Autonomic Functioning among Cocaine-Exposed Kindergarten-Aged Children: Examination of Child Sex and Caregiving Environmental Risk as Potential Moderators

Abstract

The purpose of this study was to examine the hypothesis that child sex moderates the association between prenatal cocaine exposure (PCE) and autonomic functioning as well as to examine the role that caregiving environmental risk playes in sex differences in autonomic functioning among exposed children. Measures of the parasympathetic nervous system (indexed by respiratory sinus arrhythmia [RSA]) and the sympathetic nervous system (indexed by skin conductance level [SCL]) were obtained from 146 (75 cocaine-exposed, 38 male; and 71 nonexposed, 36 male) children during baseline and a task designed to elicit negative affect (NA). We also examined the role of caregiving environmental risk as a moderator of the association between PCE and autonomic functioning separately for boys and girls. PCE boys had a significantly higher baseline RSA and lower baseline SCL than PCE girls or nonexposed children. Environmental risk also moderated the association between PCE and baseline RSA for boys, but not girls, such that boys with PCE and high environmental risk had the highest baseline RSA. These findings indicate that exposed boys had significantly lower levels of sympathetic activation while at rest. However, for autonomic reactivity, the exposed girls had a larger change in both RSA and SCL relative to nonexposed girls while exposed boys had significantly smaller increases in SCL during environmental challenge. Finally, girls with both PCE and high environmental risk had the highest levels of parasympathetic reactivity during challenge. These results underscore the importance of examining sex differences and considering comorbid environmental risk factors when examining developmental outcomes in cocaine-exposed children and highlight the complexity involved with understanding individual differences in cocaine-exposed populations.

The prenatal period is a time of enhanced vulnerability for both physical and behavioral development due to rapid neurological development. As suggested by the adaptive calibration model (Del Giudice et al., 2011), increased fetal stress as a result of prenatal cocaine exposure and comorbid risks is likely to alter the developmental trajectory to enhance adaptation of the child to their physical environment. Consequently, infants are born with stress response systems that respond in ways that allow the maximal possible adaptation of physiological and behavioral responses to information obtained from stressful environmental conditions. Sex differences have also been suggested by the adaptive calibration model (Del Giudice et al., 2011), indicating that a shift toward physiological dysregulation would be more prevalent among males given the evolutionary benefits of such a pattern under conditions of high prenatal and environmental risk. Although limited, there is some indication that the autonomic functioning of boys is more sensitive to the environmental context in which they are raised. For example, El-Sheik and Hinnant (2011) found decreases in baseline RSA among 8–11-year-old boys who experienced high levels of marital conflict.

Despite the large body of evidence documenting an association between prenatal cocaine exposure (PCE) and a wide range of developmental outcomes, relatively few studies have examined sex differences among cocaine-exposed children. In general, studies have found enhanced developmental vulnerability for boys (Campbell,Shaw & Gilliom, 2000; Gualitieri & Hicks, 1985) and have found that these sex differences are stronger among children with prenatal substance exposure (e.g., Hay, 1997; Moe & Slining, 2001; Weinberg, Tronick, Cohn, & Olsen, 1999. For example, relative to exposed girls, substance-exposed boys show greater cognitive deficits (Moe & Slining, 2001) and a higher incidence of conduct disorder (Wakschlag & Hans, 2002). In studies of children with PCE, findings of sex differences are mixed. Whereas one study found the highest level of externalizing behaviors in 10 year-old boys with PCE (Bennet et al., 2013), another study found that parents reported that cocaine exposed adolescent girls, but not boys, had more difficulties with behavioral regulation at age 12 (Minnes et al., 2016), perhaps due to earlier social maturation of girls and earlier engagement with risky peers. In the same sample as used in the current study, we have found that substance-exposed boys have higher-levels of autonomic dysregulation during frustration during infancy (Schuetze, Eiden & Danielewicz, 2009). Specifically, we found that exposed boys had an increase in RSA during a task designed to elicit NA rather than the typical pattern of RSA suppression. Taken together, these studies suggest that sex differences among substance exposed children may be sensitive to type of prenatal risk and may differ across various developmental outcomes. To date, it is unclear if sex moderates the association between cocaine exposure and autonomic regulation in childhood and if these associations are further moderated by levels of caregiving environment risk. Based on previous literature regarding greater biological vulnerability among males, we hypothesized that the association between PCE and autonomic dysregulation would be stronger for boys compared to girls, and stronger for cocaine exposed boys experiencing higher levels of caregiving environmental risk than nonexposed boys or cocaine exposed boys experiencing lower levels of risk.

Cocaine binds to monoaminergic transporters and prevents the re-uptake of monoamines into the presynaptic cell, resulting in higher concentrations of norepinephrine, serotonin, and dopamine in the synaptic cleft and excess stimulation of dopamine receptors (Gawin & Ellinwood, 1988; Nassogne, Evrard, & Courtoy, 1998). The dopamine system develops early in gestation and is associated with regulatory processes (Robbins, 1997; Tucker & Williamson, 1984; Mayes, 2002). Thus, maternal cocaine use during pregnancy may have a teratological effect on the development of regulatory processes in the developing fetus.

