The effect of antipsychotic medications on acoustic startle latency in schizophrenia

Prepulse inhibition of the acoustic startle reflex (PPI) is extensively studied as a biomarker of schizophrenia (SCZ); however, antipsychotic medication can confound the measure. Latency, the time between the startling stimulus and the reflexive eye blink, provides an index of neural processing speed and is 90% heritable. SCZ subjects have slower latency than controls (CON). This study examined the effects of antipsychotic medication on startle latency. 108 CON and 132 SCZ subjects in three medication subgroups (94 on second-generation antipsychotics (SGA), 25 on first-generation antipsychotics (FGA), 13 unmedicated (NoMed)) were tested on a standard acoustic startle paradigm designed to measure startle magnitude, PPI, and latency. Latency was slower in SCZ compared to CON subjects (p=0.005). Latency did not differ between the three SCZ medication groups. When CON were added to that model, both the NoMed subjects (p=0.04) and the SGA subjects (p=0.003) were slower than CON subjects. For PPI, CON did not differ from SCZ analyzed as a single group. When SCZ subjects were divided into medication groups, PPI was lower in NoMed subjects than the CON group (p=0.03), the SGA group (p=0.02) and the FGA group (p=0.05). SCZ subjects on any medication did not differ from CON. Thus, latency was partially normalized by antipsychotic medication, but this did not obscure the slower latency in SCZ compared to CON. Therefore latency is both trait and state related, whereas medication normalized PPI and obscured any difference between SCZ and CON.