Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2019 Jun 9;40(14):4058–4071. doi: 10.1002/hbm.24685

Impaired structural connectivity between dorsal attention network and pulvinar mediates the impact of premature birth on adult visual–spatial abilities

Maria Berndt 1,2, Josef G Bäuml 1,2, Aurore Menegaux 1,2,3,4, Chun Meng 2,5, Marcel Daamen 7,8, Nicole Baumann 6, Claus Zimmer 1, Henning Boecker 7, Peter Bartmann 8, Dieter Wolke 6,9, Christian Sorg 1,2,10,✉
PMCID: PMC6865666  PMID: 31179600

Abstract

The dorsal attention network (DAN), including frontal eye fields and posterior parietal cortices, and its link with the posterior thalamus, contribute to visual–spatial abilities. Very premature birth impairs both visual–spatial abilities and cortico‐thalamic structural connectivity. We hypothesized that impaired structural DAN‐pulvinar connectivity mediates the effect of very premature birth on adult visual–spatial abilities. Seventy very premature (median age 26.6 years) and 57 mature born adults (median age 26.6 years) were assessed with cognitive tests and diffusion tensor imaging. Perceptual organization (PO) index of the Wechsler Adult Intelligence Scale‐III was used as a proxy for visual–spatial abilities, and connection probability maps in the thalamus, derived from probabilistic tractography from the DAN, were used as a proxy for DAN‐thalamic connectivity. Premature born adults showed decreases in both PO‐index and connection probability from DAN into the pulvinar, with both changes being positively correlated. Moreover, path analysis revealed that DAN‐pulvinar connectivity mediates the relationship between very premature birth and PO‐index. Results provide evidence for long‐term effects of very premature birth on structural DAN‐pulvinar connectivity, mediating the effect of prematurity on adult visual–spatial impairments. Data suggest DAN‐pulvinar connectivity as a specific target of prognostic and diagnostic procedures for visual–spatial abilities after premature birth.

Keywords: cortico‐thalamic connectivity, perceptual organization index, premature birth, probabilistic tractography, visual–spatial abilities

1. INTRODUCTION

Very premature birth (VPB), defined as birth before the completion of 32 weeks of gestation and/or with birth weight below 1,500 g, affects about 1–2% of all live births (Martin et al., 2008; Volpe, 2009a). VPB infants are at increased risk of perinatal brain injury due to hypoxia‐ischemia, infections, inflammatory processes and/or drug exposure (Deng, 2010; Penn, Gressens, Fleiss, Back, & Gallo, 2016). This increased risk has been attributed to the fact that premature infants are born at a time when many body systems (e.g., the respiratory system, the cardiovascular system, the immune system, and the central nervous system) are not fully developed making them vulnerable to superimposed injury (Volpe, 2009b). While for 5–10% of surviving VPB children brain changes lead to major motor deficits (e.g., cerebral palsy), 25–50% of VPB infants do not have major motor impairments but cognitive, attentional, behavioral, and social deficits; therefore, cognitive deficits without major motor deficits are the dominant neurodevelopmental sequelae after VPB (Anderson et al., 2003; Breeman, Jaekel, Baumann, Bartmann, & Wolke, 2015; Eryigit Madzwamuse, Baumann, Jaekel, Bartmann, & Wolke, 2015; Larroque et al., 2008; Platt et al., 2007; Volpe, 2009b; Woodward, Edgin, Thompson, & Inder, 2005). In particular, premature born individuals often show impaired visual–spatial abilities, for example impairments in visual attention, nonverbal reasoning, visual‐motor integration, and visual–spatial problem‐solving (Bohm, Lundequist, & Smedler, 2010; Finke et al., 2015; Foulder‐Hughes & Cooke, 2003; Leung, Thompson, Black, Dai, & Alsweiler, 2018; Marlow et al., 2007; Menegaux et al., 2017). Visual–spatial abilities, in turn, are an important prerequisite for learning and educational and academic performance, and thus highly relevant for socioeconomic success and life quality (Aarnoudse‐Moens, Oosterlaan, Duivenvoorden, van Goudoever, & Weisglas‐Kuperus, 2011; Johnson, Wolke, Hennessy, & Marlow, 2011; Molloy et al., 2017). A detailed understanding of brain mechanisms that mediate the impact of premature birth on visual–spatial abilities is necessary, as it might be a critical starting point for the development of targeted treatments or prognostic procedures. These brain mechanisms, however, are poorly understood. Thus, the aim of the present study was to identify brain systems that may mediate between very premature birth and adult visual–spatial abilities.

Previous studies indicated a relation between premature birth, the development of visual–spatial abilities, and brain regions of the dorsal visual stream, that is, posterior occipital and parietal regions being involved in visual–spatial attention and visual‐motor integration (Atkinson & Braddick, 2007; Chaminade, Leutcher, Millet, & Deruelle, 2013). In terms of large‐scale functional brain networks, posterior parietal regions of the visual stream overlap most with the dorsal attention network (DAN). The DAN is an intrinsic brain network of coherent ongoing brain activity of the frontal eye fields and intra‐parietal sulci, and it underpins attentional top‐down control, spatial attention, and visual‐motor control (Corbetta, Patel, & Shulman, 2008; Corbetta & Shulman, 2002; Vossel, Geng, & Fink, 2014). Critically, the development of visual–spatial abilities is linked with the DAN. For example, a recent study demonstrated that the development of selective visual attention in 4 to 7 years‐old children are reflected in functional connectivity changes between frontal eye fields and intra‐parietal sulci (Rohr et al., 2017). Consequences of premature birth, which impair visual–spatial abilities, might therefore be mediated by impaired integrity of the DAN. Toward formulating a more specific hypothesis of how prematurity might impact the DAN, we have to recall two points—the primary impact of prematurity on corticothalamic connectivity and the role of corticothalamic connectivity for cortico‐cortical communication within the DAN. (a) Previous studies showed that premature birth has a preferential and lasting impact on corticothalamic connectivity (Ball et al., 2013; Ball et al., 2015; Meng et al., 2016). These connections are formed typically late in intra‐uterine development in gestational weeks of about 18–35 under the guidance of both subplate neurons and oligodendrocyte‐progenitor cells (Hoerder‐Suabedissen & Molnar, 2015; Salmaso, Jablonska, Scafidi, Vaccarino, & Gallo, 2014; Volpe, 2009a). The selective vulnerability of subplate neurons and oligodendrocyte‐progenitor cells to hypoxic–ischemic events is thought to lead to impaired corticothalamic connectivity and brain network development after premature birth (Back et al., 2001; Bystron, Blakemore, & Rakic, 2008; Lopez‐Bendito & Molnar, 2003; Volpe, 2009a), with connectivity and network impairments persisting into adulthood (Bauml et al., 2015; Meng et al., 2016; Sripada et al., 2015). (b) Coordinated cortico‐cortical communication within the DAN depends on corticothalamic control, and thereby on structural corticothalamic connectivity (Daitch et al., 2013; Saalmann, 2014; Sherman, 2016). For example, Saalmann, Pinsk, Wang, Li, and Kastner (2012) demonstrated in monkeys, that the pulvinar's structural connectivity with distinct cortical regions of the DAN regulates cortico‐cortical communication according to visual attention selection, demonstrating the pulvinar as part of the DAN.

Based on both the impact of premature birth on corticothalamic connectivity and the role of the pulvinar for DAN cortico‐cortical communication, we hypothesized that structural connectivity between pulvinar and the DAN mediates the impact of premature birth on visual–spatial abilities. As a proxy for visual–spatial functioning, we used the perceptual organization (PO) index of the Wechsler Adult Intelligence Scale (WAIS) (Von Aster & Horn, 2006)—an integrative measure which reflects visual–spatial processes, problem‐solving, nonverbal reasoning and visual‐motor integration (Lange, 2011). As a proxy for structural connectivity between the pulvinar and the DAN, we used connection probability maps in the thalamus, which were derived from diffusion tensor imaging (DTI) data, and probabilistic tractography from the DAN into the thalamus. The DAN was defined for each participant by individual resting‐state functional connectivity maps, derived from resting‐state functional MRI (resting‐state fMRI); the thalamus was defined by canonical anatomical atlas. PO‐index assessment, resting‐state fMRI, and DTI were applied in a sample of 70 VPB and 57 mature‐born (MB) adults at the age of 26 years. Connection probability values were associated with prematurity at birth and PO‐index at the age of 26 years via correlation and mediation analysis. Finally, to investigate the relevance of PO‐index and DAN structural connectivity for educational performance, we explored the possible link between structural alterations, PO‐index, and educational success at the age of 26 years as well as the role of structural alterations as a potential mediator between prematurity and educational success.

