- Email: [email protected]
The timing of norplant insertion and postpartum depression in teenagers
JOURNAL OF ADOLESCENT HEALTH 2000;26:408– 413
ORIGINAL ARTICLE
The Timing of Norplant Insertion and Postpartum Depression in Teenagers CATHERINE STEVENS-SIMON, M.D., LISA KELLY, C.H.A., P.A., AND JOAN WALLIS, C.N.M.
Background: This study tested the hypothesis that teenagers who have Norplant inserted during the puerperium report more depressive symptoms during the first postpartum year than their peers who do not receive Norplant. Methods: We studied the prevalence of depressive symptoms in a group of 212 mothers aged 19 years less, in relation to the timing of Norplant insertion. The participants were divided into 3 groups: 100 (47%) had Norplant inserted during the puerperium (early Norplant users); 72 (34%) had Norplant inserted during the next 10 months (late Norplant users); and 40 (19%) used other contraceptives (40% oral contraceptives; 17% condoms; 43% nothing). Depressive symptoms were measured with the Center for Epidemiologic Studies - Depression Scale. Postpartum depression was defined as a scale score >16, 6 –12 months after Norplant insertion or delivery. Variables examined as potential confounders were identified a priori from a review of the literature and controlled for in analysis of variance. Results: At delivery, members of the 3 contraceptive groups did not differ significantly with regard to age, race, parity, educational, marital, or socioeconomic status. Late Norplant users were, however, more apt to have new boyfriends (p ⴝ .03), to rate the support they received from the baby’s father as poor (p ⴝ .004), and experience depression prior to Norplant insertion (p ⴝ .02). Contrary to the study hypothesis, late rather than early Norplant insertion was associated with postpartum depression.
From the Department of Pediatrics, Division of Adolescent Medicine, University of Colorado Health Science Center (C.S.-S.), Denver, Colorado, Department of Pediatrics, University of Colorado Health Science Center (L.K.), Denver, Colorado, and Department of Obstetrics & Gynecology, University of Colorado Health Science Center (J.W.), Denver, Colorado Address reprint requests to: Catherine Stevens-Simon, M.D., Department of Pediatrics - Division of Adolescent Medicine, University of Colorado Health Science Center - The Children’s Hospital, 1056 East 19th Street, Denver, CO 80218. Manuscript accepted August 3, 1999. 1054-139X/00/$–see front matter PII S1054-139X(99)00091-9
Multivariate analyses identified 3 independent predictors of the severity of depressive symptoms at follow-up (depression prior to Norplant insertion, a new boyfriend at delivery, and late Norplant insertion); R2 ⴝ 41.3%. Conclusions: Contrary to the study hypothesis, puerperal Norplant insertion did not exacerbate postpartum depression. Delaying Norplant insertion may increase the risk of depression during the first postpartum year, particularly in teenagers with other psychosocial risk factors. © Society for Adolescent Medicine, 2000
KEY WORDS: Adolescent Pregnancy Postpartum Depression Norplant
Repeat adolescent pregnancies pose a serious public health problem in this country (1,2). Even among teenagers who are enrolled in special, comprehensive, adolescent-oriented, maternity programs, the rate of preterm and low birthweight deliveries increases and the likelihood of completing high school, having a job, and being self-supporting decreases with each additional pregnancy at this age. The frequency and rapidity with which repeat pregnancies occur, even in these intensive reproductive health care settings, makes long-acting contraceptive agents like Norplant a particularly attractive postpartum intervention strategy. Although postpartum Norplant insertion is the only intervention that has been consistently associated with a significant reduction in the repeat adolescent pregnancy rate in the United States (3–5), there are concerns about the added risk of Norplant side effects during the puerperium. Specifically, the tendency for childbirth to trigger depression in young, single women living in
© Society for Adolescent Medicine, 2000 Published by Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010
June 2000
stressful, unsupportive environments (6 –10) may increase vulnerability to the moodiness and irritability that have been associated with Norplant use (11). This study of the prevalence of depressive symptoms in relation to the timing of postpartum Norplant insertion was designed to test the hypothesis that adolescent mothers who have Norplant inserted during the puerperium will report more symptoms of depression during the first postpartum year than their peers who do not receive Norplant.
Methods Patients The data were collected as part of a large study of contraceptive use and repeat pregnancy among patients in the Colorado Adolescent Maternity Program (CAMP). CAMP is a comprehensive, multidisciplinary, adolescent-oriented prenatal, delivery, postpartum, and infant care program in Denver, Colorado (11–14). The original study population consisted of a multiracial (50% white, 26% black, 22% Hispanic and 2% other) group of 354, poor (94% Medicaid-dependent), predominantly primiparous (84%), 13- to 19year-olds (mean ⫾ sd ⫽ 17.0 ⫾ 1.2 years). Participants were enrolled consecutively at delivery during the 1992 and 1993 calendar years (11). We chose this period for study because Norplant was extremely popular during this time and because it antedated major concerns about the Norplant removal process and the immense popularity of the other long-acting, progesterone-only contraceptive agent, Depo-Provera. Teenagers who used DepoProvera following delivery (n ⫽ 28) were excluded from this analysis because, like Norplant, DepoProvera is a long-acting, progesterone-only contraceptive agent, the use of which has been associated with moodiness, irritability, and depression (15). We also excluded teenagers who had no quantitative assessment of depressive symptoms (n ⫽ 49) and those lost to follow-up (n ⫽ 65). Analyses for this study were based on the remaining 212 young mothers. The study sample (n ⫽ 212) did not differ significantly from those excluded with regard to any of the social demographic variables related to the risk of postpartum depression we studied.
