Introduction
Hydrocephalus is the pathological disruption of the production, absorption, or flow of cerebrospinal fluid (CSF), resulting in enlargement of the cerebral ventricles at the expense of brain tissue (Barkovich & Raybaud, Reference Barkovich and Raybaud2012; Del Bigio, Reference Del Bigio2010). In children, hydrocephalus most often occurs in congenital conditions with characteristic brain malformations that obstruct the flow of CSF, leading to the development of hydrocephalus at or shortly following birth (Menkes, Sarnat, & Maria, Reference Menkes, Sarnat and Maria2006). Hydrocephalus involves tissue-damaging mechanical forces from expansion of the ventricles, as well as secondary reactive changes, both causing gradual destruction of periventricular white matter (Del Bigio, Reference Del Bigio2010).
In the three primary forms of congenital hydrocephalus, the characteristic brain malformations are relatively specific. Aqueductal stenosis (AS) is characterized by narrowing of the cerebral aqueduct (Barkovich & Raybaud, Reference Barkovich and Raybaud2012); the Dandy Walker malformation (DWM) involves agenesis or hypoplasia of the cerebellar vermis, cystic expansion of the fourth ventricle, and enlargement of the posterior fossa (Barkovich & Raybaud, Reference Barkovich and Raybaud2012); and spina bifida myelomeningocele (SBM) is characterized by a small posterior fossa causing mechanical abnormalities of the medulla, downward herniation of the cerebellum and hindbrain, and midbrain abnormalities, including beaking of the tectum (Raybaud & Miller, Reference Raybaud and Miller2008).
It remains unclear whether neuropsychological function varies with the specific brain malformations of congenital hydrocephalus conditions. A comparison of function in AS, DWM, and SBM relative to typically developing (TD) individuals can address this issue.
Three Etiologies of Congenital Hydrocephalus
SBM is the most common severely disabling congenital birth defect affecting the central nervous system, with recent prevalence estimates ranging from 3 to 7 of every 10,000 live births (Au, Ashley-Koch, & Northrup, Reference Au, Ashley-Koch and Northrup2010). It is also the leading cause of congenital hydrocephalus, accounting for approximately 70% of hydrocephalus in children (Menkes et al., Reference Menkes, Sarnat and Maria2006). Caused by a complex pattern of gene–environment interactions, SBM involves incomplete formation of the neural tube early in gestation resulting in characteristic brain malformations, most notably the Chiari II malformation of the posterior fossa, cerebellum, and midbrain. Tectal beaking (posterior and inferior stretching of the midbrain tectum, forming a “beak”) occurs in 75% of patients. The Chiari II malformation is virtually ubiquitous in SBM, blocking the flow of CSF and necessitating treatment with a diversionary shunt.
Other etiologies of congenital hydrocephalus are rarer. Congenital AS occurs in 5–10 per 100,000 live births (Moffitt, Abiri, Scheuerle, & Langlois, Reference Moffitt, Abiri, Scheuerle and Langlois2011) and accounts for approximately 20% of children born with congenital hydrocephalus (Menkes et al., Reference Menkes, Sarnat and Maria2006). Hydrocephalus occurs in AS because of narrowing of the aqueduct of Sylvius (Barkovich & Raybaud, Reference Barkovich and Raybaud2012) and may represent a mild neural tube defect (Menkes et al., Reference Menkes, Sarnat and Maria2006). Midbrain and cerebellar dysmorphology is infrequent in AS, unless accompanied by another neurological condition (Barkovich & Raybaud, Reference Barkovich and Raybaud2012).
The prevalence of DWM is approximately 1 per 25,000 births (Hirsch, Pierre-Kahn, Renier, Sainte-Rose, & Hoppe-Hirsch, Reference Hirsch, Pierre-Kahn, Renier, Sainte-Rose and Hoppe-Hirsch1984), accounting for 5–10% of all congenital hydrocephalus (Menkes et al., Reference Menkes, Sarnat and Maria2006). The malformation consists of an enlarged posterior fossa with a raised tentorium, hypoplasia or agenesis of the cerebellar vermis, and extreme cystic dilation of the fourth ventricle filling nearly the entire posterior fossa (Barkovich & Raybaud, Reference Barkovich and Raybaud2012; Parisi & Dobyns, Reference Parisi and Dobyns2003). Hydrocephalus develops in approximately 70%, often necessitating shunting around the posterior fossa (Barkovich & Raybaud, Reference Barkovich and Raybaud2012). Anomalies of the midbrain are rare (Bolduc & Limperopoulos, Reference Bolduc and Limperopoulos2009).
