Figures and data in Reevaluating the neural noise in dyslexia using biomarkers from electroencephalography and high-resolution magnetic resonance spectroscopy

Figures
Tables
Additional files

4 figures, 11 tables and 1 additional file

Figures

Figure 1

Download asset Open asset

Selected results for the electroencephalography (EEG) excitatory and inhibitory (E/I) balance biomarkers.

n = 119 (DYS n = 59, CON n = 60). (A) Power spectral densities averaged across three electrodes (T7, TP7, TP9)corresponding to the left superior temporal sulcus (STS) separately for dyslexic (DYS) and control (CON) groups at rest and (C) during the language task. (B) Plots illustrating results for the exponent, offset, and the beta power from the left STS electrodes at rest and (D) during the language task.

Figure 2 with 1 supplement

Download asset Open asset

Selected results for the magnetic resonance spectroscopy (MRS) excitatory and inhibitory (E/I) balance biomarkers.

(A) Sample MRS voxel localization. (B) Sample MRS spectrum with individual total creatine (tCr), Glu, and GABA+ contributions plotted below. (C) Group results (CON>DYS) from the fMRI localizer task for words compared to the control stimuli (p<0.05 FWE cluster threshold) and overlap of the MRS voxel placement across participants. n = 50 (DYS n = 21, CON n = 29). (D) MRS spectra separately for the dyslexic (DYS) and control (CON) groups. n = 50 (DYS n = 21, CON n = 29). (E) Plots illustrating results for the Glu, n = 50 (DYS n = 21, CON n = 29), GABA+, n = 47 (DYS n = 20, CON n = 27), Glu/GABA+ ratio, n = 47 (DYS n = 20, CON n = 27), and the Glu/GABA+ imbalance, n = 47 (DYS n = 20, CON n = 27).

Figure 2—figure supplement 1

Download asset Open asset

Main effect of group CON>DYS for both visual (words > false fonts) and auditory runs (words > words backward) from the fMRI localizer task.

Note: CON – control group, DYS – dyslexic group, L – left hemisphere, R – right hemisphere. Results reported at p<0.001 height threshold corrected for multiple comparisons using p<0.05 FWE cluster threshold. n = 50 (DYS n = 21, CON n = 29).

Figure 3

Download asset Open asset

Relationships between electroencephalography (EEG) and magnetic resonance spectroscopy (MRS) excitatory and inhibitory (E/I) balance biomarkers.

(A) Semi-partial correlation between offset at rest in the left superior temporal sulcus (STS) electrodes and Glu controlling for age and gray matter volume (GMV), n = 50 (DYS n = 21, CON n = 29). (B) Semi-partial correlation between exponent at rest in the left STS electrodes (controlled for age) and Glu/GABA+ imbalance (controlled for GMV), n = 47 (DYS n = 20, CON n = 27).

Figure 4

Download asset Open asset

Associations between beta power, phonological awareness, and reading.

n = 119 (DYS n = 59, CON n = 60). (A) Semi-partial correlation between phonological awareness controlling for age and beta power (in the left superior temporal sulcus [STS] electrodes) at rest and (B) during the language task. (C) Partial correlation between phonological awareness and reading controlling for age. (D) Mediation analysis results. Unstandardized b regression coefficients are presented. Age was included in the analysis as a covariate. 95% CI – 95% confidence intervals; left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9).

Tables

Table 1

Demographic and behavioral characteristics of the entire sample of 120 participants.

	DYS28 F, 32 M		CON28 F, 32 M		t(df)	p	Cohen’s d	BF₁₀
	M	SD	M	SD	t(df)	p	Cohen’s d	BF₁₀
Demographics
Age	19.41	3.18	19.54	2.96	0.25 (118)	0.806	0.05	0.20
Mother’s education (years)	17.20	3.36	16.58	2.28	–1.19 (103.89)	0.235	–0.22	0.37
Father’s education (years)	16.12^a	3.10^a	17.13^a	3.27^a	1.71 (114)	0.091	0.32	0.73
IQ	103.56^b	11.83^b	111.12	10.43	3.70 (117)	<0.001	0.68	75.31
Nonverbal IQ (scaled score)	10.40	2.94	11.62	2.57	2.42 (118)	0.017	0.44	2.63
ARHQ-PL	51.50	9.70	25.47	8.00	–16.04 (113.87)	<0.001	–2.93	>10,000
Reading and reading-related tasks
Words/min	108.38	20.93	134.57	13.29	8.18 (99.90)	<0.001*	1.49	>10,000
Pseudowords/min	56.75	14.16	83.43	17.04	9.33 (118)	<0.001*	1.70	>10,000
RAN objects (s)	32.12	5.11	28.70	4.43	–3.92 (118)	<0.001*	–0.72	149.93
RAN colors (s)	35.83	6.82	31.18	5.73	–4.04 (118)	<0.001*	–0.74	229.96
RAN digits(s)	19.32	4.61	16.25	2.94	–4.34 (100.28)	<0.001*	–0.79	642.86
RAN letters (s)	22.70	4.53	19.68	3.16	–4.23 (105.42)	<0.001*	–0.77	433.23
Reading comprehension (s)	64.47	20.13	43.72	9.63	–7.20 (84.66)	<0.001*	–1.32	>10,000
Phoneme deletion (% correct)	76.41	24.68	91.47	9.07	4.44 (74.66)	<0.001*	0.81	898.25
Spoonerisms phonemes (% correct)	54.29	35.42	82.74	22.06	5.28 (98.78)	<0.001*	0.96	>10,000
Spoonerisms syllables (% correct)	46.94	30.61	73.06	23.98	5.20 (111.62)	<0.001*	0.95	>10,000
Orthographic awareness (accuracy/time)	0.33	0.13	0.53	0.14	8.12 (118)	<0.001*	1.48	>10,000
Perception speed (sten score)	3.32	2.04	4.50	1.67	3.48 (118)	<0.001*	0.64	38.71
Digits forward	5.53	1.64	6.98	1.95	4.40 (118)	<0.001*	0.80	792.55
Digits backward	5.25	1.49	7.33	2.25	5.99 (102.59)	<0.001*	1.09	>10,000