Behaviorally, studies have demonstrated that prenatal cocaine exposure is associated with deficits in arousal regulation during cognitive and emotional tasks among cocaine-exposed children from the neonatal period into adolescence (e.g., Accornero et al., 2011; Bendersky, Bennet & Lewis, 2006; Lester, et al., 2012). There has been growing interest in studying biological markers associated with arousal regulation, including the autonomic nervous system. The ANS is comprised of the sympathetic (SNS) and parasympathetic (PNS) nervous system. The SNS initiates the “fight/flight” response by increasing heart rate and respiration. In contrast, the PNS has an inhibitory effect on the SNS and its role is to maintain homeostasis and to regulate recovery following stress by decreasing heart rate and respiration. PNS activity is often assessed by respiratory sinus arrhythmia (RSA), the heart rate variability at the frequency of respiration, which is thought to index the neural control of the heart via the vagus nerve (Porges, 2007). Two of the most commonly used measures of autonomic regulation are baseline RSA and RSA reactivity (changes in RSA during environmental challenge; Bornstein & Suess, 2000; Calkins, 1997; Porges, 1996). Baseline RSA is a measure of the child’s ability to maintain physiological homeostasis during periods of rest and readiness to respond to challenge, with moderate levels being more adaptive than over or under responsive baseline levels (El-Sheikh & Erath, 2011). In response to environmental challenge, the vagal system optimally responds by functioning as a brake (Porges, Doussard-Roosevelt, Portales, & Greenspan, 1996) and decreasing RSA (suppression). Moderate RSA suppression is believed to be functional and to reflect the ability of the child to interact appropriately with the environment (Bornstein & Suess, 2000; Porges, 1996). However, unresponsive or greater RSA reactivity may index increased emotional lability that is indicative of a “fight or flight” response (Beauchaine, 2001).

Most research examining stress reactivity in young children has focused on RSA or global measures of autonomic functioning like heart rate without specific assessments of the SNS. If withdrawal of the PNS is not sufficient to manage a stressor, SNS activity typically increases in order to prepare the body for more active stress responses. SNS functioning can be measured by assessing levels of skin conductance. SNS activity leads to excitation of the palmer exocrine sweat glands and, thus, increased conductivity in the surface of the skin in response to perceived threat (Fowles, Kochanska & Murray, 2000). Skin conductance levels (SCL) also index the behavioral inhibition system, a neurophysiological motivation system that leads to inhibition during negative-affect arousing situations (Beauchaine, 2001). Low SCL is related to reduced behavioral inhibition or fearlessness during aversive circumstances (Beauchaine, 2001; Gao et al., 2010; Raine, 2002), whereas high SCL indicates increased sympathetic functioning and has been associated with increases in anxiety and fearfulness (Beauchaine, 2001).

Increasingly, there have been calls to examine more than one biomarker of regulatory processes (e.g., Berntson et al., 1994; Buss, Jaffee, Wadsworth & Kliewer, 2018). While a pattern of reciprocal sympathetic and parasympathetic activation is often observed during environmental challenge considerable individual differences in sympathetic and parasympathetic reactivity to stressors have been observed (Berntson et al. 1994; Salomon, Matthews & Allen, 2000). In fact, measures of PNS and SNS reactivity are often not correlated, further underscoring the independence of these two autonomic measures.

To date, the preponderance of studies assessing autonomic functioning among children with PCE have focused on measures of heart rate and RSA in the neonatal period and have yielded mixed results, with some studies indicating higher PNS activity during rest among infants with PCE (Mehta et al., 1993; Regalado, Schechtman, Del Angel, & Bean, 1996; Regalado, Schechtman, Khoo, & Bean, 2001;Silvestri, Long, Weese-Mayer, & Barkov, 1991) while other studies have not found any effects of maternal cocaine use during pregnancy on infant baseline HR or RSA (DiPietro, Suess, Wheeler, Smouse & Newlin, 1995; Mehta et al., 1993). Studies assessing RSA reactivity among infants and young children with PCE typically find no RSA responses during environmental challenge (DiPietro, Suess, Wheeler, Smouse & Newlin, 1995; Schuetze, Eiden & Coles, 2007; Schuetze, Eiden, & Danielewicz, 2009; Sheinkopf et al., 2007), however findings are different when assessing RSA reactivity among older children with exposure to other substances. For example, one study found greater RSA reactivity among exposed 7–12-year old children with prenatal exposure to opiates and alcohol (Suess, Newlin, & Porges, 1997). The inconsistent findings may be due to differences in the prenatal exposure history of the children but it is also possible that the differences in the age of assessment may account for some of these inconsistencies. Finally, differences may also be due to lack of consideration of environmental risk in studies of prenatal exposure and ANS.

Because the neurodevelopmental insult associated with prenatal substance exposure and other prenatal risks is likely to be associated with additional risk factors during childhood, the impact of PCE on developmental outcomes should also be considered in the context of early environmental risk (Fisher et al., 2011). Indeed, studies of prenatal cocaine exposure highlight the importance of considering multiple environmental risks associated with maternal cocaine use in pregnancy (e.g., Brown et al., 2004). These include continued maternal substance use, greater caregiving instability, greater maternal psychopathology symptoms and exposure to violence (Eiden, Veira, & Granger, 2009). Mothers who use substances in pregnancy often continue to use in the postnatal period (Eiden, Schuetze, & Coles, 2011). Children of substance using women also experience greater instability in caregiving situations highlighted by separations from their primary caregivers, changes in caregiving situations, lack of male caregiver involvement, and lack of caregiving routines (Brown et al., 2004). In addition, women who continue to use in pregnancy often experience greater psychiatric symptom and these have significant implications for the quality of caregiving experienced by the child (Eiden, Foote, & Schuetze, 2007). Finally, substance using mothers and their children experience greater exposure to violence, and often living in contexts that have high risk for violence (Bada et al., 2011). Given the co-occurring nature of these risks, the cumulative number of these risk factors may be more predictive of dysregulated autonomic functioning compared to any one specific risk condition alone (Seifer, 1995). Thus, it is important to attempt to distinguish between the effects of prenatal insults versus postnatal environmental risk factors associated with maternal substance use during childhood on development