2. MATERIALS AND METHODS

2.1. Sample and clinical‐cognitive measures

2.1.1. Participant description

Participants were recruited as part of the prospective Bavarian longitudinal study (BLS) (Riegel, Orth, Wolke, & Osterlund, 1995; Wolke & Meyer, 1999). The BLS investigates a geographically defined whole‐population sample of neonatal at‐risk children and healthy term born controls. All live‐birth infants who were born between January 1985 and March 1986 in Southern Bavaria and required admission to a neonatal unit of 17 children's hospitals within the first 10 days of life comprised the target sample (Wolke & Meyer, 1999). A total of 7,505 children (10.6% of all live births) were classified as neonatal at‐risk children, of whom 2,759 children were born before 37 weeks of gestation (Riegel et al., 1995). During the same period, 916 healthy term infants (>36 weeks gestation; normal postnatal care) born in the same hospitals were recruited as control infants. Over the following years, subjects of both groups were repeatedly assessed via neurological and psychological tests and parental interviews. Full design of the BLS is provided elsewhere (Gutbrod, Wolke, Soehne, Ohrt, & Riegel, 2000; Wolke & Meyer, 1999). At 26 years of age and based on the study design of premature born population versus reference population, all eligible 411 surviving VPB and 308 MB adults, similar regarding the overall distribution of gender, family socioeconomic status (SES), and maternal age, were invited for a follow‐up assessment and of these 260 VBP and 229 MB adults participated at 26 years of age (Eryigit Madzwamuse et al., 2015). A total of 183 subjects underwent structural T1‐ and diffusion‐weighted MRI as well as resting‐state fMRI. MRI assessments were carried out at two different sites: The Department of Neuroradiology, Klinikum Rechts der Isar, Technische Universität München, Germany, and the Department of Radiology, University Hospital Bonn, Germany. The study was approved by the local ethics committees of the Klinikum rechts der Isar and University Hospital Bonn in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments. All study participants gave written informed consent and received travel expenses and a payment for attendance. The following inclusion criteria were applied: normal or corrected‐to‐normal vision, free from non‐correctable reduction of sight in either eye, from medication, from psychiatric or neurological diseases at the assessment or qualitative signs of brain injury (such as ventriculomegaly or polymicrogyria). Exclusion criteria were poor structural, diffusion, or functional MRI data quality, or dropout of MRI acquisition.

A full description of the sample, which underwent DTI (n = 183), and corresponding details about exclusion (n = 29) can be found in (Meng et al., 2016). Further 27 subjects were excluded because of missing resting state fMRI acquisition (n = 11) and/or head motion artifacts (n = 16) in the resting state fMRI acquisition. Thus, the current study sample consists of 70 VPB adults and 57 MB adults (see Table 1 for more details).

Table 1.

Demographics and neuropsychological data

VPB (n = 70) MB (n = 57)
Age (year) Median 26.57 (range 25.72–27.76) Median 26.60 (range 25.58–27.89)
Gender (m/f) n (41/29) n (35/22)
Mean SD IQR Mean SD IQR p‐value
Gestational age (week) 30.23 1.95 29–31 39.63 0.99 39–40 <.001
Birth weight (g) 1,300.57 321.81 1,045–1,496 3,471.39 358.01 3,210–3,715 <.001
INTI 6.44 2.09 4.27–8.09 –
Intracranial hemorrhage n = 13 –
Full‐scale IQ 94.41 12.86 86–100 102.88 12.16 95–111 <.001
PO‐index 93.78 14.11 83–104 106.39 12.11 96–114 <.001
VC‐index 104.23 14.51 95–111 109.54 15.03 98–122 .048
Educational success 2.79 0.85 2–4 3.07 0.88 2–4 .08
MRI center
Scanner MUC n = 45 n = 37 .94
Scanner BN1 n = 3 n = 8 .05
Scanner BN2 n = 22 n = 12 .19
Open in a new tab

Note: Group comparisons: two‐sample t‐tests for age, gestational age, birth weight, full‐scale IQ, PO‐index and VC‐index; Mann–Whitney U test for educational success; chi‐squared statistics for gender and MRI centers (first and second MRI scanner in Bonn as well as the scanner in Munich, respectively coded as BN1, BN2, MUC). Data are presented as mean ± SD as well as the interquartile range.

Abbreviations: VPB, very premature born; MB, mature born; INTI, intensity of neonatal treatment index; IQ, intelligence quotient; PO‐index, perceptual organization index.

2.1.2. Birth‐related variables

Gestational age was estimated from maternal reports of the last menstrual period and serial ultrasounds during pregnancy. Further clinical assessment with the Dubowitz method was applied in case of a variation of these two measures by more than 2 weeks (Dubowitz, Dubowitz, & Goldberg, 1970). Birth weight was obtained from obstetric records. The intensity of neonatal treatment index as a measure of neonatal complications was ascertained by daily assessments of care level, respiratory support, feeding dependency, and neurological status (Casaer P., 1985).

2.1.3. Assessment of cognitive performance and educational success

Cognitive performance at the age of 26 years was assessed by independently trained psychologists using the German version of the Wechsler Adult Intelligence Scale‐III (WAIS‐III), including our measure‐of‐interest, namely PO‐index (Von Aster & Horn, 2006). The psychologists were blinded to group membership. Full‐scale intelligence quotient (IQ) as a measure of global cognitive functioning as well as Verbal Comprehension Index (VC‐index) was additionally derived from WAIS‐III.

Educational and occupational accomplishments were asked at the age of 26 years in a life course interview, similar to the German Socioeconomic Panel Study (Wagner et al., 2007). Educational success was categorized into four levels: 1 = basic educational level (up to 10 years of schooling; no further education or apprenticeship), 2 = middle educational level (up to 10 years of schooling and profession‐oriented education [10+ years]), 3 = higher educational level (high school [12+ years] giving access to higher education, that is, university access qualification), and 4 = university degree (undergraduate or postgraduate).

2.2. MRI data acquisition

MRI data were acquired on 3 T Philips scanners in Munich and Bonn (Achieva [Philips Medical System, Netherlands] and Ingenia [Philips Medical System, Netherlands]) with standard eight channel head coils using consistent sequences and parameter settings across scanners. To account for different scanners in data analyses, scanner identities were included in the analyses as covariates of no interest. (a) For all subjects, diffusion images were acquired using a single‐shot spin‐echo echo‐planar imaging sequence, resulting in one non‐diffusion weighted image (b = 0 s/mm2) and 32 diffusion weighted images (b = 1,000 s/mm2, 32 noncollinear gradient directions) covering whole brain with following parameters: echo time (TE) = 47 ms, repetition time (TR) = 20,150 ms, flip angle = 90°, field of view = 224 × 224 mm2, matrix = 112 × 112, 75 transverse slices, slice thickness = 2 mm, and 0 mm interslice gap, voxel size = 2 × 2 × 2 mm3. (b) A whole‐head, high‐resolution 3D gradient echo T1‐weighted image was acquired using the following parameters: echo time (TE) = 3.9 ms, repetition time (TR) = 7.7 ms, flip angle = 15°, field of view = 256 × 256 mm2, matrix = 256 × 256, 180 sagittal slices, slice thickness = 1 mm, and 0 mm interslice gap, voxel size = 1 × 1 × 1 mm3. (c) Resting‐state fMRI data were collected for 10 min 52 s from a gradient‐echo echo‐planar sequence (TE = 35 ms, TR = 2,608 ms, flip angle = 90°, FOV = 230 mm2, matrix size = 64 × 63, 41 slices, thickness 3.58 and 0 mm interslice gap, reconstructed voxel size = 3.59 × 3.59 × 3.59 mm3) resulting in 250 volumes of blood oxygenation level dependent (BOLD) fMRI data for each subject. Immediately before undergoing the resting‐state sequence, subjects were instructed to keep their eyes closed and to restrain from falling asleep.

2.3. Individual DAN definition via resting‐state fMRI analysis

To identify for each participant individual DAN as the mask for subsequent structural connectivity tractography, we performed resting‐state fMRI data analysis, including group independent component analysis and dual regression analysis via FSL (FMRIB Software Library, Oxford, UK, https://fsl.fmrib.ox.ac.uk/fsl/fslwiki, (Jenkinson, Beckmann, Behrens, Woolrich, & Smith, 2012)). For each participant, the first five functional scans of each resting‐state fMRI session were discarded due to magnetization effects. Data were then preprocessed using FSL (Jenkinson et al., 2012). Functional volumes were realigned to correct for head motion (Jenkinson et al., 2012) and non‐brain tissue was removed using BET (Smith, 2002). Following, spatial smoothing was applied using a Gaussian kernel of FWHM 5 mm, and data were high‐pass temporal filtered (200 s). Each subject's fMRI data were subsequently coregistered to that subject's high‐resolution structural T1 image by using boundary‐based registration (Greve & Fischl, 2009). The remaining data were then transformed to Montreal Neurological Institute (MNI) space at 2 × 2 × 2‐mm resolution using nonlinear registration (Andersson, Jenkinson, & Smith, 2007). To ensure comparable data quality across groups, particularly concerning motion‐induced artifacts, point‐to‐point head motion was estimated for each subject (Van Dijk, Sabuncu, & Buckner, 2012). Point‐to‐point motion was defined as the absolute displacement of each brain volume compared to its previous volume. Two‐sample t‐tests yielded no significant differences between groups regarding mean point‐to‐point translation or rotation of any direction (p = .43).

