Adolescence (10–19 years) is a period of rapid growth and development which marks the transition from childhood to adulthood. Adolescent malnutrition, including both undernutrition and obesity, in low- and middle-income countries (LMIC) is a major public health concern. Nutritional requirements increase during adolescence to enable the growth, sexual maturation and brain re-modelling necessary(Reference Sawyer, Afifi and Bearinger1,Reference Blakemore2) . Diet and physical activity habits adopted in adolescence can influence lifelong health behaviour(Reference Sawyer, Afifi and Bearinger1). Adolescent nutrition research has been neglected in LMIC where the dual burden of malnutrition is common. Malnutrition including anaemia and stunting remains commonplace, while at the same time, urbanisation and economic and societal transitions are contributing to increasing obesity leading to later life non-communicable chronic diseases(Reference Popkin, Adair and Ng3). Optimising nutrition during adolescence has the potential to both improve the health of adolescents themselves now and in future and that of the next generation(Reference Sawyer, Afifi and Bearinger1,Reference Patton, Sawyer and Santelli4,Reference Patton, Olsson and Skirbekk5) .
Almost one-quarter of the world’s population comprises adolescents and young adults (10–24 years), with more than 80 % living in LMIC(Reference Patton, Sawyer and Santelli4). Adolescent diet and physical activity behaviours are driven by a complex interaction of factors at the individual, familial, community and environmental levels. During this time, an individual’s sphere of influence expands from their immediate family to include their peers, the community, culture, education, the media and the economy(Reference Viner, Ozer and Denny6). In addition, many LMIC are undergoing rapid urbanisation which is associated with adolescents’ engaging in obesogenic behaviours due to exposure to cheap energy-dense foods and sedentary lifestyles(Reference Popkin, Adair and Ng3).
Although it is important to understand the impact of these transitions on young people’s lives(Reference Morrow7), nutrition research with children and young people in LMIC is in its infancy. Qualitative methodologies provide an opportunity to gain insights into the lives, experiences and perspectives of adolescents living in these rapidly transitioning settings. Qualitative methods enable the complexities of everyday life to be captured and documented, providing new insights into young people’s experiences and how these shift as they move from childhood, through adolescence and into young adulthood(Reference Morrow and Crivello8). Youth-centred research views adolescents as experts of their own lives and enables the collection of rich, novel research findings whereby context is taken into account(Reference Malterud9). Given the complex nature of adolescence and the influences on their diet and physical activity behaviours, qualitative research methodologies represent a valuable method for generating insights into the key drivers behind these behaviours. Synthesising qualitative evidence allows us to aggregate findings that explore adolescent diet and physical activity behaviours across multiple contexts(Reference Bohren, Hunter and Munthe-Kaas10).
Established in 2018, Transforming Adolescent Lives through Nutrition (TALENT) is a consortium of researchers from the Europe (France, UK), India (Dervan, Mumbai, Mysore and Pune) and sub-Saharan Africa (Abidjan, Cote D’Ivoire; Jimma, Ethiopia; Kiang West, The Gambia and Soweto, South Africa) at different stages of economic and societal transition, who share a special interest in adolescent nutrition. This paper is part of a special issue (see Barker et al., this issue) focused on addressing the evidence gap in adolescent nutrition research within these settings.
Aim of this review
To review the qualitative literature exploring influences on diet and physical activity behaviours among adolescents across Côte d’Ivoire, Ethiopia, India, South Africa and The Gambia.
Specifically, this review aims to synthesise the qualitative literature exploring: (a) adolescents’ (10–19 years) perceptions of diet and physical activity behaviours; (b) body image and satisfaction/dissatisfaction in adolescents and how these may influence their diet and physical activity behaviours; and (c) perceptions of and attitudes towards healthy eating and physical activity and the barriers and/or facilitators to these behaviours in adolescents living in TALENT sites (Côte d’Ivoire, Ethiopia, India, South Africa and The Gambia).