To date, the preponderance of studies have examined the effects of prenatal substance exposure on developmental outcomes while controlling for the influence of other risk factors. However, studies exploring the impact of other prenatal insults (i.e., prematurity) on development have found interactions between biological vulnerability and associated environmental risk. For example, study examined the effects of prenatal biological insults and environmental risk on long-term developmental outcomes among adolescents who were born prematurely (Levy-Shiff, Einat, Mogilner, Lerman, & Krikler, 1994). Among adolescents who were born prematurely with very low birthweight, lower SES predicted increased hyperactivity and visual-motor coordination. However, SES did not predict outcomes among adolescents who were born full term at a normal birth weight, indicating an interaction between prenatal vulnerability and environmental risk. Similarly, Landry, Smith, Swank, Assel and Vellet (2001) found that both preterm and full-term infants with mothers who had high levels of parental responsivity showed the expected cognitive gains by age 12-months. However, among infants with less responsive mothers, preterm children has significantly less cognitive gain during the first year of life compared to full-term children indicating that the combination of prenatal insult with environmental risk was associated with more negative developmental outcomes than prenatal vulnerability alone. Moreover, not all children of cocaine using mothers experience the same environmental adversities. Understanding if associations between PCE and the autonomic nervous system differ for varying levels of environmental risks for boys and girls may present us with a more nuanced understanding of risk heterogeneity in substance exposed samples.

PCE often co-occurs with numerous other risk factors which may impact development. Prenatal risk factors that tend to be comorbid with PCE include use of other substances such as cigarettes, alcohol and marijuana, and demographic risks such as low socioeconomic status or single parenthood. The nested nature of these risk factors with PCE makes it difficult to separate the effects of PCE from other risks, making it important to consider these other risk factors when examining if PCE accounts for unique variance in autonomic functioning.

The purpose of this study was to examine whether child sex moderated any associations between PCE and measures of baseline autonomic functioning and autonomic reactivity at early school age. Based on previous findings of sex differences among cocaine-exposed children including those in this sample during infancy (Schuetze, Eiden, & Danielewicz, 2009), we hypothesized that boys with PCE would exhibit increased autonomic dysregulation including lower baseline RSA, less RSA reactivity and more SCL reactivity. We were also interested in the role that caregiving risk played in these sex differences. However, due to the moderate sample size, we were unable to directly examine three-way interactions between PCE, child sex, and environmental risk. Thus, we examined the role of a cumulative measure of caregiving environmental risk as a moderator of the association between PCE and autonomic functioning separately for boys vs. girls. Given the existing body of literature that has indicated that boys are more vulnerable to a range of developmental influence, we hypothesized that the combination of PCE and high caregiving environmental risk would be associated with autonomic dysregulation for boys but not girls.

Method

Participants

Mother–infant dyads (birth years ranged from 2000–2005) were recruited postpartum from two local hospitals into a longitudinal study of maternal substance use and child development. Mothers ranged in age from 18 to 42 (M = 29.69, SD = 5.95). 74% of mothers were African-American, 18% were Caucasian, 6% were Hispanic and the remaining were other. 76% were receiving Temporary Assistance for Needy Families and 90.3% were single at recruitment. Women were classified as either cocaine-users or ‘abstainers’. The two groups were matched on maternal education, age, race/ethnicity and infant sex; 51.9% of the infants were male (PCE: 53.2%; nonexposed: 50.6%). Additional exclusionary criteria were: maternal age under 18 years, and prenatal illicit substance use other than cocaine or marijuana. The study received approval from the institutional review boards of the hospitals and primary institutions of the authors. Participants were compensated for their time in the form of gift certificates, checks and child toys.

In the circumstance of a custody change, the legal guardian was asked to participate. By kindergarten, 55 children in the PCE group and nine children in the nonexposed group had been removed from parental care and placed in nonparental care. All assessments were conducted with the primary caregiver of the child at that time, although for ease of presentation the terms mother and maternal are used throughout the article when referring to the primary caregiver. The primary caregiver was identified as the adult who had legal guardianship of the child and accompanied the child to all appointments. The average child age at kindergarten was 5.52 years (SD = 0.36, range: 4.8–7.0).

Maternal and child assessments were conducted at 4–8 weeks as well as at 7, 13, 24, 36, and 48 months, and at kindergarten age, with additional maternal interviews conducted at 18, 30, 42, and 54 months of child ages. Measures obtained at 4–8 weeks and at kindergarten were included in the current analyses. Of the 216 children recruited into the study, 7 indicated that they were no longer interested in participating in the study, 33 did not show up for the Kindergarten assessment after repeated reschedules, 14 (6 PCE) were dropped (severe medical problems such as Fetal Alcohol Syndrome and Down Syndrome), and 1 child was unable to participate due to custodial issues. 8 children (4 PCE) refused to allow the placement of the physiological recording equipment. An additional 7 children did not have complete physiological data due to equipment failure (n =3) or research assistant error (n = 4). Thus, the final sample was 146 (75 PCE, 71 nonexposed) dyads. There were no significant differences between families with complete versus missing data at Kindergarten on demographic or substance use variables.

