Introduction

Individuals’ capacity to identify and adhere to health-related information, and to then exercise appropriate health behaviours in response to this knowledge, is engrained in their health literacy. Health literacy can be defined in terms of individuals’ understanding of the information needed to promote good health through their behaviours and health-related choices (Nutbeam, 2008). It also reflects individuals’ knowledge and understanding of appropriate health services, behaviours and treatment adherence (Jorm, 2000, 2012; Jorm et al., 1997; Kutcher et al., 2016; Nutbeam, 2008).

Mental Health Literacy

More specifically, mental health literacy refers to the knowledge, skills, and the ability to recognize, understand, and respond to mental health problems (Jorm et al., 1997). Mental health literacy focuses on building knowledge by which individuals recognise pathology, and to understand the causes and the underlying risk factors for poor mental health outcomes. Individuals with good mental health literacy are better equipped to recognize when they or others may be experiencing a mental health problem (e.g. depression, anxiety, or substance use disorder). Those with good mental health literacy also know how to access appropriate resources and support, including talking to a healthcare professional, seeking counselling or therapy, or joining a support group (Jorm, 2000, 2012; Jorm et al., 1997; Kutcher et al., 2016).

Mental health literacy improves mental health-related behaviour in several ways. Increased knowledge about pathology enables individuals to recognise symptoms of psychopathology and respond appropriately (Cotton et al., 2006; Rickwood et al., 2007). Importantly, increased literacy improves knowledge that enables individuals to seek information about mental health, understand risk factors. And by promoting positive attitudes to mental health, literacy can help to reduce stigma about recognising pathology within themselves, but also stigma associated with seeking support. Literacy also increases the likelihood of individuals seeking appropriate help-seeking (Burns & Birrell, 2014; Jorm et al., 1997). Increased knowledge therefore contributes to better treatment outcomes and reduces the likelihood of more adverse outcomes including suicide or self-harm. Improving mental health literacy is a key factor in the endeavour to improve mental health outcomes in the population, and substantial investment in public mental health campaigns has resulted in substantive increases in mental health literacy (Jorm et al., 2006; Reavley & Jorm, 2012).

Wellbeing Literacy as an Extension of Jorm’s Mental Health Literacy Framework

Whilst mental health literacy is clearly defined, the concept of wellbeing literacy is new and not as well articulated. One of the first concerted attempts at proposing a wellbeing literacy framework was described by Oades et al. (2020) and Hou et al. (2021). The framework emphasises individuals’ use of language about their own wellbeing. In contrast to mental health literacy, this concept of wellbeing literacy is defined in terms of the vocabulary, knowledge, and language individuals use to describe their wellbeing, and the wellbeing of others (Hou et al., 2021; Oades et al., 2020). An alternative wellbeing literacy framework was proposed by Chng et al. (2022) who defined wellbeing literacy in more similar ways to the concept of mental health literacy. As an initial exploration, community members’ knowledge of pre-defined wellbeing statements drawn from an extant wellbeing literature that emphasises personal and interpersonal dimensions encompassing hedonic, eudaimonic and social wellbeing (Huppert et al., 2009; Keyes, 2002), was tested by examining their capacity to discriminate statements of wellbeing from statements of psychopathology in a two-option (mental health vs. wellbeing) fixed-choice experiment (Chng et al., 2022). Chng et al’s (2022) analysis of 705 Australian adults identified that community members can discriminate between statements of wellbeing and mental health. However, when presented a second discrimination task which included a third option to allow respondents to define statements as reflecting both mental health and wellbeing, respondents overwhelming selected the 3rd option; statements of both wellbeing and mental health reflected the same underlying psychological phenomenon, namely psychological health.

The results of Chng et al. (2022) led Burns et al. (2022b) to propose the Total Psychological Health Framework which argues that community members define psychological health in terms of both absence of pathology and the presence wellbeing, and that community interpretations of what it means to flourish is reflected by both an absence of pathology and presence of wellbeing. These findings were similar to another study, comprising school students, and their parents and educators, and confirmed that community members do discriminate between psychological wellbeing and illbeing (Hou et al., 2021). Consequently, we propose that whilst mental health literacy reflects individuals’ knowledge of pathology, and understanding of the risk and protective factors for illness and potential treatment strategies, wellbeing literacy focuses on individuals’ understanding of what gives life meaning and pleasure, and similarly to mental health literacy, an understanding of the barriers and the factors that promote wellbeing, and the behavioural responses needed to improve individuals’ sense of wellbeing. And both wellbeing and mental health literacy are necessary conditions to enable individuals to flourish.

