White matter abnormalities and impaired attention abilities in children born very preterm

Highlights

• The microstructure of white matter tracts involved in attention is altered.

• Altered microstructural tract organization relates to adverse attention.

• White matter tract integrity may explain later attentional abilities.

Abstract

While attention impairments are commonly observed in very preterm (< 32 weeks' gestational age) children, neuroanatomical correlates of these difficulties are unclear. We aimed to determine whether the microstructural organization of key white matter tracts thought to be involved in attention (cingulum bundle, superior longitudinal fasciculi, reticular activating system, and corpus callosum) were altered in very preterm children compared with term-born controls. We also aimed to determine whether alterations in microstructural organization of these tracts were associated with attention functioning in very preterm children. One hundred and forty-nine very preterm children and 36 term-born controls underwent neuroimaging and assessment of their attention abilities at 7 years. Constrained spherical deconvolution and probabilistic tractography was used to identify the key white matter tracts. Altered microstructural organization and reduced tract volume within reticular activating system and corpus callosum were found in the very preterm group compared with the control group. Diffusion and volume changes in the cingulum bundle, superior longitudinal fasciculi, reticular activating system, and corpus callosum were related to variations in attention functioning in the very preterm children. These findings emphasize that white matter tract integrity is associated with later attentional abilities in very preterm children.

Introduction

While children born very preterm (VPT) are at increased risk of a spectrum of medical, neurological, cognitive, and behavioral disorders (Bhutta et al., 2002, Arpino et al., 2005, Aylward, 2005), attention problems are arguably the most commonly reported concern of both caregivers and teachers (Anderson et al., 2011). As attention is an elementary ability from which other more complex cognitive abilities develop (Anderson et al., 2001, Rose et al., 2011), it is imperative that children who are at risk of developing attention difficulties are identified to enable early intervention that may reduce the impact of attention dysfunction on the other cognitive systems.

VPT children exhibit impaired basic attentional skills, such as orienting (i.e., focusing attention to relevant stimuli) and alerting (i.e., acquiring and maintaining an alert state), compared with term-born controls (van de Weijer-Bergsma et al., 2008, Mulder et al., 2009). They also demonstrate difficulties with more complex attentional processes, such as shifting (i.e. attentional flexibility) (Bayless and Stevenson, 2007, Aarnoudse-Moens et al., 2009, Mulder et al., 2011, Murray et al., 2014) and divided attention (i.e. multi-tasking ability) (Murray et al., 2014). The increased prevalence of attention problems in the VPT population may occur as a consequence of early structural damage to the brain. Between 20 and 40 weeks of gestation the CNS undergoes substantial formative change (Volpe, 2008), thus rendering infants born preterm vulnerable to brain injury. The immature vascular system of the preterm neonate makes their brains highly susceptible to injuries such as hemorrhage, hypoxia/ischemia, excitotoxicity, oxidative stress, and inflammation (Leviton et al., 2013, Raybaud et al., 2013).

White matter injury is the most common neuropathology associated with preterm birth (Volpe, 2009). In addition to the approximately 10% of infants with focal cystic and punctate lesions, MRI studies show evidence of diffuse white matter injury in approximately 50% of VPT infants (Inder et al., 2003, Miller et al., 2005, Cheong et al., 2009). Such injury leads to a reduction in mature myelin producing cells leading to impairment in myelination and axonal development (Boardman and Dyet, 2007, Volpe, 2009). As a consequence, white matter development can be atypical in the preterm brain, with abnormalities most commonly occurring within bilateral frontal, parietal and temporal regions (Ment et al., 2009). Our group and others have reported that white matter abnormality on neonatal MRI is predictive of later cognitive functioning (Woodward et al., 2006, Woodward et al., 2012, Omizzolo et al., 2014), including attention (Murray et al., 2014).

Considering the high prevalence of both white matter injury on neonatal neuroimaging and attentional difficulties in the VPT population, an exploration of white matter tracts thought to be related to attention is worthy of investigation. White matter can be examined using diffusion-weighted MRI (dMRI), a technique which models water diffusion properties and provides several key parameters that offer information regarding white matter microstructural organization. Fractional anisotropy (FA) is the metric used to determine the proportion of directional diffusion existing within a voxel (Pierpaoli and Basser, 1996). Mean diffusivity (MD) reflects the magnitude of overall diffusion within each voxel. In order to obtain additional microstructural information, the axial diffusivity (diffusion along the prominent diffusion orientation; AD) and radial diffusivity (diffusion perpendicular to the prominent diffusion orientation; RD) parameters can also be investigated.

