Child Abuse & Neglect 38 (2014) 1778–1786
Contents lists available at ScienceDirect
Child Abuse & Neglect
Disparate plasma cortisol concentrations in sexually abused female children from Johannesburg, South Africa夽 Denise Muller a,∗ , Sheri-lee Errington b , Christopher P. Szabo c , Neville Pitts a , Lorna Jacklin d a
School of Physiology, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa Teddy Bear Clinic, Johannesburg, South Africa School of Clinical Medicine, Department of Psychiatry, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa d School of Clinical Medicine, Department of Pediatrics, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa b c
a r t i c l e
i n f o
Article history: Received 31 March 2014 Received in revised form 17 July 2014 Accepted 30 July 2014 Available online 29 August 2014 Keywords: Sexual abuse Stress reactivity Female children Girls HPA axis Cortisol
a b s t r a c t A growing body of research indicates that a bidirectional response to a stressor may occur in maltreated children and may be associated with later life psychopathology. However, few studies have investigated stress reactivity in children when they first present to a sexual abuse clinic. Thus, in order to evaluate whether HPA axis dysregulation would be evident at first presentation to a sexual abuse clinic in young girls (n = 26), between the ages of 6–12 years old, blood samples were obtained immediately following examination at a forensic sexual abuse clinic and from the matched control group of children (n = 14; 10.1 ± 0.8) immediately following a bone density scan. Stratification of the sexually abused group into those children who were reportedly abused by a stranger and had no other family stressors (n = 15, 10.4 ± 1.8) and those children whose parents reported abuse of the child by a stranger and other family stressors (n = 11; 9.5 ± 1.8) revealed differences in stress reactivity. Plasma concentrations, of the children from the forensic clinic, were significantly increased in children who reported abuse by a stranger only (322.3 ± 117.4 nmol/l) and significantly decreased in children whose histories indicated sexual abuse by a stranger and other family stressors (149.6 ± 39.7 nmol/l) when compared to the control group (225.5 ± 47.5 nmol/l). In conclusion, following sexual abuse and a secondary stressor, the forensic examination, there is evidence of divergent cortisol responses in the stratified clinical group of children. © 2014 Elsevier Ltd. All rights reserved.
Children with significant stressors in their family of origin are more likely to present with behavioral and emotional problems associated with a dysregulated stress reactivity pattern (Loman & Gunnar, 2010; Trickett, Gordis, Peckins & Susman, 2014). Moreover, increasing evidence of a dysregulated hypothalamic-pituitary-adrenal (HPA) axis has been noted in maltreated children (Trickett et al., 2014) and in adult patients who retrospectively report early life stress (Badanes, Watamura,
夽 This work was supported by Iris Ellen Hodges Trust – Stress and Emotional Problems (Jack007). Grant to Dr Lorna Jacklin. ∗ Corresponding author. http://dx.doi.org/10.1016/j.chiabu.2014.07.014 0145-2134/© 2014 Elsevier Ltd. All rights reserved.
