2005;56:1394-401.
JUNE 2007 American Journal of Obstetrics & Gynecology
544.e4
Research
Obstetrics
23. Ray WA, Gigante J, Mitchel EF Jr, Hickson GB. Perinatal outcomes following implementation of TennCare. JAMA 1998; 279:314-6. 24. US Department of Health and Human Services. A profile of Medicaid, Chartbook 2000. Electronic citation. Available at: http://www. cms.hhs.gov/TheChartSeries/downloads/ 2Tchartbk.pdf.
544.e5
www.AJOG.org 25. Hoyert DL, Mathews TJ, Menacker F, Strobino DM, Guyer B. Annual summary of vital statistics: 2004. Pediatrics 2006;117:168-83. 26. Food and Drug Administration. FDA public health advisory: treatment challenges of depression in pregnancy. Electronic citation. Available at: http://www.fda.gov/cder/drug/advisory/SSRI_ PPHN200607.htm. Accessed July 19, 2006, August 7, 2006.
American Journal of Obstetrics & Gynecology JUNE 2007
27. American College of Obstetricians and Gynecologists. Committee opinion no. 343. Psychosocial risk factors: perinatal screening and intervention. Obstet Gynecol 2006;108: 469-77. 28. Bonari L, Pinto N, Ahn E, Einarson A, Steiner M, Koren G. Perinatal risks of untreated depression during pregnancy. Can J Psychiatry 2004;49:726-35.