Food and nutrition knowledge

Nine studies reported food and nutrition knowledge to have an influence on dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23–Reference Holdsworth, Delpeuch and Landais25,Reference Van Den Berg, Okeyo and Dannhauser35,Reference Charlton, Brewitt and Bourne36,Reference Boatemaa, Badasu and De-Graft Aikins38,Reference Hiamey, Amuquandoh and Boison39)} . Several misconceptions about healthy eating were reported in qualitative terms. Healthy eating was interpreted in terms of moderate intake of fat with unclear understanding of the benefits of fruits and vegetables and diet diversification. Dietary diversity was understood from the perspective of consuming different foods rather than from concept of food group diversity^{(Reference Boatemaa, Badasu and De-Graft Aikins38)}. Knowledge of the role of healthy food to NCD prevention was scanty^{(Reference Holdsworth, Delpeuch and Landais25,Reference Charlton, Brewitt and Bourne36)} . There was a link between fat and NCD, but salt and sugar linkage along with preventive role of fruits and vegetables was not understood. The role of food and nutrition knowledge on dietary behaviours was inconclusive^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Hiamey, Amuquandoh and Boison39)} . A study among young adults in South Africa reported that despite having a good understanding of healthy eating, it was not a priority when making dietary decisions^{(Reference Phillips, Comeau and Pisa23)}. In contrast, a study among Ghanaian street food consumers found that among the reasons people choose street foods was to compose a healthy diet from the available diversity of foods^{(Reference Hiamey, Amuquandoh and Boison39)}.

Food attitude

Ten studies reported attitude to be a determinant of dietary behaviours^{(Reference Kiawi, Edwards and Shu20,Reference Cockx, Colen and De Weerdt21,Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36–Reference Hiamey, Amuquandoh and Boison39)} . All studies reported a positive attitude towards energy-dense diets that are rich in sugar, salt or fat (unhealthy). Unhealthy foods were perceived to have a better taste and to require minimal or no time for preparation^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36–Reference Hiamey, Amuquandoh and Boison39)} . Healthy foods, including whole grains, fruits and vegetables, were not preferred by consumers. The high cost of food, long time involved in food preparation and undesirable sensory attributes (tasteless, dry and insipid) explained the negative attitude towards healthy foods^{(Reference Kiawi, Edwards and Shu20,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36,Reference Boatemaa, Badasu and De-Graft Aikins38)} . Perceived health concerns also influenced dietary behaviours^{(Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36,Reference Boatemaa, Badasu and De-Graft Aikins38)} . Health concerns (adverse health risks to an individual or loved ones) were discussed as elicitors of a positive attitude towards healthy foods^{(Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36,Reference Boatemaa, Badasu and De-Graft Aikins38)} . However, this was discussed to influence older adults. Young adults perceived that health concerns only manifest in old age.

Psychological and physiological factors

Five studies found emotional satisfaction to be a determinant for consumption of foods rich in fat, salt and sugar^{(Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Zyl, Steyn and Marais34,Reference Stern, Puoane and Tsolekile37,Reference Boatemaa, Badasu and De-Graft Aikins38)} . Fruits and vegetables were also perceived to be non-satiating^{(Reference Everett-Murphy, De Villiers and Ketterer24,Reference Boatemaa, Badasu and De-Graft Aikins38)} .

Time and financial barriers

Financial difficulties and lack of time were noted as major barriers influencing dietary behaviours^{(Reference Kiawi, Edwards and Shu20,Reference Cockx, Colen and De Weerdt21,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36–Reference Hiamey, Amuquandoh and Boison39)} . Financial difficulties were mentioned as determinants of low fruit and vegetable intake^{(Reference Kiawi, Edwards and Shu20,Reference Cockx, Colen and De Weerdt21,Reference Charlton, Brewitt and Bourne36,Reference Boatemaa, Badasu and De-Graft Aikins38)} . Studies on consumption of street foods in Kenya, Ghana and South Africa found that ability to get food on credit and savings on fuel when food preparation is avoided partly explains the increased consumption of these foods^{(Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Stern, Puoane and Tsolekile37,Reference Hiamey, Amuquandoh and Boison39)} . In South Africa and Cameroon, participants perceived that a balanced diet was hard to sustain due to budget constraints.

