Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 May 1.
Published in final edited form as: Child Dev. 2012 Nov 19;84(3):826–840. doi: 10.1111/cdev.12008

Foster Mother-Infant Bonding: Associations Between Foster Mothers’ Oxytocin Production, Electrophysiological Brain Activity, Feelings of Commitment, and Caregiving Quality

Johanna Bick 1, Mary Dozier 2, Kristin Bernard 3, Robert Simons 4, Damion Grasso 5
PMCID: PMC3582747  NIHMSID: NIHMS408337  PMID: 23163703

Abstract

This study examined the biological processes associated with foster mother-infant bonding. In an examination of foster mother-infant dyads (N = 41, mean infant age = 8.5 months), foster mothers’ oxytocin production was associated with their expressions of behavioral delight toward their foster infant and their average P3 response to images of all infant faces in the first two months of the relationship. Three months later, foster mothers’ oxytocin production was still associated with delight toward their foster infant and was also specifically associated with their P3 response to an image of their foster infant. Similar to biologically-related mothers and infants, oxytocin appears to be associated with foster mothers’ brain activity and caregiving behavior, with patterns suggestive of bond formation.

Oxytocin, a neuropeptide involved in social bonding, is associated with mother-infant relationship quality in both human and non-human mammals (Galbally, Lewis, Van IJzendoorn, & Permezel, 2010; Kendrick, Keverne, & Baldwin, 1987; Neumann, 2008). Among biologically-related mothers and infants, maternal peripheral oxytocin levels are associated with brain activity in response to infant stimuli (Strathearn, Fonagy, Amico, & Montague, 2009), maternal feelings regarding their infants (Feldman, Weller, Zagoory-Sharon, & Levine, 2007), and the quality of caregiving behavior displayed toward their infants (Gordon, Zagoory-Sharon, Leckman, & Feldman, 2010a; Gordon, Zagoory-Sharon, Leckman, & Feldman, 2010b). However, to our knowledge, researchers have not examined the role of oxytocin among non-biologically-related human mother-infant dyads. Therefore, it is unclear whether oxytocin production has relevance for the formation and quality of non-biological mother-infant bonds, such as those that occur in foster care. Given that foster infants are especially in need of high quality caregivers, investigating the biological correlates of alloparental bonding and caregiving behavior seems critical.

Oxytocin and Non-human Mammal Maternal Care

Oxytocin is a neuropeptide produced in the hypothalamus and is released centrally in the brain and peripherally into the bloodstream (Landgraf & Neumann, 2004). Centrally-released oxytocin acts as a neurotransmitter or neuromodulator. When secreted into the peripheral blood system, oxytocin acts as a hormone and is best known for its role in facilitating uterine contractions during the birth process and milk ejection during nursing (Carter, 1998). The association between central and peripheral release of oxytocin is controversial and not fully understood. A minute amount of oxytocin (less than 1%) has been found to enter the brain from the bloodstream (Landgraf, Ermisch, & Hess, 1979; Mens, Witter, van Wimersma, & Greidanus, 1983) but it is unclear whether this small amount leads to significant physiological changes (Ermisch et al., 1985). Several studies among humans and animals indicate that central and peripheral oxytocin release occurs independently (Neumann, 2007). Yet, additional studies suggest that the release of central and peripheral oxytocin can be coordinated at times (Burri, Heinrichs, Schedlowski, & Kruger, 2008; Landgraf & Neumann, 2004; Wotjak, Ganster, Kohl, Holsboer, Landgraf, & Engelmann, 1998).

Among several species of non-human mammals, oxytocin is necessary for the onset and expression of maternal behavior (Farbach, Morrell, & Pfaff, 1984; Kendrick et al., 1987; Pedersen & Prange, 1979; Van Leengoed, Kerker, & Swanson, 1987). Furthermore, a proliferation of oxytocin receptors in brain regions associated with maternal care has been associated with the quality of maternal behavior expressed toward offspring in several rodent species (Champagne, Diorio, Sharma, & Meaney, 2001; Insel, 1986; Pedersen, Vadlamudi, Boccia, & Amico, 2006).

Studies involving non-human mammals suggest that oxytocin plays a role in the expression of alloparental care. Among virgin juvenile female prairie voles, the number of oxytocin receptors in the brain was found to be associated with the quality of alloparental care exhibited toward vole pups (Olazábal & Young, 2006). Oxytocin receptor density has also been associated with the quality of adult female rats’ maternal behavior displayed toward biological and non-biological cross-fostered young (Champagne et al., 2001).

Oxytocin and Human Mother-infant Bonding

Given that the oxytocin system has been linked with parental behavior across numerous mammalian species, it is not surprising that oxytocin plays a role in bonding between human parents and infants. Plasma oxytocin levels during pregnancy have been correlated with mothers’ feelings related to bonding to their infant (Levine, Zagoory-Sharon, Feldman, & Weller, 2007), attachment representations toward their infants, and maternal preoccupations regarding the health and safety of their babies (Feldman et al., 2007). Oxytocin has also been predictive of the quality of biological mothers’ and fathers’ behavior displayed toward their infants (Feldman et al., 2007; Gordon et al., 2010a; Gordon et al., 2010b).

There is also evidence that peripheral oxytocin levels are associated with brain activity in response to infant stimuli. Biological mothers’ peripheral oxytocin levels have been associated with the activation of areas of the brain involved in the dopaminergic reward system (the ventral striatum and the nucleus accumbens, Atzil, Hendler, & Feldman, 2011; Strathearn et al., 2009) and the oxytocin-associated areas (hypothalamic and pituitary regions; Strathearn et al., 2009) when presented with stimuli involving their own infant. Although the exact mechanism linking central and peripheral activity is not known, several additional studies involving humans and animals have found associations between peripheral oxytocin activity and central nervous system activity (Carson et al., 2010; Geenen, Adam, Baro, Mantanus, & Ansseau, 1988).

The Role of Oxytocin in Foster Mother-infant Bonding

In the past decade, there have been exciting advances in the understanding of oxytocin’s role in bond formation and caregiving quality among biological mothers and infants. However, to our knowledge, there have not been studies assessing the role of oxytocin in bond formation among non-biologically-related dyads. Although there are numerous forms of naturally occurring alloparental relationships among humans, the current study focused on the development of foster mothers’ relationships with their foster infants. Despite important differences between biologically-related versus foster mothers (e.g. pregnancy and breastfeeding), it seemed plausible that oxytocin might also play a role in foster mothers’ bonding with foster infants.

Oxytocin and Foster Mothers’ Cognitions

Previous research has identified important psychological components of foster mother-infant relationships. The degree to which a foster parent feels committed to maintaining a long-term relationship with a foster infant appears to have important implications for foster infants’ healthy behavioral and psychological adjustment (Dozier & Lindhiem, 2006; Lindhiem & Dozier, 2007). Relative to low levels of commitment, higher levels of commitment among foster mothers have been associated with reduced problematic behaviors among foster children (Dozier & Lindhiem, 2006).

Given prior evidence suggesting that oxytocin is associated with maternal bonding-related cognitions (Feldman et al., 2007; Levine et al., 2007), we hypothesized that foster mothers’ oxytocin production would be positively associated with their feelings of commitment toward their foster infants. This seemed particularly important for understanding why foster mothers have been found to vary in the degree to which they report feelings of closeness and commitment to foster infants (Dozier & Lindhiem, 2006; Lindhiem & Dozier, 2007).

Oxytocin and Foster Mothers’ Brain Activity

Recently, biological and foster mothers’ brain activity in response to infant faces has been examined using event-related potentials (ERPs). ERPs are electrical potentials in the brain that are associated with physical or cognitive “events” and are extracted from the ongoing electroencephalogram (EEG) using signal-averaging techniques. The averaged ERP waveform consists of series of “components” often associated with overlapping positive (P) and negative (N) peaks normally referred to by polarity and latency.

A particularly well-studied component of the ERP, the P3 or P300, is a positive deflection occurring approximately 300 ms following a significant stimulus and achieves its maximal amplitude at parietal recording sites. The P3 has been shown to increase when biological parents and foster parents view images of their “own” children versus unfamiliar children (Grasso, Moser, Dozier, & Simons, 2009; Weisman, Feldman, & Goldstein, in press). In addition, foster parents’ P3 amplitudes in response to their “own” infants have been associated with the degree to which parents are accepting of their children (by reporting pleasure, delight, and reward in the child as an individual) and feel that they are influential in their children’s development (by valuing the nurturance they provide to their children) (Grasso et al., 2009). Therefore, this ERP component seems to have particular relevance for foster mother-infant relationships.