Data Collection and Variables Information about the study participants’ medical, social, psychiatric, and reproductive histories was
NORPLANT AND POSTPARTUM DEPRESSION
409
collected during gestation, at delivery, and 6 –12 months after Norplant insertion (or delivery in the case of those who did not use Norplant) with a self-administered questionnaire. Additional medical data were obtained by reviewing the participants’ medical records. Independent variable. “Norplant-user groups”: the 212 study participants were divided into three contraceptive use groups: 1) 100 (47%) “early Norplant users” who had Norplant inserted during the first two postpartum months, (on average at 5.5 ⫾ 1.7 weeks postpartum); 2) 72 (34%) “late Norplant users” who had Norplant inserted during the next 10 months (on average at 19.1 ⫾ 10.9 weeks postpartum); and 3) 40 (19%) “other contraceptive users” of whom 40% used an oral contraceptive, 17% used a barrier method, and 43% used no method. The distinction between “early” vs. “late” Norplant users, although somewhat arbitrary, was chosen to coincide with the puerperium, the period during which studies suggest new mothers are most apt to suffer from the moodiness and emotional lability (6 –10). In addition, most women who do not breastfeed resume ovarian cycling during the puerperium (10). Dependent variable. “Depressive symptoms at follow-up”: symptoms of depression were quantified with the Center for Epidemiologic Studies - Depression Scale (CES-D Scale) (16). The 20 CES-D Scale test items are scored on a 4-point Likert scale and summed to create a total score ranging from 0 to 60. Consistent with prior studies, depressive symptoms were dichotomized as “not depressed” (0 –15) and depressed (16 – 60) for the categorical analysis (7). Since the scale was administered 6 –12 months after Norplant insertion or 6 –12 months after delivery to study participants who did not use Norplant, the mean postpartum interval at follow-up was significantly different in the three Norplant use groups: 50.2 ⫾ 11.6, 65.0 ⫾ 16.8, and 43.5 ⫾ 12.3 weeks, respectively for early-, late-, and nonusers; (p ⬍ .0001). Center for Epidemiologic Studies - Depression Scale scores did not vary significantly in relation to postpartum week at follow-up (r ⫽ .06; p ⫽ .37). Intervening variables. The specific variables examined as potential confounders of the relationship between contraceptive group and depressive symptoms were identified a priori from a review of the literature (6 –10). They included the following widely accepted medical, psychologic, and social risk factors
410
STEVENS-SIMON ET AL.
for postpartum depression: (1) adverse pregnancy outcome (e.g., newborn sick, malformed, not the desired sex and/or multiple maternal complications during gestation or delivery); (2) multiparity; (3) rapid repeat conception; (4) antenatal experience described as extremely difficult or stressful; (5) a history of depression prior to conception; (6) antecedent depression (e.g., a CES-D Scale score ⱖ 16 at any time prior to Norplant insertion); (7) dissatisfaction with the support provided by the father of the child and/or family members (the teenagers were asked to rate the quality of the support they received during gestation from the baby’s father, their mother, and other important people in their support network on a simple scale: “excessive/intrusive”; “about what I needed”; “poor”; “none”); (8) minority race or ethnicity; (9) under-education, school failure, school dropout; (10) poverty; (11) unmarried status; (12) past or current physical or sexual abuse; and (13) involvement in socially problematic behaviors such as substance abuse, fighting, or delinquency. The study was approved by the Institutional Review Board at the University of Colorado Health Sciences Center.
Data Analysis Summary statistics were used to describe the study population. Comparisons among contraceptive groups were conducted with analysis of variance (ANOVA) when the outcome variable was continuous and Chi-square analysis when the outcome variable was categorical. To simplify the model and its application in clinical practice, the intervening variables were dichotomized (“present” or “absent”). Intervening variables for which there were significant contraceptive group differences at the bivariate level were allowed to enter the multivariate model, one at a time on the basis of the statistical significance of their association with depressive symptoms at follow-up. Analyses were performed with SPSS/PC⫹ (17).
Results Members of the three contraceptive groups did not differ significantly with regard to the sociodemographic characteristics we studied. Overall, the study sample was representative of the multiracial, predominantly Medicaid-dependent, unmarried, population of urban teenagers CAMP serves. Nearly half of the study participants had dropped out of school
JOURNAL OF ADOLESCENT HEALTH Vol. 26, No. 6
Table 1. Psychosocial Characteristics (N ⫽ 212) Norplant Insertion Group N (%) Characteristic
Early
Late
None
Number Significant other Father-of-child New boyfriend* None Poor father support** Antecedent depression†,***
100
72
40
65 (65) 8 (8) 27 (27) 42 (45) 49 (49)
34 (47) 16 (22) 22 (31) 42 (71) 39 (54)
21 (55) 5 (13) 14 (35) 19 (48) 11 (28)
* p ⫽ .03. ** p ⫽ .004 (missing data: N ⫽ 193 (94;40;59)). *** p ⫽ .02. † A CESD score ⱖ 16 at any time prior to Norplant insertion.
