key: cord-0951524-0yequagp
authors: Toh, Wei Lin; Sumner, Philip J.; Meyer, Denny; Neill, Erica; Phillipou, Andrea; Tan, Eric J.; Van Rheenen, Tamsyn E.; Rossell, Susan L.
title: Investigating predictors contributing to the expression of schizotypy during the COVID-19 pandemic
date: 2022-04-03
journal: J Psychiatr Res
DOI: 10.1016/j.jpsychires.2022.03.060
sha: 7f1b3f0fe3dc36a00ed0316249268a60a647964b
doc_id: 951524
cord_uid: 0yequagp

The coronavirus (COVID-19) pandemic has caused major disruptions to social and other forms of functioning, which may influence schizotypy expression. The current study aimed to explore possible distal and proximal predictors contributing to schizotypy in a sample of the Australian general population during the COVID-19 pandemic. The COvid-19 and you: mentaL heaLth in AusTralia now survEy (COLLATE) project is an online mental health study aimed at tracking key mental health indicators over the progression of the pandemic. Adults residing in Australia were invited to take part using non-discriminative snowball sampling. Demographic-clinical information was collected for 850 participants in either October 2020 or January 2021. To assess schizotypy facets, the Launay-Slade Hallucinations Scale-Extended (LSHS-E) and Peters Delusions Inventory (PDI-21) were used to measure hallucination and delusion proneness respectively. Generalised linear models (with gamma and negative binomial distributions) were employed. Age, negative emotions and loneliness significantly contributed to both hallucination and delusion proneness; gender, education and religiosity also significantly contributed to delusion proneness, in the final regression models. Our study corroborated the specific contribution of loneliness, amongst other factors, in the prediction of schizotypy facets. Tackling loneliness represents a public health challenge that needs to be urgently addressed, especially in the face of the ongoing COVID-19 pandemic.

the progression of the pandemic. Adults residing in Australia were invited to take part using non-23 discriminative snowball sampling. Demographic-clinical information was collected for 850 participants in 24 either October 2020 or January 2021. To assess schizotypy facets, the Launay-Slade Hallucinations Scale-25

Extended (LSHS-E) and Peters Delusions Inventory (PDI-21) were used to measure hallucination and 26 delusion proneness respectively. Generalised linear models (with gamma and negative binomial 27 distributions) were employed. Age, negative emotions and loneliness significantly contributed to both 28 hallucination and delusion proneness; gender, education and religiosity also significantly contributed to 29 delusion proneness, in the final regression models. Our study corroborated the specific contribution of 30 loneliness, amongst other factors, in the prediction of schizotypy facets. Tackling loneliness represents a 31 public health challenge that needs to be urgently addressed, especially in the face of the ongoing COVID-32 19 pandemic. 33 et al., 2021; Fekih-Romdhane et al., 2021) , maladaptive coping (Bortolon et al., 2021; Fekih-Romdhane et 94 al., 2021) , loneliness (Bortolon et al., 2021) and social isolation (Alle and Berntsen, 2021; Fekih-95 Romdhane et al., 2021) significantly predicted schizotypal experiences during the COVID-19 pandemic. 96

Although the aforementioned findings are consistent with a COVID-19-related exacerbation of 97 schizotypy, the replicability of these findings and their generalisability to Australia is not yet clear. 98

Australia has been relatively successful in containing the spread of COVID-19, and mitigated the number 99 of infections and deaths, but its population has had to endure some of the strictest social isolation 100 measures in the world (described earlier). As such, it provides an excellent context for investigating 101 associations between loneliness and social isolation, and schizotypy. Furthermore, there is currently very 102 little information regarding whether some pandemic-related disruptions are associated with particular 103 schizotypy dimensions, or whether they implicate schizotypy globally. Only Knolle et al. (2021) and 104 Bortolon et al. (2021) have investigated individual schizotypy dimensions separately, and they reported 105 evidence of dimension-specific relationships. For instance, Bortolon et al. (2021) found that the frequency 106 of experiencing paranoia during lockdown was predicted by a different set of variables to those that 107 predicted the frequency of experiencing hallucinations. This difference included loneliness, which 108 predicted paranoia but not hallucinations. 109

Therefore, the aim of the current study was to investigate whether self-reported life disruptions 110 during COVID-19 were associated with hallucination-and delusion-like experiences in an Australian 111 general population sample. Our research question focused on exploring whether proximal changes in 112 employment, financial status, work location, negative emotions, social contact and loneliness are 113 significantly associated with hallucination-and delusion-like experiences, while controlling for existing 114 distal factors (i.e. age, gender, education, religiosity, living situation, and pre-existing physical and mental 115 health) that are known to influence the expression of these schizotypy facets. Based on previous findings, 116

we expected to find that more self-reported financial and employment disruptions (including not being pandemic, after accounting for the effects of the aforementioned distal influences. 120