Whether peripheral oxytocin is associated with electrophysiological brain activity in response to infant images has not yet been examined, to our knowledge. However, there is some evidence that associations between peripheral oxytocin levels and ERP responses may be found. First, systemic administration of oxytocin into the blood stream has been associated with the activation of oxytocin-receptor rich areas of the brain among rodents. One of the areas affected by peripheral oxytocin infusion is the locus coeruleus, which is also considered to be a source of the P3 component among humans (Aston-Jones & Cohen, 2005; Carson et al., 2010; Tribollet et al., 1992). Second, there is evidence among humans that peripheral oxytocin levels are associated with ERP activity in response to presentation of visual stimuli (Geenen et al., 1988). Taking the findings from prior research together, we hypothesized that foster mothers’ oxytocin production would be positively associated with their P3 amplitudes in response to viewing infant images in current study.

Oxytocin and Foster Mothers’ Caregiving Quality

In previous research, behavioral delight has been identified as an important component of foster mothers’ caregiving behavior (Bernard & Dozier, 2010). Foster mothers’ behavioral delight, defined as displays of positive affect, smiling, praising, and active encouragement of mutual interaction that demonstrate a genuine affection and joy in playing with or being with one’s child, has been associated with the degree to which foster mothers feel committed to their foster infants. Because parenting quality has been linked with peripheral oxytocin levels among biologically-related parents and infants (Feldman et al., 2007; Gordon et al., 2010a; Gordon et al., 2010b), we hypothesized that foster mothers who produced more oxytocin across a cuddle interaction with their foster infant would show greater displays of behavioral delight toward their infant during a play interaction.

The Current Study

Based on the various studies involving oxytocin and maternal behavior among humans, oxytocin appears to be an important biological correlate of maternal behavior, bonding-related cognitions, and brain activity in response to infant images. The current study examined whether foster mothers’ oxytocin production was associated with their electrophysiological (ERP) brain activity in response to their foster infants, psychological variables that have been established as particularly relevant to caring for one’s foster infant (i.e., their levels of commitment, Dozier & Lindhiem, 2007), and foster mothers’ displays of delight, a behavioral variable that has been established as significant when caring for one’s foster infant (Bernard & Dozier, 2011). We hypothesized that foster mothers’ oxytocin production would be positively associated with their electrophysiological brain activity in response to images of their foster infants, reports of commitment, and expressions of behavioral delight at early and more established phases of the relationship.

Method

Participants

Forty-three foster mother-infant dyads participated in this study. At the start of the study, infants ranged in age from 0.5 months to 35 months old (M = 8.5, SD = 9.6). Fifty-five percent of the infants were females (n = 24). Twenty-five percent of the foster infants were White/non-Hispanic, 35% were African American, and 40% were bi-racial. Foster mothers ranged in age from 28 to 68 years old (M = 42.1, SD = 10.1). Forty-seven percent of the foster mothers were White/non-Hispanic, 46% were African American, and 7% were Hispanic/Latino. Fourteen percent of the foster mothers had completed graduate school, 21% had completed college, 25% had completed some college, and 40% had completed high school. Forty-six percent of the foster mothers did not work, and 54% were employed either part-time or full-time. Seventy-nine percent of the foster mothers were married, 2% were living with a partner, 9% were divorced, and 10% were single. Twenty-four percent of the families had annual incomes greater than $100,000/year, 22% had incomes between $60,000 to $100,000/year, 29% had incomes between $40,000 and $60,000/year, 7% had incomes between $30,000 to $40,000/year, 10% between $20,000 and $40,000/year, and 8% had incomes below $20,000/year.

Procedure

After receiving consent for the foster mothers and infants, two laboratory visits (“A” and “B”) were scheduled during the first 60 days of the foster infant placement (at this first visit, foster infants had lived with foster mothers for a minimum of .47 months to a maximum of 1.96 months, M = 1.28 months, SD = .47 months). At visit A, foster mothers’ oxytocin was measured after a cuddle interaction task with their foster infant, as described below. At visit B, foster mothers’ ERP brain activity in response to images of their foster infant, verbal reports of commitment, and behavioral delight (through a ten-minute play interaction with their foster infant) were assessed. The order of these visits was counterbalanced between foster mothers. Foster mothers were asked to return to the lab to repeat visits A and B, three months following the first visits. Due to scheduling issues, the timing of this visit ranged from 2.8 months to 6.6 months following the first laboratory visits (M = 3.6 months, SD = .86 months). At the second visit, infants had lived with their foster mothers for a minimum of 3.7 months to a maximum of 8.43 months (M = 5.0 months, SD = 1.05 months).

Measures

Composite risk variables

During the laboratory visits, foster mothers were asked to report on several parent and child risk variables found to be predictive of parent-child relationship quality and maternal behavior in previous research (Bakermans-Kranenburg, Breddels-Van Baardewijk, Juffer, Velderman, & Van IJzendoorn, 2008; Blair, 2002; Dozier & Lindhiem, 2006; Fisher, Burraston, & Pears, 2005; Lindhiem & Dozier, 2007; Schuler, Black, & Starr, 1995; Zeanah et al., 2001). Composites of multiple risk factors have proven more powerful in predicting problematic outcomes, when compared with examining individual risk variables in isolation (e.g., Sameroff, Seifer, Zax, & Barocas, 1987).

A child composite risk score was calculated by combining dichotomous scores for infant temperament, premature birth, previous number of foster placements, reports of prenatal drug exposures, evidence of physical or sexual abuse or neglect prior to the current foster placement, and child age. Foster infant temperament was assessed using the Infant Behavior Questionnaire and the Early Childhood Behavior Questionnaire (Gartstein & Rothbart, 2003; Rothbart, 1981). Infants with scores that fell 2 standard deviations or more above the mean on the soothability subscale were coded as “high risk.” Prematurity was determined if the infant either weighed less than 2500 grams at birth or was born at less than 32 weeks gestation (or both), following guidelines in previous studies (Caputo et al., 1974; Minde, 1993). Two or more foster placements were coded as “high risk,” as established in previous research (Lewis et al., 2007). Prenatal drug exposure and maltreatment history (defined as physical abuse, sexual abuse, or neglect) was assessed as present or absent. These variables have been associated with parenting quality in previous research (Bakermans-Kranenburg et al., 2008; Bustan & Sagi, 1984; Dozier & Lindhiem, 2006; van den Boom, 1994).

In previous research, foster infant age has been predictive of foster mothers’ acceptance of and belief in the degree to which they could positively influence the infant in their care (Bates & Dozier, 2002) and foster mothers’ commitment toward foster infants (Lindhiem & Dozier, 2007). In the current study, foster infant age was negatively associated with foster mothers’ behavioral delight at the second laboratory visit (see Appendix 2). For this reason, foster infant age was entered as a risk factor in the child composite variable. Child age was dichotomized based on the medium split and categorized as “high risk” if they were categorized as “older” (above the median) and “low risk” if categorized as “younger” (below the median). The total foster infant composite risk score ranged from 0 to 6, with higher scores reflecting the presence of more risk factors.

Appendix 2.

Time 1 Correlations between Bonding Variables and Demographic Characteristics

com1 delight1 difoxy1 p3ave1 p3diff1 marital cgeduc income cgage chgen chage plclngth
com1 1 . . . . . . . . . . .
delight1 .31* 1 . . . . . . . . . .
difoxy1 −.12 .34* 1 . . . . . . . . .
p3ave1 .22 .04 .20 1 . . . . . . . .
p3fos1 .01 .08 −.14 .15 1 . . . . . . .
marital −.15 .02 .04 −.25 .10 1 . . . . . .
cgedu1 .10 .08 −.02 .22 .28 −.11 1 . . . . .
income .01 −.05 −.11 .29 .26 −.55** .55** 1 . . . .
cgage −.13 −.29 .19 .08 −.16 .05 −.28 −.05 1 . . .
chgen −.10 −.19 .09 −.11 .05 .32* .14 −.10 .11 1 . .
chage −.27 −.19 −.19 −.05 −.10 .17 −.11 −.11 .07 .10 1 .
plclgth .02 −.26 −.08 .07 .23 −.09 −.11 .25 .21 .19 .17 1
Open in a new tab
**

p<.001

*

p<.05

Note: com1= commitment at time 1; delight1 = delight at time 1; p3ave1 = foster mothers’ average P3 to all infant images at time 1; p3fos1 = foster mothers’ P3 amplitude to foster – average of familiar and unfamiliar P3 amplitude at time 1; marital = marital status (0 = single, 1 = married/in a relationship); cgedu1 = foster mothers’ educational status; income = foster mothers’ annual income; cgage = foster mothers’ age; chgen = foster child’s gender (0 = male, 1 = female); chage = foster child’s age; plclgth = length of placement.

A composite score for foster mother risk was calculated by combining dichotomous measures of mental health status and number of previous foster children cared for, as these variables have been found to affect foster parents’ commitment and delight (Dozier & Lindhiem, 2006). Foster mothers’ mental health status was assessed through the Brief Symptoms Index (BSI), a 53-item self-report measure administered to assess the prevalence of psychological symptoms (Derogatis & Spencer, 1982). For the composite risk variable, scores that fell 2 standard deviations or more above the mean on the Global Severity Index of psychological health functioning were classified as “high risk.” The previous number of foster children cared for was coded as “high risk” if the number of previous placements was 10 or more. The foster mother composite risk score ranged from 0 to 2, with higher scores reflecting more foster mother risk factors.