Individual DAN maps were derived from resting‐state fMRI data following a multi‐step procedure: first, we parcellated the cortex into distinct intrinsic brain networks (IBNs) using independent component analysis (ICA). For this purpose, preprocessed data from both groups were concatenated in time and entered into a single group ICA framework as implemented in FSL MELODIC (Beckmann, DeLuca, Devlin, & Smith, 2005). We used a fixed dimensionality of 25 spatially independent components. To identify the DAN component, the spatial correlations between the DAN template, provided by Allen et al. (2011), and the components identified in our sample were calculated. The sub‐component with the highest correlation score (Allen IC72; r = 0.32) was used for further analysis. The DAN component was then mapped back to each subject's dataset through dual regression (Filippini et al., 2009) to obtain subject‐specific DAN maps. Specifically, back‐projected maps were thresholded (Z > 5) and the resulting DAN maps were used as regions‐of‐interest (ROIs) for fiber tracking.

We performed the same procedure in order to identify further networks (namely, the posterior and anterior Default Mode Network (DMN, IC50/53/25), sensorimotor network (SMN, IC23/24/29), salience network (SN, IC55), auditory network (IC17), visual (IC46), and lateral visual network [IC48] and frontal networks [IC42/20]), which were used later to test for specificity of impaired structural connectivity between DAN and pulvinar and its mediating role for the effect of VPB on adult PO‐index.

2.4. Functional connectivity analysis—Control analyses

  1. In order to control that DANs across groups were comparable and do not confound group differences in structural connectivity between DAN and thalamus, we compared DANs' intra‐network connectivity. We used ICA/dual‐regression based individual DAN maps as proxies for intra‐network functional connectivity of the DAN, and applied firstly, one sample t‐tests for DANs' of MB and VPB adults, respectively, to visualize DAN of each group (see Figure S1). After that we used a two‐sample t‐test to compare DANs' functional connectivity across groups.

  2. In order to control whether structural connectivity changes between DAN and pulvinar in the VPB group were specific for modality, we analyzed the functional connectivity between DAN and thalamus and compared it across MB and VPB groups. To define corticothalamic functional connectivity, we relied on previously applied partial correlation approach of our group on functional connectivity as described in (Avram, Brandl, Bauml, & Sorg, 2018). More specifically, we used a voxel‐wise partial correlation approach to map intrinsic functional connectivity (iFC) between each voxel in the thalamus and characteristic time series associated with a specific cortical ROI, in each case regressing out the time series of all other cortical ROIs, the signal in white matter and cerebrospinal fluid as well as head motion parameters (O'Reilly, Beckmann, Tomassini, Ramnani, & Johansen‐Berg, 2010). These analyses were performed in native space to reduce spatial inaccuracy introduced by suboptimal normalization (O'Reilly et al., 2010). For each subject the component maps (i.e., the ROIs) were thresholded at Z > 5, and the remaining voxels inside the mask were used to extract the first Eigen time series from the preprocessed individual functional data. In a second step, masks representing white matter and cerebrospinal fluid were created using FSL tissue segmentation tool FAST (Zhang, Brady, & Smith, 2001) and transformed into individual functional space. Both WM and CSF masks were thresholded at p > .5, and the remaining voxels inside each mask were used to extract the first Eigen time series from the preprocessed individual functional data. For every subject, these procedures resulted in 12 component Eigen time series (i.e., one per component) and second Eigen time series representing WM and CSF. In the last step, we extracted the subject‐specific six head motion parameters, which were estimated with MCFLIRT (Jenkinson, Bannister, Brady, & Smith, 2002). Finally, a thresholded thalamus mask (p > .3) was created using the Harvard‐Oxford Subcortical Structural Atlas in FSL and transformed back into individual functional space. This partial correlation procedure resulted in one partial r‐map per component and subject (total: 12 × 176) that were subsequently converted to Z values using Fisher's r‐to‐z transformation.

Before performing group analyses, thalamic Z maps were transformed back into Montreal Neurological Institute space using nonlinear registration implemented in FMRIB's nonlinear image registration tool (FNIRT). To statistically evaluate network‐specific thalamic iFC, for each group (term and preterm) and each cortical ROI, we calculated voxel‐wise one‐sample t‐tests on participants' thalamic Z maps using SPM 8 controlling for sex and scanner‐ID (p < .05, FDR corrected). Group differences in DAN‐thalamus iFC were tested via voxel‐wise two‐sample t‐tests using SPM 8 controlling for sex and scanner‐ID (p < .05, FDR corrected).

To relate between group differences in DAN‐thalamus iFC to PO‐indices, we used the following approach: voxels whose iFC differed significantly between groups were saved as images in SPM 8. After binarization, ROI‐restricted network‐specific iFC was extracted and averaged subject‐wise with Matlab‐based in‐house software. This procedure resulted in one mean iFC Z‐value per cluster and subject. Then, this thalamocortical iFC was used to predict PO‐index within a multiple regression model using SPSS 24 with further regressors being sex and scanner‐ID.

2.5. Diffusion MRI data processing

DTI data were processed using FSL (Jenkinson et al., 2012) applying the following steps: DTI preprocessing including eddy current and head motion correction by coregistrating all diffusion‐weighted images to b0 image, DTI data quality check based on visual inspection by several independent raters (B.M., B.J., C.S.), reconstruction of the distribution of diffusion directions at each voxel, probabilistic tractography, and analysis of connection probability as explained in the following.

2.5.1. DTI preprocessing and quality check

Diffusion data were preprocessed using FSL's FDT toolbox. Eddy‐current distortion and head motion were corrected by linear registration of all diffusion‐weighted images to the first b0 volume using FSL “eddy_correct.” Brain‐tissue extraction was performed by removing the skull and non‐brain tissue using FSL BET (Smith, 2002).

Each image was visually checked by three independent raters (C.M., C.S., A.M.) and fitting residuals (the sum‐of‐squared‐error maps generated by FSL DTIFIT) were used to identify data corrupted by artifacts. Detailed information about the exclusion process can be found in Meng et al. (2016).

2.5.2. Seed and target definition for tractography

Before applying probabilistic tractography, seed and target ROIs were defined, based on the study's hypothesis. The seed for tractography was placed in individual DAN derived from resting‐state fMRI data analysis (see above), its target into the bilateral thalamus. The target ROI was taken from the Oxford thalamic connectivity atlas, based on the probability of anatomical connections between thalamus and cortex in MNI space (Behrens et al., 2003).

Each subject's native 3D T1‐weighted image was brain extracted by removing the skull and non‐brain tissue using FSL BET (Smith, 2002) as well as bias‐corrected using FMRIB's Automated Segmentation Tool (FAST). A two‐step transformation procedure was used to register these T1‐weighted images to the MNI ICBM 152 nonlinear (sixth generation) symmetric standard‐space T1‐weighted average structural template image (Grabner et al., 2006). In the first step, linear (affine) transformation was performed using FLIRT, then the generated affine transform was used as a guide for nonlinear registration using FNIRT. The output of this transformation procedure was an individual native to standard (MNI space) nonlinear warp field.

The brain‐extracted b0‐image of each subject was registered to the corresponding brain extracted and bias‐corrected T1 image in native space using linear registration (FLIRT) with 12 degrees of freedom and correlation ratio cost function. This individual diffusion‐to‐native transformation matrix was combined with the above described native‐to‐standard nonlinear transformation matrix, that resulted in diffusion‐to‐standard space transformations and their corresponding inverses. These warp‐fields were then applied to the ROI masks of seed and target (in standard MNI space), namely DAN and thalamus, to align them into each subject's individual diffusion space with the implementation of the nearest neighbor interpolation, followed by a visual check by an experienced neuroradiologist (M.B.).

2.5.3. Probabilistic fiber tracking and connection probability

Probabilistic tractography was used to get cortico‐subcortical connection probabilities between DAN and thalamus (Figure 2a). To estimate the fiber orientations, the distribution of diffusion directions at each voxel was reconstructed by the use of BEDPOSTX (Bayesian Estimation of Diffusion Parameters Obtained using Sampling Techniques, FSL) with two crossing fibers modeled per voxel (Behrens et al., 2007). Probabilistic tractography was performed using PROBTRACKX2 (probabilistic tracking with crossing fibers, FSL) using the following parameters: number of samples = 5,000, step length = 0.5 mm, curvature‐threshold = 0.2, subsidiary fiber volume fraction threshold = 0.01. By creating probabilities of an existing path in the diffusion field of each subject between individual DAN, used as seed ROI, and thalamus, used as waypoint ROI, a surrogate measure of white matter connectivity was created (Behrens et al., 2007; Jbabdi & Johansen‐Berg, 2011). As a previously used step of normalization, the resulting individual streamline maps were divided by the total number of generated tracts of each individual tractography (“waytotal”), taking into account the high inter‐subject variability of these tracts as well as ROI‐sizes (Arnold et al., 2012; Rus et al., 2017; Zhang, Snyder, Shimony, Fox, & Raichle, 2010).

Figure 2.

