View Article PDF

In New Zealand, obesity, prediabetes and T2DM are serious non-communicable (NCD) diseases that have an impact on overall health and wellbeing.1 New Zealand has the third highest obesity (ie, BMI >30kg/m2) rates (31%) (following Mexico and the US).2 These diseases pose a major challenge for healthcare in New Zealand, place a substantial social-economic burden on the health system,3–6 and are the leading drivers for health inequalities, particularly among New Zealand Pacific peoples.1 Pacific peoples in New Zealand make up 7% of the total population,7 and they have the highest rate (67%) of obesity, compared to Māori (indigenous people of New Zealand) (47%), and the non-Māori non-Pacific population (32%).8 Little is known about how New Zealanders manage the challenges imposed by these conditions on wellbeing. Reducing the incidence and disease impact remains a key issue in disease prevention, and there is a major knowledge gap in how to tailor prevention programmes for sustainable healthier lifestyle change.9

With the scale of the rising NCD problem being more evident among Pacific peoples, recent work has highlighted that a focus on health and wellbeing required in-depth knowledge of: lifestyle factors (eg, poor diet); systemic issues (eg, lack of knowledge); and further understanding of the role of cultural and family responsibilities on Pacific people’s overall health. Through better knowledge and understanding of these issues, effective prevention programmes that place health and wellbeing as a holistic focus10 are considered to be better aligned with Pacific peoples’ cultural and value systems. Researchers have also called for interventions to be ethnic-specific and culturally safe,11 and for programmes that are inclusive of health and wellbeing from a Pacific viewpoint.10 There have been many efforts to develop and implement culturally appropriate intervention programmes;12,13 however, these programmes were not planned, developed, piloted or evaluated with the Pacific communities playing an equal partnership role at the helm of the project.

Mobile health (mHealth) programmes, that is, the use of mobile and wireless tools,14 have been shown to aid the improvement in reducing NCD risk factors and develop healthy behavioural changes.15,16 The OL@-OR@ project was a culturally tailored mHealth programme,17 co-designed between New Zealand health researchers and Māori and Pasifika (defined as a collective group of people representing different Pacific Island Nations18) communities.19 We will refer to Pacific peoples as Pasifika peoples from here onwards. The project employed co-design principles and methods to develop a pragmatic mHealth intervention tool with communities to support better health and wellbeing, through improved nutrition, healthy behaviours and to build better knowledge and awareness of community-level activities, resources and social cohesion. The co-design principles aligned well with indigenous health frameworks, and therefore it was considered to be a good fit, and likely to be well accepted,9 by Pasifika communities.

This paper presents analyses of secondary outcomes of the cRCT and aims to determine the effects of the co-designed mHealth intervention on the wellbeing of Pasifika communities, and to identify the demographic and behavioural factors associated with enhanced wellbeing.

Methods

The OL@-OR@ mHealth programme focused on managing or reducing the key risk factors for NCDs (eg, diet, physical activity, smoking, alcohol). The co-design approach enabled Pasifika communities to include a cultural measure of health that was holistic, Pasifika values-based, and included family and cultural identity as the foundation of health and wellbeing. The Pacific model of health (Fonofale)20 includes four dimensions of health, namely: spiritual, physical, mental and other, and was used to inform the wellbeing measurements (Appendix 1) used in the OL@-OR@ mHealth programme.

The OL@-OR@ mHealth programme was implemented in a 12-week, community-based two-arm, cluster randomised control trial (cRCT) design, administered from between January–December 2018. Eligibility to participate in the trial included self-identification as being Māori or Pasifika, aged ≥18 years, regular mobile device access (eg, smartphone, laptop), regular internet access, and an email account. The main findings of the cRCT intervention have been published, and the trial protocol adheres to the SPIRIT guidelines, which has been published elsewhere and included as Appendix 1.21 However, briefly, the participants were recruited predominantly via face-to-face from 64 community clusters (32 Māori, 32 Pasifika), and these were defined as a distinct New Zealand community context with an average of 20 participants per cluster. For Pasifika clusters, these included groups or communities (eg, churches, sports clubs), as identified by the Pasifika community coordinators (employed by the Pasifika community research partners). All clusters were randomly allocated (1:1 ratio) to either the intervention (mHealth tool) or the control (a control version of the mHealth tool that only selected collected data) group using a computer-generated randomisation list, and block randomisation was used to stratify Pasifika clusters by locality (Auckland/urban or Waikato/rural). The risk of contamination between cluster arms was minimised by recruitment across large geographic areas and multiple diverse community.

Ethical approval for the trial was received from the Northern B Health and Disability Ethics Committee of New Zealand (OL@-OR@) in 2017. All clusters provided written informed consent, and individual participants provided informed consent via an online questionnaire completed at registration.

Pasifika participants in the initial phase of this study, provided an end-user perspective, contributing to the design of the intervention tool, design of outcome measures, analysis of qualitative data, and recruitment pathways for the cRCT. As a secondary outcome measure of the overall cRCT, we included the focus on holistic health and wellbeing status from a Pasifika perspective,22 and this was compared between trial arms.

Study outcomes and analyses

The original sample size calculation was based on the primary outcome for the overall cRCT; self-reported adherence to health-related behaviour guidelines,17 and included complete data from 69 clusters (based on 80% power at a 5% level of significance (two-sided) to detect between group absolute difference of 15% in the primary outcome at 12 weeks post-randomisation). At baseline there were 69 clusters and 1,451 participants (657 Māori and 794 Pasifika), and 84% completed the 12-week follow-up questionnaire (n=1,224).

For the current paper, the Pasifika wellbeing data was extracted as a focus for this investigation, and therefore, all Pasifika participants were included in the analyses, irrespective of whether the participants in each cluster received or used the intervention. In addition, clusters that withdrew from the study or did not register any participants at baseline were excluded. Thus, the overall sample included in this paper was 794 Pasifika (controls n=405 and intervention group n=389). Continuous and categorical variables were presented as numbers observed, means and 95% confidence limits. Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary NJ, US). All statistical tests of significance were based on paired t-tests, t-tests, using the conventional p<0.05.

Defining wellbeing

We defined five areas of ‘wellbeing’ based on the individual variables that were aligned to the four pillars of the Fonofale model of health, and according to a priori knowledge and understanding of Pasifika health and wellbeing, as developed from our previous work.10 This included; spiritual wellbeing: defined as, ‘spiritual beliefs supporting health’ measured on a 5-point Likert scale (1=not very likely to 5=very likely); physical wellbeing defined as, ‘being physically ambulant (without pain)’, all answers were measured on a 5-point Likert scale (1=not very able to 5=very able); mental wellbeing defined as ‘how likely are setting family goals’ and ‘having a positive outlook about life in general’. Each question were measured on a 5-point Likert scale (1=not very able/not very positive to 5=very able/very positive); and ‘family and community life’ was determined by ‘how families rated their capacity to support healthier choices’ and ‘environments that support healthy choices’. All answers were measured on a 5-point Likert scale (1=not very strong at all/not very well at all to 5=very strong/very well). We also aggregated other variables that had indicated significance (data not shown in this paper), to formulate a single ‘combined wellbeing’ variable. This included spiritual, physical and mental domains (including: spiritual beliefs, eating the right-sized portions at social events, mental goals and positive outlook on life). They were considered to be important aspects of wellbeing to our study participants, but not a sufficiently meaningful variable on its own, hence the aggregated approach. Each of these wellbeing variables were measured on a 5-point Likert scale (1=not very able/not very confident at all/not very positive at all to 5=very able/very confident/very positive). From here onwards, we will refer to these as the ‘domains of wellbeing’.

Complementing the findings of the overall cRCT21 this paper presents Pasifika data examining the relationship between the ‘factors of wellbeing’ (dependent variables) with demographics/behaviours (independent variables): Socio-demographic data: age, gender, ethnicity, highest education level; Anthropometry: self-reported weight (in kilograms) and height (in centimetres); Co-morbidities: self-reported health condition(s) defined as being told by a doctor that they have high blood pressure, high cholesterol, diabetes and/or heart disease; and Acculturation: Pasifika and Kiwi-New Zealand Heritage and Lifestyle: Attitudes and beliefs about Pacific and Kiwi/New Zealand heritage and lifestyle measured using an eight-item cultural affiliation questionnaire.23 The acculturation tool used in this study was developed by researchers of the Kohala Health Research Project,24 and is a validated tool for adult Pasifika peoples examining similar health outcomes (metabolic health problems).24 In accordance with the Kohala Health Research Project guidance, we analysed the responses by grouping the summed responses into the following categories: integrated (high affiliation with Pacific heritage and mainstream culture); tradition (high affiliation with Pacific heritage only); assimilated (high affiliation with mainstream culture only); and marginalised (low affiliation with both Pacific heritage and mainstream culture).

Results

Demographics

Table 1 shows the characteristics of all Pasifika study participants. The study communities were mostly located in urban centres (69.5%). The highest education qualifications obtained from the participants was at secondary school (45.0%) and tertiary (32.8%) levels, and the majority of participants were female (65.5%).

Table 1: Distribution of Pasifika participant characteristics, at baseline.

A wide range of Pacific Island nations were represented in the study, with the majority being Samoan (28.3%), Cook Island Māori (25.3%), and Tongan (19.7%). We grouped the remainder under ‘Other Pacific Islands’ because the numbers were too small to include independent island nations on their own.

Age was categorised into approximate quartiles: 18–24 (17.5%); 25–34 years (26.3%); 35–44 years (26.3%) and 45+ (29.9%).

A large proportion (67.6%) of the study participants assessed their acculturation mode as being ‘marginalised’, indicating they had a low degree of affiliation with both their Pacific heritage and the mainstream culture.

Overwhelmingly, obesity BMI (30+) was highly prevalent among the entire Pasifika study sample (69.9%), and this is analogous for both the intervention (69.6%) and control (70.3%) groups. The participants presented with co-morbidities based on known diagnosis: high blood pressure being the most commonly reported. We also included a grouped variable (‘any’) to include all known morbidities. There were no significant differences indicated between the co-morbidity groups.

Table 2 examines the group means (standard deviations) for intervention and control participants at baseline and at 12 weeks. At baseline, there was no significant difference between these two groups. However, at 12 weeks, there was a significant difference between intervention and control groups, for the ‘family and community’ wellbeing (t-test p-value=0.007).

Table 2: Effect of wellbeing factors based on the mHealth programme, at 12 weeks from baseline.

a,b=aggregate variables (see earlier for composition); SD=standard deviation; *=Difference between means, 95%CI= 95% confidence intervals.

Relative to baseline, based on the mean differences (95%CI), both groups showed no change for ‘spiritual’, ‘physical’ and ‘mental’ and ‘combined’ factors of wellbeing. However, there was a significant difference between intervention and control groups for the ‘family and community’, at the end of the 12-week trial (t-test p-value=0.006).

Table 3 summarises the univariate analyses, examining the relationships between each ‘factor of wellbeing’ and the demographic and behavioural variables, at baseline.

Table 3: Wellbeing model: univariate analyses at baseline, by participant characteristics.

a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, dietary portions, mental wellbeing goals for family and positive view on life; R=referent group; *=p<0.05; **=p<0.001.