Methods
Search strategy and selection criteria
Search terms were defined by scanning key words from relevant papers in an initial scoping search. Keywords were divided into categories using a modified version of the PICO formula. Terms were therefore divided into population, context (location and study design, separately) and outcome. Comprehensive literature searches were conducted in January 2019 by two researchers (PD and PHJ), in two stages: (1) scanning of titles and abstracts and (2) full-text reviewing of studies against the exclusion/inclusion criteria. Searches identified all available studies published between January 2000 and January 2019. The search strategy was developed based on the PICO method(Reference Santos, Pimenta and Nobre11) and search terms through meetings amongst the TALENT team and keywords identified from a scoping review of potentially relevant articles (see Table 1). The following limits were added to all searches: (1) studies conducted in India, South Africa, Ethiopia, Cote D’Ivoire and The Gambia and (2) studies that have been published between 2000 and 2019 in English.
Search terms used to identify relevant studies exploring the influences on adolescent diet and physical activity
Searches were conducted using five key electronic databases: Medline, Web of Science, EMBASE, CINAHL and PsychINFO. To supplement the electronic searches, a search of the online publication list of a collaboration, Young Lives(Reference Barnett, Ariana and Petrou12), a longitudinal, global study into children and youth was conducted. The reference lists of all included studies were also scanned to identify any potentially relevant studies. Search results were imported into EndNote (bibliographic software) where, following the PRISMA(Reference Moher, Liberati and Tetzlaff13) chart (see Fig. 1), two researchers (PD and PHJ) independently screened potentially relevant titles and abstracts. Following this, the same researchers independently screened full-text articles against the inclusion/exclusion criteria:
PRISMA flow diagram(Reference Moher, Liberati and Tetzlaff13)
Inclusion criteria
Studies were included in this review if they (1) explored adolescents’ perspectives and/or experiences of diet and/or physical activity behaviours; (2) were conducted in India, South Africa, Ethiopia, Cote D’Ivoire or The Gambia; (3) collected data from boys and/or girls aged 10–19 years; (4) were published between 2000 and January 2019, in English; (5) used a qualitative study design including observations, face-to-face interviews or focus group discussions and (6) published in peer-reviewed journals.
Studies were excluded from this review if they (1) did not use qualitative methods; (2) did not collect data from adolescents aged 10–19 years; (3) collected and analysed data from parents, teachers or other stakeholders and (4) were not available in English.
Once included studies had been identified, data were extracted from each included study on aims, participant details (age, sample size), design (method of data collection), outcomes related to diet and physical activity and conclusions. The standardised Critical Appraisal Skills Programme(14) checklist recommended by the University of York Centre for Reviews Guidelines for writing systematic reviews was used to consider the quality of studies being synthesised(15). The Critical Appraisal Skills Programme checklist contains 10 items based on the research aims, methods used, the role of the researcher, data analysis method and the clarity of the findings and value of the research (http://www.casp-uk.net/#!casp-tools-checklists/c18f8). The quality assessment was undertaken independently by two reviewers (PD, PHJ), with disagreements resolved through discussion. The outcome of each Critical Appraisal Skills Programme item is reported for each included study in Appendix A.
Qualitative synthesis
This qualitative evidence synthesis report is written using the ENTREQ statement guidelines to enhance transparency(Reference Tong, Flemming and McInnes16). The challenges associated with synthesising qualitative literature are different to those concerning quantitative publications. A qualitative synthesis is similar to qualitative research whereby the findings from other qualitative studies are pooled and synthesised. The findings from the included studies in this review have been pooled, uploaded to the qualitative analysis software NVIVO v12 and thematically analysed based on Centre for Reviews guidance(15), following Braun and Clarke’s guide(Reference Braun and Clarke17) to develop an overall synthesis of main findings(Reference Thomas and Harden18). PHJ coded the findings of the included papers deductively, focussing on findings related to adolescents’ perceptions of health, diet and physical activity behaviours. These codes were reviewed, revised and discussed amongst the research team. Once the coding framework was finalised (see except in Appendix B), codes were merged with other similar codes and eventually categorised under broad headings. These categories were then reviewed, revised and developed into three broad themes that represented the main areas of research focus in the literature so far.