Foster mothers’ oxytocin measurements

Oxytocin was measured through urine samples as part of a laboratory visit at the Wisconsin National Primate Research Center. Foster mothers were encouraged to drink liquids before and during the visits to facilitate the process of voiding on three occasions. They were asked to avoid caffeine, nicotine, and alcohol before the laboratory visit as these substances can interfere with oxytocin production. All foster mothers reportedly complied with this request and no samples were excluded on this basis.

At the beginning of each lab visit, foster mothers and infants were separated from each other for 45 minutes so that post-interaction samples of oxytocin were not affected by physical contact apart from the experimental manipulation. Foster mothers were asked to sit in a room where they could not hear or see their foster infants. A staff member with extensive experience with infants cared for the foster infant in another room. After the first 20 minutes of this 45-minute separation, foster mothers were asked to void in a private bathroom in order to rid their systems of oxytocin that may have accumulated peripherally prior to being separated from their infants. Oxytocin has been found to have a half-life of approximately 10 minutes and metabolic clearance rate of 1.4 liters/min (Dawood, Ylikorkala, Trivedi, & Gupta, 1980). Therefore, it was expected that this time interval would allow for oxytocin levels to be sufficiently cleared from foster mothers’ peripheral systems, and would allow for a valid estimation of baseline oxytocin levels.

At the end of the 45-minute separation, foster mothers were asked to provide a pre-interaction (baseline) urine sample. Following the collection of the sample, foster mothers were instructed to cuddle with their foster infants for 30 minutes. After the completion of this task, foster mothers were asked to provide the post-interaction urine samples as soon as they were able to do so. Given that oxytocin is secreted in a pulsatile fashion into the blood stream in as little as five minutes following physical contact (Turner, Altemus, Enos, Cooper, & McGuiness, 1999), this duration of 30 minutes was thought to be sufficient for measuring peripheral oxytocin production in response to cuddling. Pre- and post-interaction samples were then transferred to sterile cryogenic vials and stored in a freezer at −80 °C until analyses by the Assay Services Unit. Urine samples were frozen for no longer than six months and were sent to the Wisconsin National Primate Research Center for assay.

Oxytocin assays

After controlled thawing, urinary samples were subjected to solid-phase extraction (SPE) using 1 ml SepPak C18 cartridges (Waters, cat no. WAT023590). Each column was pretreated with 1 ml of methanol and then 1 ml of water before application of 1 ml of urine. Samples were washed with 1 ml of 10 percent acetonitrile, 1 percent TFA (trifluroacetic acid). Oxytocin samples were eluted in 1 ml of 80/20 percent acetonitrile solution. Samples were then dried down in a water bath with air stream and reconstituted in the assay-appropriate buffer supplied in the 96-well ELISA assay kit used (Assay Designs, Cat no. 901-153). Intra- and inter-coefficients of variation (CV) were determined by a human urine pool. Intra CV = 2.3 and Inter CV = 8.4, respectively. Oxytocin standards were used to determine recoveries from the extraction method and were 85%.

Oxytocin values

Oxytocin values that were more than 3 standard deviations from the mean for each of the four data collection time points (pre- and post-interaction collections at Time 1 and Time 2) were considered outliers and were not included in the analyses. This excluded only mothers with high oxytocin values because the lowest possible value (0) was within 3 standard deviations of the mean for each context. Three oxytocin values were identified as outliers across the four measurement points (one at the Time 1 pre-interaction assessment, one at the Time 2 pre-interaction assessment, and one at the Time 2 post-interaction assessment) and were therefore excluded from analyses.

The normality of the distributions of oxytocin values was also assessed. Due to the presence of significant skew, a square root transformation was employed to normalize the distributions for each time point. Total oxytocin production in response to the cuddle interaction was calculated by subtracting the transformed pre-interaction oxytocin values from the transformed post-interaction values at Time 1 and Time 2. Foster mothers’ baseline oxytocin values (i.e. pre-interaction oxytocin values), were marginally stable from Time 1 to Time 2, r = .30, p = .09. Foster mothers’ oxytocin production during the cuddle interaction (i.e. the difference between post-interaction oxytocin values and pre-interaction oxytocin values) was not stable from Time 1 to Time 2, r = .10, p = .56. Time elapsed between pre- and post-interaction measurements and time of day were not associated with foster mothers’ oxytocin values and were therefore not included as covariates in subsequent analyses.

Foster mothers’ P3 amplitude

The parietally-maximal P3, previously associated with motivated attention to emotionally evocative stimuli and “loved ones” (Grasso et al., 2009; Langeslag, Jansma, Franken, & Van Strien, 2008; Vico, Guerra, Robles, Vila, Anllo-Vento, 2010) was the ERP component of interest in the current study. Prior to a passive viewing task, foster mothers were familiarized with an image of a child they had not previously seen. The familiarization method was adapted from a prior study that examined the impact of familiar faces on ERP activity (Claypool, Hugenberg, Housley, & Mackie, 2007). Next, foster mothers participated in a passive viewing task. Foster mothers were presented with a series of faces in random order. For each foster mother, stimuli consisted of images of her foster child, the child to whom she was “familiarized,” and an unfamiliar child. Familiar and unfamiliar faces were matched by age, gender, and facial expression to the foster mother’s infant who participated in the study. Each face appeared on the computer screen for 2000 ms with an inter-trial interval of 1500 ms using Presentation software (Neurobehavioral Systems Inc.). Pictures of the foster child, the familiar child, and the unfamiliar child were presented 25 times each in a pseudo-random order, for a total of 75 trials.

Data collection, processing, and analyses

ERP data were recorded from an electrode cap consisting of 32 Ag/AgCl electrodes. Placement of electrodes in the cap followed the International 10–20 System (Cooper, Osselton, & Shaw, 1969; Jasper, 1958). Advanced Neuro Technology Acquisition hardware (ANT, Enschede, The Netherlands) was used for EEG data collection with an average electrode reference and forehead ground. The continuous EEG was digitized at 512 samples per second and processed using Advanced Source Analysis (ASA), a software package designed for ANT ERP systems. Re-referencing to the averaged mastoid was performed offline. The EEG was corrected for artifacts such as eye blinks and EKG, and bandpass filtered from .1 Hz to 30 Hz. EEG data that exceeded −75μV or +75 μV were rejected. Average ERP waveforms, locked to picture onset, were computed for each condition (foster child, familiar child, or unfamiliar child) at each electrode. Averages were baseline corrected by subtracting the average voltage occurring during the 200 ms before the presentation of the infant image from the entire average. The P3 component was defined as the average amplitude within time windows of 300–650 ms following stimulus presentation. For the purposes of these analyses, the Pz electrode was examined, as the P3 component associated with motivated attention has been found to be maximal at this electrode site (Schupp et al., 2004).

ERP P3 values

Prior to analyses, the validity of ERP data was examined for each participant at Time 1 and Time 2. After the processing of ERP data, two participants at Time 1 had fewer than 15 trials in the “foster child” condition due to excessive noise that occurred during data collection. Therefore, the ERP data for these participants were excluded from analyses. In addition, one foster mother did not complete the ERP portion of the laboratory visits at Time 1 because her child left her care. There were no missing data for ERP activity at Time 2.

Differences between foster mothers’ P3 responses to pictures of their foster infants and to pictures of familiar and unfamiliar infants were explored at Time 1 and Time 2 (see Table 1 for descriptive statistics). At Time 1 and Time 2, results of separate repeated measures ANOVAs revealed a significant effect of stimulus type (i.e., foster, familiar, or unfamiliar) on P3 amplitude at Time 1, F(2, 78) = 9.22, p < .001, and Time 2, F(2, 64) = 9.89, p < .01 (see Appendix 1). Post-hoc comparisons revealed that the foster mothers’ P3 responses to their foster infants were significantly larger than their P3 to the familiar and unfamiliar infants at Time 1 and Time 2 (p values < .01). Foster mothers’ P3 responses to the familiar infants did not differ significantly from their P3 responses to the unfamiliar infants at Time 1 or Time 2.

Table 1.

Ranges, Means, and Standard Deviations of Oxytocin, P3 Amplitudes, Delight, and Commitment at Time 1 and Time 2

Time 1 Variables Min. Max. M SD
T1 Oxytocin production (N = 42) −8.09 6.58 0.89 3.71
T1 P3 “unique” foster amplitude (difference score) (N = 40) −1.12 2.34 0.53 0.89
T1 P3 average amplitude to all infants (N=40) 0.24 4.23 1.96 0.97
T1 Foster mother delight (N=41) 1.00 5.00 3.95 1.07
T1 Foster mother commitment (N=39) 1.00 5.00 3.45 0.70
Time 2 Variables Min. Max. M SD
T2 Oxytocin production (N = 32) −5.01 10.32 1.22 3.24
T2 P3 “unique” foster amplitude (difference score) (N=33) −1.80 2.47 0.64 1.01
T2 P3 average amplitude to all infants (N = 33) 0.42 4.99 2.14 0.94
T2 Foster mother delight (N=33) 1.00 5.00 4.07 1.19
T2 Foster mother commitment (N=33) 2.00 5.00 4.18 0.76
Open in a new tab
Appendix 1.