For spiritual wellbeing: the strongest associations were age (oldest group) (p=0.0001); Other Ethnicity (p=0.0001); being assimilated (p=0.0001) and marginalised (p=0.0001).

As to the physical wellbeing factor, the strongest associations were: the older age groups (35–44 (p=0.0005); and 45 years (p=0.0008); those reporting extreme obesity (BMI 40+) (p=0.0001), and missing obesity data (p=0.015); and participants who identified as being assimilated (p=0.035); traditional (p=0.0005); and marginalised (p=0.0001), and having a co-morbidity (p=0.004).

For the mental wellbeing factor, the relationships were evident among those: in the oldest age group (45+years), p=0.011; participants who were from the Other Pacific Island nations (p=0.009) and Others (p=0.007); having only secondary school qualifications (p=0.010); and participants whose scores identified them as being assimilated (p=0.0001), traditional (p=0.009) and ‘marginalised’ (p=0.0001).

Under the family/community wellbeing factor, the following positive associations with demographic factors were highlighted: from those among the oldest age group (p=0.0001); being from ‘Other Pacific Island’ nations (p=0.002); participants with the lowest education qualifications: ‘none’ and ‘secondary’, p=0.021 and p=0.006, respectively; those who aligned with being ‘assimilated’ (p=0.004); ‘traditional’ (p=0.0007); and ‘marginalised’ (p=0.0001), and participants from the ‘rural’ cluster localities (p=0.021).

The combined wellbeing factor showed significant positive relationships with participants: in the oldest age group (45+yrs) (p=0.0011); being from Other Pacific Island’ nations (p=0.037), and ‘Other’ (p=0.002) ethnic groups; and those who rated their acculturation status as being assimilated (p=0.0001); traditional (p=0.007), and marginalised (p=0.0001), all significantly reported alignment with this wellbeing factor.

Informed by our univariate analyses (Table 3), Table 4 includes the potential independent variables in our multivariate analyses of all participants that provided data at both baseline and at 12 weeks. We excluded BMI and obesity class variables from this analyses as (from earlier models) their significant levels consistently diminished and it was no longer meaningful to retain them in the model. The independent variables were examined by way of mean differences (95CI) from 12 weeks to baseline, for each factor of wellbeing. Notably, only the significant relationships are highlighted in the table.

Table 4: Wellbeing model: multivariate regression (mean differences), from 12 weeks to baseline.

R=Referent group; a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, diet, mental wellbeing goals for family and positive view on life; *=p<0.05; **=p<0.001.

For the spiritual wellbeing factor, after adjusting for all co-variates: being of young age (25–34 years) p=0.031; and acculturation (assimilation and marginalised) p=0.008 and 0.003, respectively, sustained significant improved relationships with this wellbeing factor.

Under the physical wellbeing factor, after adjusting for sex, age, ethnicity, education, cluster region and having any comorbidity, only the participants who were aged 35–44 years (p=0.030) retained a positive association with physical wellbeing. Conversely, those who rated as being ‘marginalised’ (p=0.001) had a very small negative mean difference that was significant, albeit indicating no improvement (-0.09) compared to the ‘integrated’ group, by the end of the trial.

As for the mental wellbeing, after adjusting for all variables, only acculturation sustained a significant relationship with this wellbeing factor. Participants that aligned with being ‘assimilated’ had shown a very small negative association, indicating no improvement (-0.01, 95CI: -0.51–0.49, p=0.002) at the end of the trial. Those participants that affiliated with being ‘marginalised’ had sustained a positive significant relationship and they reported a mild improvement with mental wellbeing (0.27, 95CI: -0.07–0.61, p=0.004), compared to the ‘integrated’ group.

For the family/community wellbeing, after controlling for all co-variates, the participants with ‘no education’ qualifications (1.06, 95CI: 0.57–1.55, p=0.019) showed a large significant (positive) improvement, compared to those participants with any ‘tertiary level’ qualifications (0.53, 95CI: 0.11–0.95). Regarding all acculturation modes, there were significant positive improvements for the ‘marginalised’ group (p=0.0004), followed by the ‘assimilated’ group (p=0.008), compared to the ‘integrated’ group.

Finally, the combined wellbeing factor, after adjusting for all variables, the significant relationships were evident among those participants that corresponded to being ‘assimilated’ (-0.10, 95CI: -0.99–0.79, p=0.0001) – showing no improvement for this wellbeing; and being ‘marginalised’ (0.20, 95CI: -0.40–0.81, p=0.0001), when compared to the ‘integrated’ group.

Discussion

In our large mHealth cRCT programme, we defined the ‘domains of wellbeing’ as being: spiritual, physical, mental, family/community and a combined wellbeing domain, which was an aggregate of various wellbeing measurement scores (Appendix 2). These wellbeing factors were arbitrarily defined by how well the Pasifika participants rated their wellbeing status according to a range of individual characteristics (Table 1).

Principal findings

There are three major findings from our analyses. Firstly, Table 2 showed significant differences between the intervention and control groups for ‘family/community’ wellbeing factor, by the end of the 12-week trial. This is not surprising, given that Pasifika peoples traditionally and have continue to live and participate in social cohesion. There were no differences between intervention and control groups for the remaining wellbeing factors, and this was analogous with the findings from the overall study,21 that also demonstrated that the mHealth programme did not significantly improve adherence to health-related behaviours for all participants. This finding may be explained by the short duration of the trial (12 weeks), and it is possible that a longer duration may have provided more meaningful information.21

The remaining major findings were based on our multivariate analyses (Table 4). The second major finding was ‘acculturation’ as being a major determinant of wellbeing for our Pasifika participants. In particular, the acculturated modes, of being ‘assimilated’ (high affiliation with mainstream culture only) and ‘marginalised’ (low affiliation with both Pacific heritage and mainstream culture) were independently negatively associated with all wellbeing factors. Specifically, those participants who classified themselves as being ‘assimilated’ showed either little or no association with ‘spiritual’, ‘physical’, ‘mental’ and ‘combined’ wellbeing factors. A possible explanation could relate to issues of adapting to the changing dynamics of traditional and cultural practices and values. On the other hand, significant positive associations were evident for those who classified themselves as being ‘marginalised’ for all wellbeing factors (but not physical—very small negative association), and this could be an indicator of cultural resilience. Previous research has shown that some groups facing chronic stresses created by poverty, racism and discrimination due to a lack of security in identity and traditional values,25 and therefore the scores in our study may reflect a lack of bicultural and societal identity.

Of note, the young and working age participants (25–34 years and 35–44 years) showed significant associations with the ‘spiritual’ and ‘physical’ wellbeing factors, which characterises their level of participation in community and church activities.

The third major finding of our study showed clear positive relationships between: ‘no education’, and acculturation modes: ‘assimilation’, and being ‘marginalised’, with the ‘family/community’ wellbeing factor. The high scores highlighted significant positive improvements by the end of the 12-week trial. A possible explanation for having a strong and diverse relationship of acculturation to this wellbeing factor could be related to how Pasifika peoples in our study connect to the Pasifika way of life (ie, cultural values and protocol). This may indicate the growing disconnect between and within Pasifika communities.26 For example, symptoms of living in diasporic communities may be manifested in the way Pasifika peoples view and define their cultural identity as being ‘born’ or ‘raised’,27 and the degree of ‘how well’ they affiliate with the mainstream and, or their Pasifika heritage.22 In relation to education, participants with ‘no education’ had improved because of the programme, and this was evident in our qualitative data (not published), where participants reportedly learnt a lot about healthy lifestyles, because the mHealth tool was relevant to Pasifika culture and values.

Implications of study

Acculturation has recently been redefined from a linear process in which one ethnic/cultural group adopt the beliefs and behaviours of another group,28 to a multi-dimensional process where people engage in different ways.29,30 The finding of associations among ‘marginalised’ participants and ‘family/community’ wellbeing may be indicative of other complex psychosocial factors, such as attitudes, beliefs, emotions and learned behaviours, that have not been catered for in the current study.

Additionally, as the family/community context is a primary environment in which its members grow up and develop their identity, it is possible that the participants in this study experienced different distress and intra-familial stressors as a result of acculturation.25 Therefore, acculturation responses are likely to be different, or conflicting based on personal experiences of acculturation and family/community cohesion.31–33 Thus, the acculturation modes used in this study may only be representative of the participants’ perspectives in relation to how we have defined ‘family/community’ wellbeing. Alternatively, the acculturation tool may not be adequately sufficient to measure the degree and variation of cultural heritage and affiliation. Observing how family and community members function as a nucleus or an extended network system of shared interests, values and experiences maybe a better alternative to understand wellbeing. Unfortunately, our study was not able to gauge participants’ in-depth understanding of acculturation and family/community cohesion.

Study limitations and future work

A major limitation is the potential for selection bias of study participants that may have led to the high proportion of participants indicating their acculturation status as being predominantly ‘marginalised’, and lower education background. Also, there is the potential for participation bias based on the limited use of the mobile/electronic platform of the intervention tool and due to the duration of the mHealth trial (12 weeks), that may have been too short to be able to measure the wellbeing factors at a comprehensive level. Finally, to better understand wellbeing from a Pasifika perspective, further research will be needed to include other domains outside of established health models, including the role of family and community.

Conclusion

Our study utilised Pasifika-only participant data from a large cRCT21 study, to examine the relationship between demographic and behavioural factors and Pasifika wellbeing. From the cRCT findings (Table 2), the programme appears to have supported positive changes, particularly for the intervention participants in ‘family and community’ wellbeing, compared with the controls. Additionally, it was clear from our multivariate analyses that at an individual level, the study participants who identified as being ‘marginalised’ had significantly positive associations with family/community wellbeing. Although the study findings do not fully explain the reasons behind the acculturation, education and age characteristics associations, it does point to the importance of ‘family/community’ as being the most important wellbeing factor for Pasifika peoples. Future work could focus on more in-depth understanding of the psychosocial factors and an up-to-date knowledge of intra-familial and inter-generational perception of acculturation, and its effect on overall wellbeing.

Appendix

Appendix Table 1: Wellbeing questions: Pasifika version.

Pacific and Kiwi/New Zealand heritage and lifestyle

Appendix Table 2:

Summary

Abstract

Aim

The primary objective of this study was to determine the effect of a mobile health (mHealth) intervention on the wellbeing of Pasifika peoples, and to explore factors associated with Pasifika wellbeing.

Method

The OL@-OR@ mHealth programme was a co-designed smartphone app. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The concept of wellbeing was examined as part of a two-arm, cluster randomised trial, using only the Pasifika data: 389 (of 726) Pasifika adults were randomised to receive the mHealth intervention, while 405 (of 725) Pasifika adults were randomised to receive a control version of the intervention. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The intervention effects and the association of demographic and behavioural relationships with wellbeing, was examined using logistic regression analyses.

Results

Relative to baseline, there were significant differences between the intervention and control groups for the ‘family/community’ wellbeing, at the end of the 12-week trial. There were no significant differences observed for all other wellbeing domains for both groups. Based on our multivariate regression analyses, education and acculturation (assimilation and marginalisation) were identified as positively strong factors associated to Pasifika ‘family and community’ wellbeing.

Conclusion

Our study provides new insights on how Pasifika peoples’ characteristics and behaviours align to wellbeing. Our findings point to ‘family and community’ as being the most important wellbeing factor for Pasifika peoples.

Author Information

Ridvan Firestone, Senior Research Officer/Co-Principal Investigator, Centre for Public Health Research, Massey University, Wellington; Soo Cheng, Biostatistician, Centre for Public Health Research, Massey University, Wellington; Sally Dalhousie, Community Partner, The Fono, Auckland; Emily Hughes, Community Coordinator, The Fono, Auckland; Tevita Funaki, Chief Executive, Community Partner, The Fono, Auckland; Akarere Henry, Chief Executive, Community Partner, South Waikato Pacific Islands Community Services, Tokoroa; Mereaumate Vano, Community Coordinator, South Waikato Pacific Islands Community Services, Tokoroa; Jacqui Grey, Project Manager, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Jodie Schumacher, Project Coordinator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Andrew Jull, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; School of Nursing, University of Auckland, Auckland; Robyn Whittaker, Associate Professor, School of Nursing, University of Auckland, Auckland; Lisa Te Morenga, Senior Lecturer/Co-Principal Investigator, School of Health, Faculty of Health, Victoria University of Wellington, Wellington; Cliona Ni Mhurchu, Professor/Co-Principal Investigator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland.

Acknowledgements

We are indebted to the Pasifika communities and their members for participating in this study. We also acknowledge the efforts of everyone that were involved in the overall cRCT. This includes the Māori community partner and their participants from Toi Tangata.

Correspondence

Dr Ridvan Tupai-Firestone, Centre for Public Health Research, Massey University, PO Box 756, Wellington.

Correspondence Email

r.t.firestone@massey.ac.nz

Competing Interests

Nil.

1. Ministry of Health. Long term conditions Wellington: Ministry of Health; 2016 [http://www.health.govt.nz/our-work/diseases-and-conditions/long-term-conditions].

2. OECD Health Data. Obesity and the Economics of Prevention: Fit not Fat. 2011.

3. Ministry of Health. Health Loss in New Zealand: A report from the New Zealand Burden of Diseases, Injuries and Risk Factors Study, 2006–2016. Wellington: Ministry of Health; 2013.

4. Ministry of Health. National Diabetes Work Programme 2014/15. Wellington: Ministry of Health; 2014.

5. Ministry of Health. Annual Update of Key Results 2014/15: New Zealand Health Survey Wellington: Ministry of Health; 2015 [http://www.health.govt.nz/publication/annual-update-key-results-2014-15-new-zealand-health-survey].

6. National Health Committee. Meeting the needs of people with chronic conditions: Hapai te whanau mo ake ake tonu. Wellington: Ministry of Health; 2007.

7. Statistics New Zealand. 2013 Census QuickStats about culture and identity. Wellington: Statistics New Zealand; 2014.

8. Ministry of Health. Annual Update of Key Results 2016/17: New Zealand Health Survey. Wellington: Ministry of Health; 2017.

9. Eyles H, Jull A, Dobson R, Firestone R, Whittaker R, Te Morenga L, et al. Co-design of mHealth Delivered Interventions: A Systematic Review to Assess Key Methods and Processes. Current Nutrition Reports. 2016:1–8.

10. Firestone R FT, Dalhousie S, Henry A, Vano M, Grey J, Jull A, Whittaker R, Te Morenga L, Ni Mhurchu C,. Identifying and overcoming barriers to healthier lives. Pacific Health Dialog. 2018; 21(2):54–66.

11. Wilson D. New perspectives on health disparities and obesity interventions in youth. J Pediatric Psychology. 2009; 34(3):231–44.

12. Counties Manukau District Health Board. LotuMoui Health Programme. Auckland: Counties Manukau District Health Board; 2010.

13. Scragg R, et al. Obesity in Pacific Island Communities. University of Auckland: Health Research Council of New Zealand; 2002.

14. World Health Organisation. mHealth: New horizons for health through mobile technologies. Geneva: World Health Organisation; 2011.

15. Ghorai K AS, Khatun F, et al. mHealth for Smoking Cessation Programs: A Systematic Review. J Pers Med. 2014; 4(3):412–23.

16. Whittaker R MH, Bullen C, et al. Mobile phone-based interventions for smoking cessation. Cochrane Database Syst Rev. 2016; 4(CD006611).

17. Verbiest M SB, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Hughes E, Humpfrey G, Jiang Y, Jull A, Pekepo C, Schumacher J, Te Morenga L, Tunks M, Vano M, Whittaker R, Ni Mhurchu C. A Codesigned, Culturally-Tailored mHealth Tool to Support Healthy Lifestyles in Māori and Pasifika Communities in New Zealand: Protocol for a Cluster Randomized Controlled Trial. JMIR Research Protocols 2018; 7(8):e10789.

18. Statistics New Zealand. QuickStats about Pacific peoples: 2006 Census. Wellington: Statistics New Zealand; 2010.

19. Verbiest M CC, Dalhousie S, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Humpfrey G, Jull A, Vano M, Pekepo C, Te Morenga L, Whittaker R, Ni Mhurchu C,. Using co-design to develop a culturally-tailored, behavior change mHealth intervention for indigenous and other priority communities: a case study in New Zealand. Transl Behav Med. 2018; 8:1–17.

20. Pulotu-Endemann K. Fonofale Model of Health. Auckland: Health Promotion; 2001.

21. Ni Mhurchu C, Te Morenga L, Tupai-Firestone R, Grey J, Jiang Y, Jull A, et al. A co-designed mHealth programme to support healthy lifestyles in Māori and Pasifika peoples in New Zealand (OL@-OR@): a cluster-randomised controlled trial. The Lancet Digital Health. 2019; 1(6):e298–e307.

22. Firestone R TH, Manukia M, Kaholokula K, Foliaki S, Kingi TK, Kruger R, Breier B, O’Connell A, Borman B, Ellison-Loschmann L. Understanding Pasifika youth and the obesogenic environment, Auckland & Wellington, New Zealand. NZ Med J. 2016; 129(1434).

23. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

24. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

25. Rothe E, Tzuang D, Pumariega AJ. Acculturation, Development, and Adaptation Child Adolesc Psychiatric Clin N Am 2010; 19:681–96.

26. Bedford R. Pacific Islanders in New Zealand. Espace Populations Sociétés. 1994:187–200.

27. Tiatia J. Pasifika Cultural Competencies: A literature review. Wellington: Ministry of Health; 2008.

28. Birman D. Acculturation and human diversity in a multicultural society. In: Trickett E, Watts R, Birman D, editors. Human diversity: Perspectives on people in context San Fancisco: JossevBass; 1994:261–84.

29. Berry JW. Conceptual approaches to acculturation. In: Chun KM, Organista PM, Mar´ın G, editors. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

30. Chun KM, Organista PB, Mar´ın G. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

31. Cespedes YM, Huey SJ Jr. Depression in Latino adolescents: a cultural discrepancy perspective. Cultur Divers Ethnic Minor Psychol 2008; 14(2):168–72.

32. Rogler LH. International migrations: a framework for directing research. Am Psychol 1994; 49:701–8.

33. Saldana DH. Acculturative stress and minority status. Journal of Behavioral Health Sciences 1994; 16:117–25.

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

In New Zealand, obesity, prediabetes and T2DM are serious non-communicable (NCD) diseases that have an impact on overall health and wellbeing.1 New Zealand has the third highest obesity (ie, BMI >30kg/m2) rates (31%) (following Mexico and the US).2 These diseases pose a major challenge for healthcare in New Zealand, place a substantial social-economic burden on the health system,3–6 and are the leading drivers for health inequalities, particularly among New Zealand Pacific peoples.1 Pacific peoples in New Zealand make up 7% of the total population,7 and they have the highest rate (67%) of obesity, compared to Māori (indigenous people of New Zealand) (47%), and the non-Māori non-Pacific population (32%).8 Little is known about how New Zealanders manage the challenges imposed by these conditions on wellbeing. Reducing the incidence and disease impact remains a key issue in disease prevention, and there is a major knowledge gap in how to tailor prevention programmes for sustainable healthier lifestyle change.9

With the scale of the rising NCD problem being more evident among Pacific peoples, recent work has highlighted that a focus on health and wellbeing required in-depth knowledge of: lifestyle factors (eg, poor diet); systemic issues (eg, lack of knowledge); and further understanding of the role of cultural and family responsibilities on Pacific people’s overall health. Through better knowledge and understanding of these issues, effective prevention programmes that place health and wellbeing as a holistic focus10 are considered to be better aligned with Pacific peoples’ cultural and value systems. Researchers have also called for interventions to be ethnic-specific and culturally safe,11 and for programmes that are inclusive of health and wellbeing from a Pacific viewpoint.10 There have been many efforts to develop and implement culturally appropriate intervention programmes;12,13 however, these programmes were not planned, developed, piloted or evaluated with the Pacific communities playing an equal partnership role at the helm of the project.

Mobile health (mHealth) programmes, that is, the use of mobile and wireless tools,14 have been shown to aid the improvement in reducing NCD risk factors and develop healthy behavioural changes.15,16 The OL@-OR@ project was a culturally tailored mHealth programme,17 co-designed between New Zealand health researchers and Māori and Pasifika (defined as a collective group of people representing different Pacific Island Nations18) communities.19 We will refer to Pacific peoples as Pasifika peoples from here onwards. The project employed co-design principles and methods to develop a pragmatic mHealth intervention tool with communities to support better health and wellbeing, through improved nutrition, healthy behaviours and to build better knowledge and awareness of community-level activities, resources and social cohesion. The co-design principles aligned well with indigenous health frameworks, and therefore it was considered to be a good fit, and likely to be well accepted,9 by Pasifika communities.

This paper presents analyses of secondary outcomes of the cRCT and aims to determine the effects of the co-designed mHealth intervention on the wellbeing of Pasifika communities, and to identify the demographic and behavioural factors associated with enhanced wellbeing.

Methods

The OL@-OR@ mHealth programme focused on managing or reducing the key risk factors for NCDs (eg, diet, physical activity, smoking, alcohol). The co-design approach enabled Pasifika communities to include a cultural measure of health that was holistic, Pasifika values-based, and included family and cultural identity as the foundation of health and wellbeing. The Pacific model of health (Fonofale)20 includes four dimensions of health, namely: spiritual, physical, mental and other, and was used to inform the wellbeing measurements (Appendix 1) used in the OL@-OR@ mHealth programme.

The OL@-OR@ mHealth programme was implemented in a 12-week, community-based two-arm, cluster randomised control trial (cRCT) design, administered from between January–December 2018. Eligibility to participate in the trial included self-identification as being Māori or Pasifika, aged ≥18 years, regular mobile device access (eg, smartphone, laptop), regular internet access, and an email account. The main findings of the cRCT intervention have been published, and the trial protocol adheres to the SPIRIT guidelines, which has been published elsewhere and included as Appendix 1.21 However, briefly, the participants were recruited predominantly via face-to-face from 64 community clusters (32 Māori, 32 Pasifika), and these were defined as a distinct New Zealand community context with an average of 20 participants per cluster. For Pasifika clusters, these included groups or communities (eg, churches, sports clubs), as identified by the Pasifika community coordinators (employed by the Pasifika community research partners). All clusters were randomly allocated (1:1 ratio) to either the intervention (mHealth tool) or the control (a control version of the mHealth tool that only selected collected data) group using a computer-generated randomisation list, and block randomisation was used to stratify Pasifika clusters by locality (Auckland/urban or Waikato/rural). The risk of contamination between cluster arms was minimised by recruitment across large geographic areas and multiple diverse community.

Ethical approval for the trial was received from the Northern B Health and Disability Ethics Committee of New Zealand (OL@-OR@) in 2017. All clusters provided written informed consent, and individual participants provided informed consent via an online questionnaire completed at registration.

Pasifika participants in the initial phase of this study, provided an end-user perspective, contributing to the design of the intervention tool, design of outcome measures, analysis of qualitative data, and recruitment pathways for the cRCT. As a secondary outcome measure of the overall cRCT, we included the focus on holistic health and wellbeing status from a Pasifika perspective,22 and this was compared between trial arms.

Study outcomes and analyses

The original sample size calculation was based on the primary outcome for the overall cRCT; self-reported adherence to health-related behaviour guidelines,17 and included complete data from 69 clusters (based on 80% power at a 5% level of significance (two-sided) to detect between group absolute difference of 15% in the primary outcome at 12 weeks post-randomisation). At baseline there were 69 clusters and 1,451 participants (657 Māori and 794 Pasifika), and 84% completed the 12-week follow-up questionnaire (n=1,224).

For the current paper, the Pasifika wellbeing data was extracted as a focus for this investigation, and therefore, all Pasifika participants were included in the analyses, irrespective of whether the participants in each cluster received or used the intervention. In addition, clusters that withdrew from the study or did not register any participants at baseline were excluded. Thus, the overall sample included in this paper was 794 Pasifika (controls n=405 and intervention group n=389). Continuous and categorical variables were presented as numbers observed, means and 95% confidence limits. Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary NJ, US). All statistical tests of significance were based on paired t-tests, t-tests, using the conventional p<0.05.

Defining wellbeing

We defined five areas of ‘wellbeing’ based on the individual variables that were aligned to the four pillars of the Fonofale model of health, and according to a priori knowledge and understanding of Pasifika health and wellbeing, as developed from our previous work.10 This included; spiritual wellbeing: defined as, ‘spiritual beliefs supporting health’ measured on a 5-point Likert scale (1=not very likely to 5=very likely); physical wellbeing defined as, ‘being physically ambulant (without pain)’, all answers were measured on a 5-point Likert scale (1=not very able to 5=very able); mental wellbeing defined as ‘how likely are setting family goals’ and ‘having a positive outlook about life in general’. Each question were measured on a 5-point Likert scale (1=not very able/not very positive to 5=very able/very positive); and ‘family and community life’ was determined by ‘how families rated their capacity to support healthier choices’ and ‘environments that support healthy choices’. All answers were measured on a 5-point Likert scale (1=not very strong at all/not very well at all to 5=very strong/very well). We also aggregated other variables that had indicated significance (data not shown in this paper), to formulate a single ‘combined wellbeing’ variable. This included spiritual, physical and mental domains (including: spiritual beliefs, eating the right-sized portions at social events, mental goals and positive outlook on life). They were considered to be important aspects of wellbeing to our study participants, but not a sufficiently meaningful variable on its own, hence the aggregated approach. Each of these wellbeing variables were measured on a 5-point Likert scale (1=not very able/not very confident at all/not very positive at all to 5=very able/very confident/very positive). From here onwards, we will refer to these as the ‘domains of wellbeing’.

Complementing the findings of the overall cRCT21 this paper presents Pasifika data examining the relationship between the ‘factors of wellbeing’ (dependent variables) with demographics/behaviours (independent variables): Socio-demographic data: age, gender, ethnicity, highest education level; Anthropometry: self-reported weight (in kilograms) and height (in centimetres); Co-morbidities: self-reported health condition(s) defined as being told by a doctor that they have high blood pressure, high cholesterol, diabetes and/or heart disease; and Acculturation: Pasifika and Kiwi-New Zealand Heritage and Lifestyle: Attitudes and beliefs about Pacific and Kiwi/New Zealand heritage and lifestyle measured using an eight-item cultural affiliation questionnaire.23 The acculturation tool used in this study was developed by researchers of the Kohala Health Research Project,24 and is a validated tool for adult Pasifika peoples examining similar health outcomes (metabolic health problems).24 In accordance with the Kohala Health Research Project guidance, we analysed the responses by grouping the summed responses into the following categories: integrated (high affiliation with Pacific heritage and mainstream culture); tradition (high affiliation with Pacific heritage only); assimilated (high affiliation with mainstream culture only); and marginalised (low affiliation with both Pacific heritage and mainstream culture).

Results

Demographics

Table 1 shows the characteristics of all Pasifika study participants. The study communities were mostly located in urban centres (69.5%). The highest education qualifications obtained from the participants was at secondary school (45.0%) and tertiary (32.8%) levels, and the majority of participants were female (65.5%).

Table 1: Distribution of Pasifika participant characteristics, at baseline.

A wide range of Pacific Island nations were represented in the study, with the majority being Samoan (28.3%), Cook Island Māori (25.3%), and Tongan (19.7%). We grouped the remainder under ‘Other Pacific Islands’ because the numbers were too small to include independent island nations on their own.

Age was categorised into approximate quartiles: 18–24 (17.5%); 25–34 years (26.3%); 35–44 years (26.3%) and 45+ (29.9%).

A large proportion (67.6%) of the study participants assessed their acculturation mode as being ‘marginalised’, indicating they had a low degree of affiliation with both their Pacific heritage and the mainstream culture.

Overwhelmingly, obesity BMI (30+) was highly prevalent among the entire Pasifika study sample (69.9%), and this is analogous for both the intervention (69.6%) and control (70.3%) groups. The participants presented with co-morbidities based on known diagnosis: high blood pressure being the most commonly reported. We also included a grouped variable (‘any’) to include all known morbidities. There were no significant differences indicated between the co-morbidity groups.

Table 2 examines the group means (standard deviations) for intervention and control participants at baseline and at 12 weeks. At baseline, there was no significant difference between these two groups. However, at 12 weeks, there was a significant difference between intervention and control groups, for the ‘family and community’ wellbeing (t-test p-value=0.007).

Table 2: Effect of wellbeing factors based on the mHealth programme, at 12 weeks from baseline.

a,b=aggregate variables (see earlier for composition); SD=standard deviation; *=Difference between means, 95%CI= 95% confidence intervals.

Relative to baseline, based on the mean differences (95%CI), both groups showed no change for ‘spiritual’, ‘physical’ and ‘mental’ and ‘combined’ factors of wellbeing. However, there was a significant difference between intervention and control groups for the ‘family and community’, at the end of the 12-week trial (t-test p-value=0.006).

Table 3 summarises the univariate analyses, examining the relationships between each ‘factor of wellbeing’ and the demographic and behavioural variables, at baseline.

Table 3: Wellbeing model: univariate analyses at baseline, by participant characteristics.

a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, dietary portions, mental wellbeing goals for family and positive view on life; R=referent group; *=p<0.05; **=p<0.001.

For spiritual wellbeing: the strongest associations were age (oldest group) (p=0.0001); Other Ethnicity (p=0.0001); being assimilated (p=0.0001) and marginalised (p=0.0001).

As to the physical wellbeing factor, the strongest associations were: the older age groups (35–44 (p=0.0005); and 45 years (p=0.0008); those reporting extreme obesity (BMI 40+) (p=0.0001), and missing obesity data (p=0.015); and participants who identified as being assimilated (p=0.035); traditional (p=0.0005); and marginalised (p=0.0001), and having a co-morbidity (p=0.004).

For the mental wellbeing factor, the relationships were evident among those: in the oldest age group (45+years), p=0.011; participants who were from the Other Pacific Island nations (p=0.009) and Others (p=0.007); having only secondary school qualifications (p=0.010); and participants whose scores identified them as being assimilated (p=0.0001), traditional (p=0.009) and ‘marginalised’ (p=0.0001).

Under the family/community wellbeing factor, the following positive associations with demographic factors were highlighted: from those among the oldest age group (p=0.0001); being from ‘Other Pacific Island’ nations (p=0.002); participants with the lowest education qualifications: ‘none’ and ‘secondary’, p=0.021 and p=0.006, respectively; those who aligned with being ‘assimilated’ (p=0.004); ‘traditional’ (p=0.0007); and ‘marginalised’ (p=0.0001), and participants from the ‘rural’ cluster localities (p=0.021).

The combined wellbeing factor showed significant positive relationships with participants: in the oldest age group (45+yrs) (p=0.0011); being from Other Pacific Island’ nations (p=0.037), and ‘Other’ (p=0.002) ethnic groups; and those who rated their acculturation status as being assimilated (p=0.0001); traditional (p=0.007), and marginalised (p=0.0001), all significantly reported alignment with this wellbeing factor.

Informed by our univariate analyses (Table 3), Table 4 includes the potential independent variables in our multivariate analyses of all participants that provided data at both baseline and at 12 weeks. We excluded BMI and obesity class variables from this analyses as (from earlier models) their significant levels consistently diminished and it was no longer meaningful to retain them in the model. The independent variables were examined by way of mean differences (95CI) from 12 weeks to baseline, for each factor of wellbeing. Notably, only the significant relationships are highlighted in the table.

Table 4: Wellbeing model: multivariate regression (mean differences), from 12 weeks to baseline.

R=Referent group; a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, diet, mental wellbeing goals for family and positive view on life; *=p<0.05; **=p<0.001.

For the spiritual wellbeing factor, after adjusting for all co-variates: being of young age (25–34 years) p=0.031; and acculturation (assimilation and marginalised) p=0.008 and 0.003, respectively, sustained significant improved relationships with this wellbeing factor.

Under the physical wellbeing factor, after adjusting for sex, age, ethnicity, education, cluster region and having any comorbidity, only the participants who were aged 35–44 years (p=0.030) retained a positive association with physical wellbeing. Conversely, those who rated as being ‘marginalised’ (p=0.001) had a very small negative mean difference that was significant, albeit indicating no improvement (-0.09) compared to the ‘integrated’ group, by the end of the trial.

As for the mental wellbeing, after adjusting for all variables, only acculturation sustained a significant relationship with this wellbeing factor. Participants that aligned with being ‘assimilated’ had shown a very small negative association, indicating no improvement (-0.01, 95CI: -0.51–0.49, p=0.002) at the end of the trial. Those participants that affiliated with being ‘marginalised’ had sustained a positive significant relationship and they reported a mild improvement with mental wellbeing (0.27, 95CI: -0.07–0.61, p=0.004), compared to the ‘integrated’ group.

For the family/community wellbeing, after controlling for all co-variates, the participants with ‘no education’ qualifications (1.06, 95CI: 0.57–1.55, p=0.019) showed a large significant (positive) improvement, compared to those participants with any ‘tertiary level’ qualifications (0.53, 95CI: 0.11–0.95). Regarding all acculturation modes, there were significant positive improvements for the ‘marginalised’ group (p=0.0004), followed by the ‘assimilated’ group (p=0.008), compared to the ‘integrated’ group.

Finally, the combined wellbeing factor, after adjusting for all variables, the significant relationships were evident among those participants that corresponded to being ‘assimilated’ (-0.10, 95CI: -0.99–0.79, p=0.0001) – showing no improvement for this wellbeing; and being ‘marginalised’ (0.20, 95CI: -0.40–0.81, p=0.0001), when compared to the ‘integrated’ group.

Discussion

In our large mHealth cRCT programme, we defined the ‘domains of wellbeing’ as being: spiritual, physical, mental, family/community and a combined wellbeing domain, which was an aggregate of various wellbeing measurement scores (Appendix 2). These wellbeing factors were arbitrarily defined by how well the Pasifika participants rated their wellbeing status according to a range of individual characteristics (Table 1).

Principal findings

There are three major findings from our analyses. Firstly, Table 2 showed significant differences between the intervention and control groups for ‘family/community’ wellbeing factor, by the end of the 12-week trial. This is not surprising, given that Pasifika peoples traditionally and have continue to live and participate in social cohesion. There were no differences between intervention and control groups for the remaining wellbeing factors, and this was analogous with the findings from the overall study,21 that also demonstrated that the mHealth programme did not significantly improve adherence to health-related behaviours for all participants. This finding may be explained by the short duration of the trial (12 weeks), and it is possible that a longer duration may have provided more meaningful information.21

The remaining major findings were based on our multivariate analyses (Table 4). The second major finding was ‘acculturation’ as being a major determinant of wellbeing for our Pasifika participants. In particular, the acculturated modes, of being ‘assimilated’ (high affiliation with mainstream culture only) and ‘marginalised’ (low affiliation with both Pacific heritage and mainstream culture) were independently negatively associated with all wellbeing factors. Specifically, those participants who classified themselves as being ‘assimilated’ showed either little or no association with ‘spiritual’, ‘physical’, ‘mental’ and ‘combined’ wellbeing factors. A possible explanation could relate to issues of adapting to the changing dynamics of traditional and cultural practices and values. On the other hand, significant positive associations were evident for those who classified themselves as being ‘marginalised’ for all wellbeing factors (but not physical—very small negative association), and this could be an indicator of cultural resilience. Previous research has shown that some groups facing chronic stresses created by poverty, racism and discrimination due to a lack of security in identity and traditional values,25 and therefore the scores in our study may reflect a lack of bicultural and societal identity.

Of note, the young and working age participants (25–34 years and 35–44 years) showed significant associations with the ‘spiritual’ and ‘physical’ wellbeing factors, which characterises their level of participation in community and church activities.

The third major finding of our study showed clear positive relationships between: ‘no education’, and acculturation modes: ‘assimilation’, and being ‘marginalised’, with the ‘family/community’ wellbeing factor. The high scores highlighted significant positive improvements by the end of the 12-week trial. A possible explanation for having a strong and diverse relationship of acculturation to this wellbeing factor could be related to how Pasifika peoples in our study connect to the Pasifika way of life (ie, cultural values and protocol). This may indicate the growing disconnect between and within Pasifika communities.26 For example, symptoms of living in diasporic communities may be manifested in the way Pasifika peoples view and define their cultural identity as being ‘born’ or ‘raised’,27 and the degree of ‘how well’ they affiliate with the mainstream and, or their Pasifika heritage.22 In relation to education, participants with ‘no education’ had improved because of the programme, and this was evident in our qualitative data (not published), where participants reportedly learnt a lot about healthy lifestyles, because the mHealth tool was relevant to Pasifika culture and values.

Implications of study

Acculturation has recently been redefined from a linear process in which one ethnic/cultural group adopt the beliefs and behaviours of another group,28 to a multi-dimensional process where people engage in different ways.29,30 The finding of associations among ‘marginalised’ participants and ‘family/community’ wellbeing may be indicative of other complex psychosocial factors, such as attitudes, beliefs, emotions and learned behaviours, that have not been catered for in the current study.

Additionally, as the family/community context is a primary environment in which its members grow up and develop their identity, it is possible that the participants in this study experienced different distress and intra-familial stressors as a result of acculturation.25 Therefore, acculturation responses are likely to be different, or conflicting based on personal experiences of acculturation and family/community cohesion.31–33 Thus, the acculturation modes used in this study may only be representative of the participants’ perspectives in relation to how we have defined ‘family/community’ wellbeing. Alternatively, the acculturation tool may not be adequately sufficient to measure the degree and variation of cultural heritage and affiliation. Observing how family and community members function as a nucleus or an extended network system of shared interests, values and experiences maybe a better alternative to understand wellbeing. Unfortunately, our study was not able to gauge participants’ in-depth understanding of acculturation and family/community cohesion.

Study limitations and future work

A major limitation is the potential for selection bias of study participants that may have led to the high proportion of participants indicating their acculturation status as being predominantly ‘marginalised’, and lower education background. Also, there is the potential for participation bias based on the limited use of the mobile/electronic platform of the intervention tool and due to the duration of the mHealth trial (12 weeks), that may have been too short to be able to measure the wellbeing factors at a comprehensive level. Finally, to better understand wellbeing from a Pasifika perspective, further research will be needed to include other domains outside of established health models, including the role of family and community.

Conclusion

Our study utilised Pasifika-only participant data from a large cRCT21 study, to examine the relationship between demographic and behavioural factors and Pasifika wellbeing. From the cRCT findings (Table 2), the programme appears to have supported positive changes, particularly for the intervention participants in ‘family and community’ wellbeing, compared with the controls. Additionally, it was clear from our multivariate analyses that at an individual level, the study participants who identified as being ‘marginalised’ had significantly positive associations with family/community wellbeing. Although the study findings do not fully explain the reasons behind the acculturation, education and age characteristics associations, it does point to the importance of ‘family/community’ as being the most important wellbeing factor for Pasifika peoples. Future work could focus on more in-depth understanding of the psychosocial factors and an up-to-date knowledge of intra-familial and inter-generational perception of acculturation, and its effect on overall wellbeing.

Appendix

Appendix Table 1: Wellbeing questions: Pasifika version.

Pacific and Kiwi/New Zealand heritage and lifestyle

Appendix Table 2:

Summary

Abstract

Aim

The primary objective of this study was to determine the effect of a mobile health (mHealth) intervention on the wellbeing of Pasifika peoples, and to explore factors associated with Pasifika wellbeing.

Method

The OL@-OR@ mHealth programme was a co-designed smartphone app. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The concept of wellbeing was examined as part of a two-arm, cluster randomised trial, using only the Pasifika data: 389 (of 726) Pasifika adults were randomised to receive the mHealth intervention, while 405 (of 725) Pasifika adults were randomised to receive a control version of the intervention. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The intervention effects and the association of demographic and behavioural relationships with wellbeing, was examined using logistic regression analyses.

Results

Relative to baseline, there were significant differences between the intervention and control groups for the ‘family/community’ wellbeing, at the end of the 12-week trial. There were no significant differences observed for all other wellbeing domains for both groups. Based on our multivariate regression analyses, education and acculturation (assimilation and marginalisation) were identified as positively strong factors associated to Pasifika ‘family and community’ wellbeing.

Conclusion

Our study provides new insights on how Pasifika peoples’ characteristics and behaviours align to wellbeing. Our findings point to ‘family and community’ as being the most important wellbeing factor for Pasifika peoples.

Author Information

Ridvan Firestone, Senior Research Officer/Co-Principal Investigator, Centre for Public Health Research, Massey University, Wellington; Soo Cheng, Biostatistician, Centre for Public Health Research, Massey University, Wellington; Sally Dalhousie, Community Partner, The Fono, Auckland; Emily Hughes, Community Coordinator, The Fono, Auckland; Tevita Funaki, Chief Executive, Community Partner, The Fono, Auckland; Akarere Henry, Chief Executive, Community Partner, South Waikato Pacific Islands Community Services, Tokoroa; Mereaumate Vano, Community Coordinator, South Waikato Pacific Islands Community Services, Tokoroa; Jacqui Grey, Project Manager, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Jodie Schumacher, Project Coordinator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Andrew Jull, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; School of Nursing, University of Auckland, Auckland; Robyn Whittaker, Associate Professor, School of Nursing, University of Auckland, Auckland; Lisa Te Morenga, Senior Lecturer/Co-Principal Investigator, School of Health, Faculty of Health, Victoria University of Wellington, Wellington; Cliona Ni Mhurchu, Professor/Co-Principal Investigator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland.

Acknowledgements

We are indebted to the Pasifika communities and their members for participating in this study. We also acknowledge the efforts of everyone that were involved in the overall cRCT. This includes the Māori community partner and their participants from Toi Tangata.

Correspondence

Dr Ridvan Tupai-Firestone, Centre for Public Health Research, Massey University, PO Box 756, Wellington.

Correspondence Email

r.t.firestone@massey.ac.nz

Competing Interests

Nil.

1. Ministry of Health. Long term conditions Wellington: Ministry of Health; 2016 [http://www.health.govt.nz/our-work/diseases-and-conditions/long-term-conditions].

2. OECD Health Data. Obesity and the Economics of Prevention: Fit not Fat. 2011.

3. Ministry of Health. Health Loss in New Zealand: A report from the New Zealand Burden of Diseases, Injuries and Risk Factors Study, 2006–2016. Wellington: Ministry of Health; 2013.

4. Ministry of Health. National Diabetes Work Programme 2014/15. Wellington: Ministry of Health; 2014.

5. Ministry of Health. Annual Update of Key Results 2014/15: New Zealand Health Survey Wellington: Ministry of Health; 2015 [http://www.health.govt.nz/publication/annual-update-key-results-2014-15-new-zealand-health-survey].

6. National Health Committee. Meeting the needs of people with chronic conditions: Hapai te whanau mo ake ake tonu. Wellington: Ministry of Health; 2007.

7. Statistics New Zealand. 2013 Census QuickStats about culture and identity. Wellington: Statistics New Zealand; 2014.

8. Ministry of Health. Annual Update of Key Results 2016/17: New Zealand Health Survey. Wellington: Ministry of Health; 2017.

9. Eyles H, Jull A, Dobson R, Firestone R, Whittaker R, Te Morenga L, et al. Co-design of mHealth Delivered Interventions: A Systematic Review to Assess Key Methods and Processes. Current Nutrition Reports. 2016:1–8.

10. Firestone R FT, Dalhousie S, Henry A, Vano M, Grey J, Jull A, Whittaker R, Te Morenga L, Ni Mhurchu C,. Identifying and overcoming barriers to healthier lives. Pacific Health Dialog. 2018; 21(2):54–66.

11. Wilson D. New perspectives on health disparities and obesity interventions in youth. J Pediatric Psychology. 2009; 34(3):231–44.

12. Counties Manukau District Health Board. LotuMoui Health Programme. Auckland: Counties Manukau District Health Board; 2010.

13. Scragg R, et al. Obesity in Pacific Island Communities. University of Auckland: Health Research Council of New Zealand; 2002.

14. World Health Organisation. mHealth: New horizons for health through mobile technologies. Geneva: World Health Organisation; 2011.

15. Ghorai K AS, Khatun F, et al. mHealth for Smoking Cessation Programs: A Systematic Review. J Pers Med. 2014; 4(3):412–23.

16. Whittaker R MH, Bullen C, et al. Mobile phone-based interventions for smoking cessation. Cochrane Database Syst Rev. 2016; 4(CD006611).

17. Verbiest M SB, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Hughes E, Humpfrey G, Jiang Y, Jull A, Pekepo C, Schumacher J, Te Morenga L, Tunks M, Vano M, Whittaker R, Ni Mhurchu C. A Codesigned, Culturally-Tailored mHealth Tool to Support Healthy Lifestyles in Māori and Pasifika Communities in New Zealand: Protocol for a Cluster Randomized Controlled Trial. JMIR Research Protocols 2018; 7(8):e10789.

18. Statistics New Zealand. QuickStats about Pacific peoples: 2006 Census. Wellington: Statistics New Zealand; 2010.

19. Verbiest M CC, Dalhousie S, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Humpfrey G, Jull A, Vano M, Pekepo C, Te Morenga L, Whittaker R, Ni Mhurchu C,. Using co-design to develop a culturally-tailored, behavior change mHealth intervention for indigenous and other priority communities: a case study in New Zealand. Transl Behav Med. 2018; 8:1–17.

20. Pulotu-Endemann K. Fonofale Model of Health. Auckland: Health Promotion; 2001.

21. Ni Mhurchu C, Te Morenga L, Tupai-Firestone R, Grey J, Jiang Y, Jull A, et al. A co-designed mHealth programme to support healthy lifestyles in Māori and Pasifika peoples in New Zealand (OL@-OR@): a cluster-randomised controlled trial. The Lancet Digital Health. 2019; 1(6):e298–e307.

22. Firestone R TH, Manukia M, Kaholokula K, Foliaki S, Kingi TK, Kruger R, Breier B, O’Connell A, Borman B, Ellison-Loschmann L. Understanding Pasifika youth and the obesogenic environment, Auckland & Wellington, New Zealand. NZ Med J. 2016; 129(1434).

23. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

24. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

25. Rothe E, Tzuang D, Pumariega AJ. Acculturation, Development, and Adaptation Child Adolesc Psychiatric Clin N Am 2010; 19:681–96.

26. Bedford R. Pacific Islanders in New Zealand. Espace Populations Sociétés. 1994:187–200.

27. Tiatia J. Pasifika Cultural Competencies: A literature review. Wellington: Ministry of Health; 2008.

28. Birman D. Acculturation and human diversity in a multicultural society. In: Trickett E, Watts R, Birman D, editors. Human diversity: Perspectives on people in context San Fancisco: JossevBass; 1994:261–84.

29. Berry JW. Conceptual approaches to acculturation. In: Chun KM, Organista PM, Mar´ın G, editors. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

30. Chun KM, Organista PB, Mar´ın G. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

31. Cespedes YM, Huey SJ Jr. Depression in Latino adolescents: a cultural discrepancy perspective. Cultur Divers Ethnic Minor Psychol 2008; 14(2):168–72.

32. Rogler LH. International migrations: a framework for directing research. Am Psychol 1994; 49:701–8.

33. Saldana DH. Acculturative stress and minority status. Journal of Behavioral Health Sciences 1994; 16:117–25.

Contact diana@nzma.org.nz
for the PDF of this article

View Article PDF

In New Zealand, obesity, prediabetes and T2DM are serious non-communicable (NCD) diseases that have an impact on overall health and wellbeing.1 New Zealand has the third highest obesity (ie, BMI >30kg/m2) rates (31%) (following Mexico and the US).2 These diseases pose a major challenge for healthcare in New Zealand, place a substantial social-economic burden on the health system,3–6 and are the leading drivers for health inequalities, particularly among New Zealand Pacific peoples.1 Pacific peoples in New Zealand make up 7% of the total population,7 and they have the highest rate (67%) of obesity, compared to Māori (indigenous people of New Zealand) (47%), and the non-Māori non-Pacific population (32%).8 Little is known about how New Zealanders manage the challenges imposed by these conditions on wellbeing. Reducing the incidence and disease impact remains a key issue in disease prevention, and there is a major knowledge gap in how to tailor prevention programmes for sustainable healthier lifestyle change.9

With the scale of the rising NCD problem being more evident among Pacific peoples, recent work has highlighted that a focus on health and wellbeing required in-depth knowledge of: lifestyle factors (eg, poor diet); systemic issues (eg, lack of knowledge); and further understanding of the role of cultural and family responsibilities on Pacific people’s overall health. Through better knowledge and understanding of these issues, effective prevention programmes that place health and wellbeing as a holistic focus10 are considered to be better aligned with Pacific peoples’ cultural and value systems. Researchers have also called for interventions to be ethnic-specific and culturally safe,11 and for programmes that are inclusive of health and wellbeing from a Pacific viewpoint.10 There have been many efforts to develop and implement culturally appropriate intervention programmes;12,13 however, these programmes were not planned, developed, piloted or evaluated with the Pacific communities playing an equal partnership role at the helm of the project.

Mobile health (mHealth) programmes, that is, the use of mobile and wireless tools,14 have been shown to aid the improvement in reducing NCD risk factors and develop healthy behavioural changes.15,16 The OL@-OR@ project was a culturally tailored mHealth programme,17 co-designed between New Zealand health researchers and Māori and Pasifika (defined as a collective group of people representing different Pacific Island Nations18) communities.19 We will refer to Pacific peoples as Pasifika peoples from here onwards. The project employed co-design principles and methods to develop a pragmatic mHealth intervention tool with communities to support better health and wellbeing, through improved nutrition, healthy behaviours and to build better knowledge and awareness of community-level activities, resources and social cohesion. The co-design principles aligned well with indigenous health frameworks, and therefore it was considered to be a good fit, and likely to be well accepted,9 by Pasifika communities.

This paper presents analyses of secondary outcomes of the cRCT and aims to determine the effects of the co-designed mHealth intervention on the wellbeing of Pasifika communities, and to identify the demographic and behavioural factors associated with enhanced wellbeing.

Methods

The OL@-OR@ mHealth programme focused on managing or reducing the key risk factors for NCDs (eg, diet, physical activity, smoking, alcohol). The co-design approach enabled Pasifika communities to include a cultural measure of health that was holistic, Pasifika values-based, and included family and cultural identity as the foundation of health and wellbeing. The Pacific model of health (Fonofale)20 includes four dimensions of health, namely: spiritual, physical, mental and other, and was used to inform the wellbeing measurements (Appendix 1) used in the OL@-OR@ mHealth programme.

The OL@-OR@ mHealth programme was implemented in a 12-week, community-based two-arm, cluster randomised control trial (cRCT) design, administered from between January–December 2018. Eligibility to participate in the trial included self-identification as being Māori or Pasifika, aged ≥18 years, regular mobile device access (eg, smartphone, laptop), regular internet access, and an email account. The main findings of the cRCT intervention have been published, and the trial protocol adheres to the SPIRIT guidelines, which has been published elsewhere and included as Appendix 1.21 However, briefly, the participants were recruited predominantly via face-to-face from 64 community clusters (32 Māori, 32 Pasifika), and these were defined as a distinct New Zealand community context with an average of 20 participants per cluster. For Pasifika clusters, these included groups or communities (eg, churches, sports clubs), as identified by the Pasifika community coordinators (employed by the Pasifika community research partners). All clusters were randomly allocated (1:1 ratio) to either the intervention (mHealth tool) or the control (a control version of the mHealth tool that only selected collected data) group using a computer-generated randomisation list, and block randomisation was used to stratify Pasifika clusters by locality (Auckland/urban or Waikato/rural). The risk of contamination between cluster arms was minimised by recruitment across large geographic areas and multiple diverse community.

Ethical approval for the trial was received from the Northern B Health and Disability Ethics Committee of New Zealand (OL@-OR@) in 2017. All clusters provided written informed consent, and individual participants provided informed consent via an online questionnaire completed at registration.

Pasifika participants in the initial phase of this study, provided an end-user perspective, contributing to the design of the intervention tool, design of outcome measures, analysis of qualitative data, and recruitment pathways for the cRCT. As a secondary outcome measure of the overall cRCT, we included the focus on holistic health and wellbeing status from a Pasifika perspective,22 and this was compared between trial arms.

Study outcomes and analyses

The original sample size calculation was based on the primary outcome for the overall cRCT; self-reported adherence to health-related behaviour guidelines,17 and included complete data from 69 clusters (based on 80% power at a 5% level of significance (two-sided) to detect between group absolute difference of 15% in the primary outcome at 12 weeks post-randomisation). At baseline there were 69 clusters and 1,451 participants (657 Māori and 794 Pasifika), and 84% completed the 12-week follow-up questionnaire (n=1,224).

For the current paper, the Pasifika wellbeing data was extracted as a focus for this investigation, and therefore, all Pasifika participants were included in the analyses, irrespective of whether the participants in each cluster received or used the intervention. In addition, clusters that withdrew from the study or did not register any participants at baseline were excluded. Thus, the overall sample included in this paper was 794 Pasifika (controls n=405 and intervention group n=389). Continuous and categorical variables were presented as numbers observed, means and 95% confidence limits. Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary NJ, US). All statistical tests of significance were based on paired t-tests, t-tests, using the conventional p<0.05.

Defining wellbeing

We defined five areas of ‘wellbeing’ based on the individual variables that were aligned to the four pillars of the Fonofale model of health, and according to a priori knowledge and understanding of Pasifika health and wellbeing, as developed from our previous work.10 This included; spiritual wellbeing: defined as, ‘spiritual beliefs supporting health’ measured on a 5-point Likert scale (1=not very likely to 5=very likely); physical wellbeing defined as, ‘being physically ambulant (without pain)’, all answers were measured on a 5-point Likert scale (1=not very able to 5=very able); mental wellbeing defined as ‘how likely are setting family goals’ and ‘having a positive outlook about life in general’. Each question were measured on a 5-point Likert scale (1=not very able/not very positive to 5=very able/very positive); and ‘family and community life’ was determined by ‘how families rated their capacity to support healthier choices’ and ‘environments that support healthy choices’. All answers were measured on a 5-point Likert scale (1=not very strong at all/not very well at all to 5=very strong/very well). We also aggregated other variables that had indicated significance (data not shown in this paper), to formulate a single ‘combined wellbeing’ variable. This included spiritual, physical and mental domains (including: spiritual beliefs, eating the right-sized portions at social events, mental goals and positive outlook on life). They were considered to be important aspects of wellbeing to our study participants, but not a sufficiently meaningful variable on its own, hence the aggregated approach. Each of these wellbeing variables were measured on a 5-point Likert scale (1=not very able/not very confident at all/not very positive at all to 5=very able/very confident/very positive). From here onwards, we will refer to these as the ‘domains of wellbeing’.

Complementing the findings of the overall cRCT21 this paper presents Pasifika data examining the relationship between the ‘factors of wellbeing’ (dependent variables) with demographics/behaviours (independent variables): Socio-demographic data: age, gender, ethnicity, highest education level; Anthropometry: self-reported weight (in kilograms) and height (in centimetres); Co-morbidities: self-reported health condition(s) defined as being told by a doctor that they have high blood pressure, high cholesterol, diabetes and/or heart disease; and Acculturation: Pasifika and Kiwi-New Zealand Heritage and Lifestyle: Attitudes and beliefs about Pacific and Kiwi/New Zealand heritage and lifestyle measured using an eight-item cultural affiliation questionnaire.23 The acculturation tool used in this study was developed by researchers of the Kohala Health Research Project,24 and is a validated tool for adult Pasifika peoples examining similar health outcomes (metabolic health problems).24 In accordance with the Kohala Health Research Project guidance, we analysed the responses by grouping the summed responses into the following categories: integrated (high affiliation with Pacific heritage and mainstream culture); tradition (high affiliation with Pacific heritage only); assimilated (high affiliation with mainstream culture only); and marginalised (low affiliation with both Pacific heritage and mainstream culture).

Results

Demographics

Table 1 shows the characteristics of all Pasifika study participants. The study communities were mostly located in urban centres (69.5%). The highest education qualifications obtained from the participants was at secondary school (45.0%) and tertiary (32.8%) levels, and the majority of participants were female (65.5%).

Table 1: Distribution of Pasifika participant characteristics, at baseline.

A wide range of Pacific Island nations were represented in the study, with the majority being Samoan (28.3%), Cook Island Māori (25.3%), and Tongan (19.7%). We grouped the remainder under ‘Other Pacific Islands’ because the numbers were too small to include independent island nations on their own.

Age was categorised into approximate quartiles: 18–24 (17.5%); 25–34 years (26.3%); 35–44 years (26.3%) and 45+ (29.9%).

A large proportion (67.6%) of the study participants assessed their acculturation mode as being ‘marginalised’, indicating they had a low degree of affiliation with both their Pacific heritage and the mainstream culture.

Overwhelmingly, obesity BMI (30+) was highly prevalent among the entire Pasifika study sample (69.9%), and this is analogous for both the intervention (69.6%) and control (70.3%) groups. The participants presented with co-morbidities based on known diagnosis: high blood pressure being the most commonly reported. We also included a grouped variable (‘any’) to include all known morbidities. There were no significant differences indicated between the co-morbidity groups.

Table 2 examines the group means (standard deviations) for intervention and control participants at baseline and at 12 weeks. At baseline, there was no significant difference between these two groups. However, at 12 weeks, there was a significant difference between intervention and control groups, for the ‘family and community’ wellbeing (t-test p-value=0.007).

Table 2: Effect of wellbeing factors based on the mHealth programme, at 12 weeks from baseline.

a,b=aggregate variables (see earlier for composition); SD=standard deviation; *=Difference between means, 95%CI= 95% confidence intervals.

Relative to baseline, based on the mean differences (95%CI), both groups showed no change for ‘spiritual’, ‘physical’ and ‘mental’ and ‘combined’ factors of wellbeing. However, there was a significant difference between intervention and control groups for the ‘family and community’, at the end of the 12-week trial (t-test p-value=0.006).

Table 3 summarises the univariate analyses, examining the relationships between each ‘factor of wellbeing’ and the demographic and behavioural variables, at baseline.

Table 3: Wellbeing model: univariate analyses at baseline, by participant characteristics.

a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, dietary portions, mental wellbeing goals for family and positive view on life; R=referent group; *=p<0.05; **=p<0.001.

For spiritual wellbeing: the strongest associations were age (oldest group) (p=0.0001); Other Ethnicity (p=0.0001); being assimilated (p=0.0001) and marginalised (p=0.0001).

As to the physical wellbeing factor, the strongest associations were: the older age groups (35–44 (p=0.0005); and 45 years (p=0.0008); those reporting extreme obesity (BMI 40+) (p=0.0001), and missing obesity data (p=0.015); and participants who identified as being assimilated (p=0.035); traditional (p=0.0005); and marginalised (p=0.0001), and having a co-morbidity (p=0.004).

For the mental wellbeing factor, the relationships were evident among those: in the oldest age group (45+years), p=0.011; participants who were from the Other Pacific Island nations (p=0.009) and Others (p=0.007); having only secondary school qualifications (p=0.010); and participants whose scores identified them as being assimilated (p=0.0001), traditional (p=0.009) and ‘marginalised’ (p=0.0001).

Under the family/community wellbeing factor, the following positive associations with demographic factors were highlighted: from those among the oldest age group (p=0.0001); being from ‘Other Pacific Island’ nations (p=0.002); participants with the lowest education qualifications: ‘none’ and ‘secondary’, p=0.021 and p=0.006, respectively; those who aligned with being ‘assimilated’ (p=0.004); ‘traditional’ (p=0.0007); and ‘marginalised’ (p=0.0001), and participants from the ‘rural’ cluster localities (p=0.021).

The combined wellbeing factor showed significant positive relationships with participants: in the oldest age group (45+yrs) (p=0.0011); being from Other Pacific Island’ nations (p=0.037), and ‘Other’ (p=0.002) ethnic groups; and those who rated their acculturation status as being assimilated (p=0.0001); traditional (p=0.007), and marginalised (p=0.0001), all significantly reported alignment with this wellbeing factor.

Informed by our univariate analyses (Table 3), Table 4 includes the potential independent variables in our multivariate analyses of all participants that provided data at both baseline and at 12 weeks. We excluded BMI and obesity class variables from this analyses as (from earlier models) their significant levels consistently diminished and it was no longer meaningful to retain them in the model. The independent variables were examined by way of mean differences (95CI) from 12 weeks to baseline, for each factor of wellbeing. Notably, only the significant relationships are highlighted in the table.

Table 4: Wellbeing model: multivariate regression (mean differences), from 12 weeks to baseline.

R=Referent group; a=Aggregate of family and community goals, active participation in community life, and rating of family’s ability to make healthy choices; b=Aggregate of spiritual, diet, mental wellbeing goals for family and positive view on life; *=p<0.05; **=p<0.001.

For the spiritual wellbeing factor, after adjusting for all co-variates: being of young age (25–34 years) p=0.031; and acculturation (assimilation and marginalised) p=0.008 and 0.003, respectively, sustained significant improved relationships with this wellbeing factor.

Under the physical wellbeing factor, after adjusting for sex, age, ethnicity, education, cluster region and having any comorbidity, only the participants who were aged 35–44 years (p=0.030) retained a positive association with physical wellbeing. Conversely, those who rated as being ‘marginalised’ (p=0.001) had a very small negative mean difference that was significant, albeit indicating no improvement (-0.09) compared to the ‘integrated’ group, by the end of the trial.

As for the mental wellbeing, after adjusting for all variables, only acculturation sustained a significant relationship with this wellbeing factor. Participants that aligned with being ‘assimilated’ had shown a very small negative association, indicating no improvement (-0.01, 95CI: -0.51–0.49, p=0.002) at the end of the trial. Those participants that affiliated with being ‘marginalised’ had sustained a positive significant relationship and they reported a mild improvement with mental wellbeing (0.27, 95CI: -0.07–0.61, p=0.004), compared to the ‘integrated’ group.

For the family/community wellbeing, after controlling for all co-variates, the participants with ‘no education’ qualifications (1.06, 95CI: 0.57–1.55, p=0.019) showed a large significant (positive) improvement, compared to those participants with any ‘tertiary level’ qualifications (0.53, 95CI: 0.11–0.95). Regarding all acculturation modes, there were significant positive improvements for the ‘marginalised’ group (p=0.0004), followed by the ‘assimilated’ group (p=0.008), compared to the ‘integrated’ group.

Finally, the combined wellbeing factor, after adjusting for all variables, the significant relationships were evident among those participants that corresponded to being ‘assimilated’ (-0.10, 95CI: -0.99–0.79, p=0.0001) – showing no improvement for this wellbeing; and being ‘marginalised’ (0.20, 95CI: -0.40–0.81, p=0.0001), when compared to the ‘integrated’ group.

Discussion

In our large mHealth cRCT programme, we defined the ‘domains of wellbeing’ as being: spiritual, physical, mental, family/community and a combined wellbeing domain, which was an aggregate of various wellbeing measurement scores (Appendix 2). These wellbeing factors were arbitrarily defined by how well the Pasifika participants rated their wellbeing status according to a range of individual characteristics (Table 1).

Principal findings

There are three major findings from our analyses. Firstly, Table 2 showed significant differences between the intervention and control groups for ‘family/community’ wellbeing factor, by the end of the 12-week trial. This is not surprising, given that Pasifika peoples traditionally and have continue to live and participate in social cohesion. There were no differences between intervention and control groups for the remaining wellbeing factors, and this was analogous with the findings from the overall study,21 that also demonstrated that the mHealth programme did not significantly improve adherence to health-related behaviours for all participants. This finding may be explained by the short duration of the trial (12 weeks), and it is possible that a longer duration may have provided more meaningful information.21

The remaining major findings were based on our multivariate analyses (Table 4). The second major finding was ‘acculturation’ as being a major determinant of wellbeing for our Pasifika participants. In particular, the acculturated modes, of being ‘assimilated’ (high affiliation with mainstream culture only) and ‘marginalised’ (low affiliation with both Pacific heritage and mainstream culture) were independently negatively associated with all wellbeing factors. Specifically, those participants who classified themselves as being ‘assimilated’ showed either little or no association with ‘spiritual’, ‘physical’, ‘mental’ and ‘combined’ wellbeing factors. A possible explanation could relate to issues of adapting to the changing dynamics of traditional and cultural practices and values. On the other hand, significant positive associations were evident for those who classified themselves as being ‘marginalised’ for all wellbeing factors (but not physical—very small negative association), and this could be an indicator of cultural resilience. Previous research has shown that some groups facing chronic stresses created by poverty, racism and discrimination due to a lack of security in identity and traditional values,25 and therefore the scores in our study may reflect a lack of bicultural and societal identity.

Of note, the young and working age participants (25–34 years and 35–44 years) showed significant associations with the ‘spiritual’ and ‘physical’ wellbeing factors, which characterises their level of participation in community and church activities.

The third major finding of our study showed clear positive relationships between: ‘no education’, and acculturation modes: ‘assimilation’, and being ‘marginalised’, with the ‘family/community’ wellbeing factor. The high scores highlighted significant positive improvements by the end of the 12-week trial. A possible explanation for having a strong and diverse relationship of acculturation to this wellbeing factor could be related to how Pasifika peoples in our study connect to the Pasifika way of life (ie, cultural values and protocol). This may indicate the growing disconnect between and within Pasifika communities.26 For example, symptoms of living in diasporic communities may be manifested in the way Pasifika peoples view and define their cultural identity as being ‘born’ or ‘raised’,27 and the degree of ‘how well’ they affiliate with the mainstream and, or their Pasifika heritage.22 In relation to education, participants with ‘no education’ had improved because of the programme, and this was evident in our qualitative data (not published), where participants reportedly learnt a lot about healthy lifestyles, because the mHealth tool was relevant to Pasifika culture and values.

Implications of study

Acculturation has recently been redefined from a linear process in which one ethnic/cultural group adopt the beliefs and behaviours of another group,28 to a multi-dimensional process where people engage in different ways.29,30 The finding of associations among ‘marginalised’ participants and ‘family/community’ wellbeing may be indicative of other complex psychosocial factors, such as attitudes, beliefs, emotions and learned behaviours, that have not been catered for in the current study.

Additionally, as the family/community context is a primary environment in which its members grow up and develop their identity, it is possible that the participants in this study experienced different distress and intra-familial stressors as a result of acculturation.25 Therefore, acculturation responses are likely to be different, or conflicting based on personal experiences of acculturation and family/community cohesion.31–33 Thus, the acculturation modes used in this study may only be representative of the participants’ perspectives in relation to how we have defined ‘family/community’ wellbeing. Alternatively, the acculturation tool may not be adequately sufficient to measure the degree and variation of cultural heritage and affiliation. Observing how family and community members function as a nucleus or an extended network system of shared interests, values and experiences maybe a better alternative to understand wellbeing. Unfortunately, our study was not able to gauge participants’ in-depth understanding of acculturation and family/community cohesion.

Study limitations and future work

A major limitation is the potential for selection bias of study participants that may have led to the high proportion of participants indicating their acculturation status as being predominantly ‘marginalised’, and lower education background. Also, there is the potential for participation bias based on the limited use of the mobile/electronic platform of the intervention tool and due to the duration of the mHealth trial (12 weeks), that may have been too short to be able to measure the wellbeing factors at a comprehensive level. Finally, to better understand wellbeing from a Pasifika perspective, further research will be needed to include other domains outside of established health models, including the role of family and community.

Conclusion

Our study utilised Pasifika-only participant data from a large cRCT21 study, to examine the relationship between demographic and behavioural factors and Pasifika wellbeing. From the cRCT findings (Table 2), the programme appears to have supported positive changes, particularly for the intervention participants in ‘family and community’ wellbeing, compared with the controls. Additionally, it was clear from our multivariate analyses that at an individual level, the study participants who identified as being ‘marginalised’ had significantly positive associations with family/community wellbeing. Although the study findings do not fully explain the reasons behind the acculturation, education and age characteristics associations, it does point to the importance of ‘family/community’ as being the most important wellbeing factor for Pasifika peoples. Future work could focus on more in-depth understanding of the psychosocial factors and an up-to-date knowledge of intra-familial and inter-generational perception of acculturation, and its effect on overall wellbeing.

Appendix

Appendix Table 1: Wellbeing questions: Pasifika version.

Pacific and Kiwi/New Zealand heritage and lifestyle

Appendix Table 2:

Summary

Abstract

Aim

The primary objective of this study was to determine the effect of a mobile health (mHealth) intervention on the wellbeing of Pasifika peoples, and to explore factors associated with Pasifika wellbeing.

Method

The OL@-OR@ mHealth programme was a co-designed smartphone app. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The concept of wellbeing was examined as part of a two-arm, cluster randomised trial, using only the Pasifika data: 389 (of 726) Pasifika adults were randomised to receive the mHealth intervention, while 405 (of 725) Pasifika adults were randomised to receive a control version of the intervention. Culturally relevant data was collected to examine holistic health and wellbeing status, at baseline, and at 12 weeks (end of the trial). The intervention effects and the association of demographic and behavioural relationships with wellbeing, was examined using logistic regression analyses.

Results

Relative to baseline, there were significant differences between the intervention and control groups for the ‘family/community’ wellbeing, at the end of the 12-week trial. There were no significant differences observed for all other wellbeing domains for both groups. Based on our multivariate regression analyses, education and acculturation (assimilation and marginalisation) were identified as positively strong factors associated to Pasifika ‘family and community’ wellbeing.

Conclusion

Our study provides new insights on how Pasifika peoples’ characteristics and behaviours align to wellbeing. Our findings point to ‘family and community’ as being the most important wellbeing factor for Pasifika peoples.

Author Information

Ridvan Firestone, Senior Research Officer/Co-Principal Investigator, Centre for Public Health Research, Massey University, Wellington; Soo Cheng, Biostatistician, Centre for Public Health Research, Massey University, Wellington; Sally Dalhousie, Community Partner, The Fono, Auckland; Emily Hughes, Community Coordinator, The Fono, Auckland; Tevita Funaki, Chief Executive, Community Partner, The Fono, Auckland; Akarere Henry, Chief Executive, Community Partner, South Waikato Pacific Islands Community Services, Tokoroa; Mereaumate Vano, Community Coordinator, South Waikato Pacific Islands Community Services, Tokoroa; Jacqui Grey, Project Manager, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Jodie Schumacher, Project Coordinator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; Andrew Jull, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland; School of Nursing, University of Auckland, Auckland; Robyn Whittaker, Associate Professor, School of Nursing, University of Auckland, Auckland; Lisa Te Morenga, Senior Lecturer/Co-Principal Investigator, School of Health, Faculty of Health, Victoria University of Wellington, Wellington; Cliona Ni Mhurchu, Professor/Co-Principal Investigator, National Institute for Health Innovation, School of Population Health, University of Auckland, Auckland.

Acknowledgements

We are indebted to the Pasifika communities and their members for participating in this study. We also acknowledge the efforts of everyone that were involved in the overall cRCT. This includes the Māori community partner and their participants from Toi Tangata.

Correspondence

Dr Ridvan Tupai-Firestone, Centre for Public Health Research, Massey University, PO Box 756, Wellington.

Correspondence Email

r.t.firestone@massey.ac.nz

Competing Interests

Nil.

1. Ministry of Health. Long term conditions Wellington: Ministry of Health; 2016 [http://www.health.govt.nz/our-work/diseases-and-conditions/long-term-conditions].

2. OECD Health Data. Obesity and the Economics of Prevention: Fit not Fat. 2011.

3. Ministry of Health. Health Loss in New Zealand: A report from the New Zealand Burden of Diseases, Injuries and Risk Factors Study, 2006–2016. Wellington: Ministry of Health; 2013.

4. Ministry of Health. National Diabetes Work Programme 2014/15. Wellington: Ministry of Health; 2014.

5. Ministry of Health. Annual Update of Key Results 2014/15: New Zealand Health Survey Wellington: Ministry of Health; 2015 [http://www.health.govt.nz/publication/annual-update-key-results-2014-15-new-zealand-health-survey].

6. National Health Committee. Meeting the needs of people with chronic conditions: Hapai te whanau mo ake ake tonu. Wellington: Ministry of Health; 2007.

7. Statistics New Zealand. 2013 Census QuickStats about culture and identity. Wellington: Statistics New Zealand; 2014.

8. Ministry of Health. Annual Update of Key Results 2016/17: New Zealand Health Survey. Wellington: Ministry of Health; 2017.

9. Eyles H, Jull A, Dobson R, Firestone R, Whittaker R, Te Morenga L, et al. Co-design of mHealth Delivered Interventions: A Systematic Review to Assess Key Methods and Processes. Current Nutrition Reports. 2016:1–8.

10. Firestone R FT, Dalhousie S, Henry A, Vano M, Grey J, Jull A, Whittaker R, Te Morenga L, Ni Mhurchu C,. Identifying and overcoming barriers to healthier lives. Pacific Health Dialog. 2018; 21(2):54–66.

11. Wilson D. New perspectives on health disparities and obesity interventions in youth. J Pediatric Psychology. 2009; 34(3):231–44.

12. Counties Manukau District Health Board. LotuMoui Health Programme. Auckland: Counties Manukau District Health Board; 2010.

13. Scragg R, et al. Obesity in Pacific Island Communities. University of Auckland: Health Research Council of New Zealand; 2002.

14. World Health Organisation. mHealth: New horizons for health through mobile technologies. Geneva: World Health Organisation; 2011.

15. Ghorai K AS, Khatun F, et al. mHealth for Smoking Cessation Programs: A Systematic Review. J Pers Med. 2014; 4(3):412–23.

16. Whittaker R MH, Bullen C, et al. Mobile phone-based interventions for smoking cessation. Cochrane Database Syst Rev. 2016; 4(CD006611).

17. Verbiest M SB, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Hughes E, Humpfrey G, Jiang Y, Jull A, Pekepo C, Schumacher J, Te Morenga L, Tunks M, Vano M, Whittaker R, Ni Mhurchu C. A Codesigned, Culturally-Tailored mHealth Tool to Support Healthy Lifestyles in Māori and Pasifika Communities in New Zealand: Protocol for a Cluster Randomized Controlled Trial. JMIR Research Protocols 2018; 7(8):e10789.

18. Statistics New Zealand. QuickStats about Pacific peoples: 2006 Census. Wellington: Statistics New Zealand; 2010.

19. Verbiest M CC, Dalhousie S, Firestone R, Funaki T, Goodwin D, Grey J, Henry A, Humpfrey G, Jull A, Vano M, Pekepo C, Te Morenga L, Whittaker R, Ni Mhurchu C,. Using co-design to develop a culturally-tailored, behavior change mHealth intervention for indigenous and other priority communities: a case study in New Zealand. Transl Behav Med. 2018; 8:1–17.

20. Pulotu-Endemann K. Fonofale Model of Health. Auckland: Health Promotion; 2001.

21. Ni Mhurchu C, Te Morenga L, Tupai-Firestone R, Grey J, Jiang Y, Jull A, et al. A co-designed mHealth programme to support healthy lifestyles in Māori and Pasifika peoples in New Zealand (OL@-OR@): a cluster-randomised controlled trial. The Lancet Digital Health. 2019; 1(6):e298–e307.

22. Firestone R TH, Manukia M, Kaholokula K, Foliaki S, Kingi TK, Kruger R, Breier B, O’Connell A, Borman B, Ellison-Loschmann L. Understanding Pasifika youth and the obesogenic environment, Auckland & Wellington, New Zealand. NZ Med J. 2016; 129(1434).

23. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

24. Kaholokula JK, Nacapoy AH, Grandinetti A, Chang HK. Association between acculturation modes and type 2 diabetes among Native Hawaiians. Diabetes Care. 2008; 31(4):698–700.

25. Rothe E, Tzuang D, Pumariega AJ. Acculturation, Development, and Adaptation Child Adolesc Psychiatric Clin N Am 2010; 19:681–96.

26. Bedford R. Pacific Islanders in New Zealand. Espace Populations Sociétés. 1994:187–200.

27. Tiatia J. Pasifika Cultural Competencies: A literature review. Wellington: Ministry of Health; 2008.

28. Birman D. Acculturation and human diversity in a multicultural society. In: Trickett E, Watts R, Birman D, editors. Human diversity: Perspectives on people in context San Fancisco: JossevBass; 1994:261–84.

29. Berry JW. Conceptual approaches to acculturation. In: Chun KM, Organista PM, Mar´ın G, editors. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

30. Chun KM, Organista PB, Mar´ın G. Acculturation: Advances in theory, measurement, and applied research. Washington DC: American Psychological Association; 2003.

31. Cespedes YM, Huey SJ Jr. Depression in Latino adolescents: a cultural discrepancy perspective. Cultur Divers Ethnic Minor Psychol 2008; 14(2):168–72.

32. Rogler LH. International migrations: a framework for directing research. Am Psychol 1994; 49:701–8.

33. Saldana DH. Acculturative stress and minority status. Journal of Behavioral Health Sciences 1994; 16:117–25.

Contact diana@nzma.org.nz
for the PDF of this article

Subscriber Content

The full contents of this pages only available to subscribers.

LOGINSUBSCRIBE