or were failing. Almost all participants (92%) had a teen parent as a close friend and most (76%) had a relative who was or had been a teen parent. The majority (80%) of the young mothers was primiparas who lived in supportive social environments. Over half lived with at least one biologic parent. Most felt they had received about the right amount of support from family members (90%) and peers (76%) during gestation and almost everyone (96%) had concrete plans for assistance with childcare. Past or current physical and/or sexual abuse and involvement in socially problematic behaviors such as substance abuse, fighting, or delinquency was reported infrequently. However, no objective measure of these events or behaviors was available. Because of the low quality of self-reports concerning abuse and involvement in risk behaviors during gestation and the bias that ensues from adjustment for a confounder that has been measured with error, these variables were not included in the analysis. Pregnancy outcomes were also similar in the three contraceptive groups. Most of the young women had uncomplicated pregnancies with good outcomes. However, 37% of the teenagers found the antepartum experience harder and more stressful than anticipated; 12% of the newborns were admitted to the neonatal intensive care nursery and 17% were not the desired sex. Importantly, none of the teenagers who used Norplant became pregnant again during the follow-up period compared to 13% of their peers who used other contraceptive methods. Table 1 describes some potentially important group differences in these young women’s relationships with their significant others and the prevalence of depression at baseline. Teenagers who had late Norplant insertions were most apt to have new boyfriends and to rate the support they received
June 2000
NORPLANT AND POSTPARTUM DEPRESSION
Table 2. Depressive Symptoms at Follow-upa (N ⫽ 212) Norplant Insertion Group Characteristic Number CESD Score (mean ⫾ SD)* Depressed [N(%)]** Became Depressed [N(%)]b,***
Early
Late
None
100 13.2 ⫾ 10.9
72 18.0 ⫾ 13.4
40 11.5 ⫾ 7.7
32 (32) 6 (6)
35 (49) 13 (18)#
10 (25) 5 (13)
* p ⫽ .004. ** p ⫽ .02. *** p ⫽ .05. # p ⫽ .02 for early vs late Norplant users. a Follow-up occurred 6 to 12 months after Norplant insertion or 6-to-12 months after delivery (for non-users); the mean postpartum interval at follow-up was: 50.2 ⫾ 11.6, 65.0 ⫾ 16.8, and 43.5 ⫾ 12.3 weeks, respectively (p ⬍.0001). b Antecedent CESD score ⬍16 and CESD score ⱖ16 at followup.
from their baby’s father during pregnancy as poor or nonexistent. There were no statistically significant contraceptive group differences in reported suicide attempts prior to conception (19%) or prolonged periods (ⱖ 5 consecutive days) of sadness (19%) and/or suicide attempts (5%) during gestation. Nevertheless, Norplant users were more likely to have experienced antecedent depression, that is to have scored in the depressed range (ⱖ 16) on the CES-D Scale at some time prior to Norplant insertion. Complaints of dysphoria and moodiness were common during the follow-up period. Center for Epidemiologic Studies—Depression Scale scores ranged from 0 to 57 (mean ⫾ sd: 14.5 ⫾ 11.5) and 77 (36%) of the 212 study participants had CES-D Scale scores indicative of depression. However, contrary to the study hypothesis, the data presented in Table 2 show that late, rather than early, postpartum Norplant insertion was associated with an increase in the severity of depressive symptoms during the follow-up period. The mean CES-D Scale score and prevalence of depression was significantly higher among late Norplant users than among members of the other two contraceptive use groups. Within the study group as a whole, CES-D Scale scores rose in tandem with postpartum week at Norplant insertion (r ⫽ .2; p ⫽ .01). However, among late Norplant users, there was no relationship between the timing of Norplant insertion and the duration of Norplant use and either the mean CES-D Scale score or the prevalence of depression. Center for Epidemiologic Studies - Depression Scale scores at follow-up exceeded antecedent CES-D Scale scores, particularly among late Norplant users,
411
Table 3. Predictors of Depressive Symptoms at Followup Characteristic
F (P)
Main effects* Antecedent depression** New boyfriend Late Norplant insertion
24.8 (⬍.0001) 7.0 (.009) 6.1 (.01)
Two-way interactions Norplant by depression Norplant by boyfriend
5.4 (.02) 4.5 (.04)
* Degrees of freedom (1,210); no Significant 3-way Interactions. ** A CESD score ⱖ16 during gestation or prior to Norplant insertion. R2 ⫽ .41.
[mean increase in CES-D Scale score: 5 ⫾ 13 points compared to 2 ⫾ 10 and 1 ⫾ 7 points, respectively among early and non-Norplant users (p ⫽ .04)]. The data presented in Table 2 show that early Norplant users were the least likely, and late Norplant users the most likely to become depressed at follow-up (p ⫽ .02). The prevalence of other side effects was unrelated to timing of Norplant insertion. Multivariate analyses controlling for preexisting contraceptive group differences in characteristics with a hypothesized relationship to postpartum depression (Table 1) revealed that antecedent depression was the most powerful predictor of the severity of depressive symptomatology at follow-up (R2 ⫽ .33). Other factors that predicted depression at follow-up are shown in Table 3 in order of decreasing significance. Late Norplant insertion remained a statistically significant, independent predictor of the severity of depressive symptoms and interacted with the other two variables in the model such that teenagers who had been depressed prior to Norplant insertion and teenagers who had new boyfriends at delivery experienced the most severe depressive symptoms following Norplant insertion (mean CES-D Scale scores ⫽ 19.6 and 26.6, respectively). Collectively, these three characteristics accounted for 41.3% of the variance in the severity of depressive symptomatology at follow-up.
Discussion The puerperium is a time of major psychologic upheaval during which women tend to be particularly emotionally labile (8,9). Many women experience a brief period of excessive irritability and moodiness, known as the “maternity blues”, and the small proportion of these mildly dysphoric women who
412
STEVENS-SIMON ET AL.
