Prenatal Cocaine Use and Maternal Depression: Effects on Infant Neurobehavior

1. INTRODUCTION

A 2003 National Survey on Drug Use and Health (NSDUH) estimated the rate of illicit drug use among pregnant women at 4.3%; over 130,000 infants were exposed to illicit drugs in utero each year [1]. Substance abuse remains a major public health problem that affects millions of children and places enormous financial and social burdens on society.

The NSDUH reported that adults who used illicit drugs were more than twice as likely to have a serious mental illness as non-using adults; 26.9% of illicit drug users reported having a serious mental illness in the past year. Major depressive disorder (MDD) is the predominant mental illness reported by women, with a lifetime occurrence of 17% [34]. The first onset of MDD is most likely to occur during the childbearing years [19, 34]. Postpartum depression occurs at a rate of 7.4% to 25%, with the highest rates occurring in lower socioeconomic samples [10, 40, 68, 91]. Thus the probability that a woman who used cocaine during her pregnancy will experience a depressive episode in the postpartum period is at least 1 in 6 (based on an average 15% prevalence)[34, 73]. Yet, little is known about the association between maternal depression and prenatal substance abuse and the effects on infant outcomes.

Prenatal cocaine exposure is often associated with a subtle but statistically significant decrease in infant birthweight, gestational age and head circumference [6, 13, 30, 71, 84, 93]. There is emerging evidence supporting a neurobehavioral profile of the cocaine exposed infant, namely, that prenatally cocaine exposed infants tend to have difficulty with arousal and state regulation, including increased or impaired behavioral and physiological reactivity to stress [12, 58, 59, 61, 64, 81]. In a previous report, we found results consistent with these prior studies. One-month old cocaine exposed infants had lower arousal, poorer quality of movement and self-regulation, as well as increased motor tone and lower auditory brain response measures after controlling for other drug use and maternal caregiving variables [50, 51]. Data conflicting with these findings have also been reported with no differences found in young infants prenatally exposed to cocaine [65, 92]. The collective data on prenatal cocaine exposure effects on infant neurobehavior has been inconclusive and warrants further research.

Maternal depression in the first 30 days postpartum is an integral component of the care giving environment of the infant and may have a significant impact on infant development, particularly in an already vulnerable environment [48]. Consistent evidence documents an association between postpartum depression and less positive mother-infant interactions during play and care-taking routines. In addition, women who are depressed immediately postpartum are more likely to have been depressed during the pregnancy, presenting another potential exposure to the fetus [18, 33]. Infants born to mothers with depression during pregnancy were reported to be more aroused, reactive, harder to console, had poorer self-regulation and sleep and less mature neuromotor development. [3, 39, 94].

Despite the growing evidence that depression is highly prevalent in women using cocaine and that both cocaine exposure and maternal depression is associated with adverse infant outcomes, the impact of postpartum depression on infant neurobehavioral development in cocaine exposed samples has not been previously reported. Therefore, the purpose of this study was to examine the effects of postpartum depression on infant neurobehavior outcomes in infants prenatally exposed to cocaine compared to non-exposed infants.

The infants were originally recruited for the Maternal Lifestyles Study (MLS) [48]. The MLS was developed in the early 1990s as an interagency collaborative effort involving the National Institute of Child Health and Human Development (NICHD); the National Institute on Drug Abuse (NIDA); the Administration on Children, Youth and Families; and the Center for Substance Abuse Treatment and is the largest prospective longitudinal study of acute neonatal events and long-term health and developmental outcomes associated with cocaine use during pregnancy.

We hypothesized that the infants whose mothers were experiencing depression in the early postpartum would perform less optimally on a structured examination of infant neurobehavior, and that infants in the maternal depression condition who were prenatally exposed to cocaine would have the poorest performance, particularly in self-regulation, arousal, and reactivity (more handling needed, more excitability) than infants who were not exposed to cocaine prenatally.

