Table 3. Comparison of extreme groups with high and low numbers of accumulated proautistic genotypes in the male GRAS sample regarding autism‐relevant measures and control variables
1.5–2.5 proautistic genotypes N = 69–73a5.5–6.5 proautistic genotypes N = 57–68aZ, T value P‐value
Autism‐relevant measures (mean ± SD)
Social withdrawal2.10 ± 1.223.10 ± 1.67 = −3.71
=   0.0002
Mannerism1.27 ± 0.791.68 ± 1.22 = −2.88
=   0.003
Blunted affect & poor rapport2.53 ± 1.073.16 ± 1.38 = −2.80
=   0.005
Speech production1.81 ± 1.202.49 ± 1.48 = −2.84
=   0.003
Stereotyped thinking & preoccupation1.99 ± 1.042.48 ± 1.13 = −2.84
=   0.005
Difficulties in abstract thinking2.58 ± 1.592.96 ± 1.62 = −1.45
= 0.146
PAUSSb−0.22 ± 0.520.21 ± 0.74 = −3.69
=   0.0002
Control variables (mean ± SD)
FMR1 repeat length29.33 ± 4.7928.90 ± 4.63 = −1.44
= 0.151
FMR2 repeat length17.41 ± 3.1017.89 ± 3.72 = −0.41
= 0.681
Age36.97 ± 10.7938.74 ± 14.01 = −0.69
= 0.488
PANSS positive11.99 ± 4.7713.70 ± 6.41 = −1.39
= 0.165
Cognitive composite scorec−0.07 ± 0.84−0.08 ± 1.02 = 0.44
= 0.661
CNIc0.06 ± 1.060.14 ± 0.95 = −0.58
= 0.562
  • Statistically significant P‐values are set in boldface (Bonferroni‐corrected significance threshold: P < 0.007).

  • Mean ± SD presented. For statistical analysis, Mann–Whitney U‐test or t‐test for normally distributed variables was used. PANSS, positive and negative syndrome scale; CNI, Cambridge Neurological Inventory.

  • a Because of missing data, sample sizes vary.

  • b Z‐standardized score.

  • c Corrected for age and chlorpromazine equivalents (standardized residual after linear regression).