In this issue of JAMA Psychiatry, Gur et al1 evaluated differences in age-related neurocognitive test performance and psychosis symptoms in the context of a large populationbased cohort. The rationale for this study grew out of observations that psychotic symptoms are relatively common in the general population2 and that individuals who have had a recent emergence of psychoticlike symptoms, but of subpsychotic intensity, are at increased risk for progressing to full psychosis within a few years of ascertainment.3,4 Given that such prodromal cases show neurocognitive deficits similar to those observed in firstepisode schizophrenia,4,5 the Gur et al1 study sought to extend evidence of this link to participants with psychotic symptoms as ascertained through primary care, independently of seeking treatment for psychiatric indications. The relationships (ie, regression weights) between test performance and age in typically developing children (TD; N = 1963) were used to model predicted neurocognitive age scores in the participants who reported significant psychotic symptoms (psychosis spectrum [PS]; n = 1423), more limited psychotic symptoms (psychosis limited [PL]; n = 898), and symptoms of other, nonpsychotic disorders (other psychiatric disorders [OP]; n = 981). Plotting predicted neurocognitive age against chronological age bymeans of growth charts, the PS group was observed to have lower predicted age compared with the TD and OP groups across all age levels represented (8-21 years), a pattern that held generally but that was most pronounced for predicted age scores based on measures of complex cognition and social cognition. ThePSgroup alsohad lowerpredictedneurocognitiveage than thePLgroup. The results are interpreted as evidence of developmental lags amongyouth reportingpsychotic symptomsandas augmenting the literatureshowingcognitivedeficits inschool-agedchildren later diagnosed as having psychosis.6 That the relationship between psychotic symptoms and neurocognitive performance extends to samples ascertained from the general population helps to support the generalizability of findings from studies that evaluate clinic-derived samples, whether those be samples at clinical high risk for psychosis4,5 or thosewith established schizophrenia.7 The results also support the notion that the risk for psychosis varies in thegeneral population8; therefore, the search for causal factors contributing to the risk for psychosis (eg, common genetic variants) could be done in large samples screened from the general population.9 Do the results of the Gur et al1 study indicate that neurocognitive deficits precede psychosis or show a progressively deviant trajectory fromnormal development? Because this is a cross-sectional studyandthere isno information (yet)on longer-term outcome, there is no ability of the current results to comment on the prediction of psychosis nor do the authors make this claim. Criteria for inclusion in the PS group includedbothsubthresholdandthreshold-levelpsychotic symptoms; it is not clearwhether the samepattern of resultswould be observed when limiting observation to those with subthreshold symptomsor to thosemeeting criteria for a prodromal risk syndrome.10 Theoretically, one could use the growth chartmethod to evaluatewhether the age trajectories of those who progress to aDSM-5 psychotic disorder differ from those whodonot (and toTD), but doing sowould require having information on long-termoutcome of the samples (and excluding anyonewho had already reached a fully psychotic state at baseline). As the results now stand, they provide a means of showing that the presence of psychotic symptoms is associated with poorer neurocognitive function at every age in the samplingframeof thestudy. In thissense, theyprovideameans to test the robustness of thepsychosis-neurocognitive impairment link; that is, how consistently is this relationship observed across childhood and adolescence? The answerwould appear to be quite consistently, at least for measures of complex and social cognition. The growth charts presented do not show a pattern of increasing divergence of predicted neurocognitive age with increasing chronological age; that is, there is generally the same degree of age-related performance decrement across childhood and adolescence for those who report psychotic symptoms. Does this mean that neurocognitive deficits do not progress among those who develop psychosis? It is important to keep in mind that the cross-sectional nature of the study designprovides abasis for evaluatingonly thecontemporaneous, rather thanpredictive, relationshipsbetweencognitionandpsychosis. However, if informationwere available on the duration of psychotic symptoms in this cohort, it could be possible to evaluate whether neurocognitive age is more deviant among caseswith longerdurationsof symptoms.Suchapatternwould beconsistentwith thenotionofaprogressivedeteriorativeprocess in psychosis but would still not be interpretable as evidence that such deterioration occurs prior to psychosis onset, for which a prospective, longitudinal design is required. Theuseofgrowthchartsprovidesanefficient andcompellingbasis forsummarizingtheconsistencyofcognitivedysfunction in those with psychotic symptoms across childhood and Related article page 366 Opinion