Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jul 1.
Published in final edited form as: Neuropsychology. 2022 Apr 14;36(5):419–432. doi: 10.1037/neu0000806

Academic Skills after Brain Injury: A Lifespan Perspective

Alyssa W Sullivan 1, Mark D Bowren 1, Joel Bruss 2, Daniel Tranel 1,2,3, Ö Ece Demir-Lira 2,3,4
PMCID: PMC9631230  NIHMSID: NIHMS1838059  PMID: 35420857

Abstract

Objectives:

This study investigated academic skills outcomes after brain injury and identified the influence of age and injury factors across the lifespan.

Methods:

Our sample included 651 participants with focal brain lesions. Math, reading, and spelling data from the Wide Range Achievement Test were used as the academic skills outcomes. Age of lesion onset ranged from 0 to 85 years old. Linear regressions were conducted to identify the relation between age and injury factors and academic skills outcomes. Lesion-symptom mapping was conducted to identify the brain areas that, when lesioned, were associated with deficits in academic skills.

Results:

A quadratic model of age of lesion onset significantly predicted math (R2 = .28, p < .001), reading (R2 = .29, p < .001), and spelling outcomes (R2 = .32, p <.001), while accounting for various covariates. Education, sex, lesion size and laterality, etiology, and seizure history were additional reliable predictors of academic skills outcomes across the lifespan. Academic skill deficits were associated with damage to various brain areas across the left-hemisphere frontal, temporal, and parietal lobes, the insular area, and left- and right-hemisphere white matter.

Conclusions:

This study supports age of lesion onset as a relevant predictor of academic skills after brain injury in a lifespan sample. Several other variables (e.g., education, sex, lesion characteristics, and seizure history) are notable in the prediction of outcomes across the lifespan. Future work could investigate more diverse samples and emphasize recruitment of early-onset injuries to examine generalizability and potential critical periods for academic skills.

Keywords: academic skills, brain lesion, Early vulnerability hypothesis

Academic success has implications for multiple aspects of life, ranging from job prospects to physical health (Reyna et al., 2009; Ritchie & Bates, 2013). Academic skills – in essence, the core abilities of math, reading, and spelling – also have an extended developmental trajectory, making these skills relevant across wide age ranges (Chall, 1996; Geary, 1994). Thus, when examining relevant factors for predicting academic skills outcomes, taking a lifespan perspective is informative given the extended developmental trajectory of academic skills and the known influence of academic skills on outcomes across both early life and adulthood. Despite their importance, academic skills outcomes after brain injury have been less thoroughly investigated than cognitive or socio-emotional outcomes. So, questions remain regarding factors that predict academic skills outcomes after brain injury, especially across the lifespan. Of note, the current study will focus on focal (i.e., circumscribed) damage to the brain rather than diffuse injury caused by other mechanisms (e.g., diffuse axonal injury, anoxia).

Factors Associated with Academic Skills Outcomes

Age of lesion onset (AgeL) is a relevant factor for the prediction of academic skills outcomes after brain injury. Though, the influence of AgeL on academic skills has primarily been investigated in samples with childhood-onset brain injury. For instance, Westmacott and colleagues (2018) found younger age at stroke was associated with decreased performance on math calculation and Max and colleagues (2010) found a small effect of AgeL for math, reading, and spelling, with earlier onset stroke (prenatally, or up to age one year) associated with poorer performance as compared to those who sustained strokes after age one year. These findings align with another study that indicated stroke before the age of one year is associated with worse performance on tests of math, reading, and spelling (das Dores Rodrigues et al., 2011). Finally, in another study, those with onset before the age of two (i.e., especially the pre or perinatal period) performed significantly lower on measures of spelling and arithmetic when compared to those who are injured after the age of seven years old (Anderson et al., 2009).

Overall, the current literature indicates there may be increased risk for academic skills deficits when brain injury occurs before the age of two-years-old (e.g., Anderson et al., 2009; das Dores Rodrigues et al., 2011; Max et al., 2010; Westmacott et al., 2018). Earlier onset conferring increased risk for poor outcomes could align with the Early Vulnerability Hypothesis (Hebb, 1949). Specifically, the Early Vulnerability Hypothesis explains brain injury occurring early in life may confer additional risk than brain injury occurring later in life (Hebb, 1949). Importantly, if brain injury occurs during early life, this could lead to a flawed recovery and damage skills that are known to develop during or after that period (e.g., Kolb, 1995; Luciana, 2003). Another important consideration is that of critical/sensitive periods, which refer to periods of importance for the development of particular skills. Injury onset before or during critical/sensitive periods for particular skills may confer additional risk for worse outcomes when compared to later brain injury after the skills have stabilized (Thomas & Johnson, 2008; for further discussion of critical periods’ relevance for neuropsychological outcomes, see Anderson et al., 2011). Importantly, critical periods have been identified for academic outcomes and cognitive domains early in life (e.g., Anderson et al., 2009; Crowe et al., 2012).

In considering the developmental periods of relevance for academic skills, the roots of the academic skills trace back to as early as preschool period, and fundamental literacy and mathematical skills are acquired during elementary school years (Chall, 1983; Geary, 1994). However, academic skills continue to get fine-tuned throughout adolescence and in early adulthood. For example, although children typically start acquiring literacy skills around 5–7 years of age, visual word processing becomes more efficient and automatic with age and reading experience. Even after adolescence and extending into early adulthood, changes within specialized visual word processing networks are observed (Brem et al., 2006). Vocabulary, especially acquisition of rare words that are more common in print, similarly extends well beyond K-12 years. In math, although basic, arithmetic skills are acquired during early school years, and individuals continue to gain competency in a range of areas, such as algebra or calculus throughout higher education. Neuroimaging studies also associate the protracted structural (e.g., changes in white matter tracts) and functional changes (e.g., accelerated neuronal processing) observed in early adulthood with academic skills (Qiu et al., 2008; Torre et al., 2020). Taken together, damage before or during these periods could confer increased risk for worse academic skills outcomes when compared to those who sustain injuries after the skills have stabilized in early adulthood.

Despite the extended developmental trajectories of academic skills, little work has identified the influence of AgeL on academic skills across both child- and adult-onset samples. Mosch and colleagues (2005) identified those with childhood-onset lesions had higher rates of impairments in academic skills when compared to adult-onset lesions, though these findings were descriptive in nature and quantitative analyses indicated similarity between children and adults who sustained strokes in similar locations for outcomes. While the identification of specific periods of risk is of value, a broader perspective of the contribution of AgeL on outcomes is also useful. For instance, a small body of past work has identified AgeL as a relevant variable of interest for the prediction of cognitive outcomes across the lifespan (e.g., Duval et al., 2008).

Overall, there is a need to identify the potential relationship of AgeL and academic skills outcomes across the lifespan, while accounting for variables known to potentially contribute to outcomes. For instance, demographic (e.g., education, sex), injury (e.g., etiology, lesion location, lesion size, lesion laterality, history of seizures), and age (e.g., age of lesion onset, time since lesion onset, age of assessment) factors have all been identified as relevant in the prediction of outcomes after brain injury (see Anderson et al., 2011 and Fuentes et al., 2016 for reviews). However, few studies have jointly considered these potentially relevant factors while examining academic outcomes (e.g., Anderson et al., 2009; Westmacott et al., 2018) and no studies to our knowledge have considered these factors while examining academic skills in a lifespan sample. Overall, a lifespan approach would expand the literature in a meaningful way by contributing to a more comprehensive understanding of the influence of AgeL and other relevant factors on academic skills.

It is important not only to identify the influence of age, injury, and demographic variables in the sample of brain-injured patients, but to also contextualize these findings by comparing the sample’s outcomes with normative expectations. In other words, does brain injury pose a risk for decreased academic skills when compared to what would be expected? Some work has shown that early brain injury is a risk factor for deficits in academic skills (e.g., Aram & Ekelman, 1988; Ballantyne et al., 2008; Champigny et al., 2020; Deotto et al., 2019; Jacomb et al., 2018). For instance, children with early brain injury (i.e., pre-birth to late childhood) tend to perform significantly lower on measures of academic skills when compared to normative populations (Anderson et al., 2009; Champigny et al., 2020). However, these differences are often small (i.e., one standard deviation; Anderson et al., 2009). Other work indicates children with pre- and perinatal brain injury can compensate after lesions that typically disrupt reading and mathematical functioning in adults, developing numerical abilities in the low normal to normal range and exhibiting normal academic achievements (Ballantyne et al., 2008; Glenn et al., 2018). These findings highlight a need to contextualize the potential impairments in brain-injured samples by examining how samples compare to normative expectations.

Another significant factor influencing academic skills after brain injury is lesion location. While neuroimaging methods utilized with typically developing populations can identify brain regions associated with behavioral outcomes, the lesion method is important for discovering brain regions necessary for skills. Past work utilizing both neuroimaging methods and the lesion method supports a wide brain network associated with academic skills (e.g., Peters & De Smedt, 2018). The brain areas correlated with math, reading, and spelling overlap but are also distinct. For math, neuroimaging studies have identified a wide, whole-brain network involved in number and calculation tasks for adults, including the inferior and superior frontal cortex, bilateral superior and inferior parietal cortex, bilateral thalamus, right insula, right claustrum, bilateral cerebellum, and occipitotemporal regions – with calculation recruiting a more widely distributed network than numerical cognition (Arsalidou & Taylor, 2011; Menon, 2015). Supporting neuroimaging studies, lesions to the left and right-side parietal lobe (e.g., left inferior parietal lobule; right superior parietal lobe), left occipito-temporal junction, left frontal lobe, right inferior prefrontal structures, and left subcortical structures result in deficits in calculation in adults (for review, see Denburg & Tranel, 2003). Several white matter tracts including the superior longitudinal fasciculus, inferior longitudinal fasciculus, posterior segment of the corpus callosum, corona radiata, and the corticospinal tract have also been associated with numerical and mathematical abilities (Matejko & Ansari, 2015).