The Current Study

We argue that there is a need to confirm the Chng et al. (2022) study into the wellbeing literacy of Australian adults. Chng et al’s (2022) study comprised a sample of Facebook users, and it is possible that the results of that study could be biased to social media users. Indeed, the sample was predominantly female (≈81%) and also reported slightly higher levels of distress (≈26%) in comparison with the Australian population (≈20%), and a significant proportion were not in employment (≈39%). There is a need to replicate these initial findings in more diverse populations and our primary aim therefore is to replicate Chng et al. (2022) by undertaking a cross-national study of community members’ capacity to identify and discriminate the statements from wellbeing and mental health domains. A limitation of Chng et al. (2022) was that the survey comprised only Australian adults; here we seek to replicate their method in another Australian community sample, and also complement with samples drawn from other nations (i.e. United Kingdom, Singapore, Malaysia, and South Africa,) which vary in terms of being High-Medium–Low Income nations, and various levels of socio-political development, but comprise populations with population-level sufficiency in English to complete the survey.. There is only limited research about the perceptions of wellbeing between cultures (e.g., Wilson Fadiji et al. (2019)) and a systematic comparison between nations is needed. Also, the Chng et al. (2022) sample was skewed with 81% of the sample being female and we propose to utilize age and sex-quota based sampling to ensure the sample captures the broader community on these key demographics. Following Chng et al. (2022), we expect that participants will generally be able to correctly discriminate between wellbeing and mental health statements. However, as the cross-national comparisons is exploratory, we can make no pre-defined hypotheses about potential cross-national differences in mental health and wellbeing literacy. Our second aim will be to examine a range of socio-demographic (e.g. sex, gender, age) and mental health characteristics (psychological distress, wellbeing, level of contact with mental health) that may be related to individuals’ literacy.

Consequently, we provide 3 research objectives for this study that guide our analytical approach:

  1. 1)

    identify the frequency of correct endorsement of mental health and wellbeing statements in the whole sample,

  2. 2)

    identify differences between countries in the correct endorsement of mental health and wellbeing statements, and

  3. 3)

    assess the socio-demographic characteristics associated with wellbeing and mental health literacy.

Methods

Participants

Participants were recruited via Qualtrics’ online sampling panels. Eligibility for participation were that participants were aged 18 years and over, and resided either in Australia, the United Kingdom, Singapore, South Africa or Malaysia. Owing to age and sex-related differences in mental health and wellbeing, we employed quota sampling to derive approximately evenly distributed sex and age-groups (18 to 39 years, 40 to 59 years, and 60 years and over) within nations. Based on Chng et al. (2022) we estimated the sample size required to detect a difference of > 10% between two groups for proportions of 50% and less since items were generally reported at 50% and below in Chng et al. All estimated group sizes were N < 200 and the maximum total sample size for the largest delta was 170. Therefore, we chose to recruit approximately 216 respondents per country with a quota sampling to derive nation samples with similar age-sex profiles (n = 36 in each age*sex cell). Owing to non-completed responses our final sample size N = 1,040 with N > 200 for each country (See Table 1).

Table 1 Socio-demographic, psychological health and literacy characteristics by Country and Condition
Full size table

Measures

Mental Health and Wellbeing Statements

To examine mental health and wellbeing literacy, we followed the methodology described by Chng et al. (2022). Participants were randomly assigned one of two conditions of the survey reflecting, 1) a positive valence condition, and 2) a negative valence condition. In both conditions, a list of 21 statements of wellbeing and mental health were presented to participants. The statements were derived from the European Social Survey Wellbeing module, and the Diagnostic Statistical Manual (DSM) 5 diagnoses for Major Depressive Disorder and Generalized Anxiety Disorder (American Psychiatric Association, 2013).

