Metab Brain Dis (2014) 29:341–349 DOI 10.1007/s11011-014-9500-0
ORIGINAL PAPER
Structural brain changes in prenatal methamphetamine-exposed children Annerine Roos & Gaby Jones & Fleur M. Howells & Dan J. Stein & Kirsten A. Donald
Received: 15 October 2013 / Accepted: 28 January 2014 / Published online: 20 February 2014 # Springer Science+Business Media New York 2014
Abstract The global use of methamphetamine (MA) has increased substantially in recent years, but the effect of MA on brain structure in prenatally exposed children is understudied. Here we aimed to investigate potential changes in brain volumes and cortical thickness of children with prenatal MA-exposure compared to unexposed controls. Eighteen 6-year old children with MA-exposure during pregnancy and 18 healthy controls matched for age, gender and socioeconomic background underwent structural imaging. Brain volumes and cortical thickness were assessed using Freesurfer and compared using ANOVA. Left putamen volume was significantly increased, and reduced cortical thickness was observed in the left hemisphere of the inferior parietal, parsopercularis and precuneus areas of MA-exposed children compared to controls. Compared to control males, prenatal MA-exposed males had greater volumes in striatal and associated areas, whereas MA-exposed females predominantly had greater cortical thickness compared to control females. In utero exposure to MA results in changes in the striatum of the developing child. In addition, changes within the striatal, frontal, and parietal areas are in part gender dependent.
A. Roos (*) : G. Jones : D. J. Stein MRC Unit on Anxiety & Stress Disorders, Department of Psychiatry, Stellenbosch University, P.O. Box 19063, Tygerberg, 7505 Cape Town, South Africa e-mail:
[email protected] F. M. Howells Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa K. A. Donald Division of Developmental Paediatrics, University of Cape Town, Cape Town, South Africa
Keywords Methamphetamine . Brain structure . Prenatal . Dopamine
Introduction The global use of methamphetamine (MA) has increased substantially in recent years, including in pregnant women. In the United States an increase of 8 % in 1994 to 24 % in 2006 was reported in pregnant women (Terplan et al. 2009). Another study conducted in South Africa found that during 2006, of 58 % of daily MA users, women mainly of mixed race made up a quarter (24 %–29 %), and more than 90 % of female subjects were women of childbearing age (Pluddemann et al. 2008). In a recent South African study (of a local mixed race population in the Western Cape), 238 out of 356 non-pregnant women in their 20’s (66 %) used MA, whereas 24 out of 26 pregnant women (92 %) used MA (Jones et al. 2011). It is thus likely that a significant number of women will use MA during pregnancy, with resultant risks for both mother and child. MA has specific effects on neurotransmitter systems, increasing the release and blocking the reuptake of dopamine in dopamine-rich striatal areas of the brain. Animal studies have shown that in MA-exposed pregnant rodents, MA crosses the placenta and has global effects on the developing fetus and in particular the central nervous system (Wouldes et al. 2004). Pregnant mice exposed to MA, showed a two to three times increase in MA levels of the fetal striatum after MA was subcutaneously injected into the female (Heller et al. 2001; Won et al. 2001). In humans, MA also crosses the placenta as well as the blood–brain barrier to exert effects on the fetal brain (Won et al. 2001). The blood–brain barrier of the human fetus is more permeable than that of children and adults, and the fetus is unable to efficiently metabolise and detoxify drugs (Rozman and Klaassen 1996). This may potentiate damaging
342
effects of MA to the human brain, particularly in the striatum, during fetal development. MA also appears to have specific effects on growth and neurodevelopment. A prospective study conducted in the US found that prenatal MA-exposed infants were more likely to be smaller for gestational age than were controls, even after adjusting for demographic and clinical information and substance use (Nguyen et al. 2010). These authors also demonstrated that MA-exposure in utero was associated with central nervous system stress, in addition to poor tone, decreased arousal, and poor quality of movement in neonates. Limited evidence in children suggests that there may be a number of neuroanatomical sequelae of prenatal exposure to MA. A small number of studies including that of Chang et al. (2004) and Sowell et al. (2010) who studied children in the age range of 3–16 years found volumetric changes mainly in dopamine-rich striatal areas, including frontal and parietal areas. Colby et al. (2012) also found alterations in white matter microstructure within these areas in prenatal MA- and alcohol-exposed children aged 10. However, there is presently no data on the effects of in utero MA-exposure on cortical thickness. The aim of this study was to investigate potential changes in brain volumes and cortical thickness of children with prenatal MA-exposure compared to unexposed controls. Our hypotheses are that there will be changes in brain volumes and cortical thickness particularly of striatal and associated frontal and parietal areas.
