Systematic mapping review of the factors influencing dietary behaviour in ethnic minority groups living in Europe: a DEDIPAC study

Background Europe has a growing population of ethnic minority groups whose dietary behaviours are potentially of public health concern. To promote healthier diets, the factors driving dietary behaviours need to be understood. This review mapped the broad range of factors influencing dietary behaviour among ethnic minority groups living in Europe, in order to identify research gaps in the literature to guide future research. Methods A systematic mapping review was conducted (protocol registered with PROSPERO 2014: CRD42014013549). Nine databases were searched for quantitative and qualitative primary research published between 1999 and 2014. Ethnic minority groups were defined as immigrants/populations of immigrant background from low and middle income countries, population groups from former Eastern Bloc countries and minority indigenous populations. In synthesizing the findings, all factors were sorted and structured into emerging clusters according to how they were seen to relate to each other. Results Thirty-seven of 2965 studies met the inclusion criteria (n = 18 quantitative; n = 19 qualitative). Most studies were conducted in Northern Europe and were limited to specific European countries, and focused on a selected number of ethnic minority groups, predominantly among populations of South Asian origin. The 63 factors influencing dietary behaviour that emerged were sorted into seven clusters: social and cultural environment (16 factors), food beliefs and perceptions (11 factors), psychosocial (9 factors), social and material resources (5 factors), accessibility of food (10 factors), migration context (7 factors), and the body (5 factors). Conclusion This review identified a broad range of factors and clusters influencing dietary behaviour among ethnic minority groups. Gaps in the literature identified a need for researchers to explore the underlying mechanisms that shape dietary behaviours, which can be gleaned from more holistic, systems-based studies exploring relationships between factors and clusters. The dominance of studies exploring ‘differences’ between ethnic minority groups and the majority population in terms of the socio-cultural environment and food beliefs suggests a need for research exploring ‘similarities’. The evidence from this review will feed into developing a framework for the study of factors influencing dietary behaviours in ethnic minority groups in Europe. Electronic supplementary material The online version of this article (doi:10.1186/s12966-016-0412-8) contains supplementary material, which is available to authorized users.


Background
During the last few decades, migration in Europe has increased and many immigrant-origin groups have been reported to have a higher prevalence of diet-related non-communicable diseases (NCDs) and poorer dietary habits than the native born European populations [1,2]. In addition, Europe has a number of indigenous minority groups such as the Sami and the Roma, who have historically suffered from discrimination and marginalization accompanied by diet-related NCDs [3]. Given the rise in ethnic minority groups and the high prevalence of NCDs among these populations [4], a clear understanding of factors influencing dietary behaviour is warranted in order to assess the needs of these populations and to develop effective public health interventions that also reach ethnic minority groups.
Dietary behaviours have been found to vary widely among and within different ethnic minority groups compared to host populations [5], indicating that factors influencing dietary behaviour may differ in ethnic minority populations as compared to the majority population [6]. However, most studies have focused either on a selected number of ethnic minority groups or are limited to specific European countries [7,8]. For instance, research in the UK has focused on South Asians [9] and African Caribbeans [7] and in the Netherlands the focus has been on Surinamese [10]. This emphasis is also reflected in reviews of dietary behaviours; a 2008 review focused specifically on dietary change among the largest ethnic minority groups in Europe [11], whilst two reviews concentrated only on ethnic minority groups in the UK [5] and France [12]. Indigenous minorities have not been included in any reviews. Thus there is a lack of insight into the broad range of factors influencing dietary behaviour among a wide range of ethnic minority groups in Europe. In addition, there has been little attempt to study factors influencing dietary behaviour in a holistic way. In the obesity foresight map [13] for instance, the complexity of factors driving dietary behaviours are illustrated, but this was not prepared through the lens of ethnic minority populations.
This review fills these gaps by systematically reviewing primary studies on a wide range of ethnic minority groups and by considering a variety of dietary behaviours over the whole life course, using a holistic and data driven approach, by clustering emerging factors across these groups. The aims of this review were to identify the broad range of factors influencing dietary behaviour among ethnic minority groups living in Europe in order to identify gaps in the literature to guide future research. The evidence from this review will also feed into developing a framework for the study of factors influencing dietary behaviours in ethnic minority groups in Europe [14], as part of the work of the DEDIPAC-KH (DEterminants of DIet and Physical Activity Knowledge Hub) [15] for European populations.