While the literature on dMRI and prematurity is limited, some studies have found reduced fractional anisotropy (Anjari et al., 2007) and increased axial, radial, and mean diffusivity in the white matter of VPT infants, children and adolescents compared with term-born controls (Bonifacio et al., 2010, Feldman et al., 2012, Thompson et al., 2014). Studies also show abnormal microstructural organization in a number of white matter pathways in the preterm brain, including the internal capsule, external capsule, inferior fronto-occipital fasciculus, uncinate fasciculus, superior longitudinal fasciculus, inferior longitudinal fasciculus, centrum semiovale (Nagy et al., 2003, Skranes et al., 2007, Constable et al., 2008, Mullen et al., 2011, Jo et al., 2012, Lee et al., 2013, Groeschel et al., 2014), and corpus callosum (Thompson et al., 2014), which has previously been reported for the current cohort. dMRI studies exploring structure–function associations in preterm cohorts have found that alterations in white matter microstructure are related to motor deficits (Skranes et al., 2007, Counsell et al., 2008, Rose et al., 2009, Thompson et al., 2012, van Kooij et al., 2012, Northam et al., 2012a, Northam et al., 2012b), neurosensory impairments (Bassi et al., 2008, Berman et al., 2009, Reiman et al., 2009, Glass et al., 2010, Groppo et al., 2014), and behavioral dysfunction (Nagy et al., 2003, Skranes et al., 2007, Rogers et al., 2012). Altered microstructural organization has also been associated with general cognitive ability (Skranes et al., 2007, Counsell et al., 2008, Allin et al., 2011, van Kooij et al., 2012), language (Northam et al., 2012a, Northam et al., 2012b), memory (Allin et al., 2011), and executive functioning (Skranes et al., 2009). However, no study to date relates to objectively administered cognitive measures of attention.

While the functional networks of attention are well-established (Petersen and Posner, 2012), few studies have investigated the structural networks (i.e., white matter tracts) involved in attention, and most of the published studies have reported on neurotypical or adult populations. Such studies have demonstrated a link between orienting performance and the superior longitudinal fasciculus and corpus callosum (Niogi et al., 2010, de Schotten et al., 2011, Bennett et al., 2012), and alerting performance and the right superior longitudinal fasciculus and corpus callosum (Mabbott et al., 2006, Takahashi et al., 2010, Klarborg et al., 2013). Complex attention processes (i.e., shifting and divided attention) have been associated with the cingulum bundle, superior longitudinal fasciculus and corpus callosum (Schulte et al., 2008, Schulte et al., 2012, Kubicki et al., 2009, Salo et al., 2009, Takei et al., 2009, Lebel et al., 2013, Peters et al., 2014).

A fiber pathway known as the reticular activating system is also believed to play a key role in attention. The reticular activating system is made up of various source nuclei in the brainstem responsible for arousal, such as the serotoninergic raphe nuclei (Azmitia and Gannon, 1986), the noradrenergic locus coeruleus (Aston-Jones and Cohen, 2005), and the cholinergic pedunculo-pontine nucleus and laterodorsal tegmental nucleus (Jones, 2004). While the reticular activating system has been identified many times in the animal brain (Moruzzi and Magoun, 1949, Glenn and Steriade, 1982, Steriade and Glenn, 1982, Steriade et al., 1982), it is difficult to identify in the human brain due to a number of methodological barriers, such as insufficient resolution on conventional MRI to identify the source nuclei, and lack of angular resolution on diffusion tensor MRI to identify its prominent crossing fibers. To date, only one study has detailed this tract using high angular resolution diffusion imaging, and images were obtained using a high field 4.7 Tesla MRI scanner (Edlow et al., 2012).

In summary, the attention dysfunction typically observed in preterm populations may be explained, at least in part, by alterations in the white matter microstructure of the cingulum bundle, superior longitudinal fasciculus, reticular activating system and corpus callosum, and examination of these tracts in VPT children is needed. Thus, we have studied the relationship between the microstructural organization of these tracts and four aspects of attention functioning in VPT 7 year-olds. We hypothesized that VPT children would display less mature microstructural organization of attention tracts compared with controls, and among VPT children, microstructural organization of these tracts would relate to attention performance.