Visual Occipito-Temporal N1 Sensitivity to Digits Across Elementary School

Abstract

Number processing abilities are important for academic and personal development. The course of initial specialization of ventral occipito-temporal cortex (vOTC) sensitivity to visual number processing is crucial for the acquisition of numeric and arithmetic skills. We examined the visual N1, the electrophysiological correlate of vOTC activation across five time points in kindergarten (T1, mean age 6.60 years), middle and end of first grade (T2, 7.38 years; T3, 7.68 years), second grade (T4, 8.28 years), and fifth grade (T5, 11.40 years). A combination of cross-sectional and longitudinal EEG data of a total of 62 children (35 female) at varying familial risk for dyslexia were available to form groups of 23, 22, 27, 27, and 42 participants for each of the five time points. The children performed a target detection task which included visual presentation of single digits (DIG), false fonts (FF), and letters (LET) to derive measures for coarse (DIG vs. FF) and fine (DIG vs. LET) digit sensitive processing across development. The N1 amplitude analyses indicated coarse and fine sensitivity characterized by a stronger N1 to digits than false fonts across all five time points, and stronger N1 to digits than letters at all but the second (T2) time point. In addition, lower arithmetic skills were associated with stronger coarse N1 digit sensitivity over the left hemisphere in second grade (T4), possibly reflecting allocation of more attentional resources or stronger reliance on the verbal system in children with poorer arithmetic skills. To summarize, our results show persistent visual N1 sensitivity to digits that is already present early on in pre-school and remains stable until fifth grade. This pattern of digit sensitivity development clearly differs from the relatively sharp rise and fall of the visual N1 sensitivity to words or letters between kindergarten and middle of elementary school and suggests unique developmental trajectories for visual processing of written characters that are relevant to numeracy and literacy.

Keywords: ERP; arithmetic skills; development; number processing; numeracy; occipito-temporal cortex; visual N1.

FIGURE 1

Numerical skills across measurements. Y-axis in left panel shows the percentage of correctly named numbers and in the right panel, the total percentile score for arithmetic operations. Asterisks indicate significant differences at p < 0.05 between test times. Boxplots show median and interquartile range. Error bars indicate mean and 95% CIs.

FIGURE 2

Target detection task. (A) Digits, false fonts, letters, and target stimuli. (B) Trial structure and presentation times.

FIGURE 3

(A) ERPs (y-axis: μV, x-axis: time in ms) for digits (blue lines), letters (red lines), and false fonts (black lines) with ribbons indicating 95% CIs (within-subject). Left hemispheric amplitudes are depicted on the left side and right hemispheric amplitudes on the right. The N1 time-window is highlighted in yellow. (B) Scalp topographies for digits, letters, and false fonts. (C) GFP for the average of the three conditions (black line), digits (dashed blue line), letters (dashed red line), false fonts (dashed black line). (D) Scalp maps show the distribution of t values for the difference between conditions tested against zero (color steps display only values from 6 to 4 and 4 to 2, corresponding to ±t values significant for current sample sizes at p < 0.05). The horizontal bars below show p values resulting from TANOVA permutation tests of differences in topographies between conditions across each data point of the N1 time intervals (T1: 198–258 ms, T2: 188–248 ms, T3: 190–250 ms, T4: 180–240 ms, and T5: 178–238 ms). The bar color map from light to darker orange indicates ps ranging from 0.05 to 0; the blank periods indicate non-significant ps > 0.05. ERPs, event-related potentials; CIs, confidence intervals; GFP, global field power; TANOVA, topographic analysis of variance; DIG, digits; LET, letters; FF, false fonts; LOT, left occipito-temporal; ROT, right occipito-temporal.

FIGURE 4

Mean amplitudes (μV) for the N1 interval over the five time points (T1, T2, T3, T4, and T5) for digits (blue), letters (red), and false fonts (black). Left (LOT) and right (ROT) electrode clusters are shown in separate panels. Error bars represent 95% CI. Normalized residuals from the main model exceeding ±3 threshold were excluded (7 data points).

FIGURE 5

Differences in N1 mean amplitudes (μV) between digits and false fonts (coarse digit sensitivity: top plots) and digits and letters (fine digits sensitivity: bottom plots) per time point. Error bars within the density plot indicate mean and 95% CIs. More negative values indicate stronger negative amplitudes for digits than false fonts (top) or letters (bottom), respectively. Left (LOT) hemispheric cluster is depicted on the left panel, right (ROT) hemispheric cluster on the right.

FIGURE 6

Spearman correlation and regression line of trends between total percentile score of arithmetic operations (HRT total) and N1 digits-false font amplitude differences in the left and right ventral-occipital electrodes. Box plots are displayed in the margins. LOT, left occipito-temporal; ROT, right occipito-temporal; DIG, digits; FF, false fonts; HRT, Heidelberger Rechentest; PR, percentile.