Introduction
Schizophrenia and schizophrenia spectrum disorders (SSD) remain a
prominent public health risk. There are persistent gaps in knowledge
surrounding the biological mechanisms of symptoms affecting cognition,
behavior, and emotion. To track the emergence and progression of
symptoms, there is a need for reliable biomarkers of schizophrenia and
SSD. These biomarkers can reveal underlying mechanisms as
well as complementing existing clinical interviewing methods. Biomarkers
may also be used to predict conversion from high-risk to diagnosis
(Näätänen et al., 2016). Currently there is no gold standard biomarker
that reliably tracks symptom severity (Lieberman et al., 2019), although
there is convergence regarding risk factors and cognitive deficits
(Mohn-Haugen et al., 2022). This has led to many calls for multivariate
biomarker metrics to more comprehensively assess an individual’s risk
level (e.g., (Cannon et al., 2016; Price et al., 2006; Seidman, Shapiro,
et al., 2016). Although this opinion is not controversial,
there remains a tendency to focus on single biomarkers and their
predictive power. This may be due to the simplicity of understanding how
a single biomarker reflects underlying processing schizophrenia and SSD,
and to the added technical difficulties and financial challenges
involved when measuring multiple biomarkers. Here, we
support the call for a multivariate biomarker, and we extend this view
by suggesting that to accelerate our understanding of the neural
mechanisms associated with symptoms, it would be valuable to test across
the entire SSD, including neurotypical individuals with high
subclinical schizotypy symptomatology in
the general population . This suggestion has the added
benefit of increasing the pool of researchers available to investigate
questions of clinical relevance, but who do not have access to clinical
populations.
Here, we define schizotypy as the subclinicalcharacteristics associated with schizophrenia that are prevalent within
the general population. Schizotypy in this context is not a diagnosis as
the schizophrenia-related traits are not severe enough to warrant
intervention. It is unknown if participants with high schizotypy traits
will develop schizophrenia and/or psychosis in the future, as many of
these studies are cross-sectional in design and recruit
from the general population . However, one value of focusing on
schizotypy is that participants with psychiatric or neurological
diagnosis can be removed. This reduces the likelihood of comorbid
conditions and medications that may confound the findings. Therefore, it
would be expeditious to include a larger swath of the
spectrum to include the larger subclinical population, and to
combine discrete biomarkers that track symptom severity.
Existing biomarkers include: genetic polymorphisms (Ettinger et al.,
2014), oculomotor abnormalities (reviewed in: (Levy et al., 2010)), and
atypical neural responses detected using EEG and fMRI. Unfortunately,
many of these biomarkers were identified in isolation rather than in
tandem with other biomarkers associated with a particular sensory or
cognitive domain. This is despite success when taking a multifactorial
approach to monitoring multiple neural (Kent et al., 2004; Price et al.,
2006; Ranlund et al., 2018; Taylor et al., 2017) and behavioral
(Seidman, Shapiro, et al., 2016) measures to help distinguish between
schizophrenia and other psychosis-related conditions such as bipolar
disorder. Due to the success seen in those diagnosed with
schizophrenia, we recommend the same multifactorial approach for the
subclinical schizotypy population. For the current review, we revisit
EEG biomarkers from several related literatures: abnormal sensorymemory (SM) and working memory (WM), in both auditory and visual
modalities. These topics have well-developed literatures with a
substantial foundation in the clinical schizophrenia and SSD
populations. We focused on sensory and working memory specifically, to
explore how these biomarkers may interact due to downstream deficits
impacting later processing. We discuss the use of sensory and working
memory measures as biomarkers of symptom severity from a cross-sectional
perspective - how do these markers reflect on their internal processes
at this moment? We conclude by discussing how some of the
biomarkers we discuss may be promising as predictive biomarkers of the
onset of psychosis and schizophrenia, and briefly summarize some of the
literature up to this point.