Personality, adrenal steroid hormones, and resilience in maltreated children: A multi-level perspective

Participants

The participants in this investigation included 677 children aged six to twelve (M age = 9.54, sd = 2.26) who attended a summer camp research program designed for school-aged low-income children during the years 1999 to 2004. Some children attended the camp for multiple years, and the data from their first year of attendance during the indicated span of years were selected for inclusion in the current study. The sample included both maltreated children (n = 347) and nonmaltreated children (n = 330). Among the participants, 54.8 % were boys. The maltreated and nonmaltreated children were comparable in terms of racial/ethnic diversity and family characteristics; 61.9% of the sample was African American, 19.1% was white, 17.0% was Hispanic, and 2.1% was from other racial/ethnic groups. The families of the children were low income, with 94.8% of the families having a history of receiving welfare benefits. Single mothers headed 62.9% of the families.

Parents of all maltreated and nonmaltreated children provided informed consent for their child’s participation, as well as consent for examination of any Department of Human Services (DHS) records pertaining to the family. Children in the maltreated group had been identified by the county DHS as having experienced child abuse and/or neglect, and the sample was representative of the children in families receiving services from the DHS. Comprehensive searches of DHS records were completed, and maltreatment information was coded utilizing operational criteria from maltreatment nosology specified in the Maltreatment Classification System (MCS: Barnett, Manly, & Cicchetti, 1993).

The MCS utilizes DHS records detailing investigations and findings involving maltreatment in identified families over time. Rather than relying on official designations and case dispositions, the MCS codes all available information from DHS records, making independent determinations of maltreatment experiences. Based on operational criteria, the MCS designates all of the subtypes of maltreatment children have experienced (i.e., neglect, emotional maltreatment, physical abuse, sexual abuse). Coding of the DHS records was conducted by trained research assistants, doctoral students, and clinical psychologists. Adequate reliability has been obtained (weighted kappas ranging from .86 to .98) (Manly, Kim, Rogosch, & Cicchetti, 2001). Other investigators have demonstrated that the MCS is reliable and valid in classifying maltreatment (Bolger & Patterson, 2001; Bolger, Patterson, & Kupersmidt, 1998; English et al., 2005; Manly, Cicchetti, & Barnett, 1994; Manly et al., 2001; Smith & Thornberry, 1995; Stouthamer-Loeber, Loeber, Homish, & Wei, 2001).

In terms of the subtypes of maltreatment, neglect involves failure to provide for the child’s basic physical needs for adequate food, clothing, shelter, and medical treatment. In addition to inadequate attention to physical needs, forms of this subtype include lack of supervision, moral-legal neglect, and education neglect. Emotional maltreatment involves extreme thwarting of children’s basic emotional needs for psychological safety and security, acceptance and self-esteem, and age-appropriate autonomy. Examples of emotional maltreatment of increasing severity include belittling and ridiculing the child, extreme negativity and hostility, exposure to severe marital violence, abandoning the child, and suicidal or homicidal threats. Physical abuse involves the non-accidental infliction of physical injury on the child (e.g., bruises, welts, burns, choking, broken bones). Injuries range from minor and temporary to permanently disfiguring. Finally, sexual abuse involves attempted or actual sexual contact between the child and caregiver for purposes of the caregiver’s sexual satisfaction or financial benefit. Events range from exposure to pornography or adult sexual activity, to sexual touching and fondling, to forced intercourse with the child.

Children in the maltreatment group all had documented histories of abuse and/or neglect occurring in their families according to DHS records. However, DHS record information was not complete enough to code maltreatment subtype information for 50 (14.4%) of the maltreated children. Among the remaining maltreated children, 76.4% had experienced neglect, 68.0% had experienced emotional maltreatment, 32.7% had experienced physical abuse, and 9.1% had experienced sexual abuse. As is typical in maltreated populations (Bolger et al., 1998; Manly et al., 1994; 2001), the majority of children had experienced multiple subtypes of maltreatment. Specifically, 63.9% of the maltreated children had experienced two or more maltreatment subtypes, and 13 of the 15 possible combinations of presence/absence for the four maltreatment subtypes were observed in the sample.

In order to consolidate subtype groups for comparison purposes, a hierarchical classification system was used, based on the degree to which the subtype violates cultural standards and the relative frequency of the different forms of maltreatment, as indicated above. Accordingly, any child who had experienced sexual abuse was categorized as sexually abused (n = 27). Children who had been physically abused, but not sexually abused were included in a physical abuse group (n = 87). Children experiencing neglect but not physical or sexual abuse were classified as neglected (n =144), whereas children who had experienced emotional maltreated alone were included in an emotional maltreatment group (n = 39).

