Exploration of 6-scanner FA values

Dataset characteristics

SSD

CMH, n=82 (35%) CMP, n=3 (1%) MRC, n=42 (18%) MRP, n=36 (15%) ZHH, n=29 (12%) ZHP, n=43 (18%) p
Sex (F:M) 26:56 2:1 10:32 8:28 13:16 17:26 .168
Age 28.24 (7.79) 27.33 (4.73) 37.67 (10.83) 34.25 (9.88) 37.03 (9.20) 32.19 (11.63) .000
Education 13.51 (2.00) 14.33 (0.58) 13.05 (2.07) 13.69 (2.25) 13.41 (2.47) 14.26 (2.26) .190
Parental education 15.15 (3.17) 12.33 (6.35) 14.91 (2.85) 16.39 (2.75) 14.79 (3.40) 14.88 (2.83) .142
SANS total score 21.66 (12.01) 14.33 (11.59) 30.62 (13.07) 27.17 (11.37) 23.62 (9.37) 24.47 (12.06) .002
Average FA 0.43 (0.02) 0.37 (0.01) 0.40 (0.02) 0.37 (0.02) 0.43 (0.02) 0.37 (0.02) .000

HC

CMH, n=45 (30%) CMP, n=1 (1%) MRC, n=25 (16%) MRP, n=28 (18%) ZHH, n=19 (12%) ZHP, n=34 (22%) p
Sex (F:M) 19:26 1:0 9:16 14:14 10:9 18:16 .612
Age 27.20 (8.45) 20.00 (NA) 37.04 (11.14) 34.36 (11.74) 33.16 (9.45) 35.12 (12.31) .002
Education 15.36 (1.85) 15.00 (NA) 15.84 (1.75) 15.64 (2.09) 15.16 (2.46) 15.82 (1.98) .781
Parental education 16.20 (2.48) 16.00 (NA) 15.92 (2.73) 15.24 (2.55) 16.17 (2.79) 15.78 (2.49) .774
Average FA 0.43 (0.01) 0.37 (NA) 0.41 (0.01) 0.38 (0.02) 0.44 (0.01) 0.38 (0.02) .000

Statistics

anova(lm(AverageFA_FA~scanner+age+sex+education+par_education, data = df))
## Analysis of Variance Table
## 
## Response: AverageFA_FA
##                Df   Sum Sq  Mean Sq  F value    Pr(>F)    
## scanner         5 0.248198 0.049640 176.5171 < 2.2e-16 ***
## age             1 0.006122 0.006122  21.7686 4.439e-06 ***
## sex             1 0.001190 0.001190   4.2331  0.040410 *  
## education       1 0.002811 0.002811   9.9963  0.001711 ** 
## par_education   1 0.000612 0.000612   2.1776  0.140962    
## Residuals     338 0.095051 0.000281                       
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

This result shows that group (diagnosis), scanner, age, and sex have significant effect on whole-brain FA (p values <.05).

I also checked if there are any significant interactions between any combination of the variables included in the model: there was one interaction between sex and parental education.

Visualization of raw data

FA boxplots by scanner

FA density plots by scanner

Overlaid FA density plots

```