Note: DYS – dyslexic group; CON – control group; F – females, M – males. BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0). ARHQ-PL – Polish version of the Adult Reading History Questionnaire. RAN – rapid automatized naming. Boldface indicates statistical significance at p<0.05 level (uncorrected).
*Significance after Bonferroni correction for 14 planned comparisons for reading and reading-related tasks
^an = 58 (two participants did not provide information about the father’s education)
^bn = 59 (one participant did not attempt a verbal subtest of the scale, thus we were not able to calculate overall IQ)

Table 2

Semi-partial correlations between magnetic resonance spectroscopy (MRS) and electroencephalography (EEG) biomarkers of excitatory-inhibitory balance.

For Glu the effects of age and gray matter volume (GMV) were regressed, for GABA+, Glu/GABA+ ratio, and Glu/GABA+ imbalance the effect of GMV was regressed, while for exponents and offsets the effect of age was regressed.

Variable	1. r(BF₁₀)	2.	3.	4.	5.	6.	7.	8.
EEG resting state
1. Glu	–
2. GABA+	0.32*^a (1.82)	–
3. Glu/GABA+ ratio	–0.08^a (0.21)	–0.91***^a (>10,000)	–
4. Glu/GABA+ imbalance	0.12^a (0.25)	0.31*^a (1.63)	–0.18^a (0.37)	–
5. Exponent mean (rest)	–0.03^b (0.18)	0.04^a (0.19)	–0.11^a (0.23)	0.29*^a (1.21)	–
6. Offset mean (rest)	–0.21^b (0.49)	0.08^a (0.21)	–0.17^a (0.35)	0.25^a (0.69)	0.68***^c (>10,000)	–
7. Exponent left STS (rest)	–0.16^b (0.32)	0.01^a (0.18)	–0.07^a (0.20)	0.35*^a (2.87)	0.68***^c (>10,000)	0.45***^c (>10,000)	–
8. Offset left STS (rest)	–0.38**^b (6.28)	–0.10^a (0.23)	0.02^a (0.18)	0.18^a (0.38)	0.18*^c (0.80)	0.47***^c (>10,000)	0.66***^c (>10,000)	–
9. Beta power left STS (rest)	–0.12^b (0.25)	0.17^a (0.33)	–0.25^a (0.74)	0.01^a (0.18)	0.18*^c (0.80)	0.21*^c (1.40)	0.51***^c (>10,000)	0.59***^c (>10,000)
EEG language task
5. Exponent mean (task)	–0.10^b (0.22)	0.06^a (0.20)	–0.15^a (0.29)	0.21^a (0.49)	–
6. Offset mean (task)	–0.26^b (0.92)	0.09^a (0.22)	–0.20^a (0.43)	0.22^a (0.53)	0.72***^c (>10,000)	–
7. Exponent left STS (task)	–0.20^b (0.44)	0.01^a (0.18)	–0.09^a (0.22)	0.22^a (0.51)	0.65***^c (>10000)	0.51***^c (>10,000)	–
8. Offset left STS (task)	–0.37**^b (5.05)	–0.08^a (0.21)	0.00^a (0.18)	0.15^a (0.30)	0.29**^c (18.31)	0.55***^c (>10,000)	0.77***^c (>10,000)	–
9. Beta power left STS (task)	–0.19^b (0.42)	0.16^a (0.32)	–0.22^a (0.55)	–0.03^a (0.19)	0.05^c (0.13)	0.13^c (0.31)	0.47***^c (>10,000)	0.60***^c (>10,000)

Note: r – Pearson’s correlation coefficient; BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0); mean – values averaged across all electrodes; left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9).
***p < 0.001 (uncorrected); **p < 0.01 (uncorrected); *p < 0.05 (uncorrected)
^an = 47; ^bn = 50; ^cn = 119

Table 3

Semi-partial correlations between reading, phonological awareness, rapid automatized naming (RAN), multisensory integration, and biomarkers of excitatory-inhibitory balance.