Figure 2

Open in a new tab

Structural connectivity between dorsal attention network and thalamus–probabilistic tractography and group comparisons. (a) Seeds of probabilistic tractography in individual DAN (in red, spatial maps are generated via independent component analysis and dual regression analysis of resting state fMRI data) and target in the thalamus (in blue, atlas based on Behrens et al., 2003), and an example of connectivity distributions (produced via probtrackx [FSL] (Behrens, Berg, Jbabdi, Rushworth, & Woolrich, 2007)) on the right, displayed in individual diffusion space of a representative VPB adult. (b) Connection probability (between DAN and thalamus) with DAN of the thalamic voxels (one sample t‐test for MB and VPB adults on a significance level of .05, FWE corrected, extracted from SPM8), the peak is marked in the pulvinar, overlaid to a T1‐weighted sequence in MNI space. (c) Thalamic voxels of reduced structural connection probability with the DAN in VPB adults compared to MB adults, located in the right pulvinar (two‐sample t‐test, p < .05, FWE corrected), overlaid to a T1‐weighted sequence in MNI space. Abbreviations: VPB, very premature born; MB, mature born [Color figure can be viewed at http://wileyonlinelibrary.com]

Based on Behrens et al. (2003), a map of thalamic voxels containing information about the probability of connectivity with the DAN was extracted for each subject. That was used for identification of thalamic substructures that exhibit group differences, and also for correlation analyses between these connectivity values and clinical‐neuropsychological data. To enable inter‐subject comparisons, the thalamic connectivity maps were warped into MNI space using the above‐mentioned diffusion‐to‐standard space transformations with the implementation of the nearest neighbor interpolation.

2.6. Statistical analysis

Concerning voxel‐wise connection probability maps in the thalamus, voxel‐wise one‐ and two‐sample t‐tests were performed for VPB and MB adults, using SPM 8 and controlling for sex and scanner identities (p < .05, FWE‐cluster corrected). Connection probability values of group difference clusters were extracted and averaged using Matlab‐based in‐house software and used for further correlation and mediation analyses.

2.6.1. Mediation analysis

Mediation analysis was carried out to test whether structural connectivity alterations between individual DAN and pulvinar mediate the effect of group (VPB vs. MB adults) on PO‐index. In the mediation model, group affiliation was entered as the causal variable, PO‐index as the outcome variable and residuals from structural connectivity values as the mediator variable. These residuals were calculated by regressing out the effects of sex and scanner identities, in order to control for effects of these variables on mediation. Path coefficients were estimated using (unstandardized) regression coefficients of multiple regression analyses. Statistical significance of the indirect pathway, reflecting the impact of mediation, was evaluated using a non‐parametric bootstrap approach with 10,000 replication samples to obtain a 95% confidence interval (Hayes & Rockwood, 2017; Preacher & Hayes, 2004).

2.6.2. Analyzing the relevance for educational success

To explore, whether adult PO‐index and DAN‐pulvinar connectivity might be relevant for educational performance after premature birth, we tested the relationship between prematurity, educational success, PO‐index, and DAN‐pulvinar connectivity, respectively. Mann–Whitney‐U‐test was used to estimate the impact of premature birth on educational success. Spearman rank correlation was applied to quantify the relationship between PO‐index and educational success. After performing linear regression to get residuals of the connectivity values (corrected for sex and scanner identities), Spearman rank bivariate correlation was performed with PO‐index and educational success. Mediation analysis was carried out to test whether structural connectivity alterations between individual DAN and pulvinar mediate the effect of group (VPB vs. MB adults) on educational success.

3. RESULTS

3.1. Lowered perceptual organization index in very premature born adults

PO‐index in VPB adults (93.78 ± 14.11) was lower than in MB adults (106.39 ± 12.11, p < .001, two‐sample t‐test) (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Perceptual organization index in very premature and mature born adults. The boxplot shows the median and interquartile range of PO‐indices, a sub‐scale of the Wechsler Adult Intelligence Scale‐III, for VPB adults on the left (96, 83–104) and MB adults on the right (107, 96–114). Groups are different in PO‐index (two‐sample t‐test, p < .001). Abbreviations: VPB, very premature born; MB, mature born; PO‐index, perceptual organization index

3.2. Reduced connection probabilities from the DAN to the pulvinar of very premature born adults

Qualitatively, connection probability maps of the DAN into the thalamus were similar between VPB and MB groups, justifying group comparison (Figure 2b; one‐sample t‐tests, p < .05 FWE‐cluster corrected). In more detail, peak voxels of both groups were located in the pulvinar and the overall pattern covered posterior thalamic parts. However, quantitative two‐sample t‐test revealed decreased connection probability from the DAN to the pulvinar in VPB adults (p < .05, FWE‐cluster corrected) (Figure 2c). To control that group different DAN‐pulvinar structural connectivity was not due to group different DANs as defined by functional connectivity (see Figure S1), we compared DAN functional connectivity maps across groups. We did not find any group difference, suggesting that DANs were comparable across groups, with distinct spatial outline of DANs being unlikely to confound group differences in DAN‐pulvinar structural connectivity.

3.3. Structural connectivity changes mediate the effect of prematurity on perception

In a mediation analysis (Figure 3), the effect of group (VPB vs. MB) on adult PO‐index (total effect c = 12.61 ± 2.38, p < .001) was still present, but reduced when controlling for DAN‐pulvinar connection probability (direct effect c' = 10.84 ± 2.55, p < .001); critically, the bootstrapped 95% confidence interval for the indirect effect (i.e., mediation: total—direct effect) was different from zero (CI: 0.26–4.03), indicating that DAN‐pulvinar connectivity significantly mediated the relationship between very premature birth and PO‐index performance.

Figure 3.

Figure 3

Open in a new tab

Structural connectivity changes mediate the effect of prematurity on perception. Structural connectivity alterations between thalamus (located in right pulvinar) and individual DAN mediate the effect of group (VPB vs. MB adults) on PO‐index. Paths are labeled with the (unstandardized) regression coefficients of the respective effect (± SE). a, effect of causal variable on mediator; b, direct effect of mediator on outcome; c, total effect of causal variable on outcome; c', direct effect of causal variable on outcome. Abbreviations: VPB, very premature born; MB, mature born; PO‐index, perceptual organization index

3.3.1. Control analyses—Specificity

  1. Next, we tested whether the mediating role of impaired DAN‐pulvinar connectivity for the association between premature birth and visual–spatial functioning was specific for visual–spatial functioning that is, PO‐index. Therefore, we repeated our mediation approach on DAN‐pulvinar connectivity but now focusing on verbal instead of nonverbal visual–spatial abilities. Specifically, we performed again a mediation analysis for DAN‐pulvinar connectivity but with the VC‐index of the WAIS as outcome/dependent variable, which reflects verbal abilities. Verbal functioning is thought to depend more on lateral fronto‐parieto‐temporal networks than the dorsal fronto‐parietal DAN (Lau, Phillips, & Poeppel, 2008). First, we found a significant reduction of the VC‐index in the VPB group (p < .05; Table 1). Then we performed path analysis for VC index in the exact same way as for PO index. Critically, we found, that the bootstrapped 95% confidence interval for the indirect effect of DAN‐pulvinar on VC index covered the zero (CI: −1.35 – 3.27), indicating that DAN‐pulvinar connectivity does not mediate the link between prematurity and verbal intelligence. This finding supports the notion of the specificity of DAN‐pulvinar connectivity to mediate prematurity effects on PO‐index, with respect to other cognitive impairments such as verbal abilities in VPB.

  2. Next we tested whether DAN‐pulvinar connectivity plays a specific role for distinguishing VPB from MB adults. Therefore, we tested global network connectivity, defined by averaging the connectivities of all networks, except DAN‐pulvinar, for each individuum at thalamus connectivity map level, and searched for group difference in the global connectivity map between VPB and MB adults. No group difference (p < .05, FWE‐cluster corrected) could be found, indicating no main difference in the global cortico‐thalamic connectivity between VPB and MB adults. By analyzing the single networks (such as posterior and anterior DMN, sensorimotor network [SMN], salience network [SN], auditory network, visual networks, and frontal network) and their connectivity to the thalamus, we found decreased connection probability from the lateral visual network, the salience network, and the anterior DMN into the posterior thalamus for VPB adults (p < .05, FWE‐cluster corrected).

However, the relationships between PO‐index and the group different structural connectivity probability clusters from the lateral visual network into the thalamus (r = 0.05, p = .66), anterior DMN (r = 0.001, p = .99), and the salience network (r = −0.23, p = .06), respectively, were not significant, supporting the specificity of DAN‐pulvinar connectivity and its link with PO‐indices.

  • 3.

    Finally, we tested whether the mediating role of DAN‐pulvinar structural connectivity for the VPB effect on PO‐index was specific with respect to DAN‐pulvinar structural connectivity. As control modality of connectivity, we chose functional connectivity between DAN and the thalamus. Voxel‐wise two‐sample t‐tests indicated that DAN‐thalamus iFC was not significantly different between MB and VPB adults (see Figure S2). However, our results suggest a trend to significance (p = .071) where VPB adults show an increased DAN‐pulvinar iFC. Moreover, correlation analyses indicated that aberrant DAN‐pulvinar iFC was not associated with PO‐index measures, neither within term (r = −0.11, p = .40), nor within VPB adults (r = −.01, p = .94). This result indicates that specifically the impaired structural connectivity between DAN and pulvinar is linked with PO‐indices.

