DEVELOPMENTAL DISABILITIES RESEARCH REVIEWS 15: 35 44 (2009)

MATHEMATICAL LEARNING DISABILITY IN GIRLS WITH TURNER SYNDROME: A CHALLENGE TO DEFINING MLD AND ITS SUBTYPES
Mich ele M.M. Mazzocco*
Math Skills Development Project, Kennedy Krieger Institute, and Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland

Turner syndrome is a common disorder with a prevalence of 1:2,500 live female births. Although not associated with mental retardation, there is an increased risk of learning difficulties in this population. In particular, mathematical learning difficulties among girls with Turner syndrome are prevalent, significant, and persistent. As such, the study of mathematical performance in girls with Turner syndrome presents opportunities to advance our knowledge of mathematics ability, disability, and disability subtypes. Moreover, the Turner syndrome phenotype illustrates the challenges faced when defining mathematical learning disability (MLD) and characterizing MLD subtypes because the cognitive phenotype is aligned with several proposed MLD subtypes. There is some evidence linking MLD in Turner syndrome with spatial deficits, with executive dysfunction, and with deficient numerosity skills. Yet there is also conflicting evidence as to whether any of these explanations underlies MLD in Turner syndrome. Most mathematical difficulties in girls with Turner syndrome, as a group, occur on timed tests or on complex problems. On untimed tests, achievement test scores may be age appropriate. Therefore, the inclusion of MLD in the Turner syndrome cognitive phenotype reminds us that we cannot rule out MLD solely on the basis of performance on an untimed calculations subtest, and it poses a challenge to the widespread practice in which many researchers engage, that is, defining MLD on the basis of broad mathematics achievement ' 2009 Wiley-Liss, Inc. test outcomes.
Dev Disabil Res Rev 2009;15:3544.

in persons with Turner syndrome is mixed. Following a brief overview of Turner syndrome and its cognitive phenotype, I present information on the mathematical difficulties associated with Turner syndrome in the context of these proposed subtypes and potential core deficits of MLD. AN OVERVIEW OF TURNER SYNDROME Turner syndrome results from the complete or partial loss of the second X chromosome that is typically present in a female, with a variety of karyotype outcomes [reviewed by Davenport et al., 2007]. It occurs in approximately 1:2,000 to 1:5,000 live female births [Nielsen and Wohlert, 1991; Davenport et al., 2007], and is sporadic rather than familial in nature. The well-documented physical phenotype of Turner syndrome includes slowed and disproportionate growth, cardiovascular abnormalities, and endocrine disruptions including delayed puberty or, more frequently, ovarian dysgenesis. The most common featureshort statureis evident in most persons with this disorder [S avendahl and Davenport, 2000, p. 457]. Although lifespan expectancy is normal for individuals affected by Turner syndrome, quality of life issues prevail, and health management can be challenging [Bondy and The Turner Syndrome Consensus Study Group, 2007]. The Turner Syndrome Cognitive Phenotype The notion that there is a cognitive phenotype of Turner syndrome is well established, having been a topic of research since 1962 [Shaffer, 1962]. At the most global level of functioning, most females with Turner syndrome have full scale IQ (FSIQ) scores well within the average range; thus mental retardation is not a characteristic of the phenotype [Rovet, 1993; Temple and Carney, 1993; Elliott et al., 1996;

Key Words: Turner syndrome; mathematical learning disability; nonverbal learning disability; cognitive phenotype

he premise of this review is that the study of mathematical performance in girls with Turner syndrome presents opportunities to advance our knowledge of mathematics ability, disability, and disability subtypes. Phenotype studies can contribute to delineating the neurocognitive pathways underlying specific behavioral characteristics, including mathematical learning disabilities (MLD). However, an interesting outcome of studying the Turner syndrome phenotype is how well the phenotype illustrates the challenges faced when defining MLD and characterizing MLD subtypes. The cognitive phenotype is aligned with more than one proposed MLD subtype, and yet there is conflicting evidence as to whether any of the potential subtypes offers appropriate explanations for why MLD occurs in Turner syndrome. Likewise, the evidence for deficient or intact number processing

Grant sponsor: National Institute of Child Health and Human Development; Grant number: HD 34061. *Correspondence to: Mich ele M. M. Mazzocco, Math Skills Development Project, 3825 Greenspring Avenue, Painter Building, Top Floor, Baltimore, MD 21211. E-mail: Mazzocco@kennedykrieger.org Received 30 November 2008; Accepted 22 December 2008 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ddrr.50

' 2009 Wiley -Liss, Inc.

Mazzocco, 1998]. The hallmark cognitive feature is a marked discrepancy between Verbal (VIQ) and Performance IQ (PIQ) scores [LaHood and Bacon, 1985; Pennington et al., 1985; Ross et al., 2000; Tamm et al., 2003], which ranges from 0 to 40 points (scores based on M 5 100, SD 5 15), with an average of 1215 points [according to Davenport et al., 2007]. Thus, despite much variability, even the average V-P discrepancy in this population is significant. Still, the V-P split fails to sufficiently describe individuals with the disorder (much less the population as a whole) because performance strengths and weaknesses are evident within both verbal and nonverbal skills. Moreover, the relationship between VIQ and PIQ differs among girls with Turner syndrome, relative to the general population. These scores are highly correlated with each other in girls with Turner syndrome, but not among girls without Turner syndrome; and in girls with Turner syndrome, PIQ is not correlated with response time on complex visual spatial tasks, but it is among girls without Turner syndrome [Rovet and Netley, 1982]. As such, knowledge of the Verbal-Performance IQ split is, at best, a starting point for investigating the cognitive phenotype of Turner syndrome, rather than an explanation for difficulties learning mathematics. Below, we use this starting point to describe the reported verbal strengths and nonverbal weaknesses, as well as the inconsistencies within each of these domains. Verbal skills in females with Turner syndrome As a group, females with Turner syndrome have average to above average performance on most, but not all verbal tasks [Waber, 1979; Temple, 2002]. For example, in Temples [2002] study of 9to 12-year-old girls with Turner syndrome, she found that girls with Turner syndrome had significantly better receptive vocabularies than their peers, as measured by picture vocabulary; and did not differ from their peers on confrontational naming. Likewise, Rovet and Netley [1982] found no group differences between 15-year-old girls with versus without Turner syndrome, on a sentence verification task that required comprehension of actions rather than only single words. Girls with Turner syndrome did not differ from their peers when asked to produce narrative descriptions of pictures, or when asked to describe the steps needed to organize a party [Temple, 2002]. Verbal strengths 36

were also evident in verbal short-term memory, as measured by digit span forward tests; and in reading skills such as letterword identification, phonological decoding, and reading comprehension, all of which met or exceeded age-level expectations [Waber, 1979; Rovet, 1993; Temple and Carney, 1996; Murphy and Mazzocco, 2008]. Although dyslexia has been reported in some case of Turner syndrome, its prevalence does not appear to exceed rates reported for the general population; instead, hyperlexia is more representative of the phenotype [Temple and Carney, 1996]. In contrast, Temple [2002] also found selective difficulties on verbal tasks. In her study, girls with Turner syndrome produced significantly fewer words than their peers on a timed oral fluency task that required generating