Review typology
A systematic mapping review [16] of the factors influencing dietary behaviour among ethnic minority groups in Europe was conducted. A mapping review was selected because it allows the mapping and categorisation of existing literature and identification of the gaps in research literature [16]. To avoid research bias during the review process, the review protocol was registered with PROSPERO (PROSPERO 2014: CRD42014013549) before commencing.

Search strategy
An initial scoping search was undertaken through PubMed PubReMiner [17] with the aim of assessing the amount of available literature and identifying appropriate search terms to be used in the main searches. A search strategy was constructed in consultation with information specialists from the University of Sheffield and the Academic Medical Centre, Amsterdam. The search strategy was based on search terms within three concepts: (i) dietary behaviours and its synonyms: diet, food habits, nutritional status, food preferences and nutrition; (ii) ethnic minority groups; and (iii) all countries that are listed by the World Bank as low and middle income countries. In addition, countries from the former Eastern European Bloc [18] from where groups commonly migrate to other parts of Europe, were included so that the review captured all ethnic minority groups living in Europe including indigenous populations. Other search terms that were used to capture potential studies were: emigrants, immigrants, cultural diversity, minority groups, migrants, ethnic groups, multiculturalism, ethnic minority, BME (Black and Minority Ethnic), black, minority ethnic, asylum seeker, refugee, non-white, coloured population or black; and (iii) Europe, all European countries by name. The search strategy contained free text and subject headings.
The following nine electronic databases were searched: MEDLINE, EMBASE, Web of Science, Cochrane Library, CINAHL, ProQuest, Psychinfo, ASSIA, and Campbell Collaboration Library of Systematic Reviews. The search strategy was modified where necessary for use in different electronic databases. The complete MEDLINE search strategy is shown in Additional file 1.
Databases were searched from 1999 to 2014, as it was expected that any factor identified before 1999 would also be referred to in more recent literature. Spot checks on results from the scoping review indicated that key papers emerged after 1999. Searches were conducted between May and July 2014. The citation follow-up technique and contacting of experts in the field was undertaken to identify additional relevant articles. In addition, the reference lists of all included articles were scanned for articles that met the inclusion criteria. All citations were downloaded into an Endnote web library and duplicates were removed.
Ethnic minority population is a concept used for very heterogeneous groups that may share minority status in their country of residence due to ethnicity, place of birth, language, religion, citizenship as well as other cultural differences [19]. This definition may include groups from newly arrived immigrants to (minority) groups that have been part of a country's history, for instance the Sami people.

Inclusion and exclusion criteria
Observational and intervention studies, using quantitative, qualitative or mixed methods that examine dietary behaviour among ethnic minority groups in Europe were included. Other studies that focused on nutrition-related conditions, for example, obesity and contained relevant data on dietary behaviour in ethnic minority groups were also included. All studies that identified an association between a factor (including correlates, predictors, moderators, determinants and mediators) and dietary behaviour of minority groups living in Europe were retained.
All primary studies that analysed diet as a confounder in a relationship between ethnicity and disease were excluded. As were studies that explored whether ethnicity is a determinant of diet and did not attempt to explain why, non-human studies/laboratory based studies and studies examining beliefs and practices around breastfeeding and weaning. Studies examining the nutrient status of particular groups without mention of diet and studies presenting descriptive information about diet were also excluded.