enter a more protracted period of melancholy are given the diagnosis of “postpartum depression” or “postpartum psychosis”, depending on the severity of their symptoms (8 –10). The incidence of these serious forms of postpartum psychopathology is extremely variable, but the literature suggests that teenagers are more vulnerable to postpartum depression than adults (6 –10). Thus, it was reassuring that the results of this study did not support the study hypothesis; puerperal Norplant insertion did not exacerbate postpartum depression in this population of poor, young, unmarried mothers. The significance of our finding that late Norplant insertion was associated with an increase in the severity of depressive symptoms during the follow-up period is more difficult to interpret. The association does not appear to reflect group differences in the timing of the postpartum follow-up, as there was no correlation between postpartum week at follow-up and CES-D Scale score. The association does, however, appear to be partially psychosocial in origin, as the effect of late Norplant insertion on the severity of depressive symptoms was moderated by controlling for the presence of antecedent depression and/or the acquisition of a new boyfriend prior to delivery. This finding is consistent with the results of most other studies and the notion that postpartum depression is a multifactorial problem, resulting from a complex interaction between biologic, psychologic, and social factors (6 –10). Within the context of this model, the single factor that has been most consistently and strongly related to the development of postpartum depression is social support (6 –10). High satisfaction with support, particularly from the father of the child and, in the case of teenagers, their own mother, decreases the likelihood of depressive symptoms (6). Conversely, conflict with, and/or rejection by, these important persons has been consistently associated with an increased likelihood of depression (6,7). A personal and/or family history of depression and the presence of anxiety and/or depression during pregnancy have also been associated with an increased risk of postpartum depression in several studies (6 –10). It is still unclear precisely how and why ovarian steroids contribute to the mental and emotional instability of some new mothers. The increased emotional lability that is characteristic of the puerperium tends to coincide temporally with rapid changes in serum levels of gonadal hormone that are capable of altering neuronal activity and neurotransmitter levels in portions of the brain that affect mood. Thus, there has been a tendency to draw a causal associa-
JOURNAL OF ADOLESCENT HEALTH Vol. 26, No. 6
tion between the rapid postpartum withdrawal of these hormones, particularly progesterone, and the development of postpartum depression (8,10). Indeed, women who develop maternity blues have been found to experience more precipitous drops in serum progesterone levels from the antenatal to the postpartum period and there have been scattered, anecdotal, reports that treatment with progesterone ameliorates the symptoms of postpartum depression (10). Like all studies, this one has its limitations. First, since all of the participants were enrolled in a comprehensive adolescent-oriented maternity program, unknown differences between pregnant and parenting teenagers who are cared for in these settings and the larger population of pregnant and parenting teenagers means that it is necessary to use caution in generalizing our findings. Second, even though we used multivariate analyses to control for antecedent “Norplant-use group” differences regarding numerous widely recognized risk factors for postpartum depression, Norplant use was not a random event in the population we studied. Thus, there may have been other, unanticipated differences between early, late, and non-Norplant users that confounded our analysis and biased our findings regarding the association between the timing of Norplant insertion and the severity of postpartum depression. Nevertheless, since we found that early Norplant users were significantly less apt to become depressed at follow-up, it is tempting to speculate about the therapeutic role of puerperal Norplant insertion. The authors wish to thank the CAMP staff and patients for their help with data collection. Supported in part by: Grant #APH000166 –5; Office of Adolescent Pregnancy, NIH; a grant from the Colorado Trust, Denver, CO and Grant #5 MO1 RR00069 General Clinical Research Centers Program, National Center for Research Resources, NIH. Presented in part: The Society for Pediatric Research Annual Meeting New Orleans, May 1998.
References 1. Stevens-Simon C, White M. Adolescent pregnancy. Pediatr Annals 1991;20:322–31. 2. Brown SS, Eisenberg L. The Best Intentions. Washington DC: National Academy Press, 1995. 3. Polaneczky M, Forke C, Armstrong J. Norplant vs oral contraceptive use by adolescent mothers: Long term outcome. J Adol Health 1996;18:120 (Abstract). 4. O’Dell CM, Forke CM, Polaneczky MM, et al. Depot Medroxyprogesterone Acetate or oral contraception in postpartum adolescents. Obstet & Gynecol 1998;91:609 –14. 5. Ricketts SA. Repeat fertility and contraceptive implant use among Medicaid recipients in Colorado. Fam Plann Perspect 1996;28:278 – 80.
June 2000
6. Barnet B, Joffe A, Duggan A, et al. Depressive symptoms, stress, and social support in pregnant and postpartum adolescents. Arch Pediatr Adoles Med 1996;150:64 –9. 7. Deal LW, Holt VL. Young maternal age and depressive symptoms: Results from the 1988 National Maternal and Infant Health Survey. Am J Pub Health 1998;88:266 –70. 8. Stowe ZN, Nemeroff CB. Women at risk for postpartumonset major depression. Am J Obstet Gynecol 1995;173:639 – 45. 9. Wolman WL, Chalmers B, Hofmeyr J, Nikodem V. Postpartum depression and companionship in the clinical birth environment: A randomized, controlled study. Am J Obstet Gynecol 1993;168:1388 –93. 10. Harris B, Lovett L, Newcombe RG, Read GF, Walker R, Riad-Fahmy D. Maternity blues and major endocrine changes: Cardiff puerperal mood and hormone study II. Br Med J 1994;308:949 –53. 11. Stevens-Simon C, Kelly LS, Singer D, et al. Correlates and consequences of early removal of Levonorgestral implants
NORPLANT AND POSTPARTUM DEPRESSION
12. 13.
14.
15.
16.
17.
413
among adolescent mothers. Arch Pediatr Adolesc Med 1998; 152:893– 8. Stevens-Simon C, Wallis J, Allan-Davis J. Which teen mothers choose Norplant? J Adol Health 1995;16:350 –3. Stevens-Simon C, Wallis J, Allan-Davis J. Antecedents of preterm delivery among adolescents: Relationship to type of prenatal care. J Mat Fetal Med 1995;4:186 –93. Stevens-Simon C, Kelly LS, Singer D. Reasons pregnant adolescents give for not using contraceptives prior to conception: Implications for recidivism. Arch Ped Adol Med 1996;150: 1037– 43. Sangi-Haghpeykar H, Poindexter III AN, Bateman L, Ditmore JR. Experiences of injectable contraceptive users in an urban setting. Obstet Gynecol 1996;88:227–33. Radloff L. The CES-D scale: A self-report depression scale for research in the general population. J Applied Psychological Measures 1977;1:385– 401. Norusis MJ. Statistical Package for the Social Sciences. Chicago, IL: SPSS/PC⫹, 1990.