Neuropsychological Outcomes
Neither AS nor SBM is typically associated with intellectual disabilities, although each condition involves patterns of neuropsychological strengths and weaknesses. Individuals with SBM show strengths in the derivation of meaning through learned associations (e.g., procedural motor learning, word decoding, vocabulary, grammar) and weaknesses in the assembly, construction, and representation of information (e.g., coordinate visual perception, spatial construction, reading and language comprehension, mathematics computation, constructional and adaptive aspects of memory and executive function; Dennis, Landry, Barnes, & Fletcher, Reference Dennis, Landry, Barnes and Fletcher2006). There is little data on outcomes in AS and DWM beyond IQ tests (Fletcher & Dennis, Reference Fletcher and Dennis2010). A recent study comparing children with SBM and AS showed reduced concept formation, memory, and fine motor domains in children with AS relative to TD children, but higher levels of performance than children with SBM (Hampton et al., Reference Hampton, Fletcher, Cirino, Blaser, Kramer and Dennis2013). Although higher severity of hydrocephalus, with presumably greater destruction of periventricular white matter, has been associated with poorer neuropsychological function in congenital hydrocephalus (Fletcher & Dennis, Reference Fletcher and Dennis2010; Hampton et al., Reference Hampton, Fletcher, Cirino, Blaser, Kramer, Drake and Dennis2011), it does not adequately account for all neurobehavioral deficits in AS and SBM (Hampton et al., Reference Hampton, Fletcher, Cirino, Blaser, Kramer and Dennis2013; Swartwout et al., Reference Swartwout, Cirino, Hampson, Fletcher, Brandt and Dennis2008), and there are specific impairments associated with the characteristic brain malformations of SBM (Dennis et al., Reference Dennis, Hopyan, Juranek, Cirino, Hasan and Fletcher2009; Dennis, Salman, Juranek, & Fletcher, Reference Dennis, Salman, Juranek and Fletcher2010).
Attention in Congenital Hydrocephalus
Attention problems represent a core neuropsychological deficit in individuals with SBM, who demonstrate dysfunction of the “posterior” attention system, but relative preservation of the regulatory “anterior” attention system (Dennis et al., Reference Dennis, Landry, Barnes and Fletcher2006; Posner & Petersen, Reference Posner and Petersen1990; Swartwout et al., Reference Swartwout, Cirino, Hampson, Fletcher, Brandt and Dennis2008). Covert orienting is one component of the larger posterior attention network, is characteristically impaired in SBM, and has been related to structural abnormalities of the midbrain and posterior cortex (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). Covert orienting has not yet been investigated in individuals with AS or DWM.
The ability to orient to salient cues in the environment is a fundamental component of survival as it allows for direction of attention toward potential threats, as well as the selection of relevant stimuli in one's environment as the focus of subsequent cognitive processing. Overt orienting involves the redirecting of eye, head, or body movements toward interesting or important stimuli, whereas covert orienting is associated with unobservable, internal shifts of attention without engaging eye, head, or body movements (Klein, Reference Klein2004). Covert visual orienting comprises three processes—engaging, disengaging, and shifting of attention (Posner, Walker, Friedrich, & Rafal, Reference Posner, Walker, Friedrich and Rafal1984).