The COvid-19 and you: mentaL heaLth in AusTralia now survEy (COLLATE) project was 123 launched on 1 April 2020, as a nationwide study aimed at tracking the mental health of Australians amidst 124 the COVID-19 pandemic. This project has been described elsewhere Tan et al., 125 2020) but in brief, comprises 13 online surveys, activated for 72 hours at the start of each month, followed 126 by a series of follow-up surveys over the next four years. Members of the general public residing in 127

Australia, aged 18 years or older, were invited to complete the survey via social media advertising and 128 other online networks, participant registries held by Swinburne University of Technology as well as non-129 discriminative snowball sampling stemming from these initial recruitment methods. Past respondents were 130 encouraged to participate in each new round of surveys, but new respondents who had not previously 131 taken part were also accepted. This serial cross-sectional design permitted timely snapshots across 132 multiple points to gain a broad understanding of population mental health as the COVID-19 situation 133

The current study utilised data collected in October 2020 and January 2021, as questions about 135 hallucination-and delusion-like experiences, and loneliness were asked in these two months. This was 136 based on a pre-determined survey design, which involved a brief core battery of key survey questions, 137 alongside a secondary list of questionnaires that were inserted/removed in line with a regular rotation 138 schedule. All January 2021 participants as well as unique October 2020 participants (who did not respond 139 to January survey) were retained. This was done because the January response rate was lower, and we 140 wanted to match sample sizes for the two time points as much as possible, whilst ensuring each participant 141 was included only once. To provide additional context around these time points, Victoria was coming study received ethics approval from the Swinburne University Human Research Ethics Committee 147 (#20202917-4107), and complied with the Declaration of Helsinki. Respondents provided online informed 148 consent, and collected responses were anonymous. 149

Two broad areas were examined: i) sociodemographic information, and ii) mental health status. 151

Basic sociodemographic information was collected, including: age, gender, education, religiosity, 152 employment status (whether adversely impacted by COVID-19) and work location (whether working from 153 home), living situation (whether residing alone or with others), household income in the past fortnight, 154 reduction in social contact owing to COVID-19 (in terms of number of hours), as well as whether 155 respondents had a pre-existing physical/medical condition (yes/no), and/or were a person with lived 156 experience of a mental illness (yes/no). The response categories for each of these sociodemographic 157 variables are presented in Table A (supplementary section) . 158

Mental health status was assessed further using several measures. Negative emotions were evaluated using 159 the Depression Anxiety Stress Scales (DASS-21), a 21-item self-report measure, rated on four-point Likert 160 scales (0-3), comprising three subscales: depression, anxiety, stress (Lovibond and Lovibond, 1995) . 161

Loneliness was gauged by the abbreviated University of California, Los Angeles Loneliness Scale 162 (UCLA-LS), comprising two positively worded and two negatively worded items, rated on four-point 163

Likert scales (1-4; Russell et al., 1980) . Hallucination and delusion proneness were assessed using the 164

respectively. The LSHS-E is a 16-item measure, rated on five-point Likert scales (0-4), assessing 166 multisensory hallucinatory experiences in the general population, with higher summed scores indicating 167 increased hallucination proneness (Vellante et al., 2012) . The PDI is a 21-item multidimensional measure 168 of the propensity for delusional thinking based on atypical beliefs or vivid mental experiences, tapping 169 into themes involving reference, persecution, grandiosity, religion-supernatural, mind-reading, control, 170 jealousy, sin-guilt, somatic, thought alienation and nihilism (Peters et al., 2004) . Questions describe