ORIGINAL ARTICLE
The Timing of Norplant Insertion and Postpartum Depression in Teenagers CATHERINE STEVENS-SIMON, M.D., LISA KELLY, C.H.A., P.A., AND JOAN WALLIS, C.N.M.
Background: This study tested the hypothesis that teenagers who have Norplant inserted during the puerperium report more depressive symptoms during the first postpartum year than their peers who do not receive Norplant. Methods: We studied the prevalence of depressive symptoms in a group of 212 mothers aged 19 years less, in relation to the timing of Norplant insertion. The participants were divided into 3 groups: 100 (47%) had Norplant inserted during the puerperium (early Norplant users); 72 (34%) had Norplant inserted during the next 10 months (late Norplant users); and 40 (19%) used other contraceptives (40% oral contraceptives; 17% condoms; 43% nothing). Depressive symptoms were measured with the Center for Epidemiologic Studies - Depression Scale. Postpartum depression was defined as a scale score >16, 6 –12 months after Norplant insertion or delivery. Variables examined as potential confounders were identified a priori from a review of the literature and controlled for in analysis of variance. Results: At delivery, members of the 3 contraceptive groups did not differ significantly with regard to age, race, parity, educational, marital, or socioeconomic status. Late Norplant users were, however, more apt to have new boyfriends (p ⴝ .03), to rate the support they received from the baby’s father as poor (p ⴝ .004), and experience depression prior to Norplant insertion (p ⴝ .02). Contrary to the study hypothesis, late rather than early Norplant insertion was associated with postpartum depression.
From the Department of Pediatrics, Division of Adolescent Medicine, University of Colorado Health Science Center (C.S.-S.), Denver, Colorado, Department of Pediatrics, University of Colorado Health Science Center (L.K.), Denver, Colorado, and Department of Obstetrics & Gynecology, University of Colorado Health Science Center (J.W.), Denver, Colorado Address reprint requests to: Catherine Stevens-Simon, M.D., Department of Pediatrics - Division of Adolescent Medicine, University of Colorado Health Science Center - The Children’s Hospital, 1056 East 19th Street, Denver, CO 80218. Manuscript accepted August 3, 1999. 1054-139X/00/$–see front matter PII S1054-139X(99)00091-9
Multivariate analyses identified 3 independent predictors of the severity of depressive symptoms at follow-up (depression prior to Norplant insertion, a new boyfriend at delivery, and late Norplant insertion); R2 ⴝ 41.3%. Conclusions: Contrary to the study hypothesis, puerperal Norplant insertion did not exacerbate postpartum depression. Delaying Norplant insertion may increase the risk of depression during the first postpartum year, particularly in teenagers with other psychosocial risk factors. © Society for Adolescent Medicine, 2000
KEY WORDS: Adolescent Pregnancy Postpartum Depression Norplant
Repeat adolescent pregnancies pose a serious public health problem in this country (1,2). Even among teenagers who are enrolled in special, comprehensive, adolescent-oriented, maternity programs, the rate of preterm and low birthweight deliveries increases and the likelihood of completing high school, having a job, and being self-supporting decreases with each additional pregnancy at this age. The frequency and rapidity with which repeat pregnancies occur, even in these intensive reproductive health care settings, makes long-acting contraceptive agents like Norplant a particularly attractive postpartum intervention strategy. Although postpartum Norplant insertion is the only intervention that has been consistently associated with a significant reduction in the repeat adolescent pregnancy rate in the United States (3–5), there are concerns about the added risk of Norplant side effects during the puerperium. Specifically, the tendency for childbirth to trigger depression in young, single women living in
© Society for Adolescent Medicine, 2000 Published by Elsevier Science Inc., 655 Avenue of the Americas, New York, NY 10010
June 2000
stressful, unsupportive environments (6 –10) may increase vulnerability to the moodiness and irritability that have been associated with Norplant use (11). This study of the prevalence of depressive symptoms in relation to the timing of postpartum Norplant insertion was designed to test the hypothesis that adolescent mothers who have Norplant inserted during the puerperium will report more symptoms of depression during the first postpartum year than their peers who do not receive Norplant.
Methods Patients The data were collected as part of a large study of contraceptive use and repeat pregnancy among patients in the Colorado Adolescent Maternity Program (CAMP). CAMP is a comprehensive, multidisciplinary, adolescent-oriented prenatal, delivery, postpartum, and infant care program in Denver, Colorado (11–14). The original study population consisted of a multiracial (50% white, 26% black, 22% Hispanic and 2% other) group of 354, poor (94% Medicaid-dependent), predominantly primiparous (84%), 13- to 19year-olds (mean ⫾ sd ⫽ 17.0 ⫾ 1.2 years). Participants were enrolled consecutively at delivery during the 1992 and 1993 calendar years (11). We chose this period for study because Norplant was extremely popular during this time and because it antedated major concerns about the Norplant removal process and the immense popularity of the other long-acting, progesterone-only contraceptive agent, Depo-Provera. Teenagers who used DepoProvera following delivery (n ⫽ 28) were excluded from this analysis because, like Norplant, DepoProvera is a long-acting, progesterone-only contraceptive agent, the use of which has been associated with moodiness, irritability, and depression (15). We also excluded teenagers who had no quantitative assessment of depressive symptoms (n ⫽ 49) and those lost to follow-up (n ⫽ 65). Analyses for this study were based on the remaining 212 young mothers. The study sample (n ⫽ 212) did not differ significantly from those excluded with regard to any of the social demographic variables related to the risk of postpartum depression we studied.