Successful reading relies on a complex process of audiovisual integration which consists of accessing linguistic information from written forms. The left fusiform gyrus, inferior temporal gyrus, bilateral motor and superior temporal cortices, pre-supplementary motor area, inferior frontal gyrus, angular gyrus, and cerebellum are recruited in integration of print and speech sounds (e.g., Cohen et al., 2002; Dehaene, 2009; Mani et al., 2008; Turkeltaub et al., 2002). White matter pathways associated with reading skills include the superior longitudinal fasciculus, splenium, calcarine, occipital and temporal callosal fibers, corona radiata, dorsal fibers that connect the posterior occipito-temporal sulcus with the lateral cortical surface, and paraventricular white matter (Ben-Shachar et al., 2007; Binder & Mohr, 1992; Damasio & Damasio, 1983; Niogi & McCandliss, 2006). Specifically, left temporo-parietal white matter is relevant for reading abilities in adults (Klingberg et al., 2000), the dorsal temporo-parietal is involved in phonological processing and integrating visual and phonological information and the ventral occipito-temporal route is associated with recognition of visual forms (Younger et al., 2017).

Meta-analyses of neuroimaging studies investigating spelling have identified the left inferior and middle frontal gyri, left inferior temporal/fusiform gyri, left intraparietal sulcus, and right cerebellum as particularly relevant during spelling tasks (Planton et al., 2013; Purcell et al., 2011). Studies utilizing neuroimaging have additionally identified peripheral writing structures for spelling, including the left precentral gyrus and superior frontal gyrus and sulcus, left post central gyrus, left superior parietal lobule/intraparietal sulcus, left supplementary motor area, and cerebellum (e.g., Purcell et al., 2011). Lesion studies have identified several other structures as relevant for spelling in adults, including the perisylvian cortical regions, supramarginal gyrus, middle temporal gyrus, superior temporal gyrus, and left temporo-occipital regions (e.g., Philipose et al., 2007; Rapcsak & Beeson, 2002, 2004; Rapcsak et al., 2009).

Given the wide brain networks associated with academic skills, it is important that lesion studies utilize methods that allow for multiple brain regions to be implicated in academic performance after injury (Uddin et al., 2010). One useful method for leveraging lesion data to indicate multiple critical areas of relevance for a skill is lesion-symptom mapping. Advanced lesion-symptom mapping approaches allow for the identification of multiple regions (i.e., white and gray matter) that, when lesioned, are associated with lower performance in a particular domain. Importantly, voxel-based lesion-symptom mapping offers a more formal analysis of brain areas critical for a particular behavior when compared to other previously utilized lesion methods, such as the subtraction method, and use the same voxel-based procedures as functional neuroimaging data (e.g., Forkel, 2020). While original lesion-symptom mapping approaches utilized a mass-univariate approach (i.e., applying statistical methods on individual brain voxels), this approach has been shown to have several limitations, including an increased potential for false positives and variation in statistical values (e.g., Inoue et al., 2014; Mah et al., 2014; Nachev, 2015; Pustina et al., 2018; Sperber & Karnath, 2017). To address these limitations, more recent work has proposed utilizing sparse canonical correlation analysis (SCCAN), a multivariate method. In a study comparing a mass-univariate approach with SCCAN, SCCAN produced more accurate results than the mass-univariate method (Pustina et al., 2018). For further information on the benefits of the SCCAN approach versus mass-univariate method, see Pustina and colleagues (2018).

Importantly, lesion studies can offer supplemental information to neuroimaging studies, supporting a convergent methods approach to understanding the neural correlates of the behavior. Combining neuroimaging and lesion-symptom mapping studies works particularly well, given the different limitations to each method. For instance, neuroimaging data (e.g., fMRI) can provide compelling evidence that any observed changes in signal are due to the experimental task and identify activation in areas of the brain associated with behaviors. Lesion data, utilizing techniques such as lesion-symptom mapping, indicate that differences in behavior (e.g., deficits in academic skills) are due to the brain injury, identifying areas of the brain critical for a particular skill. As such lesion-based studies are supplementary to neuroimaging studies and can build on knowledge of relevant areas associated with skills and further identify those areas of the brain that are critical for a particular skill (Kimberg et al., 2007). As neuroimaging studies have been disproportionately represented in the literature, there is a need for lesion-based studies to support and expand upon the findings of neuroimaging studies (Fellows et al., 2005).

Regarding academic skills, lesion-symptom mapping studies with adult-onset lesions have identified deficits in arithmetic ability are associated with lesions in the left inferior parietal lobule, the inferior frontal cortex, and pre- and post-central gyri (Baldo & Dronkers, 2007), deficits in reading are associated with damage to a frontal network including the superior part of Broca’s area and the premotor cortex (Piras & Marangolo, 2009), and the left ventral occipito-temporal cortex (Baldo et al., 2018), and deficits in single-word writing are associated with the left supramarginal gyrus (Baldo et al., 2018). To our knowledge no lesion-symptom mapping studies have investigated math, reading, and spelling within the same sample. Importantly, examining all academic skills in the same sample allows for further investigation of overlap between the academic skills to identify common brain areas for all academic skills.

Current Study

Our overarching objective is to examine factors that influence academic skills after focal brain injury, using a lifespan approach. Of note, in this study we include focal brain lesions caused by heterogeneous etiologies (e.g., stroke, resection, focal traumatic brain injury). All included subjects have lesions visible on a brain scan. Importantly, conditions for which there could be a “brain injury” or “dysfunction” but no focal, visible brain lesions (e.g., degenerative diseases) are not included.

We first aim to examine if our sample differs from normative expectations (i.e., the population test mean). We predict there will be significant differences between the performance in our lesioned sample and normative expectations based on past work (e.g., Anderson et al., 2009; Champigny et al., 2020), but acknowledge these differences may be small following work that has found consistent, small differences between lesioned populations and normative expectations (e.g., Peterson et al., 2019). We then aim to expand knowledge of the contribution of AgeL for outcomes and examine if AgeL is a reliable predictor of academic skills outcomes in a lifespan sample, while accounting for relevant covariates. Potentially relevant covariates include previously identified factors of interest such as demographic (i.e., sex, education), age (i.e., age of assessment, chronicity), and injury (i.e., lesion laterality, lesion size, etiology, history of seizures) factors. Following prior work that examined outcomes across early development (e.g., Anderson et al., 2009; Max et al., 2010), we hypothesize that earlier AgeL will predict worse academic skills. It is important to note that while a large majority of the literature has found a linear relationship between AgeL and outcomes, it is possible that the relationship is non-linear as non-linear relationships have been found in studies examining the relationship between AgeL and cognitive outcomes (e.g., Allman & Scott, 2013; Chang et al., 2016; Goodman & Yude, 1996). After identifying if AgeL is a significant predictor of academic skills while accounting for covariates, we will identify which of the various demographic, age, and injury factors are also significant predictors in the model to contribute to broader depiction of academic skills outcomes after brain injury.

The final main goal of the study is to investigate neural regions associated with math, reading, and spelling, utilizing lesion-symptom mapping. Following prior literature, we hypothesize that (i) damage to parietal structures, specifically the intraparietal sulcus and associated white matter tracts (such as the superior longitudinal fasciculus; SLF) will be associated with deficits in math, (ii) damage to left-hemisphere temporo-parietal structures, including white matter tracts will be associated with deficits in reading, and (iii) damage to the left angular gyrus and left temporo-occipital regions will be associated with deficits in spelling.

Methods

Sample

Participants were selected from the Patient Registry of the Division of Cognitive Neuroscience at the University of Iowa. All individuals included in the registry have been assessed following protocols outlined in Tranel (2009). Individuals are included in the registry if they have a stable lesion of at least three months duration, and no history of significant alcohol abuse or substance abuse, psychiatric disorder, intellectual disabilities, or neuropathological processes unrelated to the lesion (e.g., dementia, movement disorder). Written consent was obtained from either the subjects or legal guardians, depending on what was appropriate given the age of the individual, in accordance with IRB #200105018 at the University of Iowa. Verbal consent was obtained from all patients regardless of age. Individuals were included in this study if they had focal damage to the brain (i.e., a lesion visualized on imaging) and had completed at least one of the desired measures, listed under Assessments. For the purposes of this study, neuropsychological data were collected when patients were in the chronic phase of recovery (i.e., ≥ 3 months post-brain injury, with ten exceptions. Specifically, ten individuals were assessed between two to three months after onset). In cases with multiple post-injury assessments, the first time point the patient received each assessment in the chronic period after injury onset was used in analyses. Overall, 651 participants met inclusion criteria for the current study. AgeL for these participants ranged from pre-birth to 85 years old. Table 1 presents detailed demographic information about the sample.

Table 1.

Demographic and Assessment Information

Injury onset period (years) Total sample
(pre-birth-85)
n 651
Demographic information
Gender (female), n (%) 309 (47%)
Handedness (right), n (%) 578 (89%)
Mean education in years at assessment (SD) 13.16 (3.15)
Race and ethnicity (White, non-Hispanic), n (%) 630 (97%)
Assessment information
Mean age at assessment in years (SD) 49.35 (17.08)
Math, n 526
Reading, n 644
Spelling, n 519
Open in a new tab

Note. Gender was assessed using self-report forms during testing; SD = Standard Deviation.

Assessments

The Wide Range Achievement Test (WRAT) was used in this study as a measure of math, reading, and spelling. Participants were included if they had at least one of the relevant subtests of the WRAT (i.e., math, reading, or spelling). Standard scores were used in analyses (M = 100, SD = 15). Due to patients being assessed and added to the Patient Registry over a long period (i.e., from 1982 to 2020) different versions of the WRAT were used (i.e., the WRAT-R and the WRAT-4; Jastak, 1984; Wilkinson & Robertson, 2006). There is substantial overlap between different versions of the WRAT and thus we included both versions. For example, 73.6% of the WRAT-4 word reading items are the same as the WRAT-3 (Wilkinson & Robertson, 2006) and the WRAT-3 and WRAT-R reading standard scores are highly correlated (r = .94; Wilkinson, 1993). Overall, using different versions of a standard test has been common practice with special patient populations (e.g., Peterson et al., 2019; Westmacott et al., 2010). Notably, the ranges for the standard score conversions changed across different WRAT versions. For instance, the lowest possible standard score on the WRAT-4 is a standard score of 55, whereas the lowest possible standard score on the WRAT-R is a 45. To ensure that all scores included in analyses were on the same scale all standard scores below a score of 55, recorded from the WRAT-R, were changed to a standard score of 55 (i.e., 16 scores for reading, four scores for math, four scores for spelling).