Results
Description of included studies
The searches retrieved 7787 titles and abstracts to be screened, following which, seventy-nine full-text papers were screened (see Fig. 1). Twelve studies (from which thirteen papers were produced) were deemed eligible for inclusion in this review (see Table 2). It is important to note that the findings of this review are based heavily on studies conducted in just two countries, India and South Africa. Five studies were carried out in South Africa(Reference Voorend, Norris and Griffiths19–Reference Tshililo, Netshikweta and Tshitangano23), of which, there were two in urban settings(Reference Voorend, Norris and Griffiths19,Reference Morris and Szabo20) , two in rural settings(Reference Sedibe, Kahn and Edin21,Reference Kinsman, Norris and Kahn22) and one in which the setting was not reported(Reference Tshililo, Netshikweta and Tshitangano23). All five of these papers recruited girls only. Five studies were carried out in India(Reference Ahanonu and Jooste24–Reference Satija, Khandpur and Satija27) in which the settings were either not reported(Reference Ahanonu and Jooste24,Reference Satija, Khandpur and Satija27) or included participants from both rural and urban settings(Reference Gavaravarapu, Vemula and Rao25,Reference Correa, Rajaraman and Swaminathan26) . Three of the five Indian studies collected data from both boys and girls and one recruited girls only. One study presented results from data collected from both girls and boys from urban settings in both South Africa and India(Reference Mmari, Blum and Sonenstein28). Finally, one study that was conducted in rural Ethiopia with data from both boys and girls included in this review(Reference Morrow, Tafere and Chuta29). Of the included studies, data collection methods comprised focus group discussions(Reference Morris and Szabo20,Reference Kinsman, Norris and Kahn22,Reference Ahanonu and Jooste24–Reference Mmari, Blum and Sonenstein28) , in-depth duo interviews(Reference Voorend, Norris and Griffiths19,Reference Sedibe, Kahn and Edin21) , in-depth individual interviews(Reference Tshililo, Netshikweta and Tshitangano23,Reference Mmari, Blum and Sonenstein28) , community mapping(Reference Mmari, Blum and Sonenstein28), photo voice(Reference Mmari, Blum and Sonenstein28) and case study reports(Reference Morrow, Tafere and Chuta29).
Characteristics of included studies
The quality of the included studies was considered using the Critical Appraisal Skills Programme checklist and presented in Appendix A. Overall, most studies met the quality criteria outlined in the Critical Appraisal Skills Programme checklist. Most studies clearly described the aims of the research, used an appropriate methodology to address their aims and outlined the approach to recruitment and data collection in a manner appropriate to the research issue/questions. Data analysis was clearly described by most studies, with two studies lacking in some clarity and one study giving no detail about analytical methods beyond citing the name of a method. Findings were clearly and sufficiently described in all studies. A key weakness for most the studies was the lack of description regarding the role of the researchers and the relationship between researchers and participants, which was only partially achieved in two studies, and not clearly addressed in the rest of the studies. Ethical issues, such as how the research was explained to participants, and issues around informed consent and confidentiality were fully and clearly described in only five studies, with seven studies presenting insufficient or unclear information.
Narrative synthesis
The qualitative literature reviewed focused on three main themes in relation to adolescent diet and physical activity: (1) the influence of body image and self-esteem; (2) social and environmental influences and (3) poverty (see Fig. 2). See Fig. 2 for a visual display of the themes and subthemes. Table 3 depicts the themes that were expressed by adolescents from each research site.
A visual display of synthesised findings across the included studies
A table to indicate which themes were discussed by adolescents in each country
Theme 1: the influence of body image and self-esteem on diet and physical activity
Desire to be thin v. ‘curvy’
The impact of body image and self-esteem on the diet of adolescent girls was a major theme, particularly in the most economically transitioned research site, South Africa. There was a clear contrast between girls in the South Africa studies who wanted to be ‘curvy’ and those who desired a thin body shape. There was a stigma about thinness, as many of those who strived for a fuller body shape-associated thinness with stress and illness, particularly AIDS. However, there was a distinction made in one of the studies between what girls referred to as ‘sick skinny’ and ‘nice skinny’(Reference Morris and Szabo20). In contrast, many girls felt enormous pressure to be thin from external sources, including clothing stores(Reference Morris and Szabo20). There was also a perception that thinness was a requirement for being ‘popular’ at school and for obtaining a boyfriend(Reference Morris and Szabo20,Reference Kinsman, Norris and Kahn22) .