2. METHODS

2.1 Study Design

The Maternal Lifestyles Study (MLS) was conducted at four NICHD Neonatal Research Network sites (Brown University, University of Miami, Wayne State University, and the University of Tennessee at Memphis). Mother-infant dyads were screened and recruited immediately after delivery (N=11,811 consented to participate). A trained nurse or social worker conducted a semi-structured interview with the mother to determine past and current drug use and sociodemographic information. A physical examination of the infant was conducted, and the infant’s meconium collected. Before mothers and infants were discharged from the hospital, their charts were abstracted to collect selected medical data. Procedures used to determine exposure status have been described previously[24, 49]. Briefly, meconium samples were collected in the nursery and analyzed in a central laboratory. The study definition of “exposure” was maternal admission of cocaine or opiate use during this pregnancy based on the hospital interview or positive gas chromatography/mass spectroscopy (GC/MS) confirmation of cocaine metabolites. “Unexposed” was defined as denial of cocaine or opiate use during this pregnancy and a negative enzyme-multiplied immunoassay technique screen for cocaine and opiate metabolites. Once exposure status was confirmed by GC/MS analyses, the eligible participants were invited to come into the center for follow-up visit.

2.2 Participants

The infant’s first follow-up visit took place between 42 and 44 weeks post-menstrual age. An approximate time of one-month post delivery was determined to be a period that would allow for exposure status confirmation and a time when potential birth effects of labor and delivery stress and substance withdrawal are less likely to impact infant neurobehavior. Mothers were eligible to participate in the study if they were at least 18 years old, had no identified psychosis, were not institutionalized for mental retardation or mental illness, had no language barriers, and planned to stay in the catchment area. Their singleton infants were eligible if they were at least 501g and less than 42 weeks gestational age at birth without chromosomal abnormalities, fatal illnesses or TORCH infection. All eligible exposed infants who attended the one month visit were matched with an unexposed infant on race, gender, and gestational age [51]. At the one month visit, 1388 mother-infant dyads (658 in the exposed group and 730 in the comparison group) were enrolled in the MLS longitudinal study.

The analyses conducted for the present study were limited to those infants still residing with their biological mothers at the follow-up visit to control for a consistent physiological and care giving environment from prenatal to postnatal life (N=1260). Only dyads without prenatal opiate exposure were included in these analyses to limit potential substance exposure confounds (N=1165). 1,053 mother-infant dyads met these criteria and accurately completed the semi-structured interview of psychiatric symptoms. The methods used to determine the final sample of 1,053 are detailed in the maternal measures section (2.3.1).

2.3 Measures

The one month visit included neurobehavioral, medical and physical status measures of the infant; social and demographic questionnaires; The Maternal Interview of Substance Use (MISU), and the Addiction Severity Index (ASI).

2.3.1 Maternal Measures

The MISU was developed for the MLS and is a structured interview that provides information about the frequency and quantity of substance use for each trimester during this pregnancy for cocaine, alcohol, marijuana, and nicotine.

The ASI is a structured clinical interview that evaluates treatment outcomes of alcohol and drug problems [42, 78]. The respondents are asked about current and lifetime substance use, psychiatric symptoms, diagnoses and hospitalizations. In addition, the interview elicits severity and duration for each psychiatric symptom present in the past 30 days to produce a summary score for psychiatric status [42, 78].

The ASI was used to determine participants who “experienced serious depression” in the past 30 days and those that experienced “a lifetime history of depression”. Although this measure is not diagnostic, the ASI as an indicator of depression has been correlated with diagnoses of depression attained via semi-structured interview and has been a strong predictor of functional outcome in opiate and cocaine abusing cohorts [5, 78 ]. In comparison to semi-structured interview diagnoses of depression, the ASI was found to separate depressed from non-depressed addicts with a sensitivity of .83 and a specificity of .55 [78]. The ASI has been used successfully as a screening tool for depression in pregnant and postpartum women [60].

