Disparities in Psychosis Risk Symptoms for New Zealand Māori May Be Explained by Systemic Stressors and Inappropriate Conceptualization of Culturally Normative Experiences

Abstract

Background and Hypothesis

Māori, the indigenous peoples of New Zealand, experience increased rates of psychotic disorders and first-episode psychosis. However, it is unclear whether they also present with increased psychosis risk symptoms, such as subclinical psychotic-like experiences (PLEs). Measurement of risk symptoms is key for early intervention. Further, it is unclear if systemic factors such as the increased rates of social adversity and discrimination or cultural biases contribute to this disparity in psychosis rates.

Study Design

This study surveyed 466 18- to 30-year olds in New Zealand, and compared Māori to non-Māori participant responses on the Prodromal Questionnaire Brief, alongside the history of childhood trauma, discrimination, and financial adversity.

Study Results

Māori reported a higher number of PLEs compared to non-Māori—however, this was not associated with increased distress related to these experiences. The increased number of psychosis-like experiences reported by Māori was likely explained by systemic factors such as childhood trauma, discrimination, and financial stress. Māori were more likely to report that the PLEs were positive.

Conclusions

Measurement of psychosis risk for Māori is nuanced, and increased scores on these tools may reflect pathologizing potentially normative experiences for Māori, such as spiritual encounters or discrimination, alongside the impact of increased rates of systemic discrimination, trauma, and financial stress.

Māori, the indigenous peoples of New Zealand Aotearoa are overrepresented in the first episode psychosis population. Latest population estimates suggest 17.1% of the national population identify as Māori¹, and yet they represent 40% of those presenting with first-episode psychosis.² Māori also receive higher rates of schizophrenia diagnoses³ and appear to be over represented in certain groups who are at higher risk for psychosis, such as those with schizotypy.⁴ These data mirror heightened rates of psychotic diagnoses in other indigenous and minority groups internationally. For example, Black Americans,⁵ Hispanic Americans,⁵ and Indigenous Australians⁶ are over represented in those with schizophrenia diagnoses. Despite this clear disparity, we know very little about the types of early psychosis risk symptoms experienced by Māori, or why these disparities might be occurring.

The duration of time that psychosis symptoms are experienced by an individual before receiving treatment is the best predictor of later outcomes, and as such early intervention for psychosis is vital.⁷ Early intervention requires the identification of early psychosis risk symptoms and syndromes⁸ occurring prior to the first full threshold episode of psychosis. An example of such symptoms is psychotic-like experiences (PLEs), subclinical psychosis symptoms that are similar to the positive symptoms seen in full-threshold psychotic disorders, such as hallucinations but are less severe and debilitating (eg, hearing whispers when no one is present⁹). PLEs often occur in the psychosis prodrome, prior to a first episode. Heightened rates of PLEs are reported by Black and Hispanic Americans^10,11, but rates of PLEs are unknown for Māori.

There are two common hypotheses for the causes of psychosis disparities between cultural groups that may apply to Māori. The colonization of New Zealand resulted in land alienation, economic impoverishment, cultural marginalization, and systemic racism for Māori.¹² As a result, they continue to experience increased rates of social adversity, discrimination, and intergenerational trauma.¹³ The first hypothesis for psychosis disparities is that systemic factors such as those Māori have experienced result in increased stress, which is known to be linked to the development of psychosis.^14–16 This hypothesis has been supported by a recent international meta-analysis comparing ethnic minority position and psychosis symptoms,¹⁷ and other recent findings in the United States showing racial disparities in rates of PLEs were associated with discrimination, police violence exposure, and lower educational attainment.¹¹ The second hypothesis is that increased rates of psychosis may reflect ethnic bias, or poor cultural knowledge in diagnostic systems¹⁸ and healthcare models.¹⁹ Diagnostic systems and mental health treatment systems were developed based on western worldviews and historically have been incongruent with Māori values such as holistic health care, unity, and the importance of whānau (family in its broadest sense, including wider community or support system and ancestors).²⁰ Diagnostic systems rely on consideration of what experiences are “normative” or “problematic” and do not necessarily account for cultural differences. For example, some Māori may interpret experiences such as hearing voices of loved ones who have died as being culturally appropriate and a gift,²¹ whereas for New Zealand Europeans, this experience would typically be considered a PLE. Further, diagnoses are given by health care practitioners, who can experience both implicit and explicit bias^19,22 and may lack adequate cultural knowledge.²³ Importantly, international disparities in psychosis are apparent when examining differences in diagnosis rates and differences in non-diagnostic symptoms, such PLEs.¹¹ This suggests such disparities cannot purely be explained by diagnostic biases. However, it could be argued that the measurement of non-diagnostic psychosis experiences such as PLEs still relies on cultural norms, as subclinical experiences are identified if they are often seen in prodromal psychosis, a concept emerging from a western worldview. It is unclear whether systemic factors might explain rates of early psychosis symptoms for Māori, and/or whether cultural biases might impact rates of diagnoses or symptoms.

