Successful implementation of the low risk design, the most powerful approach that is not contingent upon parental phenotypes or the presence/absence of a particular outcome (the usual scenario in prevention trials), presents several barriers which are easy to overcome. The first is the establishment of common measures of resistance. An important way of defining a high resistance phenotype is via developing a continuous index of liability to the disorder and selecting individuals from the low end of the distribution. Such an index, for instance, has been developed for measurement liability to drug addiction in children, before any exposure to drugs (Vanyukov, Tarter et al., 2003; Vanyukov, Kirisci et al., 2003; Vanyukov et al., 2009; Vanyukov et al., 2016). A second related barrier is the development and use of items, scales and constructs that measure not just the “bad” but also the “good”. A vestige of the disease era of research is the widespread existence of items, test and scales designed to measure only a small part of the liability distribution, usually most informative around the threshold between “normal” and “disease”.