Lesion Characteristics

Individuals who experienced focal brain insult from a range of etiologies were included in the sample. Included cases had neuroimaging and identifiable focal lesions (see Table 2 for lesion characteristics). One limitation of past work has been the reduced inclusion of ranges of lesion etiologies. While this is useful in limiting potential confounding effects of type of injury, it prevents investigators from painting a more generalizable picture of the impact of AgeL on academic skills, because many etiologies are age specific. For instance, strokes are more likely to occur in later life (Kissela et al., 2012). To have a comprehensive picture of the influence of brain injury on the development of academic skills, it is useful to incorporate conditions that are seen from the perinatal period through late childhood and into adulthood (Taylor & Alden, 1997).

Table 2.

Lesion Characteristics

Lesion onset period (years) Total sample
(pre-birth-85)
n 651
Age at lesion onset mean (SD) 45.27 (18.72)
Chronicity mean (SD) 3.64 (5.51)
Lesion etiology
Ischemic stroke, n (%) 279 (43%)
Hemorrhagic stroke, n (%) 95 (15%)
Subarachnoid hemorrhage with a clip or coil, n (%) 22 (3%)
Benign tumor resection, n (%) 74 (11%)
Epilepsy resection, n (%) 44 (7%)
Other resection, n (%) 34 (5%)
Head trauma, with focal contusion, n (%) 22 (3%)
Herpes simplex encephalitis, n (%) 15 (2%)
Encephalitis, non-HSE, n (%) 2 (<1%)
Urbach-Wiethe disease, n (%) 2 (<1%)
Other, n (%) 3 (<1%)
Multiple etiologies, n (%) 59 (9%)
Relevant medical history
History of epilepsy, n (%) 105 (16%)
Lesion laterality
Left, n (%) 284 (44%)
Right, n (%) 251 (39%)
Bilateral, n (%) 116 (18%)
Open in a new tab

Note. All percentages are rounded to the nearest whole percentage.

Definition of Age and Injury Variables

In this study, AgeL refers to the age at which the lesion occurred and was calculated based on review of the patient’s medical records by subtracting the date of birth from the date of lesion onset. AgeL was leveraged as a continuous predictor. Age of assessment refers to the age at which academic skills testing was completed and was represented as a continuous variable. Chronicity refers to the time elapsed from the lesion onset to the date of assessment (i.e., age of assessment – AgeL) and was leveraged as a continuous variable. Lesion size is a calculation of the voxels (measured in cubic mm) impacted by a lesion on imaging and was represented as a continuous variable. Lesion laterality was represented as a categorical variable (i.e., left, right, or bilateral). Etiology was coded as a categorical variable based on review of the patient’s medical records. A history of epilepsy was coded as a categorical variable (i.e., a history of seizures, or not) from the patient’s medical records.

Analysis Plan

To investigate if the full sample differed from expected population mean scores for the measures of academic skills, the sample was compared to published test norms using one sample t-tests. Next, we ran a linear regression with all covariates of interest and AgeL. The set of relevant variables considered included education, sex, etiology, history of seizures, lesion size, lesion laterality, age of assessment, and time since lesion (i.e., chronicity). Specifically, for each academic skill, all relevant covariates were first entered into the model. Then AgeL was added into the model. Next, to find the best fitting model to describe the relation between AgeL and each academic skill, AgeL squared, and finally AgeL cubed were also included as predictors of math, reading, and spelling, while accounting for relevant covariates. Finally, the full model was reduced to only those predictors which most reliably accounted for variance in academic skills outcomes.

Lesion Location Analyses

A Pearson correlation was run to identify the association between the academic skills (i.e., math, reading, and spelling) to contextualize lesion-symptom mapping overlap results. Multivariate lesion-symptom mapping analyses were then performed to examine areas of the brain that, when lesioned, were associated with math, reading, and spelling performance. Given the potential differences between those with adult-onset and child-onset lesions all lesion-symptom maps were derived from adult-onset participants. As there is a protracted development of academic skills, the cut-off for adult-onset was made cautiously based on the age brackets in the normative tables of the WRAT (i.e., AgeL > 24 years old; Wilkinson & Robertson, 2006).

First, structural MRI or CT data were used to identify lesion location, depending on what was available for each individual. All neuroimaging was acquired at least three months after the onset of the lesion, with seven exceptions who had scans between 51 days and three months after the injury. Overall, neuroimaging largely took place between nine months and four years after brain injury, with a median time of 1.41 years and an interquartile range of 0.72 to 4.68 years. Importantly, lesions tend to be stable after the three-month recovery period and don’t change over time. This has been well-documented in the Iowa Neurological Patient Registry (e.g., Damasio, 2000; Tranel, 2019).

Identified lesion boundaries were manually traced on the native T1-weighted scans using FSL (Smith et al., 2004). For any scans that were acquired before 2006, Brainvox and Map-3 techniques were used to trace the lesion volume directly onto a template brain, rather than the native scans (Damasio & Frank, 1992; Fiez et al., 2000). After lesion tracing was completed, the tracings were transformed into MNI152 space, using nonlinear registration from ANTs software (Avants et al., 2014). A neurologist reviewed both the native trace and transformation in MNI space, to ensure accuracy.

We utilized a lesion-symptom mapping package in R (LESYMAP), to perform the lesion-behavior mapping analyses (Pustina et al., 2018). LESYMAP uses sparse canonical correlation analysis (SCCAN) to identify the normalized voxel weights that are most highly correlated with behavioral/performance scores (i.e., the three WRAT subtests). Statistical significance of the resulting map of voxel weights was determined through a 4-fold cross-validation correlation between the scores predicted by the model and the observed scores on measures of academic skills. Each individual voxel received a weight between 0 and 1. Higher-weighted values indicate that particular group of voxels was more highly associated with a score on a behavioral measure.

Each SCCAN result was then binarized and utilized for a proportional analysis to identify the relevant gray matter corresponding with the lesion-symptom map results. Specifically, as all SCCAN results were significant, they were made binary and then divided into left and right hemisphere masks. Then we calculated the proportion of the regions of interest (ROIs) occupied by each of the MMP1 parcels in the Glasser and colleagues (2016) atlas. There are 180 MMP1 parcels in the HCP-MMP1 atlas per hemisphere, for 360 total parcels (Glasser et al., 2016). All areas that overlapped greater than 1% (0.01 proportion) with ROIs for academic skills are listed. The same process was used to examine relevant white matter using the Yeh and colleagues (2018) HCP842 tractography atlas. We identified relevant gray and white matter for all individual academic skills and overlap between academic skills.

Transparency and Openness

In this manuscript we report how we determined our sample size, any data exclusion criteria, all manipulations, and all measures included in the study. We follow JARS (Kazak, 2018). Due to the sensitive nature of our data (protected health information), the data are not available for public use. Quantitative analyses were conducted using SPSS (Version 28.0) and R. This study was not pre-registered.

Results

Sample Characteristics

Our lifespan sample had significantly lower scores on measures of math, reading, and spelling when compared to a population mean standard score of 100 (Table 3). Of note, while the sample scored significantly lower than normative expectations, the mean score of the sample was still within the average range and the mean differences were rather small (Table 3).

Table 3.

Academic Outcomes, Difference from Normative Population

Measure (WRAT) Test mean Sample mean SD t df p Mean difference 95% CI
Math 100 96.06 14.39 −6.28 525 <.001 −3.94 −5.17, −2.71
Reading 100 94.72 15.44 −8.67 643 <.001 −5.28 −6.47, −4.08
Spelling 100 94.79 15.04 −7.90 518 <.001 −5.21 −6.51, −3.92
Open in a new tab

Note. SD = Standard Deviation

Impact of Age of Lesion Onset across the Lifespan

Correlations revealed age of assessment and AgeL were highly correlated in this sample (r = .94, 95% CI [.94, .95]) and could induce multicollinearity. Given a primary aim was to identify the potential contribution of AgeL in a lifespan sample, AgeL was retained in analyses and age of assessment was removed.

Math Results

A linear regression was run predicting math standard scores first as a function of all covariates of interest. Then to obtain the best fitting model for the data, models with AgeL, AgeL squared, and AgeL cubed were run successively. Tests to see if the data met the assumption of collinearity indicated multicollinearity was not a concern (AgeL VIF = 1.45; Chronicity VIF = 1.32; Epilepsy VIF = 1.94; Etiology VIF = 2.00; Lesion Size VIF = 1.02; Lesion Laterality VIF = 1.01; Sex VIF = 1.01; Education VIF = 1.09). The initial model including covariates was significant (F(11, 504) = 13.50, p < .001), with an R2 of .23. In the second model, AgeL was a significant predictor (B = .13, SE = 0.04, p < .001), and it significantly improved the model (p < .001) with an R2 change of .02 (R2= .25) from the initial model. Next, we added AgeLsquared to see if a quadratic model would fit the data better. Indeed, AgeLsquared was a significant predictor (B = .01, SE = 0.001, p < .001) with a significant R2 change of .05 (R2 = .29, p < .001) above the relevant covariates and AgeL, and this model was significant (F(13, 502) = 15.95, p < .001). This quadratic model revealed an upward parabola (Figure 1). In this model, education (B = 2.05, SE = 0.19, p < .001), lesion size (B = .001, SE = 0.001, p < .001), and sex (B = 2.66, SE = 1.09, p = 0.02) all remained as significant predictors of math performance. Specifically, male sex was associated with better outcomes (M = 97.52, SD = 15.11) when compared to female sex (M = 94.54, SD = 13.50). Smaller lesion size and higher education were also associated with better math outcomes. AgeLcubed was not significant and did not significantly add to the model (p = .46, R2 change =.001) and thus was not included in the model. The final model including only those predictors that remained significant (i.e., education, lesion size, sex, AgeL and AgeLsquared) had an R2 of .28, (F(5, 511) = 39.88, p < .001).

Figure 1.