The neural correlates of covert orienting involve the posterior attention network, including connections between the midbrain, thalamus, and posterior parietal cortex (Posner, Reference Posner1980; Posner & Petersen, Reference Posner and Petersen1990; Rafal, Posner, Friedman, Inhoff, & Bernstein, Reference Rafal, Posner, Friedman, Inhoff and Bernstein1988). The lateral pulvinar of the thalamus is involved in engaging attention, and the posterior parietal cortex is associated with disengaging attention. Classic lesion studies (Posner et al., Reference Posner, Walker, Friedrich and Rafal1984; Rafal et al., Reference Rafal, Posner, Friedman, Inhoff and Bernstein1988), and more recent functional neuroimaging (Gitelman, Parrish, Friston, & Mesulam, Reference Gitelman, Parrish, Friston and Mesulam2002) and primate (Ignashchenkova, Dicke, Haarmeier, & Thier, Reference Ignashchenkova, Dicke, Haarmeier and Thier2004; Lovejoy & Krauzlis, Reference Lovejoy and Krauzlis2010; Nummela & Krauzlis, Reference Nummela and Krauzlis2010) investigations, highlight the importance of the superior colliculus in the midbrain tectum as a guiding and selecting mechanism for the covert orienting network. In addition, patients with damage to the superior colliculus redirect toward new locations at a slower pace and experience difficulty ignoring previously attended locations (Posner, Rafal, Choate, & Vaughn, Reference Posner, Rafal, Choate and Vaughn1985; Posner et al., Reference Posner, Walker, Friedrich and Rafal1984; Rafal et al., Reference Rafal, Posner, Friedman, Inhoff and Bernstein1988).
Relative to TD children, children with SBM show deficits in disengaging and shifting of attention (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). Consistent with clinical (e.g., Parinaud syndrome involving paralysis of upward gaze during shunt blocks) and experimental (e.g., more difficulty bisecting lines in the vertical plane and altitudinal neglect; Dennis, Edelstein, Frederick, et al., Reference Dennis, Edelstein, Frederick, Copeland, Francis, Blaser and Fletcher2005) evidence that the vertical plane is especially vulnerable to disruption in SBM, covert orienting deficits tend to occur in the vertical, but not horizontal, plane. In addition, impairments in covert orienting are most significant in individuals with SBM and tectal beaking (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b).
Present Study
It remains unclear whether deficits in covert orienting in SBM result predominantly from the effects of characteristic brain malformations or from the effects of hydrocephalus. By comparing performance across three etiologies of congenital hydrocephalus, each with a unique spectrum of midbrain and posterior fossa malformations ranging from isolated (AS) to extensive but very different (DWM, SBM), the present study evaluated the effects of specific midbrain and posterior fossa malformations on covert orienting in individuals who share a history of hydrocephalus. Given the importance of the midbrain tectum for covert orienting, we hypothesized a stepwise order of group performance reflecting the degree of midbrain tectum dysmorphology. Thus, we expected TD individuals and individuals with AS to demonstrate intact covert orienting, whereas we expected individuals with SBM and tectal beaking (SBM-tectal beaking) to show the greatest impairment. Although individuals with SBM and a normal tectum (SBM-no tectal beaking) and individuals with DWM were not expected to show significantly impaired covert orienting performance, we expected they might perform better than the SBM-tectal beaking group, but less well than the TD and AS groups.
Method
Participants
Participants included 68 individuals with SBM-tectal beaking, 17 individuals with SBM-no tectal beaking, 9 individuals with DWM, 16 individuals with AS, and 41 TD individuals involved in a larger study of neuropsychological function in SBM and related disorders (Fletcher et al., Reference Fletcher, Northrup, Landry, Kramer, Brandt, Dennis and Francis2004). Most data from the TD and SBM groups was previously reported (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). The present study adds previously unpublished data from the DWM and AS groups, with the addition of 16 individuals with SBM and four TD individuals due to recruitment subsequent to the previous publications. Participants in the clinical groups were recruited from clinics and neurosurgical practices in Houston and Toronto. TD participants were volunteers and friends (not siblings) of the families of the participants with SBM. Participants were English-speaking and had Verbal Reasoning and/or Abstract/Visual Reasoning scores of at least 70 on the Stanford-Binet Test of Intelligence (4th ed.; Thorndike, Hagen, & Sattler, Reference Thorndike, Hagen and Sattler1986). Individuals were excluded if they had neurological disorders unrelated to SBM, DWM, or AS, prematurity or low birth weight, severe psychiatric disorders, or uncontrolled seizure disorder, all through review of history at screening. Exclusions were rare, but the number is not known. In addition, participants were excluded for uncorrected vision or hearing impairment through vision screening and formal audiometry, the latter leading to two exclusions. Finally, there was one exclusion due to inability to control the upper limbs. The study was approved by the human participants review boards at all institutions. Parents and participants gave written consent unless the participant was under 13, in which case the parent consented and the child assented.