Average Amplitude of P3 Response to Foster, Familiar, and Unfamiliar Images (Microvolts)

P3 at Time 1 Min. Max. M SD
Foster .09 4.94 2.31 1.21
Familiar .02 4.22 1.78 1.03
Unfamiliar −.51 4.08 1.78 1.04
P3 at Time 2 Min. Max. M SD
Foster .67 6.47 2.57 1.26
Familiar .18 3.82 1.86 .97
Unfamiliar .39 4.73 1.98 1.04
Open in a new tab

For the primary analyses, we were interested in examining foster mothers’ P3 responses to their foster infants when controlling for their P3 responses to the familiar and unfamiliar infant faces. Therefore, we calculated a difference score by subtracting the average of each foster mother’s P3 response to the familiar and unfamiliar infant faces from her P3 response to her foster infant’s face. For the sake of clarity, we will refer to this difference score as foster mothers’ “unique P3 amplitude to their foster infant” for the remainder of this manuscript. We were also interested in examining foster mothers’ average P3 responses to all infant faces, regardless of familiarity. Therefore, we averaged each foster mother’s P3 amplitude to the foster, unfamiliar, and familiar conditions. Hereinafter, we will refer to this variable as foster mothers’ “average P3 amplitude to all infant faces”. The difference score and average score were both used as variables in primary analyses.

In the current study, foster mothers’ P3 amplitude to one’s foster infant, r = .52, p < .01, the familiar infant face, r = .54, p < .01, and the unfamiliar infant face, r = .57, p < .01, remained highly stable over time. Foster mothers’ average P3 amplitudes to all infant faces also remained highly stable over time, r = .71, p < .001. However, foster mothers’ P3 amplitudes to their foster infants (i.e., the difference between foster mothers’ P3 to their own foster infant minus their average P3 to an unfamiliar and familiarized infant) did not remain stable from Time 1 to Time 2, r = .07, p = .70.

Foster mother commitment

Foster mother commitment was assessed using the TIMB (This Is My Baby) Interview developed by Bates and Dozier (1998). The TIMB interview is a semi-structured interview, in which foster parents are asked about their long-term goals for the foster child, how influential they feel they are in shaping the child’s long-term development, and how much they would miss the child if he or she left their care. Interviews were transcribed and coded by two independent coders. Commitment is rated on a 5-point Likert scale. Prior research has provided evidence of the TIMB’s predictive validity and test-retest reliability (Bernard & Dozier, 2011; Dozier & Lindhiem, 2006; Lindhiem & Dozier, 2007). Good inter-rater reliability has been established for this measure in prior studies and in the current study, Intra Class Correlation (ICC) = .82, p < .001.

At Time 1, commitment data were missing for four participants (two participants’ foster infants were removed from their foster placements prior to the measurement of commitment and two participants’ data were not able to be coded due to audio-recorder problems). There were no missing data for commitment at Time 2 among the remaining 33 participants. In the current study, foster mothers’ commitment scores were moderately stable from Time 1 to Time 2, r = .47, p < .01.

Foster mother delight

Foster mothers’ behavioral expressions of delight were assessed by coding foster mothers’ behavior during a 10-minute play interaction with their foster infants. Foster mothers were asked to participate in a standardized play interaction. Foster infants were seated in an infant seat and foster mothers sat directly in front of them. Foster mothers and infants were provided with a set of standardized toys. If the foster infant was less than 20 months of age, a rattle, a set of cups, and soft “squeaky” toy were provided. If the foster infant was 20 months or older, a set of blocks was provided. Foster mothers were asked to play with their foster infants “as they typically would” for ten minutes. The interaction was videotaped from behind a two-way mirror.

Displays of delight toward their foster infant were coded from the videotaped play interactions. Delight was rated on a 5-point Likert scale. Examples of high ratings of delight included displays of genuine affection, physical closeness, and obvious joy when interacting with the child. Low ratings were assigned to foster mothers who interacted with their foster infants without affective involvement or who appeared uninterested in interacting with their foster infants. This measure, which was adapted from the Maternal Delight Scale used by Ainsworth et al., 1978, has been shown to have concurrent validity and good inter-rater reliability (Bernard & Dozier, 2011). Inter-rater reliability in the present study was good (ICC = .84, p < .001). At Time 1, behavioral delight data were missing for two participants (two participants’ foster infants were removed from their foster placements prior to the measurement of behavioral delight). There were no missing data for behavioral delight scores at Time 2 among the remaining 33 participants. Foster mothers’ displays of behavioral delight were moderately stable from the first to second laboratory assessment (r = .41, p < .05).

Results

Preliminary Analyses

The current sample consisted of 43 foster mother-infant dyads at Time 1. However, only 41 foster mothers were included in primary analyses due to the exclusion of 1 foster mother with outlying oxytocin data and 1 foster mother with both outlying oxytocin and P3 data (reasons for missing data are described in methods section). Ten of the 43 foster mothers and infants did not return for the Time 2 visit because the foster infants had left their care. Therefore, the Time 2 sample consisted of 33 foster mothers and infants. However, only 32 foster mothers were included in primary analyses due to the exclusion of 1 foster mother with outlying oxytocin values at Time 2. There were no significant differences in foster mothers’ oxytocin production, F(1, 41) = .706, p = .41, delight, F(1, 42) = .448, p = .51, unique P3 amplitude to their foster infant, F(1, 39) = .489, p = .49, average P3 amplitude to all infant images, F(1, 39) = .585, p = .45, and commitment, F(1, 38) = 1.91, p = .17, between mothers who did and did not return to follow-up assessments at Time 2. Given the small sample size, mean imputation was performed on missing values of foster mothers delight (two foster mothers at Time 1), commitment (4 foster mothers at Time 1) and P3 activity (1 foster mother at Time 1) in order to conserve power in primary analyses.

Descriptive characteristics at Time 1 and 2 for foster mother commitment, delight, oxytocin production during the cuddle interaction, and P3 amplitude are presented in Table 1. Correlations between primary variables of interest (foster mothers’ commitment, delight, oxytocin production during the cuddle interaction, unique P3 amplitude to their foster infant, and average P3 amplitude to all infants) and demographic characteristics are presented in Appendix 2 and 3. Because caregiver age, education, marital status, and income were not significantly associated with the primary variables of interest, these variables were not included as covariates in primary analyses (see Appendix 2 and 3). There were no significant associations between placement length and variables of interest; therefore, placement length was not entered as a covariate in primary analyses at Time 1 or Time 2. Finally, we explored whether the range of time elapsed between assessment visits was associated with the primary variables of interest. The amount of time between Time 1 and Time 2 assessments ranged from 2.79 months to 6.64 months (M = 3.6, SD = .86). However, this variability was not significantly associated with foster mothers’ change in oxytocin production during the cuddle interaction, r = .23, p = .24 at Time 2. Therefore, time elapsed between the first and second laboratory visits was not included as a covariate in analyses.

Appendix 3.

Time 2 Correlations between Bonding Variables and Demographic Characteristics

com2 delight2 difoxy2 p3ave2 p3diff2 marital cgedu income cgage chgen chage plclngth
com2 1 . . . . . . . . . . .
delight2 −.09 1 . . . . . . . . . .
difoxy2 −.01 .48* 1 . . . . . . . . .
p3ave2 .20 .02 .17 1 . . . . . . . .
p3fos2 .13 −.12 .27 .19 1 . . . . . . .
marital −.21 .18 −.05 −.08 −.07 1 . . . . . .
cgedu .31 −.09 .26 .27 .29 −.11 1 . . . . .
income .05 −.06 .07 .01 .17 −.55** .55** 1 . . . .
cgage −.16 .10 .13 −.07 −.15 .05 −.27 −.05 1 . . .
chgen −.08 −.33 .40* .04 .02 .32* .14 −.10 .11 1 . .
chage −.31 −.36* .12 −.07 .143 .30 −.11 −.11 .07 .12 1 .
plclngth2 −.10 −.03 .14 .23 .19 −.13 .16 .36* .36* −.07 .06 1
Open in a new tab
**

p<.001,

*

p<05,

Inline graphic

p<.10

Note: com2= commitment at time 2; delight2 = delight at time 2; p3ave2 = foster mothers’ average P3 to all infant images at time 2; p3fos2 = foster mothers’ P3 amplitude to foster minus average of familiar and unfamiliar P3 amplitude at time 2; marital = marital status (0 = single, 1 = married/living together); cgedu2 = foster mothers’ educational status; income = foster mothers’ annual income; cgage = foster mothers’ age; chgen = foster child’s gender (0=male, 1=female); chage = foster child’s age; plclgth = length of placement

Oxytocin Production across the Cuddle Interaction

A series of separate repeated measures ANOVAs were conducted to examine patterns of oxytocin production across the cuddle interaction at Time 1 and Time 2. Foster mothers’ oxytocin levels were not significantly different at post interaction assessments when compared with pre interaction assessments at Time 1, F(1, 42) = .85, p = .36, or Time 2, F(1, 31) = 2.20, p = .15. Consistent with previous research (Feldman et al., 2010), foster mothers varied in whether their oxytocin increased across the cuddle interaction (see Table 1 for descriptive statistics). We also examined whether foster mothers’ oxytocin production varied from the first to the second laboratory assessment. Results from a repeated measures ANOVA revealed that foster mothers’ oxytocin production across the cuddle interaction at Time 1 was not significantly different from their oxytocin production at Time 2, F(1, 30) = .01, p = .95.