Because maltreated children are predominantly from low socioeconomic status families, demographically comparable nonmaltreated children were recruited from families receiving Temporary Assistance for Needy Families (TANF). DHS record searches were completed for these families to verify the absence of any record of child maltreatment. Trained research assistants also interviewed mothers of children recruited for the nonmaltreatment group to confirm a lack of DHS involvement and prior maltreatment experiences. Subsequently, record searches were conducted in the year following camp attendance to verify that all available information had been accessed. Only children from families without any history of documented abuse and/or neglect were retained in the nonmaltreatment group. In addition, families who had received preventive services through DHS due to concerns over risk for maltreatment also were excluded from the nonmaltreated comparison group to reduce the potential for unidentified maltreatment existing within this group.

Procedure

Children attended a week-long day camp program and participated in research assessments. At the camp, children were assigned to groups of eight same-age and same-sex peers; half of the children assigned to each group were maltreated. Each group was conducted by three trained camp counselors, who were unaware of the maltreatment status of children and the hypotheses of the study. Each day of camp lasted for seven hours, providing 35 hours of interaction between children and counselors. In addition to the recreational activities, after providing assent, children participated in various research assessments (see Cicchetti & Manly, 1990, for detailed descriptions of camp procedures) and provided morning and afternoon saliva samples. Trained research assistants, who also were unaware of research hypotheses and maltreatment status, conducted individual research sessions with children, in which questionnaires and other research measures were administered. Clinical consultation and intervention occurred if any concerns over danger to self or others emerged during research sessions. At the end of the week, children in each group completed sociometric ratings of their peers. The counselors, who had been trained extensively for two weeks prior to the camp, also completed assessment measures on individual children, based on their observations and interactions with children in their respective groups. Counselors also were unaware of the research hypotheses and children’s maltreatment status.

Measures

The measures described below constitute a subset of assessments conducted during the research camp. The camp context and associated measurement battery provide a multi-informant, multi-perspective view of child adaptive and personality functioning. Measures include child self-report, peer evaluations, counselor observations, and counselor-report assessments of individual children, as well as school record data obtained from children’s school districts. Assay determinations also were made to determine children’s cortisol and DHEA levels during the morning and afternoon across the days of the camp.

Saliva Collection and Preparation

The camp context allowed for a setting where a consistent collection of saliva samples could occur at uniform times across the camp week. Each day upon arrival to the camp at 9:00 a.m., trained research assistants obtained saliva samples from each child. Samples also were obtained each day prior to the children’s departure from the camp at 4:00 p.m. Cortisol was assayed from the saliva samples for each day and time across the week. Larger saliva samples were obtained from children on Tuesday and Thursday mornings and afternoons so that DHEA also could be assayed.

The children had not consumed food or drink for at least 30 minutes before each saliva sample was obtained. Samples were collected following Granger et al. (1999). The children were asked to chew Trident® sugarless original flavor gum to stimulate saliva flow, and then passively drool though a short drinking straw into a 20 ml plastic vial. The samples were immediately frozen and stored at −40° C. Each week, the samples were shipped overnight on dry ice for next day delivery to Salimetrics Laboratories (State College, PA) for assay. Each sample was thawed, 4–5 1 mL aliquots were placed into 1.8 ml cryogenic storage vials and all aliquots were then frozen at −80°C until assayed. On the day of testing, samples were brought to room temperature, centrifuged at 3,000 RPM for 15 minutes, and the clear top-phase of the sample was pipetted into appropriate test tubes/wells.

Salivary Hormone Determination

Salivary cortisol was assayed using a commercially available enzyme immunoassay kit (Salimetrics, State College, PA). The test uses 25 μl of saliva, has a lower limit of sensitivity of 0.007 μg/dl (range up to 1.8 μg/dl), and average intra-and inter-assay coefficients of variation less than 5.0% and 10.0%, respectively. There is no detectable cross-reactivity of the antibody used in this assay with DHEA. Units of cortisol are expressed in μg/dl (micrograms per deciliter). The distributions of cortisol values for each day and time period were substantially skewed. Thus, prior to analyses all values were log transformed. The log transformed cortisol values for the morning samples were averaged to obtain a mean morning cortisol measurement. Similarly, the log transformed late afternoon samples values were averaged across days to derive an afternoon cortisol measure for analysis.

Salivary dehydroepiandrosterone determinations were made using an enzyme immunoassay kit (Salimetrics, State College, PA). The test uses 50 μl of saliva and has a minimum detection limit of 10.0 pg/mL (range up to 1000 pg/ml). Average intra-and inter-assay coefficients of variation were less than 5.0% and 15.0%, respectively. There is no cross-reactivity of the antibody used in this assay with cortisol. Units of DHEA are expressed as pg/ml (picograms per milliliter). The distributions of DHEA values also were found to be highly skewed, and all values were log transformed for analyses. DHEA log-transformed values for the Tuesday and Thursday morning samples were averaged to obtain a mean morning DHEA measurement. Similarly, the log-transformed later afternoon sample values on Tuesday and Thursday were averaged to obtain a mean afternoon DHEA value.

