LAUSR

In March, Walter et al. (2018) published much-needed findings on schizotypy-influences on stressand cortisolresponse using a socialevaluative stressor. Their findings showed no effect of schizotypy on perceived stress, but blunted cortisol-responses in highly schizotypal individuals. These results are highly relevant for understanding schizophreniadevelopment, but have important limitations due to the used psychometric inventory: A core-aspect of schizophrenia-liability is cognitive slippage (Grant et al., 2018), which is not assessed by the inventory the authors used. Furthermore, the dimensions and items of said inventory conflate schizotypy with aspects related to Neuroticism or distressproneness; e.g., the positive dimension consists of items simultaneously tapping into both schizotypy and Neuroticism (paranoid ideation). The results, thus, cannot account for influences of cognitive slippage, which is of relevance, as the sub-facets of schizotypy are not linearly independent. We therefore attempted to replicate these findings, but using an inventory that, additionally, captures cognitive slippage (O-LIFE: Mason et al., 1995), using the same stress-paradigm as Walter et al. in a mixed-sex student sample (55 females, 30 males; aged 18–37 [mean age 22.5+/− 3.8 years]). An investigation of the influences of schizotypal facets on the areas under the curve for cortisol and stress (both baselinecorrected and not baseline-corrected) showed that “Cognitive Disorganisation” (CogDis; tapping both into cognitive slippage and Neuroticism) explained major variance regarding both perceived stress and cortisol-response: With respect to the baseline, CogDis was linearly associated with higher perceived stress throughout the paradigm (β= .23; p= .027) but not cortisol-response. Increases in cortisol, however, were linearly associated with CogDis similarly as in Walter at al.: increasing CogDis-values predicted more blunted cortisol-responses (β=−.22; p= .045). Ex post single item analyses, however, showed that the stress-association was primarily due to items associated with cognitive slippage, while the blunted cortisol-response was more influenced by items tapping into Neuroticism. As those questionnaire-dimensions most strongly associated with blunted cortisol-responses in Walter et al (cognitiveperceptual, interpersonal) are also closely associated with distressproneness/Neuroticism, we would argue that our results support their findings, albeit that the effect is not one of schizotypy but of Neuroticism. Vice versa, our findings of an association between cognitive slippage and perceived stress are not in conflict with those by Walter et al., as their null-results can simply be explained by the lack of items capturing cognitive slippage. In conclusion, further research is surely called for, but our results strongly emphasise the necessity for inclusion of measures of cognitive slippage in schizotypy-research as well as the necessity to consider the lack of such items in certain inventories when interpreting schizotypyrelated results.