Procedure

All mothers were screened after delivery for initial eligibility and matching criteria. Interested and eligible mothers were given detailed information about the study and asked to sign consent forms. About 2 weeks after delivery, mothers were contacted and scheduled for their first laboratory visit, which took place at the time that their infant was approximately 4–8 weeks old. All visits consisted of a combination of maternal interviews, observations of mother-child interactions, and child assessments. In the circumstance of a change in custody arrangements, the person who had legal guardianship of the child was contacted and asked to participate. Biological mothers were interviewed at the 4–8-week assessment in addition to the foster mother in order to obtain accurate information about prenatal substance use. Once a family was recruited into the PCE group, the closest matching nonexposed group family (based on maternal education, race and ethnicity, and infant gender) was recruited. However, a significantly higher proportion of mothers in the nonexposed group declined participation or withdrew before formal enrollment, resulting in a smaller number of families in the control group. Mothers in the comparison group reported not having used any illicit substances other than marijuana. They also tested negative for cocaine or illicit substances other than marijuana, based on urine and hair analysis results. The final sample consisted of 216 (116 PCE, 100 nonexposed) mother-child dyads (see Finger, Schuetze, & Eiden, 2014 for additional details on sample recruitment).

Identification of Maternal Substance Use

Cocaine status was determined through maternal self-report, chart review which provided results of urine toxicology, and maternal hair analysis. Urine toxicologies were routinely conducted by participating hospitals. Self-reports of maternal substance use before, during and after pregnancy were obtained using the Timeline Follow-back Interview administered to the biological mother (TLFB; Sobell, Sobell, Klajner, Pavan, & Basian, 1986) at each time point beginning at 1-month of infant age. The TLFB yielded data about the average number of days per week cocaine was used, the average number of joints of marijuana, the average number of cigarettes (tobacco), and the average number of standard drinks consumed per week, as well as the mean standard drinks per drinking day and number of alcohol binges (5 or more standard drinks) during pregnancy and postpartum.

Urine toxicologies consisted of standard screening for drug levels/metabolites of cocaine, opiates, benzodiazepines, and tetrahydrocannabinol. Urine was rated positive if the quantity of drug/metabolite was >300 g/ml. Hair samples were collected from all mothers at the 1-month visit and sent to the Psychemedics Corporation for radioimmunoanalyses (RIAH). Hair samples were screened for cocaine followed by a gas chromatography/mass spectrometry (GC/MS) confirmation for positive cocaine screens.

Of the mothers in the cocaine group, 46% (n = 36) had positive urine toxicologies at delivery, 61% (n = 48) had hair samples that tested positive for cocaine during pregnancy, and 84% (n = 66) admitted having used cocaine in the brief self-report screening instrument administered after delivery.

The majority of mothers in the cocaine group met multiple criteria for inclusion into the cocaine group. Mothers in the comparison group reported not having used any illicit substances other than marijuana and did not test positive for cocaine or other illicit substances other than marijuana on any biomarker.

Assessment of growth and risk status

Three measures of growth were taken by obstetrical nurses in the delivery room: birth weight (gm), birth length (cm), and head circumference (cm). Medical chart review at the time of recruitment was used to complete the Obstetrical Complications Scale (OCS; Littman & Parmelee, 1978), a scale designed to assess perinatal risk factors. Higher numbers indicate a more optimal score.

Assessment of Autonomic Functioning

The physiological assessment of reactivity and regulation was recorded during a 2-minute baseline period, and a negative affect (NA) episode by examiners blind to child group status. Recording of the physiological data began once the child was observed to be in a stable, quiet, alert state which was induced by having the child watch a 2-minute segment of a neutral videotape. The NA paradigm consisted of the “Wrong Gift”, a frustration paradigm from the school age version of the Laboratory Temperament Assessment Battery (LABTAB; Gagne et al., 2011). Children were asked to rank order five prizes from best to worst (e.g., slinky, yo-yo, top, rubber toy, worn broken white crayon) and told that they would get to keep one of the prizes, left in the room by themselves for 2 minutes while the gift was wrapped. They were then presented with the broken white crayon (rated as the worst prize by most children). After a minute, children were told we had made a mistake and were asked to pick out the prize they liked the best.

A Bioamp (James Long Company, Caroga Lake, NY) recorded respiration, electrocardiograph and skin conductance data. Disposable electrodes were triangulated on the child’s chest. A respiration bellows was placed at the level of the zyphoid process to measure inspiration and expiration. Ag/AgCl electrodes, filled with isotonic citrate salt electrode gel with gel contact area limited to a 1 cm diameter circle by double-sided adhesive collars, were attached to the distal phalanges of the participant’s non dominant hand. IBI Analysis software (James Long Company, Caroga Lake, NY) was used to process the heartrate data and to calculate RSA. The software synchronizes with respiration and is relatively insensitive to arrhythmia due to tonic shifts in heartrate, thermoregulation, and baroreceptor.

Average RSA and SCL was calculated for the 2-minute baseline period and for the NA task which begain once the child saw the broken crayon. Change scores from baseline to the NA task, were calculated for RSA and for SCL to assess reactivity. Negative scores indicate a decrease in RSA or SCL. A negative RSA Reactivity score (RSA suppression) is indicative of more optimal parasympathetic reactivity and a negative SCL Reactivity score indicates less sympathetic activity in response to frustration.

Assessment of Other Prenatal Substance Exposure

The prenatal substance exposure risk score was comprised of 9 indicators representing prenatal cigarette, marijuana, and alcohol exposure for each trimester of pregnancy. Substance use data was extracted from the TLFB (described above) in order to calculate the average number of cigarettes, joints, and standard drinks consumed per week during each trimester of pregnancy. Those averages were then divided by the sample maximum in order to create a proportion score for each of the 9 indicators. For the final risk index, the 9 proportions scores (cigarettes, alcohol, and marijuana in the first, second, and third trimester) were averaged. The mean composite substance exposure risk score for the sample was .046 (SD = .078), with a range of .00 to .51.