Figure 1

Open in a new tab

Age of Lesion Onset and Math Outcomes

Reading Results

The same set of analyses were conducted for reading scores. A linear regression was run predicting reading standard scores as a function of all covariates, followed by AgeL, AgeL squared, and AgeL cubed. Tests to see if the data met the assumption of collinearity indicated multicollinearity was not a concern (AgeL VIF = 1.42; Chronicity VIF = 1.28; Epilepsy VIF = 1.94; Etiology VIF = 2.02; Lesion Size VIF = 1.02; Lesion Laterality VIF = 1.01; Sex VIF = 1.00; Education VIF = 1.07). The initial model including covariates was significant (F(11, 619) = 18.95, p < .001), with an R2 of .25. In the second model, AgeL improved the model with marginal significance (p = .06, R2 change of .004; R2 = .26). Next, we tested if a quadratic model would fit the data better. AgeLsquared was a significant predictor (B = .01, SE = 0.001, p < .001) and it significantly improved the model (p < .001) with an R2 change of .04 above the relevant covariates and AgeL (R2 = .30). The quadratic model suggested an upward parabola (see Figure 2). In this model, a history of epilepsy (B = −5.66, SE = 1.80, p = .002), etiology (i.e., specifically a history of a resection; B = 5.22, SE = 1.54, p < .001), lesion size (B = .001, SE = 0.001, p = .002), lesion laterality (i.e., specifically right laterality; B = 3.23, SE = 1.16, p = .01), and education (B = 2.45, SE = 0.19, p < .001) all remained as significant predictors of reading performance. Specifically, those with a history of epilepsy had lower reading scores (M = 92.63, SD = 14.17) than those without a history of epilepsy (M = 95.22, SD = 15.59) and those with a history of a resection had higher mean scores (M = 96.12, SD = 13.02), than those with other injury etiologies (M = 94.08, SD =16.40). Additionally, right-sided lesions tended to have better outcomes (M = 95.99, SD = 14.58) when compared to lesions that were left-sided (M = 93.34, SD = 16.40; p = .05). There was no significant difference between bilateral lesions (M = 95.30, SD = 14.72) and left-sided lesions (p = .27), or bilateral lesions and right-sided lesions (p = .68). Higher education and smaller lesion size were also associated with better reading outcomes. AgeLcubed did not significantly add to the model (p = .18, R2 change =.002; R2 = .30). The final quadratic model including only those significant predictors (i.e., a history of epilepsy, etiology, lesion size, lesion laterality, education, AgeL, and AgeLsquared) was significant (F(11, 620) = 23.20, p < .001), with an R2 of .29.

Figure 2.

Figure 2

Open in a new tab

Age of Lesion Onset and Reading Outcomes

Spelling Results

A linear regression was then run predicting spelling standard scores as a function of all potential covariates, followed by AgeL, AgeL squared, and AgeL cubed. Tests to see if the data met the assumption of collinearity indicated multicollinearity was not a concern (AgeL VIF = 1.46; Chronicity VIF = 1.31; Epilepsy VIF = 1.93; Etiology VIF = 2.00; Lesion Size VIF = 1.02; Lesion Laterality VIF = 1.01; Sex VIF = 1.01; Education VIF = 1.08). The initial model was significant (F(11, 497) = 15.91, p <. 001), with an R2 of .26. Adding AgeL did not significantly improve the model (p = .10, R2 change = .004; R2 = .26). As before, AgeLsquared did significantly improve the model (p < .001) with an R2 change of .06 (R2 = .32) above the relevant covariates and AgeL. In the quadratic model, results revealed a significant quadratic estimate (B = .01, SE = 0.002, p < .001), suggesting an upward parabola (see Figure 3). In the quadratic model, etiology (i.e., a history of resection; B = 4.40, SE = 1.63, p = .007), lesion size (B = .001, SE = .001, p = .02), sex (B = 5.72, SE = 1.12. p < .001), and education (B = 2.45, SE = .20, p < .001) all remained as significant predictors of spelling performance. Specifically, female sex was associated with higher spelling outcomes (M = 97.52, SD = 13.96) than male sex (M = 92.08, SD = 15.60), as were higher education, and smaller lesion size. Those with a history of a resection had higher spelling scores on average (M = 96.37, SD = 14.63) when compared to those who had another etiology of injury (M = 93.90, SD = 15.21). AgeLcubed did not significantly add to the model (p = .42, R2 change =.001; R2 of .32). The final quadratic model including only these predictors (i.e., etiology, lesion size, sex, education, AgeL, and AgeLsquared) was F(9, 500) = 25.58, p <. 001), with an R2 of .32.

Figure 3.

Figure 3

Open in a new tab

Age of Lesion Onset and Spelling Outcomes

Interim Summary

Overall, those who have brain injury are more likely to be significantly lower on measures of math, reading, and spelling than would be expected based on population mean scores, though mean differences were rather small. Age of lesion onset is a relevant variable for predicting academic skills outcomes after brain injury in a lifespan sample with a quadratic model fitting best for predicting outcomes. In addition to age of lesion onset, demographic variables (i.e., education, and sex) and injury variables (i.e., lesion size, lesion laterality, etiology, history of seizures) remained as significant predictors for academic skills.

Lesion Location Analyses

Next, we examined which brain areas are associated with math, reading, and spelling performance utilizing lesion-symptom mapping in those with lesion-onset in the adult period. All lesion-symptom maps were significant, for math (r = 0.14, p = 0.003, optimal sparseness = 0.16), reading (r = 0.35, p < .0001, optimal sparseness = −0.37), and spelling (r = 0.16, p = 0.001, optimal sparseness = −0.68). Correlations were run to investigate the association between academic skills in our sample. All academic skills were significantly correlated (math and reading r(422) = .55, p <.001; math and spelling r(476) = .58, p <.001; reading and spelling r(429) = .78, p <.001). We additionally found that the areas of the brain relevant for academic skills overlapped. Areas of overlap area detailed first, followed by additional areas of relevance identified in the individual lesion-symptom maps.

Overlap

Overlap for all three academic skills included the left hemisphere frontal opercular area and left-hemisphere white matter including projection fibers (i.e., corticospinal tract, corticostriatal pathway, temporopontine tract, occipitopontine tract, corticothalamic pathway), association fibers (i.e., inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, U-fiber, arcuate fasciculus, frontal aslant tract, acoustic radiation), and commissural fibers (i.e., corpus callosum). See Figure 4 for depiction of overlapping and individual academic skills lesion-symptom map results.

Figure 4.

Figure 4

Open in a new tab

Lesion-Symptom Map Analyses for Math, Reading, and Spelling, Including Overlap

Math and Reading.

For math and reading additional structures of relevance included the left hemisphere temporal lobe (i.e.., auditory association cortex) and left hemisphere white matter association fibers (i.e., superior longitudinal fasciculus).

Math and Spelling.

Additional relevant right-hemisphere white matter for math and spelling included projection fibers (i.e., corticostriatal pathway, corticospinal tract, corticothalamic pathway, frontopontine and parietopontine tracts) and association fibers (i.e., U-fiber, extreme capsule).

Reading and Spelling.

Additional areas of relevance for overlap of reading and spelling included left hemisphere gray matter in the frontal lobe (i.e., rostral area 6 of the inferior frontal cortex including Broca’s area; the premotor cortex), insular areas (i.e., the middle insular area, and posterior insular area), temporal lobe (i.e., auditory association cortex and lateral temporal cortex), and parietal lobe (i.e., inferior parietal cortex). Overlapping left hemisphere white matter included projection fibers (i.e., frontopontine tract).

Individual Skills

Areas of relevance specifically for math not listed in overlapping sections included right-hemisphere association fibers (i.e., arcuate fasciculus, superior longitudinal fasciculus; see supplemental Figure 1 for lesion coverage and the individual math lesion-symptom map).

Reading.

Additional areas of relevance for reading included left-hemisphere insular area (i.e., posterior operculum of the sylvian fissure), temporal lobe (i.e., hippocampus, perirhinal ectorhinal cortex, auditory 4 and 5 complex), and parietal lobe structures (i.e., the PF complex, Brodmann area 40) and right hemisphere third and fourth visual area. Additional relevant left hemisphere white matter pathways included association fibers (i.e., vertical occipital fasciculus), and projection fibers (i.e., parietopontine tract) and right hemisphere association fibers (i.e., inferior fronto-occipital fasciculus, inferior longitudinal fasciculus) and projection fibers (i.e., temporopontine tract, optic radiation) and the cerebellum. See supplemental Figure 2 for lesion coverage and the individual reading lesion-symptom map.

Spelling.

The lesion-symptom map of the spelling subtest demonstrated that this measure was also associated with right hemisphere white matter including association fibers (i.e., middle longitudinal fasciculus). See supplemental Figure 3 for lesion coverage and the individual spelling lesion-symptom map.

Discussion

The first aim of the current study was to examine how our sample compared to normative expectations. While our full sample had significantly lower scores on measures of math, reading, and spelling when compared to population norms for the WRAT it is important to note that the mean differences between our sample and the normative population were relatively small (i.e., less than one standard deviation). These small differences between those with brain injuries and normative expectations align with some other past work investigating math problem solving and single word reading (Peterson et al., 2019). The relatively small differences between our sample and population norms can be taken as an indication of the resilience of academic skills post-brain injury. There has been conflicting evidence regarding the resilience of academic skills after brain injury and this question could benefit from more investigation with other samples, including a longitudinal approach, which may provide important insight regarding recovery across time (e.g., Ballantyne et al., 2008; Duval et al., 2002). Overall, our data point strongly to resilience, albeit with some consistent indication of very mild reduction in academic skills after brain damage.