Study selection
The title and abstracts of a total of 2965 articles identified references were imported into Endnote and 730 duplicates removed. The remaining 2235 articles were equally divided between five independent reviewers (HO, MN, KP, MH and LT) against the inclusion criteria. Of these abstracts, 1956 articles did not meet the inclusion criteria. The main reasons for excluding studies were because they contained no empirical data on ethnic minority groups, presented only descriptive information on diet or where outside the review time limit. Full text articles were retrieved by the five reviewers for the remaining 279 articles and the inclusion/exclusion criteria were applied. This yielded 68 potentially relevant papers for data extraction. Spot checks were conducted on a sample of 10 % of the excluded papers to assess the extent of agreement between reviewers. During the spot checks, there was a good degree of concordance. There was disagreement between two reviewers on two papers, therefore a third reviewer in the team was consulted. The outcome in both cases was to exclude the papers. The most common reason for excluding studies from this review during the data extraction stage was because they were focused on describing dietary differences between populations without examining the factors driving dietary behaviour.
During the data extraction process 37 studies met the inclusion criteria (Fig. 1).

Data extraction and synthesis
Data extraction was performed by the five reviewers according to the following study characteristics: study design, sampling population, sample characteristics, number of participants, country where the study was conducted, sampling method, dietary behaviour measured, factors reported to influence dietary behaviour.
As the main aim of this mapping review was to gather a broad base of evidence to map the conceptual domain of factors potentially influencing dietary behaviours it was decided to include all factors reported by authors and not to restrict to those factors where a statistical relationship had been demonstrated. For quantitative studies, types of analysis were also extracted, whilst direct quotes were extracted from qualitative studies.

Quality assessment
Quality assessment of quantitative and qualitative studies was undertaken using the standard quality assessment criteria for evaluating primary research papers [20]. Each reviewer in the process of extracting data assessed the paper for quality as well. To ensure consistency in quality assessment, an independent reviewer (LM) cross-checked the quality assessment between the five reviewers. Additional files (Additional files 2 and Additional file 3) summarise the quality assessment scores given to the included studies.

Data analysis and emerging clusters
Two stages were used in the analysis. In the first stage, all factors influencing dietary behaviour identified in the selected papers were extracted and their association described with the different dietary behaviours observed. The second stage involved sorting and structuring the factors into clusters, during which, the list of factors emerging from the review were clustered according to how they were seen to relate to each other, in a data driven approach [21]. This clustering step was part of a larger concept mapping process leading to the development of a systems-based framework of factors influencing dietary behaviour and physical activity/sedentary behaviour in ethnic minority groups living in Europe. The use of the concept mapping methodology [21] is explored in detail in a separate paper [14]. The concept mapping approach is guided by systems thinking, which can simply be defined as "looking at things in terms of the bigger picture" [22]. This concept mapping approach included steps on generating a list of factors, and sorting and structuring these factors into clusters by consensus based on how they relate to each other [19].
Within the DEDIPAC-KH, an inter-disciplinary group focused on the determinants of dietary and physical activity behaviours. This comprised a team specifically focussing on the factors influencing the behaviours of ethnic minority populations ('DEDIPAC ethnic minority team'). Other parts of the DEDIPAC-KH focussed on the factors influencing dietary [23,24] (Condello G, Ling FCM, Bianco A, Chastin S, Cardon G, Ciarapica D, et al. Using Concept Mapping in the Development of the EU-PAD Framework (European Physical Activity Determinants across the Life Course): a DEDIPAC-Study. Submitted) and physical activity behaviours of the general European population ('DEDIPAC general population team').
All 63 factors were grouped into seven clusters (Table 1) in two steps -firstly by a meeting of n = 5 members of the DEDIPAC ethnic minority team, followed by a confirmatory workshop of n = 21 members of DEDIPAC general population team. Experts representing a range of disciplines participated in the clustering process: nutrition, public health, epidemiology, anthropology, social demography, economics, sociology, dietetics, psychology, exercise physiology, health promotion and physical activity.