preparation]. However, at least during the primary school years (an age when RAN performance is predictive of dyslexia), the lack of deficient RAN performance in girls with Turner syndrome is consistent with the absence of dyslexia in the phenotype. In summary, most verbal skills are age appropriate in girls with Turner syndrome during early and later development, and collectively represent mostly typical function and development. Selected areas of weakness are observed on oral fluency, unrelated to deficits in lexical storage or access. These skills may influence mathematical learning and function with respect to semantic comprehension of number words, and word problem comprehension. Although age appropriate verbal encoding, decoding, and short-term memory should promote fact retrieval skills, slow oral fluency may impact this area of mathematical performance. Visual spatial skills in females with Turner syndrome Deficits in visual spatial skills are among the earliest and most consistently reported findings from studies of the Turner syndrome cognitive phenotype [e.g., Shafer, 1962; Alexander et al., 1966; Waber, 1979; Rovet and Netley, 1982], leading to claims that the Turner syndrome deficit appears to be confined primarily to processes involved in dealing with spatial information [Rovet and Netley, 1982, p. 90]. Still, even the earliest studies produced inconsistent evidence across components of this highly complex domain [e.g., Silbert et al., 1977; Waber, 1979], and conflicting support for specific right hemisphere [Alexander et al., 1966] or bilateral [Pennington et al., 1985] dysfunction. Shafer [1962] reported poor visual memory and poor perceptual organizational scores in girls and women with Turner syndrome, further supported by some [e.g., Alexander et al., 1966] but not all [e.g., Waber, 1979] of the later studies. Shafer found that females with Turner syndrome showed below age-level performance on the Wechsler Block Design and Object Assembly subtests, both of which have a visuo-constructional component; and also on the Benton Visual Retention that requires immediate reproduction of geometric shapes. More recent studies also show construction deficits in young primary age school girls with Turner syndrome, using similar block construction tasks such as the Copying and Pattern Analysis tasks from the Stanford
MLD
in

Knowledge of the Verbal Performance IQ split is a starting point for investigating the cognitive phenotype of Turner syndrome, rather than an explanation for observable deficits such as difficulties learning mathematics.
words beginning with a specific initial consonant. Considered in the context of superior vocabulary, Temple argued that these oral fluency deficits are not linked to lexical storage, but that lexical access is implicated by the weaker (i.e., age appropriate) confrontational naming relative to superior receptive vocabulary. However, she added that intact naming skills fail to support this latter notion, which is further refuted by findings that girls with Turner syndrome are unimpaired on rapid automatized naming (RAN) of numbers and letters. On these RAN subtests, girls with Turner syndrome are as fast and as accurate as their peers, at least through elementary school [Mazzocco 2001; Murphy and Mazzocco, 2008]; although there is some evidence of slower performance than peers by the end of middle school [Mazzocco and Murphy, manuscript, in
Dev Disabil Res Rev


Turner Syndrome

Mazzocco

Binet Fourth Edition (SBFE), and pencil and paper copying tasks such as the Beery visual motor integration task [e.g., Mazzocco, 2001]. Still, visuo-constructional tasks overall do not lead to the most severely deficient spatial skills scores seen in Turner syndrome [e.g., Rovet and Buchanan, 1999], unlike drawing tasks [e.g., Temple and Carney, 1995], for which the evidence of deficits is more robust. Likewise, not all researchers have reported evidence of visual memory deficits among girls with Turner syndrome. For instance, Alexander et al. [1966] concluded that the poor reproductions on the Benton Visual Retention test resulted from poor shape-form perception rather than poor memory per se. This notion received further support from Rovet and Netley [1982], who argued that encoding of (spatial) information was not affected by Turner syndrome, whereas the transformation of spatial information was affected. In their study, girls with Turner syndrome were less accurate than their peers on a classic, multiple choice, mental rotation task [Shepard and Metzler, 1971]; they were also slower on the task, and yet the strategies in which they engaged were similar to those of their peers. For both groups, accuracy varied as a function of angular separation between the target and correct response, and to similar degrees. Yet in a later study that involved letter stimuli rather than the classic cube patterns, no such mental rotation deficits were observed among girl with Turner syndrome [Murphy et al., 1994]. Additional evidence for selective slowed response times on visual spatial tasks is drawn from computerized ocular motor studies [Lasker et al., 2007]. When directing their gaze from a central fixation point on a computer screen, girls with Turner syndrome took significantly longer than their peers to initiate visually guided saccades (i.e., when asked to look quickly at an appearing target)particularly leftdirected saccadesdespite comparable response accuracy across groups; they took significantly longer to initiate memory guided saccades (i.e., when asked to look at where a target had appeared, but only after the target disappeared). On an anti-saccade task (i.e., when asked to suppress eye movements towards an appearing stimulus by directing gaze to the opposite location), there were no group differences in response time onset or duration; however, girls with Turner syndrome failed to suppress
Dev Disabil Res Rev


responses (i.e., they made more premature saccades); and made more directional errors. Thus not all spatial tasks lead to slower response times in girls with Turner syndrome. Attempts to further delineate the spatial deficits in Turner syndrome drove several researchers to develop and use experimental tasks, such as those believed to differentiate posterior parietal pathways that underlie object location skills from inferior temporal pathways presumed to underlie object identification skills. Buchanan et al. [1998] administered computerized same or different (object, or location) judgment tasks to assess these skills in adolescents with Turner syndrome. Rather than showing differential response profiles, the tasks revealed that girls with Turner syndrome were slower and less accurate than their peers on both spatial tasks. In a later study, we also found that girls with Turner syndrome were less accurate on object perception, object identification, and object location, but were as accurate at their peers at identifying the gestalt (versus the internal features) of an array of geometric shapes [Mazzocco et al., 2006]. Thus, this global impairment in visuospatial processing [Buchanan et al., 1998; p. 361] is nonetheless limited, at least within the object location domain. The potential influences that spatial skills have on math range from approximate number estimation to deciphering outcomes of mathematical functions. For young children, nonsymbolic magnitude judgments and number line skills may be influenced by spatial skills, and among school age children, spatially mediated difficulties may include alignment errors in vertically presented math calculations, or other place value confusion. It is difficult to predict how the spatial deficits in Turner syndrome will affect mathematical learning and function, because the deficits reported lack clear delineation despite compelling evidence that spatial deficits are a component of the phenotype. An additional complication stems from the overlapping demands of visuospatial and executive function processes, including working memory, on spatial tasks such as mental rotation. Executive functions are considered briefly in the section that follows. Executive function skills in females with Turner syndrome Executive function skills include a wide range of intentional, goal-oriented tasks such as planning, self regulation


and response inhibition, sustained attention, mental flexibility, and working memory. Evidence for selective executive dysfunction among girls with Turner syndrome is drawn not only from studies focused specifically on executive functions in this population, but also from studies of verbal and visual skills (such as those summarized earlier). For instance, in Temples [2002] oral fluency study (discussed previously), girls with Turner syndrome and their peers produced narratives of comparable lengths when asked to describe planning a party; but, when asked to describe what happened yesterday, girls with Turner syndrome produced significantly shorter narratives than their peers. Temple concluded that this finding exemplifies that girls with Turner syndrome are as competent as their peers on tasks that involve planning (such as the Tower of London, a classic planning task) despite difficulties in other executive functions (like the Stroop Test, which requires response inhibition and cognitive flexibility, but no planning) [Temple et al., 1996]; and also show deficits in episodic memory. When analyzing the content of either their narratives or verbal fluency responses, Temple [2002] found that girls with Turner syndrome produced a higher rate of low frequency words, relative to their peers and relative to their own rate of high frequency words. Such findings suggest differences in lexical storage, in the effects of word frequency on retrieval thresholds, or on search and retrieval functions. In addition, girls with Turner syndrome showed less switching across categories during oral fluency, a profile Temple described as more rigid and inflexible. As was the case for select verbal and visual spatial tasks, working memory tasks reveal selective impairments in Turner syndrome, although response times are slow on most of these tasks. The Stroop and the Contingency Naming Test (CNT) both reveal this pattern [Temple et al., 1996; Kirk et al., 2005], on baseline and experimental trials. For example, the CNT is a naming task that involves maintaining and operating on information during a task that requires a response decision (i.e., whether to name a stimulus by its shape, or by its color), while inhibiting a competing response. Across three trials, naming rules increase in complexity (and thus working memory load increases), from no working memory load and no response competition (name only colors; or name only shapes), to operating on one rule during naming (i.e., name 37