Table 1 contains demographic information and IQ scores by group. Participants ranged in age from 6 to 26 years, with the inclusion of young adults essential for recruiting sufficient sample sizes of the rarer conditions of hydrocephalus. There were no significant differences among groups in SES, F(4,143) = 1.12, p = .35, gender, χ 2 (4) = 2.78, p = .60, or ethnicity, Fisher's exact, p = .16. The mean age of participants in the SBM-tectal beaking, F(1,146) = 14.03, p < .001, and TD, F(1,146) = 9.62, p = .002, groups was lower than participants in the AS group. Participants in the TD group had higher IQ relative to all other groups, F(1,145) = 30.42, p < .001, and participants in the AS group had higher IQ than the SBM-tectal beaking group, F(1,145) = 11.15, p = .001.
Demographic information and IQ by group
Note. AS = aqueductal stenosis; DWM = Dandy-Walker malformation; SBM-no tectal beaking = spina bifida myelomeningocele-no tectal beaking; SBM-tectal beaking = spina bifida myelomeningocele-no tectal beaking; TD = typically developing.
aMissing SES data for 1 participant with SBM-TB and 2 participants with AS.
bMissing IQ data for 1 participant with AS.
*p < .05.
All participants in the clinical groups had shunted or arrested hydrocephalus. Participants with SBM were identified at birth, and diagnosis was confirmed by neuroradiology and pathology; participants with DWM and AS were diagnosed in the first year of life by neuroradiology. Three individuals with AS and tectal dysmorphology were excluded to avoid confounding etiology with the key variable of tectal dysmorphology (on which our hypotheses were based). Only participants with classic variants of DWM were included.
Table 2 presents clinical markers, obtained from medical records, commonly used to characterize individuals with congenital hydrocephalus. The large majority of participants had a history of shunted hydrocephalus, although the small group with DWM had a larger proportion of participants with a history of arrested hydrocephalus relative to the other groups, Fisher's exact, p = .04. There were no significant differences between groups with regard to history of shunt complications, Fisher's exact, p = .39, shunt revisions, Fisher's exact, p = .35, oculomotor disorder, χ 2 (3) = 5.46, p = .14, or history of seizure disorder, Fisher's exact, p = .52. As expected, the groups with SBM had higher proportions of participants with impaired ambulation and neurogenic bladder relative to the DWM and AS groups, Fisher's exact, p < .001. In addition, a history of nystagmus was present only in the SBM-tectal beaking group, Fisher's exact, p = .03, accounting for 15% of individuals. Because nystagmus might affect covert orienting, we compared performance of individuals with SBM with and without nystagmus and found no significant differences (ps > .50). Thus, a history of nystagmus was not expected to affect the results.
Number (percentage) of participants with clinical markers by group
Note. AS = aqueductal stenosis; DWM = Dandy-Walker malformation; SBM-no tectal beaking = spina bifida myelomeningocele-no tectal beaking; SBM-tectal beaking = spina bifida myelomeningocele-no tectal beaking; TD = typically developing.
*p < .05.
MRI scans of the brain were available for all participants in the clinical groups. Two pediatric radiologists blinded to condition coded tectal beaking as present or absent using a formal set of conventions. To establish reliability of the coding scheme, the first 20 scans were coded by both radiologists. Disagreements were discussed; however, a formal reliability study was not conducted because of the time and expense of coding across sites. We attempted to code for degree of impairment of several small structures, such as the tectum, but did not find such coding to be reliable. We instead focused on simple, obvious visual distinctions of present versus absent. The validity of this coding scheme is demonstrated by its sensitivity to differences in connectivity using diffusion tensor imaging (DTI) in a recent study on a different cohort (Williams et al., Reference Williams, Juranek, Stuebing, Cirino, Dennis and Fletcher2013).
MRI abnormalities by etiology group are presented in Table 3. There were no significant differences between groups with regard to hydrocephalus at study participation, Fisher's exact, p = .78, type of ventricular dilation, Fisher's exact, p = .70, lateral ventricles, Fisher's exact, p = .42, or third ventricle, Fisher's exact, p = .17. Clinical groups differed on several key variables. The fourth ventricle was small in most participants in the two SBM groups, enlarged in the DWM group, and normal in the AS group, Fisher's exact, p < .001. The majority of participants in the two groups with SBM had Type II Chiari malformations. Three participants with SBM were identified as having only some features of a Chiari malformation. The two participants with SBM who did not have Chiari malformations had low lumbar spinal lesions and the source of their hydrocephalus was AS. None of the participants with DWM, and only one of the participants with AS, had a Chiari malformation, Fisher's exact, p < .001. The cerebellum was abnormal in all participants with SBM-tectal beaking and DWM and the majority of participants with SBM-no tectal beaking, but was normal in the majority of participants with AS, Fisher's exact, p < .001. The group with DWM had lower proportions of participants with mild ventricular dilation, and higher proportions of participants with severe ventricular dilation, relative to the SBM-no tectal beaking group only, Fisher's exact, p = .010. In addition, the group with DWM had a higher proportion of participants with a normal corpus callosum relative to the SBM-tectal beaking group only, Fisher's exact, p = .001.