Associations between Oxytocin Producton and Delight, Commitment and Brain Activity

Associations between foster mothers’ oxytocin production across the cuddle interaction and foster mothers’ average P3 amplitude to all infant images, unique P3 amplitude to their foster infant, commitment, and behavioral delight were examined using multiple regression. Two separate multiple regression models were computed for Time 1 and Time 2 data.

At Time 1, foster mothers’ oxytocin production across the cuddle interaction was regressed on their average P3 to all infants, unique P3 amplitude to their foster infant, commitment and delight. Caregiver and child composite risk variables were also included in the model as covariates. The overall model at Time 1 was statistically significant, F(6, 40) = 2.37, p = .05, and accounted for 29.5 percent of the variance in foster mothers’ oxytocin production across the cuddle interaction (see Table 2). A post hoc power analysis was conducted for R2 = .295. Power was found to be .83 with six predictors in the model, for N = 41, p = .05.

Table 2.

Time 1 and Time 2 Linear Regression for Oxytocin Production across the Cuddle Interaction

T1 Oxytocin regressed on B SE β Est./SE P-value
T1 Delight 1.55 0.54 0.43 2.86 .007**
T1 Commitment −0.99 0.63 −0.25 −1.58 .124
T1 P3 to Foster −0.49 0.65 −0.12 −0.76 .453
T1 P3 Average 1.32 0.60 0.34 2.21 .034*
T1 Foster Parent Composite Risk 0.32 0.58 0.08 0.56 .582
T1 Foster Infant Composite Risk 1.65 1.03 0.24 1.61 .117
(Constant) −5.18 3.12 - −1.67 .105
T2 Oxytocin regressed on B SE β Est./SE P-value
T2 Delight 1.03 0.44 0.38 2.33 .028*
T2 Commitment 0.16 0.65 0.04 0.25 .803
T2 P3 to Foster 1.10 0.48 0.35 2.30 .030*
T2 P3 Average 0.08 0.54 0.02 0.14 .889
T2 Foster Parent Composite Risk 1.13 0.83 0.21 1.37 .182
T2 Foster Infant Composite Risk 1.05 0.56 0.31 1.87 .073
(Constant) −0.56 3.30 - −1.77 .090
Open in a new tab

Note: T1 Adjusted R2 = .17, T2 Adjusted R2 = .33

*

= p < .05,

**

= p < .01

A second regression analysis was performed for the Time 2 data. Foster mothers’ oxytocin production across the cuddle interaction at Time 2 was regressed on their average P3 amplitude to all infants, unique P3 amplitude to their foster infant, commitment and delight collected at Time 2. Similar to Time 1 analyses, caregiver and child composite risk variables were also included in the model as covariates. The overall model at Time 2 was statistically significant, F(6, 31) = 3.53, p < .05, and accounted for 45.9 percent of the variance in foster mothers’ oxytocin production across the cuddle interaction (see Table 2). A post hoc power analysis was also conducted for R2 = .459. Power was found to be .96 with six predictors in the model, for N = 32, p = .05.

Associations between Oxytocin and Foster Mothers’ Bonding Variables

Is Oxytocin Production Associated with Foster Mother Delight and Commitment?

At Time 1, foster mothers’ total oxytocin production in response to the cuddle interaction was significantly associated with the amount of behavioral delight exhibited during a play interaction, when controlling for foster parent and infant composite risk factors (β = .43, p < .05; see Table 2). At Time 2, foster mothers’ total oxytocin production in response to the cuddle interaction continued to be associated with foster mothers’ displays of behavioral delight toward their foster infants, when controlling for foster parent and infant risk variables (β = .38, p < .05; see Table 2). Foster mothers who produced more oxytocin while cuddling with their foster infants displayed higher levels of behavioral delight during a play interaction with their infant. This association was observed at the first assessment, which occurred in the first two months of the foster mother infant relationship, and at an assessment that occurred three months later.

Foster mothers’ total oxytocin production was not associated with levels of commitment at Time 1 (β = −.25, p = .12; see table 2). Similar to Time 1, foster mothers’ total oxytocin production in response to the cuddle interaction was not significantly associated with reports of commitment at Time 2 (β = .04, p = .80; see Table 2).

Is Oxytocin Production Associated with Foster Mother P3 Activity?

At Time 1, foster mothers’ total oxytocin production in response to the cuddle interaction was significantly associated with the amplitude of their average P3 to all infant faces (β = .34, p < .05) but was not significantly associated with the unique P3 amplitude to their foster infant (β = −.115, p = .45; see Table 2). Foster mothers who showed greater increases in oxytocin after cuddling with their foster infants showed larger P3 components in response to viewing all babies’ pictures. At Time 1, foster mothers’ oxytocin production was not related to the unique P3 amplitude to their foster infant.

At Time 2, foster mothers’ total oxytocin production in response to the cuddle interaction with their foster infants was no longer associated with their average P3 amplitude to all infants (β = .02, p = .89); however, foster mothers’ total oxytocin production was now significantly associated with the amplitude of the unique P3 amplitude to their foster infants (β = .35, p < .05; see Table 2). At this later assessment, foster mothers who produced more oxytocin after cuddling with their foster infants showed larger unique P3 amplitudes to their foster infant, when compared with foster mothers who produced less oxytocin. Unlike Time 1, oxytocin was no longer associated with foster mothers’ average P3 amplitude to all babies.

Discussion

Oxytocin has been studied extensively among biologically-related parent-infant dyads and has been associated with maternal caregiving behavior, attachment-related thoughts, and brain activity in response to infant stimuli. Although foster parents do not experience the physical conditions of childbirth and lactation known to affect oxytocin production, previous research on maternal caregiving among humans and animals suggests that oxytocin may play an important role in bond formation toward non-biological infants. Our results are exciting in showing that, even when biological relatedness is eliminated as a factor (and issues of pregnancy and breastfeeding are eliminated), oxytocin may still relate to bond formation in meaningful ways. Specifically, individual differences in foster mothers’ oxytocin levels are associated with the amount of behavioral delight exhibited toward one’s foster infant and the pattern of brain activity that emerges in response to viewing infant images.

Results from this study revealed that oxytocin, a neuro-hormone associated with affiliative behavior, was associated with foster mothers’ brain activation in response to infant stimuli and with the quality of maternal behavior, with the pattern of associations changing over time. First, foster mothers’ oxytocin levels were associated with their electrophysiological brain activity in response to infant stimuli. Interestingly, the manner in which foster mothers’ oxytocin production was associated with their brain activity changed over the course of the foster mother-infant bond. In the first two months of the relationship, foster mothers’ oxytocin production in response to a cuddle interaction was associated with their brain activity in response to infants in general, but not to the specific brain activity elicited in response to their “own” foster infant. However, over time, foster mothers’ oxytocin production in response to a cuddle interaction became significantly linked to their brain activity in response to viewing their “own” foster infant. These associations over the first months of the foster mother-infant relationship may depict the biological processes that are associated with foster mothers’ bond formation with non-biological children.

Furthermore, foster mothers’ oxytocin production in response to a cuddle interaction was significantly associated with their behavioral expressions of delight toward their foster infants in a play interaction. The association between foster mothers’ oxytocin production and behavioral delight emerged during the first two months of the foster mother-infant relationship and at a follow up assessment that occurred three months later. Whereas foster mothers’ oxytocin production in response to a cuddle interaction was not found to be stable over time when examined independently, its association with foster mothers’ behavior (when controlling for composite foster parent and child risk factors) appeared to persist over time, at least through the initial phase of the foster mother-infant relationship.

Associations between oxytocin systems and alloparental behavior have been demonstrated in studies involving non-human mammals (Olazabal & Young, 2006). The animal literature has focused on associations between oxytocin receptor density and alloparental caregiving quality (Champagne et al., 2001; Olazábal & Young, 2006), but such associations have not yet been demonstrated among humans, to our knowledge. Even less is known about peripheral oxytocin activity and alloparental caregiving in human or animal species. The findings from the current study therefore contribute to the literature in suggesting that neuro-hormonal mechanisms related to peripheral oxytocin production are involved in non-biological parenting among humans.