MLD

in

Turner Syndrome

Mazzocco

shapes if condition A is met, else name colors), to maintaining two rules during naming (name shapes if condition A is met, else name colors; however, reverse this rule for stimuli that appear underneath an arrow). Kirk et al. [2005] showed that third graders with Turner syndrome had less efficient performance than their peers on all three CNT trials, with respect to their speed/accuracy trade off, despite the fact that the first trial did not involve working memory demands (according to conventional definitions of working memory). However, the poorer performance of girls with Turner syndrome was driven primarily by longer response times on the warm up trial, mostly by errors on the two-rule trial, and by the ratio of both measures on the one-rule trial. From first to seventh grades, girls with Turner syndrome show slower rates of growth on overall CNT performance efficiency, which takes into account the combination of speed and accuracy [Mazzocco and Murphy, manuscript in preparation]. Slower performance has also been reported on the Stroop baseline trial (that involves naming colors) [Temple et al., 1996], and on RAN colors at primary school [Mazzocco, 2001], late elementary school [Murphy and Mazzocco, 2008], and middle school [Mazzocco and Murphy, manuscript in preparation], despite the lack of slower RAN performance on letter and number subtests (discussed previously). Considered together, these findings reflect differential degrees to which relatively simple taskslike RAN colorsappear to tax working memory in girls with Turner syndrome [Murphy and Mazzocco, 2008]. This, in combination with additional evidence for difficulty inhibiting a prepotent response set [Tamm et al., 2003], may be particularly relevant to arithmetic skills. Within the construct of working memory, Rovet and coworkers examined whether visual and verbal working memory demands differentially affected visuospatial performance, among girls with Turner syndrome [Buchanan et al., 1998]. In addition to the tasks described from their study of object location versus object identification judgments, they administered tasks to assess location and identification performance following a delay. Performance was not compromised when a delay involved a verbal working memory interference task; but performance accuracy did suffer when the delay task involved visual working memory. These differences were not evident in the peer comparison group. 38

On the basis of these and other findings [e.g., LaHood and Bacon, 1985], Rovet suggested that visual working memory is a hallmark deficit of the Turner syndrome phenotype, rather than visual perception or visual memory. Several additional studies have supported her claim. Haberecht et al. [2001] administered two visual working memory tasks that differed in complexity. In their study, girls with or without Turner syndrome were asked to indicate whether a target stimulus on a computer screen was in the same or different location as a stimulus that had either appeared immediately prior to, or two steps prior to, the appearance of the target stimulus. Girls with Turner syndrome were less accurate (and slower) than their peers on both the one-back and two-back tasks; but their performance on the more challenging task was no worse than their performance on the less challenging task. Note that this is a pattern that we observed on the CNTamong 7th graders with Turner syndrome, so such findings may not be limited to visual working memory tasks [Mazzocco and Murphy, manuscript in preparation]. It remains unclear what role these visual working memory and other executive skills play in mathematical competence. There is empirical support for the relationship between these skills in mathematical learning, and between spatial working memory and mathematical ability [e.g., Bull and Scerif, 2001; Espy et al., 2004; Bull et al., 2008]. Working memory also appears to mediate the effects of cognitive supporting skills underlying mathematical difficulties in children with MLD [Geary et al., 2007]. The ability to attend to, recall, and operate on instructions is essential for mathematics even in the early years of schooling, such as may be involved in comprehending, recalling, and following the steps involved in regrouping algorithms during multi-digit addition or subtracting. What pathways to MLD are implicated by the Turner syndrome cognitive phenotype? Inconsistencies such as those discussed within each of the three domains described earlier have been reported since the onset of Turner syndrome research. They partially account for the opposing views as to whether the Turner syndrome phenotype represents a specific right hemisphere deficit or, as Waber [1979] suggested, involvement of both hemispheres as well as parietal and frontal areas (p. 65).
Dev Disabil Res Rev


Consistent with the behavioral data presented thus far, findings from neuroanatomical, functional imaging, and ocular motor studies of Turner syndrome are mixed with regard to evidence of neuropathway involvement. For instance, the aforementioned increased latency on visually guided saccades implicates right posterior parietal cortex involvement [Law et al., 1997], whereas failure to inhibit reflective saccades during the antisaccades task implications dorsolateral prefrontal cortex involvement [Pierrot-Deseilligny et al., 1991]. Pathways involved in memory guided saccades also involve dorsal lateral prefrontal cortex, in addition to the parietal and frontal eye fields and multiple subcortical regions [as reviewed by Lasker et al., 2007]. Finding from neuroanatomical studies have been subtle, but statistically significant. For instance, Reiss et al. [1995] found group differences in the distribution of grey and white matter, such that girls with Turner syndrome had a bilateral decrease in proportions of both gray and white matter in parietal regions, and increased proportions of gray and white matter in the right inferior parietal-occipital region. Yet Murphy et al. [1993] reported decreased volume in right parietal-occipital regions, and Clark [1990] reported increased glucose metabolism in parietal and occipital regions, bilaterally. Despite behavioral deficiencies in spatial location tasks, functional imaging studies of performance on a visual working memory task, by girls with Turner syndrome [Haberechts et al., 2001], failed to reveal any activation deficits in the superior parietal region that is believed to be involved with spatial location. During this study, bilateral activation deficits were observed during the aforementioned two-back task, but in the dorsolatoral prefrontal, inferior parietal cortex and caudate regions. The authors of this study concluded that their results provide evidence for impairments in executive as well as storage/retrieval operations underlying higher-level cognition. Considered together, these studies provide sufficient reason to suspect that mathematics may be an area of difficulty for girls with Turner syndrome. Poor mathematics achievement as a component of the Turner syndrome phenotype Several researchers have demonstrated poor mathematics performance in girls with Turner syndrome [e.g., Siegel et al., 1998]. Rovet [1993] found
MLD
in