The presence or absence of a reported serious depression in the past 30 days (lasting at least 2 weeks), coupled with the symptom severity scores on the ASI “psychological problems” subscale (0-5) was used to determine study depression groups (Table 1):

• Depressed group: experienced serious depression in the past 30 days AND psychiatric symptom score of 3 or more.

• Non-Depressed group: denied serious depression in the past 30 days AND psychiatric symptom score less than 3, regardless of psychiatric history.

TABLE 1.

Number of Participants in the Depression Groups by Cocaine Exposure

	Cocaine Exposed	Non Cocaine Exposed
	( COC )	(NonCOC)
	N (%)	N (%)	Total in Depression Groups
Depressed (DEP)	76 (19.3)	104 (15.5)	180
Non-Depressed (NonDEP)	309 (80.7)	564 (84.6)	873
Total in Cocaine Groups	385	668	1053

Participants that did not meet these criteria were eliminated from the analyses, as their psychiatric status could not be determined from their ASI data (N=112).

2.3.2 Infant Measures

In addition to medical and demographic characteristics of the sample, the NICU Network Neurobehavioral Scale (NNNS) was used to examine infant neurobehavior at 42-44 weeks post-menstrual age. The NNNS was specifically designed for the MLS [52] and used in studies of intrauterine exposure to cocaine, opiates, and nicotine [25, 38, 44, 46, 47, 64]. The NNNS is composed of three parts: neurological, behavioral, and stress/abstinence. The neurological component includes active and passive tone, primitive reflexes, and items that reflect the integrity of the central nervous system and maturity of the infant. The behavioral component is based on items from the Neonatal Behavioral Assessment Scale (NBAS) modified to be sensitive to putative drug effects [8]. The stress/abstinence component is a checklist of signs observed throughout the examination organized by organ system based primarily on the work of Finnegan that are indicative of stress, hyper-stimulation, or nervous system instability [29]. The NNNS follows a relatively invariant sequence of administration that starts with a pre-examination observation, followed by the neurological and behavioral components. The individual NNNS items are aggregated into summary scales (see Table 2). The summary scores were formed from a conceptual and statistical approach and have psychometric properties with coefficient α ranging from 0.56 to 0.85 [53]. The habituation data were not used, as infants are required to be asleep for this part of the assessment and most infants were awake at the beginning of the examination. The actual sequence of administration and the means used by the examiner to maintain an infant’s participation in the examination were recorded. All infants were examined between 42 and 44 weeks post-menstrual age by certified psychometrists at each site who were masked to infant exposure status. The certification process for the NNNS ensures initial reliability with certified trainers. A gold-standard NNNS trainer visited each site for the training and certification process to ensure reliability across sites.

TABLE 2.

Summary Scores of the NICU Network Neurobehavioral Scale (NNNS)

Attention	Indicates an infant’s ability to attend and respond to auditory and visual stimulation; high scores on this scale show good response and sustained alertness
Handling	Indicates the number and type of maneuvers necessary to keep the infant in the appropriate state to administer items; high scores indicate infants who need substantial input from the examiner in order to elicit attention and response to stimuli
Quality Movement	A measure of motor control including smoothness, maturity, modulation of movement of the arms and legs as well as startles and tremors; high scores indicate good quality of movement
Self Regulation	Indicates how the infant copes with the demands of the exam, with higher scores indicating better regulation
Non-Optimal Refl.	The number of non-optimal reflex responses
Stress/Abstinence	The number of stress/abstinence signs displayed by the infant across five organ systems
Arousal	Indicates the infant’s overall level of arousal and associated motor activity during the exam; high scores indicate high arousal, activity, and fussing and crying during the exam
Hypertonia	Indicates increased muscle tone in the arms, legs, and trunk; a high score describes an infant whose overall tone is consistently hypertonic
Hypotonia	Indicates decreased muscle tone in the arms, legs, and trunk; a high score on this scale describes an infant who was consistently hypotonic
Asymmetrical Refl.	A count of the number of times a reflex on one side of the body is stronger or weaker than on the other side
Excitability	A measure of high levels of motor, state and physiological reactivity; high scores indicate high levels of irritability even with attempts to soothe
Lethargy	A measure of low levels of motor, state and physiological reactivity; high scores indicate extreme under-arousal despite the necessary handling during the exam

