Impact statement
There is growing recognition of significant psychological distress among Arab populations, underscoring the need for contextualized and culturally sensitive prevention approaches that focus on subjective well-being (SWB) in Arab countries. However, the cross-country validity of well-being scales remains to date underexplored in the region. The current study is the first to explore the cross-country validity of the WHO-5 Well-Being Index among young adults in six Arab countries across the Middle East and North Africa region. Our findings demonstrated that the scale’s unidimensional structure was consistent across genders, age groups and respondents from different nations. Convergent and divergent validity were good, and reliability was excellent. Overall, these findings suggest that the Arabic WHO-5 measures the originally intended SWB construct in the specific context conditions of Arabic-speaking populations. By establishing the cross-country validity of the Arabic WHO-5, this study supports its broader application in epidemiological research to explore SWB among Arabic-speaking young adults across diverse geographic areas.
Introduction
A state of good mental health is not limited to the absence of mental illnesses, but is also described as a state of well-being in all bodily, psychological and social domains (Diener et al., Reference Diener, Scollon, Lucas and Diener2009). The concept of subjective well-being (SWB) encompasses both negative (e.g., depression and anxiety) and positive aspects (e.g., happiness, satisfaction and contentment) (McDowell, Reference McDowell2010; Barden et al., Reference Barden, Conley and Young2015). The SWB construct is complex, as it concerns the cognitive, behavioral, emotional, social and personal spheres of human experience, and their optimal functioning (Keyes, Reference Keyes2002; Huppert and Ruggeri, Reference Huppert, Ruggeri, Bhugra, Bhui, Wong and Gilman2018). SWB may have various connotations for different populations and cultures (World Health Organization, 2017). It has universally and consistently been proven to be a key outcome and predictor of several major life domains, and to contribute to both physical and mental health (Kansky, Reference Kansky2017). SWB has been found to be closely related to a range of important life domains, including positive development, successful learning (Diener et al., Reference Diener, Pressman, Hunter and Delgadillo-Chase2017), high-quality social relationships, better academic/work performance, less mental distress, and increased resilience in the face of stressors (Kansky, Reference Kansky2017). Given the well-established impact of SWB on health, several researchers have called for its inclusion as an outcome measure of mental health programs (Thornicroft and Slade, Reference Thornicroft and Slade2014). Therefore, several countries have already included SWB as a routine assessment to inform government decisions and public policy (Dolan et al., Reference Dolan, Layard and Metcalfe2011; Helliwell et al., Reference Helliwell, Layard, Sachs and Neve2021). In recent years, particular emphasis has been placed on collecting self-rated SWB data in clinical settings (Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015), in the general population (De Kock et al., Reference De Kock, Latham, Leslie, Grindle, Munoz, Ellis, Polson and O’Malley2021), and in research (Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015; Lara-Cabrera et al., Reference Lara-Cabrera, Bjørkly, De las, Pedersen and Mundal2020) in an attempt to deepen understanding of the SWB concept and its applications.
One of the well-known, free-to-use and most widely used scales for assessing SWB is the WHO 5-item (WHO-5) Well-Being Index (World Health Organization, 1998; Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015). The WHO-5 Well-Being Index allows for a simple, brief self-report evaluation of the SWB construct over a 2-week period. It contains five positively worded items scored on a six-point scale. All items focus on positive health statements (Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015) and measure a global hedonic dimension of SWB (Bech, Reference Bech2012). The WHO-5 Well-Being Index has demonstrated good psychometric qualities in a unidimensional structure, with high internal consistency and high convergent associations with other well-being measures (e.g., Bech et al., Reference Bech, Olsen, Kjoller and Rasmussen2003). Since its development, the WHO-5 Well-Being Index has gained global popularity and has been translated into more than 30 languages (World Health Organization, 2024), predominantly in high-income Western and Asia-Pacific settings. The different linguistic versions of the WHO-5 Well-Being Index include Icelandic (Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014), Swedish (Löve et al., Reference Löve, Andersson, Moore and Hensing2014, Spanish (Bonnín et al., Reference Bonnín, Yatham, Michalak, Martínez-Arán, Dhanoa, Torres, Santos-Pascual, Valls, Carvalho, Sánchez-Moreno, Valentí, Grande, Hidalgo-Mazzei, Vieta and Reinares2018), Polish (Cichoń et al., Reference Cichoń, Kiejna, Kokoszka, Gondek, Rajba, Lloyd and Sartorius2020), Italian (Nicolucci et al., Reference Nicolucci, Giorgino, Cucinotta, Zoppini, Muggeo, Squatrito, Corsi, Lostia, Pappalardo, Benaduce, Girelli, Galeone, Maldonato, Perriello, Pata, Marra and Coronel2004), Romanian (Preoteasa and Preoteasa, Reference Preoteasa and Preoteasa2015), Danish (Schougaard et al., Reference Schougaard, de Thurah, Bech, Hjollund and Christiansen2018), Sinhala (Perera et al., Reference Perera, Jayasuriya, Caldera and Wickremasinghe2020), Brazilian Portuguese (de Souza and Hidalgo, Reference De Souza and Hidalgo2012), Farsi (Dadfar et al., Reference Dadfar, Momeni Safarabad, Asgharnejad Farid, Nemati Shirzy and Ghazie Pour Abarghouie2018), Turkish (Eser et al., Reference Eser, Çevik, Baydur, Güneş, Esgin, Öztekin Ç, Eker, Gümüşsoy, Eser and Özyurt2019), Malay (Suhaimi et al., Reference Suhaimi, Makki, Tan, Silim and Ibrahim2022), Thai (Saipanish et al., Reference Saipanish, Lotrakul and Sumrithe2009), Taiwanese (Lin et al., Reference Lin, Lee, Wu, Huang, Sun and Tsen2013), Bangla (Faruk et al., Reference Faruk, Alam, Chowdhury and Soron2021), Japanese (Awata et al., Reference Awata, Bech, Yoshida, Hirai, Suzuki, Yamashita, Ohara, Hinokio, Matsuoka and Oka2007), Korean (Moon et al., Reference Moon, Kim and Kim2014), Chinese (Fung et al., Reference Fung, Kong, Liu, Huang, Xiong, Jiang, Zhu, Chen, Sun, Zhao and Yu2022), and Swahili Kenyan (Chongwo et al., Reference Chongwo, Ssewanyana, Nasambu, Mwangala, Mwangi, Nyongesa, Newton and Abubakar2018). All these versions confirmed the robustness of the WHO-5 Well-Being Index and its utility in different research settings and across different geographical contexts (Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015). Over the years, the WHO-5 Well-Being Index has been increasingly and largely adopted for epidemiological research in various fields, including pediatrics (Allgaier et al., Reference Allgaier, Pietsch, Frühe, Prast, Sigl-Glöckner and Schulte-Körne2012), adolescentology (Rose et al., Reference Rose, Joe, Williams, Harris, Betz and Stewart-Brown2017), geriatrics (Allgaier et al., Reference Allgaier, Kramer, Saravo, Mergl, Fejtkova and Hegerl2013), occupational psychology (Sischka et al., Reference Sischka, Schmidt and Steffgen2018), and coronavirus disease 2019 (COVID-19)-related research (Lara-Cabrera et al., Reference Lara-Cabrera, Betancort, Muñoz-Rubilar, Rodríguez-Novo, Bjerkeset and De Las Cuevas2022). Furthermore, numerous studies have indicated that the WHO-5 Well-Being Index is suitable as a measure to screen for depression (Allgaier et al., Reference Allgaier, Kramer, Saravo, Mergl, Fejtkova and Hegerl2013; Omani-Samani et al., Reference Omani-Samani, Maroufizadeh, Almasi-Hashiani, Sepidarkish and Amini2019) and to monitor treatment response (Newnham et al., Reference Newnham, Hooke and Page2010a, Reference Newnham, Hooke and Page2010b).
