Table 7.
Comparison of minimally detectable effect (MDE) and reported effect sizes of the largest studies (Ntotal and Nautism(f)) by neuroimaging modality.
| Imaging Modality (Ntotal and/or Nautism(f)) | Study | Method (‘sex’ or ‘gender’ is based on the term used in the study) | Metrics/Outcome measure | Sample size (M:F) | Minimally detectable effect size (MDE; Cohen’s d)aα = 0.05 power = 0.8 | Reported effect size (Cohen’s d) where P-values are significant at a level <0.05 |
||
|---|---|---|---|---|---|---|---|---|
| sex/gender*dx NOTE: F converted to Cohen’s d (Lenhard and Lenhard, 2016) | sex/gender-stratified (M) | sex/gender-stratified (F) | ||||||
| DTIb Ntotal = 213 |
Zeestraten et al., 2017 (MRC-AIMS) United Kingdom |
ROI sex*dx + sex stratified |
FA | 98 autism (61:37) 115 TD (61:54) |
sex*dx (0.39) sex-stratified (males, 0.51) (females, 0.60) |
Anterior segment of AF left (0.41); right (0.46) Long segment of AF left (0.32) CING left (0.38); right (0.33) Uncinate left (0.43); right (0.49) IFOF left (0.34); right (0.34) |
Anterior segment of AF left (0.49); right (0.41) Long segment of AF left (0.37) CING left (0.50) Right (0.43) Uncinate left (0.52); right (0.48) IFOF left (0.47); right (0.45) Posterior segment of AF left (0.31); right (0.29) ILF left (0.41); right (0.34) |
No significant diagnostic effect in females |
| DTIb Nautism(f) = 55 |
Irimia et al., 2017 (GENDAAR) | whole-brain sex*dx | GM thickness, volume, cortical area, mean curvature, CD | 110 autism (55:55) 83 TD (43:40) |
sex*dx (0.41) | CD (0.19) | N/A | N/A |
| MEG Ntotal = 75 Nautism(f) = 8 |
Yoshimura et al., 2021 | (whole-brain) gender* dx |
bilateral auditory cortical response (P1m) | 29 autism (21:8) 46 TD (41:5) |
gender*dx (0.66) | No significant gender*dx interaction found | N/A | N/A |
| MRS Ntotal = 174 Nautism(f)=15 |
O’Neill et al., 2020 | Near whole-brain sex*dx |
metabolite concentration (N-acetyl compounds, glutamate + glutamine, creatine + phosphor-creatine, choline compounds) | 78 autism (63:15) 96 TD (69:27) |
sex*dx (0.43) | No effect size reported (only reported p = 0.001) | N/A | N/A |
| PET Ntotal = 121 Nautism(f) = 4 |
Mitelman et al., 2018 | ROI group* sex 3 × 2 ANOVA |
GM and WM metabolic rates | 25 autism (21:4) 41 schizo-phrenia (32:9) 55 TD (29:26) |
group*sex (0.57)a | no significant group*sex interaction; d = 0.068; p = 0.79 | N/A | N/A |
| rs-EEG Ntotal = 46 Nautism(f) = 3 | Saunders et al., 2016 | whole-brain group* gender 4 × 2 ANOVA |
128-channel EEG oscillation coherence |
13 autism (10:3) 10 anxiety (2:8) 11 ADHD (7:4) 12 TD (7:5) |
group*gender (1.02)a | Interaction between gender and experimental group for interhemispheric coherence scores that was approaching significance: Alpha eyes closed frontal-frontal d = 0.231; p = 0.053 Alpha eyes open central-central d = 0.219; p = 0.072 Theta eyes open central-central d = 0.218; p = 0.074 |
N/A | N/A |
|
rs-fMRI Ntotal =1491 Nautism(f) =91 |
Tomasi and Volkow, 2019 (ABIDE) |
whole-brain sex stratified |
lFCD, LI, ALFF, seed-voxel correlation maps |
656 autism (565:91) 835 TD (602:233) |
sex-stratified (males, 0.16) (females, 0.35) |
N/A |
lFCD in anterior thalamus (ASD < TD) (PFWE < 0.005, effect size: 0.1807 < d < 0.3034, df = 1162) |
none reported; no significant group differences in thalamic ln(IFCD) between autistic and TD females (P = 0.55) |
| sMRI Ntotal =3607 Nautism(f) =274 |
Postema et al., 2019 (ENIGMA) |
whole-brain sex*dx + sex stratified |
structural brain asymmetry for multiple brain regional and global hemispheric measures (i.e., cortical thickness, cortical surface area, subcortical volume) |
1778 autism (1504:274) 1829 TD (1400:429) |
sex*dx (0.09) sex-stratified (males, 0.10) (females, 0.22) |
rostral anterior cingulate thickness asymmetry index (d = 0.11) |
rostral anterior cingulate thickness asymmetry index (d = -0.17, P = 1.4 × 10–5) |
no group differences in rostral anterior cingulate thickness asymmetry index (d = 0.11, P = 0.165) |
|
task fMRI Ntotal =154 Nautism(f) =39 |
Lawrence et al., 2020b (GENDAAR) |
whole-brain and ROI sex*dx + sex stratified |
social reward processing during instrumental implicit learning task |
82 autism (43:39) 72 TD (39:33) |
sex*dx (0.45) sex-stratified (males, 0.63) (females, 0.67) |
ROI analyses: no sex*dx interaction found Whole-brain analysis: no sex*dx interaction found |
ROI analyses: no differences found between autism and control groups when stratified by sex Whole-brain analysis: no differences found between autism and control groups when stratified by sex |
ROI analyses: no differences found between autism and control groups when stratified by sex Whole-brain analysis: Left frontal/insular cluster (d = 0.87) Right insular/temporal cluster (d = 0.89) Left frontal cluster (d = 0.80) |
| TMS-EEG Ntotal =42 Nautism(f) =12 |
Kirkovski et al., 2016b | ROI sex stratified |
cortical function and connectivity | 22 autism (10:12) 20 TD (11:9) |
sex stratified (males, 1.29) (females, 1.30) |
N/A | no differences found between autism and control groups when stratified by sex | no differences found between autism and control groups when stratified by sex |
Abbreviations: ABIDE, Autism Brain Imaging Data Exchange; AF, arcuate fasciculus; ALFF, amplitude of low-frequency fluctuations; CD, connectivity density; CING, cingulum; CSF, cerebrospinal fluid; ENIGMA, Enhancing Neuro Imaging Genetics through Meta-Analysis; GABA, gamma-aminobutyric acid; GENDAAR, Gender Exploration of Neurogenetics and Development to Advance Autism Research; GM, gray matter; IFOF, interior frontal occipital fasciculus; ILF, inferior longitudinal fasciculus; lFCD, local functional connection density; LI, laterality index; MRC-AIMS, Medical Research Council Autism Imaging Multicentre Study; PFC, prefrontal cortex; ROI, region-of-interest; TPJ, temporo-parietal junction; WM, white matter.
Due to limitations of the R pwr package for MDE calculations for 3 × 2 and 4 × 2 ANOVA designs, the MDE was calculated assuming equal sample sizes using WebPower (Zhang & Yuan, 2018). This value is expected to be smaller than the MDE for a study with unequal sample sizes.
DTI study conducted by Zeestraten et al., (2017) had the largest total sample size but did not have the largest autistic female sample. A second DTI study (Irimia et al., 2017) which had the largest autistic female sample size was therefore also included.