Multivariate characterization of brain white matter maturation related to intellectual ability in children
Yannan Cheng1, Chao Jin2, Xianjun Li2, Congcong Liu2, Miaomiao Wang2, Xiaocheng Wei3, Yuli Zhang2, Fan Wu2, Mengxuan Li2, and Jian Yang2

1Department of Radiology, the First Affiliated Hospital, Xi’an Jiaotong University, Xi'an, China, Xi’an, China, 2the First Affiliated Hospital, Xi’an Jiaotong University, Xi'an, China, Xi'an, China, 3MR Research China, GE Healthcare, Bei Jing, People's Republic of China, Bei Jing, China


Brain maturations are thought to relate to behavioral acquisitions and cognitive development. Nevertheless, in vivo investigations of such relationships remain scarce in childhood. To bridge this gap, a multivariate index (DM), which delineates the “maturational distance” between children and adults and leverages DTI-metrics complementarity, was utilized to characterize WM variation. We found that DM showed significantly negative correlations with FSIQ in children aged 4-12 yr, especially in cingulum and superior longitudinal fasciculus. Besides, left hemispheric lateralization (higher correlations with FSIQ) was also observed. Our findings suggest DM as a useful biomarker in detailing the brain WM maturation related to intelligence.


From birth, brain structure undergoes remarkable maturation changes. These maturation processes are thought to relate to behavioral acquisitions and cognitive development. Nevertheless, in vivo investigations of such relationships remain scarce in childhood. Mutiparametric neuroimaging (e.g. DTI) allows in vivo quantification of white matter (WM) maturation. However, various complex metrics (FA, MD, AD and RD) challenges the clinicians and limited its further clinical application. To bridge this gap, a multivariate index, which delineates the “maturational distance” between children and adults and leverages DTI-metrics complementarity, was utilized to characterize individual WM variation1. By using this index, the correlations between brain WM maturation and full scale intelligence quotient (FSIQ) in children aged 4-12 yr were explored.

Materials and Methods

The Institutional Review Broad of the first author’s affiliation approved this study and written informed consent were obtained from parents of all the children. Subjects Twenty-three children with no abnormality on MRI participated in the study. Intellectual ability of all children were in normal range (full scale IQ >85 scores). Intellectual ability was assessed by the Chinese Wechsler Young Children scale of Intelligence (C-WYCSI) for children aged 4 to 6 years and Chinese Wechsler Intelligence Scale for Children (C-WISC) children aged 6 to 18 years. MR Protocols All subjects were examined by using a 3.0T scanner (Signa HDxt, General Electric Medical System, Milwaukee, WI, USA) with an 8-channel head coil. Data acquisition included three-dimensional fast spoiled gradient-echo T1-weighted sequence (TR/TE, 10.2ms/4.6ms; NEX, 1; isotropic 1×1×1mm3; FOV, 24cm) and transverse fast spin-echo T2-weighted sequence (TR/TE, 4200ms/113ms; NEX of 1.5; matrix, 320×320; thickness, 4mm; FOV, 24cm), and DTI (18 directions; b value, 1000s/mm2; TR/TE, 11000ms/67.4ms; NEX of 1; thickness, 2.5mm; FOV, 24cm; matrix, 172×172). Data and statistical analysis DTI data were preprocessed by FMRIB software library (FSL; http://www.fmrib.ox.ac.uk/fsl) and Matlab software (MathWorks, Natick, Massachusetts). Four parameters of fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were calculated by using the above tools based on ROIs. Here we use the Johns Hopkins white matter tracts atlas as a reference. DM was calculated by Mahalanobis distance according to Ref. 1. Considering the possible impacts of age, partial Pearson’s correlation analysis was used to explore the relationships between FSIQ and DM, with age controlled.

All statistical analysis was performed by using SPSS 18.0 (SPSS, Chicago, IL, USA); p<0.05 was considered as statistically significant difference.


Twenty-three children were included in this study (Table 1).

Results indicated that significant negative correlations between DM and FSIQ scores were found in bilateral superior longitudinal fasciculus (SLF: rleft=0.506, P=0.016; rright=0.594, P=0.004), inferior fronto-occipital fasciculus (IFO: rleft=0.472, P=0.027; rright=0.425, P=0.048) and forceps_major (r=-0.450, P=0.036). In addition, strong correlation also existed in left cingulum (r=-0.731, P<0.001) and temporal part of SLF (r=-0.675, P=0.001) (Figure 1).


In this study, a multivariate index DM that leverages DTI-metrics complementarity was used to explore the associations between WM maturation and intelligence. In contrast to univariate metric, this index was more reliable as it takes into account complementary dependencies of DTI-metrics on different maturational processes, notably the decrease in water content and the myelination1.

Previous study reported the close link between intelligence and anterior cingulate in children2. Consistently, we also observed significant correlation of FSIQ with DM in cingulum. Cingulum is thought to correlate with cognitive control in later life. Besides, more mature long-range tracts such as IFO, SLF and forceps_major have been demonstrated to show higher ability of working memory. These WM tracts were suggested as critical region in supporting the intelligence development3,4. Similarly, we also found that a shorter DM in these WM tracts corresponds to a higher level of FSIQ, further suggesting the importance of WM maturation in these regions related to intelligence.

In addition, left cingulum and temporal part of SLF presented remarkably higher correlations with FSIQ than right ones. This may suggest a left cerebral hemispheric lateralization in relation to intellectual ability. Although evidence of hemispheric specialization in left rostral cingulum, e.g. in the context of verbal or symbolic rules have been reported4, more details concerning these WM maturation with intelligence should also been explored.


Our findings further demonstrated the specific correlations between WM maturation and intellectual ability development in children and suggested DM as a useful marker for WM maturation’s characterization.


This work was supported by the National Key Research and Development Program of China (2016YFC0100300), National Natural Science Foundation of China (No. 81471631, 81771810 and 51706178), the 2011 New Century Excellent Talent Support Plan of the Ministry of Education, China (NCET-11-0438) the Clinical Research Award of the First Affiliated Hospital of Xi’an Jiaotong University (No.XJTU1AF-CRF-2015-004)


  1. Kulikova S, Hertz-Pannier L, Dehaene-Lambertz G, et al. Multi-parametric evaluation of the white matter maturation. Brain Structure and Function, 2015, 220(6): 3657-3672.
  2. Casey B J, Trainor R, Giedd J, et al. The role of the anterior cingulate in automatic and controlled processes: a developmental neuroanatomical study. Dev Psychobiol, 1997, 30(1): 61-69.
  3. Chiang M C, Barysheva M, Shattuck D W, et al. Genetics of brain fiber architecture and intellectual performance. J Neurosci, 2009, 29(7): 2212-2224.
  4. Tamnes C K, Østby Y, Walhovd K B, et al. Intellectual abilities and white matter microstructure in development: a diffusion tensor imaging study. Human Brain Mapping, 2010, 31(10): 1609-1625.


Table 1 Participant demographics

Figure 1 Scatterplots for correlation analysis between FSIQ score and DM in white matter tracts. SLF = superior longitudinal fasciculus; SLF_tem = temporal part of superior longitudinal fasciculus; IFO = inferior fronto-occipital fasciculus.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)