Number (percentage) of participants with MRI abnormalities by group
Note. AS = aqueductal stenosis; DWM = Dandy-Walker malformation; SBM-no tectal beaking = spina bifida myelomeningocele-no tectal beaking; SBM-tectal beaking = spina bifida myelomeningocele-no tectal beaking; TD = typically developing.
*p < .05.
Cued Orienting Task
The cued orienting task, designed to measure disengagement cost and inhibition of return, was based on Posner's paradigm, described in Rafal et al. (Reference Rafal, Posner, Friedman, Inhoff and Bernstein1988), and previously used by Dennis et al. (Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). Covert orienting depends on: (1) the nature of cue validity, and (2) the time between a cue and a forthcoming target (the stimulus onset asynchrony [SOA]). Disengagement cost involves: (1) a shorter reaction time for detection of a peripheral target when preceded by a cue validly indicating the location of the forthcoming target, and (2) a longer reaction time for detection of a peripheral target when an invalid cue precedes a target (Akhtar & Enns, Reference Akhtar and Enns1989). Disengagement cost measures the cost of misdirecting attention on invalidly cued trials. Inhibition of return (IOR) is a phenomenon in which new targets are favored over previously encountered targets at certain SOAs (Posner et al., Reference Posner, Rafal, Choate and Vaughn1985). If a target occurs in close time proximity (e.g., 200 ms) to a valid cue indicating the location of an upcoming target, the reaction time required to provide a response is decreased; however, as the SOA becomes longer, the response to validly cued targets is no longer facilitated and becomes slower (inhibited).
Procedure
The task was administered on an IBM computer, coded in Micro Experimental Laboratory (Version 2.0), and scored in SAS 9.3. A chin rest with a forehead bar established distance between the screen and the respondent's eye level (47.5 cm). Figure 1 shows a depiction of the task. Participants were instructed to fixate on a black cross in the middle of a white background. The cross was continuously present throughout the task. The goal of the task was to detect a target occurring in one of four equally probable peripheral locations (left, right, above, or below the central fixation point) while maintaining central fixation, and to respond with a button press upon detection. The target was always primed by an exogenous cue in the form of a brightness change in one of the peripheral locations. The cue validly forecasted an upcoming target in 50% of trials. On each trial, one of two SOAs was used: (1) a short SOA equal to 200 ms, or (2) a long SOA equal to 1000 ms. When the task operates as intended, the implementation of the two different SOAs brings about the occurrence of a disengagement cost at the short SOA in response to invalidly cued targets and IOR at the long SOA in response to validly cued targets.
Depiction of cued orienting task. Participant fixates on continuously present centrally-located cross. A cue in the form of a brightness change is presented for 150 ms to the left, right, above, or below the fixation point. A stimulus onset asynchrony of 200 ms or 1000 ms occurs between the cue and the forthcoming target. A target is then presented to the left, right, above, or below the fixation point and remains present until the participant indicates detection with a button press. The cue validly forecasts the location of the target in 50% of trials.
Participants were familiarized with the task by completing 5 practice trials before every block. The trials were presented in two blocks with no fewer than 80 trials each. The interval between trials was equal to 1500 ms. Trials with RTs shorter than 150 ms or longer than 2000 ms were considered unreliable and, therefore, not included in analyses. For each block, spoiled trials were repeated after the 80th trial.
Computation of disengagement cost and IOR
The estimation of disengagement cost and IOR was consistent with computations in Dennis et al. (Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). Median RTs for each participant were separately computed across trials for validly and invalidly cued targets with regard to plane (vertical and horizontal) and SOA (short and long). Disengagement cost was the relative difference in median RTs between validly and invalidly cued targets at the short SOA. IOR was the relative difference in median RTs between validly and invalidly cued targets at the long SOA.