The cortisol to DHEA ratio also was calculated. Cortisol values in μg/dl units and DHEA values in pg/mL units were respectively transformed into moles per liter (nmol/L) values. This transformation was done for the Tuesday and Thursday, morning and afternoon, cortisol and DHEA values, and ratios were calculated for each day and time. As with the cortisol and DHEA distributions, the cortisol/DHEA ratio distributions evinced high skew. For analytic purposes, the ratios were log transformed prior to analysis. The two log transformed morning values were averaged, as were the two log transformed afternoon values to yield morning and afternoon cortisol/DHEA ratio measurements.

Child self-report measures

Peer measures

Counselor Measures

School Record Data

Composite of Resilient Functioning

Consistent with prior work examining resilience in maltreated and nonmaltreated children (Cicchetti & Rogosch, 1997; Cicchetti et al., 1993), a composite of resilient functioning was derived from indicators of children functioning well in domains of particular developmental importance for school-aged children. Initially, three composite indicators of social competence, combining the perspectives of peers and adults, were generated. First, a prosocial composite combined peer nomination scores for leader and cooperative with counselor behavior rating scores for prosocial behavior and counselor PEI scores for likeability. Internal consistencies of this composite were internally consistent with alphas ranging from .66 to .75 (M = .70). Second, a disruptive-aggressive composite was comprised of peer nominations of disruptive and fights, and counselor behavior ratings of aggressive and counselor PEI scores for aggressiveness. The internal consistencies for this composite across the years ranged from .80 to .87 (M = .85). Third, a withdrawn composite included peer nomination scores for shy, counselor behavior ratings for withdrawn, and counselor PEI scores for the withdrawn subscale. Alphas ranged from .68 to .78 (M = .69). For each of these internally consistent composites, component variables were standardized within each hear of camp, and the standardized variables were averaged to generate the composite scores for each child.

In addition to these three indicators of social competence, the child’s self-report of depressive symptoms on the CDI, counselor assessment of internalizing and externalizing behavior problems from the TRF, and the school risk index were other components of the resilience composite. For each of the seven indicators, criteria for the most competent functioning was established. More specifically, children were considered as demonstrating competent functioning if they had scores in the top third of the distribution for the prosocial composite, scored in the lowest one third of the distribution for the aggressive composite, the withdrawn composite, the CDI score, and TRF internalizing and externalizing scores, and had no school risk indicator. Children meeting the criterion on a given dimension were given a score of 1; all other children were given a score of 0 for that dimension. Summing across the seven indicators produced a composite index of resilient functioning with a possible range of scores ranging from 0 to 7.

Plan of analysis

Initially, differences between maltreated children in average levels of resilient functions scores are examined, as well as differences in the rate of high and low resilience. In subsequent analyses, we begin to build a multi-level model of influences on resilient functioning, incorporating the effects of maltreatment, cortisol, and DHEA. The relations among these domains are examined in two ways. First, each of the hormones and maltreatment are examined as predictors of mean level differences in resilient functioning. Second, at a more person-centered level, groups of maltreated and nonmaltreated children with varying levels of resilient functioning are examined to determine if there are systematic differences in hormone levels for different types of children. The joint effects of cortisol and DHEA on resilient functioning also are examined. Then, the influences of the personality variables, ego resiliency and ego control, are examined as predictors of resilient functioning. Finally, from a multi-level perspective, the multiple influences of maltreatment, cortisol, DHEA, and the personality variables are examined together to investigate potential overlap of influence versus independence of their contributions.

Maltreatment and Resilience

As expected, maltreated (n = 347) and nonmaltreated children (n = 330) differed significantly on the composite of resilient functioning, t (675) = 6.02, p < .001. Maltreated children (M = 1.76, sd = 1.51) had lower average resilience scores than nonmaltreated children (M = 2.48, sd = 1.60). Maltreatment subtype groups also demonstrated significant differences, F (4,622) = 9.92, p < .001. Tukey post hoc comparisons indicated that physical abused children (M = 1.51, sd = 1.40) and sexually abused children (M = 1.59, sd = 1.31) had significantly lower resilient functioning scores than nonmaltreated children. Emotionally maltreated children (M = 1.90, sd = 1.76) and neglected children (M = 1.85, sd = 1.53) had intermediate scores between the nonmaltreated children and the two abuse groups, did not differ significantly from either the physically or sexually abused group or from the nonmaltreated group.