Data Collection and Variables Information about the study participants’ medical, social, psychiatric, and reproductive histories was
NORPLANT AND POSTPARTUM DEPRESSION
409
collected during gestation, at delivery, and 6 –12 months after Norplant insertion (or delivery in the case of those who did not use Norplant) with a self-administered questionnaire. Additional medical data were obtained by reviewing the participants’ medical records. Independent variable. “Norplant-user groups”: the 212 study participants were divided into three contraceptive use groups: 1) 100 (47%) “early Norplant users” who had Norplant inserted during the first two postpartum months, (on average at 5.5 ⫾ 1.7 weeks postpartum); 2) 72 (34%) “late Norplant users” who had Norplant inserted during the next 10 months (on average at 19.1 ⫾ 10.9 weeks postpartum); and 3) 40 (19%) “other contraceptive users” of whom 40% used an oral contraceptive, 17% used a barrier method, and 43% used no method. The distinction between “early” vs. “late” Norplant users, although somewhat arbitrary, was chosen to coincide with the puerperium, the period during which studies suggest new mothers are most apt to suffer from the moodiness and emotional lability (6 –10). In addition, most women who do not breastfeed resume ovarian cycling during the puerperium (10). Dependent variable. “Depressive symptoms at follow-up”: symptoms of depression were quantified with the Center for Epidemiologic Studies - Depression Scale (CES-D Scale) (16). The 20 CES-D Scale test items are scored on a 4-point Likert scale and summed to create a total score ranging from 0 to 60. Consistent with prior studies, depressive symptoms were dichotomized as “not depressed” (0 –15) and depressed (16 – 60) for the categorical analysis (7). Since the scale was administered 6 –12 months after Norplant insertion or 6 –12 months after delivery to study participants who did not use Norplant, the mean postpartum interval at follow-up was significantly different in the three Norplant use groups: 50.2 ⫾ 11.6, 65.0 ⫾ 16.8, and 43.5 ⫾ 12.3 weeks, respectively for early-, late-, and nonusers; (p ⬍ .0001). Center for Epidemiologic Studies - Depression Scale scores did not vary significantly in relation to postpartum week at follow-up (r ⫽ .06; p ⫽ .37). Intervening variables. The specific variables examined as potential confounders of the relationship between contraceptive group and depressive symptoms were identified a priori from a review of the literature (6 –10). They included the following widely accepted medical, psychologic, and social risk factors
410
STEVENS-SIMON ET AL.
for postpartum depression: (1) adverse pregnancy outcome (e.g., newborn sick, malformed, not the desired sex and/or multiple maternal complications during gestation or delivery); (2) multiparity; (3) rapid repeat conception; (4) antenatal experience described as extremely difficult or stressful; (5) a history of depression prior to conception; (6) antecedent depression (e.g., a CES-D Scale score ⱖ 16 at any time prior to Norplant insertion); (7) dissatisfaction with the support provided by the father of the child and/or family members (the teenagers were asked to rate the quality of the support they received during gestation from the baby’s father, their mother, and other important people in their support network on a simple scale: “excessive/intrusive”; “about what I needed”; “poor”; “none”); (8) minority race or ethnicity; (9) under-education, school failure, school dropout; (10) poverty; (11) unmarried status; (12) past or current physical or sexual abuse; and (13) involvement in socially problematic behaviors such as substance abuse, fighting, or delinquency. The study was approved by the Institutional Review Board at the University of Colorado Health Sciences Center.
Data Analysis Summary statistics were used to describe the study population. Comparisons among contraceptive groups were conducted with analysis of variance (ANOVA) when the outcome variable was continuous and Chi-square analysis when the outcome variable was categorical. To simplify the model and its application in clinical practice, the intervening variables were dichotomized (“present” or “absent”). Intervening variables for which there were significant contraceptive group differences at the bivariate level were allowed to enter the multivariate model, one at a time on the basis of the statistical significance of their association with depressive symptoms at follow-up. Analyses were performed with SPSS/PC⫹ (17).
Results Members of the three contraceptive groups did not differ significantly with regard to the sociodemographic characteristics we studied. Overall, the study sample was representative of the multiracial, predominantly Medicaid-dependent, unmarried, population of urban teenagers CAMP serves. Nearly half of the study participants had dropped out of school
JOURNAL OF ADOLESCENT HEALTH Vol. 26, No. 6
Table 1. Psychosocial Characteristics (N ⫽ 212) Norplant Insertion Group N (%) Characteristic
Early
Late
None
Number Significant other Father-of-child New boyfriend* None Poor father support** Antecedent depression†,***
100
72
40
65 (65) 8 (8) 27 (27) 42 (45) 49 (49)
34 (47) 16 (22) 22 (31) 42 (71) 39 (54)
21 (55) 5 (13) 14 (35) 19 (48) 11 (28)
* p ⫽ .03. ** p ⫽ .004 (missing data: N ⫽ 193 (94;40;59)). *** p ⫽ .02. † A CESD score ⱖ 16 at any time prior to Norplant insertion.