We found three previous validations of the WHO-5 Well-Being Index in the Arabic language. The first one was performed in Lebanon among a relatively small sample (N = 121) and a gender-disproportionate group (75.2% females) composed of both community-dwelling and outpatient older individuals (Sibai et al., Reference Sibai, Chaaya, Tohme, Mahfoud and Al-Amin2009). Results indicated that the Arabic WHO-5 Well-Being Index had satisfactory external and internal validity in detecting depression among Lebanese older adults (Sibai et al., Reference Sibai, Chaaya, Tohme, Mahfoud and Al-Amin2009). The second validation was performed among a small sample of Saudi adults (N = 190, 59.5% females) and revealed a unidimensional latent structure of the scale, as well as high reliability and good convergent/divergent validity (Kassab Alshayea, Reference Kassab Alshayea2023. The third validation was performed in a sample of patients with schizophrenia from Lebanon, in whom the WHO-5 Well-Being Index showed a unidimensional structure, good internal consistency reliability (α = .80), cross-gender measurement invariance and good concurrent validity (Fekih-Romdhane et al., Reference Fekih-Romdhane, Al Mouzakzak, Abilmona, Dahdouh and Hallit2024).
Well-being: Arab perspectives
People from Arab countries have been struggling over the past years with a high burden of mental health problems (Ibrahim, Reference Ibrahim and Laher2021). Mental disorder rates have exceeded the expected levels in Eastern Mediterranean Arab countries, resulting in a steadily increasing burden of disease (Mokdad et al., Reference Mokdad, Charara, El Bcheraoui, Khalil, Moradi-Lakeh, Afshin, Kassebaum, Collison, Krohn and Chew2018). This burden is expected to rise due to the unstable political, economic and social climate in the Arab region (e.g., Charlson et al., Reference Charlson, Steel, Degenhardt, Chey, Silove, Marnane and Whiteford2012; Farran, Reference Farran2021), and mental health will likely pose major challenges and strains on the already fragile resources in the coming years (Charara et al., Reference Charara, Forouzanfar, Naghavi, Moradi-Lakeh, Afshin, Vos, Daoud, Wang, El Bcheraoui and Khalil2017). Despite these alarming predictions, mental health care systems in Arab countries continue to be centralized, hospital-based and mainly focused on secondary care and disease treatment, thus neglecting the crucial role that SWB may play in alleviating mental health issues and promoting adaptive psychological outcomes (Basurrah et al., Reference Basurrah, Al-Haj Baddar and Di Blasi2022). Such strategies are inappropriate and ineffective for dealing with mental health in the Arab population. Therefore, contextualized and culturally sensitive prevention approaches focused on SWB are urgently needed in Arab countries.
Recently, growing attention has been directed to the positive psychology field, and initial local research initiatives aiming at promoting SWB have begun to emerge (Basurrah et al., Reference Basurrah, Al-Haj Baddar and Di Blasi2021). However, emerging studies are in no way comparable to non-Arab research in this field, both in terms of quality and quantity (Basurrah et al., Reference Basurrah, Al-Haj Baddar and Di Blasi2021). In addition, experimental research on SWB in the Arab region is still in its infancy and suffers from major methodological flaws (Basurrah et al., Reference Basurrah, Al-Haj Baddar and Di Blasi2021). We found only limited information available on SWB among Arab people, and very few studies using the WHO-5 Well-Being Index while focusing on specific populations (e.g., Youth in Jordan; Jamaluddine and Sieverding, Reference Jamaluddine and Sieverding2022), Saudi women (Jradi and Abouabbas, Reference Jradi and Abouabbas2017), Emirati and other Arabic-speaking adults (Elbarazi et al., Reference Elbarazi, Saddik, Grivna, Aziz, Elsori, Stip and Bendak2022) and aid workers exposed to cumulative trauma in Palestine (Veronese et al., Reference Veronese, Pepe, Massaiu, De and Robbins2017). One of the main factors that hampers advances in mental health research and access to evidence-informed care in Arab countries is the lack of valid and reliable measurement tools (Zeinoun et al., Reference Zeinoun, Akl, Maalouf and Meho2020). Providing psychometrically sound measures of the SWB construct could aid in designing and implementing evidence-informed interventions aimed at improving Arab people’s well-being.