We contributed to past work examining age of lesion onset (AgeL) by determining if AgeL was a significant predictor of academic skills outcomes after brain injury, in a lifespan sample. AgeL significantly predicted math, reading, and spelling outcomes, even when accounting for relevant covariates, providing evidence that AgeL is a relevant variable in the prediction of academic skills outcomes after brain injury across the lifespan. Further, a quadratic model of AgeL fit best, with both earlier and later AgeL predicting better performance in the full sample than the middle years. To confirm that the quadratic relation observed cannot solely be explained by distribution of AgeL in our sample, participants were divided into quartiles based on AgeL, which ensured an equal number of participants in each age of lesion onset group. We then replicated all analyses with the categorical AgeL variable. The pattern of results remained unchanged, whereby AgeL showed a quadratic relation to all three academic variables (reading, math, spelling), confirming the analyses using AgeL as a continuous variable. While most work has focused on a linear model of the influence of AgeL on outcomes, non-linear effects have been supported by a few studies (e.g., Allman & Scott, 2013; Chang et al., 2016; Goodman & Yude, 1996).

Importantly, the aim of this study was not to identify specific periods of risk for academic skills outcomes, but rather to examine a broader overview of academic skills after brain injury in a lifespan sample. In work that has examined specific periods of increased risk for poor outcomes, lesion onset in a period before the age of two has been shown to confer increased risk for academic skills outcomes (e.g., Anderson et al., 2009; das Dores Rodrigues et al., 2011; Max et al., 2010; Westmacott et al., 2018). Increased vulnerability during early life could align with the Early Vulnerability Hypothesis and indicate damage during or before this period in early life could confer increased risk for poor outcomes after injury.

Overall, our nonlinear finding likely indicates a combination between plasticity and early vulnerability is relevant for academic skills outcomes. When examining results in a more fine-grained way in our study, preliminary analyses evidenced a potential quadratic relationship within the developmental years. However, we were unable to test this finding quantitatively due to our relatively small sample size within the developmental years. Deotto and colleagues (2019) also reported a nonlinear finding for academic skills and those with stroke during the ages of 6–14 had the poorest academic outcomes, though these findings were descriptive in nature. Importantly, greater impairment to academic skills might be observed if a lesion is acquired during times that correspond to when these skills are being acquired or fine-tuned, aligning with a critical period (e.g., Kolb, 1995; Luciana, 2003). A combination between plasticity and early vulnerability has also been identified in some work that found injury onset in a period from one to sixty months confers more risk for poor IQ outcomes when compared to before or after that time, potentially due to increased plasticity during periods before that time and stabilization of skills after that time (Goodman & Yude, 1996). Increased plasticity, or the increased ability of a young brain for anatomical reorganization or regrowth when compared to an adult brain (e.g., Kolb & Gibb, 1993; Kolb et al., 1994) resulting in better functional outcomes, has been supported in classic work examining motor and language outcomes (e.g., Kennard, 1938, 1940, 1942; Lenneberg, 1967; Tompkins, 1990).

Overall, we again emphasize our goal was to identify the potential utility of the age of lesion onset variable in a lifespan sample and we cannot make specific claims about periods of vulnerability versus resilience. Future work could further refine periods of potential vulnerability during the early life period for academic skills by including more individuals with early age of lesion onset while also including later age of onset ranges (i.e., into adulthood) and allowing for a quadratic relationship to expand work that has previously been done. Future work could also utilize the lesion method to determine structures necessary for academic skills during transitional time periods.

Beyond age of lesion, various other factors, including age, injury characteristics, and demographic factors were associated with academic skills after brain injury in a lifespan sample. Specifically, education, sex, lesion size and laterality, etiology, and a history of epilepsy emerged as the most reliable predictors of academic skills outcomes. These are among well-known demographic and injury factors that are associated with a wide range of cognitive and academic outcomes (e.g., Anderson et al., 2011; Anderson et al., 2014; Banich et al., 1990; Braun et al., 2000; Duval et al., 2008; Levine et al., 1987; Montour-Proulx et al., 2004; Westmacott et al., 2010; Westmacott et al., 2018). In examining these findings, one could question if higher education could be correlated with chronicity resulting in chronicity ceasing to be a reliable predictor. To check this possibility, follow-up analyses indicated chronicity was not significantly associated with education (r(751) = .03, p = .38). Overall, this study expands on past work by accounting for these potentially relevant variables while examining academic skill outcomes in a lifespan sample and identifying the most reliable predictors in a lifespan sample. The unique contribution of the current paper is the examination of the contribution of all these factors in a single lifespan sample.

Lesion-Symptom Mapping of Academic Skills

Our final aim was to utilize lesion-symptom mapping to examine which neural structures are most associated with deficits in academic skills, in a large population of lesion patients, which has not previously been produced for math, reading, and spelling in the same sample. Importantly, as we conducted these analyses in the same sample, we were able to identify both areas critical for each academic skill and those shared neural correlates between academic skills. The lesion-symptom map for the math subtest indicated the measure was associated with bilateral long-range association fibers including fronto-parietal white matter, temporo-parietal white matter, and the left temporal lobe (i.e., auditory association cortex). Much work has supported the relevance of the parietal and prefrontal regions for math in adults (Arsalidou & Taylor, 2011; Menon, 2015), including a previous study utilizing lesion-symptom mapping (Baldo & Dronkers, 2007). The finding of the bilateral fronto-parietal white matter as relevant for math in our study also aligns with the previous work showing that the connectivity between the frontal and parietal regions is related to math abilities (e.g., Chang et al., 2016; Rosenberg-Lee et al., 2011) and work showing that abnormalities in white matter tracts such as the superior longitudinal fasciculus, corpus callosum, inferior longitudinal fasciculus, corona radiata, and corticospinal tracts and hyper-connectivity between the intraparietal sulcus, and lateral frontoparietal, and default mode networks are associated with impairments in math (e.g., Matejko & Ansari, 2015; Rosenberg-Lee et al., 2015).

The lesion-symptom map for reading indicated that the measure was associated with long-range association fibers, including left frontal white matter, left temporo-parietal white matter, left occipito-temporal white matter, and the left frontal lobe (i.e., inferior frontal cortex and premotor cortex), temporal lobe (e.g., hippocampus, auditory association cortex, lateral temporal cortex), parietal lobe (i.e., inferior parietal cortex, PF complex), and insular area (i.e., posterior operculum of the sylvian fissure, posterior insular area). The left frontal and temporal cortex have been supported as relevant for reading by numerous studies, as has temporo-parietal and occipito-temporal white matter, so results largely align with what was expected (e.g., Dehaene, 2009; Mani et al., 2008; Piras & Marangolo, 2009).

Finally, the lesion-symptom map for spelling indicated that the measure was associated with left long-range association fibers including frontal white matter, left temporo-parietal white matter, left occipito-temporal white matter, right fronto-parietal white matter, and the left frontal lobe (i.e., inferior frontal cortex, premotor cortex, frontal operculum), temporal lobe (i.e., auditory association cortex, lateral temporal cortex), parietal lobe (i.e., inferior parietal cortex), and insular area (i.e., middle and posterior insular area). These findings are largely supported by past work investigating the neural correlates of spelling tasks (e.g., Baldo et al., 2018; Planton et al., 2013).

Additionally, significant overlap existed between relevant areas for math, reading, and spelling, with the left hemisphere frontal opercular area and left hemisphere white matter significantly associated with deficits in all three academic skills. Some work has identified overlap in other academic skills using lesion-symptom mapping including between word-reading and writing (Baldo et al., 2018) and math skills and language comprehension (Baldo & Dronkers, 2007). Our findings support past work that has examined academic skills individually and those that have identified overlap in other academic skills by reporting relevant neural areas for three core academic skills when measured in the same sample. Our findings also add to prior neuroimaging with typically developing individuals to investigate the neural correlates of academic skills by highlighting areas that are necessary for optimal functioning.

Limitations and Future Directions

This study included participants with a wide range of etiologies of brain injury. This is potentially a limitation as some etiologies may be more focal (e.g., stroke), or diffuse (e.g., TBI). To account for this potential limitation, we examined the contribution of etiology as a factor in our models and all lesions were identified as focal on neuroimaging. Importantly, recent work has shown patients who sustained TBIs, identified on neuroimaging, and those who sustained strokes have equivalent outcomes when matched for lesion volume and location (see Harris et al., in press). This work supports the use of heterogeneous etiologies when accounting for relevant lesion variables. Past work also utilizing samples with heterogeneous etiologies (e.g., Anderson et al., 2014) employed a similar method of identifying those with focal lesions and accounting for relevant factors including lesion size and laterality. As such, we accounted for such relevant variables and included a range of etiologies to paint a comprehensive picture.

In this study we utilized a broad measure of academic skills (i.e., the WRAT) and compared our participants to a very broad measure of the population’s performance (i.e., the mean standard score). Importantly, the WRAT is a screening tool of academic abilities and there are several other measures that may offer a more in-depth analysis of academic abilities beyond word-reading, spelling, and math. The WRAT has been used in several other studies to examine academic abilities after brain injury (e.g., Anderson et al., 2009; Ballantyne et al., 2008; Champigny et al., 2020; Max et al., 2010; Mosch et al., 2005) and the use of this measure allows us to expand on these studies and offer a lifespan perspective. Future work could benefit from examining more detailed subsections of academic skills. Also, it is possible the population norms are not a good comparison for our sample and rather that our sample should be performing above the population mean, since those educated in Iowa perform higher on standardized testing than the average (National Center for Education Statistics). Future work should verify the potential resilience of academic skills by identifying tests with more refined norms based on the geographical area of interest (e.g., state testing).

Another avenue of interest for future work may be investigating longitudinal outcomes on academic skills tests before and after injury onset. In this study, no information was available on pre-diagnosis or pre-injury academic skills. Identifying pre-injury academic skills information would allow investigators to compare individuals to themselves, rather than a population norm. As pre-injury academic ability has been identified as a predictor of academic success after brain injury, accounting for a large portion of the variance of academic outcomes after brain injury (e.g., Catroppa & Anderson, 2007), future work could benefit from identifying change across time (i.e., before and after brain injury) within individuals.

Finally, the wide age range included in the study adds additional considerations when interpreting findings. It is likely there are age-related differences in academic skills performance (e.g., the elderly group may evidence deterioration in skills due to normal aging). To limit the influence of effects on academic skills due to expected age-related changes, age-corrected standard scores were utilized in this study. Age-corrected standard scores are calculated with an aim to account for variance explained by typical aging and place all individuals on the same scale (i.e., comparing them to their peers, rather than to others not in their age groups). However, it is possible there are differences across the sample due to typical aging that are not perfectly accounted for using standard scores and results should be interpreted with this consideration of the wide age range in mind.