At the first assessment (which occurred in the first two months of the foster parent-infant relationship), foster mothers’ oxytocin production in response to a cuddle interaction was associated with their average P3 amplitude to all infants. Therefore, foster mothers who produced more oxytocin in response to cuddling with their infant were also more likely to show a greater P3 response when presented with infant stimuli relative to foster mothers who produced less oxytocin during a cuddle interaction. The convergence in foster mothers’ oxytocin production and average P3 amplitude to all infant faces may illustrate individual differences in the degree to which foster mothers respond to infant stimuli. This convergence early on in the foster mother-infant relationship may reflect the degree to which foster mothers find infant stimuli rewarding and may indicate their potential for providing high quality alloparental care. Although it is not possible to infer the direction of association (given the cross-sectional nature of the association), it is interesting to consider the mechanism that may explain the association between these variables. It is possible foster mothers’ greater P3 amplitude, or motivated attention, to all infant faces leads them to produce more oxytocin when cuddling with an infant, even when the relationship is relatively new. However, given previous research indicating the effects of synthetic oxytocin on non-parents’ brain activity in response to hearing unfamiliar infants’ cries or laughter (Riem, Bakermans-Kranenburg, et al., 2011; Riem, van IJzendoorn, et al., in press), it is equally possible that foster mothers’ higher levels of oxytocin production in response to cuddling may cause increased motivated attention to infant-related stimuli.

Interestingly, the nature by which foster mothers’ oxytocin production was associated with their brain activity changed over time. At the follow-up assessment, which occurred approximately three months following the initial assessment, foster mothers’ oxytocin production in response to a cuddle interaction was no longer associated with their average P3 amplitude in response to all infant images. However, unlike the initial assessment, foster mothers’ oxytocin production in response to a cuddle interaction at the follow-up assessment was associated with the magnitude of their unique P3 amplitude in response to their own foster infants. Therefore, foster mothers’ oxytocin production became linked specifically with their unique P3 amplitude to their foster infant over time. When visually examining the strength of the associations, it appeared that the magnitude of the associations between foster mothers’ oxytocin production in response to a cuddle interaction and average P3 activity to all infants decreased over the course of the relationship. In contrast, the associations between foster mothers’ oxytocin production and unique P3 amplitude to one’s own infant appeared to increase over the course of the relationship.

When interpreting these findings, it is important to consider the biological systems that may be associated with the co-activation of foster mothers’ oxytocin production and P3 responses to infant stimuli. Both P3 amplitudes and peripheral oxytocin levels have been linked to the activation of the dopaminergic reward system. The P3 component has been hypothesized as at least partially reflective of activation of the locus coeruleus, which receives input from limbic areas via projections from dopaminergic reward systems in the brain, and projects to cortical areas involved in increased attention to motivational and emotional stimuli (Aston-Jones & Cohen, 2005). In two recent studies, peripheral oxytocin release during mothers’ play interactions with their biological infants was associated with activation of reward regions in the brain (Atzil et al., 2011; Strathearn et al., 2009). Therefore, the association between foster mothers’ oxytocin and average P3 amplitude to all infant faces early on in the relationship may signify an activation of foster mothers’ reward system in response to all infant stimuli, possibly signaling individual differences in foster parents’ capabilities for exhibiting high quality alloparental care. The associations between foster mothers’ oxytocin production and specific P3 activity to their foster infant, observed later on in the relationship, may reflect changes in the degree to which their specific foster infant becomes motivating or rewarding. Therefore, P3 activity and oxytocin may be influenced by the activation of the dopaminergic reward system in response to general infant stimuli in the early stages of the mother-infant relationship and specific stimuli related to their own foster infant after the relationship becomes more established.

There are limitations that should be kept in mind when interpreting the results of this study. Several human and animal studies have indicated that urinary concentrations of oxytocin are connected with social behavior and bonding (Seltzer & Ziegler, 2007; Seltzer, Ziegler, & Pollack, 2010). However, it is currently unclear how oxytocin levels in urine are related with levels of oxytocin in other peripheral systems such as blood and saliva (Feldman, Gordon, Zagoory-Sharon, 2010). Therefore, we cannot conclude, based on the results of this study, that all peripheral levels of oxytocin are equally associated with foster mothers’ brain activity and behavior in response to their foster infants. Second, due to the small sample size, it is likely that many of our analyses were underpowered to detect statistically significant differences. The small sample size, especially at Time 2, also limits the replicability of these findings. Therefore, the results from this study should be interpreted with caution. Third, although the results from the current study suggest that peripheral oxytocin is associated with bonding among non-biologically related foster mothers and infants, it is equally possible that these patterns would be observed during the bonding process among biological mothers and infants, or even among adults who interact with unfamiliar infants. In fact, in previous research, we have found that adult female mothers produce oxytocin even when participating in a close cuddle interaction with an unfamiliar child (Bick & Dozier, 2010). To understand the unique manner in which oxytocin is associated with alloparental care in human relationships (such as those that occur in foster care), future research should include comparison groups of biologically-related mother-infant and unfamiliar women-infant dyads.

Fourth, we interpreted foster mothers’ increased P3 activity in response to their own foster infants as reflective of their increased “motivated attention” to their own children. Although we found foster mothers’ unique P3 amplitude to the image of their foster infants to be greater than the P3 activity to a familiar face, we cannot rule out the possibility that the increased activity to one’s own foster child was merely due to an increase in the foster parents’ familiarity with their foster infants’ face. Future studies might improve upon this method by examining foster parents’ P3 amplitude in response to a familiar child’s face that is not cared for by the foster parent in combination with stimuli used in the current study.

Finally, the results from this study generalize to a sample of foster mothers and infants who remain together for approximately six months or less. However, it is unclear whether the associations observed in the current study might generalize to a population of foster mothers and infants whose relationships last for shorter or longer periods of time. Given that numerous studies have suggested that oxytocin plays a significant role in the expression of parental behavior among fathers (Feldman et al., 2010; Gordon et al., 2010a; Gordon et al, 2010b; Naber, van IJzendoorn, Deschamps, van Engeland, & Bakersman-Kranenburg, 2010), it also seems important to assess whether these results generalize to foster parenting among father-infant dyads. Furthermore, infants are a subset of the total number of children who are cared for by foster parents; it would be interesting to examine whether these findings extend to a population of foster parents caring for older children.

Despite these limitations, the results from the current study provide information on the biological variables that are associated with the expression of high quality parenting behavior toward non-biologically-related infants. This study suggests that biological systems may be associated with the development of a maternal attachment to foster infants. This has important clinical implications for populations of non-biologically related and biologically related parents and infants. In summary, we believe that these results not only help us understand why individuals vary in the quality of caregiving provided to a specialized group of at risk infants, but also may aid in the development of interventions designed to promote healthy parent-infant relationships more broadly.

Figure 1.

Figure 1

Figure 1

Open in a new tab

Foster mothers’ P3 amplitude to foster, unfamiliar, and familiar images: Time 1 and time 2.

Please note that the convention is to plot ERP waveforms with negative voltages upwards and positive voltages downwards.

Acknowledgments

This work was supported by National Institute of Mental Health Grants F31MH085423to the first author and R01MH084135, R01MH052135, and R01MH074374to the second author. We acknowledge the support of the Delaware Division of Family Services and the caseworkers, foster families, and children involved in this project.

Contributor Information

Johanna Bick, Yale University, Child Study Center.

Mary Dozier, University of Delaware, Department of Psychology.

Kristin Bernard, University of Delaware, Department of Psychology.

Robert Simons, University of Delaware, Department of Psychology.

Damion Grasso, University of Connecticut, Health Center.