Data Analyses
To investigate the effect of group on covert orienting, median RT differences between validly and invalidly cued targets were compared among the five groups (SBM-tectal beaking, SBM-no tectal beaking, DWM, AS, and TD) using a Group × SOA repeated-measures analysis of covariance (ANCOVA) controlling for age. Age was not a significant covariate (p > .05) in any models and is, therefore, not further reported. Analyses were computed separately for the vertical and horizontal planes. Because many of the groups had small sample sizes, we focused on the magnitude of the effect sizes rather than on tests of statistical significance. The effect sizes for raw RT differences were computed and interpreted when small to medium (d ≥ 0.35) or large (d ≥ 0.8; Cohen, Reference Cohen1988).
Results
Figure 2 presents the median RT differences between validly and invalidly cued targets for each group at the short and long SOA in the vertical and horizontal planes. Visual inspection shows the hypothesized stepwise ordering of groups for the 200 ms interval; however, the repeated measures ANCOVA did not reveal a significant group by SOA interaction in the vertical, F(4,141) = .60, p = .66, or horizontal, F(4,141) = .23, p = .92, planes. There was a significant main effect of SOA in both the vertical, F(1,141) = 11.52, p < .001, and horizontal, F(1,141) = 18.72, p < .001, planes, indicating that, collapsed across groups, RT differences for validly and invalidly cued targets were longer at the 200 ms SOA than at the 1000 ms SOA. What is more important in Figure 2 is the direction of the differences; RT differences at the short SOA were negative, confirming the occurrence of the disengagement cost associated with invalidly cued targets (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a), whereas RT differences at the long SOA were positive, confirming the occurrence IOR (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b; Posner et al., Reference Posner, Rafal, Choate and Vaughn1985). These effects indicate that the task operated as intended across groups.
Median RT differences between validly and invalidly cued targets by group at the short and long stimulus onset asynchrony (SOA) in the vertical and horizontal planes. Negative RT differences at 200 ms SOA represent the disengagement costs associated with misdirecting attention on invalidly cued trials; positive RT differences at 1000 ms SOA represent shorter RTs to invalidly cued targets, which are evidence of inhibition of return (IOR). Error bars reflect standard error. AS = aqueductal stenosis; DWM = Dandy-Walker malformation; SBM-no tectal beaking =spina bifida myelomeningocele-no tectal beaking; SBM-tectal beaking = spina bifida myelomeningocele-no tectal beaking; TD = typically developing.
Table 4 provides effect sizes for the interaction of group by SOA on RT differences between validly and invalidly cued targets. The group with SBM-tectal beaking demonstrated the highest disengagement cost in the vertical plane, with a large effect size relative to the TD group (RTdifference =115.54 ms), and a medium effect size relative to the group with AS (RTdifference = 79.48 ms). In addition, the groups with SBM-no tectal beaking (RTdifference = 52.40 ms) and DWM (RTdifference = −79.74 ms) also demonstrated a higher disengagement cost relative to the TD group with medium effect sizes. There were no notable effects for disengagement cost in the horizontal plane.
Effect sizes by group and SOA for reaction time differences between validly and invalidly cued targets
Note. AS = aqueductal stenosis; DWM = Dandy-Walker malformation; SBM-no tectal beaking = spina bifida myelomeningocele-no tectal beaking; SBM-tectal beaking = spina bifida myelomeningocele-no tectal beaking; TD = typically developing.
Effect sizes of small to medium (d ≥ 0.35) or large (d ≥ 0.8) in bold.
For IOR in the vertical plane, the groups with SBM-tectal beaking (RTdifference = 51.51 ms, small effect size), DWM (RTdifference = −52.65 ms, medium effect size), and AS (RTdifference = −34.99 ms, small effect size) showed more attenuated (more impaired) IOR relative to the TD group. In the horizontal plane, the group with SBM-no tectal beaking demonstrated less attenuated (less impaired) IOR relative to the groups with DWM (RTdifference = −72.36 ms, large effect size), SBM-tectal beaking (RTdifference = 58.17 ms, medium effect size), and AS (RTdifference = −41.81 ms, medium effect size).