Resilience groups were defined as low (scores of 0 or 1), medium (scores of 2, 3 or 4), and high (scores of 5 or more) based on resilient functioning scores. The maltreated and nonmaltreated children differed in terms of their distribution across these resilience groups. A greater proportion of nonmaltreated children (11.8%) were in the high resilience group as compared to the maltreated children (6.1%). In contrast, a greater proportion of maltreated children (49.9%) than nonmaltreated children (29.4%) were in the low resilience group. Among the maltreated children, subtype groups did not differ in the rates of low, medium, or high resilience, χ² (6, N = 297) = 4.63, ns.

Cortisol and maltreatment in relation to resilient functioning scores

Resilience groups and maltreatment in relation to cortisol

Whether maltreated and nonmaltreated children with different levels of resilient functioning (low, medium, and high) differed in their average level of morning and afternoon cortisol was examined. An ANCOVA, with age and gender as covariates, was conducted with maltreatment status and resilience group as main effects to examine relations with morning cortisol. Maltreatment status, resilience group, and their interaction were all nonsignificant. The same ANCOVA was again conducted with afternoon cortisol as the dependent variable, and the main and interaction effects were nonsignificant. Finally, diurnal variation in cortisol in relation to maltreatment or resilience groups also was examined. A repeated measures ANCOVA was conducted with morning and afternoon cortisol levels as the within-subjects dependent variables, maltreatment status and resilience group as the main effects, and age and gender as covariates. The main effects and interaction of maltreatment and resilience group were not significant. Thus, resilience groups in the maltreated and nonmaltreated groups were not distinguished based on their levels of morning or afternoon cortisol, or diurnal variation in cortisol from morning to afternoon.

DHEA and maltreatment in relation to resilient functioning scores

Hierarchical regression analyses were conducted to determine if maltreatment, morning and afternoon DHEA, and their interaction were related to resilient functioning scores. In these analyses, age and gender were included as covariates. For morning DHEA levels, significant effects were obtained for maltreatment status, β = −.21, p < .001, and morning DHEA level, β = −.13, p = .001, adjusted R² = .068, p < .001. Nonmaltreatment status and lower morning DHEA were independently related to higher resilient functioning scores. The interaction of maltreatment and morning DHEA was not significant.

Similar effects were found for afternoon DHEA levels. In the hierarchical regression analysis, maltreatment status, β = −.21, p < .001, and afternoon DHEA levels, β = −.13, p = .001, were both significant predictors of resilient functioning scores, adjusted R² = .067, p < .001. The interaction of maltreatment status and afternoon DHEA levels was not significant. Thus, in addition to maltreatment status, both lower morning DHEA levels and lower afternoon DHEA levels were found to relate to higher resilient functioning scores, and this did not differ by maltreatment status.

Resilience groups and maltreatment status in relation to morning and afternoon DHEA

Maltreated and nonmaltreated children in the low, medium, and high resilience groups also were examined to determine if there were differences related to variation in average levels of morning and afternoon DHEA. For morning DHEA, an ANCOVA was conducted with maltreatment status and resilience group as main effects, and age and gender as covariates. The resilience group effect was significant, F (2,667) = 5.30, p = .005, whereas the effect of maltreatment and the maltreatment by resilience group interaction were not significant. An analogous ANCOVA was conducted for afternoon cortisol, and similar results were obtained. The main effect for resilience group was significant, F (2, 662) = 5.52, p = .004. However, maltreatment status and the interaction of maltreatment status and resilience group were not significant. Based on Tukey post hoc tests on covariate-adjusted means, for both morning DHEA and afternoon DHEA, the high resilience group had significantly lower morning and afternoon DHEA than the low resilience group. The medium resilience group did not differ from the low or high resilience groups.

A more elaborate picture emerged when diurnal variation in DHEA was examined for the maltreatment and resilience groups. A repeated measures ANCOVA was conducted, with morning and afternoon DHEA as the within-subjects factor, and maltreatment status and resilience group as between subjects factors. Age and gender were included as covariates. Significant within-subjects effects were obtained for resilience groups, F (2,651) = 3.06, p = .048, but this effect was clarified by a significant maltreatment by resilience group interaction, F (2, 651) = 4.07, p = .018. The pattern of this interaction effect is depicted in Figure 3. Among nonmaltreated children only, a repeated measures ANCOVA for resilience group did not result in a significant within-subjects effect for morning to afternoon DHEA, F (3,318) = .68, ns. In the figure, there is a relative stability in DHEA from morning to afternoon with a subtle decline in DHEA for the medium and high resilience groups. Among the maltreated children, the within subjects effect was significant, F (2,331) = 5.03, p = .007. As shown in Figure 3, the low and medium resilience groups show a similar pattern of decreasing DHEA from morning to afternoon and a decrease in DHEA level with higher levels of resilience. However, the high resilience maltreated children exhibited a distinctive pattern of diurnal variation. Rather than showing a decrease in DHEA from morning to afternoon, the high resilience maltreated children had a notable increase in afternoon DHEA.

