**7. Results**

According to the studies so far discussed, different techniques have proven to be useful in the identification of endophenotypes or potentially promising biomarkers. Social deficits, as well as cognitive, anatomical, and/or neurophysiological findings are encouraging and in some cases are associated with candidate genes. Endophenotypes may help detect children who are at higher risk for ASD. This could promote better and earlier diagnosis and treatment, as well as the establishment of prevention strategies. It is also probable that new drug targets can be identified and that personalized treatments can be available.

The precise genetic, environmental, and epigenetic factors that increase susceptibility to ASD may vary in several ways from one affected child to another. For this reason, it is very important to continue the search for means that allow us to form less heterogenous groups of patients. Endophenotypes, at least in some cases, could contribute to achieve this goal. They could help recognize risk factors that participate or affect common pathways or processes, such as chromatin remodeling and transcription that may affect neurodevelopment.

In some instances, de novo mutations associated with ASD may be highly penetrant, and in absence of a family history of ASD, endophenotypes may not be observed in relatives, but could still contribute to an earlier identification of symptoms in the carrier of the mutation; the next generations could also benefit from early assessments and prevention strategies. On the other hand, there will be families in which relatives will probably share enough risk factors with the patient, as to exhibit one or more endophenotypes or ASD.

### **8. Future research**

Further studies in patients and relatives, preferably of first and second degree, will help determine if biomarkers fulfill the criteria of heritability and cosegregation.

Knowledge about the neuroetiology of ASD will be enriched by research in areas such as neuroanatomy, physiology, and social behavior. Multiple investigations are underway, and risk factors for ASD will continue to be identified, possibly in more homogenous groups of patients.

Future research should ideally involve longitudinal research in large samples that include twin pairs, nuclear families and/or extended pedigrees. These conditions may facilitate the identi‐ fication or confirmation of endophenotypes, in which the criteria of heritability and cosegre‐ gation are fulfilled. Whenever possible, it will be useful to obtain detailed information about clinical and demographic aspects, not only regarding the individual with ASD, but about first and second degree relatives. While the grouping of individuals in a dichotomous way, as affected or unaffected will continue to be evaluated, and offers advantages, this approach can be complemented with the evaluation of endophenotypes [46].