Discussion
In this study, we explored the specificity of the relation between covert orienting and the midbrain by comparing covert orienting in three etiologies of congenital hydrocephalus (AS, DWM, and SBM) that varied in midbrain and posterior fossa malformations. The results bear on the role of hydrocephalus in covert orienting, and on the relation of covert orienting deficits to specific dysmorphologies of the midbrain and posterior fossa. All groups exhibited the expected task effects of a disengagement cost at the short SOA and IOR at the long SOA. The results revealed a stepwise order of group performance in the vertical plane that helps disentangle the relative effects of hydrocephalus and midbrain and posterior fossa dysmorphology on covert orienting deficits.
When the effects of hydrocephalus were examined by comparing TD individuals to individuals with AS who have a history of hydrocephalus but minimal midbrain and posterior fossa dysmorphology, attentional disengagement did not differ and the two groups performed similarly. This finding suggests that impairments in attentional disengagement were not attributable to general effects of hydrocephalus.
Disengagement deficits were also related to the specific type of structural anomaly in the midbrain and posterior fossa. As hypothesized, the group with SBM and tectal beaking of the midbrain demonstrated the greatest impairment in disengagement of attention in the vertical plane, consistent with previous investigations (Dennis et al., Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005a, Reference Dennis, Edelstein, Copeland, Frederick, Francis, Hetherington and Fletcher2005b). Thus, the specific type of neuropathology, not hydrocephalus alone, is important for covert orienting.
Other studies have linked neuropsychological impairments to specific brain malformations in SBM rather than to hydrocephalus alone. Swartwout et al. (Reference Swartwout, Cirino, Hampson, Fletcher, Brandt and Dennis2008) found comparable performances across TD children and children with AS on a measure of sustained attention, whereas children with SBM demonstrated a higher number of omission errors and slower reaction times relative to the children with AS. When compared across several neuropsychological domains, Hampton et al. (Reference Hampton, Fletcher, Cirino, Blaser, Kramer and Dennis2013) showed poorer performance in children with SBM on most domains relative to children with AS, although both groups performed generally less well than a TD comparison group. Together, the results of these studies along with the present findings indicate that neuropsychological dysfunction in conditions of congenital hydrocephalus cannot be attributed solely to the effects of hydrocephalus on the brain, but that the various characteristic brain malformations in these conditions can also contribute to specific neuropsychological impairments.
The data also bear on the specificity of the covert orienting deficit within a neurodevelopmental disorder involving hydrocephalus. Consistent with our stepwise order of group performance hypothesis, the groups with SBM-no tectal beaking and DWM demonstrated mid-range performance, showing significantly greater disengagement costs in the vertical plane relative to the TD group, though slightly better performance relative to the group with SBM and tectal beaking. These results suggest that other factors may be exerting a mildly impairing effect on covert orienting in these individuals. One such factor could be dysmorphology of the cerebellum, which affects timing and sequencing of motor movements and may depress response speed. In addition, although the cortex has not been closely examined in individuals with DWM, individuals with SBM have regions of aberrant cortical thickness and cortical complexity that are associated with more impaired performance on broad measures of cognitive function (Treble, Juranek, Stuebing, Dennis, & Fletcher, Reference Treble, Juranek, Stuebing, Dennis and Fletcher2012). Neuroimaging tools, especially those examining white matter pathways, could be used to examine the integrity of neural networks underlying covert orienting in individuals with various etiologies of congenital hydrocephalus. In a recent study with a different sample, Williams et al. (Reference Williams, Juranek, Stuebing, Cirino, Dennis and Fletcher2013) used DTI to show reduced integrity of white matter pathways from the midbrain in SBM, with less integrity in parietal than frontal pathways in those with tectal beaking.
The patterns of results for IOR were less clear than those for attentional disengagement as the pattern of effect sizes was not consistent. Although there was a small to medium effect size difference between TD individuals and individuals with SBM-tectal beaking for the measure of IOR, the effect size was large for the comparison of DWM to TD individuals. In the horizontal plane, the SBM-no tectal beaking group had better IOR than the AS, DWM, and SB-tectal beaking groups, reflecting surprisingly strong performance by the former group. The basis for the pattern is not clear. The TD, SBM-no tectal beaking, and AS groups performed similarly, but better than DWM and SB-tectal beaking. Why this pattern would emerge just in the groups with posterior fossa and cerebellar malformations is not clear and difficult to address with a small sample of DWM and no additional high quality neuroimaging data, especially addressing connectivity issues involving the midbrain in this cohort.
