Figure 4. ACE and TD effects on within-network connectivity in replication timeseries.

a. To probe the validity of our results, we tested our hypotheses regarding the impact of ACE and TD on functional connectivity using a separate set of task-regressed timeseries data acquired in a subsample of our participants. As in the discovery dataset, we used a model that evaluated the impact of ACE group, TD status, and ACE group x TD status on within-network connectivity to test our hypotheses that ACE would be associated with lower connectivity and that tryptophan depletion would selectively lower connectivity only in the high ACE group. The effect of ACE group was robust and consistent with discovery results. Within-network connectivity was significantly lower in the high ACE group during both sham TD (p=0.002) and active TD (p=0.05). The effect of TD in the low ACE group was also replicated and remained non-significant. However in contrast to discovery results, there was no significant ACE group x TD status interaction and no effect of TD in the high ACE group. Bars represent least-square means adjusted for age, time since last menstrual period, estradiol level, global connectivity, and motion. Error bars represent standard error. b. We also attempted to replicate our findings regarding the continuous effects of ACE using a model that evaluated the impact of total ACE score, TD status, and ACE score x TD status. Consistent with our hypothesis and analyses using ACE group, total ACE score was significantly negatively associated with within-network connectivity during sham TD (p=0.008). As in the discovery dataset, this association did not reach significance during active TD. However, it was present at a trend level (p=0.1), and the interaction between ACE score and TD status was not significant. The effect of all covariates included in analyses have been regressed from the connectivity values depicted. ACE= adverse childhood experiences; TD = tryptophan depletion; *p ≤ 0.1, **p ≤ 0.5, ***p ≤ 0.01