Statistical analyses were conducted using IBM SPSS Statistics, version 27. To contextualise our 176 results, we first reported pertinent demographic and clinical information relevant to our sample, and 177 compared our mean hallucination and delusion proneness scores with those of the respective original 178 validation studies (Peters et al., 2004; Vellante et al., 2012) . To examine the influence of distal and 179 proximal factors, two sets of generalised linear models were employed to identify predictive factors 180 contributing to hallucination and delusion proneness. A gamma distribution for LSHS-E and a negative 181 binomial distribution for PDI-21 were assumed to account for the high degree of skewness in these 182 distributions. Analyses were performed across the entire sample collapsed across the two time points (with 183 duplicate respondents removed) to gain a continuum understanding of these experiences and beliefs during 184 the COVID-19 outbreak. Variables of interest were assigned as distal (i.e. pre-existing factors typically 185 associated with sociodemographics) or proximal (i.e. relatively state-based and assessed over a recent 186 period of time) to COVID-19, prior to being entered into the model. Distal predictors (age, gender, 187 education, religiosity, living situation, and pre-existing physical and mental health conditions) were 188 entered in Block 1; and proximal predictors (finances and employment status, work location, reduced 189 social contact, negative emotions and loneliness) were entered in Block 2; with hallucination-and 190 generalised linear model of LSHS-E scores and 805 observations for the generalised linear model of Participants had a mean age of 35.9 years, with a standard deviation of 13.0 years (range 18-84 201 years; see Table A Scores from all three DASS subscales were highly inter-correlated in the current dataset (.65 ≤ rs 208 ≤ .72, all p < .001). Thus, only total DASS scores were entered into subsequent analyses. Mean LSHS-E 209 was slightly higher in our sample (M = 11.3, SD = 11.4) than the mean of 10.7 obtained in the original 210 validation study (Peters et al., 2004; Vellante et al., 2012) . Conversely, mean PDI-21 was somewhat lower 211 in our sample (M = 3.4, SD = 3.5) than the mean of 6.7 obtained in the original validation study (Peters et 212 al., 2004; Vellante et al., 2012) . Notably, these schizotypy dimensions declined with age and higher 213 education level, and increased with physical and mental illness, loneliness and negative emotions. PDI-21 214 was also significantly higher for females than males, and for those who endorsed greater religiosity, lower 215 income and not working from home (see Table C in the supplementary section for the relevant correlation 216 matrix). 217 Table 1 shows the results of the two generalised linear models elucidating which predictive factors 218 significantly contributed to hallucination-and delusion-like experiences. For LSHS-E, age, education, 219 religiosity, living situation, physical health and mental illness were significant distal predictors in Block 1. 220

Gender did not significantly predict LSHS-E scores. Of the significant predictors form Block 1, only age 221 remained significant in Block 2, with negative emotions and loneliness being significant proximal 222

predictors. The final model was significant (χ 2 [14] = 264.1, p < .001). For PDI-21, age, gender, education, 223 religiosity, physical health and mental illness were significant distal predictors in Block 1. Living situation 224 did not significantly predict PDI-21 scores. Age, gender, education and religiosity remained significant in

The current study aimed to investigate how the expression of schizotypy facets, specifically 229 hallucination-and delusion-like experiences, were associated with factors that were distal and proximal to 230 the COVID-19 pandemic . Our hypothesis that financial and employment disruptions, loneliness and 231 reduced social contact, and negative emotions would be associated with more hallucination-and delusion-232 like experiences was only partly supported . Loneliness and negative emotions predicted both facets of 233 schizotypy independently of sociodemographic factors distal to the pandemic. However, whilst household 234 income and the ability to work from home were both associated with more delusion-like experiences (but 235 not hallucination-like experiences), these associations were weak and were not independent of the distal 236 sociodemographic factors.. 237

the final regression models, while female gender, higher education levels and reduced religiosity were also 239 associated with delusion-like experiences. These findings are broadly consistent with previous literature. 240

For instance, younger age and lower education are often associated with increased schizotypy (Binbay et 241 al., 2012; van Os et al., 2009) , including hallucination-and delusion-like experiences specifically (Knolle 242 et al., 2021) , and these relationships continue to be found during the pandemic . 243

Moreover, whilst schizotypal experiences tend to be slightly more common in males than females (Binbay 244 et al., 2012; van Os et al., 2009) , Knolle et al. (2021) reported more anomalous experiences and beliefs in 245 females than males during the pandemic when modelled with other demographic predictors. The fact that 246 pre-existing physical and mental health predictors (significant in the first step of both regressions) were no 247 longer significant in the final models signifies possible mediation effects of proximal predictors entered in 248 the second step, as supported by preliminary studies (Daimer et al., 2021; Knolle et al., 2021) . It is noted 249 that slightly different factors were involved in predicting hallucination-like experiences versus delusion-250 like experiences. In particular, certain sociodemographic factors (gender, religiosity, household income 251 and working from home) seemingly influenced the development of delusion, but not hallucination, Of the significant predictors identified in the final models, proximal factors associated with 255 negative emotions, reduced social contact and loneliness were of special interest, as these represent 256 possible psychological outcomes stemming from the pandemic. Preliminary research supports the notion 257 that negative affect (Daalman and Diederen, 2013; Laroi et al., 2012) and loneliness (Le et al., 2019; 258 Michalska da Rocha et al., 2018; Narita et al., 2020) may increase the likelihood of transition to psychotic 259 illness in the face of heightened schizotypy. Although loneliness and social isolation are often conflated or 260 used interchangeably, it is of note that these constructs are distinct; being socially isolated does not 261 necessarily equate to feeling lonely and vice versa. This is exemplified in our analysis, where loneliness, 262