The most common method for measuring whether someone is at risk for psychosis is through a clinical interview by a trained health professional using measures such as the Structured Interview for Prodromal Symptoms (SIPS), which assesses subclinical psychosis symptoms (eg, hearing whispered voices or suspiciousness) and changes in functioning.²⁴ As the SIPS requires a full clinical interview, it is time consuming and has a high resource load and is not suited to screening large groups for psychosis. Early psychosis screening tools include the Prodromal Questionnaire – Brief Version (PQ-B; Loewy et al ²⁵), which is a self-report tool measuring subclinical psychosis symptoms. However, the PQ-B has not been validated for indigenous groups.²⁶ It is unclear if such tools measure experiences that are culturally normative and occur as part of regular life for Māori.

This study aims to examine whether Māori report higher rates of psychosis risk symptoms on the PQ-B compared to other ethnic groups within a community sample from New Zealand, and if so, (1) whether this difference is explained by increased rates of systemic factors such as childhood trauma, discrimination or financial stress, and (2) whether the difference may be explained by inappropriate conceptualization of culturally normative PLEs. The latter will be measured by analyzing if any items/experiences are reported as more or less distressing by Māori, and if Māori are more likely to report certain items as positive experiences.

Methods

Sample and Procedures

Data were collected from a community sample of young adults (aged 18–30 to capture adults at the highest risk for early psychosis) living in New Zealand using an online survey. Participants were recruited using widespread social media recruitment using advertisements targeted to this age range through Facebook and Instagram. The survey was run using Qualtrics from May 16, 2022 to July 27, 2022 and was accessed by participants through QR code. Participation was voluntary, and there was no compensation or reward for taking part. Participants were excluded if PQ-B was not completed (n = 80), age requirements were not met (n = 5), age information was missing (n = 34), and if more than two measures were incomplete (n = 2), leaving a final sample of 466 participants. There was no difference in ethnicity between those excluded or included in the final analysis (P = .45). Ethical approval was granted by the Victoria University of Wellington Human Ethics Committee.

Measures

Demographics.

Participants self-reported ethnicity data and could identify as multiple ethnicity groups. Ethnicity data were distilled to several binary variables representing the largest ethnicity groups in our sample (variable included whether they identified as that ethnic group or not), including New Zealand European, Māori, Pacific, Asian, and Other European. Participants also reported on their sex at birth, gender identity, age, whether or not they were born in New Zealand and if they had been given a mental health diagnosis before.

Childhood Trauma.

The history of childhood trauma was measured using the Adverse Childhood Experiences questionnaire.²⁷ This is a 17-item self-report questionnaire that asks about traumatic events prior to the age 18. A sum score was created ranging from 0 to 17. This questionnaire has been widely validated across cultures and generally has good reliability and validity.²⁸

Discrimination.

Discrimination was measured using one item that asked, Have you ever experienced discrimination due to your ethnic, gender or sexual identity? If yes, how often has this occurred in the past 12 months? Responses were scored as a 0 for never, a 1 for monthly, a 2 for every 2 weeks, a 3 for weekly, and a 4 for every day.

Financial Adversity.

Financial adversity was measured using a list of commonly experienced financial stressors, occurring in the last 12 months: Could not fill or collect a prescription medicine, Could not get a medical test, treatment, or follow-up that was recommended by a doctor, Ran out of food and could not afford to buy more, Could not pay a bill, such as a gas, electricity or telephone bill on time, Could not pay the mortgage or rent on time, Asked for financial help from friends or family, Could not fill my car with petrol. This was scored as a binary variable.

Psychotic-Like Experiences.

Participants completed the Prodromal Questionnaire Brief (PQ-B).²⁵ The PQ-B is a 21-item self-report questionnaire that measures subclinical PLEs (eg, hearing whispers, seeing shadows). The total score is a sum of the number of items endorsed (possible scores range from 0 to 21), and the distress score is the total number of endorsed items weighted by distress with possible scores ranging from 0 to 126 (a score of zero indicates no PLEs experienced). PLEs that were endorsed are scored on the distress scale from 1 to 5, where 5 strongly agrees the experience was distressing and 1 strongly disagrees the experience was distressing. To assess whether the PQ-B may include items that are experienced as positive for Māori, a positive experience scale was added. Scores were calculated on a scale from 1 to 5, where 5 strongly agrees the experience was positive and 1 strongly disagrees the experience was positive.