Participants in both conditions completed two tasks. The first task asked participants to categorise the statements from the wellbeing and mental health scales into one of two columns, reflecting either mental health OR wellbeing statements. In the second task, the same statements were presented, but participants were now asked to categorise them into one of three columns reflecting mental health OR wellbeing or BOTH mental health and wellbeing.

Because there is a negative valence to the statements drawn from the mental health statements and a positive valence to the wellbeing statements, the statements were rephrased to be consistent in each of the different conditions. In the positive condition, the mental health statements were phrased in a positive valence and in terms of an absence of pathology. For example, “experiencing feelings of worthlessness” was reworded as “not experiencing feelings of worthlessness”, i.e. the absence of pathology. In the negative condition, the wellbeing statements were stated in a negative valence, for example “feeling close to community and people in the local area” became “not feeling close to community and people in the local area”, i.e. the absence of social wellbeing. For participants in the positive condition, the sorting columns were labelled “positive mental health” or “positive wellbeing”, and for participants in the negative condition, the columns were labelled “poor mental ill-health” or “poor wellbeing”.

Covariates

We collected multiple social demographic and health characteristics to potentially identify predictors of wellbeing and mental health literacy. These are listed below:

Socio-Demographics

Self-reported sex, chronological age in years, age-group, partner status, employment status, highest education, self-rated health and ethnicity. Age-group was coded as 18 to 39 years, 40 to 59 years, and 60 + years. Partner status was coded as currently partnered versus not partnered, employment status as currently employed versus not employed, and highest education was coded as no higher than secondary school versus higher than secondary school. Self-rated health was assessed with the question “In general, how would you rate your health?”. Responses were collapsed into excellent to good vs fair to poor. Participants were asked the open question “How would you define your ethnicity or ancestry background?”. Answers were coded using the broad (single digit) classifications of the Australian Bureau of Statistics standard classification of Cultural and Ethnic groups (Australian Bureau of Statistics, 2019).

Mental Health Contact and Experience Measures

Level of Contact with Mental Health

Familiarity with mental illness was measured using the 12-item Level of Contact Report (Holmes et al., 1999). The scale was adapted by Holmes et al (1999) from other stigma scales, and lists 12 situations of contact with people with mental illness, ranging in levels of intimacy (i.e. “I have observed in passing a person I believe may have had mental illness” and “I live with a person who has mental illness”). Participants check each statement that is true for them, with a higher score indicating greater exposure to persons with mental illness. Holmes et al (1999) reported good interrater reliability (0.83) based on three experts in mental illness and psychiatric rehabilitation.

Kessler Psychological Distress Scale

Non-specific symptoms of psychological distress experienced in the past 30 days was measured using the Kessler Psychological Distress Scale (K-10; Kessler et al., 2002). The 10-item questionnaire is scored on a scale from one (none of the time) to five (all of the time), with higher scores indicating higher distress levels. Cronbach’s alpha for the current study was α = 0.95.

Warwick Edinburgh Mental Wellbeing Scale

The Warwick Edinburgh Mental Wellbeing Scale (WEMWBS) was used to measure hedonic and eudaimonic aspects of wellbeing (Deary et al., 2013; Tennant et al., 2007). Participants were asked their experience of 14 statements relating to thoughts and feelings (e.g. “I’ve been feeling optimistic about the future” and “I’ve been dealing with problems well”) in the past two weeks, with scores scaled from 1 (none of the time) to 5 (al of the time). Higher scores reflect higher levels of wellbeing. Cronbach’s alpha for the current study was α = 0.95.

Statistical Analysis

Differences between nations and conditions on the socio-demographic and health covariates were compared with general linear models (GLM) with Gaussian and Logit distributions for continuous and binary variables, respectively. Wald chi-square are reported and reflect the main effect for nation or conditions separately. Post-hoc multiple pairwise tests were undertaken with Bonferroni correction to compare differences between all nations. To assess our first research objective, the frequency and proportions in which respondents classified statements in the two discrimination tasks was tabulated and plotted. To assess our second objective, differences between-nations were tested with a GLM and an overall Wald chi-square test of difference was reported. A sensitivity analysis also examined differences between conditions to ensure results of the main analyses were consistent between conditions. To assess our third objective, separate total mental health and wellbeing literacy scores for Task 1 were computed for each participant and reflected a summation of the number of statements correctly endorsed as mental health or wellbeing. These associations were tested with a GLM which was estimated with a Poisson distribution and reports the Incidence Rate Ratio, to reflect the increase or decline in number of statements correctly reported. For all analyses, we were sensitive to the sample size, and so, to minimise Type 1 error, we focus on differences that are reported with p < 0.01. Even then, we balance statistical probability with a consideration of the magnitude of the difference between groups.