Assessment of Environmental Risk

A composite caregiving environmental risk was computed from the following measures collected at each time point with more specific details published previously (e.g., Eiden, Godleski, Schuetze, & Colder, 2015). Maternal psychopathology symptoms were assessed using the Brief Symptom Inventory (BSI; (Derogatis, 1993), consisting of 53 items rated on a five-point scale. A positive symptom distress index was computed by summing all items and dividing by number of items endorsed with a positive response and higher scores indicating higher maternal psychopathology at each of 7 time points. Maternal exposure to violence was assessed at every time point using the Timeline Followback Interview (TLFB, Sobell & Sobell, 1996) as used previously (e.g., Mignone, Klostermann, & Chen, 2009). The total number of days women witnessed, experienced, or perpetrated violence were summed within each time point and was dummy-coded (i.e., 0, 1; distribution was bimodal) at each time point. Caregiving instability was assessed using a structured caregiver interview (Platzman, Coles, Lynch, Bard, & Brown, 2001), and administered to the child‟s caregiver by a trained examiner. Items included at each time point were: no male adult in the household, major separations from primary caregiver/custody changes, and child does not see primary caregiver on a regular basis. Additional items until 24 months included baby is fed and sleeps significantly less than average. At 36, 48 months, and kindergarten age, an additional item of lack of routine health care was included. The caregiving instability score within each time point was a count variable of each dummy coded risk. Postnatal substance use risk included average number of joints/week, standard drinks/week, and cigarettes/week from the TLFB administered at each time point. The cumulative environmental risk variable was created by computing a count variable which included the BSI, exposure to violence, and caregiving instability for each time point that each variable was assessed based on scores in the upper quartile. The risk variable also included postnatal substance use with cut-offs for alcohol (greater than or equal to 4 drinks/day) and marijuana (greater than or equal to one day/week) based on prior research (Yumoto, Jacobson & Jacobson, 2008), and cut-off for cigarettes based on an average of higher than 10 cigarettes/day. BSI, postnatal substance use risk, and caregiving instability were assessed at seven time points (i.e., 1, 7, 13, 24, 36 48, and 60 months), exposure to violence was assessed at eleven time points (i.e., 1, 7, 13, 18, 24, 30, 36, 42, 48, 54, and 60 months). Thus, scores for cumulative caregiving environmental risk could theoretically range from 0 to 32. The scores for cumulative caregiving environmental risk ranged from 0 to 24 in this sample. Children were grouped into the high environmental risk and low environmental risk groups using the median of 5.69 as the cut-off.

Analytic Strategy

Analyses of potential confounds were conducted using correlations or analysis of variance (ANOVAs) as appropriate. Confounds with significant bivariate associations with autonomic functioning were included as covariates in all remaining analyses (if they were associated at p < .10). PCE by child sex interactions in autonomic functioning were examined using ANCOVAs for baseline and reactivity measures. Chi Square analyses were used to examine group status by sex interactions for the patterns of autonomic functioning. Environmental risk was dichotomized using a median split. ANCOVAs were then used to examine the interactions between postnatal environmental risk and PCE on autonomic measures separately for boys and girls.

Results

Descriptive Statistics

Individual analyses of variance (ANOVA) were conducted to examine group differences for demographic and prenatal substance exposure variables for boys and girls in the two exposure groups (see Table 1 for results from univariate analyses).

Table 1.

Group Differences

Exposure Group:	Cocaine (PCE)		Nonexposed (NE)		F value for Sex Differences	F value for Exposure Status Group Differences
Child Sex	Boys	Girls	Boys	Girls
	M (SD) n=39	M (SD) n = 35	M (SD) n = 37	M (SD) n = 35
Demographics:
BM age	31.7 (6.3)	30.44 (5.81)	27.44 (6.01)	28.11 (5.28)	.36	16.56**
BM parity	4.32 (2.88)	4.1 (2.17)	3.14 (1.55)	3.18 (1.78)	.001	.11.90**
Socioeconomic Status	.44 (.14)	.39 (.21)	.47 (.22)	.37 (.22)	5.92*	1.01
Prenatal Substance Use:
# days cocaine/week	1.24 (1.71)	0.68 (1.39)	0	0	3.31	38.04**
# cigarettes/week	46.83 (52.19)	29.48 (45.32)	10.9 (25.47)	14.89 (26.33)	1.85	26.5**
# standard drinks/week	6.25 (14.61)	1.36 (6.08)	.19 (.80)	.20 (.85)	4.85*	10.61**
# joints/week	1.52 (3.75)	1.07 (4.50)	0.32 (1.19)	2.62 (10.26)	1.36	.047
Cumulative Environmental Risk	6.68 (4.74)	7.07 (4.39)	6.19 (4.56)	5.66 (4.41)	1.66	38.85**
Birth outcomes:
Gestational age (weeks)	38.4 (1.92)	38.7 (1.95)	39.4 (1.37)	39.2 (1.41)	0.89	6.91*
Birth weight (gm)	2934.71 (579.54)	2911.52 (485.52)	3289.08 (448.97)	3379.82 (554.07)	.23	33.6**
Birth length (cm)	48.6 (2.77)	47.72 (3.29)	49.5 (3.21)	50.47 (2.5)	.04	20.05**
Head circumference (cm)	33.41 (1.36)	32.77 (2.52)	33.49 (1.5)	33.73 (1.27)	.67	4.62*
Child Age at Kindergarten Visit	67.14(4.7)	65.85(4.48)	67.14(4.7)	66.53(4.73)	1.644	.16
Autonomic Measures:
Baseline RSA	.13 (.08)	.11 (.06)	.09 (.05)	.10 (.08)	.08	4.32*
RSA Reactivity	−.014 (.04)	−.014 (.03)	−.004 (.03)	.10 (.08)	.01	2.89*
Baseline SCL	5.71 (4.06)	5.80 (2.76)	6.55 (4.36)	5.46 (3.36)	.54	.78
SCL Reactivity	.82 (.99)	.96 (.88)	1.21 (1.58)	.83 (.73)	.43	.56