D. Muller et al. / Child Abuse & Neglect 38 (2014) 1778–1786
1779
& Hankin, 2011; Carpenter et al., 2007; Danese & McEwen, 2012) and may lead to the development of allostatic load (Katz, Sprang & Cooke, 2012; McEwen, 1998, 2006; McEwen & Seeman, 1999). Allostasis has been defined as the return to normal baseline levels of cortisol, a glucocorticoid and the final hormone in the HPA axis cascade, following a response to a stressor (McEwen, 2006). However, stress that occurs too early in life has been associated with dysregulation of the HPA response (Johnson et al., 2002; Trickett et al., 2014) and the consequent long-term dysregulation in reactivity to future stressors (Burghy et al., 2012; Wachs, Georgieff, Cusick, & McEwen, 2014). Allostatic load needs to be understood in terms of the initial physiological demand which primes the stress response for maladaptive responses to later stressors resulting in an imbalance on the physiology of an individual (Katz et al., 2012). McEwen (2006) documents four maladaptive responses that may persist beyond the original stressful event: (a) frequent increases in high levels of glucocorticoids because of continuing stressors in the individuals life, (b) frequent increases in high levels of glucocorticoids in the absence of any real stressful event, (c) prolonged increases of glucocorticoids after a stressful event, and (d) an inadequate response in the event of a new stressor. Prolonged increases in plasma cortisol concentrations are associated with neurodegeneration in the hypothalamus, the hippocampus and neurons in the prefrontal cortex (McEwen, 2006; Wachs et al., 2014). However, not all children subjected to early life stress present with increased cortisol concentrations following later life stressors (Ayer et al., 2013; Heim, Ehlert, & Hellhammer, 2000; Trickett et al., 2014). The possibility exists that a repeat stressor in vulnerable individuals results in different physiological responses than in non-vulnerable individuals (Goodman, New, & Siever, 2004; Heim et al., 2000). Thus, it would also appear that there may be a dichotomous response to stress that would categorize some patients as either high or low stress responders as measured by circulating cortisol in response to a stressor (Fries, Hesse, Hellhammer, & Hellhammer, 2005; Goldman-Mellor, Hamer & Steptoe, 2012; Heim et al., 2000; Lovallo, Farag, Sorocco, Cohoon, & Vincent, 2012; Pervanidou, 2008). A burgeoning body of evidence suggests that attenuated cortisol responses to a severe stressor are associated with multiple early life stressors (Goldman-Mellor et al., 2012; Heim et al., 2000; Miller et al., 2013; Raison & Miller, 2003; Trickett et al., 2014). Moreover, growing research indicates that adult clinical patients who present with a variety of psychopathologies frequently report retrospective associations with a sexual abuse stressor in their formative years (Burghy et al., 2012; Cohen, 2008; Heim, Plotsky, & Nemeroff, 2004). Stress associated with childhood sexual abuse is considered to be a significant risk factor for the development of long term psychiatric disorders such as Post-Traumatic Stress Disorder (PTSD; Chen et al., 2010; Mathews, Abrahams & Jewkes, 2013; Runyon, Deblinger, & Steer, 2014; Walsh et al., 2013), anxiety (Chen et al., 2010; Mathews et al., 2013), depression (Chen et al., 2010) and Borderline Personality Disorder (BPD; Kahl et al., 2006; Rinne et al., 2002, 2003; Zanarini et al., 1997). Therefore, the growing understanding of the physiological load associated with early life trauma, the lasting impact on the hypothalamic-pituitary-adrenal (HPA) axis and the resultant impact on cortisol reactivity to a stressor is changing how stress responses in young children are viewed (Heim, Newport, Mletzko, Miller, & Nemeroff, 2008; Heim et al., 2004). However, as Trickett et al. (2014) note, few studies have evaluated cortisol reactivity in abused and maltreated children. In light of the possible long term implications to health it is critical to determine if there is evidence for dysregulated responses to stress as early as possible following severe early life stress (Wachs et al., 2014). Kapur, Phillips, and Insel (2012) note that there is also a need to identify and stratify at risk children into relevant sub-groups when they first present with a trauma. Currently, however, we cannot identify, using an objective measure such as physiological biomarkers, which children might develop