Results

Table 1 provides socio-demographic descriptive statistics and scores on the Level of Contact, Psychological Distress (K-10) and Wellbeing (WEMWBS) scales by country and condition. There were no between-country differences for age (χ2 = 4.01; p = 0.405) or sex (χ2 = 0.22, p = 0.995). There were between-country differences on the K-10 (χ2 = 19.63; p = 0.001), WEMWBS (χ2 = 33.46; p < 0.001), level of contact (χ2 = 64.71; p < 0.001), partner status (χ2 = 10.83, = p = 0.029), education (χ2 = 33.18, p < 0.001), employment (χ2 = 34.99, p < 0.001), and self-rated health (χ2 = 46.25, p < 0.001).

Pairwise comparison with Bonferroni correction identified where nation differences existed. Malaysian respondents reported lower K-10 scores than Australian (t = 3.14, p = 0.017) and United Kingdom (t = 3.93, p = 0.001) respondents although noting that these differences were small reflecting only 2–3 points difference on the total K-10. Malaysian and South African respondents reported higher wellbeing than Australian (Malaysian: t = 3.49, p = 0.005; South African: t = 3.66, p = 0.003) and United Kingdom (Malaysian: t = 4.32, p < 0.001; South African: t = 4.49, p < 0.001) respondents. As with the K-10, these differences reflected only between 2 to 6 points on the WEMWB scale. Singaporeans and Malaysians reported lower level of contact than Australian (Malaysian: t = 3.66, p = 0.003; Singaporean: t = 3.82, p = 0.001) and United Kingdom (Malaysian: t = 6.33, p < 0.001; Singaporean: t = 6.49, p < 0.001) respondents. Malaysians also reported lower level of contact than South Africa (t = 4.41, p < 0.001) respondents. Singaporeans and Malaysians reported higher rates of education in comparison with Australian (Singaporean: t = 3.97, p = 0.001; Malaysian: t = 3.70, p = 0.002) and United Kingdom (Singaporean: t = 4.52, p < 0.001; Malaysian: t = 4.25, p < 0.001) respondents. Education rates ranged from 55 to 82%. Singaporeans reported higher employment rates than Australian (t = 5.99, p < 0.001), United Kingdom (t = 3.81, p = 0.001), South African (t = 4.23, p < 0.001) and Malaysian (t = 3.28, p = 0.011) respondents. Employment rates varied from 54 to 87%. South African and Malaysian respondents reported higher proportions with excellent-good self-rated health in comparison with Australian (South Africa: t = 5.46, p < 0.001; Malaysian; t = 5.15, p < 0.001), United Kingdom (South Africa: t = 4.81, p < 0.001; Malaysian; t = 4.49, p < 0.001), and Singaporean (South Africa: t = 3.80, p = 0.002; Malaysian; t = 3.48, p = 0.005) respondents; rates varied from 68 to 93%. Differences between individual countries on rates of partner status were not reported.

There were no between conditions differences for any of the socio-demographic or health characteristics of respondents: age (χ2 = 0.00; p = 0.999), K-10 (χ2 = 0.26; p = 0.613), Wellbeing (χ2 = 2.59; p = 0.108), level of contact (χ2 = 0.06; p = 0.807), sex (χ2 = 0.02, p = 0.901), partner status (χ2 = 0.77, p = 0.380), higher education (χ2 = 2.38, p = 0.123), employment (χ2 = 0.24, p = 0.625), self-rated health (χ2 = 0.46, p = 0.497). Also, there were no differences as a function of country and condition interactions: age (χ2 = 3.26; p = 0.515), K-10 (χ2 = 2.09; p = 0.720), wellbeing (χ2 = 2.00; p = 0.735), level of contact (χ2 = 2.17; p = 0.705), sex (χ2 = 6.02, p = 0.198), partner status (χ2 = 7.48, p = 0.113), education (χ2 = 8.85, p = 0.065), employment (χ2 = 5.71, p = 0.222), self-rated health (χ2 = 3.17, p = 0.529) suggesting any between-country differences were consistent between conditions.