References

  1. Aston-Jones G, Cohen JD. An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance. Annual Review of Neuroscience. 2005;28:403–450. doi: 10.1146/annurev.neuro.28.061604.135709. [DOI] [PubMed] [Google Scholar]
  2. Atzil S, Hendler T, Feldman R. Specifying the neurobiological basis of human attachment: Brain, hormones, and behavior in synchronous and intrusive mothers. Neuropsychopharmacology. 2011;36:2603–2615. doi: 10.1038/npp.2011.172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bakermans-Kranenburg M, Breddels-van Baardewijk P, Juffer F, Velderman MK, van IJzendoorn MH. Insecure mothers with temperamentally reactive infants: A chance for intervention. In: Juffer F, Bakermans-Kranenburg MJ, van IJzendoorn MH, editors. Promoting positive parenting: An attachment-based intervention. New York, NY: Taylor & Francis Group/Lawrence Erlbaum Associates; 2008. pp. 75–90. [Google Scholar]
  4. Bates B, Dozier M. Unpublished manuscript. University of Delaware; Newark: 1998. “This is My Baby” coding manual. [Google Scholar]
  5. Bates BC, Dozier M. The importance of maternal state of mind regarding attachment and infant age at placement to foster mothers’ representations of their foster infants. Infant Mental Health Journal. 2002;23:417–431. doi: 10.1002/imhj.10022. [DOI] [Google Scholar]
  6. Bernard K, Dozier M. This is my baby: Foster parents’ feelings of commitment and displays of delight. Infant Mental Health Journal. 2011;32:251–262. doi: 10.1002/imhj.20293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bick J, Dozier M. Mothers’ concentrations of oxytocin following close, physical interactions with biological and nonbiological children. Developmental Psychobiology. 2010;52:100–107. doi: 10.1002/dev.20411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Blair C. Early intervention for low birth weight, preterm infants: The role of negative emotionality in the specification of effects. Development And Psychopathology. 2002;14:311–332. doi: 10.1017/S0954579402002079. [DOI] [PubMed] [Google Scholar]
  9. Burri A, Heinrichs M, Schedlowski M, Kruger TH. The acute effects of intranasal oxytocin administration on endocrine and sexual function in males. Psychoneuroendocrinology. 2008;33:591–600. doi: 10.1016/j.psyneuen.2008.01.014. [DOI] [PubMed] [Google Scholar]
  10. Bustan D, Sagi A. Effects of early hospital-based intervention on mothers and their preterm infants. Journal of Applied Developmental Psychology. 1984;5:305–317. doi: 10.1016/0193-3973(84)90004-2. [DOI] [Google Scholar]
  11. Caputo DV, Taub HB, Goldstein KM, Smith N, Dalack JD, Pursner JP, Silberstein RM. An evaluation of various parameters of maturity at birth as predictors of development at one year. Perceptual and Motor Skills. 1974;39:631–652. doi: 10.2466/pms.1974.39.1.631. [DOI] [PubMed] [Google Scholar]
  12. Carson DS, Hunt GE, Guastella AJ, Barber L, Cornish JL, Arnold JC, McGregor IS. Systemically administered oxytocin decreases methamphetamine activation of the subthalamic nucleus and accumbens core and stimulates oxytocinergic neurons in the hypothalamus. Addiction Biology. 2010;15:448–463. doi: 10.1111/j.1369-1600.2010.00247.x. [DOI] [PubMed] [Google Scholar]
  13. Carter CS. Neuroendocrine perspectives on social attachment and love. Psychoneuroendocrinology. 1998;23:779–818. doi: 10.1016/S0306-4530(98)00055-9. [DOI] [PubMed] [Google Scholar]
  14. Champagne F, Diorio J, Sharma S, Meaney M. Naturally occurring variations in maternal behavior in the rat are associated with differences in estrogen-inducible central oxytocin receptors. Proceedings of National Academy of Science USA. 2001;98:12736–12741. doi: 10.1073/pnas.221224598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Claypool HM, Hugenberg K, Housley MK, Mackie DM. Familiar eyes are smiling: On the role of familiarity in the perception of facial affect. European Journal of Social Psychology. Special Issue: Familiarity Impacts Person Perception. 2007;37:856–866. doi: 10.1002/ejsp.422. [DOI] [Google Scholar]
  16. Cooper R, Osselton JW, Shaw JC. EEG Technology. 2. Butterworths; London: 1969. [Google Scholar]
  17. Dawood MY, Ylikorkala O, Trivedi D, Gupta R. Oxytocin levels and disappearance rate and plasma follicle-stimulating hormone and luteinizing hormone after oxytocin infusion in men. Journal of Clinical Endocrinology and Metabolism. 1980;50:397–400. doi: 10.1210/jcem-50-2-397. [DOI] [PubMed] [Google Scholar]
  18. Derogatis LR, Spencer PM. Administration and Procedures: BSI Manual I. Baltimore, MD: Clinical Psychometrics Research; 1982. [Google Scholar]
  19. Dozier M, Lindhiem O. This is my child: Differences among foster parents in commitment to their young children. Child Maltreatment. 2006;11(4):338–345. doi: 10.1177/1077559506291263. [DOI] [PubMed] [Google Scholar]
  20. Ermisch A, Barth T, Ruhle HJ, Skopkova J, Hrbas P, Landgraf R. On the blood-brain barrier to peptides: accumulation of labeled vasopressin, DesGlyNH2-vasopressin and oxytocin by brain regions. Endocrinologia Expermentalis. 1985;19:29–37. [PubMed] [Google Scholar]
  21. Farbach SE, Morrell JI, Pfaff DW. Oxytocin induction of short-latency maternal behavior in nulliparous, estrogen-primed female rats. Hormones and Behavior. 1984;18:267–286. doi: 10.1016/0018-506x(84)90016-3. [DOI] [PubMed] [Google Scholar]
  22. Feldman R, Gordon I, Schneiderman I, Weisman O, Zagoory-Sharon O. Natural variations in maternal and paternal care are associated with systematic changes in oxytocin following parent–infant contact. Psychoneuroendocrinology. 2010;35:1133–1141. doi: 10.1016/j.psyneuen.2010.01.013. [DOI] [PubMed] [Google Scholar]
  23. Feldman R, Gordon I, Zagoory-Sharon O. Maternal and paternal plasma, salivary, and urinary oxytocin and parent–infant synchrony: Considering stress and affiliation components of human bonding. Developmental Science. 2011;14:752–761. doi: 10.1111/j.1467&#x02013;7687.2010.01021.x. [DOI] [PubMed] [Google Scholar]
  24. Feldman R, Weller A, Zagoory-Sharon O, Levine A. Evidence for a neuroendocrinological foundation of human affiliation: Plasma oxytocin levels across pregnancy and the postpartum period predict mother–infant bonding. Psychological Science. 2007;18:965–970. doi: 10.1111/j.1467-9280.2007.02010.x. [DOI] [PubMed] [Google Scholar]
  25. Fisher PA, Burraston B, Pears K. The Early Intervention Foster Care Program: Permanent Placement Outcomes From a Randomized Trial. Child Maltreatment. 2005;10:61–71. doi: 10.1177/1077559504271561. [DOI] [PubMed] [Google Scholar]
  26. Galbally M, Lewis A, van IJzendoorn M, Permezel M. The role of oxytocin in mother-infant relations: A systematic review of human studies. Harvard Review of Psychiatry. 2010;19:1–14. doi: 10.3109/10673229.2011.549771. [DOI] [PubMed] [Google Scholar]
  27. Gartstein MA, Rothbart MK. Studying infant temperament via the revised infant behavior questionnaire. Infant Behavior & Development. 2003;26:64–86. doi: 10.1016/S0163-6383(02)00169-8. [DOI] [Google Scholar]
  28. Geenen V, Adam F, Baro V, Mantanus H, Ansseau M. Inhibitory influence of oxytocin infusion on contingent negative variation and some memory tasks in normal men. Psychoneuroendocrinology. 1988;5:367–375. doi: 10.1016/0306-4530(88)90043-1. [DOI] [PubMed] [Google Scholar]
  29. Gordon I, Zagoory-Sharon O, Leckman JF, Feldman R. Oxytocin and the development of parenting in humans. Biological Psychiatry. 2010a;68:377–382. doi: 10.1016/j.biopsych.2010.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Gordon I, Zagoory-Sharon O, Leckman JF, Feldman R. Oxytocin, cortisol, and triadic family interactions. Physiology & Behavior. 2010b;101:679–684. doi: 10.1016/j.physbeh.2010.08.008. [DOI] [PubMed] [Google Scholar]
  31. Grasso DJ, Moser JS, Dozier M, Simons R. ERP correlates of attention allocation in mothers’ processing faces of their children. Biological Psychology. 2009;81:95–102. doi: 10.1016/j.biopsycho.2009.03.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Insel TR. Postpartum increases in brain oxytocin binding. Neuroendocrinology. 1986;44:515–518. doi: 10.1159/000124694. [DOI] [PubMed] [Google Scholar]
  33. Jasper HH. Report of the committee on methods of clinical examination in electroencephalography. Electroencephalography and Clinical Neurophysiology. 1958;10:370–371. [Google Scholar]