Constraints on Generality

The generalizability of our findings is limited by the demographics of our sample, as the sample contains mostly White, non-Hispanic, and relatively highly educated participants. This sample is not representative of all groups. Thus, our results are most applicable to samples with similar demographics and caution should be exercised when generalizing findings to samples with different demographics. For the purposes of this research, we included all patients who were identified as having sustained a focal brain lesion and who were willing to participate in research (from our Patient Registry). However, our sample has limited racial/ethnic diversity and thus it is important that future work address these limitations by investigating if these results are replicated in a more diverse sample.

Conclusion

Overall, several age, injury, and demographic factors are relevant in the prediction of academic skills after brain injury in a lifespan sample. This study showed AgeL is a significant predictor of academic skills across the lifespan, expanding on past work that primarily investigated the contribution of AgeL on academic skills in developmental samples. The current study adds to past work in a number of ways. It is the first study to use a large sample of lesion patients to investigate the contribution of AgeL in predicting academic skills outcomes, across the lifespan, while examining the contribution of other relevant age, injury, and demographic factors. The utilization of lesion-symptom mapping techniques to identify brain areas associated with academic skills is another critical addition to past work largely using neuroimaging techniques to identify neural correlates of academic skills. Finally, the examination of all academic skills in the same sample allowed for identification of overlapping areas of relevance for academic skills.

Supplementary Material

Supplementary Material

Key Points:

Question:

This paper examines academic skills outcomes after brain injury by identifying how age and injury factors influence outcomes, with a special focus on the age at which a brain injury occurs, in a lifespan sample.

Findings:

A distributed brain network is relevant for academic skills and age of lesion onset, education, sex, lesion size and side, etiology and seizure history are significant predictors of academic skills in a lifespan sample.

Importance:

The age at which a brain injury occurs, and the location of the injury could be utilized to inform potential risk for academic skills deficits after brain injury.

Next Steps:

Future work could benefit from more precisely defining periods of increased risk for deficits in academic skills after brain injury and utilize methods commonly used in those with adult-onset injuries (i.e., lesion-symptom mapping) in those with developmental-onset lesions to further inform lesion-deficit relations for those with early-onset brain injury.

Acknowledgments

This research was supported in part by a National Institutes of Health Behavioral-Biomedical Interface T32 Predoctoral Training Grant (T32GM108540) and the Kiwanis Neuroscience Research Foundation for Daniel Tranel. Thank you to Aaron Boes for offering relevant insights regarding the results of this study. Due to the sensitive nature of our data (protected health information), the data are not available for public use. Some results were presented during a poster presentation at the 2021 American Academy of Clinical Neuropsychology Conference.