^*p<.05

^**p<.01

Note: BM: biological mother

Because physiological responses may be influenced by the Law of Initial Value (Lacey, 1956; Wilder, 1956), the association between baseline RSA and RSA Reactivity was examined. Baseline RSA was significantly associated with RSA during the frustration task (r =−.43, p = .001). Thus, consistent with the Law of Initial Value, RSA change was adjusted for baseline levels by including baseline RSA as a covariate in the analysis of RSA Reactivity. Similarly, the association between baseline SCL and SCL Reactivity was examined. Baseline SCL was significantly associated with SCL during the frustration task (r=.40, p=.001). Thus, baseline SCL was included as a covariate in all analyses of SCL change.

Potential covariates were examined using analyses of bivariate correlations. We first examined associations between demographics (matenral age, parity, maternal education level, family income) and the autonomic measures. Family income was associated with Baseline RSA, r = .21, p = .008, and parity was associated with RSA Reactivity, r = −.21, p =.01. We then examined associations between perinatal risk measures (gestational age, birth weight, birth length and birth head circumference). None of these were associated with the autonomic measures. Maternal prenatal alcohol and marijuana use were positively associated with baseline RSA and SCL and maternal prenatal alcohol and cigarette use were negatively associated with RSA reactivity (all r’s <.05). In addition, the composite prenatal substance exposure variable was positively associated with baseline RSA, r = .16, p = .05, and negatively associated with RSA Reactivity, r= −.17, p = .044. Consequently, income, parity and the composite variable of other prenatal substance use were included as covariates in the respective analyses of autonomic functioning.

PCE Group by Sex Differences in RSA

Separate 2 (group status) × 2 (child sex) ANCOVAs were performed on the RSA variables: Baseline RSA and RSA Reactivity. The covariate, prenatal substance use composite, was marginally related to baseline RSA, F(1, 139) = 2.87, p = .092, eta²_P = .02, but not to RSA Reactivity, F(1, 139) = 0.37, p = .54, eta²_p = 0.3. The analyses on baseline RSA and RSA Reactivity both yielded interaction effects for group status and child sex, F(1, 139) = 4.06, p = .046, eta²_p = .0.3, F(1, 139) = 4.01, p = .049, eta²_p = 0.3, respectively (see Figures 1 and 2). PCE boys had a significantly higher baseline RSA than PCE girls or nonexposed boys or girls, however, for RSA Reactivity, the exposed girls had a larger change in RSA relative to nonexposed girls.

Figure 1.

Interaction effect for child sex and cocaine exposure group status on Baseline RSA.

F(1, 139) = 4.06, p = .046, eta²_p = .03. PCE boys had a significantly higher baseline RSA than PCE girls or nonexposed boys or girls.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

Figure 2.

Interaction effect for child sex and cocaine exposure group status on RSA Reactivity.

F(1, 139) = 4.01, p = .049, eta²_p = .03. PCE girls had a larger change in RSA relative to nonexposed girls.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

PCE Group by Sex Differences in SCL

Similarly, separate 2 (group status) × 2 (child sex) ANCOVAs were performed on the SCL variables: Baseline SCL and SCL Reactivity. The covariate, prenatal substance use composite, was not significantly associated with baseline SCL, F(1, 139) = 0.14, p = .71, eta²_p = .001, or SCL Reactivity, F(1, 139) = 0.26, p = .64, eta²_p = .001. The analyses on Baseline SCL and SCL Reactivity both yielded significant interaction effects, for PCE group status by child sex, F(1, 137) = 3.43, p = .049, eta²_p = .03, F(1, 137) = 3.59, p = .043, eta²_p = .03, respectively (see Figures 3 and 4). Nonexposed boys had significantly higher baseline SCL levels relative to exposed boys but there were no differences in baseline SCL between exposed and nonexposed girls. For SCL Reactivity, exposed boys had significantly smaller increases in SCL during environmental challenge than nonexposed boys. However, for girls, it was the nonexposed girls that had a significantly smaller increase in SCL than exposed girls.

Figure 3.

Interaction effect for child sex and cocaine exposure group status on Baseline SCL.

F(1, 137) = 4.29, p = .039,. eta²_p = .04. Nonexposed boys had significantly higher baseline SCL levels relative to exposed boys. No group differences between exposed and nonexposed girls.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

Figure 4.

Interaction effect for child sex and cocaine exposure group status on SCL Reactivity.

F(1, 139) = 4.23, p = .042, eta²_p = .04. Exposed boys had significantly smaller increases in SCL than nonexposed boys. Nonexposed girls had significantly smaller increases in SCL than exposed girls.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

PCE Group by Environmental Risk Interactions

2 (PCE group status) × 2 (level of environmental risk) ANCOVAs were performed on the RSA and SCL variables first for boys and then for girls. There were no significant main or interaction effects for either of the SCL variables for either boys or girls. However, there was a significant interaction effect for Baseline RSA for boys, F(1,71) = 13.766, p = .001, eta²_p = .06 (see Figure 5), but not for girls, F(1,64) = .15, p = .71, eta²_p = .002. The covariate, prenatal substance use composite, was associated with baseline RSA, F(1, 71) = 7.56, p = .008, eta²_p = .02, but not to RSA Reactivity, F(1, 71) = 1.31, p = .26, eta²_p = .02. Boys with PCE and high environmental risk had significantly higher levels of Baseline RSA than boys with PCE and low environmental risk or nonexposed boys with any level of environmental risk. For RSA Reactivity, there was a significant interaction effect for RSA Reactivity for cocaine exposed girls, F(1,71) = 13.766, p = .001, , eta²_p = .16. (see Figure 6), but not for boys, F(1,62) = 3.96, p = .049, eta²_p = .06. The covariate, prenatal substance use composite, was not associated with baseline RSA, F(1, 69) = 2.49, p = .12, eta²_p = .038, or RSA Reactivity, F(1, 69) = 1.1, p = .29, eta²_p = .016. Girls with both PCE and high environmental risk had significantly larger decreases in RSA Reactivity than any of the other groups.