Rationale of the present study
SWB is a culturally dependent and context-driven concept (Rice and Steele, Reference Rice and Steele2004; Tov and Diener, Reference Tov, Diener and Diener2009). There is evidence that individuals from collectivist cultures tend to exhibit lower ratings compared to those from individualistic cultures, which may result in distinct levels of functioning for the WHO-5 Well-Being Items (Brailovskaia et al., Reference Brailovskaia, Lin, Scholten, Zhu, Fu, Shao, Hu, Li, Guo, Cai, Lu and Margraf2022). Despite this evidence, the cross-country validity of well-being scales remains underexplored (Cooke et al., Reference Cooke, Melchert and Connor2016). The vast majority of previous validation and adaptation studies of the WHO-5 Well-Being Index were performed in Western countries with individualistic backgrounds (Zhang et al., Reference Zhang, Balloo, Hosein and Medland2024). In addition, the limited body of research available on the cross-country validity of the WHO-5 Well-Being Index has mainly involved Western and Asian countries. For example, Carrozzino et al. (Reference Carrozzino, Christensen, Patierno, Woźniewicz, Møller, Arendt, Zhang, Yuan, Sasaki, Nishi, Berrocal Montiel, Ceccatelli, Mansueto and Cosci2022) investigated the validity of the WHO-5 Well-Being Index in a sample of 3,762 adults from 5 European (i.e., Italy, Poland and Denmark) and non-European (i.e., China and Japan) countries. Sischka et al. (Reference Sischka, Costa, Steffgen and Schmidt2020) demonstrated that the WHO-5 Well-Being Index is psychometrically appropriate and cross-nationally applicable in different nationally representative samples of individuals (N = 43,469) across 35 European countries. Another study also found that the WHO-5 Well-Being Index showed good validity and reliability across Spain, Chile and Norway in nurses who worked during the COVID-19 pandemic (Lara-Cabrera et al., Reference Lara-Cabrera, Betancort, Muñoz-Rubilar, Rodríguez-Novo, Bjerkeset and De Las Cuevas2022). More recently, a large multinational study confirmed the unidimensional measurement structure of the WHO-5 Well-Being Index in a sample of adolescents from 43 countries (in Europe, Central Asia and North America) (Sischka et al., Reference Sischka, Martin, Residori, Hammami, Page, Schnohr and Cosma2025). The study also demonstrated configural and metric cross-country invariance, as well as appropriate patterns of correlations with life satisfaction, self-rated health, loneliness and psychosomatic complaints (Sischka et al., Reference Sischka, Martin, Residori, Hammami, Page, Schnohr and Cosma2025). Cross-country validation studies are crucial to prove that the measure covers transcultural components of the subjective well-being construct, and can be used for cross-country comparison purposes in international multicenter research.
Although people from different Arab countries share similarities (including the language, geography, collectivist identity, religion and a young age structure; Harb, Reference Harb, Amer and Awad2016), diversity also exists. Large cross-country studies have shown that the way Arab people view and behave toward mental health issues is not uniform and appears to be largely shaped by the local context of each Arab country (Fekih-Romdhane et al., Reference Fekih-Romdhane, Daher-Nashif, Stambouli, Alhuwailah, Helmy, Shuwiekh, Mohamed Lemine, Radwan, Saquib, Saquib, Fawaz, Zarrouq, Naser, Obeid, Saleh, Haider, Miloud, Badrasawi, Hamdan-Mansour, Barbato, Bakhiet, Sayem, Adawi, Grein, Loch, Cheour and Hallit2023a, Reference Fekih-Romdhane, Jahrami, Stambouli, Alhuwailah, Helmy, Shuwiekh, Lemine, Radwan, Saquib, Saquib, Fawaz, Zarrouq, Naser, Obeid, Hallit, Saleh, Haider, Daher-Nashif, Miloud, Badrasawi, Hamdan-Mansour, Barbato, Bakhiet, Sayem, Adawi, Grein, Cherif, Chalghaf, Husni, Alrasheed and Cheour2023b). Taking into consideration these cultural disparities, it is necessary to examine whether the WHO-5 Well-Being Index measures the SWB construct accurately in different Arab countries and cultural backgrounds. In this article, we aimed to contribute to the literature on SWB in different ways. First, we propose to investigate, for the first time, the cross-country validity of the WHO-5 Well-Being Index across different Arab countries to ensure its suitability for capturing and providing reliable information on the SWB construct in different Arab contexts. Second, as the two previous validations were conducted in Arab Middle East countries, we intended to expand our investigation to North African countries (i.e., Tunisia and Morocco) that have not been the subject of previous validation studies of the WHO-5 Well-Being Index. Third, we sought to examine psychometric properties that have not been previously examined, such as measurement invariance across genders. Gender differences in SWB are culturally determined, as they may be substantially affected by social norms and adherence to traditional gender roles (Matud et al., Reference Matud, López-Curbelo and Fortes2019). However, variations across genders may also be largely driven by methodological factors (Graham and Chattopadhyay, Reference Graham and Chattopadhyay2013). For this reason, we sought to verify that the WHO-5 Well-Being Index invariantly measures the SWB factor across gender groups. Fourth, we aimed to explore its concurrent validity by calculating Pearson’s correlation coefficients between the WHO-5 Well-Being Index and measures of depression, anxiety, stress, suicidal ideation and insomnia. We hypothesized that the Arabic version of the WHO-5 Well-Being Index would show a single-factor structure and satisfactory composite reliability in all samples from different countries, and would be invariant across gender groups. We also expected that the concurrent validity of the Arabic WHO-5 Well-Being Index would be supported through significant negative correlations with depression and other psychopathology measures.
Methods
Study design and participants
This was a multi-country, web-based, cross-sectional study. Several researchers from different institutions in the 22 Arab countries were invited to collaborate in our multinational project and join our team as co-investigators and co-authors. Researchers from six Arab countries accepted our invitation: Tunisia, Lebanon, Egypt, Jordan, Morocco and Kuwait. Arabic-speaking individuals from the general population, aged between 18 and 35 years and residing in an Arab country during the study period, were considered eligible to participate. This age range was chosen to guarantee homogeneous sampling and eliminate any differences resulting from age. Young adults aged 18–35 years have been found to display a worse health profile than both adolescents and those in their late 30s (Stroud et al., Reference Stroud, Walker, Davis and Irwin2015). The committee on Improving the Health, Safety and Well-Being of Young Adults (convened by the National Research Council and the Institute of Medicine) concluded in their report that young adulthood is developmentally “of critical nature” within the life course (Committee on Improving the Health, Safety, and Well-Being of Young Adults et al., Reference Bonnie, Stroud and Breiner2015). Accordingly, the committee recommended that “outcomes should be measured specifically for young adults,” and that young adults should be treated as a distinct subpopulation in programming, planning, policy and research (Committee on Improving the Health, Safety, and Well-Being of Young Adults et al., Reference Bonnie, Stroud and Breiner2015). Following these recommendations, we aimed to test the psychometric properties of the WHO-5 Well-Being Index exclusively in young adults within a relatively narrow age range. The survey was open between February and June 2022, and all responses collected during that period were included in the analysis.