lasting complications from early life stress. A small study on cortisol responses to the stress of a forensic examination at a sexual abuse clinic (Muller, Errington, Szabo, Pitts, & Jacklin, 2014) showed a significant difference in plasma cortisol concentrations in young girls (n = 11) who lived at home compared to children from a residential care facility. This proof of concept study, however, lacked a matched, non-sexually abused control group. Thus, the hypothesis for the current research was that there would be a significant difference in cortisol concentrations in the children from the forensic clinic when compared to a control group of matched children. The current study had two specific objectives: (a) identify whether a differential pattern of cortisol response occurs in young girls aged between 6 and 12 years of age whose histories include other stressors in the family home and (b) identify cortisol responses in a matched control group of children under a novel condition. Methods Ethics Approval Clinical Sample. Adhering to the principles of the Declaration of Helsinki, ethics approval was obtained from the University of the Witwatersrand’s Human Research Ethics Committee (M070724 and M060250) for the study to take place at the Teddy Bear Clinic, Charlotte Maxeke Hospital in Johannesburg, under the supervision of a pediatrician. The Teddy Bear Clinic, a forensic abuse clinic, identifies the abuse a child has suffered and prepares documentation for court proceedings. Permission for inclusion in the study was granted by either parents or guardians from care facilities. Prior to the forensic examination the pediatrician informed all children about all the tests that would be conducted including blood tests to evaluate stress responses. Control Group. Children enrolled in a bone density study (Meiring, Avidon, Norris, & McVeigh, 2013) were recruited as a control group. Ethics approval was obtained from the University of the Witwatersrand’s Human Research Ethics Committee
1780
D. Muller et al. / Child Abuse & Neglect 38 (2014) 1778–1786
(M10635). Permission for inclusion in our study was granted by either parents or guardians. The children were informed about the stress study and signed assent forms for inclusion in the research project. Procedure Clinical Sample. Twenty-six female children between the ages of 6 and 12 years, primarily of African descent and from low income families, were recruited into this study. These children presented on different mornings to a general hospital forensic child sexual abuse clinic, The Teddy Bear Clinic, at the Charlotte Maxeke Hospitial in Johannesburg. The forensic examination involved a general interview with a qualified intake worker, a forensic medical examination by the pediatrician with specific focus on the genital area to ascertain whether macroscopic evidence of sexual abuse is present, and a blood sample (by venesection). Exclusion criteria for this research included infections, fevers, or any other illnesses. Importantly, as the children are being assessed for a possible court appearance, only the forensic assessment is done at the initial interview. The assessment provides information of the severity of the abuse, physical characteristics of the abuse, developmental level of the child, veracity of disclosure, and the capacity to testify. As the forensic examination is an established protocol for the criminal court proceedings there are very specific guidelines which currently do not include psychological assessments for PTSD, depression, or anxiety. Thus, in light of possible court appearances, no psychometric testing or psychological assessments are conducted at this assessment. Furthermore, in order to exclude sexually transmitted diseases investigative blood samples are routinely drawn from the children at the Teddy Bear Clinic. An additional 5 ml blood was drawn into a heparinized vacutainer tube (BD-Plymouth) by the pediatrician to enable cortisol concentrations to be determined. All blood samples were taken in the morning between 11:00 a.m. and 12:30 p.m., immediately following the interview by the pediatrician and a physical examination. Because of the stressful nature of the forensic examination, an increase in plasma cortisol concentration was to be expected (McEwen, 2006). Control Group. The children (n = 14) from the control group were recruited from Afrika Tikkun, a community center in Diepsloot, Johannesburg. Diepsloot is a low-income housing area