Data Analysis

Demographic differences across Māori and non-Māori were calculated using chi-square or t-test. To examine whether PQ-B scores were higher for Māori, linear regression was used to predict PQ-B total scores, distress scores, and positive scores for Māori compared to non-Māori. To examine whether systemic factors explained this relationship, these regressions were completed with and without controlling for trauma, discrimination and financial adversity. Given missing data for trauma (n = 8) and discrimination scores (n = 104), regression analyses were performed across five multiply imputed datasets with regression estimates pooled following Rubin’s standard rules²⁹ in SPSS version 28.0.1.0. All variables used across models were included. To examine whether the PQ-B was culturally appropriate for Māori, item-level analyses were undertaken to examine (1) what percentage of Māori compared to non-Māori endorsed an item, (2) of the endorsed items, what percentage were considered distressing by Māori compared to non-Māori, and (3) of the endorsed items, what percentage were considered positive experiences by Māori compared to non-Māori. Items were compared using chi-square tests.

Results

Participants included 466 people aged between 18 and 30 years (M = 24.8, SD = 3.63). This group included 58.6% of participants identifying they were assigned female at birth (n = 273), 40.6% identifying they were assigned male at birth (n = 189), 0.04% identifying they were assigned intersex at birth (n = 2), and 0.04% missing data (n = 2). In terms of gender, 81.8% identified as cis gender (n = 381), and 15.5% reported they were gender diverse (n = 72), and 2.8% were missing data (n = 13). The sample was 82.0% New Zealand born (n = 382), with 80.0% identifying as New Zealand European (n = 373), 18.7% identifying as Māori (n = 87), 4.5% identifying as Pacific Islanders (n = 21), 6.7% identifying as Asian (n=31), 21.9% identifying as other European ethnicities (n = 102), and 1.7% identifying as other ethnic groups (n = 8). Of note, 28.8% of participants identified as two ethnic groups (n = 134), and 3.0% identified as more than two (n = 14). The ethnicity proportions of the sample appear comparable to recent national statistics in New Zealand.¹ Demographic differences between Māori and non-Māori participants are displayed in table 1. Participants reported high rates of past mental health diagnoses, with Māori participants reporting higher rates than non-Māori. While slightly more females participated overall, Māori did not differ from non-Māori by sex, gender, or age. Of the total sample, 78.5% endorsed at least one PLE (n = 449), 70.7% endorsed at least one traumatic event (n = 386), and 43.6% endorsed discrimination related to ethnicity, sexuality, or gender occurring at least monthly (n = 238). The PQ-B had good internal consistency for the total (α = 0.84) distress (α = 0.86) and positive (α = 0.80) scores. Differences across key variables between Māori and non-Māori are shown in table 2.

Table 1.

Demographic Data for Māori and Non-Māori Participants

	Māori	Non-Māori	Comparison of Groups
Sex at birth	62.1% Female 0.0% Intersex 33.0% Male	57.8% Female 0.5% Intersex 41.2% Male	χ 2 (2, N = 464) = 0.85 P = .651
Gender	12.6% Gender diverse	16.1% Gender diverse	χ 2 (1, N = 453) = 0.53 P = .467
Born in New Zealand	3.4% were born overseas	21.4% were born overseas	χ 2 (1, N = 466) = 15.38, P < .001
Age range	18–30 (M = 25.23, SD = 3.58)	18–30 (M = 24.69, SD = 3.64)	t(464) = −1.26, P = .104
Diagnoses	81.6% had a prior diagnosis	70.4% had a prior diagnosis	χ 2 (1, N = 466) = 4.42 P = .035

Table 2.

Differences Across Key Variables Māori and Non-Māori Participants

	Māori	Non-Māori	Comparison of Groups
PQ-B (total)	M = 8.67, SD = 4.43	M = 7.30, SD = 4.61	t(464) = −2.5, P = .01
PQ-B (distress)	M = 28.17, SD = 16.62	M = 24.46, SD = 17.35	t(464) = −1.78, P = .04
PQ-B (positive)	M = 18.57, SD = 10.42	M = 15.04, SD = 10.87	t(464) = −2.70, P < .01
Childhood trauma	M = 6.22, SD = 4.35	M = 3.52, SD = 3.40	t(110.27) = −5.28, P < .001
Financial adversity	77% reported at least 1	54.6% reported at least 1	χ 2 (1, n = 466) = 14.65, P < .001
Discrimination	39.1% monthly, 22.9% more than monthly	34.3% monthly, 9.2% more than monthly	χ 2 (4, n = 363) = 19.83, P < .001