Frequency of Correct Endorsement of Mental Health and Wellbeing Statements

For Task 1, six of the 20 statements had moderate to good accuracy with 61–80% of respondents correctly identifying: Depressed Mood (MH), Feeling Worthless (MH), Worries & Anxieties (MH), Close to Community (WB), Being Energized (WB), and People Care (WB)). Six statements showed fair accuracy with 50–60% of respondents correctly identifying: Pleasure in Activities (MH), Irritability (MH), Ability to Concentrate (MH), Optimism (WB), Sense of Accomplishment (WB) and Learning new Things (WB)). The remaining eight statements reported poor accuracy (< 50%; See Supplementary Tables 1 and 2).

For Task 2, there was a preponderance of respondents who classified statements as reflecting both mental health and wellbeing. In terms of mental health or wellbeing, only one statement reported fair accuracy: 50–60% correctly identified Close to Community (WB)), leaving 19 of the 21 statements showing accuracy of less than 50%, with Fatigue (MH) and Muscle Tension (MH) both reporting less than 15% accuracy (See Supplementary Tables 3 and 4). The clear trend was for participants to classify statements as reflecting ‘both mental health and wellbeing’.

Country Differences in the Frequency of Correct Endorsement of Mental Health and Wellbeing Statements

Endorsement of each statement by country are reported in Table 2 for mental health, and in Table 3 for wellbeing. Overall, frequency of correct endorsement between countries was consistent for Task 1. The only between-country difference on the mental health statements was reported for ‘Fatigue’ (χ2 (4) = 13.47, p = 0.009), where UK reported the highest endorsement (34.2%) and Singapore the lowest (18.6%). Between-country differences were reported for three wellbeing statements: ‘Being Energized (χ2 (4) = 14.08, p = 0.007; endorsement ranged from 68.7% (Australia) to 80.9% (Singapore)); ‘Positive about Self’ (χ2 (4) = 15.38, p = 0.004, endorsement ranged from 37.7% (Malaysia) to 51.2% (UK)); and ‘Learning new Things’ (χ2 (4) = 13.46, p = 0.009, endorsement ranged from 51.7% (UK) to 63.6% (South Africa)).

Table 2 Proportion of endorsement of Mental Health statements in Discrimination Task 1 by Country
Full size table
Table 3 Proportion of endorsement of Wellbeing statements in Discrimination Task 1 by Country
Full size table

There were between-country differences reported for Task 2 as well (see Table 4 (mental health statements) and Table 5 (wellbeing statements)), though these differences should not be over-emphasised given the overwhelming preponderance of respondents to define statements as reflecting both wellbeing and mental health in task 2, and the low rate to which respondents identified statements separately as mental health or wellbeing. Differences were reported for the mental health statements ‘Pleasure in Activities’(χ2 (4) = 20.91, p = 0.007, endorsement ranged from 18.9% (Malaysia) to 25.4% (UK)), and ‘Fatigue’ (χ2 (4) = 24.60, p = 0.002, endorsement ranged from 11.2% (UK) to 17.5% (Malaysia)). It is important to emphasise that these differences are not of a large magnitude. More substantive differences in the wellbeing statements were reported and included ‘Close to Community’ (χ2 (4) = 23.28, p = 0.003, endorsement ranged from 38.0% (Australia) to 57.6% (Malaysia)), ‘Being Energized’ (χ2 (4) = 40.04, p < 0.001, endorsement ranged from 37.0% (Australia) to 57.4% (South Africa)), and ‘Positive about Self’ (χ2 (4) = 24.23, p = 0.002, endorsement ranged from 19.7% (Australia) to 36.8% (South Africa)).