in the north of Johannesburg (Meiring et al., 2013). The children were matched to the experimental group based on age, ethnicity, and socio-economic status. Only children from intact families who were known to have not experienced any recent trauma and were considered to have no behavioral problems were recruited into the control group. Within 15 min of their Dual energy X-ray absorptiometry scan (DXA; Meiring et al., 2013) children were escorted to an examination room for blood collection. However, this examination area was designed to reduce stress by allowing them to associate with their friends and the promise of refreshments and watching a film. Although some children were visibly stressed by the procedure, their friends were encouraged to stay with them and hold their hand. The blood sample was immediately transferred to the laboratory for centrifugation at 1000 × g for 10 min at 4 ◦ C. Plasma was removed, aliquoted and frozen at −70 ◦ C until assayed. Measures Cortisol Assay. Cortisol concentrations were determined using 125 I-labled cortisol solid-phase radiommunoassay (RIA), Coat-A-Count, Human Cortisol (Siemens Medical Solutions Diagnostics, Los Angeles, USA). Detection limits were 0.2 g/dl (5.5 nmol/l) with within-run sample coefficient of variation of
Contents lists available at ScienceDirect
Child Abuse & Neglect
Disparate plasma cortisol concentrations in sexually abused female children from Johannesburg, South Africa夽 Denise Muller a,∗ , Sheri-lee Errington b , Christopher P. Szabo c , Neville Pitts a , Lorna Jacklin d a
School of Physiology, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa Teddy Bear Clinic, Johannesburg, South Africa School of Clinical Medicine, Department of Psychiatry, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa d School of Clinical Medicine, Department of Pediatrics, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa b c
a r t i c l e
i n f o
Article history: Received 31 March 2014 Received in revised form 17 July 2014 Accepted 30 July 2014 Available online 29 August 2014 Keywords: Sexual abuse Stress reactivity Female children Girls HPA axis Cortisol
a b s t r a c t A growing body of research indicates that a bidirectional response to a stressor may occur in maltreated children and may be associated with later life psychopathology. However, few studies have investigated stress reactivity in children when they first present to a sexual abuse clinic. Thus, in order to evaluate whether HPA axis dysregulation would be evident at first presentation to a sexual abuse clinic in young girls (n = 26), between the ages of 6–12 years old, blood samples were obtained immediately following examination at a forensic sexual abuse clinic and from the matched control group of children (n = 14; 10.1 ± 0.8) immediately following a bone density scan. Stratification of the sexually abused group into those children who were reportedly abused by a stranger and had no other family stressors (n = 15, 10.4 ± 1.8) and those children whose parents reported abuse of the child by a stranger and other family stressors (n = 11; 9.5 ± 1.8) revealed differences in stress reactivity. Plasma concentrations, of the children from the forensic clinic, were significantly increased in children who reported abuse by a stranger only (322.3 ± 117.4 nmol/l) and significantly decreased in children whose histories indicated sexual abuse by a stranger and other family stressors (149.6 ± 39.7 nmol/l) when compared to the control group (225.5 ± 47.5 nmol/l). In conclusion, following sexual abuse and a secondary stressor, the forensic examination, there is evidence of divergent cortisol responses in the stratified clinical group of children. © 2014 Elsevier Ltd. All rights reserved.
Children with significant stressors in their family of origin are more likely to present with behavioral and emotional problems associated with a dysregulated stress reactivity pattern (Loman & Gunnar, 2010; Trickett, Gordis, Peckins & Susman, 2014). Moreover, increasing evidence of a dysregulated hypothalamic-pituitary-adrenal (HPA) axis has been noted in maltreated children (Trickett et al., 2014) and in adult patients who retrospectively report early life stress (Badanes, Watamura,
夽 This work was supported by Iris Ellen Hodges Trust – Stress and Emotional Problems (Jack007). Grant to Dr Lorna Jacklin. ∗ Corresponding author. http://dx.doi.org/10.1016/j.chiabu.2014.07.014 0145-2134/© 2014 Elsevier Ltd. All rights reserved.