Linear regression (see supplementary materials for tables) indicated Māori had significantly higher total scores on the PQ-B (B = 1.36, P = .01); on average, they reported experiencing 1.36 more psychosis symptoms than people who did not identify as Māori. At an item level, Māori were more likely to report the following items: Q1 “Do familiar surroundings sometimes seem strange, confusing, threatening or unreal to you” (χ² (1, N = 466) = 4.33, P = .04), Q11 “Have you had the sense that some person or force is around you, although you couldn’t see anyone” (χ² (1, N = 466) = 10.19, P = .001), Q14 “Have you been confused at times whether something you experienced is real or imaginary” (χ² (1, N = 466) = 8.83, P < .01), Q18 “Do you find yourself feeling mistrustful or suspicious of other people” (χ² (1, N = 466) = 6.18, P = .01) and Q21 “Do people sometimes find it hard to understand what you are saying” (χ² (1, N = 466) = 5.93, P = .02).

When a history of trauma, discrimination, and financial stress was added to the regression, the effect of being Māori on total PQ-B score was no longer significant (P > .05), and the effect of trauma (B = 0.45, P < .001), discrimination (B = 1.66, P < .001) and financial stress (B = 2.60, P < .001) on PQ-B score were all significant. People with financial stress, on average, reported 2.60 more psychosis symptoms than those without financial stress. Experiencing an additional trauma resulted in 0.45 more psychosis symptoms on average, and experiencing more discrimination resulted in 1.66 more psychosis symptoms on average. This indicates that the increased PQ-B total scores for Māori may be explained by increased rates of trauma, discrimination and financial stress.

Interestingly, Māori did not report higher overall PQ-B distress scores. This indicates that while they experienced more of the experiences, this was not associated with more distress. At the item level, of those endorsing an item, the percentage of Māori reporting an item as distressing did not differ from the percentage of non-Māori reporting the item as distressing for any of the items (P > .05) apart from item 17 “Are your thoughts sometimes so strong that you can almost hear them?” which was less likely to be rated distressing by Māori. These data are displayed in figure 1.

Fig. 1.

Percentage of endorsed items considered distressing for Māori and non-Māori.

Finally, Māori had higher positive scores on the PQ-B (B = 3.46, P < .01). On average, Māori had positive PQ-B scores 3.46 points higher than non-Māori. However, at an item level (of those endorsing an item), the percentage of Māori reporting the item was a positive experience did not differ from the percentage of non-Māori reporting the item was a positive experience for any of the items (P > .05), except for item 21 “Do people sometimes find it hard to understand what you are saying.” This experience was only ever reported to be positive by Māori (figure 2). Regression findings were consistent, controlling for mental health diagnosis, and other ethnicity groups.

Fig. 2.

Percentage of endorsed items considered positive for Māori and non-Māori.

Discussion

Māori reported experiencing more PLEs than non-Māori. However, this relationship was eliminated when systemic factors such as trauma, discrimination, and financial stress were controlled for, suggesting systemic factors may contribute to increased psychosis risk symptoms for Māori. Interestingly, despite the higher numbers of experiences reported, Māori did not experience more distress resulting from PLEs than non-Māori. Of those endorsing PLEs, Māori were not more or less likely to report an item was distressing when compared to non-Māori. Māori were more likely to endorse items as positive on the whole, but on an item level, there was only one item they endorsed significantly more frequently as positive than non-Māori (“Do people sometimes find it hard to understand what you are saying”) which was only reported by Māori. This finding suggests patterns of increased psychosis symptoms for Māori may be nuanced, and while likely associated with systemic factors, increased symptoms may not always reflect increased distress.

Our findings align with other research suggesting Māori are over represented in psychosis risk groups⁴ and are diagnosed more frequently with psychotic disorders.^2,3 However, our data suggest that there is nuance to rates of early psychosis symptoms for Māori. While Māori reported more overall PLEs in our survey, this was not associated with more distress resulting from PLEs compared to non-Māori. Given distress is a key element of a psychotic disorder, alongside functioning impairments, this is an important consideration. This may reflect that Māori hold important protective factors that make them more resilient to PLEs. Examples of possible protective factors could include interpreting symptoms as a gift or spiritual interpretations.²¹ Supporting this idea, Māori had higher overall positive scores for psychosis experiences, indicating these were positive experiences for Māori more often than for non-Māori. Future longitudinal research will need to examine whether positive PLEs predict psychosis risk in the same way as distressing PLEs.