In contrast, in the least economically transitioned research setting Ethiopia, obesity was desirable as it was associated with a person who is living a good life. Weight was rarely mentioned in all but one of the studies conducted in India(Reference Gavaravarapu, Vemula and Rao25). The overweight adolescents felt that their healthy weight peers were overly and ‘unnecessarily’ concerned with ‘being skinny’. Although most of the adolescents in the studies conducted in India did not mention weight, the salience of appearance to adolescents was clear, though this largely focussed on cleanliness, hygiene and clothing(Reference Ahanonu and Jooste24).
Peers and family influencing body image
The motivation to become ‘thin’ was rarely health related. In the studies from South Africa, the social element of losing weight with friends and strong desire to avoid teasing and humiliation by peers and family heavily influenced weight ideals(Reference Voorend, Norris and Griffiths19,Reference Kinsman, Norris and Kahn22) . Cultural differences in idealised body shapes also contributed to body image concerns among black South African girls. Socialising with white South African girls, who themselves would strive for thinness, increased the perceived pressure for them to do the same. In contrast, girls from the Kinsman (2015) study felt pressured by peers to gain weight because they would be ridiculed for being thin(Reference Kinsman, Norris and Kahn22). However, although curves were desirable, girls in the current study indicated that muscle definition was not. For example, defined arms were associated with household chores which indicated poverty and exposed the girls to the possibility of further bullying.
Dysfunctional eating practices
There were multiple types of dysfunctional eating practices reported by girls in the South African studies focussing on body image including skipping meals, smoking to suppress hunger, using diet pills, drinking vinegar, over exercising and purging which included both vomiting and using laxatives(Reference Tshililo, Netshikweta and Tshitangano23). Peers swapped knowledge of various weight loss methods and encouraged their friends to lose weight. Other adolescents with ‘good figures’ were seen as role models(Reference Tshililo, Netshikweta and Tshitangano23). These dysfunctional eating practices were mostly used to lose weight; however, several other motivations were reported. For example, meals were often skipped due to the embarrassment of eating cultural-specific home-cooked lunch in front of peers(Reference Morris and Szabo20). Some cultural traditions also lead to girls participating in regular purging. This was often encouraged by parents, some of which would buy laxatives for family members to purge any consumed food which was perceived to contain toxins(Reference Morris and Szabo20).
The research exploring the influence of body image and self-esteem on diet was gendered with a focus on adolescent girls. This compares with findings from high-income countries, where body dissatisfaction has been found to be more commonly reported amongst girls than boys(Reference Locke and Lloyd-Sherlock30). In this review, disordered eating practices were only explored among adolescent girls living in the most economically transitioned setting (South Africa). This is also comparable with qualtiative studies finding that peers and a dissatisfaction with body image influence ‘crash dieting’ behaviours(Reference Hulme31).
Theme 2: social and environmental influences on diet and physical activity
Safety of outside spaces and physical activity
Across both the South African and Indian literature, adolescents’ concerns for their own safety, particularly among girls, were apparent. Fear of violent attacks and sexual exploitation had a negative impact on adolescents’ use of outside spaces. This was particularly prominent in the South African literature where many girls reported feeling unsafe in their environment(Reference Mmari, Blum and Sonenstein28). Indian adolescent girls were often forbidden from leaving their homes. This was due to fears from both adolescents and their parents that girls leaving the home would inevitably be at risk of sexual exploitation, which would ‘defame’ their families. This restriction was gendered, with boys having complete freedom to use outside spaces(Reference Mmari, Blum and Sonenstein28). Safety concerns were only highlighted in the settings undergoing economic transition and urbanisation in this review. Perceived safety is often associated with fear of crime in urban settings(Reference Caradas, Lambert and Charlton32).