or were failing. Almost all participants (92%) had a teen parent as a close friend and most (76%) had a relative who was or had been a teen parent. The majority (80%) of the young mothers was primiparas who lived in supportive social environments. Over half lived with at least one biologic parent. Most felt they had received about the right amount of support from family members (90%) and peers (76%) during gestation and almost everyone (96%) had concrete plans for assistance with childcare. Past or current physical and/or sexual abuse and involvement in socially problematic behaviors such as substance abuse, fighting, or delinquency was reported infrequently. However, no objective measure of these events or behaviors was available. Because of the low quality of self-reports concerning abuse and involvement in risk behaviors during gestation and the bias that ensues from adjustment for a confounder that has been measured with error, these variables were not included in the analysis. Pregnancy outcomes were also similar in the three contraceptive groups. Most of the young women had uncomplicated pregnancies with good outcomes. However, 37% of the teenagers found the antepartum experience harder and more stressful than anticipated; 12% of the newborns were admitted to the neonatal intensive care nursery and 17% were not the desired sex. Importantly, none of the teenagers who used Norplant became pregnant again during the follow-up period compared to 13% of their peers who used other contraceptive methods. Table 1 describes some potentially important group differences in these young women’s relationships with their significant others and the prevalence of depression at baseline. Teenagers who had late Norplant insertions were most apt to have new boyfriends and to rate the support they received
June 2000
NORPLANT AND POSTPARTUM DEPRESSION
Table 2. Depressive Symptoms at Follow-upa (N ⫽ 212) Norplant Insertion Group Characteristic Number CESD Score (mean ⫾ SD)* Depressed [N(%)]** Became Depressed [N(%)]b,***
Early
Late
None
100 13.2 ⫾ 10.9
72 18.0 ⫾ 13.4
40 11.5 ⫾ 7.7
32 (32) 6 (6)
35 (49) 13 (18)#
10 (25) 5 (13)
* p ⫽ .004. ** p ⫽ .02. *** p ⫽ .05. # p ⫽ .02 for early vs late Norplant users. a Follow-up occurred 6 to 12 months after Norplant insertion or 6-to-12 months after delivery (for non-users); the mean postpartum interval at follow-up was: 50.2 ⫾ 11.6, 65.0 ⫾ 16.8, and 43.5 ⫾ 12.3 weeks, respectively (p ⬍.0001). b Antecedent CESD score ⬍16 and CESD score ⱖ16 at followup.
from their baby’s father during pregnancy as poor or nonexistent. There were no statistically significant contraceptive group differences in reported suicide attempts prior to conception (19%) or prolonged periods (ⱖ 5 consecutive days) of sadness (19%) and/or suicide attempts (5%) during gestation. Nevertheless, Norplant users were more likely to have experienced antecedent depression, that is to have scored in the depressed range (ⱖ 16) on the CES-D Scale at some time prior to Norplant insertion. Complaints of dysphoria and moodiness were common during the follow-up period. Center for Epidemiologic Studies—Depression Scale scores ranged from 0 to 57 (mean ⫾ sd: 14.5 ⫾ 11.5) and 77 (36%) of the 212 study participants had CES-D Scale scores indicative of depression. However, contrary to the study hypothesis, the data presented in Table 2 show that late, rather than early, postpartum Norplant insertion was associated with an increase in the severity of depressive symptoms during the follow-up period. The mean CES-D Scale score and prevalence of depression was significantly higher among late Norplant users than among members of the other two contraceptive use groups. Within the study group as a whole, CES-D Scale scores rose in tandem with postpartum week at Norplant insertion (r ⫽ .2; p ⫽ .01). However, among late Norplant users, there was no relationship between the timing of Norplant insertion and the duration of Norplant use and either the mean CES-D Scale score or the prevalence of depression. Center for Epidemiologic Studies - Depression Scale scores at follow-up exceeded antecedent CES-D Scale scores, particularly among late Norplant users,
411
Table 3. Predictors of Depressive Symptoms at Followup Characteristic
F (P)
Main effects* Antecedent depression** New boyfriend Late Norplant insertion
24.8 (⬍.0001) 7.0 (.009) 6.1 (.01)
Two-way interactions Norplant by depression Norplant by boyfriend
5.4 (.02) 4.5 (.04)
* Degrees of freedom (1,210); no Significant 3-way Interactions. ** A CESD score ⱖ16 during gestation or prior to Norplant insertion. R2 ⫽ .41.
[mean increase in CES-D Scale score: 5 ⫾ 13 points compared to 2 ⫾ 10 and 1 ⫾ 7 points, respectively among early and non-Norplant users (p ⫽ .04)]. The data presented in Table 2 show that early Norplant users were the least likely, and late Norplant users the most likely to become depressed at follow-up (p ⫽ .02). The prevalence of other side effects was unrelated to timing of Norplant insertion. Multivariate analyses controlling for preexisting contraceptive group differences in characteristics with a hypothesized relationship to postpartum depression (Table 1) revealed that antecedent depression was the most powerful predictor of the severity of depressive symptomatology at follow-up (R2 ⫽ .33). Other factors that predicted depression at follow-up are shown in Table 3 in order of decreasing significance. Late Norplant insertion remained a statistically significant, independent predictor of the severity of depressive symptoms and interacted with the other two variables in the model such that teenagers who had been depressed prior to Norplant insertion and teenagers who had new boyfriends at delivery experienced the most severe depressive symptoms following Norplant insertion (mean CES-D Scale scores ⫽ 19.6 and 26.6, respectively). Collectively, these three characteristics accounted for 41.3% of the variance in the severity of depressive symptomatology at follow-up.
Discussion The puerperium is a time of major psychologic upheaval during which women tend to be particularly emotionally labile (8,9). Many women experience a brief period of excessive irritability and moodiness, known as the “maternity blues”, and the small proportion of these mildly dysphoric women who
412
STEVENS-SIMON ET AL.