but not reduced social contact, significantly contributed to our two schizotypy dimensions. Indeed, the 263 correlation between loneliness and reduced social isolation did not survive correction for multiple 264 comparisons (Table B , supplementary section), though we do note that our measure of reduced social 265 contact since the onset of the pandemic (relative to pre-pandemic levels) does not necessarily equate to 266 social isolation per se. Regardless, there are steps people can still take to tackle loneliness, including 267 regularly engaging with loved ones virtually or through other means. In light of this, raising public 268 awareness about the importance of managing negative emotions and feelings of loneliness during these 269 challenging times might be of benefit. 270

Relative to figures reported in the respective original validation research (Peters et al., 2004; 271 Vellante et al., 2012) , our mean scores for LSHS-E were similar, but our mean scores for PDI-21 were 272 somewhat lower. This could be attributed to demographic differences of samples involved across these 273 studies (e.g. Australia versus Italy and the UK, general population versus student cohorts, cohort effects 274 over time, etc.). By contrast, the DASS scores currently sampled were similar to those reported in a 275 previous study earlier in the pandemic, and these levels of negative emotions were substantially high 276 compared to pre-pandemic normative data from Australia . Psychological research in 277 previous pandemics has suggested that some adverse impacts to mental health may only emerge after a 278 prolonged time lag, and could persist for significant periods thereafter (Ayers and Yellowlees, 2008) . impact on schizotypy expression will likely depend on a myriad of factors moving forward, such as 282 personal, social or economic losses and other unfavourable events, such as further lockdowns. Moreover, 283 the impact of the COVID pandemic on people's mental health may be exacerbated in those experiencing 284 socioeconomic disadvantage (O'Sullivan et al., 2020) . Further research is thus required to elucidate these 285 complex interrelationships. 286

The current study had several limitations. First, we did not measure schizotypy scores before the 287 onset of the COVID-19 pandemic in the current sample. Moreover, although we categorised loneliness 288 and negative affect as being proximal to the pandemic, the grouping of these variables encompasses their 289 potential for change (relative to the distal factors) based on the timeframe used for their assessment (i.e. 290 self-reporting the occurrence of experiences within the past week or the past four weeks). In the absence 291 of pre-pandemic data, we cannot determine with certainty whether the reported levels of schizotypy, 292 loneliness and negative affect represent changes that have occurred since the onset of the pandemic (i.e. 293 COVID-related), or whether they have remained stable despite the pandemic (i.e. intransient despite 294 COVID). Indeed, even comparisons with pre-pandemic normative data is limited for the reasons 295 mentioned above. Second, the fluctuating time course and severity of differing pockets of COVID-19 296 outbreaks across the various Australian states meant that we were unable to accurately account for state-297 wise variations, even though it was apparent that Victoria (where the majority of our respondents resided) 298 had borne the brunt of COVID-infected numbers, related fatalities and lockdowns. As an added point of 299 consideration, our combined data across two time points may help to balance out some of the differences 300 owing to disparities in location/timing of COVID impacts in Australia, although loneliness, negative affect 301 and schizotypal experiences appeared relatively stable amongst people who completed the survey at both 302 time points. Third, our variable involving reduced social contact attributed to COVID restrictions was 303 employed as a proxy for social isolation, but the validity of this assumption may be questioned, depending 304 on how social isolation has been defined within the context under investigation. Finally, we did not rate 305 dimensional distress, preoccupation and conviction for the PDI-21 in the current study, owing to 306 constraints around study design and administration time. Having this information would have been COVID-19 pandemic. 309

The current study design permitted a series of cross-sectional indicators of population mental 310 health over the course of the COVID-19 outbreak, but was not longitudinal in nature. Future research 311 would benefit from focusing on further longitudinal studies aimed at fully elucidating the complex 312 interplay amongst schizotypy expression, negative affect and loneliness. Despite the devastation wreaked 313 by the pandemic, ensuing lockdowns and other social restrictions enacted do offer an unfortunate, but 314 unique, opportunity to study resultant effects on constructs such as schizotypy expression, impacted by 315 these events. This is especially so in countries like Australia, where medical aspects of COVID-19 have 316 been relatively well-managed. In fact, this begets the question of whether nations facing less severe 317 COVID-19 outbreaks may observe smaller changes in population schizotypy levels, with continued 318 longitudinal research imperative in capturing these longer-term effects. Related to this, future studies may 319 also examine other predictors not assessed in the current study, for instance involving the consumption of 320 alcohol, tobacco or other illicit substances and social media use, where preliminary results suggesting 321 significant influences exist (e.g. Knolle et al., 2021) . Constructive findings from this avenue of research 322 may be applied to mitigate the potentially adverse impact of negative psychological and social variables 323 spurring the transition of typical schizotypy expression to serious psychotic illness. 324

