Brain structures in pediatric maltreatment-related posttraumatic stress disorder: a sociodemographically matched study

Abstract

Background

Previous investigations suggest that maltreated children evidence alterations of chemical mediators of stress and adverse brain development. Previous anatomical magnetic resonance imaging (MRI) brain studies have not controlled for socioeconomic status.

Methods

In this study, 28 psychotropic naïve children and adolescents with maltreatment-related posttraumatic stress disorder (PTSD) and 66 sociodemographically similar healthy control subjects underwent comprehensive clinical assessments and anatomical MRI brain scans.

Results

Compared with control subjects, subjects with PTSD had smaller intracranial, cerebral, and prefrontal cortex, prefrontal cortical white matter, and right temporal lobe volumes and areas of the corpus callosum and its subregions (2, 4, 5, 6, and 7), and larger frontal lobe cerebrospinal fluid (CSF) volumes than control subjects. The total midsagittal area of corpus callosum and middle and posterior regions remained smaller in subjects with PTSD, whereas right, left, and total lateral ventricles and frontal lobe CSF were proportionally larger than in control subjects, after adjustment for cerebral volume. Brain volumes positively correlated with age of onset of PTSD trauma and negatively correlated with duration of abuse. Significant gender × group effect demonstrated greater lateral ventricular volume increases in maltreated male subjects with PTSD than maltreated female subjects with PTSD. No hippocampal differences were seen.

Conclusions

These data provide further evidence to suggest that maltreatment-related PTSD is associated with adverse brain development. These data also suggest that male children may be more vulnerable to these effects.

Introduction

Child maltreatment, defined as neglect, physical abuse, sexual abuse, and emotional maltreatment, is a common contributor to child and adult mental illness in this country (Felitti et al 1998). Posttraumatic stress disorder (PTSD) is commonly seen in maltreated children, especially during the period immediately following maltreatment disclosure Famularo et al 1993, Famularo et al 1996, McLeer et al 1998. Furthermore, partial PTSD responses are commonly seen in victims of childhood maltreatment Armsworth and Holaday 1993, De Bellis et al 1999a, Hillary and Schare 1993, Mannarino et al 1994, Wolfe and Charney 1991, Wolfe et al 1994. These partial symptoms may also contribute to substantial functional impairment and distress (Carrion et al 2001b).

Although limited, the psychobiological data in maltreated children suggest that maltreated children and adolescents with mood and anxiety symptoms (i.e., PTSD symptoms) show evidence of altered catecholamines and hypothalamic–pituitary–adrenal (HPA) axis activity. These include findings of greater 24-hour urinary norepinephrine concentrations in neglected depressed male subjects (Queiroz et al 1991) and greater 24-hour urinary catecholamine and catecholamine metabolite concentrations in dysthymic, sexually abused girls (De Bellis et al 1994b). Results from pediatric studies suggest that maltreated children show evidence of corticotrophin-releasing hormone or factor hypersecretion. These include findings of hypersecretion of morning cortisol in sexually abused girls (Putnam et al 1991) and dysregulation of the HPA axis in depressed, maltreated children Hart et al 1996, Kaufman et al 1997b and dysthymic, sexually abused girls (De Bellis et al 1994a). Children and adolescents with maltreatment-related PTSD show evidence of increased catecholamine and cortisol activity. A recent study identified significantly greater concentrations of urinary dopamine and norepinephrine concentrations over 24 hours in children with maltreatment-related PTSD than non-maltreated children with overanxious disorder and control subjects, as well as greater concentrations of 24-hour urinary free cortisol in children with maltreatment-related PTSD compared to control subjects (De Bellis et al 1999a). Other investigators reported higher levels of salivary cortisol throughout the day in children with maltreatment-related PTSD or subthreshold PTSD (Carrion et al 2002) and findings of decreased platelet adrenergic receptors and increased heart rate following orthostatic challenge in physically and sexually abused children with PTSD compared with non-maltreated subjects (Perry 1994).

In the developing brain, elevated levels of catecholamines and cortisol may lead to adverse brain development through the mechanisms of accelerated loss (or metabolism) of neurons Edwards et al 1990, Sapolsky et al 1990, Simantov et al 1996, Smythies 1997, delays in myelination (Dunlop et al 1997), abnormalities in developmentally appropriate pruning Lauder 1988, Todd 1992, and/or the inhibition of neurogenesis Gould et al 1997, Gould et al 1998, Tanapat et al 1998. Furthermore, stress decreases brain-derived neurotrophic factor expression (Smith et al 1995). Thus, the overwhelming stress of child maltreatment experiences may have adverse influences on a child’s brain maturation. Until recently, investigators have generally studied childhood brain function with psychoeducational instruments (i.e., intelligence and achievement tests). The results of these studies suggest that maltreated children demonstrate a variety of intellectual and academic impairments, including lower intelligence quotient (IQ) Augoustinos 1987, Carrey et al 1995, Money et al 1983, Perez and Widom 1994, Pianta et al 1989, Trickett et al 1994; however, studies applying neuropsychological methods suggest that children and adolescents with PTSD show deficits in executive functioning and attention (Beers and De Bellis 2002) and everyday memory (Moradi et al 1999).

More recently, magnetic resonance imaging (MRI) has provided a safe and novel approach to measuring brain maturation in healthy and psychologically traumatized living children. To date, only two studies involving maltreated children have been reported. The results of these studies suggest that pediatric maltreatment-related PTSD is associated with adverse brain development. These include findings of smaller intracranial and cerebral volumes, and smaller total midsagittal area of corpus callosum and middle and posterior regions and larger lateral ventricular volumes than non-maltreated control subjects (De Bellis et al 1999b) and smaller brain and cerebral volumes and attenuation of frontal lobe asymmetry in children with maltreatment-related PTSD or subthreshold PTSD compared with archival control subjects (Carrion et al 2001a). In these cross-sectional studies, however, causal relationships between maltreatment, psychiatric symptoms, and brain changes cannot be ascertained. Studies in developmental traumatology are inherently complicated, because it is difficult to separate out the effects of heterogeneous sources of maltreatment from other confounding factors that are commonly present in maltreating families. These include low socioeconomic status (SES), substance abuse, low educational levels, poor parenting skills, and legal and social service entanglements. For example, these previous MRI studies did not control for socioeconomic status, which may also influence brain maturation through ecological variables. An important mission for the field of developmental traumatology research is to unravel these complex interactions. Thus in this brain maturation study, we recruited an independent and unique sample of children and adolescents with maltreatment-related PTSD for brain maturation studies who were medically healthy, psychotropic naïve, free of significant prenatal substance exposure and adolescent-onset substance abuse or dependence, and compared them to a relatively large group of sociodemographically matched, healthy, non-maltreated control subjects. It was hypothesized that maltreated children with PTSD would show decreases in volumes of structures that may be vulnerable to stress during developmental processes, such as the cerebral, frontal, and temporal cortex, amygdala and hippocampus, and corpus callosum. The basal ganglia were also measured as comparison structures. It was further hypothesized that PTSD symptoms and trauma characteristics would significantly correlate with anatomical brain measurements.