D. Muller et al. / Child Abuse & Neglect 38 (2014) 1778–1786
1779
& Hankin, 2011; Carpenter et al., 2007; Danese & McEwen, 2012) and may lead to the development of allostatic load (Katz, Sprang & Cooke, 2012; McEwen, 1998, 2006; McEwen & Seeman, 1999). Allostasis has been defined as the return to normal baseline levels of cortisol, a glucocorticoid and the final hormone in the HPA axis cascade, following a response to a stressor (McEwen, 2006). However, stress that occurs too early in life has been associated with dysregulation of the HPA response (Johnson et al., 2002; Trickett et al., 2014) and the consequent long-term dysregulation in reactivity to future stressors (Burghy et al., 2012; Wachs, Georgieff, Cusick, & McEwen, 2014). Allostatic load needs to be understood in terms of the initial physiological demand which primes the stress response for maladaptive responses to later stressors resulting in an imbalance on the physiology of an individual (Katz et al., 2012). McEwen (2006) documents four maladaptive responses that may persist beyond the original stressful event: (a) frequent increases in high levels of glucocorticoids because of continuing stressors in the individuals life, (b) frequent increases in high levels of glucocorticoids in the absence of any real stressful event, (c) prolonged increases of glucocorticoids after a stressful event, and (d) an inadequate response in the event of a new stressor. Prolonged increases in plasma cortisol concentrations are associated with neurodegeneration in the hypothalamus, the hippocampus and neurons in the prefrontal cortex (McEwen, 2006; Wachs et al., 2014). However, not all children subjected to early life stress present with increased cortisol concentrations following later life stressors (Ayer et al., 2013; Heim, Ehlert, & Hellhammer, 2000; Trickett et al., 2014). The possibility exists that a repeat stressor in vulnerable individuals results in different physiological responses than in non-vulnerable individuals (Goodman, New, & Siever, 2004; Heim et al., 2000). Thus, it would also appear that there may be a dichotomous response to stress that would categorize some patients as either high or low stress responders as measured by circulating cortisol in response to a stressor (Fries, Hesse, Hellhammer, & Hellhammer, 2005; Goldman-Mellor, Hamer & Steptoe, 2012; Heim et al., 2000; Lovallo, Farag, Sorocco, Cohoon, & Vincent, 2012; Pervanidou, 2008). A burgeoning body of evidence suggests that attenuated cortisol responses to a severe stressor are associated with multiple early life stressors (Goldman-Mellor et al., 2012; Heim et al., 2000; Miller et al., 2013; Raison & Miller, 2003; Trickett et al., 2014). Moreover, growing research indicates that adult clinical patients who present with a variety of psychopathologies frequently report retrospective associations with a sexual abuse stressor in their formative years (Burghy et al., 2012; Cohen, 2008; Heim, Plotsky, & Nemeroff, 2004). Stress associated with childhood sexual abuse is considered to be a significant risk factor for the development of long term psychiatric disorders such as Post-Traumatic Stress Disorder (PTSD; Chen et al., 2010; Mathews, Abrahams & Jewkes, 2013; Runyon, Deblinger, & Steer, 2014; Walsh et al., 2013), anxiety (Chen et al., 2010; Mathews et al., 2013), depression (Chen et al., 2010) and Borderline Personality Disorder (BPD; Kahl et al., 2006; Rinne et al., 2002, 2003; Zanarini et al., 1997). Therefore, the growing understanding of the physiological load associated with early life trauma, the lasting impact on the hypothalamic-pituitary-adrenal (HPA) axis and the resultant impact on cortisol reactivity to a stressor is changing how stress responses in young children are viewed (Heim, Newport, Mletzko, Miller, & Nemeroff, 2008; Heim et al., 2004). However, as Trickett et al. (2014) note, few studies have evaluated cortisol reactivity in abused and maltreated children. In light of the possible long term implications to health it is critical to determine if there is evidence for dysregulated responses to stress as early as possible following severe early life stress (Wachs et al., 2014). Kapur, Phillips, and Insel (2012) note that there is also a need to identify and stratify at risk children into relevant sub-groups when they first present with a trauma. Currently, however, we cannot identify, using an objective measure such as physiological biomarkers, which children might develop lasting complications from early life stress. A small study on cortisol responses to the stress of a forensic examination at a sexual abuse clinic (Muller, Errington, Szabo, Pitts, & Jacklin, 2014) showed a significant difference in plasma cortisol concentrations in young girls (n = 11) who lived at home compared to children from a residential care facility. This proof of concept study, however, lacked a matched, non-sexually abused control group. Thus, the hypothesis for the current research was that there would be a significant difference in cortisol concentrations in the children from the forensic clinic when compared to a control group