For reading, phonological awareness, and multisensory integration the effect of age was regressed, for Glu the effects of age and gray matter volume (GMV) were regressed, for GABA+, Glu/GABA+ ratio, and Glu/GABA+ imbalance the effect of GMV was regressed, while for exponents and offsets the effect of age was regressed.

Variable	1. r(BF₁₀)	2.	3.	4.
EEG resting state
1. Reading	–
2. Phonological awareness	0.60***^c (>10,000)	–
3. RAN	0.71***^c (>10,000)	0.48***^c (>10,000)	–
4. Multisensory integration	0.16^d (0.41)	0.25*^d (2.09)	0.02^d (0.14)	–
5. Glu	–0.03^b (0.18)	–0.12^b (0.25)	–0.05^b (0.19)	0.03^b (0.18)
6. GABA+	–0.18^a (0.36)	–0.06^a (0.20)	–0.23^a (0.56)	0.31*^a (1.62)
7. Glu/GABA+ ratio	0.07^a (0.20)	0.04^a (0.19)	0.09^a (0.22)	–0.32*^a (1.84)
8. Glu/GABA+ imbalance	–0.21^a (0.47)	–0.08^a (0.21)	–0.20^a (0.44)	0.17^a (0.33)
9. Exponent mean (rest)	–0.08^c (0.17)	0.10^c (0.20)	–0.06^c (0.14)	0.02^d (0.14)
10. Offset mean (rest)	0.06^c (0.14)	0.14^c (0.35)	0.03^c (0.12)	0.16^d (0.38)
11. Exponent left STS (rest)	–0.08^c (0.16)	0.06^c (0.14)	–0.04^c (0.12)	–0.04^d (0.14)
12. Offset left STS (rest)	0.12^c (0.25)	0.18*^c (0.77)	0.08^c (0.17)	0.08^d (0.18)
13. Beta power left STS (rest)	0.04^c (0.13)	0.23*^c (2.73)	–0.04^c (0.12)	0.05^d (0.15)
EEG language task
9. Exponent mean (task)	–0.07^c (0.16)	0.13^c (0.30)	–0.10^c (0.21)	0.01^d (0.14)
10. Offset mean (task)	0.05^c (0.13)	0.14^c (0.34)	0.01^c (0.12)	0.18^d (0.50)
11. Exponent left STS (task)	–0.03^c (0.12)	0.09^c (0.18)	–0.07^c (0.15)	–0.04^d (0.14)
12. Offset left STS (task)	0.13^c (0.28)	0.18^c (0.71)	0.07^c (0.15)	0.09^d (0.19)
13. Beta power left STS (task)	0.07^c (0.15)	0.23*^c (2.84)	0.02^c (0.12)	0.15^d (0.33)

Note: r – Pearson’s correlation coefficient; BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0); mean – values averaged across all electrodes; left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9).
***p < 0.001 (uncorrected); **p < 0.01 (uncorrected); *p < 0.05 (uncorrected)
^an = 47 ^bn = 50; ^cn = 119; ^dn = 87

Appendix 1—table 1

Descriptive statistics for electroencephalography (EEG) and magnetic resonance spectroscopy (MRS) results separately for the groups.

	DYS		CON
	M	SD	M	SD
EEG resting state^a
Exponent mean (rest)	1.54	0.14	1.54	0.18
Exponent left IFG (rest)	1.54	0.16	1.53	0.18
Exponent left STS (rest)	1.50	0.18	1.47	0.22
Exponent right IFG (rest)	1.54	0.15	1.54	0.18
Exponent right STS (rest)	1.48	0.18	1.45	0.22
Offset mean (rest)	–10.80	0.19	–10.80	0.24
Offset left IFG (rest)	–10.72	0.34	–10.74	0.33
Offset left STS (rest)	–10.97	0.38	–10.98	0.37
Offset right IFG (rest)	–10.79	0.29	–10.81	0.32
Offset right STS (rest)	–10.99	0.31	–11.04	0.36
Beta power left IFG (rest)	0.48	0.18	0.48	0.20
Beta power left STS (rest)	0.49	0.19	0.48	0.21
Beta power right IFG (rest)	0.49	0.18	0.50	0.19
Beta power right STS (rest)	0.51	0.20	0.50	0.21
EEG language task^a
Exponent mean (task)	1.55	0.15	1.56	0.18
Exponent left IFG (task)	1.55	0.16	1.55	0.19
Exponent left STS (task)	1.50	0.20	1.47	0.21
Exponent right IFG (task)	1.54	0.17	1.55	0.19
Exponent right STS (task)	1.47	0.19	1.45	0.22
Offset mean (task)	–10.67	0.25	–10.67	0.28
Offset left IFG (task)	–10.58	0.39	–10.60	0.37
Offset left STS (task)	–10.86	0.44	–10.87	0.42
Offset right IFG (task)	–10.65	0.36	–10.68	0.37
Offset right STS (task)	–10.88	0.36	–10.94	0.41
Beta power left IFG (task)^b	0.50	0.23	0.51	0.21
Beta power left STS (task)	0.54	0.24	0.53	0.23
Beta power right IFG (task)	0.51	0.23	0.52	0.21
Beta power right STS (task)	0.55	0.26	0.55	0.23
MRS
Glu^c	1.11	0.12	1.07	0.10
GABA+^d	0.46	0.14	0.44	0.12
Glu/GABA +ratio^d	2.67	0.87	2.68	0.75
Glu/GABA +imbalance^d	0.27	0.11	0.24	0.12