All participants fulfilling these criteria were sampled using a convenience sampling technique and were invited to respond to a uniform, anonymous web-based questionnaire through social media platforms (including Instagram, Facebook and WhatsApp). Recruitment via Instagram was done using posts and stories shared by the research team and collaborators. The posts included a description of the study, the eligibility criteria and a link to the survey. No specific hashtags were used. Engagement was driven through reposts and snowball sampling. Participants were also asked to forward the link to other eligible people they might know, using the snowball technique (Parker et al., Reference Parker, Scott, Geddes, Atkinson, Delamont, Cernat, Sakshaug and Williams2019). Snowballing techniques and online recruitment of non-help-seeking participants are typically adopted for research in this area (e.g., see Preti et al., Reference Preti, Raballo, Kotzalidis, Scanu, Muratore, Gabbrielli, Tronci, Masala, Petretto and Carta2018). This recruitment approach was also chosen because several Arab countries boast high Internet penetration rates (varying from 72% in Tunisia to 100% in Kuwait) (World Bank, 2023), and some of the highest rates of social media usage in the world (Radcliffe et al., Reference Radcliffe, Abuhmaid and Mahliaire2023). Eight out of 10 Arab youth aged 18–24 years reported daily usage of messaging apps, including Facebook (72%), Instagram (61%) and YouTube (53%) (Radcliffe et al., Reference Radcliffe, Abuhmaid and Mahliaire2023). The questionnaire was administered using the free online survey tool provided by Google Forms. The study information and answering instructions were provided online via text; participants were asked to read them and give their informed consent before filling out the survey. Participants did not receive any incentives for participation. The study was performed in accordance with the Declaration of Helsinki for human research. The research protocol was approved by the Ethics Committees of the home institutions of the Principal Investigators, the Psychiatric Hospital of the Cross Ethics Committee, Jal Eddib, Lebanon (Ref: HPC-012-2022), and the Ethics Committee of the Razi Psychiatric Hospital, Manouba, Tunisia (Ref: ECRPH-2022-0019).
The total sample consisted of 3,247 participants, with a mean age of 23.36 ± 4.62 years. The majority of the participants were females (71.6%), single (75.9%) and had a university level of education (79.5%). The details of the sample by country are summarized in Supplementary Material (Supplementary Table S1).
Minimum sample size
As a rule of thumb, simulation studies show that with normally distributed indicator variables and no missing data, a reasonable sample size for a simple Confirmatory Factor Analysis (CFA) model is about N = 150 (Muthén and Muthén, Reference Muthén and Muthén2002), which was far exceeded in our sample. The resulting sample size was sufficiently large to provide adequate statistical power for all our analyses, including measurement invariance testing across groups (Meade et al., Reference Meade, Johnson and Braddy2008).
Measures
The WHO-5 Well-Being Index
This instrument was developed in 1998 and has been translated into 30 different languages. The WHO-5 consists of five items and assesses subjective psychological well-being. Each item is scored on a 5-point Likert scale with 5 = all of the time to 0 = none of the time. Therefore, the total score ranges from 0 (absence of well-being) to 25 (maximum well-being) (World Health Organization 2019). Raw scores are then multiplied by 4 to obtain a percentage score ranging from 0 (worst) to 100 (best). The Arabic version of this instrument was validated in Lebanon among elderly people (Sibai et al., Reference Sibai, Chaaya, Tohme, Mahfoud and Al-Amin2009).
Columbia-Suicide Severity Rating Scale (C-SSRS)
This scale is composed of five items, rated as a no/yes type of answer. It evaluates suicidality over the past month. Higher scores indicate higher suicidal ideation. This scale has been validated in the Arabic language among Arabic-speaking adults from Lebanon, where it showed a unidimensional factor structure, good internal consistency (α = .797) and appropriate convergent validity with measures of depression, anxiety and self-esteem (Zakhour et al., Reference Zakhour, Haddad, Sacre, Fares, Akel, Obeid, Salameh and Hallit2021). In the present sample, the C-SSRS yielded a McDonald’s ω of .79 and a Cronbach’s α of .79.
Insomnia Severity Index (ISI)
This scale is composed of seven items, rated on a 4-point Likert scale. Higher scores reflect more severe insomnia. The Arabic validated version of the ISI was used, which demonstrated good reliability (α = .833) and good validity in a sample of Arabic-speaking community-dwelling adults from Lebanon (Hallit et al., Reference Hallit, Haddad, Hallit, Al Karaki, Malaeb, Sacre, Kheir, Hajj and Salameh2019). In the present sample, the ISI yielded a McDonald’s ω of .82 and a Cronbach’s α of .82.
Depression, Anxiety, and Stress Scale 8 items (DASS-8)
The DASS-8 is composed of eight items measuring depression (three items), anxiety (three items) and stress (two items). Items are rated on a 4-point Likert scale. Higher scores reflect higher depression, anxiety and stress. The Arabic-validated DASS-8 was used in this study, which showed excellent psychometric properties in terms of internal consistency (α = .94), convergent validity, predictive validity and discriminant validity (Ali et al., Reference Ali, Alkhamees, Hallit, Al-Dwaikat, Khatatbeh and Al-Dossary2024). In this study, the DASS-8 showed good internal consistency reliability for all three dimensions: depression (ω = .91/α = .91), anxiety (ω = .90/α = .90) and stress (ω = .73/α = .73).
Demographics
Participants were asked to provide their demographic details, including age, gender and education level.
