{Reference Type}: Journal Article {Title}: "Average is good, extremes are bad" - Non-linear inverted U-shaped relationship between neural mechanisms and functionality of mental features. {Author}: Northoff G;Tumati S; {Journal}: Neurosci Biobehav Rev {Volume}: 104 {Issue}: 0 {Year}: 09 2019 {Factor}: 9.052 {DOI}: 10.1016/j.neubiorev.2019.06.030 {Abstract}: Traditionally, studies emphasize differences in neural measures between pathological and healthy groups, assuming a binary distinction between the groups, and a linear relationship between neural measures and symptoms. Here, we present four examples that show a continuous relation across the divide of normal and pathological states between neural measures and mental functions. This relation can be characterized by a non-linear inverted-U shaped curve. Along this curve, mid-range or average expression of a neural measure is associated with optimal function of a mental feature (in healthy states), whereas extreme expression, either high or low, is associated with sub-optimal function, and occurs in different neural disorders. Neural expression between the optimal or intermediate and pathological or extreme values is associated with sub-optimal function and at-risk mental states. Thus, this model of neuro-mental relationship can be summarized as "average is good, extremes are bad". By focussing on neuro-mental relationships, this model can facilitate the transition of psychiatry from a categorical to a dimensional and individualized approach needed in the era of precision medicine.