In the present study, we have shown that a high level of schizotypy predicts a poor quality of life in adults with ASD. Moreover, we found that the negative effect of schizotypy was indirect and, apparently, largely mediated by anxiety and depressive symptoms. Although the causal direction of these relationships cannot be settled from our cross-sectional design, it seems highly plausible that more schizotypal traits result in poor quality of life, rather than the reverse. Our study is also one of the first to assess schizotypy in subjects with ASD. To our knowledge, only Spek & Wouters (2010) and Barneveld et al. (2011) have published on this, using SPQ in small samples consisting of 21 and 27 individuals with autism, respectively.
It is well known that adults with ASD experience low overall quality of life (van Heijst et al., 2015; Hesselmark et al., 2013; Mason et al., 2018). Compared to studies using the same quality of life measures as we did, our participants scored considerably lower than healthy controls (Lindner et al., 2016; Janssen and Szende, 2013). In addition, our participants reported lower quality of life when compared to autistic children (Ten Hoopen et al., 2020). Presumably, with age people gain insight on what to expect of life and subsequently become disillusioned, which is reflected by the low scores on quality-of-life measures. Expectedly, another contributor to poor quality of life was, apart from severity of autistic and schizotypy symptoms, psychiatric comorbidity. Mental health comorbidity predicts poor quality of life in people with ASD (Mason et al., 2018) and psychiatric comorbidity could even be viewed as a sign of overall severity (Kessler et al., 2005), likewise linked to quality of life.
Many studies report strong correlations between anxiety and depression, and poor quality of life (González-Blanch et al., 2018; Rapaport et al., 2005, Schonfeld et al., 1997). Depression is hypothetically a common pathway for quality of life, i.e., a mechanism by which all factors found to affect quality of life are mediated. In fact, depression has been shown to impact quality of life across conditions and influence life quality to a greater extent than physical illness does (Hsu et al., 2014; Martino et al., 2019; Son et al., 2009). Our findings could be understood analogously. We found that a large part of the negative effect on quality of life, driven by schizotypal traits, was indeed mediated through depression and anxiety symptoms. Our participants presented high scores on the anxiety and depressive measure HADS, which was strongly associated with increase in schizotypy and autism scores. HADS is intended to assess state, but in ASD, HADS is perhaps rather a trait measure. Depressive, anxiety and obsessive-compulsive symptoms are presumably inherent to the pathological mechanisms behind ASD and may explain why the effect of antidepressants often is disappointing (Howes et al., 2018; Reddihough et al., 2019; Williams et al., 2013). Very few controlled pharmacological treatment studies for depression and anxiety have been carried out on patients with ASD (Williams et al., 2013), which may be associated with low expectations on drug response in the ASD population.
With an international outlook, the SPQ-BR scores reported by our participants may seem modest. However, to decide whether a particular score is high or not we should only compare with people of similar cultural background. One reason for the low scores among Swedes could be religion. Belief in God is positively correlated with positive schizotypal traits (Crespi et al., 2019), and Sweden is considered as the least religious country in the world (https://worldpopulationreview.com). Furthermore, in a recent study we investigated SPQ-B, a 22-item version of the SPQ, and found that Swedes rate themselves considerably lower than most other populations across 14 nations, worldwide (Bejerot et al., 2020). Consequently, in a Swedish cultural context, individuals with ASD experience schizotypy symptoms to a high degree.
We found a substantial association between ASD and schizotypy among our participants, i.e., individuals with more autistic traits rated themselves as having more schizotypal traits as well. Considering how much the gestalt of schizotypy resembles ASD, this is not surprising. A nosological link between ASD and schizoid/schizotypal personality traits has been observed and debated for almost a century. Clinical descriptions of schizoid psychopathy in childhood (Sukhareva, 1926), schizotypal personality disorders in children (Nagy & Szatmari 1986) and ‘schizoid’ personality of childhood (Wolff, 1991) are all relevant to this. One reason for introducing the concept of schizotypy was to identify persons with elevated risk of psychosis. Accordingly, assessment instruments for schizotypy, including the SPQ, usually convey the psychopathological structure of schizophrenia, i.e., positive, negative, and disorganised symptoms. The negative symptoms of schizophrenia (and “negative schizotypy”) are generally viewed as closely related to schizoid personality and (at least superficially) very similar to core symptoms of ASD. The positive symptoms are usually believed to discriminate better between schizophrenia and ASD, which was supported by Spek & Wouters (2010). However, Barneveld et al. (2011) argued that their ASD subjects had elevated measures on all three higher order constructs of the SPQ, to levels similar to patients with schizophrenia, and that all three were related to measures of autism. Based on a considerably larger sample of ASD subjects, our correlations of Positive and Negative, as well as Disorganised schizotypy with RAADS-14 corroborate the findings of Barneveld et al., based on a considerably larger sample of ASD subjects. However, when we take a closer look at the relationship between the SPQ-BR Positive schizotypy and Disorganized constructs, only two of their subscales (Suspiciousness and Eccentric behavior) showed a highly significant relationship with autistic traits, while no other subscales within these constructs did. This suggests that the overlap between ASD and schizotypy is mainly driven by Interpersonal schizotypy (i.e., the Negative schizotypy and the Social Anxiety constructs), supporting Spek & Wouters’ findings (2010) (Table 3).