Table 4 Proportion of endorsement of Mental Health statements in Discrimination Task 2 by Country
Full size table
Table 5 Proportion of endorsement of Wellbeing statements in Discrimination Task 2 by Country
Full size table

A Sensitivity Analysis of the Effect of Conditions in Moderating the Frequency of Correct Endorsement of Mental Health and Wellbeing Statements

A potential methodological artifact of these results could be due to the way mental health and wellbeing statements were provided. To address for the negative and positive valence of the original mental health and wellbeing statements respectively, statements were rephrased to be consistent in both a positive (absence of pathology and presence of wellbeing) and negative (presence of pathology and absence of wellbeing) condition. There was only 1 between-condition difference in endorsement of statements in Task 1 (See Figs. 1 and 2) with those in the negative condition reporting lower endorsement of ‘Depressed Mood’ as a statement of mental health (endorsement ranged from 56.3% (Negative condition) to 71.6% (Positive condition)). No between-condition differences were identified for the wellbeing statements in Task 1. For Task 2 there were no differences in the proportions to which responded classified statements between conditions (See Figs. 3 and 4). For both Discrimination Tasks 1 and 2, no significant country by condition interactions were reported (p < 0.01) for any statement (see Supplementary Tables 1 to 4).

Fig. 1
figure 1

Discrimination Task 1—Proportion of Respondents who endorsed each Mental Health Statement by Condition

Full size image
Fig. 2
figure 2

Discrimination Task 1—Proportion of Respondents who endorsed each Wellbeing Statement by Condition

Full size image
Fig. 3
figure 3

Discrimination Task 2—Proportion of Respondents who endorsed each Mental Health Statement by Condition

Full size image
Fig. 4
figure 4

Discrimination Task 2—Proportion of Respondents who endorsed each Wellbeing Statement by Condition

Full size image

Characteristics Associated with Mental Health and Wellbeing Literacy

The median number of correctly identified mental health and wellbeing statements were calculated as a measure of mental health and wellbeing literacy. Table 6 outlines these raw literacy scores, with the interquartile range (IQR) for each country. In Task 1, the median number correctly identified across all countries was 6 out of the 9 mental health statements, and between 5 and 6 out of the eleven wellbeing items. Unadjusted literacy scores were consistent between countries (mental health:—χ2 (4) = 0.83, p = 0.935; wellbeing:—χ2 (4) = 2.30, p = 0.681).

Table 6 Median number (with Inter Quartile Range: IQR) of correctly identified mental health and wellbeing statements by Country
Full size table

Literacy scores fell considerably with the inclusion of ‘both mental health and wellbeing’ option in Task 2, with the median number of correctly identified for mental health statements reduced to 2 of 9 statements for all countries, and between 3 and 4 out of the eleven wellbeing statements. Again, the literacy scores were consistent between countries in Task 2 (mental health—:χ2 (4) = 1.25, p = 0.870; wellbeing—:χ2 (4) = 4.88, p = 0.299).

To examine characteristics that may be associated with mental health and wellbeing literacy, a regression model with condition, demographics, psychological distress, wellbeing, and level of contact included was conducted on Task 1 literacy scores (see Table 7). Literacy scores between countries remained consistent after adjusting for these socio-demographic and health characteristics. None of the socio-demographic attributes, condition valence, or mental health and wellbeing predictors were significantly related to Task 1 literacy scores, suggesting literacy was consistent between key social, demographic and psychological characteristics of respondents.

Table 7 Incident Rate Ratios (IRR) with 95% Confidence Intervals (95% CI) for predictors of Total Mental Health and Wellbeing Literacy Scores, Discrimination Task 1
Full size table

To explore the potential effects of ethnicity (versus the broader “country” grouping), the regression model for Task 1 literacy scores, including socio-demographic and health characteristics, was re-run with ethnicity replacing country variables (see Supplementary Table 5). Consistency between ethnicity groups for both mental health and wellbeing literacy was again reported.