  34. Kendrick KM, Keverne EB, Baldwin BA. Intracerebroventricular oxytocin stimulates maternal behaviour in the sheep. Neuroendocrinology. 1987;46:56–61. doi: 10.1159/000124796. [DOI] [PubMed] [Google Scholar]
  35. Landgraf R, Ermisch A, Hess J. Indications for a brain uptake of labeled vasopressin and ocytocin and the problem of the blood–brain barrier. Endokrinologie. 1979;73:77–81. [PubMed] [Google Scholar]
  36. Landgraf R, Neumann I. Vasopressin and oxytocin release within the brain: A dynamic concept of multiple and variable modes of neuropeptides communication. Frontiers in Neuroendocrinology. 2004;25:150–176. doi: 10.1016/j.yfrne.2004.05.001. [DOI] [PubMed] [Google Scholar]
  37. Langeslag SJ, Jansma BM, Franken IH, Van Strien JW. Event-related potential responses to love-related facial stimuli. Biological Psychology. 2007;76(1–2):109–115. doi: 10.1016/j.biopsycho.2007.06.007. [DOI] [PubMed] [Google Scholar]
  38. Levine A, Zagoory-Sharon O, Feldman R, Weller A. Oxytocin during pregnancy and early postpartum: Individual patterns and maternal fetal attachment. Peptides. 2007;28:1162–1169. doi: 10.1016/j.peptides.2007.04.016. [DOI] [PubMed] [Google Scholar]
  39. Lewis E, Dozier M, Ackerman J, Sepulveda-Kozakowski S. The effect of caregiving instability on adopted children’s inhibitory control abilities and oppositional behavior. Developmental Psychology. 2007;43:1415–1427. doi: 10.1037/0012-1649.43.6.1415. [DOI] [PubMed] [Google Scholar]
  40. Lindhiem O, Dozier M. Caregiver commitment to foster children: The role of child behavior. Child Abuse & Neglect. 2007;31:361–374. doi: 10.1016/j.chiabu.2006.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mens WB, Witter A, van Wimersma-Greidanus TB. Penetration of neurohypophyseal hormones from plasma into cerebrospinal fluid (CSF): Half-times of disappearance of these neuropeptides from CSF. Brain Research. 1983;262:143–149. doi: 10.1016/0006-8993(83)90478-x. [DOI] [PubMed] [Google Scholar]
  42. Minde K. Prematurity and serious medical illness in infancy: Implications for development and intervention. In: Zeanah CH Jr, editor. Handbook of infant mental health. New York, NY, US: Guilford Press; 1993. pp. 87–105. [Google Scholar]
  43. Naber F, van IJzendoorn MH, Deschamps P, van Engeland H, Bakermans-Kranenburg MJ. Intranasal oxytocin increases fathers’ observed responsiveness during play with their children: A double-blind within-subject experiment. Psychoneuroendocrinology. 2010;35:1583–1586. doi: 10.1016/j.psyneuen.2010.04.007. [DOI] [PubMed] [Google Scholar]
  44. Neumann I. Stimuli and consequences of dendriticrelease of oxytocin within the brain. Biochemical Society Transactions. 2007;35:1252–1257. doi: 10.1042/BST0351252. [DOI] [PubMed] [Google Scholar]
  45. Neumann I. Brain oxytocin: A key regulator of emotional and social behaviours in both females and males. Journal of Neuroendocrinology. 2008;20:858–865. doi: 10.1111/j.1365-2826.2008.01726.x. [DOI] [PubMed] [Google Scholar]
  46. Olazábal DE, Young LJ. Species and individual differences in juvenile female alloparental care are associated with oxytocin receptor density in the striatum and the lateral septum. Hormones and Behavior. 2006;49:681–687. doi: 10.1016/j.yhbeh.2005.12.010. [DOI] [PubMed] [Google Scholar]
  47. Pedersen CA, Prange AJ. Evidence that central oxytocin plays a role in activation of maternal behavior. In: Krasnegor NA, Blass EM, Hofer MA, Smotherland WP, editors. Perinatal Development: A Psychobiological Perspective. Academic Press; Orlando: 1979. pp. 299–320. [Google Scholar]
  48. Pedersen CA, Vadlamudi SV, Boccia ML, Amico JA. Maternal behavior deficits in nulliparous oxytocin knockout mice. Genes, Brain and Behavior. 2006;5:274–281. doi: 10.1111/j.1601-183X.2005.00162.x. [DOI] [PubMed] [Google Scholar]
  49. Riem M, Bakermans-Kranenburg M, Pieper S, Tops M, Boksem MS, Vermeiren RM, Rombouts SB. Oxytocin modulates amygdala, insula, and inferior frontal gyrus responses to infant crying: A randomized controlled trial. Biological Psychiatry. 2011;70:291–297. doi: 10.1016/j.biopsych.2011.02.006. [DOI] [PubMed] [Google Scholar]
  50. Riem M, van Ijzendoorn M, Tops M, Boksem M, Rombouts S, Bakermans-Kranenburg M. No Laughing Matter: Intranasal Oxytocin Administration Changes Functional Brain Connectivity during Exposure to Infant Laughter. Neuropsychopharmacology. doi: 10.1038/npp.2011.313. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Rothbart MK. Measurement of temperament in infancy. Child Development. 1981;52:569–578. doi: 10.2307/1129176. [DOI] [Google Scholar]
  52. Sameroff A, Seifer R, Zax M, Barocas R. Early indicators of developmental risk: Rochester longitudinal study. Schizophrenia Bulletin. 1987;13:383–394. doi: 10.1093/schbul/13.3.383. [DOI] [PubMed] [Google Scholar]
  53. Schupp HT, Cuthbert BN, Bradley MM, Hillman CH, Hamm AO, Lang PJ. Brain processes in emotional perception: Motivated attention. Cognition and Emotion. 2004;18:593–611. doi: 10.1080/02699930341000239. [DOI] [Google Scholar]
  54. Schuler ME, Black MM, Starr RH. Determinants of mother-infant interaction: Effects of prenatal drug exposure, social support, and infant temperament. Journal Of Clinical Child Psychology. 1995;24:397–405. doi: 10.1207/s15374424jccp2404_4. [DOI] [Google Scholar]
  55. Seltzer LJ, Ziegler TE. Non-invasive measurement of small peptides in the common marmoset (Callithrix jacchus): A radiolabeled clearance study and endogenous excretion under varying social conditions. Hormones and Behavior. 2007;51:436–442. doi: 10.1016/j.yhbeh.2006.12.012. [DOI] [PubMed] [Google Scholar]
  56. Seltzer L, Ziegler T, Pollak S. Social vocalizations can release oxytocin in humans. Proceedings of Biological Science. 2010;7:2661–2666. doi: 10.1098/rspb.2010.0567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Strathearn L, Fonagy P, Amico J, Montague PR. Adult attachment predicts maternal brain and oxytocin response to infant cues. Neuropsychopharmacology. 2009;34:2655–2666. doi: 10.1038/npp.2009.103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tribollet E, Dubois-Dauphin M, Dreifuss J, Barberis C, Jard S. Oxytocin receptors in the central nervous system. Distribution, development, and species differences. Annual New York Academy of Science. 1992;652:29–38. doi: 10.1111/j.1749-6632.1992.tb34343.x. [DOI] [PubMed] [Google Scholar]
  59. Turner RA, Altemus M, Enos T, Cooper B, McGuinness T. Preliminary research on plasma oxytocin in normal cycling women: Investigating emotion and interpersonal distress. Psychiatry: Interpersonal and Biological Processes. 1999;62(2):97–113. doi: 10.1080/00332747.1999.11024859. [DOI] [PubMed] [Google Scholar]
  60. van den Boom DC. The influence of temperament and mothering on attachment and exploration: An experimental manipulation of sensitive responsiveness among lower-class mothers with irritable infants. Child Development. 1994;65:1457–1477. doi: 10.2307/1131511. [DOI] [PubMed] [Google Scholar]
  61. Van Leengoed E, Kerker E, Swanson HH. Inhibition of postpartum maternal behavior in the rat by injecting an oxytocin antagonist into the cerebral ventricles. Journal of Endocrinology. 1987;112:275–282. doi: 10.1677/joe.0.1120275. [DOI] [PubMed] [Google Scholar]
  62. Vico C, Guerra P, Robles H, Vila J, Anllo-Vento L. Affective processing of loved faces: Contributions from peripheral and central electrophysiology. Neuropsychologia. 2010;48:2894–2902. doi: 10.1016/j.neuropsychologia.2010.05.031. [DOI] [PubMed] [Google Scholar]
  63. Weisman O, Feldman R, Goldstein Parental and romantic attachment shape brain processing of infant cues. Biological Psychology. doi: 10.1016/j.biopsycho.2011.11.008. in press. [DOI] [PubMed] [Google Scholar]
  64. Wotjak CT, Ganster J, Kohl G, Holsboer F, Landgraf R, Engelmann M. Dissociated central and peripheral release of vasopressin, but not oxytocin, in response to repeated swim stress: New insights into the secretory capacities of peptidergic neurons. Neuroscience. 1998;85:1209–1222. doi: 10.1016/s0306-4522(97)00683-0. [DOI] [PubMed] [Google Scholar]
  65. Zeanah CH, Larrieu JA, Heller S, Valliere J, Hinshaw-Fuselier S, Aoki Y, Drilling M. Evaluation of a preventive intervention for maltreated infants and toddlers in foster care. Journal Of The American Academy Of Child & Adolescent Psychiatry. 2001;40:214–221. doi: 10.1097/00004583-200102000-00016. [DOI] [PubMed] [Google Scholar]

RESOURCES