Note: DYS – dyslexic group; CON – control group; mean – values averaged across all electrodes; left IFG – values averaged across three electrodes corresponding to the left inferior frontal gyrus (F7, FT7, FC5); left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9); right IFG – values averaged across three electrodes corresponding to the right inferior frontal gyrus (F8, FT8, FC6); right STS – values averaged across three electrodes corresponding to the right superior temporal sulcus (T8, TP8, TP10).
^an = 119 (DYS n = 59, CON n = 60); ^bn = 117 (DYS n = 57, CON n = 60); ^cn = 50 (DYS n = 21, CON n = 29); ^dn = 47 (DYS n = 20, CON n = 27).

Appendix 1—table 2

Demographic and behavioral characteristics of the subsample of 47 participants.

	DYS(n=20)		CON(n=27)		Test	p	Effect size	BF₁₀
	M	SD	M	SD	Test	p	Effect size	BF₁₀
Demographics
Sex	12 F, 8 M		11 F, 16 M		χ=1.71	0.192	phi = –0.19	0.81
Age	19.98	3.92	20.33	3.25	U=261.0	0.846	r_rb = 0.03	0.29
Mother’s education (years)	17.48	3.68	16.63	2.56	U=242.5	0.551	r_rb = –0.10	0.36
Father’s education (years)	16.74^a	3.46^a	16.41	3.49	U=237.0	0.659	r_rb = –0.08	0.31
IQ	103.80	14.30	113.56	9.23	t(30.41)=2.67	0.012	d=0.81	6.65
Nonverbal IQ (scaled score)	10.45	3.14	12.37	2.17	U=174.5	0.038	r_rb = 0.35	1.79
ARHQ-PL	52.90	10.78	24.26	6.71	U=6.0	<0.001	r_rb = –0.98	1060.02
Reading and reading-related tasks
Words/min	110.80	19.59	136.11	12.80	U=74.0	<0.001*	r_rb = 0.73	108.53
Pseudowords/min	59.15	13.91	85.59	15.83	t(45)=5.96	<0.001*	d=1.76	>10,000
RAN objects (s)	32.55	4.36	28.56	4.46	U=122.0	0.001*	r_rb = –0.55	7.92
RAN colors (s)	36.40	4.91	30.63	3.67	t(45)=–4.61	<0.001*	d=–1.36	571.37
RAN digits(s)	19.35	4.00	16.33	2.24	U=134.0	0.003*	r_rb = –0.50	3.82
RAN letters (s)	21.75	3.24	19.52	2.53	t(45)=–2.65	0.011	d=–0.78	4.52
Reading comprehension (s)	63.10	18.30	43.89	7.87	U=82.0	<0.001*	r_rb = –0.70	65.88
Phoneme deletion (% correct)	74.62	29.20	94.16	6.61	U=120.0	0.001*	r_rb = 0.56	13.46
Spoonerisms phonemes (% correct)	52.14	38.23	87.83	8.36	U=106.5	<0.001*	r_rb = 0.61	17.43
Spoonerisms syllables (% correct)	41.67	29.86	77.78	20.67	U=87.5	<0.001*	r_rb = 0.68	27.77
Orthographic awareness (accuracy/time)	0.35	0.16	0.54	0.13	t(45)=4.26	<0.001*	d=1.26	215.19
Perception speed (sten score)	2.95	1.79	4.33	1.64	U=126.0	0.002*	r_rb = 0.53	4.49
Digits forward	5.65	1.79	7.04	2.03	U=158.0	0.014	r_rb = 0.42	2.13
Digits backward	4.95	1.73	7.48	2.06	t(45)=4.45	<0.001*	d=1.31	355.64

Note: DYS – dyslexic group; CON – control group; F – females, M – males. BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0). ARHQ-PL – Polish version of the Adult Reading History Questionnaire. RAN – rapid automatized naming. Non-parametric Mann-Whitney test was performed when assumption of normal distribution was violated. r_rb – rank biserial correlation provided as an effect size parameter for Mann-Whitney test. Boldface indicates statistical significance at p<0.05 level (uncorrected).
^*Significance after Bonferroni correction for 14 planned comparisons for reading and reading-related tasks;
^an = 19 (one participant did not provide information about the father’s education).

Appendix 1—table 3

Zero-order correlations between magnetic resonance spectroscopy (MRS) and electroencephalography (EEG) biomarkers of excitatory-inhibitory balance.