In addition to mean level differences, another perspective on this diurnal variation was examined. In particular, whether children demonstrated a higher level of DHEA in the afternoon was determined. Among nonmaltreated children, the percentage of children in the three resilience groups who had a higher level of DHEA in the afternoon was not significantly different, χ² (2, n = 323) = .05, ns., with 45.4%, 45.5%, and 43.6% of children exhibiting an increase, for the low, medium, and high resilience groups, respectively. In contrast, among the maltreated children, the group difference was significant, χ² (2, n = 336) = 7.49, p = .024. Whereas 42.4% of the low and 40.0 % of the medium resilience group maltreated children had an increase in DHEA in the afternoon, among the high resilience maltreated children, this increase was found among 71.4% of the children. Thus, the high resilience group maltreated children stands out in terms of greater mean level DHEA levels in the afternoon and a higher proportion of the group exhibiting a morning to afternoon rise in DHEA levels.

Cortisol/DHEA ratio, maltreatment, and resilient functioning

The dual contributions of cortisol and DHEA to resilient functioning were examined with the cortisol/DHEA ratio. A hierarchical multiple regression analysis to predict resilient functioning was conducted with maltreatment status and the morning cortisol/DHEA ratio as main effects; age and gender were included as covariates. Significant main effects for maltreatment, β = −.20, p < .001, and for the morning cortisol/DHEA ratio, β = .11, p = .007, were obtained, adjusted R² = .063, p < .001. The interaction of maltreatment and morning cortisol/DHEA ration was not significant. Thus, in addition to the contribution of maltreatment, higher cortisol/DHEA ratios were related to higher levels of resilient functioning. Comparable findings were found for the afternoon cortisol/DHEA ratio. In a hierarchical multiple regression analysis, significant main effects for maltreatment, β = −.21, p < .001, and for the afternoon cortisol/DHEA ratio, β = .11, p = .008, were found, R² = .064, p < .001, whereas the interaction of maltreatment and the afternoon cortisol/DHEA ratio was not significant. As with the morning cortisol/DHEA ratio, higher values of the afternoon cortisol/DHEA ratio were related to higher levels of resilient functioning.

Independent effects of maltreatment, cortisol, and DHEA on resilient functioning

Given the previous findings, particularly for morning cortisol and DHEA, the independent and potentially interactive effects of these hormones on resilient functioning, in addition to maltreatment status were examined. In a hierarchical regression analysis, the covariates of age and gender were entered first, followed by the main effects of maltreatment, morning cortisol, and morning DHEA. In step three, terms for the two-way interactions among maltreatment, morning cortisol, and morning DHEA were added, followed by the three-way interaction in step four. See Table 1.

In step one, age was a significant covariate, β = .12, p =. 003. The main effects were added in step two, and consistent with prior analyses, maltreatment status, β = −.21, p < .001, and morning DHEA, β = −.13, p = .002, were significant main effects. Among the two-way interactions entered in step three, only the maltreatment by morning cortisol interaction was significant, β = −.15, p = .01. The three-way interaction of maltreatment, cortisol, and DHEA entered in step 4 was not significant. Thus, previously demonstrated relations of maltreatment, morning DHEA, and the interaction of maltreatment and cortisol in predicting resilient functioning each maintained significant effects when examined together. Additional interactive effects between cortisol and DHEA were not observed.

Ego resiliency, ego control, and maltreatment in relations to resilient functioning

Next, the influences of the personality variables, ego resiliency and ego control, in conjunction with maltreatment, were examined for their role in predicting individual differences in resilient functioning. Maltreated and nonmaltreated children differed significantly on both personality variables, ego resiliency: t (675) = 6.82, p < .001; ego control: t (670) = 4.36, p< .001. Maltreated children had lower ego resiliency, M = .23, sd = .31, and higher ego undercontrol, M = .06, sd = .25, than nonmaltreated children, ego resiliency: M = .39, sd = .29, ego control: M = −.02, sd = .22.

To investigate the joint influences, a hierarchical regression was conducted, with age and gender as covariates. In step two, maltreatment was added, whereas the personality variables of ego resiliency and ego control were added in step three. The two-way interaction terms were included in step four, and the three-way interaction term in step five. See Table 2.