Figure 5.

Interaction effect for child sex and environmental risk on Baseline RSA for boys.F(1, 71) = 13.766, p = .001, eta²_p = .16. Boys with PCE and high environmental risk has significantly higher levels of Baseline RSA than other groups.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

Figure 6.

Interaction effect for child sex and environmental risk on RSA Reactivity for girls.

F(1, 64) = .13.47, p = .001, eta²_p = .15. Girls with both PCE and high environmental risk had significantly larger decreases in RSA Reactivity than other groups.

Note: PCE=prenatal cocaine exposure; NE = nonexposed

Discussion

A growing body of research has found that cocaine-exposed children are at a higher risk for dysregulation. Although most studies have examined behavioral regulation, several studies have indicated that this dysregulation extends to physiological measures of regulation during infancy and into childhood. Furthermore, evidence has been accumulating suggesting that exposed boys may be particularly vulnerable to the effects of substance exposure. Thus, the purpose of the present study was to examine sex-specific effects of prenatal cocaine exposure on autonomic functioning in kindergarten-aged children. We examined whether child sex moderated the association between prenatal cocaine exposure and measures of autonomic functioning after controlling for other prenatal substance exposure and environmental risk. We also explored the possibility that there may be sex differences in how boys and girls respond to the combination of prenatal cocaine exposure and environmental risk. Our findings highlight the importance of considering the impact of child sex on both baseline autonomic functioning and autonomic reactivity.

First, PCE appeared to exert its effects on baseline RSA in some unexpected ways. Contrary to our hypothesis, boys with PCE had higher baseline RSA than girls or nonexposed boys. Resting RSA is believed to be an indicator of the child’s ability to maintain homeostasis in the presence of minimal exogenous stimulation and the ability to respond to environmental challenge (Beauchaine, 2001; Bornstein & Suess, 2000). However, studies examining RSA in children with prenatal substance exposure have been mixed While some studies have found lower levels of resting RSA among substance-exposed infants at 1- and 7-months of age (Schuetze & Eiden, 2006; Schuetze, Eiden, & Edwards, 2009), other studies have found no association between prenatal substance exposure and baseline RSA (DiPietro et al., 1995) or increased resting RSA (Mehta et al., 1993). Beyond infancy, one study found higher levels of resting RSA among children with prenatal substance exposure (Conradt et al., 2014). Other studies of children in high-risk populations have also found higher levels of baseline RSA to be associated with increased early adversity (e.g., Conradt et al., 2013; Eisenberg et al., 2012) and with increased internalizing and externalizing symptoms among low socioeconomic status African-American children (Kidwell & Barnett, 2007). These studies are consistent with our findings that boys with a combination of PCE and high environmental risk have higher baseline RSA than nonexposed boys or boys with PCE and low environmental risk. There is also evidence that children with PCE employ increased neurobiological resources in the face of demands from their environment (Sheinkopf et al., 2009). Taken together, these studies suggest that higher levels of resting RSA may not always be optimal and may indicate the need for greater PNS activation to maintain homeostasis. Thus, our findings suggest that boys with PCE may be “working harder” to maintain homeostasis which is indicative of increased vulnerability to the effects of PCE relative to girls or nonexposed boys and supports the hypothesis of the adaptive calibration model (DelGiudice et al., 2011) that physiological dysregulation is likely to be more prevalent among boys. On the other hand, given that higher baseline RSA is associated with an enhanced capacity to respond to environmental perturbations (Beauchaine, 2001; Bornstein & Suess, 2000), it may be that higher baseline in the context of higher environmental risk is adaptive.

Exposed boys also had significantly lower levels of resting SCL indicating less sympathetic activation while at rest and lower levels of SCL reactivity. Reduced SCL reactivity is associated with externalizing behaviors, fearlessness and reduced behavioral inhibition during stressful situations (e.g., Beauchaine, 2001; Gao, Raine, Venables, Dawson & Mednick, 2010). Furthermore, some research has indicated that the relation between SCL underarousal and conduct problems is more robust for boys (Beauchaine, Hong, & Marsh, 2008; Calkins & Demon, 2000; Isen et al., 2010). The findings of lower SCL Reactivity among exposed boys may suggest a physiological mechanism underlying the behavioral dysregulation and increased externalizing behaviors frequently observed in substance-exposed children (e.g., Accornero et al., 2011; Bendersky, Bennet & Lewis, 2006; Lester, et al., 2012) Thus, future studies should explore the possible linkage between measures of SNS underactivation and the development of externalizing behaviors among substance-exposed boys.