Discussion

This paper sought to extend the findings reported by Chng et al. (2022) and identify whether their findings are replicated amongst community members from different nations. Drawing on a sample of respondents from Western (Australia, United Kingdom), African (South Africa) and South-East Asian (Singapore, Malaysia) nations confirmed Chng et al. (2022) initial findings. For Discrimination Task 1, there were few between-nation differences in the extent to which respondents classified statements as reflecting wellbeing or mental health The general pattern was for respondents from different nations to categorize statements consistently. For example, the overwhelming majority of respondents reported moderate to good accuracy in terms of identifying the mental health indicators ‘Depressed Mood’, ‘Worthlessness’, ‘Worries and Anxieties’, and the wellbeing indicators ‘Close to Community’, ‘Feeling Energized’, and having ‘People who Care’, with another 6 indicators of mental health (‘Pleasure in Activities’, ‘Ability to Concentrate’) and wellbeing (‘Optimism’, ‘Sense of Accomplishment’ and ‘Learning New Things’), were accurately identified by around 50–60% of respondents. Any reported between-nation differences of any substance were still of similar magnitudes. For example, the extent to which respondents classified ‘fatigue’ as a mental health indicator ranged from 18.6% for Singaporeans to 34.2% for UK respondents and reflect a preponderance to define this item incorrectly as a wellbeing statement. The results from Discrimination Task 2 also confirm with Chng et al. (2022). Community members are overwhelmingly consistent in identifying statements as reflecting both mental health and wellbeing suggesting that for community members, wellbeing and mental health probably reflect a higher order psychological health factor and it is their totality, rather than the separation of wellbeing and mental health, that is emphasized by community members.

There are three main take-away messages to these results. First, the level of mental health and wellbeing literacy in both tasks was generally consistent between nations. This conforms with Chng et al. (2022) and also with studies of mental health literacy in the general population (Reavley & Jorm, 2011). Second, there were no notable socio-economic factors, including between-nation differences, related to respondents mental health or wellbeing literacy as tested in Discrimination Task 1. That individual level characteristics were mostly unrelated to literacy suggests that it is the context within which individuals live, probably the social and public health policies which address issues of stigma around mental health and wellbeing and encourage citizens to seek support, which drive individuals’ mental health and wellbeing literacy (Monnapula-Mazabane & Petersen, 2023; Munawar et al., 2022; Tonsing, 2018). Third, although a substantive literature base discriminates between mental health and wellbeing as two related but separate dimensions, community members overwhelmingly reported that these statements, from common measures of mental health and wellbeing, reflect both domains. Whilst the overall wellbeing and mental health literacy of community members in the forced choice Discrimination Task 1 was noted, that there was a preponderance of community members to classify 20 mental health and wellbeing statements as reflecting both domains in the second Discrimination Task conforms with Chng et al. (2022). This may suggest that community members consider mental health and wellbeing as separate dimensions but that they reflect some higher order psychological health construct.

Previously, Burns et al. (2022a) proposed the Total Psychological Health Framework to reconcile how community members discriminate between mental health and wellbeing as separate dimensions of an underlying framework of psychological health. Evidence is also found in more recent factor structure studies that have utilized bi-factor analysis approaches in which multiple indicators reflect some general common factor as well as orthogonal factors. For example, several studies of wellbeing structure have utilized bi-factor modelling techniques and generally support the existence of a ‘g’ general wellbeing factor with some evidence for domain level factors (Burns, 2020; Chen et al., 2013; de Bruin & du Plessis, 2015; Gatt et al., 2014; Hides et al., 2016; Jovanović, 2015; Longo et al., 2020). We are unaware of any conjoint analyses of wellbeing and mental health indicators within a bi-factor modelling framework, but we would conjecture that bi-factor analyses may similarly find that mental health and wellbeing indicators are captured by a general psychological health factor and domain level wellbeing/mental health factors.

In Discrimination Task 1, we noted that the statements most likely to be misclassified were those emphasising physical symptoms of mental disorders (i.e. Fatigue, Muscle Tension). This pattern was consistent between nations. This corresponds with those findings reported by Chng et al. (2022), but also is consistent with the reports of community members who designate physical health in terms of wellbeing, and not mental health (Wilson Fadiji et al., 2019). In terms of wellbeing literacy however, responses to Discrimination Task 1 were less clear. Apart from “Close to Community”, “Being Energised”, “Having People who Care”, community respondents classified items equally as either mental health or wellbeing, or predominantly as a mental health indicator. Generally, differences between community members from different nations were consistent, except for Malaysian respondents who were more likely to incorrectly classify feeling “Calm and Peaceful”, and “Positive about Self” as reflecting mental health.