The results of the present study point to tectal beaking as an easily detectable biomarker of increased risk for dysfunction of the posterior attention system in SBM. Individuals with SBM and tectal beaking will be more likely to demonstrate difficulties orienting to important stimuli in the environment and difficulties disengaging and shifting attention to other important stimuli. For example, 18-month-old infants with SBM took significantly longer than TD infants to shift or disengage their attention from a salient visual stimulus when tested with a habituation-dishabituation paradigm (Taylor et al., Reference Taylor, Landry, Barnes, Swank, Cohen and Fletcher2010). Clinically, individuals with this type of attention impairment often appear lethargic and slow to respond and seemingly distracted. The behavioral presentation of “bottom-up” posterior attention impairments in SBM differs from “top-down” anterior attention system impairment, as evident in individuals with ADHD who demonstrate higher frequencies of hyperactivity and impulsivity relative to individuals with SBM (Burmeister et al., Reference Burmeister, Hannay, Copeland, Fletcher, Boudousquie and Dennis2005; Colvin, Yeates, Enrile, & Coury, Reference Colvin, Yeates, Enrile and Coury2003). As such, the findings cluster with other disorders that seem to share a more posterior profile of attention impairment, such as in children treated for acute lymphoblastic leukemia (ALL). As opposed to regulatory deficits related to voluntary attention, children with SBM and ALL are underactivated, with primary difficulties with involuntary attention (Dennis, Sinopoli, Fletcher, & Schachar, Reference Dennis, Sinopoli, Fletcher and Schachar2008). Such findings may help explain what often appears as a poor response to stimulant medication in SBM (Davidovitch et al., Reference Davidovitch, Manning-Courtney, Hartmann, Watson, Lutkenhoff and Oppenheimer1999), although this question certainly warrants more investigation. The early identification of children with SBM and tectal beaking would allow for earlier provision of interventions and accommodations targeting impairments of attention orienting.
Limitations
Limitations of the present study include the small sample sizes of the SBM-no tectal beaking, AS, and DWM groups, as well as the heterogeneity of the latter two groups. Due to the rarity of AS and DWM, recruiting adequate sample sizes is challenging. In addition, a thorough understanding of the determinants of functional outcome remains elusive in these conditions. As such, the present study contains individuals from a wide range of age and level of functioning, which may have impacted the findings. Power was not adequate to detect small and medium effect size differences. We focused on effect size differences because it would take larger samples of the rare groups to disentangle the relations of structural brain pathology and covert orienting, and to balance against Type II errors. We provided limited control for the number of comparisons because of the power issue and the rarity of the clinical groups.
Despite these limitations, the present study is one of only a few neuropsychological comparisons involving AS and DWM, which largely report IQ scores and retrospective chart reviews. Although the neuropsychological profiles of individuals with AS and DWM are rarely examined, the present findings emphasize the significance of etiology (and hence, of neuropathology) for outcome in conditions of congenital hydrocephalus and lay the groundwork for future research to explore sources of variability within these conditions.
Conclusions
The findings of the present study clarify a long-standing question regarding the relative contributions of characteristic brain malformations and hydrocephalus to neuropsychological dysfunction in conditions of congenital hydrocephalus. While the detrimental effects of hydrocephalus on white matter have received the greatest clinical and research attention, our findings show that characteristic brain malformations in conditions of congenital hydrocephalus result in specific neuropsychological impairments that are not attributable to the effects of hydrocephalus. The results are particularly informative for SBM because tectal beaking, a malformation that is exclusively but variably present in SBM, was associated with the greatest degree of impairment in covert orienting, a cognitive function fundamental for the selection of relevant stimuli in the environment as the focus of subsequent cognitive processing. Continued exploration of the effects of specific brain malformations in SBM, and the beginning of this line of research in less frequently studied conditions of congenital hydrocephalus, will allow for a more complete understanding of pathological mechanisms in these conditions and better prediction of outcomes.
Acknowledgments
This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Grant (P01 HD35946-06, “Spina Bifida: Cognitive and Neurobiological Variability”). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Eunice Kennedy Shriver National Institute of Child Health and Human Development or the National Institutes of Health. Conflict of Interest: None declared.