Description of included studies
The characteristics of the 37 included studies are presented in Tables 2 and 3 (19 qualitative; 18 quantitative). Most of the studies were conducted in Northern Europe, i.e.,(UK n = 9, Norway n = 9, The Netherlands n = 6 and Sweden n = 5). The most commonly studied ethnic minority groups were Pakistanis and Bangladeshis. Most of the studies were conducted among adults (n = 31), but studies on children/adolescents (n = 5) and on older adults (n = 1) were found. However, 11 of the studies conducted on adults included participants who were older adults and 3 other studies included children/adolescents. The number of participants ranged from 14 to 83 in the qualitative studies and 100-12,811 in the quantitative studies.
The 12 dietary behaviours (as stated by authors) examined in the included studies (Tables 2 and 3) were food intake, fruit and vegetable intake, changes in food habits, intention to change diet, product purchase, meal preparation, dietary acculturation, dietary intake (healthy/unhealthy intake), eating habits, food neophobia, dieting and diet quality. Dietary behaviour as used in this review encompasses all food related behaviours that are grouped into three main categories [23] 1. Food choice, consists of outcomes preceding the actual consumption (e.g., produce purchase and intentions); 2. Eating behaviour, comprising outcomes to do with the actual act of eating (e.g., dieting and food neophobia); and 3. Dietary intake consisting of all outcomes related to what is consumed (e.g., fruits and vegetable intake and healthy versus unhealthy.

Factors influencing dietary behaviours
There were some commonalities and differences in factors influencing dietary behaviour across different ethnic minority groups. For instance, lack of availability of     Taste preference was another frequently reported factor, irrespective of the population or setting of the study. Factors reported by specific populations included fluency in the host language identified in a study conducted among diabetic Pakistani born persons living in Oslo [25] and traditional lifestyle as deduced from occupation, e.g., reindeer herder reported in both papers addressing diet among the Sami and non-Sami groups in Norway (41,47). Also a number of studies found differential associations between socioeconomic status (SES) and dietary behaviour [9,10,[26][27][28][29]. For instance, SES was not associated with dietary behaviour among populations from East Asia, Indians and African origin in Norway [26], or in a Surinamese population in the Netherlands [10], whilst in a study conducted in the UK, SES influenced dietary intake among South Asians from Bangladesh. All factors are shown in Table 4.

Discussion
Europe has a growing population of ethnic minority groups whose dietary behaviours are potentially of public health concern and the factors driving these behaviours need to be understood. This review identified a broad range of factors and clusters influencing dietary behaviour and identified gaps in the literature to guide future research. The evidence from this review will feed into developing a framework for the study of factors influencing dietary behaviours in ethnic minority populations in Europe. This review extracted 63 individual factors that were grouped in seven clusters. Two clusters, 'social and cultural environment' and 'food beliefs and perceptions' had the highest number of factors shown to shape dietary behaviours of ethnic minority populations. These findings corroborate those of earlier reviews on ethnic minority populations [11,5], in the sense that like other reviews, most factors identified are related to the sociocultural environment, cultural beliefs and perceptions around food. In our review, factors are clustered in a way that cut across the more traditionally used socioecological levels (individual, family, community and society, see for example [51,52]). The socio-ecological model presents different layers of influence, with an underlying assumption that the layers operate in linearity. Thus community factors are presumed to influence the individual via the family, whereas certain community factors may directly influence dietary choices, bypassing the family. In addition, the socio-ecological model depicts reality as artificially separating individual and social experiences. Clustering factors into systems may provide a more adequate means of depicting interrelationships between the factors.  For example, in our analysis, social norms and identity were clustered together but would have been classified as 'community' and 'individual' factors using the socioecological model. By analysing data in this way, we aimed to explore the underlying mechanisms that shape dietary behaviours, in a holistic, systems-based approach. This is in line with recent work that seeks to understand dietary choice as a social practice [53]. Our review identified some factors that have also been shown to influence dietary behaviours of majority populations. These include food price, income [54], social networks [54], time constraints and food availability [55], although these factors are often reported in low income groups [54]. However, many of the factors known to influence diet among majority populations were not identified in our review. For instance, a recent umbrella review among adults [56] of studies conducted in Europe, Australia and North America, identified political environments, food advertising, late-shift work, behavioural regulation and sedentary behaviour as important correlates of dietary behaviour but these factors were absent in our review. This might be due to a bias in studies amongst ethnic minority groups, as there is a tendency by researchers to focus on sociocultural factors. Factors that were identified as unique to immigrant origin groups included all factors within the 'migration context' cluster, and some from other clusters, e.g.; level of acculturation, cultural identity, availability of traditional foods, familiarity with host country foods, competency in host language, perception of host culture and religious prescriptions. While the majority of studies explored 'differences' across ethnic groups, this review found a need for research exploring 'similarities'. This would be useful in adapting mainstream interventions for ethnic minority groups.
It is important to note that although dietary behaviour is generally considered to be associated with SES [54], in  [26] reported that SES using a composite measure of parental occupation, mother's education, employment status and social security status was not associated with diet quality across the study sample of South Asian and African adolescents. In contrast, in other studies [27][28][29], education was identified to be a determinant of selected food intakes, including fruit and vegetables among elderly Iranian-born residents of Stockholm. The inconsistencies in the relationship between SES and diet corroborates previous studies that have observed differences between SES and metabolic outcomes in different migrant populations [57]. It has been hypothesised that these groups might be in another stage of epidemiological transition, where diet-related NCDs (and risk factors) are still more prevalent (or equally so) in higher SES groups, as was the case among European populations in the 1950s and 60s. Almost half the studies (44 %) included in this review were focused on South Asians, which is not surprising given that this group forms the largest ethnic minority group in some countries in northern Europe, where most of the studies were conducted [11]. The 'psychosocial' cluster of factors seems relevant only to the South Asian population in our review, however, this is a reflection of only one study, therefore, there is insufficient evidence to be able to conclude that these factors are only important to South Asian populations. The findings should be interpreted with caution as not all factors may be applicable to all ethnic minority groups, as they are heterogeneous populations in terms of their acculturation level, ethnicity, socio-demographic status and religion.
Factors that were included in the 'social and cultural environment' and 'food beliefs and perceptions' clusters appeared to influence dietary behaviours amongst almost all minority groups. Among studies conducted with South Asians (Pakistani, Bangladeshi, Indians) and other migrants from predominantly Muslim countries, religious beliefs and prescriptions were identified as important factors.