Analytic strategy
Confirmatory Factor Analysis
There were no missing responses in the dataset since all questions were required in the Google Forms. Duplicate responses were screened and removed using Excel’s “Remove Duplicates” function, based on identical patterns in response time stamps and item-level data. We used data from the total sample to conduct a CFA using the SPSS AMOS v.29 software. Our intention was to test the original model of the WHO-5 Well-being scale (i.e., one-factor model). Parameter estimates were obtained using the maximum likelihood method with corresponding fit indices. To identify the model, we used the marker variable approach (Little et al., Reference Little, Cunningham, Shahar and Widaman2002), in which the factor loading of the first item (Well-being 1) was fixed to 1 to scale the latent variable. This is a common method for setting the metric of latent constructs in CFA (Schreiber, Reference Schreiber2008, Reference Schreiber2017). Multiple indices were calculated to assess model fit: the normed model chi-square (χ 2/df), the root mean square error of approximation (RMSEA), the Tucker–Lewis Index (TLI) and the comparative fit index (CFI). Values ≤5 for χ 2/df, ≤0.08 for RMSEA and 0.90 for CFI and TLI indicate good fit of the model to the data (Hu and Bentler, Reference Hu and Bentler1999). In addition to reporting global fit indices (RMSEA, CFI, TLI and standardized root mean square residual [SRMR]), we examined local fit through standardized residual covariances and modification indices, as recommended by previous authors (Steiger, Reference Steiger2007; Kline, Reference Kline2023; Goretzko et al., Reference Goretzko, Siemund and Sterner2024). These local diagnostics help identify specific areas of model misfit that global indices may obscure. Moreover, evidence of convergent validity was assessed in this subsample using the Fornell–Larcker criterion, with average variance extracted (AVE) values of ≥0.50 considered adequate (Malhotra and Dash, Reference Malhotra and Dash2011). Multivariate normality was not verified at first (Bollen-Stine bootstrap p = .002); therefore, we performed a nonparametric bootstrapping procedure.
Measurement invariance
To examine gender, country and age (dichotomized into ≤25 vs. ≥26 years (Carlucci et al., Reference Carlucci, Watkins, Sergi, Cataldi, Saggino and Balsamo2018) invariance of WHO-5 Well-being scores, we conducted multigroup CFA (Chen, Reference Chen2007) using the total sample. Measurement invariance was assessed at the configural, metric and scalar levels (Vadenberg and Lance, Reference Vadenberg and Lance2000). We accepted ΔCFI ≤ 0.010 and ΔRMSEA ≤ 0.015 or ΔSRMR ≤ 0.010 as evidence of invariance (Chen, Reference Chen2007). Differences between genders and age groups were evaluated using the Student t-test, and differences between countries were evaluated using the analysis of variance test.
Further analyses
Reliability was assessed using McDonald’s ω and Cronbach’s α, with values >.70 reflecting adequate reliability (Malkewitz et al., Reference Malkewitz, Schwall, Meesters and Hardt2023). The WHO-5 Well-being total score was considered normally distributed since the skewness and kurtosis values fell between ±1 (Hair et al., Reference Hair, Sarstedt, Ringle and Gudergan2017). Therefore, to assess concurrent and divergent validity, we examined bivariate correlations between the WHO-5 Well-Being Index and the CSRS, ISI and DASS-8 scores using the Pearson test. Based on Cohen (Reference Cohen1992), values ≤.10 were considered weak, ~.30 as moderate and ~.50 as strong correlations.
Results
The information on the distribution (mean, SD, skewness and kurtosis) of each WHO-5 Well-Being Index item, stratified by country and gender, is summarized in Table 1.
Distribution (mean, SD, skewness and kurtosis) of each WHO-5 Well-Being Index item, stratified by country and gender
CFA of the Arabic WHO-5 Well-Being Index
Except for the RMSEA, most CFA model fit indices indicated that the fit of the one-factor model of the Arabic WHO-5 Well-Being Index was acceptable: χ 2 = 223.44, df = 5 (p < .001), RMSEA = 0.116 (90% confidence interval [CI] = 0.103, 0.129), SRMR = 0.017, CFI = 0.985, TLI = 0.970. When a correlation between the residuals of items 1 and 4 was added (after showing a high modification index), the results improved further as follows: χ 2 = 57.71, df = 4 (p < .001), RMSEA = 0.064 (90% CI = 0.050, 0.080), SRMR = 0.009, CFI = 0.996, TLI = 0.991. The standardized estimates of factor loadings (Figure 1) and the AVE values (0.77) were all excellent. The same analysis was conducted for each country and showed adequate results as well (Table 2). The results of the standardized residual covariances and modification indices can be found in Supplementary Tables S2 and S3.
Standardized factor loadings derived from the Confirmatory Factor Analysis of the Arabic WHO-5 Well-Being Index in the total sample.
Confirmatory Factor Analysis and standardized loading factors of the Arabic WHO-5’s items per country
Internal and composite reliability
Internal reliability of the WHO-5 Well-being scores was adequate in the total sample (ω = .94/α = .94), in men (ω = .95/α = .95) and in women (ω = .94/α = .94), in participants aged ≤25 years (ω = .94/α = .94) and ≥ 26 years (ω = .96/α = .96) and within each country as follows: Tunisia (ω = .96/α = .96), Lebanon (ω = .95/α = .96), Kuwait (ω = .94/α = .94), Egypt (ω = .92/α = .92), Jordan (ω = .93/α = .93) and Morocco (ω = .94/α = .94).
Measurement invariance and differences by gender, age and country
As reported in Table 3, indices suggested that configural, metric and scalar invariance were supported across gender, country and age categories. The results showed that a significantly higher mean WHO-5 Well-being score was found in males compared to females (10.30 ± 6.65 vs. 8.73 ± 6.17; t (3245) = 6.37; p < .001, Cohen’s d = 0.248). Moreover, the highest mean well-being score was found in Morocco (44.2 ± 26.84) and Kuwait (40.4 ± 27.04), followed by Lebanon (39.8 ± 25.24), Jordan (36.84 ± 24.04), Tunisia (33.6 ± 25.88) and Egypt (32.2 ± 22.72), with the difference being significant (F = 15.96, p < .001). The Bonferroni post-hoc analysis showed a significant difference between Tunisia and Lebanon (p < .001), Tunisia and Kuwait (p < .001), Tunisia and Morocco (p < .001), Lebanon and Egypt (p < .001), Kuwait and Egypt (p < .001), Egypt and Morocco (p < .001), and Jordan and Morocco (p = .027).
Measurement invariance of the Arabic WHO-5 Well-Being Index in the total sample
Note: CFI, Comparative Fit Index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual.
Finally, a higher mean well-being score was found in participants aged 26 years and above compared to those aged 25 years and below (9.61 ± 6.61 vs. 9.02 ± 6.25; t (3,245) = −2.26; p < .001, Cohen’s d = .093).