Variable	1. r(BF₁₀)	2.	3.	4.	5.	6.	7.	8.
EEG resting state
1. Glu	–
2. GABA+	0.40**^a (8.08)	–
3. Glu/GABA +ratio	–0.11 (0.24)^a	–0.90***^a (>10,000)	–
4. Glu/GABA +imbalance	0.16 (0.32)^a	0.33*^a (2.14)	–0.19 (0.42)^a	–
5. Exponent mean (rest)	0.13 (0.25)^b	0.10 (0.23)^a	–0.11 (0.24)^a	0.31* (1.58)^a	–
6. Offset mean (rest)	0.10 (0.23)^b	0.17 (0.35)^a	–0.15 (0.30)^a	0.26 (0.79)^a	0.70*** (>10,000)^c	–
7. Exponent left STS (rest)	0.04 (0.18)^b	0.09 (0.22)^a	–0.08 (0.21)^a	0.37* (4.06)^a	0.70*** (>10,000)^c	0.52*** (>10,000)^c	–
8. Offset left STS (rest)	–0.09 (0.21)^b	0.01 (0.18)^a	0.00 (0.18)^a	0.21 (0.48)^a	0.24** (3.92)^c	0.54*** (>10,000)^c	0.70*** (>10,000)^c	–
9. Beta power left STS (rest)	–0.06 (0.19)^b	0.22 (0.53)^a	–0.28 (1.04)^a	0.03 (0.19)^a	0.19* (0.85)^c	0.20* (1.29)^c	0.50*** (>10,000)^c	0.56*** (>10,000)^c
EEG language task
5. Exponent mean (task)	0.14 (0.27)^b	0.15 (0.29)^a	–0.14 (0.28)^a	0.24 (0.64)^a	–
6. Offset mean (task)	0.10 (0.23)^b	0.19 (0.41)^a	–0.16 (0.31)^a	0.23 (0.59)^a	0.75*** (>10,000)^c	–
7. Exponent left STS (task)	0.05 (0.19)^b	0.09 (0.22)^a	–0.09 (0.21)^a	0.24 (0.64)^a	0.69*** (>10000)^c	0.58*** (>10,000)^c	–
8. Offset left STS (task)	–0.07 (0.20)^b	0.03 (0.19)^a	–0.01 (0.18)^a	0.18 (0.38)^a	0.36*** (375.45)^c	0.62*** (>10,000)^c	0.80*** (>10,000)^c	–
9. Beta power left STS (task)	–0.07 (0.20)^b	0.22 (0.52)^a	–0.25 (0.77)^a	–0.01 (0.18)^a	0.07 (0.15)^c	0.15 (0.43)^c	0.47*** (>10,000)^c	0.59*** (>10,000)^c

Note: r – Pearson’s correlation coefficient; BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0); mean – values averaged across all electrodes; left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9).
***p < 0.001 (uncorrected); **p < 0.01 (uncorrected); *p < 0.05 (uncorrected).
^an = 47; ^bn = 50; ^cn = 119.

Appendix 1—table 4

Zero-order correlations between reading, phonological awareness, rapid automatized naming, multisensory integration, and biomarkers of excitatory-inhibitory balance.

Variable	1. r(BF₁₀)	2.	3.	4.
EEG resting state
1. Reading	–
2. Phonological awareness	0.62*** (>10,000)^c	–
3. RAN	0.73*** (>10,000)^c	0.50*** (>10,000)^c	–
4. Multisensory integration	0.24* (1.44)^d	0.33** (16.94)^d	0.08 (0.17)^d	–
5. Glu	–0.10 (0.22)^b	–0.16 (0.32)^b	–0.08 (0.20)^b	–0.08 (0.21)^b
6. GABA+	–0.17 (0.35)^a	–0.02 (0.18)^a	–0.19 (0.41)^a	0.24 (0.62)^a
7. Glu/GABA+ ratio	0.03 (0.19)^a	–0.02 (0.18)^a	0.08 (0.21)^a	–0.31* (1.62)^a
8. Glu/GABA+ imbalance	–0.21 (0.47)^a	–0.06 (0.20)^a	–0.19 (0.40)^a	0.13 (0.27)^a
9. Exponent mean (rest)	–0.13 (0.30)^c	0.05 (0.13)^c	–0.08 (0.17)^c	–0.05 (0.15)^d
10. Offset mean (rest)	–0.03 (0.12)^c	0.04 (0.13)^c	–0.02 (0.12)^c	–0.01 (0.14)^d
11. Exponent left STS (rest)	–0.14 (0.37)^c	–0.01 (0.12)^c	–0.07 (0.16)^c	–0.15 (0.33)^d
12. Offset left STS (rest)	0.03 (0.12)^c	0.09 (0.18)^c	0.03 (0.12)^c	–0.07 (0.17)^d
13. Beta power left STS` (rest)	0.03 (0.12)^c	0.22* (1.96)^c	–0.04 (0.12)^c	0.04 (0.14)^d
EEG language task
9. Exponent mean (task)	–0.13 (0.32)^c	0.06 (0.14)^c	–0.14 (0.34)^c	–0.09 (0.19)^d
10. Offset mean (task)	–0.05 (0.13)^c	0.04 (0.12)^c	–0.05 (0.13)^c	–0.01 (0.13)^d
11. Exponent left STS (task)	–0.11 (0.23)^c	0.01 (0.12)^c	–0.11 (0.24)^c	–0.17 (0.48)^d
12. Offset left STS (task)	0.04 (0.12)^c	0.09 (0.18)^c	0.01 (0.12)^c	–0.07 (0.16)^d
13. Beta power left STS (task)	0.05 (0.13)^c	0.21* (1.61)^c	0.02 (0.12)^c	0.11 (0.22)^d

Note: r – Pearson’s correlation coefficient; BF₁₀ – Bayes factor indicating ratio of the likelihood of an alternative hypothesis (H1) to a null hypothesis (H0); mean – values averaged across all electrodes; left STS – values averaged across three electrodes corresponding to the left superior temporal sulcus (T7, TP7, TP9).
***p < 0.001 (uncorrected); **p < 0.01 (uncorrected); *p < 0.05 (uncorrected).
^an = 47; ^bn = 50; ^cn = 119; ^dn = 87.