As in prior analyses, in step 2, maltreatment had an independent effect in predicting resilient functioning, β = −.22, p < .001. However, with the entry of ego resiliency and ego control, in step 3, the independent effect of maltreatment status was no longer significant, indicating that the personality variables served to mediate the relation between maltreatment status and resilient functioning. Ego resiliency, β = .64, p < .001, and ego (under)control, β = −.08, p = .009, each made independent contributions to prediction, increasing the R² substantially. In the next step, only the two-way interaction of ego resiliency and ego control, β = −.28, p < .001, was found to be significant. The three-way interaction among maltreatment and the two personality variables was not significant. Thus, the personality variables, notably high ego resiliency and greater undercontrol were shown to have strong predictive power in relating to individual differences in children’s resilience striving.

Maltreatment, hormones, and personality in relation to resilient functioning

Based on the prior analyses, a final hierarchical multiple regression analysis to predict resilient functioning was conducted to examine the extent to which the influences of maltreatment, cortisol, DHEA, and the personality variables were maintained when the already demonstrated effects of these domains were considered within one analysis. Accordingly, age and gender were entered in step one, followed by maltreatment status in step two. In step three, morning cortisol, morning DHEA, as well as ego resiliency and ego control were entered. In the final step, the two-way interactions of maltreatment and cortisol and of ego resiliency and ego control were included. See Table 3.

As shown in prior analyses, age was a significant covariate, β = .12, p = .003. When maltreatment was entered in step 2, it also had a significant effect, β = −.21, p < .001; however, the impact of maltreatment was mediated by the personality variables, as they were entered in step 3. In this step, significant effects were found for DHEA, β = −.07, p = .033, ego resiliency, β = .64, p < .001, and ego (under)control, β = −.09, p = .003. Finally, both two-way interaction terms were found to be significant in step 4, including, maltreatment by morning cortisol, β = .11, p = .015, and ego resiliency by ego control, β = −.26, p < .001. Thus, the effects of the personality variables, morning DHEA, and the interaction of maltreatment and morning cortisol were maintained when examined in conjunction. Although not presented here, additional terms were considered for incorporation into these analyses. These terms included interactions between the personality and hormone variables, as well as higher order interactions. However, none of the interactions was found to be significant. Thus, from a multi-level perspective, the relations demonstrated for cortisol, DHEA, and ego resiliency and ego control constituted distinct and independent contributions to understanding individual differences in resilient functioning in high risk maltreated and nonmaltreated children.

Discussion

In this investigation of the developmental pathways to resilience in maltreated and nonmaltreated children, we adopted a multiple-levels-of-analysis perspective and examined the unique and interactive contributions that personality features and adrenal steroid stress hormone regulation make to the achievement of competent functioning in the face of significant stress and adversity. Utilizing a large, representative sample of maltreated children and a well-matched group of nonmaltreated children, we embarked on one of the first empirical studies of biological and psychological contributors to resilience in abused and neglected children (see also Curtis & Cicchetti, 2007).

Personality Features

The personality features, ego resiliency and ego control, and the adrenal steroid hormones associated with stress, cortisol and DHEA, made independent contributions to resilience. Although operating at different levels of analysis, behavioral/psychological and biological factors each made unique contributions to resilience.

Consistent with the results of previous investigations (Cicchetti et al., 1993; Cicchetti & Rogosch, 1997; Flores et al., 2005; Rogosch & Cicchetti, 2004), ego resiliency and ego (over)control were significant predictors of resilience. Moreover, ego resiliency and ego (over)control operated as mediators of the relation between maltreatment and resilient functioning. However, contrary to findings obtained in prior studies of the determinants of resilience in maltreated children, the interaction of ego (over)control and maltreatment status was not significant, indicating that ego control was not operating differentially in the maltreated and comparison groups of children.

Despite the fact that these personality features did not differentially predict resilient functioning in maltreated and nonmaltreated children, the finding that ego resiliency and ego (over)control were predictive of resilience in both maltreated and nonmaltreated children is noteworthy. The resilient maltreated and nonmaltreated children, through adopting a more reserved, controlled, and rational way of interacting and relating with their peers and the adults camp counselors, may be more attuned to behave in ways that are critical for adapting successfully to their stress-laden environments.

Stress Regulation

Masten (2001) has theorized that resilience is a “common phenomenon that results…from the operation of basic human adaptational systems” (p. 227) that work together optimally when confronted with stress, adversity, or trauma. Another aspect of resilience is the ability to recover positive function by bouncing back from negative experiences through adapting flexibly to life’s ongoing vicissitudes (Cicchetti & Rogosch, 1997; Luthar et al., 2000; Masten, Best, & Garmezy, 1990). When neurobiological systems are operating in an optimal fashion, organisms are able to confront and adapt to the stressors and adversity they are facing and respond while continuing to maintain homeostasis (Gunnar, 2007; Gunnar & Vazquez, 2006; Levine, 2005).