Concurrent validity (total sample)
As for concurrent validity, WHO-5 Well-being scores showed a moderate significant inverse correlation with DASS depression (r = −.28; p < .001), anxiety (r = −.29; p < .001) and stress (r = −.27; p < .001) subscales scores, suicidal ideation (r = −.16; p < .001) and insomnia severity (r = −.37; p < .001) (Table 4).
Pearson correlation matrix
*** p < .001.
Discussion
This study is the first to explore the cross-country validity of the WHO-5 Well-Being Index among young adults across six Arab countries (i.e., Tunisia, Lebanon, Egypt, Jordan, Morocco and Kuwait) in the Middle East and North Africa region. Results showed that all five items loaded onto a single latent factor in both genders and across all six countries, demonstrating adequate reliability, as well as good convergent and divergent validity. Overall, these findings suggest that the Arabic WHO-5 measures the originally intended SWB construct within the specific context conditions of Arabic-speaking populations. By verifying the cross-country validity of the Arabic WHO-5, our study supports its wider application to epidemiologically explore SWB among Arabic-speaking young adults from broad geographic areas.
We found that WHO-5 mean scores varied significantly across countries, ranging from 32.2 ± 22.72 in Egypt to 44.2 ± 26.84 in Morocco. Despite these wide variations, WHO-5 scores reported in all six Arab countries were much lower than those observed among the adult general population in other international studies (e.g., 56 in Sri Lanka [Perera et al., Reference Perera, Jayasuriya, Caldera and Wickremasinghe2020], 64.74 in Iceland [Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014], 73.37 in southern Brazil [de Souza and Hidalgo, Reference De Souza and Hidalgo2012], but were comparable to scores found in a Middle Eastern country (i.e., 35.8 in Iranian people [Dadfar et al., Reference Dadfar, Momeni Safarabad, Asgharnejad Farid, Nemati Shirzy and Ghazie Pour Abarghouie2018]). It should be noted that direct comparisons of WHO-5 Well-being mean scores between studies may not be meaningful due to differences in contextual factors, such as demographics (e.g., our sample exclusively included young adults). That said, the low mean scores observed in our present sample should serve as a warning for clinicians, researchers and policy-makers working in Arab settings, and further highlight that local culturally sensitive strategies are needed to address well-being issues among Arab young adults.
The construct validity of the WHO-5 was examined using CFA, which is consistently advocated by validation researchers as a crucial step in scale validation (Loewenthal and Lewis, Reference Loewenthal and Lewis2018; Zeng et al., Reference Zeng, Fung, Li, Hussain and Yu2020). Unlike exploratory factor analysis (EFA), CFA imposes meaningful constraints when evaluating a measure’s validity (Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014). Although the WHO-5 has been extensively validated in dozens of languages and countries, few WHO-5 Well-being assessments have used CFA (De Wit et al., Reference De Wit, Pouwer, Gemke, Delemarre-Van De Waal and Snoek2007; Fung et al., Reference Fung, Kong, Liu, Huang, Xiong, Jiang, Zhu, Chen, Sun, Zhao and Yu2022), and several validation studies have relied only on EFA (Allgaier et al., Reference Allgaier, Pietsch, Frühe, Prast, Sigl-Glöckner and Schulte-Körne2012; Awata et al., Reference Awata, Bech, Yoshida, Hirai, Suzuki, Yamashita, Ohara, Hinokio, Matsuoka and Oka2007; Bonnín et al., Reference Bonnín, Yatham, Michalak, Martínez-Arán, Dhanoa, Torres, Santos-Pascual, Valls, Carvalho, Sánchez-Moreno, Valentí, Grande, Hidalgo-Mazzei, Vieta and Reinares2018; Cichoń et al., Reference Cichoń, Kiejna, Kokoszka, Gondek, Rajba, Lloyd and Sartorius2020; Hochberg et al., Reference Hochberg, Pucheu, Kleinebreil, Halimi and Fructuoso-Voisin2012; Löve et al., Reference Löve, Andersson, Moore and Hensing2014). Analyses from the present study showed that the fit of a one-factor model to the data was acceptable in each of the six countries, thus replicating the factor structure of the original WHO-5 Well-Being Index (Bech, Reference Bech2004, Reference Bech2012; Bech et al., Reference Bech, Olsen, Kjoller and Rasmussen2003), and that obtained in other linguistic versions using CFA (e.g., Swahili Kenyan [Chongwo et al., Reference Chongwo, Ssewanyana, Nasambu, Mwangala, Mwangi, Nyongesa, Newton and Abubakar2018], Malay [Suhaimi et al., Reference Suhaimi, Makki, Tan, Silim and Ibrahim2022], Icelandic [Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014], Sinhala [Perera et al., Reference Perera, Jayasuriya, Caldera and Wickremasinghe2020], Chinese [Fung et al., Reference Fung, Kong, Liu, Huang, Xiong, Jiang, Zhu, Chen, Sun, Zhao and Yu2022] and Arabic [Kassab Alshayea, Reference Kassab Alshayea2023]). Our results support the applicability of the WHO-5 Well-Being Index as a unidimensional measure of SWB in Arab contexts. Furthermore, to assess composite reliability in our sample, McDonalds’ ω coefficients were used as they have been shown to provide more realistic estimates of a measure’s reliability than Cronbach’s α (Ravinder and Saraswathi, Reference Ravinder and Saraswathi2020). Findings revealed that the Arabic WHO-5 Well-Being Index yielded high reliability coefficients in the total sample and both genders, which is in line with previous international studies on other translations of the WHO-5 Well-Being Index that mostly relied on Cronbach’s αcoefficients (e.g., α = 0.79–0.91 in Italy [Nicolucci et al., Reference Nicolucci, Giorgino, Cucinotta, Zoppini, Muggeo, Squatrito, Corsi, Lostia, Pappalardo, Benaduce, Girelli, Galeone, Maldonato, Perriello, Pata, Marra and Coronel2004], 0.86–0.88 in Kenya [Chongwo et al., Reference Chongwo, Ssewanyana, Nasambu, Mwangala, Mwangi, Nyongesa, Newton and Abubakar2018], 0.87 in Poland [Cichoń et al., Reference Cichoń, Kiejna, Kokoszka, Gondek, Rajba, Lloyd and Sartorius2020], 0.88 in Romania [Preoteasa and Preoteasa, Reference Preoteasa and Preoteasa2015], 0.83 in Sweden [Löve et al., Reference Löve, Andersson, Moore and Hensing2014], 0.91 in Iran [Dadfar et al., Reference Dadfar, Momeni Safarabad, Asgharnejad Farid, Nemati Shirzy and Ghazie Pour Abarghouie2018], 0.81 in Turkey [Eser et al., Reference Eser, Çevik, Baydur, Güneş, Esgin, Öztekin Ç, Eker, Gümüşsoy, Eser and Özyurt2019], 0.91 in Malaysia [Suhaimi et al., Reference Suhaimi, Makki, Tan, Silim and Ibrahim2022], 0.75 in Bangladesh [Faruk et al., Reference Faruk, Alam, Chowdhury and Soron2021], 0.83 in Brazil [de Souza and Hidalgo, Reference De Souza and Hidalgo2012], 0.82–0.87 in Iceland [Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014], 0.81–0.85 in China [Fung et al., Reference Fung, Kong, Liu, Huang, Xiong, Jiang, Zhu, Chen, Sun, Zhao and Yu2022], 0.91 in Saudi Arabia [Kassab Alshayea, Reference Kassab Alshayea2023]). In the initial unidimensional CFA model, global model fit was adequate but could be improved. Based on the modification indices and in line with prior findings (Goretzko et al., Reference Goretzko, Siemund and Sterner2024), we allowed residuals of items 1 and 4 to correlate since these items reflect affective well-being and share similar emotional tone and response patterns, which may explain the residual association beyond the general well-being factor. This correlation suggests that these items may consistently share unexplained variance. This modification improved model fit substantially, while remaining theoretically justified.