Academic pressure v. diet and physical activity
Pressure to perform well academically was a common theme across the studies, particularly in the Indian studies. Indian adolescents emphasised how academic pressure, busy schedules and the resulting stress were key barriers to engaging in physical activity and eating healthily. Adolescents in Correa’s (2017) study were being academically tested every day. Academic pressures took priority over food so much so that some adolescents reported that they forgot to eat once they had finished studying. Others reported feeling too ‘tensed’ about their tests to eat(Reference Correa, Rajaraman and Swaminathan26). Age also influenced how much adolescents were encouraged to focus on academic achievement over and above physical fitness. For example, South African adolescents reported that as they got older, they were encouraged to engage in increasing academic study and discouraged from participating in extra-curricular activities, including sports(Reference Sedibe, Kahn and Edin21).
Eating and doing physical activity with friends and family
Most included studies, from all countries, found that adolescents were knowledgeable about nutrition, food hygiene and understood that home-cooked food was generally healthier than ‘outside’ food. This was not just due to nutritional content, adolescents reported that ‘outside’ food maybe be prepared in a way that was unhygienic. It was believed that street vendors might not wash their hands and utensils regularly or ‘properly’, they might not maintain good levels of personal hygiene and that they might reuse their cooking oils(Reference Gavaravarapu, Vemula and Rao25,Reference Correa, Rajaraman and Swaminathan26) . However, despite this awareness adolescents continued to eat ‘outside’ junk foods. Both Indian and South African adolescents reported that they ate more junk foods when socialising with their friends. Food was an important part of the adolescents’ social worlds whereby friendship groups or pairs would often purchase and eat these foods together(Reference Voorend, Norris and Griffiths19). Although adolescents were more likely to eat junk food when in the presence of friends, peers also encouraged each other to engage in physical activity(Reference Sedibe, Kahn and Edin21). Peers having a significant influence on adolescent diet and physical activity have been found across the globe, including adolescents living in HIC(Reference Wrottesley, Pedro and Fall33).
Theme 3: poverty, diet and physical activity
Food (in) security
One Ethiopian study was eligible for inclusion in this review, which was also the least economically transitioned setting. The main message of which was the complex interaction between poverty and diet. For those adolescents living in the most rural, resource-poor settings, being able to satisfy their hunger was key to having a good quality of life. Adolescents distinguished each other as ‘rich or poor’ by the types of food they ate. Food insecurity was also gendered in Ethiopia, due to restrictions on girls’ mobility. For example, adolescents in the current study reported that as girls got older, there was an expectation that they helped with domestic chores. Where girls mostly stayed at home and had little financial independence, the adolescent boys were encouraged to work outside of the home in agricultural practices or to move to towns to earn money(Reference Morrow, Tafere and Chuta29). Reliance on subsistence farming in Ethiopia was clear, and many families were living ‘hand-to-mouth’. As a result, adolescents valued the rain highly because it was an assurance that crops would grow and that they and their families could eat. Though food insecurity was much less of a factor in South Africa, there was mention of it and how it influenced the adolescents’ diets. For example, adolescents in Sedibe’s (2014) study reported that many young people skip breakfast due to lack of available food(Reference Sedibe, Kahn and Edin21). In South Africa, groceries were bought monthly. Once these had been consumed, many families would rely on foods they could pick from the fields. During these times, adolescents often would not be able to take a packed lunch to school. Despite this, many preferred not to eat the poor-quality free school food. Girls reported having stomach cramps from eating the free school meals. Instead, they would walk around outside and avoid eating the food.
Food safety and hygiene
Food hygiene was incredibly important to the Indian adolescents, who were concerned about the possible contamination of outside foods. Alongside the rapid economic and nutrition transition in India has seen an increase in outside food vendors and markets. There was a good understanding that unhygienic foods could cause illness and therefore should be avoided, where possible. They were particularly concerned with ‘open foods’ that were not covered and were exposed to flies(Reference Gavaravarapu, Vemula and Rao25,Reference Correa, Rajaraman and Swaminathan26) . Despite this, adolescents were still tempted to buy food from roadside vendors(Reference Correa, Rajaraman and Swaminathan26).