Methods Subjects Children residing in the Cape Metropole were included in the study. Children with known prenatal MA-exposure and unexposed children were matched for age, sex, socioeconomic profile, birth circumstances, gestation and schooling. Children were excluded from the study if there were; fetal anomalies; history of epilepsy, diabetes, or head injury; or premature birth (less than 36 weeks). Although the aim of this study was to include only prenatal MA-exposed children, in a small minority the mother used alcohol (n=3). These mothers/caregivers stated that no other drugs (e.g. cocaine, heroin) were used besides MA during pregnancy. None of the mothers of the controls used alcohol or illicit drugs. A detailed demographic, socio-economic and medical history of the child was taken during the brain imaging session (at the time of birth and current), as well as of the mother and family (including the use of other recreational drugs and alcohol). Anthropometrics were also determined including
Metab Brain Dis (2014) 29:341–349
weight, length and head circumference of the child. The child was prepared for imaging in a mock scanner emulating the MRI scanner in terms of form and sequence noise. The child chose an animated movie that they viewed during the scan. Parental/legal guardian and child consent were given before inclusion into the study. The attending caregiver was informed that participation was voluntary and that they were allowed to withdraw from the study at any time without any consequences to them, either social or medical. The study was approved by the Health Research Committee of Stellenbosch University and University of Cape Town and was conducted according to the ethical guidelines of the international Declaration of Helsinki 2008.
Brain imaging and data analyses Brain images were acquired on a Siemens Allegra 3 T MRI scanner. A high resolution motion-navigated T1 multiecho MPRAGE structural scan (Van der Kouwe et al. 2008) was acquired with the following parameters: repetition time of 2,530 ms; 4 echo times of 1.5 ms, 3.2 ms, 4.8 ms and 6.5 ms; flip angle of 7°; matrix size of 224×224×144; field of view of 224 mm; voxel size of 1.3×1.0×1.0 mm and acquisition time of 5 min 20 s. The sequence used an echoplanar imaging volumetric navigator to track and correct subject motion in real time. A stabilising head cushion was used during imaging (Howells; Stabilising kit, UK). Freesurfer 5.1.0 was implemented on a local supercomputing cluster at the Centre for High Performance Computing (CHPC, Cape Town). Freesurfer provides cerebral cortex and white matter models to reconstruct DICOM images in order to determine volumes and cortical thickness. Brain regions are segmented into different tissue classes by applying Markov random field theory to acquire volumetric information (Desikan et al. 2006; Fischl et al. 2004). The cerebral cortex is divided into different regions as defined by gyral and sulcal structure to perform thickness measurements. Cortical thickness is determined as the closest distance from the gray or white matter boundary to the gray or cerebrospinal fluid boundary at each vertex on the image (Fischl and Dale 2009). Volumetric and cortical thickness data were extracted and compared by group using individual factorial ANOVAs in Statistica 11. Upon preliminary investigation of data, variable differences in volumes were found between boys and girls in fronto-striatal regions, while girls had greater cortical thickness across lobes (Table 1). Subsequently, group and gender were included as categorical factors. Fisher’s least significant difference (or LSD) tests were performed post hoc to differentiate significant interaction effects.