The diagnostic boundaries between autism and schizophrenia spectrum disorders are still unclear and different diagnostic traditions may come to different conclusions. High levels of schizotypy may balance the diagnosis towards a schizophrenia spectrum disorder and reversely, the prodromal schizophrenia state, characterized by a broad range of cognitive deficits that predate the onset of clinical symptoms, may suggest an autism. Thus, local traditions may decide whether a person will be diagnosed with ASD or schizophrenia spectrum disorder. In Sweden, similar to the rest of the Western world, there has been a tremendous increase in the diagnosis of ASD in people with intelligence within the normal range, possibly linked to a decrease in the schizophrenia population characterised as deficit schizophrenia or a ‘neurodevelopmental’ subtype, for a review on schizophrenia subtypes, see Jablensky (2006).
Significant schizotypy traits can be expected to lead to poor treatment response, which has been clearly shown in obsessive-compulsive disorder (Moritz et al., 2004). It is presumably of similar importance in the treatment of individuals with ASD. The considerable overlap between schizotypy and ASD needs to be considered. Prominent schizotypal traits in people with ASD may constitute an endophenotype coinciding with a particularly poor quality of life.
However, quality of life can improve through various forms of interventions (Hesselmark et al., 2013; Pallathra et al., 2019), thus adults with ASD ought not be disqualified because of high levels of schizotypy.
Limitations
There are some limitations to our study. First, almost all data was collected through self-report. This may lead to unreliable responses due to e.g., misinterpretations. Although this limitation is well known, self-reports are regularly used in psychiatric research and in clinical practice. Hesselmark conducted a study investigating the reliability and validity of self-reported questionnaires in adults with ASD, and supported the use of these (Hesselmark et al., 2015). In fact, among individuals with ASD, self-reports could be preferred; it is presumably less stressful to sit peacefully by oneself and respond, without time limits and distractions, than being interviewed.
A second limitation was our skewed gender proportion with a majority of women. Although ASD is more common in men than in women, women tend to seek psychiatric treatments more often than men, which may explain the female preponderance in our study. Gender did not show any association with schizotypy scores and the sex distribution was very similar in the SPQ high and low group; thus, the gender bias should not have affected the results of our study.
Third, other researchers have suggested general health and social relationships to affect quality of life beyond depression (Villas-Boas et al., 2019). Unfortunately, we did not use any measures to quantify general health or quality of social relationships in our study, which is another limitation. However, our participants were rather young, which makes them less likely to be affected by physical illness. Moreover, several items in the SPQ-BR concern social relationships; hence, this was intertwined in our assessment.
Fourth, we did not examine executive functioning. In a recent study, sustained attention differed between individuals with schizotypal personality disorder and individuals with ASD (Abu-Akel et al., 2020). Preferably, clinical, and behavioural phenotypes should be investigated in the endeavour to delineate the underlying mechanisms in clinical populations with substantial phenotypical overlaps.
Fifth, we used a life quality questionnaire, the BBQ, which has not been validated in an autism sample. Ideally, we should have chosen the well-established World Health Organisation measure (WHOQoL-BREF) combined with autism-specific items (ASQoL) adapted for people with autism (McConachie et al., 2018), however, the ASQoL was not available when we designed our study. We were concerned about the risk of dropouts if we set the threshold for inclusion too ambitiously. BBQ covers six domains in life and consists of 12 items whereas a measure such as the WHOQoL-BREF combined with ASQoL adds up to almost three times as many items. Furthermore, the BBQ resembles the more elaborate 32-item Quality of Life Inventory (QOLI, Frisch et al., 1992). Both BBQ and QOLI measure life satisfaction rather than lack of symptoms and impairment. They ask about domains in life, which is followed by inquiring whether this domain is important to the person or not. In sum, BBQ and OQLI both ask if the person is satisfied with the parts of life that matter most, not what others may think is important. We believe this design enables inclusion of idiosyncrasies that are experienced by many autistic individuals. In a previous CBT study among adults with ASD (Hesselmark et al., 2013), the participants did not express any difficulties using the QOLI, but the instrument was nevertheless regarded as too extensive to be included in the present internet-based study.
Sixth, the evidence for the validity of a schizotypy measure in autism populations could be questioned. However, in our post-hoc analysis we tried to separate schizotypy from autistic traits by excluding comorbid, mainly negative schizotypy traits. The impact of schizotypy still influenced symptoms of depression, anxiety, and quality of life. For the purpose of studying the validity of schizotypy constructs in ASD in future studies, the use of the full item versions of RAADS-R and SPQ would be preferable, rather than the abbreviated ones that we applied in the present study.
Clinical implications
Our findings underscore the importance of identifying schizotypy in individuals with ASD. Schizotypy, anxiety and depressive symptoms, possibly inherent to ASD, contribute to the low quality of life often reported by people with ASD and may account for the poor response rate to antidepressants (Howes et al., 2018). By taking a broad approach towards the difficulties often experienced by people with ASD, instead of focusing on a limited area, we have reasons to believe that quality of life can be improved. This was done in an ambitious, group-based cognitive behavioural psychotherapy project (ALMA), resulting in improvements in quality of life among adult participants with ASD (Hesselmark et al., 2013; Bejerot and Björnstjerna, 2019). For individualised treatments, Nidotherapy (Tyrer, 2019), a method to explore life preferences and thereby improve quality of life, is presumably a good choice, although still understudied. In Nidotherapy the individual selects the goals, and these are reached by changing the environment, not the individual’s thoughts or behaviours. Because people with ASD vary greatly in interests, ambitions, and abilities, one should keep an open mind regarding which methods to choose to improve life satisfaction in this population.