INT adj problems were predicted by a three-way interaction of PH severity, squared basal cortisol, and chronic stress, which was marginally significant for parent-report, p = 0.06, and significant for self-report, p = 0.006, as shown in Table 4 .
An interaction effect of chronic stress with linear rather than squared basal cortisol approached marginal significance in predicting parent-reported INTadj given a very severe PH, p = 0.10, and was significant in predicting self-reported INTadj given a very severe PH, p = 0.013, as well as parent-reported and self-reported INTadj given no PH, p = 0.010 and p = 0.007, respectively.