Supporting international data that systemic factors may explain psychosis disparities amongst culturally diverse groups,¹¹ our data suggest that increased PLEs for Māori may result from experiencing increased rates of trauma, discrimination, and financial stress linked to the history of colonization in New Zealand. Psychosis is thought to arise from a combination of biological factors such as genetics and changes in brain functioning that interact with environmental stress,¹⁵ so it makes sense that increased rates of systemic stressors would result in increased psychosis experiences based on these models.¹⁶ This evidence is consistent with the idea that biological changes associated with stress can result in increased psychosis risk.

With respect to the second hypothesis, that Māori may report increased rates of psychosis symptoms and be diagnosed more frequently with psychotic disorders due to diagnostic biases and culturally inappropriate models and measures, our findings were mixed. Māori were more likely to endorse certain items but were not more likely to be distressed by the experiences, raising the possibility that measures of PLEs including such items may not accurately reflect risk—and rather reflect that some of these experiences are normative for Māori. When examining the items that occur more frequently for Māori, many of these may reflect experiences Māori have more so than other groups. For example, several could be related to experiences of racism, such as “Do familiar surroundings seem strange…,” “Do you find yourself feeling mistrustful … of people,” and “Do people sometimes find it hard to understand what you are saying.” This would support the hypothesis that PLE measures are picking up normative experiences for Māori. Further, Māori were more likely overall to experience PLEs as positive, again suggesting this group of experiences may not necessarily reflect the risk for psychosis accurately. However, at an item level, Māori were not more or less likely to report particular experiences were distressing or positive. This indicates that for those endorsing an experience, non-Māori and Māori reported no significant differences in emotional response to that particular experience. This suggests that, for the most part, items are not significantly less distressing for Māori, or significantly more positive, which does not support this hypothesis. In addition, given increased trauma, discrimination, and financial stress were associated with the increased rates PLEs—this suggests while certain PLEs may be more normative or experienced as positive for Māori, these PLEs still may be triggered by increased stress in the same manner as other PLEs.

These findings are vital in shedding light on factors that might contribute to increased rates of psychosis symptoms for Māori. In particular, our findings suggest that managing systemic issues such as discrimination or intergenerational trauma at a community level is likely to reduce psychosis and psychosis-related distress for Māori. While results were mixed as to whether Māori are PLEs in the same way as non-Māori, there were some data suggesting that Māori may experience more PLE-type experiences but not more distress. This may reflect that Māori cultural factors reduce distress surrounding psychotic experiences, supporting the importance of culturally appropriate care that acknowledges and champions their cultural worldviews and values. Further research is needed to better understand how this resilience might impact ongoing risk and be promoted within therapeutic settings to improve outcomes for Māori. Given certain items were endorsed by Māori more frequently, and may reflect normative experiences, it may be helpful to reconsider measures of risk for psychosis for Māori, and develop tools with such items removed.

It is worth noting the limitations of this study. First, the data are cross-sectional. As such causal relationships could not be determined, nor could formal mediation analyses occur. Future longitudinal studies will be able to expand on these findings and explore temporal occurrences of variables alongside using more in-depth measurements of factors such as trauma. Second, the overall endorsement of PLEs was unusually high, similar to other recent international findings that PLEs when self-reported in community samples are endorsed at high rates.³⁰ This may reflect that the PQ-B operates uniquely in community samples, or for those who are completing the questionnaire in self-report format report more of the experiences. Further research will need to better understand what these self-reported experiences might indicate in community samples, and how reflective they are of psychosis risk. Third, a detailed statistical analysis, including measurement invariance of the PQ-B was not undertaken due to sample size limitations, and as such, we were not able to examine how the items perform across ethnic groups. With a larger sample, detailed examinations of measurement invariance across ethnic groups could be completed to better understand the application of this tool across cultures in New Zealand. Finally, the sample was mostly representative of the New Zealand population; however, there were high rates of mental health problems and a higher than typical group of gender-diverse participants, indicating some self-selection bias with those more likely to experience mental health problems completing the study. Participants were likely excluded if they did not have access to the internet; however, this is uncommon in this age group in New Zealand, with about 1% of 15-year olds without access to the internet.³¹ Rates of those without internet are consistent between Māori and New Zealand Europeans³¹ and so this is unlikely to affect the results. In conclusion, psychosis risk for Māori is nuanced, and likely reflects increased rates of systemic discrimination, trauma, and financial stress. Conceptualizations of psychosis risk in Māori need to be better understood in order to accurately measure and target those needing support, as our current measures may pathologize normal experiences.