Strengths and limitations of the review
This is the first systematic mapping review that has mapped out factors influencing dietary behaviour among a diverse population of minority groups living in Europe. One difference between this review and others is in the method used in synthesizing the findings. Most reviews have used existing frameworks for this purpose [52,58]. The approach used in this review has resulted in clustering of factors that transcends existing models, aiming to better capture the complexity of the system of factors influencing dietary behaviour. Another strength of this review was the inclusion of indigenous groups and Eastern Europe migrants to Western Europe in the search strategy, recognising that these groups are potentially disadvantaged and marginalized and may be more vulnerable to NCDs. However, only three studies were identified that reported factors influencing dietary behaviours among the Sami [47,59,60] and minority groups from the former Eastern Bloc European countries who commonly migrate to other parts of Europe [28]. We found no studies on Roma populations.
Although migration is a phenomenon that is found in all European countries, most studies were conducted in Northern Europe, and predominantly among populations of South Asian origin. Few studies were conducted among children and older adults. Findings need to be interpreted with caution as factors might differ across age groups. Many studies excluded in the reviewing process focused on describing dietary differences and did not present findings on the factors driving behaviour, which partly accounts for the limited number of relevant studies included in the review. The inclusion of mainly cross-sectional quantitative studies in this review reflects the types of studies available, which also means that we cannot establish causal relationships between the factors identified and dietary behaviours. Whilst there was no limitation for language during the search strategy, our review consists of articles published entirely in English. This could be due to the fact that other relevant articles may not have been indexed in the electronic databases used for this review.

Implications of the findings
Future research is needed to further deepen our understanding of the interrelationships between identified factors both within and between clusters. In addition, future studies need to make direct comparisons between minority and majority populations to understand differences and commonalities in factors underlying dietary behaviours and food choice. We also recommend studies into a broader range of more 'mainstream' factors (as with the majority population). There is also a need for more studies including longitudinal data of factors influencing dietary behaviours across the life course, particularly of young people and older adults among ethnic minority groups. Finally, a gap was identified for studies comparing the drivers of dietary behaviours across a wide range of ethnic minority groups living in different contexts in Europe (including central and southern Europe) and including groups that are under-represented in national surveys such as the Roma, asylum seekers and refugees, which are increasingly relevant groups in Europe.