Appendix 1—table 5

One-sample t-tests separately for visual (words>false fonts) and auditory runs (words>words backward) within control (CON) and dyslexic groups (DYS).

Brain regions	Hemisphere	Peak of cluster coordinates			t-Value	Number of voxels
Brain regions	Hemisphere	x	y	z	t-Value	Number of voxels
CON (n=29) visual runs (FWEc = 127)
Middle temporal gyrus, superior temporal gyrus, inferior frontal gyrus (pars triangularis, orbitalis, opercularis), supramarginal gyrus, temporal pole (superior and middle temporal gyri), postcentral gyrus, hippocampus, posterior orbital gyrus, amygdala, parahippocampal gyrus, inferior parietal gyrus, rolandic operculum, lateral orbital gyrus, pallidum, anterior orbital gyrus, inferior frontal gyrus, putamen, middle frontal gyrus, insula, angular gyrus	L	–62	–54	10	9.78	5360
Precuneus, posterior cingulate gyrus, middle cingulate and paracingulate gyri, calcarine fissure, cuneus	L/R	-4	–52	20	7.17	957
Middle temporal gyrus, superior temporal gyrus, temporal pole (superior and middle temporal gyri), inferior temporal gyrus	R	60	–34	0	7.10	1117
Cuneus, superior occipital gyrus, calcarine fissure	L/R	14	–92	26	5.98	334
Superior frontal gyrus (dorsolateral and medial)	L	-8	54	34	5.95	678
Superior frontal gyrus (medial orbital), anterior cingulate cortex (pregenual), superior frontal gyrus (medial)	L/R	-8	50	-8	5.33	194
Insula, putamen, rolandic operculum, inferior frontal gyrus (pars opercularis)	L	–36	4	6	5.09	127
Angular gyrus, middle temporal gyrus, middle occipital gyrus	L	–40	–58	26	4.56	167
DYS (n=21) visual runs (FWEc = 665)
Inferior frontal gyrus (pars triangularis, orbitalis, opercularis), posterior orbital gyrus	L	–48	24	0	6.54	665
Middle temporal gyrus, superior temporal gyrus, supramarginal gyrus	L	–62	–32	2	6.40	677
CON (n=29) auditory run (FWEc = 124)
Middle temporal gyrus, postcentral gyrus, cingulate gyrus (mid part), superior parietal gyrus, precentral gyrus, precuneus, middle occipital gyrus, supramarginal gyrus, inferior parietal gyrus, inferior temporal gyrus, calcarine fissure, fusiform gyrus, superior temporal gyrus, angular gyrus, temporal pole (superior and temporal gyri), cuneus, supplementary motor area, lingual gyrus, inferior frontal gyrus (pars triangularis, orbitalis), posterior cingulate gyrus, superior frontal gyrus (dorsolateral), inferior occipital gyrus, posterior orbital gyrus, paracentral lobule, rolandic operculum, parahippocampal gyrus, lateral orbital gyrus, middle frontal gyrus, superior occipital gyrus, anterior orbital gyrus	L/R	–52	–12	–10	9.75	13,444
Middle temporal gyrus, superior temporal gyrus, temporal pole (superior temporal gyrus), inferior temporal gyrus	R	50	–4	–20	7.98	1038
Putamen, insula, rolandic operculum, pallidum, Heschl’s gyrus, amygdala, hippocampus, thalamus (lateral geniculate)	L	–34	–20	8	6.15	463
Thalamus (mediodorsal medial magnocellular, intralaminar, pulvinar medial, ventral posterolateral, mediodorsal lateral parvocellular, ventral lateral, pulvinar anterior, lateral posterior, pulvinar inferior, medial geniculate), hippocampus, lingual gyrus, parahippocampal gyrus	L	–10	–20	0	5.82	259
Superior frontal gyrus (dorsolateral and medial), middle frontal gyrus	L/R	–6,	60	34	5.81	438
Thalamus (ventral posterolateral, pulvinar medial, pulvinar anterior, lateral geniculate, intralaminar, pulvinar inferior, pulvinar lateral, ventral lateral, medial geniculate), hippocampus	R	14	–20	4	5.68	149
Middle occipital gyrus, middle temporal gyrus, angular gyrus, superior occipital gyrus, superior temporal gyrus, superior parietal gyrus, cuneus, supramarginal gyrus, inferior parietal gyrus	R	36	–76	40	5.46	978
Cerebellar hemispheres (lobules IV, V, VI, VIII, IX, crus I, crus II), vermis (lobules VI, VII, VIII, IX)	L/R	12,	–50	–30	5.35	630
Inferior frontal gyrus (pars triangularis)	L	–50	22	20	4.54	124
DYS (n=21) auditory run (FWEc = 192)
Middle temporal gyrus, superior temporal gyrus, temporal pole (superior temporal gyrus)	L	–52	–12	–12	7.00	403
Middle temporal gyrus	L	–56	–38	0	5.33	192

Note: L – left, R – right. All results are reported at p<0.001 height threshold corrected for multiple comparisons using p<0.05 FWE cluster threshold.