3.4. DAN‐pulvinar connectivity, premature birth, and educational success

Finally, to investigate the relevance of reductions in both visual–spatial abilities and related DAN‐pulvinar connectivity for educational performance after premature birth, we analyzed whether PO‐index and DAN‐pulvinar connectivity are related to educational success after premature birth. First, we found a positive correlation between PO‐index and educational success in VPB adults (r = 0.3, p = .012), demonstrating a link between visual–spatial abilities and educational success. Then, we saw that DAN‐pulvinar connectivity was associated with both PO‐index (r = 0.27, p = .03) and educational success (r = 0.28, p = .02) for VPB adults. This final result supports the idea that prematurity, educational success, and PO‐index abilities are linked by impaired DAN‐pulvinar connectivity. Indeed, in a mediation analysis, the effect of group (VPB vs. MB) on educational success (total effect c = 0.28 ± 0.15, p = .06) was still present, but reduced when controlling for DAN‐pulvinar connection probability (direct effect c' = 0.13 ± 0.16, p = .43). The bootstrapped 95% confidence interval for the indirect effect was different from zero (CI: 0.04–0.30), indicating that DAN‐pulvinar connectivity significantly mediated the relationship between very premature birth and educational success.

In order to estimate which type of ability is more linked to educational success, we compared correlation coefficients for both VC‐ and PO‐indices by the use of Fisher‐z‐testing. We found no significant differences between the two correlations (z = −1.7, p = .09). This indicates that not only visuo‐motor but also verbal comprehension abilities are relevant for educational success.

4. DISCUSSION

Using cognitive assessment and diffusion MRI in very premature and mature born adults, we tested the hypothesis that aberrant structural connectivity between the pulvinar and dorsal attention network mediates the effect of very premature birth on impaired visual–spatial functioning in adulthood. In VPB adults, the perceptual organization index, which reflects visual–spatial skills and problem‐solving, was reduced, as well as DAN‐pulvinar connectivity. Critically, DAN‐pulvinar connectivity partially mediated specifically the link between prematurity and adult PO‐index. To the best of our knowledge, this result is the first one outlining a specific neural system mediating adverse effects of very premature birth on visual–spatial functioning. Our finding defines a specific target—DAN‐pulvinar connectivity—to develop focused treatment or prognostic strategies to improve long‐term visual–spatial functioning after premature birth. Next, we discuss these points in more detail.

4.1. Aberrant DAN‐pulvinar connectivity mediates prematurity effects on adult visual–spatial abilities

The main result of the current study is that impaired DAN‐pulvinar connectivity mediates the effect of very premature birth on adult PO‐index (Figure 3). This finding is based on three elements: (a) the effect of prematurity on PO‐index, (b) impaired DAN‐pulvinar connectivity, and (c) the proper mediation function of this connectivity.

4.1.1. Impaired PO‐index after premature birth

PO‐index was reduced in VPB adults (Figure 1). PO‐index is derived from the Wechsler Adult Intelligence Scale (WAIS) (Von Aster & Horn, 2006), and measures visual–spatial processes and problem‐solving, nonverbal fluid reasoning, and visual‐motor integration (Lange, 2011). Therefore, PO‐index is an integrative measure of visual–spatial functioning, which interacts with more specific functions such as visual attention (including for example visual short‐term memory capacity or top‐down attentional weighting) or visuo‐motor integration (including for example control of eye movements) (Kimchi, 2009). Impaired visual attention and visuo‐motor integration, in turn, are common findings in premature born children (Butcher et al., 2012; Foulder‐Hughes & Cooke, 2003; Goyen, Lui, & Hummell, 2011; Marlow et al., 2007) and seem to persist into adulthood (Breeman et al., 2015). Correspondingly, specifically impaired visual short‐term memory capacity was found for a subsample of the current sample (Finke et al., 2015). It might be that these attentional and visuo‐motor deficits trace back to early impairments in newborns; for example, Papageorgiu and colleagues demonstrated that the individual differences in eye fixation duration of newborns predict attentional and visuo‐motor control deficits in childhood (Papageorgiou et al., 2014).

4.1.2. Impaired DAN‐pulvinar structural connectivity

We found reduced connection probabilities in the pulvinar from the DAN in VPB adults (Figure 2). In the mature born adults, peak values of connection probabilities from the DAN focused on the posterior thalamus (Figure 2b), being in line with both anatomical connections between the posterior parietal cortex and the pulvinar (Fischer & Whitney, 2012) and recent diffusion‐based tractography results of thalamo‐cortical connectivity (Kumar, van Oort, Scheffler, Beckmann, & Grodd, 2017). In the VPB group, connection probability maps were qualitatively similar to those of the mature born group, indicating that DAN‐thalamus connectivity was not substantially re‐organized in the VPB group, but with changes being focused on the pulvinar (Figure 2b,c). Our analysis of corticothalamic structural connectivity was based on individually defined DANs via resting‐state fMRI and corresponding ICA/dual regression analysis (Figure 2a). We performed this approach in order to increase individual specificity of DAN tractography. One should note that such an approach includes the possibility for individual DAN detection by resting‐state fMRI, a possibility that might be relevant for individual treatments of the DAN‐pulvinar system (see below for more details). The finding of impaired DAN‐pulvinar structural connectivity corresponds with more generally impaired thalamocortical structural connectivity after premature birth, as explicitly shown in premature born newborns (Kostovic & Judas, 2010) and children (Ball et al., 2013; Ball et al., 2015), and implicitly suggested in adults (Meng et al., 2016; Sripada et al., 2015). In more detail, only extended white matter changes have been demonstrated for premature born adults so far, but the current study demonstrates directly impaired thalamocortical connectivity.

4.1.3. Impaired DAN‐pulvinar connectivity as mediator for the prematurity effect on PO‐index

DAN‐pulvinar connectivity partly mediated the effect of very prematurity on PO‐index (Figure 3). The mediating role of DAN‐pulvinar structural connectivity was threefold specific—specific for PO‐index but not for reduced VC‐index, specific for the DAN but not for global connectivity or other networks—thalamic connectivities, and specific for structural connectivity but not for the functional connectivity between DAN and the thalamus. The absent group differences between VPB and MB adults for global connectivity values show that the DAN‐pulvinar connectivity changes are specific for VPB and not simply a reflection of a wider reduction of cerebral connectivity. The functional connectivity analysis showed a trend to increased DAN‐pulvinar iFC for VPB adults, maybe due to increased variance in regional BOLD fluctuations on the basis of a structural integrity change with decreased structural connectivity.

Furthermore, impaired DAN‐pulvinar connectivity was unlikely confounded by distinct DANs across groups, as we did not find any group differences for DANs' functional connectivity maps. The pulvinar is known to control cortico‐cortical communication, especially within the dorsal attention network (Saalmann, 2014; Shipp, 2003). For example, in monkeys, the pulvinar's structural connectivity with the DAN is associated with attention selection via controlling cortico‐cortical communication (Saalmann et al., 2012). Reduced structural connectivity between DAN and pulvinar after premature birth may reflect impaired development of thalamo‐cortical connectivity and consequently impaired coordination of cortico‐cortical communication. For example, Ball and colleagues showed that impaired thalamo‐cortical connectivity of neonates was predictive for cognitive performance at the age of 2 years (Ball et al., 2013; Ball et al., 2015). Future studies have to test this idea in more detail.

One should note that our analysis revealed a partial mediation effect of DAN‐pulvinar on the association between prematurity and PO‐index, suggesting further brain mechanism to mediate prematurity effects on visual–spatial functioning. Due to the complexity of visual–spatial functioning in general and PO‐index in particular, it is not surprising that multiple mechanisms underpin these abilities. For example, we previously found impaired visual attention, specifically reduced visual short‐term memory capacity in VPB adults, that was associated with impaired integrity of connections between the posterior thalamus and the primary occipital gyri (Menegaux et al., 2017). This result indicates that structural connectivity of more posterior occipito‐thalamic systems also contributes to prematurity effects on visuo‐motor functioning. Furthermore, beyond structural connectivity, coherence of ongoing slowly fluctuating activity in both DAN and occipital networks covering primary occipital gyri has been shown to modulate impaired visual short‐term memory after premature birth, indicating that physiological mechanisms on top of structural connectivity link with impaired visual–spatial abilities in the context of prematurity (Finke et al., 2015). An example for such physiological mechanism might be alterations in local temporal variability of ongoing activity within the thalamus—as being found in premature born adults (Shang et al., 2018)—which impair functional integration in corticothalamic networks (Garrett, Epp, Perry, & Lindenberger, 2018). It is clear, that this list of potential mechanisms is not exhaustive; in order to increase potential targets for “clinical” procedures such as treatment or prognostics of visual–spatial deficits, future studies are necessary to reveal further brain mechanisms mediating prematurity effects on visual–spatial functioning.

4.2. Relevance for educational success

Visual–spatial abilities are basic abilities for other cognitive functions, learning, and thereby for educational success, which in turn are important determinants of socioeconomic status (Fernandes et al., 2016; Haapala et al., 2014; Piaget, 1952). In line with this knowledge, we found a positive correlation between PO‐index and educational success in adulthood, suggesting a link between PO‐index relevant abilities such as visual–spatial skills, problem‐solving, and educational performance.