2.4 Ethics

The procedures followed for this study were approved by the Internal Review Board of each institution involved. All participants signed an IRB-approved consent form prior to participating in the study. A NIDA Certificate of Confidentiality was obtained by each site that assured confidentiality of information regarding the subjects’ drug use. The certificate superseded the mandatory reporting of illegal substance use that was in effect in the Florida and Rhode Island sites. The certificate was explained to the mother during the recruitment and informed consent procedure, including the condition that the certificate did not exclude reporting evidence of child abuse or neglect.

2.5 Statistics

Analysis of variance (ANOVA) and Chi-square were used to determine group differences on infant and maternal characteristics. The neurobehavioral dependent measures were tested for main effects and interactions using a series of ANOVA with Type III sums of squares: 1). A 2 (Cocaine: exposed/non-exposed) x 2 (Depression: depressed/non-depressed) factorial ANOVA was conducted with each NNNS summary score variable. 2). An ANCOVA with the 2 X 2 design with covariates selected because they met the following statistical criteria: the variable was correlated with the independent variables (cocaine exposure or depression group) or with the NNNS subscales (p < .05) and not correlated with other covariates (Pearson r < 0.70) [37, 43, 45, 76]. The covariates that were used in the adjusted analysis included Index of Social Position Score from the Hollingshead scale of socioeconomic status (SES), birthweight, maternal age, research site, nicotine, marijuana, and alcohol use throughout the pregnancy. 3.) An ANCOVA design using a three cocaine exposure groups: no exposure, some exposure, and heavy exposure by the 2 depression groups was conducted to determine if depression effects would differ based on amount of cocaine exposure. 4.) An ANCOVA design comparing depression effects based on trimester of exposure, no exposure, first trimester only, and second or third trimester. Post-hoc analyses were conducted on the significant variables using Tukey HSD tests.

3. RESULTS

3.1 Prevalence of reported depression

The number of participants excluded from the analyses due to lack of consistent ASI data was significantly greater in the cocaine-exposed (COC) group (N=59;13.29%) than the non cocaine-exposed (nonCOC) groups (N=53; 7.35%) (X^2/Yates = 10.48, p<.002). The 112 participants excluded from further analyses were not different on demographic characteristics or infant outcome measures than the participants that were included in the remaining analyses.

Based on the ASI, there was not a significant difference in the number of women meeting criteria for depression in the past 30 days in the COC group than the nonCOC group (X^2/Yates =2.71, p=.09) (Table 2). Only 14 of the 172 depressed (DEP) mothers reported taking psychotropic medication in the past 30 days; 3 of them were cocaine users.

3.2 Maternal Variables

3.2.1 Demographics

The cocaine groups were originally matched on demographic variables. Therefore, we analyzed demographic variables using Chi-square for DEP differences within the COC groups. Table 3 presents the maternal characteristics of the DEP and non-depressed (nonDEP) groups in both the COC and nonCOC groups. Only one significant difference was found for DEP within COC groups: more women in the nonCOC/DEP group were less than 27 years of age (X²=8.24, df=2, p<.02) than women in the nonCOC/nonDEP group.

TABLE 3.