Turner Syndrome

Mazzocco

that 55% of the 7- to 16-year-old girls with Turner syndrome who participated in her study met criteria for MLD, versus 7% of her comparison group. We reported similar rates among primary school age children, among whom 43% of girls with Turner syndrome and 10% of girls without Turner syndrome met criteria for MLD [Mazzocco, 2001]. From our longitudinal research, we found that by the end of primary school over 75% of girls with Turner syndrome met criteria for MLD [Murphy et al., 2006]; and that girls with Turner syndrome who meet criteria for MLD are more likely to continue to do so than to catch up with their peers [Murphy et al., 2006]. Among those girls with Turner syndrome who have MLD, the severity of the learning difficulties can be significant. During the school age years, math achievement is, on average, two grades below age expectations [Rovet et al., 1994]. Girls with Turner syndrome are underserved by special education, perhaps because of apparent strengths in intellectual and language skills [Rovet et al., 2004]. Thus, there is sufficient evidence that mathematical learning difficulties among girls with Turner syndrome are prevalent, persistent, and significant. It is therefore not surprising that current recommendations for management of Turner syndrome include obtaining a comprehensive psycho-educational evaluation . . . immediately preceding school entry . . .; [Bondy and The Turner Syndrome Consensus Study Group, 2007]. To provide appropriate mathematical support for those girls with Turner syndrome who have MLD, we must first identify the nature of their MLD, whether it is a deficiency or a delay, and what relative strengths may support learning mathematics. Such efforts will also contribute to ongoing debates concerning the nature of MLD and its underlying pathways. DOES THE TURNER SYNDROME PHENOTYPE ILLUSTRATE A SPECIFIC PATHWAY TO MLD? There are at least two competing approaches to identifying cognitive characteristics of children with MLD, both of which have implications for potential pathways to MLD. The first approach rests on the assumption that MLD results from deficits in one or more supporting cognitive skills [Geary, 1993; Geary and Hoard, 2005]. The alternative approach assumes that basic numeracy skills underlie MLD [e.g., Landerl et al., 2004; Butterworth and
Dev Disabil Res Rev


Reigosa Cresp, 2007]. In reality, it is likely that these approaches combined will shed the most light on subtypes or variants of MLD, and on individual differences in correlates to mathematical competence and achievement. Thus, each is considered below. Both approaches are informed by research on the neuroanatomical correlates of, and neurocircuitry underlying, specific subcomponents of mathematical skills [e.g., Dehaene et al., 1999]. Therefore, imaging studies relevant to this review are also briefly summarized in the discussion that follows. Which MLD Subtype Characterizes Turner Syndrome? In his seminal paper on MLD, Geary [1993] proposed three subtypes, each related to proposed neurodevelopmental pathways. Briefly summarized from his more recent work [Geary and

tity and number line [Dehaene et al., 1999; Siegler and Opfer, 2003], hence, with basic numeracy skills. Others have expanded upon the Spatial subtype notion as involving spatial attention [e.g., Simon et al., 2008]. Finally, the Procedural subtype reflects a delay rather than deficit trajectory, and is marked by immature use of strategies and weak conceptual understanding of the procedures used in mathematics. Among children with procedural math deficits, executive function skills such as working memory are related to procedural accuracy [Geary et al., 2004]. Each of these three subtypesSemantic, Spatial, and Proceduralis considered below as a possible model of the MLD observed in girls with Turner syndrome. Turner Syndrome: A Model of an MLD Subtype? Is Turner syndrome an etiology of the Semantic Memory subtype of MLD? Of the three MLD subtypes summarized earlier, the Semantic Memory subtype may be the least likely to be linked with Turner syndrome, in view of the age appropriate reading and decoding skills reported for this population [Rovet, 1993; Mazzocco, 2001]. Indeed, in their study of school age girls with Turner syndrome, Temple and Marriott [1998] found that girls with Turner syndrome performed as well as their peers at reading and writing number words; reading, writing, and copying numerals; and at accurately responding to addition math facts. Murphy et al. [2006] reported similar strengths in primary school age girls with Turner syndrome, who showed accurate rote counting, set enumeration, reading and writing of numerals, and magnitude judgments, relative to their peers. Bruandet et al. [2004] also found accurate reading and writing of numbers in adults with Turner syndrome. However, slow response times were frequently observed in these studies [Temple and Marriott, 1998; Murphy and Mazzocco, 2008]. Simon et al. [2008] also reported accurate (but slow) enumeration and magnitude judgment by adolescents with Turner syndrome. Thus slow math fact retrieval by girls with Turner syndrome does map onto this subtype. Although slowed retrieval performance is consistent with the Semantic Memory subtype, accuracy on fact retrieval is not. This profile, coupled with the absence of widespread language-based deficits in this group and 39

. . . mathematical learning difficulties among girls with Turner syndrome are prevalent, persistent, and significant.

Hoard, 2005], these include Semantic Memory, Procedural, and Spatial subtypes of MLD. The Semantic Memory subtype, which co-occurs with reading disability, is characterized by poor fact retrieval performance, specifically slow and often inaccurate responses to arithmetic facts. There is an ongoing debate regarding whether these fact retrieval deficits are linked to a deficit in language-related skills (such as access to linguistic information stored in longterm memory) [Geary et al., 2004], or response competition from irrelevant associates (such as 6 as a possible solution for the problem, 3 3 3). Geary describes this subtype as genetically linked and persistent over time. Geary [1993] initially proposed that the Spatial subtype is revealed by overt errors in alignment and failure to master place value concepts tied to relative positions of numerals; more recently, researchers have emphasized the spatial components involved in the representation of quan

MLD

in

Turner Syndrome

Mazzocco

the evidence that the slowed response time is unlikely related to deficient lexical access, fails to support the notion of Turner syndrome as a model for Semantic Memory subtype of MLD. However, the slower response time during addition is noteworthy, in view of evidence that response speed is not slowed for all timed tasks measured in this population (discussed later). Is Turner syndrome an etiology of the visual-spatial subtype of MLD? The cooccurrence of visual spatial and mathematics difficulties reported in early phenotype studies motivated the earliest explorations of cognitive skills leading to math deficiencies in girls with Turner syndrome [Rovet et al., 1994; Mazzocco, 1998; Temple and Marriott, 1998]. These initial studies failed to reveal a correlation between math and visual spatial skills, leading Temple and Marriott [1998] to conclude that arithmetical difficulty in Turner syndrome is not simply a consequence of a spatial deficit. Like these earlier researchers, we also have not found consistent correlations between various math scores (standardized scores of math achievement, math ability, and counting skills) and visual spatial reasoning scores in either young [Mazzocco et al., 2006] or older girls with Turner syndrome [Mazzocco, 1998]; still, we have reported positive correlations between visual perceptual and counting skills among third graders with Turner syndrome that were not evident in a peer comparison group. We have also failed to find consistent, selective impairment from subsets of math items that have a strong visual spatial component [Mazzocco, 2001]. For instance, although we did find that Geometry subtest scores on the KeyMath-Revised were significantly lower among young girls with Turner syndrome relative to their peers, this subtest does not differentiate older girls with versus without Turner syndrome from each other. At third grade, girls with Turner syndrome were as capable as their peers at determining space-relationships, such as by rank ordering items according to length or size [Murphy et al., 2006]. Nevertheless, we have reported that adolescents with Turner syndrome have a higher frequency of alignment errors on paper and pencil calculation tasks (such as 14 26 5 166), relative to age-matched peers [Mazzocco, 1998]. In an error analysis study, we found that close to half of the girls with Turner syndrome 40