Metab Brain Dis (2014) 29:341–349
343
Table 1 Significantly greater volumes and cortical thickness by gender. These differences are consistent with previous findings on brain development including frontal, striatal, parietal and temporal regions (Lenroot and Giedd 2006; Sowell et al. 2002, 2007). Notably, there are greater
Volumes Male > female L Thalamus Female > male L Caudate R Caudate R Accumbens Cortical thickness Female > male R Fusiform gyrus R Isthmus cingulate R Lateral orbitofrontal cortex R Parahippocampal R Precuneus L Superior frontal R Superior frontal L Transverse temporal R Transverse temporal L Precentral
caudate volume and thicker cortices in right parietal and temporal regions in females (Durston et al. 2001; Giedd et al. 1996; Sowell et al. 2007). Means are presented in mm F
p
Male mean (SD)
Female mean (SD)
0.0054 (0.0003)
0.0051 (0.0003)
5.43
0.026
0.0029 (0.0003) 0.0030 (0.0004) 0.0006 (0.0001)
0.0032 (0.0003) 0.0033 (0.0009) 0.0007 (0.0001)
10.83 6.57 9.74
0.002 0.015 0.004
2.9910 (0.1546) 2.6624 (0.2034) 3.1082 (0.3188) 2.5594 (0.3288) 2.8371 (0.1447) 3.2450 (0.2159) 3.1857 (0.2581) 2.5042 (0.2922) 2.6645 (0.2036) 2.6404 (0.1362)
3.0916 (0.1019) 2.8583 (0.1579) 3.2971 (0.1295) 2.8401 (0.2573) 2.9678 (0.1379) 3.4455 (0.1497) 3.3961 (0.1756) 2.7508 (0.2190) 2.8213 (0.1960) 2.7499 (0.1399)
4.69 9.32 4.77 7.24 7.74 9.43 7.58 8.19 6.46 5.41
0.038 0.005 0.036 0.011 0.009 0.004 0.010 0.007 0.016 0.027
L left hemisphere, R right hemisphere
Results Subjects A sample of 48 children aged six years was recruited to the study. The final sample included data from 36 children; there were 18 prenatal MA-exposed children (10 male, 8 female) and 18 unexposed controls (8 male, 10 female). Although we aimed to have information by trimester regarding the time period and amount of MA use, we were unable to acquire this information. This is likely because there is fear and shame among these people associated with disclosure, especially of illicit drugs, and accuracy of reporting on drug usage is often unreliable. Children (n=12) were excluded from the initial sample for the following reasons: 1) we were unable to acquire useful scans from six MA-exposed and three unexposed children; 2) one unexposed child was excluded and referred for clinical evaluation after imaging, due to explicit low cognitive ability and probable fetal alcohol syndrome, and 3) two structural scans could not be used due to motion artefacts. The children were of mixed race, and of low socioeconomic status according to profiles on income, employment and housing. The majority of children were either in their
preparatory year (~40 %) or first year of formal schooling (~50 %). There were no significant differences in any demographic or anthropometric data between the MA-exposed and unexposed groups. See Table 2. Mothers of MA-exposed children had significantly lower educational levels compared to mothers of unexposed controls [t(1,31)=2.10, p=0.044]. The unemployment rate of MAusers was high (76 %) compared to non-users (44 %), where income was comparably low in MA-users.
Volumes and cortical thickness Group effects There was significantly increased left putamen volume in MA-exposed children compared to controls [F(1,32)=6.38, p=0.017] (Fig. 1). There was also significantly reduced left hemisphere cortical thickness of the inferior parietal [F(1,32)=6.05, p=0.020], parsopercularis [F(1,32)=6.72, p=0.014] and precuneus areas [F(1,32)=6.96, p=0.013] of MA-exposed children compared to controls (Fig. 2).
344
Metab Brain Dis (2014) 29:341–349
Table 2 Demographic and anthropometric information of prenatal MAexposed and control children
Child Age (mean, SD) Sex (male/female) Weight (kg, SD) Length (cm, SD) Head circumference (cm, SD) Mother Education (years) Employed (no/yes) Income (n, %)