By comparing correlation coefficients for both VC‐ and PO‐indices, no significant differences indicate that bot visuo‐motor, and verbal comprehension abilities are relevant for educational success. This result specifies our finding for PO‐index and its dependence on DAN‐pulvinar connectivity further, in terms of partial relevance amongst other factors such as verbal comprehension.

Furthermore, DAN‐pulvinar connectivity, that was found to be impaired for VPB adults, is positively correlated with PO‐index and educational success. This finding suggests that structural connectivity alterations caused by prematurity contribute to perceptual as well as educational outcome.

A link between premature birth and less favorable academic outcome has been demonstrated for several times (Aarnoudse‐Moens et al., 2011; Johnson et al., 2011; Leung et al., 2018; Molloy et al., 2017). In the present study, this correlation between prematurity and lower educational outcome in adulthood was mediated by adult DAN‐pulvinar connectivity. This finding together with the role of DAN‐pulvinar connectivity for visual–spatial functioning after premature birth suggests the model that prematurity‐induced changes in DAN‐pulvinar connectivity lead to impaired visual–spatial functioning, which in turn is relevant for educational success. This model is supported by above‐mentioned findings of Ball and colleagues, indicating the predictive role of neonatal thalamocortical connectivity for later attention functioning (Ball et al., 2015). Furthermore, changes in functional connectivity within the DAN were linked with both DAN development from 4–7 years and corresponding development in selective attention of 4–7 years old girls (Rohr et al., 2017).

Integrating these findings, we suggest that DAN‐pulvinar connectivity after premature birth might have some potential for targeted clinical procedures such as specific treatment of developing visual–spatial and visual‐motor functioning or predicting the development of visual–spatial abilities at birth. (a) Concerning treatment: Targeted brain stimulation, especially transcranial magnetic stimulation (TMS) of cortical DAN structures might be of interest when combined with appropriate visual‐motor training. (b) Concerning prediction of visual–spatial abilities at birth: use at birth DAN‐pulvinar structural connectivity could be helpful to predict later visual–spatial abilities, that is in line with the above‐mentioned studies. However, based on these preceding as well as current findings, further studies are needed to derive a concrete concept useful for the prediction of visual–spatial abilities.

4.3. Strength and limitations

Some points should be considered when interpreting our results. First, the current sample might be biased to VPB adults with less severe neonatal complications and less functional impairments. Individuals with stronger birth complications and/or severe lasting impairments in the initial BLS sample were more likely to be excluded in the initial screening for MRI or to reject MRI scanning or even continuation in the study. Thus, reported differences in outcomes including DAN‐pulvinar connectivity between VPB and MB adults are conservative estimates. Second, the current sample size is rather large (70 VPB, 57 MB), enhancing the generalizability of our findings.

Third, the study was performed with the existing methodologies of structural connectivity analysis to the point of time when acquiring the data and postprocessing them with well‐established software. Recently developed techniques are not possible to apply on these data but should be performed for consequential studies in the future, for example, correction for susceptibility‐induced geometric distortions.

A further limitation of the study is the use of three distinct MRI scanners to acquire diffusion data. However, all of them were 3 T Phillips scanners of the same type with standard eight channel head coils using identical sequences and parameter settings. Furthermore, to account for potential confounds induced by different scanners, scanner identities were statistically integrated as covariates of no interest in distinct data analyses, demonstrating no essential influence of scanner types in our findings.

5. CONCLUSION

Impaired DAN‐pulvinar structural connectivity makes a major contribution to the impact of very prematurity.

Supporting information

Supplemental Figure 1 One sample t‐tests of the DAN for MB and VPB adults (p < .05, FWE cluster‐corrected); Yellow: MB adults only; Red: VPB adults only; Orange: overlap of MB and VPB adults

Supplemental Figure 2: Voxel‐wise two sample t‐tests of DAN‐thalamus functional connectivity between MB and VPB adults; p = .071, FWE‐cluster corrected

Click here for additional data file. (342.2KB, docx)

ACKNOWLEDGMENTS

We thank all current and former members of the Bavarian Longitudinal Study Group who contributed to general study organization, recruitment, and data collection, management, and subsequent analyses, including (in alphabetical order): Barbara Busch, Stephan Czeschka, Claudia Grünzinger, Christian Koch, Diana Kurze, Sonja Perk, Andrea Schreier, Antje Strasser, Julia Trummer, and Eva van Rossum. We are grateful to the staff of the Department of Neuroradiology in Munich and the Department of Radiology in Bonn for their help in data collection. Most importantly, we thank all our study participants and their families for their efforts to take part in this study. This work was supported by Deutsche Forschungsgemeinschaft “(SO 1336/1‐1 to C.S.)”, German Federal Ministry of Education and Science (BMBF 01ER0801 to P.B. and D.W., BMBF 01ER0803 to C.S.) and the Kommission für Klinische Forschung, Technische Universität München (KKF 8765162 to C.S). The authors declare no conflict of interest. Data is available on request from the authors.

Berndt M, Bäuml JG, Menegaux A, et al. Impaired structural connectivity between dorsal attention network and pulvinar mediates the impact of premature birth on adult visual–spatial abilities. Hum Brain Mapp. 2019;40:4058–4071. 10.1002/hbm.24685

Funding information Bundesministerium für Bildung und Forschung, Grant/Award Number: BMBF 01ER0801; Deutsche Forschungsgemeinschaft, Grant/Award Number: SO 1336/1‐1