In summary, negative emotions and loneliness were associated the expression of hallucination-325 and delusional-like experiences during the COVID-19 pandemic. Given the adverse influences on our 326 general health, as well as specific impact in relation to increased schizotypy, public health campaigns to 327 tackle these negative psychological outcomes, with dedicated interventions targeting loneliness, might be 328 warranted moving forward. 329 Table 1 Generalised linear models elucidating significant predictors contributing to hallucination-and delusion-like experiences (N=805-850)

Delusion 

Self-isolation, psychotic symptoms and cognitive problems during the COVID-19 worldwide outbreak

Mental health considerations during a pandemic influenza outbreak

The role of schizotypy in the study of the etiology of schizophrenia spectrum disorders

Auditory verbal hallucinations and continuum models of psychosis: A systematic review of the healthy voice-hearer literature

Testing the psychosis continuum: differential impact of genetic and nongenetic risk factors and comorbid psychopathology across the entire spectrum of psychosis

Persecutory ideation and anomalous perceptual experiences in the context of the COVID-19 outbreak in France: what's left one month later?

The prevalence of domestic violence among women during the COVID-19 pandemic

A final common pathway to hearing voices: Examining differences and similarities in clinical and non-clinical individuals

Are we back to normal yet? The impact of the COVID-19 pandemic on mental health with a specific focus on schizotypal traits in the general population of Germany and the UK, comparing responses from

A comprehensive review of auditory verbal hallucinations: lifetime prevalence, correlates and mechanisms in healthy and clinical individuals. Front Hum Neurosci 7, 367. Department of Health, 2020. Framework for National Reopening

Loneliness in schizophrenia: Construct clarification, measurement, and clinical relevance

Culture counts: the diverse effects of culture and society on mental health amidst COVID-19 outbreak in Australia

A cognitive model of the positive symptoms of psychosis

Social networks, support and early psychosis: a systematic review

A social deafferentation hypothesis for induction of active schizophrenia

Loneliness and Psychotic Symptoms: The Mediating Role of Depression

The impact of the COVID-19 pandemic on mental health in the general population: a comparison between Germany and the UK

The characteristic features of auditory verbal hallucinations in clinical and nonclinical groups: Stateof-the-art overview and future directions

The importance of loneliness in psychotic-like symptoms: Data from three studies

The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories

Schizotaxia, schizotypy and schizophrenia

Loneliness in Psychosis: A Meta-analytical Review

Testing the Validity of Taxonic Schizotypy Using Genetic and Environmental Risk Variables

The context of delusional experiences in the daily life of patients with schizophrenia

Loneliness and psychotic experiences in a general population sample

The Impact and Implications

An Australian Perspective

Measuring delusional ideation: The 21-item Peters et al. delusions inventory (PDI)

The SPQ: a scale for the assessment of schizotypal personality based on DSM-III-R criteria

An overview of current mental health in the general population of Australia during the COVID-19 pandemic: Results from the COLLATE project

The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence

Social defeat: risk factor for schizophrenia?

The social defeat hypothesis of schizophrenia: an update

How the COVID-19 pandemic is focusing attention on loneliness and social isolation

Exploring the associations between dimensions of schizotypy and social defeat

The impact of the COVID-19 pandemic on negative symptoms in individuals at clinical high-risk for psychosis and outpatients with chronic schizophrenia

Social networks and support in first-episode psychosis: exploring the role of loneliness and anxiety

Using multivariate statistics Pearson Education

Considerations for assessing the impact of the COVID-19 pandemic on mental health in Australia

Family violence and COVID-19: Increased vulnerability and reduced options for support

A systematic review and meta-analysis of the psychosis continuum: evidence for a psychosis pronenesspersistence-impairment model of psychotic disorder

The authors would like to thank all the participants who took the time and effort to take part in this study, especially during these challenging and unprecedented times. WLT (GNT1161609) and AP (GNT1159953) are supported by National Health and Medical Research Council (NHMRC) New Investigator Project Grants. SLR holds a NHMRC Senior Fellowship (GNT1154651), and EJT (GNT1142424) and TVR (GNT1088785) hold NHMRC Early Career Fellowships.