Our findings regarding sex differences for RSA reactivity indicate that, although both exposed and nonexposed boys and girls showed some suppression of RSA during environmental challenge, it is exposed girls, not boys, that differ from nonexposed boys and girls. Based on our finding of less autonomic reactivity among exposed children in this same sample during infancy (Schuetze, Eiden & Coles, 2007; Schuetze, Eiden, & Danielewicz, 2009), we had hypothesized that there would be “blunting” of the parasympathetic response among exposed children and that this pattern would be enhanced among the exposed boys. Instead the exposed boys showed patterns of PNS responses that were similar to nonexposed girls and nonexposed boys. RSA has been shown to increase signifantly over time during infancy and early childhood (Alkon et al., 2006; Bar-Haim, Marshall, & Fox, 2000; Bornstein & Suess, 2000; Calkins & Keane, 2004) with some studies (Alkon et al., 2003: Marshall & Stevenson-Hinde, 1998), but not all (El-Sheikh, 2005; Salomon, 2005), indicating that RSA levels continue to increase in children up to 7 years of age in low risk populations. There is also considerable evidence that resting RSA and RSA reactivity are susceptible to environmental influence. Thus, our findings that the exposed boys who showed increases in RSA during environmental challenge as infants no longer differ from nonexposed boys in kindergarten suggest that the developmental trajectories for RSA differs between high risk and low risk girls and boys, perhaps due to socialized environmental differences. Furthermore, other studies have also found greater RSA reactivity among exposed infants (Haley, Handmaker & Lowe, 2006) and young children (Conradt et al., 2014) with one study finding the greater RSA reactivity among opiate exposed boys during middle childhood (Suess, Newlin & Porges, 1997). One explanation for the mixed findings in the existing literature may be the failure of many studies to consider individual differences in adversity among exposed children. Our findings that the greatest RSA Reactivity was among exposed girls who also had high levels of postnatal environmental risk suggest girls may be particularly vulnerable to the combination of prenatal substance exposure with environmental adversity early in life. Increased reactivity in the context of high environmental risk has beenc conceptualized as a risk factor for externalizing behaviors (Boyce & Ellis, 2005) and previous research has found that, the combination of increased reactivity with other risk (i.e., social and/or cognitive risk) is associated with physical aggression in girls (Sijtsema, Shoulberg & Murray-Close, 2011). Thus, future research should explore the possibility that increased parasympathetic reactivity when paired with more adverse environments may be one pathway to externalizing behaviors among girls with PCE.

Importantly, the present findings provide additional evidence that child sex should be considered when exploring developmental effects of prenatal substance exposure. These results clearly indicate an interaction between prenatal substance exposure and child sex in measures of autonomic functioning, both at rest and in reaction to environmental challenge. However, unlike previous findings that suggest boys are more vulnerable to the effects of prenatal substance exposure, our findings suggest that the interaction between child sex and prenatal substance exposure is more complex. Additional research using person-centered methods is needed to see if the patterns of SNS and PNS functioning among exposed children predicts future outcomes such as externalizing behaviors later in childhood and into adolescence.

In interpreting our findings, it is important to note some limitations of this study. First, although care was taken in the present study to identify substance use in this sample, the accurate assessment of substance use is always difficult, particularly among pregnant women. Pregnant women are often hesitant to divulge information regarding the use of substances during pregnancy. To address this issue, multiple indices of substance use were used including self-report using the reliable TLFB interview as well as analysis of medical records, maternal hair samples and urine samples. Each of these measures has its own limitations. However, when used in combination, they greatly increase the likelihood of accurately identifying prepartum substance use. Second, our measures of autonomic regulation were limited to one frustration task. It is unclear if these findings would extend to other contexts including other frustration taks or tasks designed to elicit other emotions such as fear. It is also important to note that children with PCE had higher levels of environmental risk than nonexposed children and, thus, the analyses of interactions between PCE and environmental risk should be considered with caution. At the same time, higher levels of environmental risk are consistently reported in samples of children with PCE underscoring the ecological validity of these findings. Finally, due to our relatively small sample size, we were unable to examine interactions between prenatal cocaine group status, child sex and environmental risk in the same analyses. Instead, we chose to examine PCE by environmental risk separately for boys and girls.

This study also had numerous strengths including its prospective design, diverse sample, retention of participants, and statistical control for potential confounds. A particular strength of the current study is the consideration of the impact of cumulative environmental risk given that many children are exposed to nested risk factors rather than experiencing risk factors individually. In addition, the present study adds to a growing literature on the role of cumulative risk in the development of regulatory processes (e.g., Appleyard et al., 2005; Ashford et al., 2008; Deater-Deckard et al., 1998).

In conclusion, our results suggest children who are biologically vulnerable due to prenatal exposure to cocaine may be particularly sensitive to cumulative environmental risk. Thus, intervention efforts focusing on reducing maternal substance use during pregnancy as well as reducing environmental risk may be productive. Furthermore, the combination of PCE and cumulative environmental risk appears to impact autonomic functioning differently for boys and girls. In general, increased sympathetic reactivity and baseline autonomic measures were associated with prenatal cocaine exposure for boys possibly indicating enhanced sympathetic activation during challenge as well as the need for boys to utilize more autonomic resources to maintain physiological homeostasis. Conversely, parasympathetic reactivity was higher among girls with higher environmental risk and prenatal exposure to cocaine, indicating potential risk for higher externalizing problems (Boyce & Ellis, 2005) and aggression (Sijtsema, Shoulberg & Murray-Close, 2011). Thus, future research examining autonomic regulation among substance-exposed children should consider systematically exploring the role of child sex.

Highlights.

• Child sex interacts with prenatal cocaine exposure (PCE) for autonomic functioning

• PCE boys had significantly lower levels of sympathetic activation while at rest, less sympathetic reactivity and have the need for greater parasympathetic activation to maintain homeostasis while at rest

• PCE girls with high environmental risk had the highest levels of parasympathetic reactivity during challenge.