enter a more protracted period of melancholy are given the diagnosis of “postpartum depression” or “postpartum psychosis”, depending on the severity of their symptoms (8 –10). The incidence of these serious forms of postpartum psychopathology is extremely variable, but the literature suggests that teenagers are more vulnerable to postpartum depression than adults (6 –10). Thus, it was reassuring that the results of this study did not support the study hypothesis; puerperal Norplant insertion did not exacerbate postpartum depression in this population of poor, young, unmarried mothers. The significance of our finding that late Norplant insertion was associated with an increase in the severity of depressive symptoms during the follow-up period is more difficult to interpret. The association does not appear to reflect group differences in the timing of the postpartum follow-up, as there was no correlation between postpartum week at follow-up and CES-D Scale score. The association does, however, appear to be partially psychosocial in origin, as the effect of late Norplant insertion on the severity of depressive symptoms was moderated by controlling for the presence of antecedent depression and/or the acquisition of a new boyfriend prior to delivery. This finding is consistent with the results of most other studies and the notion that postpartum depression is a multifactorial problem, resulting from a complex interaction between biologic, psychologic, and social factors (6 –10). Within the context of this model, the single factor that has been most consistently and strongly related to the development of postpartum depression is social support (6 –10). High satisfaction with support, particularly from the father of the child and, in the case of teenagers, their own mother, decreases the likelihood of depressive symptoms (6). Conversely, conflict with, and/or rejection by, these important persons has been consistently associated with an increased likelihood of depression (6,7). A personal and/or family history of depression and the presence of anxiety and/or depression during pregnancy have also been associated with an increased risk of postpartum depression in several studies (6 –10). It is still unclear precisely how and why ovarian steroids contribute to the mental and emotional instability of some new mothers. The increased emotional lability that is characteristic of the puerperium tends to coincide temporally with rapid changes in serum levels of gonadal hormone that are capable of altering neuronal activity and neurotransmitter levels in portions of the brain that affect mood. Thus, there has been a tendency to draw a causal associa-
JOURNAL OF ADOLESCENT HEALTH Vol. 26, No. 6
tion between the rapid postpartum withdrawal of these hormones, particularly progesterone, and the development of postpartum depression (8,10). Indeed, women who develop maternity blues have been found to experience more precipitous drops in serum progesterone levels from the antenatal to the postpartum period and there have been scattered, anecdotal, reports that treatment with progesterone ameliorates the symptoms of postpartum depression (10). Like all studies, this one has its limitations. First, since all of the participants were enrolled in a comprehensive adolescent-oriented maternity program, unknown differences between pregnant and parenting teenagers who are cared for in these settings and the larger population of pregnant and parenting teenagers means that it is necessary to use caution in generalizing our findings. Second, even though we used multivariate analyses to control for antecedent “Norplant-use group” differences regarding numerous widely recognized risk factors for postpartum depression, Norplant use was not a random event in the population we studied. Thus, there may have been other, unanticipated differences between early, late, and non-Norplant users that confounded our analysis and biased our findings regarding the association between the timing of Norplant insertion and the severity of postpartum depression. Nevertheless, since we found that early Norplant users were significantly less apt to become depressed at follow-up, it is tempting to speculate about the therapeutic role of puerperal Norplant insertion. The authors wish to thank the CAMP staff and patients for their help with data collection. Supported in part by: Grant #APH000166 –5; Office of Adolescent Pregnancy, NIH; a grant from the Colorado Trust, Denver, CO and Grant #5 MO1 RR00069 General Clinical Research Centers Program, National Center for Research Resources, NIH. Presented in part: The Society for Pediatric Research Annual Meeting New Orleans, May 1998.
References 1. Stevens-Simon C, White M. Adolescent pregnancy. Pediatr Annals 1991;20:322–31. 2. Brown SS, Eisenberg L. The Best Intentions. Washington DC: National Academy Press, 1995. 3. Polaneczky M, Forke C, Armstrong J. Norplant vs oral contraceptive use by adolescent mothers: Long term outcome. J Adol Health 1996;18:120 (Abstract). 4. O’Dell CM, Forke CM, Polaneczky MM, et al. Depot Medroxyprogesterone Acetate or oral contraception in postpartum adolescents. Obstet & Gynecol 1998;91:609 –14. 5. Ricketts SA. Repeat fertility and contraceptive implant use among Medicaid recipients in Colorado. Fam Plann Perspect 1996;28:278 – 80.
June 2000
6. Barnet B, Joffe A, Duggan A, et al. Depressive symptoms, stress, and social support in pregnant and postpartum adolescents. Arch Pediatr Adoles Med 1996;150:64 –9. 7. Deal LW, Holt VL. Young maternal age and depressive symptoms: Results from the 1988 National Maternal and Infant Health Survey. Am J Pub Health 1998;88:266 –70. 8. Stowe ZN, Nemeroff CB. Women at risk for postpartumonset major depression. Am J Obstet Gynecol 1995;173:639 – 45. 9. Wolman WL, Chalmers B, Hofmeyr J, Nikodem V. Postpartum depression and companionship in the clinical birth environment: A randomized, controlled study. Am J Obstet Gynecol 1993;168:1388 –93. 10. Harris B, Lovett L, Newcombe RG, Read GF, Walker R, Riad-Fahmy D. Maternity blues and major endocrine changes: Cardiff puerperal mood and hormone study II. Br Med J 1994;308:949 –53. 11. Stevens-Simon C, Kelly LS, Singer D, et al. Correlates and consequences of early removal of Levonorgestral implants
NORPLANT AND POSTPARTUM DEPRESSION
12. 13.
14.
15.
16.
17.
413
among adolescent mothers. Arch Pediatr Adolesc Med 1998; 152:893– 8. Stevens-Simon C, Wallis J, Allan-Davis J. Which teen mothers choose Norplant? J Adol Health 1995;16:350 –3. Stevens-Simon C, Wallis J, Allan-Davis J. Antecedents of preterm delivery among adolescents: Relationship to type of prenatal care. J Mat Fetal Med 1995;4:186 –93. Stevens-Simon C, Kelly LS, Singer D. Reasons pregnant adolescents give for not using contraceptives prior to conception: Implications for recidivism. Arch Ped Adol Med 1996;150: 1037– 43. Sangi-Haghpeykar H, Poindexter III AN, Bateman L, Ditmore JR. Experiences of injectable contraceptive users in an urban setting. Obstet Gynecol 1996;88:227–33. Radloff L. The CES-D scale: A self-report depression scale for research in the general population. J Applied Psychological Measures 1977;1:385– 401. Norusis MJ. Statistical Package for the Social Sciences. Chicago, IL: SPSS/PC⫹, 1990.