made alignment (48%) or operation errors (57%), in contrast with girls in a comparison group (of whom 14% and 19% made such errors, respectively). (The comparison group in our study was another group of children also at risk for poor mathematical achievement [i.e., girls with fragile X syndrome; see Murphy, this issue].) Note that although these errors could be conceptualized as spatial, they could also reflect weak executive function skills; that is, poor organization or inadequate tracking of the steps needed during complex problem solving (with complex defined as involving more than one step, versus single digit arithmetic with solutions less than 10). Before we dismiss the possibility that MLD in Turner syndrome has a spatial component, it is worth considering that the failure to find support may be an artifact of the components of mathematics, or of visual spatial functioning, targeted for investigation. Both are multifaceted constructs that cannot be generalized by broad standardized assessment tools often used in early research. Simon et al. [2008] recently examined spatial-attention and numeric skills, which he referred to as start state (basic cognitive underpinnings) rather than end state level skills (such as achievement test outcomes). In a series of studies, they asked school-age girls (ages 714 years) with Turner syndrome to enumerate sets of one to eight items by quickly calling out responses to visually presented arrays; to judge relative magnitude of two pictorially or nonsymbolically represented quantities; and to perform a simple motor task used to assess motor response times. The researchers found that girls with Turner were slower (but as accurate) as their peers on the enumeration and on the magnitude comparison task. On both the symbolic (numeric) and pictorial (stacked blocks) versions of the magnitude judgment, girls with Turner syndrome showed a comparable distance effect by responding more quickly when the difference of the two quantities being compared was larger versus smaller, a finding consistent with Dehaenes research on adults with Turner syndrome [Molko et al., 2003; Bruandet et al., 2004]. Plateaus for the distance effect occurred for different (larger) values for the Turner syndrome group, relative to peers, particularly for the blocks version of the task, which Simon describes as most dependent on visuospatial attention and spatial relational representations and processes
Dev Disabil Res Rev


(p. 90). Simon and his colleagues concluded that impaired spatial/attentional skills underlie numerical deficits in girls with Turner syndrome. (Note that Simon found comparable deficits in children with 22q deletion syndrome [see De Shmedt et al., this issue].) What mechanism could underlie these complex relationships? Molko reported that women with Turner syndrome showed atypical activation in the intraparietal sulcus, a region involved in number representation, during a problem verification task. Unlike their adult peers, women with Turner syndrome failed to show an increase in activation in this region as number size increased (e.g., 3 2, versus 5 6). However, it was unclear whether this finding was specific to arithmetic, because a similar effect was noted for a letter detection task. Indeed, the authors concluded that the neuroanatomical and functional findings could reflect disruption of spatial, attentional, and/or working memory (p. 855). Likewise, Haberechtss imaging research with the one- and two-back visual working memory task (discussed previously) revealed that frontal-striatal and frontal-parietal circuitry seemed to underlie the performance difficulties observed in girls with Turner syndrome, suggestive of executive dysfunction. When Kesler et al. [2005] asked girls with Turner syndrome to complete two- and three-operand arithmetic problems (addition and/or subtraction), girls with Turner syndrome recruited more neural resources than their peers on the two-operand task, and fewer resources than their peers on the three-operand task, including from frontal and parietal regions. The authors concluded that, unlike controls who used more of the same mechanisms for the three- versus two-operand problems, girls with Turner syndrome used different strategies on these two tasks. Greater temporal lobe activation during the two-operand tasks suggested the recruitment of verbal skills, which would reflect an appropriate compensatory mechanism for this population characterized by verbal strengths. Yet, when the problem became more complex, different strategies may have been implemented, and those strategies involved less activation in posterior visuospatial networks on which the comparison group drew from to solve three-operand problems. Considered together, findings such as these pose a challenge for determining whether spatial or executive skills underlie MLD in girls with Turner syndrome, thus
MLD
in

Turner Syndrome

Mazzocco

whether Turner syndrome is a model of the Spatial or Procedural MLD subtype. Is MLD in Turner syndrome an example of Procedural MLD? Geary has proposed that the Procedural subtype of MLD may be linked to prefrontal dysfunction, particularly when the children with this subtest manifest difficulty on complex math procedures. This is aligned with the Turner syndrome phenotype; because studies that reveal weak calculation skills in Turner syndrome (relative to their peers) further show that problems leading to calculation errors are often limited to complex (multi-digit) problems rather than single digit arithmetic. Another primary feature of this subtype is the use of immature strategies, and procedural deficits believed linked to poor working memory and other executive skills. These characteristics are aligned with the Turner syndrome phenotype, and may account for the alignment and operation errors observed during math calculations, or even the procedural rigidness observed among girls whose calculations may be accurate but completed very slowly (e.g., see Figs. 1 and 2). Consistent with this view, Rovet et al. [1994] proposed that poor executive functions account for the greater frequency of girls with Turner syndrome neglecting to apply or complete correct regrouping procedures during paper and pencil calculations, relative to age, grade, and verbal IQ matched peers. However, a hallmark feature of the Procedural MLD subtype is delayed acquisition of mathematics, rather than persistent challenges over time. Through longitudinal studies we have demonstrated that poor mathematics achievement in girls with Turner syndrome is evident by kindergarten [Mazzocco, 2001], persists through primary school [Mazzocco et al., 2006; Murphy et al., 2006], continues into fifth grade [Murphy and Mazzocco, 2008], and through middle school [Mazzocco and Murphy, manuscript in preparation). Cross-sectional studies reveal that math is an area of difficulty for adolescents [Mazzocco, 1998; Kesler et al., 2006] and women with Turner syndrome [Molko et al., 2003; Bruandet et al., 2004]. These findings counter the notion of math delays that will disappear over time. Persistence over time may be linked to ongoing executive function or visual working memory deficits reported earlier. Alternatively, numerical representations have been
Dev Disabil Res Rev


Fig. 1. An example of the strategies used to solve a multiplication problem. A 20-yearold girl with Turner syndrome completed the problem correctly, and accurately, but commented, I know I dont have to put all these zeros here, but I need to do it that way. Her response time was 58 sec.

Fig. 2. Example of the strategies used to solve simple arithmetic word problems. A 13-year-old girl with Turner syndrome used tally marks to determine her response to the question If you have 10 rows of chairs, and five chairs in each row, how many chairs will there be altogether? Her response was accurate; her response time was over 1 min.

implicated as an underpinning of MLD in Turner syndrome. Number Sense and Number Knowledge in Turner Syndrome Deficits in number processing are a proposed core deficit of MLD, leading Molko and colleagues to evaluated numeric representation systems among women with Turner syndrome. They contrasted performance on a semantic system involved in the representation of exact quantities (true of false: 7 5 5 12) that is associated with rote solution retrieval; with performance on a nonverbal system involved in representation of approximate quantities (is the solution