Allostasis has been conceptualized as “maintaining stability (or homeostasis) through change” (McEwen, 1998, p. 32). In the process of allostasis, biological homeostatic “set points” are modified to generate the physiological resources necessary to promote successful adaptation in response to environmental challenges (Davies, Sturge-Apple, Cicchetti, & Cummings, in press; Korte, Koolhaas, Wingfield, & McEwen, 2005; Lupien, Ouellet-Morin, Hupbach, Tu, & Buss, et al., 2006; McEwen, 1998). The concept of allostatic load (McEwen & Stellar, 1993) has been invoked to describe the wear-and-tear that the body experiences due to repeated cycles of allostasis that are inefficiently handled. Allostatic load refers to cumulative physiological dysregulation across multiple biological systems (Cooper, Feder, Southwick, & Charney, 2007; Davies et al., in press; McEwen, 1998). Whenever biological systems become activated or overstimulated too frequently, when these systems do not return to baseline following exposure to stressors, or when interconnected biological systems do not respond to one another in an appropriate fashion, then physiological dysregulation eventuates (Cooper et al., 2007; Lupien et al., 2006; McEwen, 1998; Miller, Chen, & Zhou, 2007). Prolonged stress, as is often the case in child maltreatment, can lead to allostatic load through a cascade of causes and sequelae that can damage the brain, organ systems, and the neurochemical balance that undergirds cognition, emotion, mood, personality, and behavior (Cicchetti & Rogosch, 2001a; 2001b; Cooper et al., 2007; Lupien et al., 2006). Interacting genetic, developmental, and experiential factors, including child maltreatment, play prominent roles in the wide array of individual differences among individuals in coping with challenges throughout their lives (Gunnar, 2007; Heim & Nemeroff, 2002; McEwen, 1998).

To date, the vast majority of research on the developmental sequelae of severe psychological stress in humans has focused on discovering the pathways to maladaptation and psychopathology. This investigation is one of the first studies to examine the role that neurobiological responses to stress play in the development of resilient adaptation in maltreated and nonmaltreated children.

In addition to personality features, which operate at the behavioral level, the adrenal steroid hormones, cortisol and DHEA, were found to have associations with resilient functioning at the biological level. On the one hand, some broad general relations of these hormones were demonstrated. On the other hand, evidence for some atypical relations also was found when particular subgroups of children were examined, suggesting complex associations between experience, biology, and striving for competence.

In terms of the broader findings obtained when examining the continuum of resilient functioning, cortisol and maltreatment were found to interact in predicting resilience. For the nonmaltreated comparison children, as cortisol increased, resilience was shown to decrease. In contrast, for the maltreated children, basal levels of cortisol in the morning were generally not found to be related to increasing levels of resilience. Among nonmaltreated children, high basal cortisol may be indicative of children who are undergoing greater stress exposure and as a consequence are constrained in their capacity to adapt competently. This association appears to be attenuated among maltreated children.

Further evidence that different levels of analysis made unique contributions to resilience can be observed through examining the findings of the relation between DHEA and resilience. Regression analyses indicated that maltreatment and both morning and afternoon levels of DHEA were associated with the continuous resilient functioning dimension. Maltreatment status did not interact with morning or afternoon DHEA levels in predicting resilience. Thus, in general, lower DHEA in the morning and afternoon was associated with higher resilience scores for both maltreated and nonmaltreated children.

Finally, neither the morning nor afternoon cortisol/DHEA ratios were related to maltreatment. However, maltreatment and morning cortisol/DHEA levels, and maltreatment and afternoon cortisol/DHEA ratios contributed to prediction of resilient functioning scores, and maltreatment and cortisol/DHEA ratios did not interact. Thus, higher cortisol/DHEA ratios related to higher resilient functioning among both maltreated and nonmaltreated children. Given the associations demonstrated above for DHEA levels, higher ratio scores are obtained with lower DHEA levels, and as such these relations for cortisol/DHEA ratios may not be unique.

DHEA may generate effects that are associated with resilience. For example, DHEA counteracts the harmful effects of hypercortisolism, therein potentially conferring neuroprotection to the brain (Charney, 2004). In addition, several empirical investigations exist that provide suggestive evidence that DHEA may promote resilience. These include negative correlations between the cortisol/DHEA ratio and dissociation, between DHEA and depressive symptomatology, between DHEA and PTSD and the antidepressant effects of stress (Charney, 2004; Haglund, Nestadt, Cooper, Southwick, & Charney, 2007).

In addition to these broad findings for cortisol, DHEA, and the cortisol/DHEA ratio, more person-centered analyses revealed additional more complex relationships. Because adversity and trauma are typically associated with high risk for maladaptation and psychopathology, resilience in the face of extreme stressors runs counter to expectations. There may be evidence of atypical processes operating to promote positive adaptation in some children evincing resilience. In the current investigation, two examples of associations that stood out as unusual were noteworthy. First, in looking beyond overall maltreatment effects, the observation that physically abused children with high morning cortisol had higher resilient functioning than children with lower levels of cortisol was divergent from the more general pattern of higher cortisol being related to lower resilient functioning as was seen in nonmaltreated and sexually abused children.

