Then, each group (occasional/regular) was stratified according to the age of first use (before/after 18 years of age) in order to assess whether the age at which cannabis use started influences the magnitude of the decreased gray matter volume localized with the two-sample t-test. The four subgroups were composed of 13 occasional smokers who started cannabis consumption before the age of 18 (called here ‘Occasional early'), and 9 who started after (called ‘Occasional late'); 16 regular users who started cannabis consumption before the age of 18 (called here ‘Regular early'), and 9 who started after (called ‘Regular late'). Gray matter volume, represented by the value of the modulated GM image, was compared between the so-defined subgroups. We tested significant differences between the ‘Occasional late' group and all of the other groups (ie, three tests for each region) at a level of p<0.05 (corrected for multiple comparisons). For each comparison between subgroups, we also computed the effect size (d) by dividing the difference between mean values by the common SD of the two groups (Cohen, 1988).