Appendix 1—table 6

Logical conjunction results from paired t-tests for both visual (words>false fonts) and auditory runs (words>words backward) within control (CON) and dyslexic groups (DYS).

Brain regions	Hemisphere	Peak of cluster coordinates			t-Value	Number of voxels
Brain regions	Hemisphere	x	y	z	t-Value	Number of voxels
CON (n=29) visual and auditory runs conjunction (FWEc = 143)
Middle temporal gyrus, inferior frontal gyrus (pars triangularis, orbitalis, opercularis), superior temporal gyrus, temporal pole (superior and middle temporal gyri), supramarginal gyrus, posterior orbital gyrus, inferior parietal gyrus, angular gyrus, lateral orbital gyrus, anterior orbital gyrus, middle occipital gyrus	L	–54	–58	12	9.58	3164
Middle temporal gyrus, superior temporal gyrus	R	50	–34	–2	7.98	186
Superior frontal gyrus (dorsolateral and medial)	L	-6	56	36	5.94	430
Superior temporal gyrus, temporal pole (superior and middle temporal gyri), middle temporal gyrus	R	52	10	–16	5.76	175
Precuneus, posterior cingulate gyrus, middle cingulate and paracingulate gyri	L/R	–2	–56	30	5.48	448
Supramarginal gyrus, postcentral gyrus, rolandic operculum, superior temporal gyrus	L	–50	–24	22	5.35	217
Middle cingulate and paracingulate gyri	L/R	–2	–10	40	5.17	143
DYS (n=21) visual and auditory runs conjunction (FWEc = 545)
Middle temporal gyrus, superior temporal gyrus	L	–52	–12	–6	6.73	545

Note: L – left, R – right. All results are reported at p<0.001 height threshold corrected for multiple comparisons using p<0.05 FWE cluster threshold.

Appendix 1—table 7

Results from the flexible factorial model for both visual (words>false fonts) and auditory runs (words>words backward) between control (CON) and dyslexic groups (DYS).

Brain regions	Hemisphere	Peak of cluster coordinates			t-Value	Number of voxels
Brain regions	Hemisphere	x	y	z	t-Value	Number of voxels
CON>DYS main effect of group for visual and auditory runs (FWEc = 121)
Middle temporal gyrus, supramarginal gyrus, superior temporal gyrus, inferior temporal gyrus, inferior parietal gyrus, middle occipital gyrus, fusiform gyrus, angular gyrus, postcentral gyrus, rolandic operculum	L	–58	–30	30	7.19	2181
Rolandic operculum, superior temporal gyrus, middle temporal gyrus, supramarginal gyrus, insula, postcentral gyrus, precentral gyrus, inferior frontal gyrus (pars opercularis), putamen, inferior temporal gyrus	R	46	–4	14	7.07	1512
Superior frontal gyrus (dorsolateral and medial), middle frontal gyrus	R	22	34	54	6.56	464
Insula, rolandic operculum, superior temporal gyrus, Heschl’s gyrus	L	–40	–6	18	6.21	266
Middle cingulate and paracingulate gyri, paracentral lobule, supplementary motor area, precuneus	L/R	–8	–26	42	5.96	1108
Precentral gyrus, superior frontal gyrus (dorsolateral)	L	–32	–8	68	5.70	191
Angular gyrus, supramarginal gyrus, inferior parietal gyrus, middle occipital gyrus	R	52	–62	40	5.37	516
Superior frontal gyrus (dorsolateral and medial), middle frontal gyrus	L	–18	38	34	5.13	241
Precentral gyrus, postcentral gyrus, middle frontal gyrus	R	58	–12	46	5.09	140
Lingual gyrus, cerebellar hemisphere (lobule VI), fusiform gyrus	R	14	–68	–10	5.04	209
Temporal pole (superior temporal gyrus), middle temporal gyrus, superior temporal gyrus	L	–54	10	–16	5.00	156
Supplementary motor area, middle cingulate and paracingulate gyri	L/R	4	6	48	4.93	174
Postcentral gyrus, superior parietal gyrus	R	24	–48	58	4.85	121
Postcentral gyrus, supramarginal gyrus	R	38	–32	36	4.78	137
DYS>CON main effect of group for visual and auditory runs
No suprathreshold clusters

Note: L – left, R – right. All results are reported at p<0.001 height threshold corrected for multiple comparisons using p<0.05 FWE cluster threshold.

Appendix 1—table 8

Magnetic resonance spectroscopy (MRS) checklist.