For Morocco, the fit indices showed unusually high values (TLI = 1.004, CFI = 1.000, RMSEA CI < 0.001–0.085), suggesting potential overfitting, model saturation, limited variance or a small sample size; these conditions can inflate fit indices and underestimate model misfit. In this study, this can be explained by the small sample from Morocco (n = 202), which is supported in the literature (Xia and Yang, Reference Xia and Yang2019). For Tunisia and Jordan, the RMSEA values were beyond the acceptable limits, indicating poor fit, although CFI and TLI suggested otherwise. This should be interpreted with caution, as RMSEA is sensitive to low degrees of freedom and can falsely indicate poor fit in small or simple models (Kenny et al., Reference Kenny, Kaniskan and McCoach2015).
Another relevant contribution of this study was to examine the measurement invariance of the WHO-5 Well-being scores across gender and countries. Results from multigroup analyses demonstrated that configural, metric and scalar invariance were supported across gender in the total sample and by country. Evidence of invariance across gender groups has also been reported in other validation studies and different linguistic contexts (e.g., Icelandic national and patient samples [Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014] and Sri Lankan people from the general population [Perera et al., Reference Perera, Jayasuriya, Caldera and Wickremasinghe2020]). These findings imply that WHO-5 Well-being mean differences in SWB between male and female respondents, as well as between respondents from various Arab countries, are not attributable to group-level variations in understanding or responding to items, but to real differences in the construct level (Putnick and Bornstein, Reference Putnick and Bornstein2016). We, therefore, suggest that the Arabic WHO-5 Well-Being Index can be used reliably to compare mean differences between gender and country groups.
In this regard, we found that males displayed higher WHO-5 Well-being scores than females, which is in accordance with previous studies (e.g., Nicolucci et al., Reference Nicolucci, Giorgino, Cucinotta, Zoppini, Muggeo, Squatrito, Corsi, Lostia, Pappalardo, Benaduce, Girelli, Galeone, Maldonato, Perriello, Pata, Marra and Coronel2004; Lin et al., Reference Lin, Lee, Wu, Huang, Sun and Tsen2013; Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014; Löve et al., Reference Löve, Andersson, Moore and Hensing2014; Preoteasa and Preoteasa, Reference Preoteasa and Preoteasa2015). Furthermore, the highest well-being scores were exhibited by Moroccan and Kuwaiti participants compared to those from other nationalities. Comparative studies on mental health and well-being between the different Arab countries are lacking. A study published in 2012 reported that Morocco has one of the lowest numbers of psychiatrists and higher prevalence rates of mental health problems compared to other Arab countries (Okasha et al., Reference Okasha, Karam and Okasha2012). One plausible explanation for our findings could be the sociopolitical unrest that took place in some Arab countries over the past years, such as the Arab spring in Tunisia and Egypt, or the economic crises and conflicts in Lebanon (Al-ghzawi et al., Reference Al-ghzawi, ALBashtawy, Azzeghaiby and Alzoghaibi2014).
As for concurrent validity, WHO-5 Well-being scores showed a strong, significant inverse correlation with DASS-8 depression subscores, which is consistent with several previous validations using various depression measures (e.g., Cichoń et al., Reference Cichoń, Kiejna, Kokoszka, Gondek, Rajba, Lloyd and Sartorius2020; Dadfar et al., Reference Dadfar, Momeni Safarabad, Asgharnejad Farid, Nemati Shirzy and Ghazie Pour Abarghouie2018; Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014; Lucas-Carrasco, Reference Lucas-Carrasco2012; Perera et al., Reference Perera, Jayasuriya, Caldera and Wickremasinghe2020; Saipanish et al., Reference Saipanish, Lotrakul and Sumrithe2009). Similarly, the WHO-5 Well-Being Index has consistently shown high sensitivity and specificity in detecting depression, and has been extensively applied as a screening tool for this condition (Topp et al., Reference Topp, Østergaard, Søndergaard and Bech2015). Negative correlations have also been demonstrated between WHO-5 Well-being scores and different symptoms of mental health problems (anxiety, stress, suicidal ideation and insomnia) in our sample. These findings also concur with prior research comparing the correlations between the WHO-5 Well-Being Index and different measures of mental health problems (e.g., anxiety [Awata et al., Reference Awata, Bech, Yoshida, Hirai, Suzuki, Yamashita, Ohara, Hinokio, Matsuoka and Oka2007; Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014], stress [Guðmundsdóttir et al., Reference Guðmundsdóttir, Olason, Guðmundsdóttir and Sigurðsson2014; Faruk et al., Reference Faruk, Alam, Chowdhury and Soron2021] and sleep problems [Löve et al., Reference Löve, Andersson, Moore and Hensing2014]). In sum, the current results strongly support the validity of the Arabic WHO-5 Well-Being Index and offer additional confirmation that it serves the purposes for which it was originally developed.