Limitations and Future Directions

We note the limitations of using a between-groups (conditions) design to assess the potential impact of statement valence. The study design included two conditions in which respondents were randomly assigned statements either with a positive valence (presence of wellbeing and absence of pathology vs. absence of wellbeing and presence of pathology). However, there were very few differences in the endorsement of individual indicators between conditions. In Task 1, those in the Positive Condition endorsed the absence of “Depressed Mood” as a mental health statement at a higher rate. No differences in the endorsement of wellbeing statements by conditions was reported. Future research should consider differences between valence conditions within a within-subjects design to account for potential person-level differences in endorsement by statement valence. However, we note that Chng et al. (2022) also identified a lack of substantive differences between valence conditions. And as the current study included 5 samples derived from different populations, we would note that the lack of differences between conditions were consistent between nations and could conclude that the valence of the presentation does not adversely impact the results.

Other potential limitations should be highlighted. The samples were drawn from Qualtrics Panels and it is important to recognise the bias introduced by participants who have elected to join these types of online panels for monetary or other reward. There is, however, substantial evidence that online research panels are not inherently biased (Batterham, 2014; Thornton et al., 2016). In contrast to previous literacy studies which have utilised vignettes of individuals with mental disorder symptoms to evaluate mental health literacy (Reavley & Jorm, 2011), this study used individual indicators of mental health and wellbeing to measure literacy. Future research should consider using vignettes which comprise individuals with complex mental health and wellbeing patterns. However, it is also important to recognise the limitations of vignettes since respondents may focus only prominent features and may not understand that less commonly discussed symptoms (e.g. somatic symptoms of mental health) that were not correctly endorsed in this study and that of Chng et al. (2022) and which may be missed in vignettes.

This study has further extended a body of evidence (Burns et al., 2022a; Chng et al., 2022) that suggests that whilst community members may discriminate between mental health in terms of pathological symptoms of mental health vs. indicators of wellbeing that reflects experiences of positive affect and functional capacity, these constructs ultimately reflect an underlying overall psychological health construct since the results from the second discrimination task indicated that most indicators were categorised as reflecting both mental health and wellbeing. We note that the failure to discriminate between mental health and wellbeing contrasts with a substantive literature base which has proposed a system of positive mental health or wellbeing to complement traditional models of psychopathology. That community members endorsed most mental health and wellbeing indicators as reflecting both mental health and wellbeing clearly adds complexity to the wellbeing literature which views wellbeing and mental health as ‘distinct but related’ (Huppert & So, 2013; Keyes, 2002). As Chng et al. (2022) proposed there is clearly a need to reconile current theoretical frameworks of mental health and wellbeing. Simply, discriminating between mental health and wellbeing appears less important to community members. We therefore propose instead that both dimensions (wellbeing and mental health) be considered reflective of overall psychological health. This suggests that public health campaigns of mental health should incorporate both mental health and wellbeing domains concurrently. Indeed, there is considerable evidence for the protective role of wellbeing and flourishing against incidence psychopathology, specifically anxiety and depression (Burns et al., 2022a, 2022b, 2022c; Trompetter et al., 2017; Wood & Joseph, 2010). And in a therapeutic context, clinicians could harness the community’s perception of psychological health and tailor therapies to address components of well-being based on clients’ presentation. For example, well-being therapy (WBT; Fava, 2016) was developed as adjuvant to Cognitive Behavioral Therapy to specifically address residual symptomology by focusing on eudaimonic well-being components. Compared to CBT only, WBT lowers relapse rates (Fava et al., 2004). A failure to incorporate wellbeing literacy with mental health literacy may explain the mental health “prevention gap” (Jorm et al., 2017) and lack of improvement in the levels of community poor mental health (Burns et al., 2020; Jorm, 2014).

Conclusion

The current study presents the findings of 1,044 community members from 5 culturally diverse middle- and high-income nations. Despite their cultural differences, the responses from community members were consistent; respondents can generally delineate statements of mental health from indicators of wellbeing. But when provided the option, respondents overwhelmingly categorise statements as reflecting both mental health and wellbeing dimensions. This provides further evidence that community members ultimately view mental health and wellbeing as reflecting an over-arching psychological health framework. It may be that researchers’ attempts to clearly articulate and delineate between mental health and wellbeing may be less important. Community members instead simply focus on overall psychological health which reflects both absence of pathology and presence of wellbeing. This suggest a need for public health messages and clinical interventions to focus on both mental health and wellbeing literacy concurrently.