Poverty as a barrier and facilitator to physical activity
Poverty both encouraged participation in and acted as a barrier for adolescent physical activity as was emphasised particularly in the studies conducted in South Africa. Many had no choice but to engage in physical domestic work at home, and others were incentivised to take part in sporting competitions by the prospect of a cash prize(Reference Kinsman, Norris and Kahn22). For those adolescents who lived in impoverished areas, sports facilities were either limited or vandalised and outdoor spaces were deemed unsafe. Opportunities to use outdoor spaces and sports facilities were gendered, with girls limited to poorly funded sports spaces compared with boys.
Discussion
Using a qualitative evidence synthesis approach, this paper reviewed and thematically synthesised the available qualitative literature exploring influences on diet and physical activity behaviours among adolescents living in five TALENT countries (India, South Africa, Côte d’Ivoire, Ethiopia and The Gambia). The existing qualitative literature focussed on three major areas in relation to adolescent diet and physical activity: (1) the influences of body image and self-esteem; (2) social and environmental influences and (3) poverty (see Fig. 2). Most studies provided good-quality reports of thoughtful and rigorous qualitative work. The main weaknesses inherent in most of the studies comprised not reflecting adequately on the role of the researchers and the relationship between researchers and participants and reporting clearly on key ethical considerations. The small number of studies identified for inclusion in this review supports previous reports that qualitative research remains in its infancy within developmental studies in LMIC and tends to lack credibility with more quantitatively orientated researchers and public policy makers(Reference Locke and Lloyd-Sherlock30). Although large-scale quantitative methodologies are important and can offer an insight into trends across populations, qualitative research can provide a more in-depth insights into the complex experiences of adolescents living in transitioning societies and how this relates to diet and physical activity(Reference Hulme31).
Implications of body image and self-esteem on diet and physical activity
The South African studies focussed heavily on adolescent body image its relation to diet and physical activity and primarily among South African adolescent girls. These studies highlighted a contrast in concerns between adolescents who desired a ‘curvy’ figure and those who desired to be ‘thin’. South Africa is home to a diverse population, and previous research has shown that ethnicity may influence these differences in perceptions of the ‘ideal’ body shape and body dissatisfaction(Reference Caradas, Lambert and Charlton32). Previous reviews have indicated white adolescent girls in South Africa were more likely to idealise a ‘thin’ body image compared with their Black South African peers(Reference Wrottesley, Pedro and Fall33). Despite this, the studies in this review showed that young Black women engaged more in disordered eating practices including purging(Reference Wrottesley, Pedro and Fall33). In the findings of the current review, South African girls expressed a desire to ‘look good’ which for many was associated with thinness. For some, this was due to the belief that boys found thin girls more attractive. Interestingly, body image was not explored among adolescent boys. In Africa, most of the previous research into adolescent body image has focussed on girls(Reference Caradas, Lambert and Charlton32,Reference Szabo and Allwood34) . This may be because dysfunctional eating practices are frequently reported among adolescent girls than boys or because adolescent overweight and obesity are much more common in girls than boys in this setting(Reference Pedro, Micklesfield and Kahn35). There have been studies, however, showing that ethnicity is also associated with differences in BMI, eating attitudes and self-esteem in boys(Reference Gitau, Micklesfield and Pettifor36).
Social and environmental influences on diet and physical activity
The findings of this review show that academic pressure, perceived safety of outside spaces and family and friends have influence over adolescent diet and physical activity behaviours. These findings support research from HIC and show that, although mothers have a strong influence over what is eaten at home(Reference Feunekes, de Graaf and Meyboom37), decision making about food is actually very complex, with many factors influencing the choice of family meals including negotiations between all family members(Reference De Bourdeaudhuij and Van Oost38). Although similar studies in HIC have highlighted the influence of the family, there are often many more factors associated with food choice, including the increasing importance of peer influence during adolescence. Interactions with peers are complex(Reference Contento, Williams and Michela39), and this is demonstrated in the findings of the current review in relation to diet. Many of the adolescents in this review felt that their peers influenced their body image, dieting and weight-loss behaviours. In contrast, however, they also felt that they influenced the consumption of energy-dense junk foods.