Study limitations and research perspectives
Despite its valuable contribution to the field of SWB, the present study has some limitations that need to be addressed in future research. First, we did not use a structured clinical interview against which the results from the self-report measure could be validated, which prevented us from assessing the specificity and sensitivity of the WHO-5 Well-Being Index as a depression screening tool. To address this limitation, future studies should include an external criterion measure. Second, our data were gathered at a single point in time, which precluded us from testing the stability and invariance of the Arabic WHO-5 Well-Being Index over time. Therefore, additional validation studies are needed to examine the test–retest reliability of the scale. Additionally, because of the cross-sectional design, the relationships examined in this study cannot be interpreted causally. Third, the use of a community sample of young adults may undermine the generalizability of our findings to clinical populations. Fourth, we employed an online survey and convenience sampling, both of which mostly attracted highly educated and female participants, thereby limiting the representativeness of our sample. Fifth, the theoretically related measures used for validity purposes (C-SSRS, ISI and DASS-8) have not been validated in the different cultural contexts and countries involved in this study. Finally, the six countries involved in our study are lower-to-middle-income countries, and cannot be considered representative of all Arab populations and the Arab world. Young adults from low-income Arab countries (such as Yemen and Syria) may have different SWB levels, and should be the subject of future validation studies. Using dynamic fit index thresholds for model evaluation, reporting effect sizes for measurement non-invariance, and examining group mean differences (e.g., by gender, age or country) within a latent variable framework may also be valuable for future work.
Conclusion
In summary, findings indicate that the WHO-5 Well-Being Index in its Arabic version has a unidimensional structure among Arabic-speaking young adults across six Arab countries, high internal consistency, good concurrent validity and measurement invariance across gender. This study contributes to the growing research in the field of positive psychology and well-being by providing a brief, valid and reliable Arabic version of the WHO-5 Well-Being Index that can be used cross-nationally with a variety of Arabic-speaking young adult populations for screening and research purposes.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2025.10051.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/gmh.2025.10051.
Data availability statement
All data generated or analyzed during this study are not publicly available due to the restrictions from the ethics committee, but are available upon a reasonable request from the corresponding authors (FFR and SH).
Acknowledgments
The authors would like to thank all participants.
Author contribution
FFR and SH designed the study. FFR and SH wrote the article. SH carried out the analysis and interpreted the results. WC, AA, MF, HAMS, MH, IHMH, AYN, BZ, MC, YE-F, GY, GS, AH-M, ER, AH and SO were involved in the data collection. MC reviewed the article for intellectual content. All authors read and approved the final manuscript.
Competing interests
The authors declare none.
Ethics approval and consent to participate
Ethics approval for this study was obtained from the Psychiatric Hospital of the Cross ethics committee, Lebanon (Ref: HPC-012-2022). Written informed consent was obtained from all subjects; the online submission of the soft copy was considered equivalent to receiving written informed consent. All methods were performed in accordance with the relevant guidelines and regulations.
Open access
Comments
Date: 15th March 2025
Dear Editor
Cambridge Prisms: Global Mental Health,
We hereby submit our paper to your esteemed journal for potential consideration and publication. Our manuscript is entitled: “Cross Country Validation of the Arabic version of the WHO-5 Well-Being Index in non-clinical young adults from six Arab countries”.
The vast majority of previous validation and adaptation studies of the WHO-5 Well-Being Index were performed in Western countries with individualistic backgrounds. However, subjective well-being (SWB) is a culturally-dependent and context-driven concept; It thus varies widely across- and within cultures, based on geographical situations. For instance, some findings indicated that individuals from collectivist cultures tend to exhibit lower ratings as compared to those from individualistic cultures, which may result in distinct levels of functioning of the WHO-5 items and overall measure. Despite these data, the cross-cultural validity of well-being scales is still an unexplored question. Some previous studies have investigated the cross-cultural validity of the WHO-5. For example, a study investigated the validity of the WHO-5 in a sample of 3762 adults from five European (i.e., Italy, Poland, Denmark) and non-European (i.e., China, Japan) countries. Another study demonstrated that the WHO-5 is psychometrically appropriate and cross-culturally applicable in different nationally representative samples of individuals (N = 43,469) across 35 European countries. Another study also found that the WHO-5 showed good validity and reliability across Spain, Chile and Norway in nurses who worked during the COVID-19 pandemic. Cross-cultural validation studies are crucial to prove that the measure covers transcultural components of the construct subjective, and can be used for cross-cultural comparison purposes in international multicenter research.
Although people from different Arab countries share similarities (including the language, geography, collectivist identity, religion, a young age structure), diversities do also exist. Large cross-cultural studies have shown that the way Arab people view and behave towards mental health issues is not uniform, and appears to be largely shaped by the local context of each Arab country. Taking into consideration these cultural disparities, it is necessary to examine whether the WHO-5 measures the SWB construct accurately in different Arab countries and cultural backgrounds. In this paper, we aimed to contribute to the literature on SWB in different ways. First, we propose to investigate, for the first time, the cross-cultural validity of the WHO-5 across different Arab countries, to ensure its suitability to capture and provide reliable information on the SWB construct in different Arab contexts. Second, as the two previous validations were conducted in Arab Middle East countries, we intended to expand our investigation to an Arab region and countries (i.e., North Africa, Tunisia and Morocco) that have not been subject of previous validation studies of the WHO-5. Third, we sought to examine important psychometric properties that have not been previously examined, such as measurement invariance across genders. Indeed, gender differences in SWB are culturally-determined, as they may be substantially affected by social norms and adherence to traditional gender roles. However, variations across genders may also be largely driven by methodological factors. It is therefore required to verify that the WHO-5 invariantly measures the SWB factor across gender groups. Fourth, we aspired to include larger samples of participants than have been used in the past in order to provide stronger and more reliable results.
The objective of the present study was to perform a cross-country validation of the Arabic version of the WHO-5. In particular, we aimed to explore its (1) factor structure and composite reliability by country, (2) cross-gender measurement invariance, and (3) concurrent validity by calculating Pearson correlation coefficients between the WHO-5 and measures of depression, anxiety, stress, suicidal ideation and insomnia.
All authors read and approved the final version and the submission.
The authors declare that they have no conflict of interest.
With high respect and appreciation,
The authors