REFERENCES

  1. Aarnoudse‐Moens, C. S. , Oosterlaan, J. , Duivenvoorden, H. J. , van Goudoever, J. B. , & Weisglas‐Kuperus, N. (2011). Development of preschool and academic skills in children born very preterm. The Journal of Pediatrics, 158(1), 51–56. [DOI] [PubMed] [Google Scholar]
  2. Allen, E. A. , Erhardt, E. B. , Damaraju, E. , Gruner, W. , Segall, J. M. , Silva, R. F. , … Calhoun, V. D. (2011). A baseline for the multivariate comparison of resting‐state networks. Frontiers in Systems Neuroscience, 5, 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anderson, P. , Doyle, L. W. , & Victorian Infant Collaborative Study, G . (2003). Neurobehavioral outcomes of school‐age children born extremely low birth weight or very preterm in the 1990s. JAMA, 289(24), 3264–3272. [DOI] [PubMed] [Google Scholar]
  4. Andersson, J. L. , Jenkinson, M. , & Smith, S. (2007). Non‐linear registration, aka spatial normalisation FMRIB technical report TR07JA2 (p. 2). FMRIB Analysis Group of the University of Oxford: Oxford, United Kingdom. [Google Scholar]
  5. Arnold, J. F. , Zwiers, M. P. , Fitzgerald, D. A. , van Eijndhoven, P. , Becker, E. S. , Rinck, M. , … Tendolkar, I. (2012). Fronto‐limbic microstructure and structural connectivity in remission from major depression. Psychiatry Research, 204(1), 40–48. [DOI] [PubMed] [Google Scholar]
  6. Atkinson, J. , & Braddick, O. (2007). Visual and visuocognitive development in children born very prematurely. Progress in Brain Research, 164, 123–149. [DOI] [PubMed] [Google Scholar]
  7. Avram, M. , Brandl, F. , Bauml, J. , & Sorg, C. (2018). Cortico‐thalamic hypo‐ and hyperconnectivity extend consistently to basal ganglia in schizophrenia. Neuropsychopharmacology, 43(11), 2239–2248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Back, S. A. , Luo, N. L. , Borenstein, N. S. , Levine, J. M. , Volpe, J. J. , & Kinney, H. C. (2001). Late oligodendrocyte progenitors coincide with the developmental window of vulnerability for human perinatal white matter injury. The Journal of Neuroscience, 21(4), 1302–1312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ball, G. , Boardman, J. P. , Aljabar, P. , Pandit, A. , Arichi, T. , Merchant, N. , … Counsell, S. J. (2013). The influence of preterm birth on the developing thalamocortical connectome. Cortex, 49(6), 1711–1721. [DOI] [PubMed] [Google Scholar]
  10. Ball, G. , Pazderova, L. , Chew, A. , Tusor, N. , Merchant, N. , Arichi, T. , … Counsell, S. J. (2015). Thalamocortical connectivity predicts cognition in children born preterm. Cerebral Cortex, 25(11), 4310–4318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Bauml, J. G. , Daamen, M. , Meng, C. , Neitzel, J. , Scheef, L. , Jaekel, J. , … Sorg, C. (2015). Correspondence between aberrant intrinsic network connectivity and gray‐matter volume in the ventral brain of preterm born adults. Cerebral Cortex, 25(11), 4135–4145. [DOI] [PubMed] [Google Scholar]
  12. Beckmann, C. F. , DeLuca, M. , Devlin, J. T. , & Smith, S. M. (2005). Investigations into resting‐state connectivity using independent component analysis. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 360(1457), 1001–1013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Behrens, T. E. , Berg, H. J. , Jbabdi, S. , Rushworth, M. F. , & Woolrich, M. W. (2007). Probabilistic diffusion tractography with multiple fibre orientations: What can we gain? NeuroImage, 34(1), 144–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Behrens, T. E. , Johansen‐Berg, H. , Woolrich, M. W. , Smith, S. M. , Wheeler‐Kingshott, C. A. , Boulby, P. A. , … Matthews, P. M. (2003). Non‐invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nature Neuroscience, 6(7), 750–757. [DOI] [PubMed] [Google Scholar]
  15. Bohm, B. , Lundequist, A. , & Smedler, A. C. (2010). Visual‐motor and executive functions in children born preterm: The bender visual motor gestalt test revisited. Scandinavian Journal of Psychology, 51(5), 376–384. [DOI] [PubMed] [Google Scholar]
  16. Breeman, L. D. , Jaekel, J. , Baumann, N. , Bartmann, P. , & Wolke, D. (2015). Preterm cognitive function into adulthood. Pediatrics, 136(3), 415–423. [DOI] [PubMed] [Google Scholar]
  17. Butcher, P. R. , Bouma, A. , Stremmelaar, E. F. , Bos, A. F. , Smithson, M. , & Van Braeckel, K. N. (2012). Visuospatial perception in children born preterm with no major neurological disorders. Neuropsychology, 26(6), 723–734. [DOI] [PubMed] [Google Scholar]
  18. Bystron, I. , Blakemore, C. , & Rakic, P. (2008). Development of the human cerebral cortex: Boulder committee revisited. Nature Reviews. Neuroscience, 9(2), 110–122. [DOI] [PubMed] [Google Scholar]
  19. Casaer, P. , & Eggermont, E. (1985). Neonatal clinical neurological assessment In H. S. A. NJ (Ed.), The at‐risk infant: Psycho/socio/medical aspects (pp. 197–220). Baltimore: Brookes. [Google Scholar]
  20. Chaminade, T. , Leutcher, R. H. , Millet, V. , & Deruelle, C. (2013). fMRI evidence for dorsal stream processing abnormality in adults born preterm. Brain and Cognition, 81(1), 67–72. [DOI] [PubMed] [Google Scholar]
  21. Corbetta, M. , Patel, G. , & Shulman, G. L. (2008). The reorienting system of the human brain: From environment to theory of mind. Neuron, 58(3), 306–324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Corbetta, M. , & Shulman, G. L. (2002). Control of goal‐directed and stimulus‐driven attention in the brain. Nature Reviews Neuroscience, 3(3), 201–215. [DOI] [PubMed] [Google Scholar]
  23. Daitch, A. L. , Sharma, M. , Roland, J. L. , Astafiev, S. V. , Bundy, D. T. , Gaona, C. M. , … Corbetta, M. (2013). Frequency‐specific mechanism links human brain networks for spatial attention. Proceedings of the National Academy of Sciences of the United States of America, 110(48), 19585–19590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Deng, W. (2010). Neurobiology of injury to the developing brain. Nature Reviews Neurology, 6(6), 328–336. [DOI] [PubMed] [Google Scholar]
  25. Dubowitz, L. M. , Dubowitz, V. , & Goldberg, C. (1970). Clinical assessment of gestational age in the newborn infant. The Journal of Pediatrics, 77(1), 1–10. [DOI] [PubMed] [Google Scholar]
  26. Eryigit Madzwamuse, S. , Baumann, N. , Jaekel, J. , Bartmann, P. , & Wolke, D. (2015). Neuro‐cognitive performance of very preterm or very low birth weight adults at 26 years. Journal of Child Psychology and Psychiatry, 56(8), 857–864. [DOI] [PubMed] [Google Scholar]
  27. Fernandes, V. R. , Ribeiro, M. L. , Melo, T. , de Tarso Maciel‐Pinheiro, P. , Guimaraes, T. T. , Araujo, N. B. , … Deslandes, A. C. (2016). Motor coordination correlates with academic achievement and cognitive function in children. Frontiers in Psychology, 7, 318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Filippini, N. , MacIntosh, B. J. , Hough, M. G. , Goodwin, G. M. , Frisoni, G. B. , Smith, S. M. , … Mackay, C. E. (2009). Distinct patterns of brain activity in young carriers of the APOE‐ε4 allele. Proceedings of the National Academy of Sciences, 106(17), 7209–7214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Finke, K. , Neitzel, J. , Bauml, J. G. , Redel, P. , Muller, H. J. , Meng, C. , … Sorg, C. (2015). Visual attention in preterm born adults: Specifically impaired attentional sub‐mechanisms that link with altered intrinsic brain networks in a compensation‐like mode. NeuroImage, 107, 95–106. [DOI] [PubMed] [Google Scholar]
  30. Fischer, J. , & Whitney, D. (2012). Attention gates visual coding in the human pulvinar. Nature Communications, 3, 1051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Foulder‐Hughes, L. A. , & Cooke, R. W. (2003). Motor, cognitive, and behavioural disorders in children born very preterm. Developmental Medicine and Child Neurology, 45(2), 97–103. [PubMed] [Google Scholar]
  32. Garrett, D. D. , Epp, S. M. , Perry, A. , & Lindenberger, U. (2018). Local temporal variability reflects functional integration in the human brain. NeuroImage, 183, 776–787. [DOI] [PubMed] [Google Scholar]
  33. Goyen, T. A. , Lui, K. , & Hummell, J. (2011). Sensorimotor skills associated with motor dysfunction in children born extremely preterm. Early Human Development, 87(7), 489–493. [DOI] [PubMed] [Google Scholar]
  34. Grabner, G. , Janke, A. L. , Budge, M. M. , Smith, D. , Pruessner, J. , & Collins, D. L. (2006). Symmetric atlasing and model based segmentation: An application to the hippocampus in older adults. International Conference on Medical Image Computing and Computer‐Assisted Intervention, 9(Pt 2), 58–66. [DOI] [PubMed] [Google Scholar]
  35. Greve, D. N. , & Fischl, B. (2009). Accurate and robust brain image alignment using boundary‐based registration. NeuroImage, 48(1), 63–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Gutbrod, T. , Wolke, D. , Soehne, B. , Ohrt, B. , & Riegel, K. (2000). Effects of gestation and birth weight on the growth and development of very low birthweight small for gestational age infants: A matched group comparison. Archives of Disease in Childhood. Fetal and Neonatal Edition, 82(3), F208–F214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Haapala, E. A. , Poikkeus, A. M. , Tompuri, T. , Kukkonen‐Harjula, K. , Leppanen, P. H. , Lindi, V. , & Lakka, T. A. (2014). Associations of motor and cardiovascular performance with academic skills in children. Medicine and Science in Sports and Exercise, 46(5), 1016–1024. [DOI] [PubMed] [Google Scholar]
  38. Hayes, A. F. , & Rockwood, N. J. (2017). Regression‐based statistical mediation and moderation analysis in clinical research: Observations, recommendations, and implementation. Behaviour Research and Therapy, 98, 39–57. [DOI] [PubMed] [Google Scholar]
  39. Hoerder‐Suabedissen, A. , & Molnar, Z. (2015). Development, evolution and pathology of neocortical subplate neurons. Nature Reviews. Neuroscience, 16(3), 133–146. [DOI] [PubMed] [Google Scholar]
  40. Jbabdi, S. , & Johansen‐Berg, H. (2011). Tractography: Where do we go from here? Brain Connectivity, 1(3), 169–183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Jenkinson, M. , Bannister, P. , Brady, M. , & Smith, S. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage, 17(2), 825–841. [DOI] [PubMed] [Google Scholar]