to 7 5 closer to 10 or to 20?). Although these tasks activate different brain regions in typical adults, there was no such distinction on functional imaging findings between tasks among women with Turner syndrome. While these and other group differences led Molko to propose a genetically-driven neuroanatomical and functional disorganization in Turner syndrome, at least as is relevant to mathematical cognition, Simon et al. [2008] disagree with this conclusion in part because of their findings of similar behavioral outcomes in disorders with very different genetic etiologies (e.g., Turner syndrome and 22 q deletion syndrome). This is an interesting counter, because we could also expect shared outcomes to emerge from different etiologies whether based on overlapping or slightly variant pathways to MLD, and/or influences on different mediating skills that lead to comparable broad outcomes. Indeed, one goal of phenotype research to not only characterize the disorder in question, but to also develop pathway models that may be shared with others without the disorder. In other words, Turner syndrome may be a model of MLD that is based on interruption in development or function of a number sense, (or of a Spatial or Procedural MLD subtype) that applies to a subset of individuals with MLD from the general population. To what does this number sense refer? Molkos use of the term number sense differs from the many referents to emerge from other fields related to MLD research, ranging from experimental psychology to mathematics education. Number sense may refer to intuitive basic functions to formal knowledge of number properties tied to instructional opportunities; and as such it is not a unitary construct [for an indepth discussion, see Berch, 2005]. The Turner syndrome phenotype may provide a compelling argument as to why this is the case. First, consider that girls with Turner syndrome have a robust knowledge of numbers. During primary school, they show age appropriate comprehension of numbers and numerals; they are as accurate as their peers in reading and writing numerals, making magnitude judgments (e.g., determining which of two numbers is larger, such as 9 or 7), making number line judgments (e.g., determining which number is closer to 7, 4, or 9?), and solving one digit addition problems (e.g., 2 3) with or without the use of manipulatives [Murphy et al., 2006]. Older school age girls with Turner syndrome 41

MLD

in

Turner Syndrome

Mazzocco

show similar strengths [Temple and Sherwood, 2002], and demonstrate a classic distance effect (an increased rate of arithmetic errors as the size of the addends within a problem increases) on problem verification and other numeric tasks. The absence of a number sense could diminish distance effects, but the magnitude of this effect for girls with Turner syndrome parallels that observed in typically achieving children [Simon et al., 2008]. Like their younger counterparts, adults with Turner syndrome also show intact number comprehension and processing skills, including counting, magnitude judgments, numeric interval bisection, rapid (exact) enumeration, and a classic distance effect [Bruandet et al., 2004]. Considered together, there is much evidence to suggest that Turner syndrome does not disrupt the number sense, in the broad sense of the term. Unfortunately, the conclusion of an intact number sense in Turner syndrome is refuted by other findings. Bruandet et al. [2004] found that adults with Turner syndrome had poor performance on cognitive estimation of quantities, such as estimating the length of a bus. Responses, which were always in the direction of underestimation, revealed an illogical sense of number (e.g., 4 m for an average length of a bus; p. 292). It is noteworthy that performance on this task seems tied to neither spatial nor executive dysfunction, the two domains for which math correlates have been previously implicated for Turner syndrome; instead, the performance of these women suggested a semantic challenge, a finding in conflict with other behavioral data (as described earlier). Specifically, it appeared that women with Turner syndrome relied on verbal mediation as a strategy on tasks that are typically solved by rote retrieval (exact arithmetic of small numbers). Thus, whereas the strong verbal memory, lexical storage, and lexical access that characterize Turner syndrome would suggest that exact arithmetic would be a strength, slow performance on these tasks suggest that reliance on these strengths occurs as a compensation for not being able to rely on intact number sense. We recently reported additional evidence that exact calculations are challenging for girls with Turner syndrome [Mazzocco and Murphy, manuscript in preparation]. We administered a three-part timed test requiring that 6th graders rapidly identify whether each of 49 pairs of numbers summed to 42

a target number that appeared on the top of a stimulus page. This required rapid composing or decomposing of numbers, and identifying only those pairs that met the criterion. The first part of this test was relatively challenging in that all foils were close to the target solution (e.g., rather than equaling 19, the foils summed to solutions like 17, or 21). The second task was easier, because all foils were distant foils (e.g., they equaled sums like 5, or 29). The third and final task introduced an interference task that increased working memory load, by requiring simultaneous completion of another numerical task (identifying pairs with sums that exceeded a second (larger) target number). Children were permitted only one minute to complete each task. Girls with Turner syndrome were slower than their peers on all three tasks, but more telling was their within-group performance profile: the typically developing 6th graders were slowest, and least accurate, on the third trial that involved an interference task; and they performed most quickly, and most accurately, on the second trial with the distant foils. If girls with Turner syndrome lacked a number sense entirely, there should be no difference in their performance on the first two trials. However, like their peers, their performance was faster and more accurate on the second trial than on the first and third trials (consistent with the distance effect they manifested, as described earlier). Yet the first and third trials were equally as challenging for girls with Turner syndrome. Quick calculation of exact solutions was not a task that appeared easy for them, as was the case for their peers. This finding is consistent with Molkos description that females with Turner syndrome are relying on different (more time consuming) strategies on an otherwise rote task. What other basic number tasks are impaired in Turner syndrome? Subitizing is one of the basic cognitive processes of interest to researchers of mathematical cognition, and thus it is not surprising that it has been examined in females with Turner syndrome. Subitizing refers to the automatic recognition (versus counting) of very small quantities. In the absence of counting, typically developing children and adults have comparable response times when identifying 1, 2, or 3 items. In contrast, Bruandet reported that women with Turner syndrome demonstrated earlier counting than their peers (as indicated by incrementally longer response times
Dev Disabil Res Rev


to identifying 3 versus 2 items), suggesting impairment in the automaticity with which subitizing judgments are made [Bruandet et al., 2004]. This counting-versus-subitizing profile is similar to the findings for children with MLD from the general population [Koontz and Berch, 1996]. An aberrant subitizing profile suggests subtle differences in very basic (start point, to use Simons terminology) number processing. However, Simon et al. [2008] did not find a subitizing deficit in his study of adolescents with Turner syndrome. Thus it is unclear to whether subitizing deficits reveal deficits in mathematical cognition in females with Turner syndrome. Are these reported weaknesses demonstrative of weak number knowledge, or number sense, or both? As difficult as it is to define [Berch, 2005], number sense can be contrasted with a more formal understanding of number, or number knowledge, that is advanced with instruction, such as understanding number line concepts, aligning quantities with numerals, and carrying out algorithms for addition and subtraction. Although these terms are often used interchangeably, and may have significant overlap, the boundaries between the two constructs are fuzzy. The mathematical performance of females with Turner syndrome may demonstrate the importance of differentiating between these two constructs. Do Girls with Turner Syndrome Have MLD? Finally, there is a fundamental challenge to the notion that Turner syndrome is a model of MLD. One could dispute this notion on the basis of repeated studies showing accurate calculation skills, indeed, age appropriate performance on mathematical achievement testson untimed tests [e.g., Rovet et al., 1994; Mazzocco, 1998; Mazzocco et al., 2006; Murphy et al., 2006]. Therefore, is MLD even an issue for girls with Turner syndrome? In view of the challenges that do existslower response times, select deficits in calculation strategy, and deficiencies on tasks that are relatively easy for their peers, the Turner syndrome cognitive phenotype reminds us why we cannot rule out MLD solely on the basis of an untimed calculations subtest. It therefore poses challenges to the widespread practice in which many of us researchers engage, of defining MLD on the basis of broad mathematics achievement test outcomes.
MLD
in