of matched children. The current study had two specific objectives: (a) identify whether a differential pattern of cortisol response occurs in young girls aged between 6 and 12 years of age whose histories include other stressors in the family home and (b) identify cortisol responses in a matched control group of children under a novel condition. Methods Ethics Approval Clinical Sample. Adhering to the principles of the Declaration of Helsinki, ethics approval was obtained from the University of the Witwatersrand’s Human Research Ethics Committee (M070724 and M060250) for the study to take place at the Teddy Bear Clinic, Charlotte Maxeke Hospital in Johannesburg, under the supervision of a pediatrician. The Teddy Bear Clinic, a forensic abuse clinic, identifies the abuse a child has suffered and prepares documentation for court proceedings. Permission for inclusion in the study was granted by either parents or guardians from care facilities. Prior to the forensic examination the pediatrician informed all children about all the tests that would be conducted including blood tests to evaluate stress responses. Control Group. Children enrolled in a bone density study (Meiring, Avidon, Norris, & McVeigh, 2013) were recruited as a control group. Ethics approval was obtained from the University of the Witwatersrand’s Human Research Ethics Committee
1780
D. Muller et al. / Child Abuse & Neglect 38 (2014) 1778–1786
(M10635). Permission for inclusion in our study was granted by either parents or guardians. The children were informed about the stress study and signed assent forms for inclusion in the research project. Procedure Clinical Sample. Twenty-six female children between the ages of 6 and 12 years, primarily of African descent and from low income families, were recruited into this study. These children presented on different mornings to a general hospital forensic child sexual abuse clinic, The Teddy Bear Clinic, at the Charlotte Maxeke Hospitial in Johannesburg. The forensic examination involved a general interview with a qualified intake worker, a forensic medical examination by the pediatrician with specific focus on the genital area to ascertain whether macroscopic evidence of sexual abuse is present, and a blood sample (by venesection). Exclusion criteria for this research included infections, fevers, or any other illnesses. Importantly, as the children are being assessed for a possible court appearance, only the forensic assessment is done at the initial interview. The assessment provides information of the severity of the abuse, physical characteristics of the abuse, developmental level of the child, veracity of disclosure, and the capacity to testify. As the forensic examination is an established protocol for the criminal court proceedings there are very specific guidelines which currently do not include psychological assessments for PTSD, depression, or anxiety. Thus, in light of possible court appearances, no psychometric testing or psychological assessments are conducted at this assessment. Furthermore, in order to exclude sexually transmitted diseases investigative blood samples are routinely drawn from the children at the Teddy Bear Clinic. An additional 5 ml blood was drawn into a heparinized vacutainer tube (BD-Plymouth) by the pediatrician to enable cortisol concentrations to be determined. All blood samples were taken in the morning between 11:00 a.m. and 12:30 p.m., immediately following the interview by the pediatrician and a physical examination. Because of the stressful nature of the forensic examination, an increase in plasma cortisol concentration was to be expected (McEwen, 2006). Control Group. The children (n = 14) from the control group were recruited from Afrika Tikkun, a community center in Diepsloot, Johannesburg. Diepsloot is a low-income housing area in the north of Johannesburg (Meiring et al., 2013). The children were matched to the experimental group based on age, ethnicity, and socio-economic status. Only children from intact families who were known to have not experienced any recent trauma and were considered to have no behavioral problems were recruited into the control group. Within 15 min of their Dual energy X-ray absorptiometry scan (DXA; Meiring et al., 2013) children were escorted to an examination room for blood collection. However, this examination area was designed to reduce stress by allowing them to associate with their friends and the promise of refreshments and watching a film. Although some children were visibly stressed by the procedure, their friends were encouraged to stay with them and hold their hand. The blood sample was immediately transferred to the laboratory for centrifugation at 1000 × g for 10 min at 4 ◦ C. Plasma was removed, aliquoted and frozen at −70 ◦ C until assayed. Measures Cortisol Assay. Cortisol concentrations were determined using 125 I-labled cortisol solid-phase radiommunoassay (RIA), Coat-A-Count, Human Cortisol (Siemens Medical Solutions Diagnostics, Los Angeles, USA). Detection limits were 0.2 g/dl (5.5 nmol/l) with within-run sample coefficient of variation of