Hardware
Field strength [T]	7T
Manufacturer	GE Healthcare
Model (software version if available)	Discovery MR 950
RF coils: nuclei (transmit/ receive), number of channels, type, body part	H¹ 32 channel head coil
Additional hardware	MR Safe Response PAD
Acquisition
Pulse sequence	Semi-Laser
Volume of interest (VOI) locations	Left superior temporal sulcus (STS)
Nominal VOI size [cm³, mm³]	15×15×15 mm³
Repetition time (TR), echo time (TE) [ms, s]	TR = 4000 ms TE = 28 ms
Total number of excitations or acquisitions per spectrum	320 averages
Additional sequence parameters (spectral width in Hz, number of spectral points, frequency offsets)	5000 Hz, 4096 points
Water suppression method	VAPOR
Shimming method, reference peak, and thresholds for ‘acceptance of shim’ chosen	Automated linear shims adjustment, 0 and first shim order only, water peak, <20 Hz
Triggering or motion correction method (respiratory, peripheral, cardiac triggering, incl. device used and delays)	N/A
Data analysis methods and outputs
Analysis software	fsl-mrs (version 2.0.7)
Processing steps deviating from quoted reference or product	fsl_mrs default pipeline+simulated basis set
Output measure (e.g. absolute concentration, institutional units, ratio) Processing steps deviating from quoted reference or product	Ratio to total creatine
Quantification references and assumptions, fitting model assumptions	The customized basis set includes: Ala, Asc, Asp, Cit, Cr, EtOH, GABA, GPC, GSH, Glc, Gln, Glu, Gly, Ins, Lac, NAA, NAAG, PCh, PCr, PE, Phenyl, Scyllo, Ser, Tau, Tyros, bHB, bHG. Macromolecules MM09, MM12, MM14, MM17, MM21 were added.
Data quality
Reported variables (SNR, linewidth (with reference peaks))	Linewidth (for metabolite group reported by fsl_mrs): mean 11.11 Hz, SD = 2.79 Hz SNR (for NAA reported by fsl_mrs): mean 7.05, SD = 23.95
Data exclusion criteria	Linewidth >20 Hz, CRLB >20% and Visual inspection (baseline, residuals) Glu – 4 subjects excluded GABA – 7 subjects excluded
Quality measures of postprocessing Model fitting (e.g. CRLB, goodness of fit, SD of residual)	%CRLB of Glu: mean 2.96, SD = 0.79 %CRLB of GABA: mean 10.59, SD = 2.76 %CRLB of NAA: 1.76 SD = 0.46
Sample spectrum

Additional files

MDAR checklist: https://cdn.elifesciences.org/articles/99920/elife-99920-mdarchecklist1-v1.docx
Download elife-99920-mdarchecklist1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Mendeley

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

Agnieszka Glica
Katarzyna Wasilewska
Julia Jurkowska
Jarosław Żygierewicz
Bartosz Kossowski
Katarzyna Jednoróg

(2025)

Reevaluating the neural noise in dyslexia using biomarkers from electroencephalography and high-resolution magnetic resonance spectroscopy

eLife 13:RP99920.

https://doi.org/10.7554/eLife.99920.4

Share this article

Cite this article

Selected results for the electroencephalography (EEG) excitatory and inhibitory (E/I) balance biomarkers.

Selected results for the magnetic resonance spectroscopy (MRS) excitatory and inhibitory (E/I) balance biomarkers.

Main effect of group CON>DYS for both visual (words > false fonts) and auditory runs (words > words backward) from the fMRI localizer task.

Relationships between electroencephalography (EEG) and magnetic resonance spectroscopy (MRS) excitatory and inhibitory (E/I) balance biomarkers.

Associations between beta power, phonological awareness, and reading.

Demographic and behavioral characteristics of the entire sample of 120 participants.

Semi-partial correlations between magnetic resonance spectroscopy (MRS) and electroencephalography (EEG) biomarkers of excitatory-inhibitory balance.

Semi-partial correlations between reading, phonological awareness, rapid automatized naming (RAN), multisensory integration, and biomarkers of excitatory-inhibitory balance.

Descriptive statistics for electroencephalography (EEG) and magnetic resonance spectroscopy (MRS) results separately for the groups.

Demographic and behavioral characteristics of the subsample of 47 participants.

Zero-order correlations between magnetic resonance spectroscopy (MRS) and electroencephalography (EEG) biomarkers of excitatory-inhibitory balance.

Zero-order correlations between reading, phonological awareness, rapid automatized naming, multisensory integration, and biomarkers of excitatory-inhibitory balance.

One-sample t-tests separately for visual (words>false fonts) and auditory runs (words>words backward) within control (CON) and dyslexic groups (DYS).

Logical conjunction results from paired t-tests for both visual (words>false fonts) and auditory runs (words>words backward) within control (CON) and dyslexic groups (DYS).

Results from the flexible factorial model for both visual (words>false fonts) and auditory runs (words>words backward) between control (CON) and dyslexic groups (DYS).

Magnetic resonance spectroscopy (MRS) checklist.

MDAR checklist

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)