Turner Syndrome

Mazzocco

As a group, girls with Turner syndrome exemplify a subset of children who struggle with mathematics even when some of their achievement scores are age appropriate, and who may struggle on the most basic tasks that are easy for many other children. This notion is supported by our recent longitudinal work [Murphy and Mazzocco, 2008] showing that fifth graders with Turner syndrome were more likely to fail items designed for primary school children (e.g., counting principles) than were their age- and FSIQ-matched peers, despite age appropriate achievement scores. But that does not mean that all girls with Turner syndrome will have difficulty learning mathematics. In that 6 years study, we administered a test of counting principles annually, until mastery on the task was achieved. Nearly all of the children in our comparison group achieved mastery on this task before the age of 10 years, but 35% of fifth graders with Turner syndrome had not. Although this rate represents a significant increase over that from the comparison group, it is important to note that these group data do not apply to all girls with Turner syndrome, many of whom have appropriate achievement in mathematics. The summaries described in this review apply to the many girls with Turner syndrome who have difficulties learning mathematics. How might MLD, in girls with Turner syndrome, appear in the classroom? There is very little research on this topic. It is possible that longer response times on mathematics tasks, which are consistently reported [e.g., Rovet et al., 1994; Temple and Sherwood, 2002; Molko et al., 2003; Bruandet et al., 2004; Murphy and Mazzocco, 2008], result in fewer problems attempted relative to peers, which may influence learning (e.g., by limiting practice time and the opportunity for repetition), assessment accuracy, and interpretations of students efforts. Immature strategies add to response slowing, and when coupled with intact performance in other areas of number processing, this behavior may create an impression of lack of effort. Slower performance despite greater activation on easy tasks may diminish mental resources (and thus performance) for long-term seatwork or assessments. Knowledge of task difficulty may increase arousal, as suggested by Keysors preliminary findings of higher psychophysiological arousal during a mental arithmetic task in adolescents with Turner syndrome relative to their peers
Dev Disabil Res Rev


[Keysor et al., 2002]. However, Roberts et al. [2006] did not find additional evidence for this association in a later study. Although it is well supported that slower calculation skills in Turner syndrome are neither linked to generally slower processing speed, nor to slower lexical access or storage (as discussed previously), it is unclear whether slowed response time during mathematics drives, or is driven by, the use of immature strategies like those exemplified in Figures 1 and 2 [e.g., Kesler et al., 2006; Mazzocco et al., 2006] Indeed, one goal of phenotype research to not only characterize the disorder in question, but to also develop pathway models that may be shared with others without the disorder. It is also unclear whether slow response times reflect ba-

Regardless of the underlying mechanisms, the Turner syndrome phenotype illustrates the complexities involved in delineating subtypes or variants of MLD, and the importance of identifying potential pathways to MLD. n ACKNOWLEDGMENTS The author would like to thank the families who have participated in the Math Skills Development Projects longitudinal study.

REFERENCES
Alexander D, Ehrhardt AA, Money J. 1966. Defective figure drawing, geometric and human, in Turners syndrome. J Nerv Ment Dis 142:161167. Berch DB. 2005. Making sense of number sense: implications for children with mathematical disabilities. J Learn Disabil 38:289384. Bondy C and The Turner Syndrome Consensus Study Group. 2007. Guidelines for the care of girls and women with Turner syndrome. J Clin Endocrinol Metabol 92:1025. Bruandet M, Molko N, Cohen L, et al. 2004. A cognitive characterization of dyscalculia in Turner syndrome. Neuropsychologia 42: 288298. Buchanan L, Pavlovic J, Rovet J. 1998. A reexamination of the visuospatial deficit in Turner syndrome: contributions of working memory. Dev Neuropsychol 14:341367. Bull R, Espy KA, Wiebe SA. 2008. Short term memory, working memory, and executive functioning in preschoolers: longitudinal predictors of mathematical achievement at age 7 years. Dev Neuropsychol 33:205228. Bull R, Scerif G. 2001. Executive functioning as a predictor of childrens mathematics ability: inhibition, switching, and working memory. Dev Neuropsychol 19:273293. Butterworth B. 2005. Developmental dyscalculia. In: Campbell J, editor. Handbook of mathematical cognition. New York: Psychology Press. p 455467. Butterworth B, Reigosa Crespo V. 2007. Information processing deficits in dyscalculia. In: Berch DB, Mazzocco MMM, editors. Why is math so hard for some children? The nature and origins of mathematical learning difficulties and disabilities. Baltimore, MD: Paul H. Brookes Publishing Co. p 6581. Clark C, Klonoff H, Hayden M. 1990. Regional cerebral glucose metabolism in Turner syndrome. Can J Neurol Sci 17:140144. Cutter WJ, Daly EM, Robertson DMW, et al. 2005. Influence of X chromosome and hormones on human brain development: a magnetic resonance imaging and proton magnetic resonance spectroscopy study of Turner syndrome. Biol Psychiatry 59: 273283. Davenport M, Hooper S, Zeger M. 2007. Turner syndrome throughout childhood. In: Mazzocco MMM, Ross JL, editors. Neurogenetic developmental disorders: variation of manifestation, in childhood. Cambridge, MA: MIT Press. p 345. Dehaene S, Spelke E, Pinel P, et al. 1999. Sources of mathematical thinking: behavioral and brain-imaging evidence. Science 284: 970974.

As a group, girls with Turner syndrome exemplify a subset of children who struggle with mathematics even when some of their achievement scores are age appropriate . . . But that does not mean that all girls with Turner syndrome will have difficulty learning mathematics.

sic number sense deficits or potential visual spatial and working memory deficits. Each of these domains suggests a different potential area of educational support and intervention, and further research is needed to determine which are most effective. Additional research will shed light on the many factors not reviewed in this article, including the biological mechanisms potentially leading to mathematics difficulties in Turner syndrome. These include direct and indirect effects linked to karyotype [Ross et al., 2000], hormone deficiency and treatment [Ross et al., 2006], and imprinting [e.g., Cutter et al., 2005].


MLD

in

Turner Syndrome

Mazzocco

43

De Smedt, Swillen A, Verschaffel L, et al. 2009. Mathematical learning disabilities in children with 22q11.2 deletion syndrome. A review. Dev Disabil Res Rev 15:410. Espy KA, McDiarmid MM, Cwik MF, et al. 2004. The contribution of executive functions to emergent mathematical skills in preschool children. Dev Neuropsychol 26: 465486. Elliott TK, Watkins JM, Messa C, et al. 1996. Positron emission tomography and neuropsychological correlations in children with Turners syndrome. Dev Neuropsychol 12: 365386. Geary DC. 1993. Mathematical disabilities: cognitive, neuropsychological, and genetic components. Psychol Bull 114:345362. Geary DC, Hoard MK. 2005. Learning disabilities in arithmetic and mathematics: theoretical and empirical perspectives. In: Campbell J, editor. Handbook of mathematical cognition. New York: Psychology Press. p 253267. Geary DC, Hoard MK, Byrd-Craven J. 2004. Strategy choices in simple and complex addition: contributions of working memory and counting knowledge for children with mathematical disability. J Exp Child Psychol 88:121151. Geary DC, Hoard MK, Nugent L, et al. 2007. Strategy use, long term memory, and working memory capacity. In: Berch DB, Mazzocco MMM, editors. Why is math so hard for some children? The nature and origins of mathematical learning difficulties and disabilities. Baltimore, MD: Paul H. Brookes Publishing Co. p 83105. Haberecht MF, Menon V, Warsofsky IS, et al. 2001. Functional neuroanatomy of visuospatial working memory in Turner syndrome. Hum Brain Mapp 14:96107. Kesler SR, Menon V, Reiss AL. 2005. Neurofunctional differences associated with arithmetic processing in Turner syndrome. Cereb Cortex 16:849856. Keysor, CS, Mazzocc, MM, McLeod, DR, et al. 2002. Physiological arousal in females with fragile X or Turner syndrome. Dev Psychobiol 41:133146. Kirk JW, Mazzocco MM, Kover ST. 2005. Assessing executive dysfunction in girls with fragile X or Turner syndrome using the Contingency Naming Test (CNT). Dev Neuropsychol 28:755777. Koontz KL, Berch DB. 1996. Identifying simple numerical stimuli: processing inefficiencies exhibited by arithmetic learning disabled children. Math Cognit 2:123. LaHood BJ, Bacon GE. 1985. Cognitive abilities of adolescent Turners syndrome patients. J Adolesc Health Care 6:358364. Landerl K, Bevan A, Butterworth B. 2004. Developmental dyscalculia and basic numerical capacities: a study of 89 year old students. Cognition 92:99125. Lasker AG, Mazzocco MMM, Zee DS. 2007. Ocular motor indicators of executive dysfunction in fragile X and Turner syndromes. Brain Cogn 63:203220. Law I, Svarer C, Holm S, et al. 1997. The activation pattern in normal humans during suppression, imagination and performance of