Characteristics of Women in the Depression Groups by Cocaine Exposure

		COC					NonCOC
		Dep		NonDep			Dep		NonDep
		N=76		N=309			N=104		N=564
		N	%	N	%		N	%	N	%
Age	18-25	18	23.68	57	18.45	n.s.	67	66.67	284	49.35 *
	26-49	58	76.32	252	81.55		37	29.89	280	42.62
Marital Status	Married	8	10.53	35	11.33	n.s.	18	17.31	154	27.30	ns
	Single	68	89.47	274	88.67		86	82.69	410	72.70
Parity	First Born	6	7.89	31	10.03	n.s.	27	25.96	175	31.08	ns
	2-3 Children	30	39.47	128	41.42		44	42.31	269	47.78
	4 or More	40	52.63	150	48.54		33	31.73	119	21.14
Race	Black	62	81.58	253	81.88	n.s.	88	84.62	439	77.84	n.s.
	White	10	13.16	34	11.00		12	11.54	84	14.89
	Other	4	5.26	22	7.12		4	3.85	41	7.27
SES	% low	22	28.95	86	27.83	n.s.	21	29.81	103	18.26	n.s.
Insurance	% Medicaid	67	88.16	276	89.32	n.s.	89	85.58	434	76.95	n.s.

Within group percentages presented. n.s.: p>.05, *p<.05

3.2.2 Maternal Drug Use during Pregnancy

The amount of cocaine use throughout pregnancy and in the first month postpartum for the DEP and nonDEP cocaine groups is shown in Figure 1. Both groups decreased their cocaine use throughout the pregnancy and into the first month postpartum (F=59.27, p<.0001). Although the DEP group reported consistently more cocaine use than the nonDEP, the difference was only marginally significant (F=3.15, p<.07). Significantly more prenatal cocaine users reported using cocaine in the first 30 days postpartum if they were in the DEP group (26.3%) than those in the nonDEP group (14.3%), X²=6.35, p<.02.

Figure 1. Amount of cocaine exposure during pregnancy reported for the DEP and nonDEP groups.

Maternal report of the mean number of days using cocaine per week.

PP=first 30 days postpartum

Table 4 shows the mean amount of other reported drug use throughout the pregnancy. The COC and nonCOC groups differed in their use of other drugs. Women who used cocaine during the pregnancy also drank more alcohol (F(1, 1039)=50.56, p<.0001), smoked more marijuana (F(1, 1039)=8.22, p<.01), and smoked more cigarettes (F(1, 1039)=103.83, p<.0001) during the pregnancy than those who did not use cocaine. There was a significant depression effect for nicotine use during pregnancy; women in the DEP group smoked more cigarettes throughout the pregnancy than the nonDEP (F (1, 1039) =11.67, p<.001).

TABLE 4.

Other Substance Use for Women in the Depression Groups by Cocaine Exposur

	COC		NonCOC
	DEP	NonDEP	DEP	NonDEP
	(N=74)	(N=285)	(N=98)	(N=546)
	M±SE	M±SE	M±SE	M±SE
Alcohol (std.drinks/day)	0.65±.09	0.59±.05	0.22±.08	0.07±.03	***COC
Marijuana (joints/day)	0.09±.04	0.17±.02	0.07±.03	0.03±.01	** COC
Nicotine (Cigs/day)	12.03±.96	9.17±.49	4.72±.83	2.57±.35	***COC, ***DEP

^**p<.01,

^***p<.001

COC, cocaine main effect; DEP, depression main effect

3.2.3 Maternal Mental Health Treatment in the first 30 Days Postpartum

There was a significant COC effect for the mean number of days women reported receiving outpatient mental health treatment in the first 30 days postpartum (Mean (SE)): 4.46 (0.33) COC/nonDEP; 4.75 (0.67) COC/DEP; 0.05 (0.03) nonCOC/nonDEP; 0.07 (0.57) nonCOC/DEP, (F (1, 1,043) = 88.85, p<.0001). More women in the COC groups were in mental health treatment versus nonCOC groups; 26.5% of women in the COC group received outpatient treatment in the postpartum period compared to only 2.3% of women in the nonCOC group.

There was not a significant difference in psychological symptom severity score derived from the ASI between COC (mean (SE) =0.834 (0.07)) and nonCOC (mean (SE)=0.692 (.05)) groups, F(1,1049)=2.68, n.s.

