The UK-based Neuroscience in Psychiatry Network (NSPN) cohort consists of over 2000 adolescents between the ages of 14 to 24 sampled from the general population. A total of ... participants completed the first round of questionnaires, ... completed the second and ... completed the third wave. Intervals between the waves ranged from ... years to ... years, with an average of ... years. These intervals do not match well with those in the IMPACT trial. Moreover attrition rates in the NSPN sample are high which makes imputation difficult if not impossible. Therefore we chose to use only baseline data and assume that in healthy adolescents, properties of the network of depressive symptoms do not change over a time interval of 36 weeks. To test this assumption, we selected all participants in the NSPN cohort that have three measurements each one year apart. For this group, we estimate the network at each of the time points and compared each level to its subsequent level. As shown in Figure SX, overall network strength does not change between t1-t2 (network strength = ... vs. ..., p=...) or between t2-t3 (network strength = ... vs. ..., p=...). Thus, in a large group of adolescents measured over three subsequent years, global network strength does not change. Therefore, using the control group for comparison to patients both before and after treatment (less than one year apart) is valid.