Poverty, diet and physical activity
Although there was only one study from Ethiopia, the findings provide a unique insight into the lives of the adolescents living there and how important their socio-economic status was in determining their diet(Reference Morrow, Tafere and Chuta29). Interestingly, although food insecurity was the focus of discussion, body image was still mentioned in the current study. Weight was used to distinguish between rich and poor, with obesity being associated with affluence. Many LMIC including, India and many countries within sub-Saharan Africa, are undergoing transitions in diet, nutritional status and life expectancy(Reference Abrahams, Mchiza and Steyn40,Reference Misra, Singhal and Sivakumar41) . These transitions from rural subsistence to urban economies are associated with a shift towards a more urbanised diet with increasing availability of energy-dense junk foods. However, many areas in India and sub-Saharan Africa, including Ethiopia are still in the early stages of transition whereby affordability and availability are key determinants of nutritional status.
Review findings in the context of nutrition transition
The findings of this review are based largely on qualitative studies conducted in South Africa and India, both of which are both rapidly undergoing economic and nutrition transition. In high-income countries, such as the United Kingdom, ‘ultra-processed’, energy-dense foods have been one of the main sources of energy consumed since the 1950s(Reference Monteiro, Cannon and Moubarac42). However, these types of foods have only recently been available globally. Urbanisation in both India and South Africa has led to a shift towards a more westernised diet, characterised by ‘ultra-processed’, energy-dense, nutrient poor foods which have a profound impact on the health of these populations. The double burden of malnutrition seen in transitioning settings presents a serious public health challenge(Reference Siddiqui, Donato and Jumrani43). For example, India is facing an epidemic of non-communicable chronic diseases that exist alongside prevalent undernutrition resulting in substantial socio-economic burden(Reference Misra, Singhal and Sivakumar41). Along with high rates of undernutrition, there has been a significant increase in the prevalence of overweight and obesity in both India and South Africa(44,Reference Wentzel-Viljoen, Lee and Laubscher45) . This nutrition transition is reflected in the findings of this review in as far as studies with adolescents in the more transitioned settings focus discussions on both over and underweight (India and South Africa) and in this way differ from the focus on food insecurity in discussions with adolescents from Ethiopia, which is at an earlier stage of transition. The findings from the more transitioned settings in this review indicate that social influences, including peers, are key influences on diet and physical activity choices. Peers and other social factors have also been shown to heavily influence the diet choices of adolescents in from HIC such as the USA(Reference Chung, Ersig and McCarthy46).
Strengths and limitations
Previous reviews have focussed on the synthesis and analysis of quantitative data, the feasibility or effectiveness of interventions(Reference Caraher, Wu and Seeley47) and data that have been collected in HIC(Reference Shepherd, Harden and Rees48). The current review is to our knowledge the first to summarise qualitative research using data from adolescents in LMIC. However, studies from only five countries were searched for in line with the focus of the TALENT consortium, and future reviews may benefit from including other LMIC countries. Finally, articles written in English were included in this review as there were no resources available to allow inclusion of articles written in other languages. Many African countries are French speaking, including Côte d’Ivoire, and some studies may have therefore been excluded.
Implications for public health interventions and future research
Qualitative research exploring the diet and physical activity behaviours of adolescents living in the five TALENT countries is in its infancy. Six of the twelve included studies were carried out with adolescent girls, many of which focussed on the way in which body image influences adolescent diet and physical activity. However, findings from the synthesis of all the included papers highlight the fact that adolescent diet and physical activity behaviours are influenced by a multitude of complex factors including social, cultural, environmental and economic. Prioritising understanding of adolescent nutritional status in both girls and boys, as well as promoting healthy attitudes and behaviours around diet and physical activity in this population, is critical to improving the intergenerational transmission of health and well-being(Reference Patton, Olsson and Skirbekk5). Therefore, there is a need for future research to describe the complex nature of the diet and physical activity behaviours of adolescents living in LMIC at different stages of transition. The qualitative studies included in this special issue aim to address this need.