  42. Jenkinson, M. , Beckmann, C. F. , Behrens, T. E. , Woolrich, M. W. , & Smith, S. M. (2012). FSL. NeuroImage, 62(2), 782–790. [DOI] [PubMed] [Google Scholar]
  43. Johnson, S. , Wolke, D. , Hennessy, E. , & Marlow, N. (2011). Educational outcomes in extremely preterm children: Neuropsychological correlates and predictors of attainment. Developmental Neuropsychology, 36(1), 74–95. [DOI] [PubMed] [Google Scholar]
  44. Kimchi, R. (2009). Perceptual organization and visual attention. Progress in Brain Research, 176, 15–33. [DOI] [PubMed] [Google Scholar]
  45. Kostovic, I. , & Judas, M. (2010). The development of the subplate and thalamocortical connections in the human foetal brain. Acta Paediatrica, 99(8), 1119–1127. [DOI] [PubMed] [Google Scholar]
  46. Kumar, V. J. , van Oort, E. , Scheffler, K. , Beckmann, C. F. , & Grodd, W. (2017). Functional anatomy of the human thalamus at rest. NeuroImage, 147, 678–691. [DOI] [PubMed] [Google Scholar]
  47. Lange, R. T. (2011). Perceptual organization index In Kreutzer J. S., DeLuca J., & Caplan B. (Eds.), Encyclopedia of clinical neuropsychology. New York, NY: Springer. [Google Scholar]
  48. Larroque, B. , Ancel, P. Y. , Marret, S. , Marchand, L. , Andre, M. , Arnaud, C. , … group, E. S . (2008). Neurodevelopmental disabilities and special care of 5‐year‐old children born before 33 weeks of gestation (the EPIPAGE study): A longitudinal cohort study. Lancet, 371(9615), 813–820. [DOI] [PubMed] [Google Scholar]
  49. Lau, E. F. , Phillips, C. , & Poeppel, D. (2008). A cortical network for semantics: (de)constructing the N400. Nature Reviews. Neuroscience, 9(12), 920–933. [DOI] [PubMed] [Google Scholar]
  50. Leung, M. P. , Thompson, B. , Black, J. , Dai, S. , & Alsweiler, J. M. (2018). The effects of preterm birth on visual development. Clinical & Experimental Optometry, 101(1), 4–12. [DOI] [PubMed] [Google Scholar]
  51. Lopez‐Bendito, G. , & Molnar, Z. (2003). Thalamocortical development: How are we going to get there? Nature Reviews. Neuroscience, 4(4), 276–289. [DOI] [PubMed] [Google Scholar]
  52. Marlow, N. , Hennessy, E. M. , Bracewell, M. A. , Wolke, D. , & Group, E. P. S . (2007). Motor and executive function at 6 years of age after extremely preterm birth. Pediatrics, 120(4), 793–804. [DOI] [PubMed] [Google Scholar]
  53. Martin, J. A. , Kung, H. C. , Mathews, T. J. , Hoyert, D. L. , Strobino, D. M. , Guyer, B. , & Sutton, S. R. (2008). Annual summary of vital statistics: 2006. Pediatrics, 121(4), 788–801. [DOI] [PubMed] [Google Scholar]
  54. Menegaux, A. , Meng, C. , Neitzel, J. , Bauml, J. G. , Muller, H. J. , Bartmann, P. , … Sorg, C. (2017). Impaired visual short‐term memory capacity is distinctively associated with structural connectivity of the posterior thalamic radiation and the splenium of the corpus callosum in preterm‐born adults. NeuroImage, 150, 68–76. [DOI] [PubMed] [Google Scholar]
  55. Meng, C. , Bauml, J. G. , Daamen, M. , Jaekel, J. , Neitzel, J. , Scheef, L. , … Sorg, C. (2016). Extensive and interrelated subcortical white and gray matter alterations in preterm‐born adults. Brain Structure & Function, 221(4), 2109–2121. [DOI] [PubMed] [Google Scholar]
  56. Molloy, C. S. , Di Battista, A. M. , Anderson, V. A. , Burnett, A. , Lee, K. J. , Roberts, G. , … Doyle, L. W. (2017). The contribution of visual processing to academic achievement in adolescents born extremely preterm or extremely low birth weight. Child Neuropsychology, 23(3), 361–379. [DOI] [PubMed] [Google Scholar]
  57. O'Reilly, J. X. , Beckmann, C. F. , Tomassini, V. , Ramnani, N. , & Johansen‐Berg, H. (2010). Distinct and overlapping functional zones in the cerebellum defined by resting state functional connectivity. Cerebral Cortex, 20(4), 953–965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Papageorgiou, K. A. , Smith, T. J. , Wu, R. , Johnson, M. H. , Kirkham, N. Z. , & Ronald, A. (2014). Individual differences in infant fixation duration relate to attention and behavioral control in childhood. Psychological Science, 25(7), 1371–1379. [DOI] [PubMed] [Google Scholar]
  59. Penn, A. A. , Gressens, P. , Fleiss, B. , Back, S. A. , & Gallo, V. (2016). Controversies in preterm brain injury. Neurobiology of Disease, 92(Pt A), 90–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Piaget, J. (1952). The origins of intelligence in children. New York: International Universities Press. [Google Scholar]
  61. Platt, M. J. , Cans, C. , Johnson, A. , Surman, G. , Topp, M. , Torrioli, M. G. , & Krageloh‐Mann, I. (2007). Trends in cerebral palsy among infants of very low birthweight (<1500 g) or born prematurely (<32 weeks) in 16 European centres: A database study. Lancet, 369(9555), 43–50. [DOI] [PubMed] [Google Scholar]
  62. Preacher, K. J. , & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4), 717–731. [DOI] [PubMed] [Google Scholar]
  63. Riegel, K. , Orth, B. , Wolke, D. , & Osterlund, K. (1995). Die Entwicklung gefährdet geborener Kinder bis zum 5 Lebensjahr. Stuttgart (Germany): Thieme. [Google Scholar]
  64. Rohr, C. S. , Vinette, S. A. , Parsons, K. A. L. , Cho, I. Y. K. , Dimond, D. , Benischek, A. , … Bray, S. (2017). Functional connectivity of the dorsal attention network predicts selective attention in 4‐7 year‐old girls. Cerebral Cortex, 27(9), 4350–4360. [DOI] [PubMed] [Google Scholar]
  65. Rus, O. G. , Reess, T. J. , Wagner, G. , Zimmer, C. , Zaudig, M. , & Koch, K. (2017). Functional and structural connectivity of the amygdala in obsessive‐compulsive disorder. NeuroImage: Clinical, 13, 246–255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Saalmann, Y. B. (2014). Intralaminar and medial thalamic influence on cortical synchrony, information transmission and cognition. Frontiers in Systems Neuroscience, 8, 83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Saalmann, Y. B. , Pinsk, M. A. , Wang, L. , Li, X. , & Kastner, S. (2012). The pulvinar regulates information transmission between cortical areas based on attention demands. Science, 337(6095), 753–756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Salmaso, N. , Jablonska, B. , Scafidi, J. , Vaccarino, F. M. , & Gallo, V. (2014). Neurobiology of premature brain injury. Nature Neuroscience, 17(3), 341–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Shang, J. , Bauml, J. G. , Koutsouleris, N. , Daamen, M. , Baumann, N. , Zimmer, C. , … Sorg, C. (2018). Decreased BOLD fluctuations in lateral temporal cortices of premature born adults. Human Brain Mapping, 39(12), 4903–4912. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Sherman, S. M. (2016). Thalamus plays a central role in ongoing cortical functioning. Nature Neuroscience, 19(4), 533–541. [DOI] [PubMed] [Google Scholar]
  71. Shipp, S. (2003). The functional logic of cortico‐pulvinar connections. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 358(1438), 1605–1624. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Sripada, K. , Lohaugen, G. C. , Eikenes, L. , Bjorlykke, K. M. , Haberg, A. K. , Skranes, J. , & Rimol, L. M. (2015). Visual‐motor deficits relate to altered gray and white matter in young adults born preterm with very low birth weight. NeuroImage, 109, 493–504. [DOI] [PubMed] [Google Scholar]
  74. Van Dijk, K. R. , Sabuncu, M. R. , & Buckner, R. L. (2012). The influence of head motion on intrinsic functional connectivity MRI. NeuroImage, 59(1), 431–438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Volpe, J. J. (2009a). Brain injury in premature infants: A complex amalgam of destructive and developmental disturbances. Lancet Neurology, 8(1), 110–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Volpe, J. J. (2009b). The encephalopathy of prematurity‐‐brain injury and impaired brain development inextricably intertwined. Seminars in Pediatric Neurology, 16(4), 167–178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Von Aster, M. , & Horn, N. A. R. (2006). Wechsler Intelligenztest für Erwachsene (WIE). Deutschsprachige Bearbeitung und Adaptation des WAIS‐III von David Wechsler. Frankfurt/Main (Germany): Harcourt Test Services. [Google Scholar]
  78. Vossel, S. , Geng, J. J. , & Fink, G. R. (2014). Dorsal and ventral attention systems: Distinct neural circuits but collaborative roles. The Neuroscientist, 20(2), 150–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Wagner, G. , Frick, J. , & Schupp, J. (2007). The German Socio‐Economic Panel Study (SOEP) – Scope, Evolution and Enhancements. Schmollers Jahrbuch : Journal of Applied Social Science Studies / Zeitschrift für Wirtschafts‐ und Sozialwissenschaften, 127(1), 139–169. [Google Scholar]
  80. Wolke, D. , & Meyer, R. (1999). Cognitive status, language attainment, and prereading skills of 6‐year‐old very preterm children and their peers: The Bavarian longitudinal study. Developmental Medicine and Child Neurology, 41(2), 94–109. [DOI] [PubMed] [Google Scholar]
  81. Woodward, L. J. , Edgin, J. O. , Thompson, D. , & Inder, T. E. (2005). Object working memory deficits predicted by early brain injury and development in the preterm infant. Brain, 128(Pt 11), 2578–2587. [DOI] [PubMed] [Google Scholar]
  82. Zhang, D. , Snyder, A. Z. , Shimony, J. S. , Fox, M. D. , & Raichle, M. E. (2010). Noninvasive functional and structural connectivity mapping of the human thalamocortical system. Cerebral Cortex, 20(5), 1187–1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Zhang, Y. , Brady, M. , & Smith, S. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation‐maximization algorithm. IEEE Transactions on Medical Imaging, 20(1), 45–57. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplemental Figure 1 One sample t‐tests of the DAN for MB and VPB adults (p < .05, FWE cluster‐corrected); Yellow: MB adults only; Red: VPB adults only; Orange: overlap of MB and VPB adults

Supplemental Figure 2: Voxel‐wise two sample t‐tests of DAN‐thalamus functional connectivity between MB and VPB adults; p = .071, FWE‐cluster corrected

Click here for additional data file. (342.2KB, docx)

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES