2.1 Design

Given the paucity of studies exploring the obesity concerns of British Pakistani women and the known barriers in engaging seldom heard ethnic groups in research, we used a feminist participatory action research³¹ design to comprehensively explore women's obesity concerns. This approach allows for the identification of women's obesity concerns from their description of lived experience and lessens the researcher impact and imbalance of power inherent in the researcher and participant relationship.³² It should be noted that the corresponding author, H. I., shares similar characteristics to the participants in that she is a Muslim, Pakistani woman also living in Bradford. The decision to undertake focus groups was informed by support for this method from public and patient involvement in the project. Before this study, public contributors who reflected the target population reported that Pakistani women may engage more with a focus group with others that shared the same characteristics as them, rather than one-to-one interviews. Focus groups are particularly useful in generating data from marginalized groups who may feel anxious about participating in research,³³ with ethnically homogeneous focus groups appearing to be more at ease discussing ethnic and cultural differences than more heterogeneous ones.³⁴ Research prioritization involving South Asian communities have solely utilized the focus group method, noting its effectiveness for eliciting data from this particular population.^35-37 Ethical approval was granted from the University of Bradford (Reference E722).

2.2 Sampling and recruitment

We used purposive sampling by location (deprived areas of Bradford), ethnicity (Pakistani) and religion (Islam) to recruit both English-speaking and non-English-speaking adult women (18+ years). Muslim, Pakistani women specifically were included because they have distinct cultural and religious needs that prevent them from living a healthy lifestyle, which is cited as one reason for failed interventions in this group.³⁸ Three community and schools settings were targeted in which to run three focus groups, which is the recommended number for data to be collected sufficiently to facilitate the emergence of patterns and themes across the focus groups.³⁹ All three settings were located within the most deprived inner-city wards of Bradford across most domains of deprivation (income, employment, education skills and training, health deprivation and disability, crime, barriers to housing and services and living environment).⁴⁰ Eligible women in two of the focus groups were existing members of scheduled, weekly meetings. One was a parents' group held at a school (n = 9) and the other at a community organization (n = 9) that actively work with Black and minority ethnic groups. These groups were also specifically targeted because the women had pre-existing relationships, which are found to be a key factor determining participation in focus groups.⁴¹ Initial contact was made with gatekeepers, both of whom were group leaders, at both settings who assisted with recruitment of women who fulfil the eligibility criteria. Pre-existing positive relationships between the gatekeeper and participants can be used by a researcher to facilitate the social access to potential participants.⁴² Information sheets and consent forms were given to all Pakistani women who expressed an interest in participating. For non-English-speaking women, the gatekeepers assisted in translating the study aims into Urdu, along with the information sheet and consent forms. The third focus group was held at an Islamic faith school (n = 5), wherein nearly all students were of Pakistani heritage. This focus group included women who had no prior relationship, and yet, their children were students at the school and the women regularly attended the monthly school assembly. The deputy head teacher at the school was approached and allowed the main author a 15-min segment at the end of assembly to discuss the study and recruit women who fit the eligibility criteria. Settings in which the target group normally congregate should be considered by the researcher,⁴³ and will increase the likelihood of participation by seldom heard communities.⁴⁴ We aimed to recruit our sample to ensure between four and eight participants, as is recommended.⁴⁵ Women provided informed consent at the start and end of the focus group to ensure that they were still happy for the inclusion of their data.

2.3 Data collection

All three focus groups were audio-recorded and conducted by H. I. in June and July 2019, lasting 1 h to 90 min each. A semistructured interview guide was developed. During its development, two other guides exploring research priorities in focus groups with South Asian communities were examined to determine the most appropriate terminology to use.^{36, 37} Public contributors also assisted in the wording of the guide. The guide was pilot-tested and iteratively refined to make it more comprehendible to women with various literacy levels. The interview guide is presented in Table 1.

2.4 Data analysis

Focus group data were transcribed verbatim, with participant information being anonymized to protect confidentiality. Transcriptions were verified for accuracy by participants who attended the existing group sessions, and to ensure that the key issues from the data were a true reflection of what was discussed during the focus groups. The transcribed data were analysed using inductive thematic analysis, and semantic codes and themes were identified in the data to reduce the risk of making assumptions from the data.⁴⁶ The transcribed data were managed using Nvivo® qualitative software (version 12.0). Themes were discussed and developed between H. I. and M. C. during group meetings. Member checking with participants from two of the focus groups (n = 15) was done to further enhance internal validity.⁴⁷

3.1 Distrust

3.2 Cultural barriers

3.3 Language barriers

4.1 Summary of key findings

We undertook this study to identify the obesity health concerns of Pakistani women living in deprived inner-city areas of Bradford. To our knowledge, this is the first study to identify and present the obesity concerns of British Pakistani women living in deprived areas. Women expressed distrust as a concern, which included the lack of trust in information provided about medication by professionals, mistrust in contradictory information around diet and a lack of trust in women-only organized physical activity. The theme of cultural barriers revealed that the obesity concerns of Pakistani women centred around the gender role of the Pakistani woman, the lack of culturally appropriate dietary advice, the lack of culturally suitable exercise facilities and pressure to conform to family and community expectations. For non-English-speaking participants, obesity concerns were the prevalence of language barriers, which meant that they were unable to understand weight management advice, the inability to read weight management advice written in Urdu, having to rely on their social networks for dietary advice and the lack of language-appropriate weight management services.

4.2 Comparison to other research, and implications for practice and research

These findings confirm and build upon prior research. Like our study, access to health promotion has been found to be hindered by the failure of health providers to offer practical diet advice that is culturally inappropriate.⁴⁸ It is noted elsewhere that access to a linguistically and culturally familiar dietician is important to South Asian women.^{49, 50} Similar to our study, language and cultural needs of Black and minority ethnic communities act as major barriers for dieticians to provide care for these communities.⁵¹ However, no previous research could be found around professionals relaying contradictory dietary advice. Research has examined contradictory messages around nutrition; yet, no research was identified in the context of conflicting dietary advice delivered by health professionals, but in terms of mixed messages in the media.⁵² The women in the current study stated that they felt that the contradictory diet advice given by professionals was due to the lack of cultural awareness of foods consumed in the Pakistani diet among these professionals. It is thus important that health practitioners such as GPs and dieticians are trained to be aware of cultural customs and traditions surrounding the food and eating practices of this population so that culturally appropriate advice can be given, which can provide value for this community.

Pakistani women in this study reported that they do not have time to engage in health-promoting activities due to their gender role in the family. Family life is considered essential and sacred within Islam, and women have been assigned the role of taking care of the family.²² As such, the interests and needs of the family are put first and exercise beyond housework is considered selfish by some South Asian women²² as well as by their families and communities.^{53, 54} We found that gender norms impeded opportunities for Pakistani women to engage in physical activity. Our findings support other studies examining the health behaviours of Pakistani women in relation to gender norms in the family.^53-55 An example of one such barrier found in both our study and the wider literature is the cultural expectation for women to remain indoors, resulting in infrequent outdoor physical activity.^{53, 56, 57} Future support could focus on helping women who do not leave their homes. One way to do this could be to educate them on how they can be more active doing home-based activities as is outlined by the NHS and Public Health England advice. An example of this is the 2020 Active at Home booklet launched by Public Health England and Sport England for older adults,⁵⁸ which may prove very useful in increasing physical activity levels in Pakistani women. In addition to promoting exercise at home for these women, wider weight management services need to consider women's home and local contexts. The development of services delivered at the grassroots level may be more impactful than weight management programmes currently being promoted through hospitals or other appointments that women are unlikely to attend due to the reasons identified in this study. These grassroots services could be home-based services or services that women could attend within frequently visited settings such as faith settings, schools, local community groups and local pharmacies. Faith settings may be particularly useful as a delivery channel for obesity interventions⁵⁹ as evidence suggests that health promotion interventions in these settings have a high reach.⁶⁰

Women in the present study indicated that they lacked knowledge of commonly consumed Asian foods that were high in fat and sugar. This is in line with other research that has found mixed knowledge in South Asians of high fat and high sugar contents of traditional South Asian foods.^{61, 62} Similarly, women in the present study felt that consumption of traditional high-fat and sugary foods was important for good health and should be regularly consumed. There is a paucity of literature on the misconceptions of this population on the benefit of these foods to their health. In addition to the lack of studies on dietary misconceptions, misconceptions surrounding physical activity in this population have also not been widely explored in the literature. It is recommended that not only should education efforts be targeted towards addressing misconceptions surrounding physical activity and diet in this population but also that educational interventions for this group are culturally appropriate to ensure that they have a greater impact. Examples of good practice can be found in the literature, in which educational interventions targeting health promotion in South Asians resulted in positive learning, highlighting the success of the intervention.⁶³ Efforts included delivering education in settings that were culturally appropriate to the attendees such as community centres, mosques and schools. In addition, embedding health ambassadors in schools, and recruiting local professionals as education champions proved effective.⁶³ It is crucial to note, however, that South Asians are not a homogeneous group.⁶⁴ Pakistani women living in deprived areas of Bradford may differ from other South Asian subpopulations. As such, interventions to address obesity in Pakistani women may need to be specifically tailored for better results.

This study identified that language and literacy problems were an issue for some women, and research has identified that health promotion materials are not always written appropriately for those with low health literacy.⁶⁵ Most individuals with low health literacy are socially deprived, have a limiting health condition or disability, lower levels of education and are from Black and minority ethnic communities.⁶⁶ Researchers and policy makers must consider literacy levels when designing obesity management information materials for Pakistani women living in deprived areas, so that those receiving the education find them appropriate to their language skills.

4.3 Study strengths and limitations

This study has several strengths. The women ranged in age and marital status, some were mothers whereas others did not have children. There were differences in their proficiency of spoken English. The sample included both English-speaking and non-English-speaking women, whereas often, non-English speakers are omitted from research due to issues surrounding language barriers.⁶⁷ Such diversity in the sample meant that a range of different perspectives were captured based on women's experiences, which increased the representativeness of this study and enriched the data. A further strength of this study lies in its reliability. The reliability of the findings was increased in a range of ways in this study. Gatekeepers translated certain words or phrases that H. I. was unaware of, despite her ability to speak Mirpuri and Urdu. Other participants also assisted in interpretation. In this respect, using existing groups in this study proved very beneficial as it was considerably easier to reconvene with these groups. The researcher returned to the participants in two of the focus groups on two separate occasions for verification: first, to verify the transcripts, and then again, to verify the relationship between the themes and codes. By doing so, the risk of misinterpreting the findings was minimized, thus further increasing the reliability of the findings.⁶⁸

This study is not without its limitations. It could be argued that selection bias may have featured in the recruitment of the sample as it is typically highly engaged individuals who engage in research, thus making them unrepresentative of the population.⁶⁹ Despite this, similar findings were reported across all focus groups, adding to the transferability of this study to Pakistani women living in deprived inner-city areas of Bradford. A further possible limitation of this study could be the risk of bias as a result of the insider status of H. I., who shared similar characteristics to participants. It is plausible that H. I. could have led the discussion based on her pre-existing knowledge of the population under study. However, the risk of this was minimized by H. I. being extra vigilant about her line of questioning and allowing the conversations to flow naturally with her limited input. This may have thus increased the credibility of this study. Lastly, it is possible that women may have omitted details under the assumption that H. I. may have insider knowledge. However, this risk was minimized by probing and asking for clarity and explicitness when women were being vague. It must also be noted that the findings from this study may not be generalizable to other Pakistani communities elsewhere or other South Asian communities more generally.