Conclusion
The findings of this review are based heavily on research conducted in two countries, South Africa and India. These findings suggest that body image is a key factor in adolescents’ (particularly girls’) diet choices and engagement in physical activity even in LMIC. This was driven by a desire to look aesthetically pleasing by creating a body that was either ‘thin’ or ‘curvy’. In some cases, the desire to look a certain way led to the uptake of dysfunctional eating practices. Other social and environmental influences included perceived safety of outdoor spaces to exercise in, academic pressures and eating as a social activity. Food insecurity, lack of availability and affordability was a key influence on the diet choices of adolescents in Ethiopia. In-depth research exploring adolescent perceptions of diet and physical activity is needed to better understand how both boys and girls, at different stages of adolescence, perceive health, diet and physical activity. More research with young people is required especially in countries where little exists to cover a wider range of issues that play a role in diet and physical activity.
Acknowledgements
Acknowledgements: The TALENT collaboration comprises: Laurence Adonis-Koffy, Yopougon University Hospital Faculty of Medical Sciences – UFHB de Cocody Abidjan Ivory Coast; Ulka Banavali, Regional Center for Adolescent Health and Nutrition, BKL Walawalkar Rural Medical College, Chiplun, India; Edna Bosire, University of the Witwatersrand, Johannesburg, South Africa; Harsha Chopra, Centre for the Study of Social Change, Mumbai, India; Meera Gandhi, Centre for the Study of Social Change, Mumbai, India; Abraham Haileamlak, College of Public Health and Medical Sciences, Jimma University, Jimma, Ethiopia; Ramatoulie Janha, MRC Keneba, MRC Unit The Gambia, MRC Keneba; Landing Jarjou, MRC Unit The Gambia; Julie Jesson, Inserm U1027, University of Toulouse, Paul Sabatier, France; Shama Joseph, Epidemiology Research Unit, CSI Holdsworth Memorial Hospital, Mysore, India; Sarah H Kehoe, MRC Lifecourse Epidemiology Unit, University of Southampton, Southampton General Hospital, UK; Kejal Joshi Reddy, Unit, KEM Hospital, Pune, India; Elizabeth Kimani-Murage, African Population and Health Research Center (APHRC), Nairobi, Kenya; Egnon-Kouakou, PAC-CI, Abidjan, Ivory Coast; GV Krishnaveni, Epidemiology Research Unit, CSI Holdsworth Memorial Hospital, Mysore, India; Kalyanaraman Kumaran, MRC Lifecourse Epidemiology Unit, University of Southampton and Head, UK and Epidemiology Research Unit, CSI Holdsworth Memorial Hospital, Mysore, India; Valeriane Leroy, Inserm U1027, University of Toulouse, Paul Sabatier, France; Mubarek Abera Mengistie, Jimma University, Ethiopia; Sophie Moore, Kings College London, London, UK; Shane Norris, Developmental Pathways Research Unit, University of the Witwatersrand, Johannesburg, South Africa; Suvarna Patil, Regional Center for Adolescent Health and Nutrition, BKL Walawalkar Rural Medical College, Chiplun, India; Sirazul Ameen Sahariah, Centre for the Study of Social Change, Mumbai, India; Kate Ward, MRC Lifecourse Epidemiology Unit, University of Southampton, UK; Stephanie Wrottesley, University of the Witwatersrand, Johannesburg, South Africa; Chittaranjan Yajnik, Diabetes Research Unit, KEM Hospital, Pune, India; Pallavi Yajnik, Diabetes Research Unit, KEM Hospital, Pune, India. Financial support: the current study was funded by a Global Challenges Research Fund/Medical Research Council pump priming grant (grant number: MC_PC_MR/R018545/1) and a University of Southampton Global Challenges Research Fund Strategic Development grant. The funding agency was not involved in the study design, data analysis or writing of this article. Conflict of interest: There are no conflicts of interest. Authorship: P.H.J. and P.D. designed the search strategy, ran the searches, sifted through the results and conducted the synthesis and drafted the manuscript. C.H.D.F., M.E.B., S.W. and S.S. contributed to the manuscript. Ethics of human subject participation: The current study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving research study participants were approved by the University of Southampton ethics committee. Written informed consent was obtained from all participants.
Supplementary material
For supplementary material accompanying this paper visit https://doi.org/10.1017/S1368980021002408