3.3 Infant Variables

3.3.1 Demographics

Table 5 presents the characteristics of the neonates at birth in the depressed (DEP) and non-depressed (nonDEP) groups by cocaine exposure. There were no significant cocaine effects on infant characteristics at birth. There was a significant interaction effect for the variables of birthweight (F(1,1043)=6.93, p<.01), length (F(1,1045)=6.13, p<.02), and head circumference (F(1,1045)=5.40, p<.03). The nonCOC/DEP infants had significantly lower birthweights (F(1,1045)=6.50, p<.01) and were shorter in length than the nonCOC/nonDEP infants (F(1,1045)=6.31p<.01). There were no significant DEP differences.

TABLE 5.

Neonate Characteristics in the Depression Groups by Cocaine Exposure

	COC		NonCOC
	Dep	Non-Dep	Dep	Non-Dep
	(N=76)	(N=306)	(N=104)	(N=560)
	Mean±SE	Mean±SE	Mean±SE	Mean±SE
GAa	36.46±0.46	36.12±0.23	35.61±0.40	36.46±0.17
BirthWt. (kg)	2.66±0.10	2.56±0.05	2.47±0.08	2.73±0.04	**Int
Length (cm)	46.91±0.58	46.44±0.29	45.70±0.50	47.27±0.21	*Int
Head Circ. (cm)	32.41±0.35	31.96±0.18	31.64±0.31	32.34±0.13	*Int
Apgar 1 (median)	7.70±0.22	7.58±0.11	7.36±0.19	7.45±0.10
Apgar 5 (median)	8.75±0.12	8.61±0.06	8.55±0.10	8.53±0.04
Male (%)	41.27	49.46	52.87	44.86

^*p<.05,

^**p<.01;

^aBest obstetrical estimate

Int, Interaction effect

3.3.2 Neurobehavioral Scores on the NNNS

There were no significant differences on any of the NNNS measures within the DEP groups for those taking psychotropic medications or receiving psychiatric treatment in the past 30 days. Therefore, psychotropic medication use prescribed by a physician and psychiatric treatment were not used in the following analyses. Infants of women with a lifetime history of depression and not currently depressed did not differ from the infants born to women that were never depressed. Therefore, women with a lifetime history of depression with the absence of a current depression were included in the nonDEP group.

Analyses were conducted in two steps. First, ANCOVA analyses were run using the factors of DEP (yes/no) and COC (yes/no) groups. All NNNS variables that were significant for a main or interaction effect were further analyzed for the trimester of cocaine exposure and the amount of cocaine use reported.

In the initial analyses there were significant main effects for depression after covariate adjustment for Self-Regulation (F(1, 996)=4.08, p<.05), Excitability (F(1, 991)=4.79, p<.03), and Hypotonia (F(1,986)=4.05, p<.05) with infants in the DEP groups having lower self-regulation, higher excitability, and more hypotonia than infants in the nonDEP groups. There were no significant cocaine main effects after covariate adjustment. There were three significant interaction effects after covariate adjustment: Self-Regulation (F(1, 996)=7.45, p<.012), Stress (F(1, 991)=4.01, p<.05), and Arousal (F(1, 980)=5.93, p<.02).

The pattern of results was the same regardless of the trimester of cocaine use; only the nonCOC groups had a significant DEP effect. Therefore, analyses examining the amount of cocaine exposure included infants who were exposed to cocaine in any trimester.

Secondary post-hoc analyses were conducted on the NNNS variables using 3 cocaine groups: none, some (<1.5 days per week) and heavy (≥1.5 days per week) cocaine use. Women who denied cocaine use but whose infants showed positive cocaine toxicology from meconium samples were eliminated from the secondary analyses. Figures 2a-d present the significant adjusted mean scores on the NNNS for the DEP groups in each of the three cocaine exposure groups.

Figure 2. NNNS scores for the DEP and nonDEP group in each COC use group.

a. Self Regulation

b. Stress

c. Arousal

d. Excitability

*p<.05, **p<.01,; Significant differences on post-hoc Tukey HSD tests

The pattern of results remained the same; only those infants without any prenatal cocaine exposure showed significant DEP differences on the NNNS. The nonCOC/DEP infants compared to nonCOC/nonDEP infants had significantly lower self-regulation (Figure 2a), more stress signs (Figure 2b), more arousal (Figure 2c), and more excitability (Figure 2d).

Due to the large discrepancy in mental health treatment received between COC and nonCOC women, the NNNS analyses were repeated with only those women not receiving mental health treatment. Three interaction effects remained significant: self-regulation, stress, and arousal.

4. DISCUSSION

This study examined the effects of early postpartum maternal depression on infant neurobehavior at one month postmenstrual age in infants prenatally exposed to cocaine compared to non-exposed infants using the NICU Network Neurobehavioral Scale (NNNS). Using the ASI as a screening tool for depression, we were able to determine neurobehavioral differences of small to medium effect sizes (0.2 to 0.5), demonstrating adequate power for the intended analyses [16]. The prevalence of a serious depression as measured by the ASI in the immediate postpartum period in this sample was substantial in both the cocaine (COC) and non-cocaine (nonCOC) groups (19.3% and 15.5%, respectively). These rates of depression are within the range typically reported in samples with low socioeconomic status and in substance abusing samples and indicate a critical need to measure maternal mood symptoms in cocaine using samples.[14, 73]

We originally hypothesized that the infants whose mothers were experiencing depression in the early postpartum would perform less optimally on a structured examination of infant neurobehavior, and that infants in the maternal depression condition who were prenatally exposed to cocaine would have the poorest performance. Our hypothesis was only partially correct. We found that infants whose mothers were experiencing a postpartum depression (DEP) were having more difficulty with self-regulation and excitability and they were more hypotonic than infants of non-depressed women (nonDEP). However, we also found significant interaction effects that did not support our hypothesis that prenatal cocaine exposure would result in an exacerbation of the DEP effects. DEP effects were found on 4 of the 12 NNNS variables only for those infants who were not prenatally exposed to cocaine: infants in the DEP group had poorer self-regulation and more stress signs, excitability, and arousal than infants in the nonDEP group. Of note is that the DEP/COC group (double exposure) did not differ on any of the neurobehavioral measures from the nonDEP/nonCOC (control) group. The pattern of results was the same regardless of the amount or trimester of cocaine exposure.

Infants in the nonCOC/DEP group also had the lowest birthweights, shortest birth-lengths, and smallest head circumferences than the other three groups. Women in the nonCOC/DEP group were also the youngest mothers. The significant neurobehavioral results persisted after controlling for the effects of maternal age, birthweight, polydrug use, SES and research site. Regression analyses were conducted to further test the relationship between maternal psychiatric symptom severity measured on the ASI and infant neurobehavior after controlling for the effect of infant birthweight. The results remained the same, namely that maternal symptom level was related to infant NNNS scores (self-regulation, stress, arousal excitation, and hypotonia) only for those in the nonCOC group.

The results indicate that for infants not prenatally exposed to cocaine, maternal depression is associated with greater difficulty with state regulation and reactivity during the examination. These data are in agreement with prior studies that reported that infants of depressed mothers were more irritable, had more sleep problems, were more tense, and reacted more under the stress of testing conditions than infants of non-depressed women [27, 36, 62, 63, 90].

We were unable to determine if the infants were exposed to maternal depression in utero as the psychiatric problems section of the ASI interview did not ask specifically about depression during the pregnancy. A significant depression in the first 30 days after delivery is likely to indicate that the depression started prior to or during the pregnancy as antenatal depression is one of the most predictive factors for postpartum depression [4, 18, 33, 40, 69, 70]. Therefore, we cannot be sure if the depression effects found in this study are due solely to postnatal maternal depression or if prenatal depression exposure contributes to the effects.

Behavioral findings in prenatally cocaine-exposed infants have not always been consistent. Some studies did not find cocaine effects on newborns and young infants, while others reported difficulty in arousal regulation, including greater excitability, poor state regulation, more rapid changes in arousal with stimulation, increased arousal from sleep, and increased physiological lability [7, 9, 20, 32, 41, 57, 75]. Maternal depression was not reported in these papers. Based on our findings, inconsistencies in results may be related to other factors that were not measured or reported, such as maternal mood and depression.

Only two studies looking at infant outcomes have reported maternal mood variables in cocaine using samples. Singer et al (2002) found that maternal psychological distress predicted smaller neonatal head circumference after controlling for the effects of prenatal cocaine use, other drug use and confounding variables [85]. In 1993, Woods and colleagues reported higher maternal depressive scores in cocaine-using women and used the depression scores as a covariate when looking at the effects of prenatal cocaine exposure on infant behavior [92]. This study did not find any differences between cocaine exposed and unexposed infants. The results of this study are consistent with ours in that controlling for maternal depression led to the absence of cocaine main effects.

Our results suggest that early postpartum maternal depressive symptoms have a significant effect on infant neurobehavior and that prenatal cocaine exposure might buffer or alter these effects. One possible explanation for the difference in results in infants exposed to cocaine before birth is the affect of cocaine on developing monoamine systems.

Evidence from rodent, rabbit and primate studies suggest that prenatal exposure to cocaine is associated with decreased dopaminergic activity and a resulting decrease in serotonin receptor density (5-HT1a) in the dorsal raphe nucleus[2, 21]. The action of serotonin in the raphe nucleus exerts an inhibitory effect on the ascending serotonin system into the cortex; therefore a depletion of raphe serotonin results in an increase in serotonin input to the cortex[22, 23, 74]. It is possible that a cocaine induced abundance of serotonin in the cortex would contribute to a “buffer” of serotonin and other monoamines for infants that are exposed to maternal depression.

Other potential mechanisms may be involved in an interaction effect of depression and cocaine, including other neurotransmitter systems, neuroendocrine alterations, and Hypothalamic-Pituitary-Adrenal Axis (HPA) responses. Glucocorticoids, the main end product of the HPA system, mediate monoaminergic activity particularly in the limbic areas of the brain and may be a primary site for further investigation of potential interactions of prenatal cocaine and depression exposure on offspring development.

Other variables we considered include the exclusion of more women in the COC group than the nonCOC group due to inadequate data on the ASI, more days in mental health treatment for women in the COC group in the first 30 days postpartum, and other demographic and neonatal growth variables. More days of treatment for the women in the COC/DEP group than the nonCOC/DEP group could influence newborn behavior if treatment was associated with greater symptom reduction and improvement in mother-infant interactions. However, there was not a significant difference in COC vs nonCOC for meeting depression criteria in the first 30 days postpartum. The interaction results remained significant when reanalyzing the data without participants in outpatient treatment. There was also no significant difference for number of days using cocaine in the postpartum period between those women receiving outpatient mental health treatment and those who were not.

There was limited power in the DEP groups to look at associations with other variables that might also contribute to the results obtained, including maternal anxiety, home environment, time between pregnancies, and general health conditions of the infant at time of testing.

In conclusion, these data provide evidence of an interaction between maternal depression and prenatal cocaine exposure on infant neurobehavioral outcomes. Cocaine exposed infants did not differ from unexposed infants on neurobehavioral scores in depressed groups. Depression effects were only found in infants not exposed to cocaine during gestation. These results suggest a possible buffering effect of cocaine on the effects of maternal depression on infants. Despite the limitations discussed, these findings suggest that past research findings as well as future research on the effects of prenatal substance exposure would benefit from consideration and measurement of maternal mood.