Table 3. Metabolite concentrations in BG, DLPFC and parietal cortex [¹H]MRS voxels and group differences.

					ANOVA		Post hoc t-tests—P-valuesb
Region	Metabolitea	Narrow ASD	Broader ASD	Control	F	P-values	Narrow ASD vs controls	Broad ASD vs controls	Broad ASD vs narrow ASD
Basal ganglia	Glx	10.12 (1.14)	10.43 (1.26)	12.34 (1.45)	12.63	<0.0001c	<0.001	0.001	0.25
	Cho	1.205 (0.19)	1.098 (0.14)	1.324 (0.19)	5.83	0.006d	0.212	0.004	0.44
	Cr	5.097 (0.65)	5.323 (0.53)	6.435 (1.25)	10.02	0.0003c	<0.001	0.004	0.47
	NAA	5.454 (0.81)	5.486 (0.42)	6.458 (1.52)	4.54	0.017d	0.030	0.045	0.32
DLPFC	Glx	7.492 (1.11)	7.156 (1.51)	8.007 (1.42)	1.36	0.27	—	—	0.32
	Cho	1.129 (0.21)	1.110 (0.16)	1.271 (0.27)	2.35	0.11	—	—	0.93
	Cr	4.074 (0.48)	4.021 (0.49)	4.602 (0.60)	5.32	0.009d	0.027	0.017	0.64
	NAA	5.954 (0.79)	5.955 (0.47)	6.660 (0.95)	4.07	0.025d	0.047	0.057	0.22
Parietal	Glx	10.607 (2.79)	9.790 (1.65)	11.028 (1.62)	1.31	0.28	—	—	0.41
	Cho	1.112 (0.37)	1.048 (0.13)	1.149 (0.25)	0.53	0.59	—	—	0.524
	Cr	5.461 (1.42)	5.140 (0.81)	5.382 (0.70)	0.39	0.68	—	—	0.19
	NAA	7.998 (2.32)	7.472 (0.76)	7.887 (1.06)	0.48	0.62	—	—	0.31

Abbreviations: ANOVA, analysis of variance; ASD, autism spectrum disorder; BG, basal ganglia; Cho, choline; Cr, creatine; DLPFC, dorsolateral prefrontal cortex; Glx, glutamate and glutamine; [¹H]MRS, proton magnetic resonance spectroscopy; NAA, N-acetylaspartate.

^aData are expressed as mean (s.d.).

^bPost hoc t-tests comparing ASD cases with controls were only performed when ANOVA was significant at uncorrected P<0.05.

^cBold figures show values significant at P=0.05 level, with Bonferroni correction over 12 comparisons, that is, uncorrected P<0.0047.

^dSignificant at P=0.05 level before Bonferroni correction.