References

  1. Allman C, & Scott RB (2013). Neuropsychological sequelae following pediatric stroke: A nonlinear model of age at lesion effects. Child Neuropsychology, 19(1), 97–107. 10.1080/09297049.2011.639756 [DOI] [PubMed] [Google Scholar]
  2. Anderson VA, Spencer-Smith MM, Coleman L, Anderson PJ, Greenham M, Jacobs R, Lee KJ, & Leventer RJ (2014). Predicting neurocognitive and behavioural outcome after early brain insult. Developmental Medicine and Child Neurology, 56(4), 329–336. 10.1111/dmcn.12387 [DOI] [PubMed] [Google Scholar]
  3. Anderson V, Spencer-Smith M, Leventer R, Coleman L, Anderson P, Williams J, … & Jacobs R (2009). Childhood brain insult: can age at insult help us predict outcome?. Brain, 132(1), 45–56. 10.1093/brain/awn293 [DOI] [PubMed] [Google Scholar]
  4. Anderson V, Spencer-Smith M, & Wood A (2011). Do children really recover better? Neurobehavioural plasticity after early brain insult. Brain, 134(8), 2197–2221. 10.1093/brain/awr103 [DOI] [PubMed] [Google Scholar]
  5. Aram DM, & Ekelman BL (1988). Scholastic aptitude and achievement among children with unilateral brain lesions. Neuropsychologia, 26(6), 903–916. 10.1016/0028-3932(88)90058-9 [DOI] [PubMed] [Google Scholar]
  6. Arsalidou M, & Taylor MJ (2011). Is 2 + 2 = 4? Meta-analyses of brain areas needed for numbers and calculations. Neuroimage, 54 (3), 2382–2393. 10.1016/j.neuroimage.2010.10.009 [DOI] [PubMed] [Google Scholar]
  7. Avants BB, Tustison NJ, Stauffer M, Song G, Wu B, & Gee JC (2014). The Insight ToolKit image registration framework. Frontiers in neuroinformatics, 8, 44. 10.3389/fninf.2014.00044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Baldo JV, & Dronkers NF (2007). Neural correlates of arithmetic and language comprehension: A common substrate?. Neuropsychologia, 45(2), 229–235. 10.1016/j.neuropsychologia.2006.07.014 [DOI] [PubMed] [Google Scholar]
  9. Baldo JV, Kacinik N, Ludy C, Paulraj S, Moncrief A, Piai V, … & Dronkers NF (2018). Voxel-based lesion analysis of brain regions underlying reading and writing. Neuropsychologia, 115, 51–59. 10.1016/j.neuropsychologia.2018.03.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ballantyne AO, Spilkin AM, Hesselink J, & Trauner DA (2008). Plasticity in the developing brain: intellectual, language and academic functions in children with ischaemic perinatal stroke. Brain, 131(11), 2975–2985. 10.1093/brain/awn176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Banich MT, Levine SC, Kim H, & Huttenlocher P (1990). The effects of developmental factors on IQ in hemiplegic children. Neuropsychologia, 28(1), 35–47. 10.1016/0028-3932(90)90084-2 [DOI] [PubMed] [Google Scholar]
  12. Ben-Shachar M, Dougherty RF, & Wandell BA (2007). White matter pathways in reading. Current opinion in neurobiology, 17(2), 258–270. 10.1016/j.conb.2007.03.006 [DOI] [PubMed] [Google Scholar]
  13. Binder JR, & Mohr JP (1992). The topography of callosal reading pathways: a case-control analysis. Brain, 115(6), 1807–1826. 10.1093/brain/115.6.1807 [DOI] [PubMed] [Google Scholar]
  14. Booth JR, Burman DD, Meyer JR, Gitelman DR, Parrish TB, & Mesulam MM (2004). Development of brain mechanisms for processing orthographic and phonologic representations. Journal of cognitive neuroscience, 16(7), 1234–1249. 10.1162/0898929041920496 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Blakemore SJ (2012). Imaging brain development: the adolescent brain. Neuroimage, 61(2), 397–406. 10.1016/j.neuroimage.2011.11.080 [DOI] [PubMed] [Google Scholar]
  16. Braun CM, Montour-Proulx I, Daigneault S, Rouleau I, Kuehn S, Piskopos M, … & Rainville C. (2000). Prevalence, and intellectual outcome of unilateral focal cortical brain damage as a function of age, sex and aetiology. Behavioural neurology, 13(3, 4), 105–116. http://www.er.uqam.ca/nobel/r31210/home.html [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Brem S, Bucher K, Halder P, Summers P, Dietrich T, Martin E, & Brandeis D (2006). Evidence for developmental changes in the visual word processing network beyond adolescence. Neuroimage, 29(3), 822–837. 10.1016/j.neuroimage.2005.09.023 [DOI] [PubMed] [Google Scholar]
  18. Catroppa C, & Anderson V (2007). Recovery in memory function, and its relationship to academic success, at 24 months following pediatric TBI. Child Neuropsychology, 13(3), 240–261. 10.1080/09297040600837362 [DOI] [PubMed] [Google Scholar]
  19. Chall JS (1996). Stages of reading development (2nd ed.). Harcourt Brace College Publishers. [Google Scholar]
  20. Champigny CM, Deotto A, Westmacott R, Dlamini N, & Desrocher M (2020). Academic outcome in pediatric ischemic stroke. Child Neuropsychology, 26(6), 817–833. 10.1080/09297049.2020.1712346 [DOI] [PubMed] [Google Scholar]
  21. Chang TT, Metcalfe AWS, Padmanabhan A, Chen T, Menon V (2016). Heterogeneous and nonlinear development of human posterior parietal cortex function. NeuroImage 126, 184–195. 10.1016/j.neuroimage.2015.11.053 [DOI] [PubMed] [Google Scholar]
  22. Cho S, Metcalfe AWS, Young CB, Ryali S, Geary DC, Menon V (2012). Hippocampal-prefrontal engagement and dynamic causal interactions in the maturation of children’s fact retrieval. J. Cogn. Neurosci, 24 (9), 1849–1866. 10.1162/jocn_a_00246 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Cohen J (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum. [Google Scholar]
  24. Cohen L, Lehéricy S, Chochon F, Lemer C, Rivaud S, & Dehaene S (2002). Language‐specific tuning of visual cortex? Functional properties of the Visual Word Form Area. Brain, 125(5), 1054–1069. 10.1093/brain/awf094 [DOI] [PubMed] [Google Scholar]
  25. Crowe LM, Catroppa C, Babl FE, Rosenfeld JV, & Anderson V (2012). Timing of traumatic brain injury in childhood and intellectual outcome. Journal of pediatric psychology, 37(7), 745–754. 10.1093/jpepsy/jss070 [DOI] [PubMed] [Google Scholar]
  26. Damasio H (2000). Chapter 4. The lesion method in cognitive neuroscience. In Boller F, Grafman J, & Rizzolatti G (Eds.). Handbook of Neuropsychology, 1, (pp. 77–102). Elsevier Science. [Google Scholar]
  27. Damasio AR, & Damasio H (1983). The anatomic basis of pure alexia. Neurology, 33(12), 1573–1573. 10.1212/WNL.33.12.1573 [DOI] [PubMed] [Google Scholar]
  28. Damasio H, & Frank R (1992). Three-dimensional in vivo mapping of brain lesions in humans. Archives of Neurology, 49(2), 137–143. 10.1001/archneur.1992.00530260037016 [DOI] [PubMed] [Google Scholar]
  29. das Dores Rodrigues S, Ciasca SM, Guimarães IE, Ibraim da Freiria Elias KM, Camargo Oliveira C, & Leme de Moura-Ribeiro MV (2011). Does stroke impair learning in children? Stroke Research and Treatment, 2011,1–6. 10.4061/2011/369836 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Dehaene S (2009). Reading in the brain Penguin. [Google Scholar]
  31. Denburg NL, & Tranel D (2003). Acalculia and disturbances of the body schema. In Heilman KM & Valenstein E (Eds.), Clinical neuropsychology (4th ed., pp. 161–184). New York: Oxford University Press. [Google Scholar]
  32. Deotto A, Westmacott R, Fuentes A, deVeber G, & Desrocher M (2019). Does stroke impair academic achievement in children? The role of metacognition in math and spelling outcomes following pediatric stroke. Journal of Clinical and Experimental Neuropsychology, 41(3), 257–269. 10.1080/13803395.2018.1533528 [DOI] [PubMed] [Google Scholar]
  33. Duval J, Braun CMJ, Montour-Proulx I, Daigneault S, Rouleau I, & Bégin J (2008). Brain lesions and IQ: Recovery versus decline depends on age of onset. Journal of Child Neurology, 23(6), 663–668. 10.1177/0883073808314161 [DOI] [PubMed] [Google Scholar]
  34. Duval J, Dumont M, Braun CM, & Montour-Proulx I (2002). Recovery of intellectual function after a brain injury: A comparison of longitudinal and cross-sectional approaches. Brain and cognition, 48(2–3), 337–342. https://pubmed.ncbi.nlm.nih.gov/12030463/ [PubMed] [Google Scholar]
  35. Fellows LK, Heberlein AS, Morales DA, Shivde G, Waller S, & Wu DH (2005). Method matters: An empirical study of impact in cognitive neuroscience. Journal of Cognitive Neuroscience, 17, 850–858. 10.1162/0898929054021139 [DOI] [PubMed] [Google Scholar]
  36. Fiez JA, Damasio H, & Grabowski TJ (2000). Lesion segmentation and manual warping to a reference brain: Intra‐and interobserver reliability. Human brain mapping, 9(4), 192–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Forkel SJ (2020). Lesion-symptom mapping: from single cases to the human disconnectome. Encyclopaedia of Behavioural Neuroscience [Google Scholar]
  38. Geary DC (1994). Children's mathematical development: Research and practical applications American Psychological Association. [Google Scholar]
  39. Glasser MF, Coalson TS, Robinson EC, Hacker CD, Harwell J, Yacoub E, … & Van Essen DC (2016). A multi-modal parcellation of human cerebral cortex. Nature, 536(7615), 171–178. 10.1038/nature18933 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Glenn DE, Demir-Lira ÖE, Gibson DJ, Congdon EL, & Levine SC (2018). Resilience in mathematics after early brain injury: The roles of parental input and early plasticity. Developmental cognitive neuroscience, 30, 304–313. 10.1016/j.dcn.2017.07.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Goodman R, & Yude C (1996). IQ and its predictors in childhood hemiplegia. Dev Med Child Neurol, 38, 881–90. 10.1111/j.1469-8749.1996.tb15045.x [DOI] [PubMed] [Google Scholar]
  42. Harris S, Bowren M, Anderson SW, & Tranel D (in press). Does brain damage caused by stroke versus trauma have different neuropsychological outcomes? A lesion-matched multiple case study. Applied Neuropsychology: Adult [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Hebb D (1942). The effects of early and late injury upon test scores and the nature of normal adult intelligence, Proc Amer Phil Soc, 85, 275–92. [Google Scholar]
  44. Inoue K, Madhyastha T, Rudrauf D, Mehta S, & Grabowski T (2014). What affects detectability of lesion–deficit relationships in lesion studies?. NeuroImage: Clinical, 6, 388–397. 10.1016/j.nicl.2014.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Jacomb I, Porter M, Brunsdon R, Mandalis A, & Parry L (2018). Cognitive outcomes of pediatric stroke. Child Neuropsychology, 24(3), 287–303. 10.1080/09297049.2016.1265102 [DOI] [PubMed] [Google Scholar]
  46. Jastak S (1984). WRAT-R: Wide Range Achievement Test Wilmington, Del.: Chicago, Ill. Jastak Associates, Inc.; Stoelting Co. [Google Scholar]
  47. Johnson MH (2011). Interactive specialization: a domain-general framework for human functional brain development?. Developmental cognitive neuroscience, 1(1), 7–21. 10.1016/j.dcn.2010.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Kazak AE (2018). Editorial: Journal article reporting standards. American Psychologist, 73, 1–2. 10.1037/amp0000263 [DOI] [PubMed] [Google Scholar]
  49. Kennard MA (1938). Reorganization of motor function in the cerebral cortex of monkeys deprived of motor and premotor areas in infancy. Journal of Neurophysiology, 1(6), 477–496. 10.1152/jn.1938.1.6.477 [DOI] [Google Scholar]
  50. Kennard MA (1940). Relation of age to motor impairment in man and in subhuman primates. Archives of Neurology & Psychiatry, 44(2), 377–397. 10.1001/archneurpsyc.1940.02280080137008 [DOI] [Google Scholar]
  51. Kennard MA (1942). Cortical reorganization of motor function: studies on series of monkeys of various ages from infancy to maturity. Archives of Neurology & Psychiatry, 48(2), 227–240. 10.1001/archneurpsyc.1942.02290080073002 [DOI] [Google Scholar]
  52. Kimberg DY, Coslett HB, & Schwartz MF (2007). Power in voxel-based lesion-symptom mapping. Journal of cognitive neuroscience, 19(7), 1067–1080. 10.1162/jocn.2007.19.7.1067. [DOI] [PubMed] [Google Scholar]
  53. Klingberg T, Hedehus M, Temple E, Salz T, Gabrieli JD, Moseley ME, & Poldrack RA (2000). Microstructure of temporo-parietal white matter as a basis for reading ability: evidence from diffusion tensor magnetic resonance imaging. Neuron, 25(2), 493–500. 10.1016/S0896-6273(00)80911-3 [DOI] [PubMed] [Google Scholar]
  54. Kolb B (1995). Brain plasticity and behavior Psychology Press. [DOI] [PubMed] [Google Scholar]
  55. Kolb B, & Gibb R (1993). Possible anatomical basis of recovery of function after neonatal frontal lesions in rats. Behavioral neuroscience, 107(5), 799. 10.1037/0735-7044.107.5.799 [DOI] [PubMed] [Google Scholar]
  56. Kolb B, Gorny G, & Gibb R (1994). Tactile stimulation enhances recovery and dendritic growth in rats with neonatal frontal lesions. Soc Neurosci Abstr, 20, 1430. [Google Scholar]