saccadic eye movements. Acta Physiol Scand 161:419434. Mazzocco MMM. 1998. A process approach to describing mathematics difficulties in girls with Turner syndrome. Pediatrics 102(2 Pt 3):492496. Mazzocco MMM. 2001. Math learning disability and math LD subtypes: evidence from studies of Turner syndrome, fragile X syndrome, and neurofibromatosis type 1. J Learn Disabil 34:520533. Mazzocco MMM, Bhatia NS, Lesniak-Karpiak K. 2006. Visuospatial skills and their association with math performance in girls with fragile X or Turner syndrome. Child Neuropsychol 12:87110. Molko N, Cachia A, Riviere D, et al. 2003. Functional and structural alterations of the intraparietal sulcus in a developmental dyscalculia of genetic origin. Neuron 40:847 858. Murphy MM. 2009. A review of mathematical learning disabilities in children with fragile X syndrome. Dev Disabil Res Rev 15:2127. Murphy DG, Allen G, Haxby JV, et al. 1994. The effects of sex steroids and the X chromosome on female brain function: a study of neuropsychology of adult Turner syndrome. Neuropsychologia 32:1309 1323. Murphy DG, DeCarli C, Daly E, et al. 1993. Xchromosome effects on female brain: a magnetic resonance imaging study of Turners syndrome. Lancet 342:11971200. Murphy MM, Mazzocco MMM. 2008. Mathematics learning disabilities in girls with fragile X or Turner syndrome during late elementary school. J Learn Disabil 41:2946. Murphy MM, Mazzocco MM, Gerner G, et al. 2006. Mathematics learning disability in girls with Turner syndrome or fragile X syndrome. Brain Cogn 61:195210. Nielsen J, Wohlert M. 1991. Chromosome abnormalities found among 34,910 newborn children: results from a 13 year incidence study in Arhus, Denmark. Hum Genet 87:8183. Pennington BF, Heaton RK, Karzmark P, et al. 1985. The neuropsychological phenotype in Turner syndrome. Cortex 121:391404. Pierrot-Deseilligny C, Rivaud S, Gaymard B, et al. 1991. Cortical control of reflexive visually-guided saccades. Brain 114:14731485. Reiss AL, Mazzocco MM, Greenlaw R, et al. 1995. Neurodevelopmental effects of X monosomy: a volumetric imaging study. Ann Neurol 38:731738. Roberts J, Mazzocco MMM, Murphy MM, et al. 2008. Arousal modulation in fragile X or Turner Syndrome. J Autism Dev Dis 38:2027. Ross JL, Roeltgen D, Kushner H, et al.. 2000. The Turner syndrome-associated neurocognitive phenotype maps to distal Xp. Am J Hum Genet 67:672681. Ross JL, Roeltgen D, Zinn A. 2006. Cognition and the sex chromosomes: studies in Turner syndrome. Horm Res 65:4756. Rovet JF. 1993. The psychoeducational characteristics of children with Turner syndrome. J Learn Disabil 26:333341.

Rovet JF, Buchanan L. 1999. Turner syndrome: a cognitive neuroscience approach. In: TagerFlusberg H, editor. Neurodevelopmental disorders. Cambridge: The MIT Press. p 223249. Rovet J, Netley C. 1982. Processing deficits in Turners syndrome. Dev Psychol 18:7794. Rovet JF, Szekely C, Hockenberry MN. 1994. Specific arithmetic calculation deficits in children with Turner syndrome. J Clin Exp Neuropsychology 16:820839. S avendahl L, Davenport ML. 2000. Delayed diagnoses of Turners syndrome: proposed guidelines for change. J Pediat 137:455459. Shaffer JW. 1962. A specific cognitive deficit observed in gonadal aplasia (Turners syndrome). J Clin Psychol 18:403406. Shepard RN, Metzler J. 1971. Mental rotation of three-dimensional objects. Science 171:701 702. Siegel PT, Clopper R, Stabler B. 1998. The psychological consequences of Turner syndrome and review of the National Cooperative Growth Study Psychosocial Substudy. Pediatrics 102:488491. Siegler RS, Opfer JE. 2003. The development of numerical estimation: evidence for multiple representations of numerical quantity. Psychol Sci 14:237243. Silbert A, Wolff PH, Lilienthal J. 1997. Spatial and temporal processing in patients with Turner syndrome. Behav Genet 7:1121. Simon TJ, Takarae Y, DeBoer T, et al. 2008. Overlapping numerical cognitive impairments in children with chromosome 22q11.2 deletion or Turner syndromes. Neuropyschologia 46:8294. Tamm L, Menon V, Reiss AL. 2003. Abnormal prefrontal cortex function during response inhibition in Turner syndrome: functional magnetic resonance imaging evidence. Biol Psychiatry 53:107111. Temple CM. 2002. Oral fluency and narrative production in children with Turners syndrome. Neuropsychologia 40:14191427. Temple CM, Carney RA. 1993. Intellectual functioning of children with Turner syndrome: a comparison of behavioural phenotypes. Dev Med Child Neurol 35:691698. Temple CM, Carney RA. 1995. Patterns of spatial functioning in Turners syndrome. Cortex 31:109118. Temple CM, Carney RA. 1996. Reading skills in children with Turners syndrome: an analysis of hyperlexia. Cortex 32:335 345. Temple CM, Carney RA, Mullarkey S. 1996. Frontal lobe function and executive skills in children with Turners syndrome. Dev Neuropsychol 12:343363. Temple CM, Marriott AJ. 1998. Arithmetical ability and disability in Turners syndrome: a cognitive neuropsychological analysis. Dev Neuropsychol 14:4767. Temple CM, Sherwood S. 2002. Representation and retrieval of arithmetical facts: developmental difficulties. Q J Exp Psychol A: Hum Exp Psychol 55:733752. Waber DP. 1979. Neuropsychological aspects of Turners syndrome. Dev Med Child Neurol 21:5870.

44

Dev Disabil Res Rev

MLD

in

Turner Syndrome

Mazzocco