Prior findings have indicated that physically abused children on the average tend to have lower morning cortisol (Cicchetti & Rogosch, 2001a). It may be that a subgroup of physically abused children who are still able to elevate cortisol to adapt to stressors in their lives are demonstrating greater striving for competence. The enhanced basal cortisol secretion may serve to mobilize and replenish energy stores and contribute to enhanced arousal and focused attention, each of which may be associated with the development of resilience (Charney, 2004). Cortisol exerts regulatory effects on the hippocampus, amygdala, and prefrontal cortex (Gold, Drevets, & Charney, 2002). Investigations of HPA axis regulation in maltreated children have revealed attenuated cortisol secretion (Gunnar & Donzella, 2002). This hyposecretion is akin to hypocortisolism, an indicator of allostatic load. The physically abused children with lower levels of morning cortisol may have developed hypocortisolism over time in response to chronic stress exposure. As a result, there may be a diminished capacity to mobilize the HPA axis to promote positive adaptation under conditions of ongoing stress. In contrast to the physically abused children, the very low level of resilience among sexually abused children with high basal cortisol may be a product of their different traumatic experiences and the consequences of chronic excessive vigilance and preoccupation, with commensurate HPA axis hyperarousal.

The second noteworthy observation was the atypical pattern of a relative diurnal increase in DHEA seen in maltreated children with high resilience. What might promote this variation is unclear. However, the general trend for lower morning DHEA and lower afternoon DHEA to be related to higher resilient functioning, as shown in regression analyses, was not borne out when we looked in more detail at the subgroup of maltreated children with high resilience. In fact, the maltreated children with high resilience were unique compared to other groups of maltreated and nonmaltreated children in terms of the greater likelihood (71.4%) of showing an increase in DHEA from morning to afternoon. It may be that maltreated children who have the capacity to increase DHEA over the course of the day are better equipped to meet the demands of high chronic exposure to stress in their lives and to adapt competently. The nonmaltreated children with high resilience did not demonstrate this pattern of diurnal DHEA increase. Rather, the high resilient nonmaltreated children exhibited to lowest levels of DHEA across the day. It may be that a more typical pattern of DHEA diurnal regulation is sufficient to address the less toxic experiences in their lives and allow their successful adaptation to unfold.

What is unclear is the extent to which there is a contribution of innate individual differences in stress hormone levels that promotes alterations in adapting to stress versus the extent to which stress exposure differentially interacts with individuals’ propensity for specific hormone levels and affects adaptation. Studying HPA regulation over the course of development is needed to differentiate the influences of innate individual differences in HPA regulation, the manner in which individuals with innate differences in HPA regulation respond differentially to stress, and the developmental course of HPA regulation under stress, particularly the potential conversion from early hypercortisolism to later hypocortisolism among individuals chronically exposed to stress. . Research has documented the role of stress hormones in the expression of genes that govern brain function (Watson & Gametchu, 1999). For example, glucocorticoid receptors in the nuclei of neurons are responsible for genomic effects, such that stress hormones influence gene expression, and affect neuronal growth, neurotransmitter synthesis, receptor density and sensitivity, and neurotransmitter reuptake. Because stress hormones can have such direct effects on genes that control brain structure and function (Watson & Gametchu, 1999), research on stress in developing organisms is crucially important. Therefore, understanding the transactions between the regulation of adrenal steroid hormones and brain processes over the course of development, particularly under conditions of stress, will be invaluable for providing insight into resilience promoting processes.

In summary, different levels of analysis contributed independently to resilience --personality, cortisol x maltreatment interaction, and DHEA each demonstrated independent effects is explaining variation in resilient functioning in maltreated and nonmaltreated children. Moreover, evidence also was found for atypical relations between adrenal steroid hormone regulation and resilience for some subgroups of maltreated children. Despite these findings, even though ego-resiliency and ego (over)control, DHEA, and the cortisol by maltreatment interaction each made unique contributions to resilience, there were no personality x hormone interactions in the prediction of resilience.

In the future, additional research on the pathways to resilience in maltreated and nonmaltreated children should be conducted within a multiple-levels-of-analysis approach (Cicchetti & Blender, 2004). Such investigations should be longitudinal in nature and incorporate a broader array of levels of analysis, both biological and psychological. The incorporation of such a perspective into the study of resilience will result in a more sophisticated and comprehensive portrayal of this phenomenon that will serve not only to advance the scientific knowledge of resilience, but also to inform efforts to translate research on positive adaptation in the face of adversity into the development of interventions to promote resilient functioning.