  57. Lenneberg EH (1967). The biological foundations of language. Hospital Practice, 2(12), 59–67. 10.1080/21548331.1967.11707799 [DOI] [Google Scholar]
  58. Levine SC, Huttenlocher P, Banich MT, & Duda E (1987). Factors affecting cognitive functioning of hemiplegic children. Developmental Medicine & Child Neurology, 29(1), 27–35. 10.1111/j.1469-8749.1987.tb02104.x [DOI] [PubMed] [Google Scholar]
  59. Luciana M (2003). Cognitive development in children born preterm: Implications for theories of brain plasticity following early injury. Development and Psychopathology, 15, 1017–1047. 10.1017/S095457940300049X [DOI] [PubMed] [Google Scholar]
  60. Mah YH, Husain M, Rees G, & Nachev P (2014). Human brain lesion-deficit inference remapped. Brain, 137(9), 2522–2531. 10.1093/brain/awu164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Mani J, Diehl B, Piao Z, Schuele SS, Lapresto E, Liu P, … & Lüders HO (2008). Evidence for a basal temporal visual language center: cortical stimulation producing pure alexia. Neurology, 71(20), 1621–1627. 10.1212/01.wnl.0000334755.32850.f0 [DOI] [PubMed] [Google Scholar]
  62. Matejko AA, & Ansari D (2015). Drawing connections between white matter and numerical and mathematical cognition: a literature review. Neurosci. Biobehav. Rev 48, 35–52. 10.1016/j.neubiorev.2014.11.006 [DOI] [PubMed] [Google Scholar]
  63. Max JE, Bruce M, Keatley E, & Delis D (2010). Pediatric stroke: plasticity, vulnerability, and AgeL. The Journal of neuropsychiatry and clinical neurosciences, 22(1), 30–39. 10.1176/jnp.2010.22.1.30 [DOI] [PubMed] [Google Scholar]
  64. Menon V (2015). Arithmetic in the child and adult brain. In: Cohen Kadosh R, Dowker A (Eds.), The Oxford Handbook Of Numerical Cognition Oxford University Press, Oxford. [Google Scholar]
  65. Montour-Proulx I, Braun CMJ, Daigneault S, Rouleau I, Kuehn S, & Bégin J (2004). Predictors of intellectual function after a unilateral cortical lesion: Study of 635 patients from infancy to adulthood. Journal of Child Neurology, 19(12), 935–943. 10.1177/08830738040190120501 [DOI] [PubMed] [Google Scholar]
  66. Mosch C, Max J, & Tranel D (2005). A matched lesion analysis of childhood versus adult-onset brain injury due to unilateral stroke: Another perspective on neural plasticity and recovery of social functioning. Cognitive and Behavioral Neurology, 18(1), 5–17. 10.1097/01.wnn.0000152207.80819.3c [DOI] [PubMed] [Google Scholar]
  67. Nachev P (2015). The first step in modern lesion-deficit analysis. Brain, 138(6), e354. 10.1093/brain/awu275 [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. National Center for Education Statistics., National Assessment of Educational Progress (Project), Educational Testing Service., & United States. (1992). NAEP … reading report card for the nation and the states Washington, D.C: National Center for Education Statistics, Office of Educational Research and Improvement, U.S. Dept. of Education. [Google Scholar]
  69. Niogi SN, & McCandliss BD (2006). Left lateralized white matter microstructure accounts for individual differences in reading ability and disability. Neuropsychologia, 44(11), 2178–2188. 10.1016/j.neuropsychologia.2006.01.011 [DOI] [PubMed] [Google Scholar]
  70. Peters L, & De Smedt B (2018). Arithmetic in the developing brain: a review of brain imaging studies. Developmental Cognitive Neuroscience, 30, 265–279. 10.1016/j.dcn.2017.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Peterson RK, Williams TS, McDonald KP, Dlamini N, & Westmacott R (2019). Cognitive and academic outcomes following childhood cortical stroke. Journal of child neurology, 34(14), 897–906. 10.1177/0883073819866609 [DOI] [PubMed] [Google Scholar]
  72. Philipose LE, Gottesman RF, Newhart M, Kleinman JT, Herskovits EH, Pawlak MA, Marsh EB, Davis C, Heidler-Gary J, & Hillis AE (2007). Neural regions essential for reading and spelling of words and pseudowords. Annals of Neurology: Official Journal of the American Neurological Association and the Child Neurology Society, 62(5), 481–492. 10.1002/ana.21182 [DOI] [PubMed] [Google Scholar]
  73. Piras F, & Marangolo P (2009). Word and number reading in the brain: evidence from a voxel-based lesion-symptom mapping study. Neuropsychologia, 47(8–9), 1944–1953. 10.1016/j.neuropsychologia.2009.03.006 [DOI] [PubMed] [Google Scholar]
  74. Planton S, Jucla M, Roux FE, & Démonet JF (2013). The “handwriting brain”: a meta-analysis of neuroimaging studies of motor versus orthographic processes. Cortex, 49(10), 2772–2787. 10.1016/j.cortex.2013.05.011 [DOI] [PubMed] [Google Scholar]
  75. Purcell J, Turkeltaub PE, Eden GF, & Rapp B (2011). Examining the central and peripheral processes of written word production through meta-analysis. Frontiers in psychology, 2, 239. 10.3389/fpsyg.2011.00239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Pustina D, Avants B, Faseyitan OK, Medaglia JD, & Coslett HB (2018). Improved accuracy of lesion to symptom mapping with multivariate sparse canonical correlations. Neuropsychologia, 115, 154–166. 10.1016/j.neuropsychologia.2017.08.027 [DOI] [PubMed] [Google Scholar]
  77. Qin S, Cho S, Chen T, Rosenberg-Lee M, Geary DC, Menon V, (2014). Hippocampalneocortical functional reorganization underlies children’s cognitive development. Nat. Neurosci, 17(9), 1263–1269. 10.1038/nn.3788 [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Qiu D, Tan LH, Zhou K, & Khong PL (2008). Diffusion tensor imaging of normal white matter maturation from late childhood to young adulthood: voxel-wise evaluation of mean diffusivity, fractional anisotropy, radial and axial diffusivities, and correlation with reading development. Neuroimage, 41(2), 223–232. 10.1016/j.neuroimage.2008.02.023 [DOI] [PubMed] [Google Scholar]
  79. Rapcsak SZ, & Beeson PM (2002). Neuroanatomical correlates of spelling and writing. (Hillis AE, Ed.). Handbook of Adult Language Disorders: Integrating Cognitive Neuropsychology, Neurology, and Rehabilitation Psychology Press, 71– 99. [Google Scholar]
  80. Rapcsak SZ, & Beeson PM (2004). The role of left posterior inferior temporal cortex in spelling. Neurology, 62(12), 2221–2229. 10.1212/01.WNL.0000130169.60752.C5 [DOI] [PubMed] [Google Scholar]
  81. Rapcsak SZ, Beeson PM, Henry ML, Leyden A, Kim E, Rising K, … & Cho H (2009). Phonological dyslexia and dysgraphia: Cognitive mechanisms and neural substrates. Cortex, 45(5), 575–591. 10.1016/j.cortex.2008.04.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Reyna VF, Nelson WL, Han PK, & Dieckmann NF (2009). How numeracy influences risk comprehension and medical decision making. Psychological Bulletin, 135, 943–973. 10.1037/a0017327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Ritchie SJ, & Bates TC (2013). Enduring links from childhood mathematics and reading achievement to adult socioeconomic status. Psychological Science, 24(7), 1301–1308. 10.1177/0956797612466268 [DOI] [PubMed] [Google Scholar]
  84. Rivera SM, Reiss AL, Eckert MA, Menon V (2005). Developmental changes in mental arithmetic: evidence for increased functional specialization in the left inferior parietal cortex. Cereb. Cortex, 15(11), 1779–1790. 10.1093/cercor/bhi055 [DOI] [PubMed] [Google Scholar]
  85. Rosenberg-Lee M, Ashkenazi S, Chen T, Young CB, Geary DC, Menon V, (2015). Brain hyper-connectivity and operation-specific deficits during arithmetic problem solving in children with developmental dyscalculia. Dev. Sci, 18(3), 351–372. 10.1111/desc.12216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Rosenberg-Lee M, Barth M, Menon V, (2011). What difference does a year of schooling make? Maturation of brain response and connectivity between 2nd and 3rd grades during arithmetic problem solving. Neuroimage 57(3), 796–808. 10.1016/j.neuroimage.2011.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. Schlaggar BL, Brown TT, Lugar HM, Visscher KM, Miezin FM, & Petersen SE (2002). Functional neuroanatomical differences between adults and school-age children in the processing of single words. Science, 296(5572), 1476–1479. 10.1126/science.1069464 [DOI] [PubMed] [Google Scholar]
  88. Schlagger BL, & McCandliss BD (2007). Development of neural systems for reading. Annu. Rev. Neurosci 30, 475–503. 10.1146/annurev.neuro.28.061604.135645 [DOI] [PubMed] [Google Scholar]
  89. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, & Matthews PM (2004). Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage, 23, S208–S219. 10.1016/j.neuroimage.2004.07.051 [DOI] [PubMed] [Google Scholar]
  90. Sperber C, & Karnath HO (2017). Impact of correction factors in human brain lesion‐behavior inference. Human Brain Mapping, 38(3), 1692–1701. 10.1002/hbm.23490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  91. Taylor HG, & Alden J (1997). Age-related differences in outcomes following childhood brain insults: an introduction and overview. Journal of the international Neuropsychological Society, 3(6), 555–567. 10.1017/S1355617797005559 [DOI] [PubMed] [Google Scholar]
  92. Thomas MS, & Johnson MH (2008). New advances in understanding sensitive periods in brain development. Current directions in psychological science, 17(1), 1–5. 10.1111/j.1467-8721.2008.00537.x [DOI] [Google Scholar]
  93. Tompkins CA (1990). Knowledge and strategies for processing lexical metaphor after right or left hemisphere brain damage. Journal of Speech, Language, and Hearing Research, 33(2), 307–316. 10.1044/jshr.3302.307 [DOI] [PubMed] [Google Scholar]
  94. Torre GA, Matejko AA, & Eden GF (2020). The relationship between brain structure and proficiency in reading and mathematics in children, adolescents, and emerging adults. Developmental cognitive neuroscience, 45, 100856. 10.1016/j.dcn.2020.100856 [DOI] [PMC free article] [PubMed] [Google Scholar]
  95. Tranel D (2019). Chapter 4. Cognitive Neuroscience and Neuropsychology. In Lim R, Fecto F, & Richerson GB (Eds.). 100 Years of Neurology at the University of Iowa (pp. 62–68). University of Iowa Carver College of Medicine. [Google Scholar]
  96. Tranel D (2009). The Iowa-Benton school of neuropsychological assessment. Neuropsychological assessment of neuropsychiatric and neuromedical disorders, 66–83. [Google Scholar]
  97. Turkeltaub PE, Eden GF, Jones KM, Zeffiro TA (2002). Meta-analysis of the functional neuroanatomy of single-word reading: method and validation. Neuroimage 16(3), 765–780. 10.1006/nimg.2002.1131 [DOI] [PubMed] [Google Scholar]
  98. Turkeltaub PE, Gareau L, Flowers DL, Zeffiro TA, & Eden GF (2003). Development of neural mechanisms for reading. Nature Neuroscience, 6, 767–773. 10.1038/nn1065 [DOI] [PubMed] [Google Scholar]
  99. Uddin LQ, Supekar K, Amin H, Rykhlevskaia E, Nguyen DA, Greicius MD, Menon V (2010). Dissociable connectivity within human angular gyrus and intraparietal sulcus: evidence from functional and structural connectivity. Cereb. Cortex 20(11), 2636–2646. 10.1093/cercor/bhq011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  100. Westmacott R, Askalan R, Macgregor D, Anderson P, & de Veber G (2010). Cognitive outcome following unilateral arterial ischaemic stroke in childhood: effects of age at stroke and lesion location. Developmental Medicine & Child Neurology, 52(4), 386–393. 10.1111/j.1469-8749.2009.03403.x [DOI] [PubMed] [Google Scholar]
  101. Westmacott R, McDonald KP, Roberts SD, deVeber G, MacGregor D, Moharir M, Dlamini N, & Williams TS (2018). Predictors of Cognitive and Academic Outcome following Childhood Subcortical Stroke. Developmental Neuropsychology, 43(8), 708–728. 10.1080/87565641.2018.1522538 [DOI] [PubMed] [Google Scholar]
  102. Wilkinson GS (1993). Wide Range Achievement Test Administration Manual (3rd ed.). Wilmingon, DE: Wide Range. [Google Scholar]
  103. Wilkinson GS, & Robertson GJ (2006). Wide range achievement test (WRAT4) Lutz, FL: Psychological Assessment Resources. [Google Scholar]
  104. Yeh FC, Panesar S, Fernandes D, Meola A, Yoshino M, Fernandez-Miranda JC, … & Verstynen T (2018). Population-averaged atlas of the macroscale human structural connectome and its network topology. Neuroimage, 178, 57–68. 10.1016/j.neuroimage.2018.05.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  105. Younger JW, Tucker-Drob E, & Booth JR (2017). Longitudinal changes in reading network connectivity related to skill improvement. Neuroimage, 158, 90–98. 10.1016/j.neuroimage.2017.06.044 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS1838059-supplement-Supplementary_Material.docx (903.6KB, docx)

RESOURCES