Food skills and self-efficacy

Dietary behaviours were discussed to be influenced by food preparation skills and self-efficacy^{(Reference Kiawi, Edwards and Shu20,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Charlton, Brewitt and Bourne36–Reference Boatemaa, Badasu and De-Graft Aikins38)} . Participants in a qualitative study in Ghana reported that the ability to improvise with available resources to prepare healthy tasty meals informs their dietary decisions^{(Reference Boatemaa, Badasu and De-Graft Aikins38)}. Everett-Murphy et al. ^{(Reference Everett-Murphy, De Villiers and Ketterer24)} in South Africa found that many individuals lack the skills and confidence to prepare healthy tasty meals. Participants reported a difficulty in breaking unhealthy entrenched dietary habits^{(Reference Everett-Murphy, De Villiers and Ketterer24)}. The difficulty was mediated by rejection of healthy foods by family especially when perceived to be un-tasty.

Food habits (habit strength)

Five studies reported food habits to have an influence on dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Kiawi, Edwards and Shu20,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Den Berg, Okeyo and Dannhauser35,Reference Hiamey, Amuquandoh and Boison39)} . High consumption of carbohydrate-rich foods and low consumption of fruits and vegetables are habitually ingrained^{(Reference Becquey, Savy and Danel17,Reference Kiawi, Edwards and Shu20,Reference Van Den Berg, Okeyo and Dannhauser35,Reference Hiamey, Amuquandoh and Boison39)} . In South Africa, sweet treats are social rituals on weekends, while deep frying is an entrenched habit nurtured from childhood^{(Reference Everett-Murphy, De Villiers and Ketterer24)}.

Age

Four studies reported age to have an influence on dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Savy, Martin-Prével and Danel41)} . In Cameroon, young adults (18–40 years) regarded fruits and vegetables to be for the older adults (≥40 years)^{(Reference Kiawi, Edwards and Shu20)}. Younger Burkinabe adults compared with older adults have a snacking behaviour with mainly energy-dense foods that are nutrient poor^{(Reference Becquey, Savy and Danel17,Reference Savy, Martin-Prével and Danel41)} .

Socio-economic status

Seven studies reported on influence of socio-economic status indices: income/economic, education and occupation on dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Cockx, Colen and De Weerdt21,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Hiamey, Amuquandoh and Boison39–Reference Savy, Martin-Prével and Danel41)} . Findings are mixed on the direction of influence. A low-income status was found to be a predictor for high consumption of unhealthy street foods in Ghana, Kenya and South Africa^{(Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Hiamey, Amuquandoh and Boison39)} . In contrast, a high-income status in Tanzania was a predictor for consumption of unhealthy foods^{(Reference Cockx, Colen and De Weerdt21)}. Three studies found higher education level to be associated with higher dietary diversity and increased likelihood to choose healthy food offers^{(Reference Becquey, Savy and Danel17,Reference Van Zyl, Steyn and Marais34,Reference Savy, Martin-Prével and Danel41)} . Women working in the informal sector (street and market vendors) were reported to depend on majorly street foods^{(Reference Hiamey, Amuquandoh and Boison39)}. In urban Burkina Faso, working in the formal sector (professional/corporate jobs) was associated with higher snacking habits^{(Reference Becquey, Savy and Danel17)}.

Household composition

Household size and composition, according to three studies, had an influence on dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Boatemaa, Badasu and De-Graft Aikins38)} . The presence of a housewife was associated with decreased dependence on out of home eating^{(Reference Van ‘T Riet, Den Hartog and Mwangi26)}. A large household size was associated with increased prioritisation of satiety-giving carbohydrate-rich foods over other food groups^{(Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Boatemaa, Badasu and De-Graft Aikins38)} .

Family and peer influence

Six studies found dietary decisions to be mediated by the family and friends^{(Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36–Reference Boatemaa, Badasu and De-Graft Aikins38)} . Dietary patterns are adopted from largely family networks^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Boatemaa, Badasu and De-Graft Aikins38)} . At the same time, the preferences of family members are a vital consideration in dietary choices^{(Reference Everett-Murphy, De Villiers and Ketterer24,Reference Charlton, Brewitt and Bourne36)} . Two qualitative studies in South Africa and Ghana found that due to peer pressure, young adults (≤ 35 years) eat more energy-dense street/fast foods^{(Reference Van Zyl, Steyn and Marais34,Reference Boatemaa, Badasu and De-Graft Aikins38)} . Community groups were also reported as determinants of dietary behaviours.

Cultural beliefs

Several cultural beliefs influenced dietary behaviours^{(Reference Becquey, Savy and Danel17,Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van Den Berg, Okeyo and Dannhauser35,Reference Stern, Puoane and Tsolekile37–Reference Hiamey, Amuquandoh and Boison39)} . Three studies reported perception of body image to have an influence on food intake^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37)} . Overweight is associated with beauty, dignity, health, wealth and good treatment by husband, while weight loss is a source of stigma and a sign of HIV/AIDS^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37)} . A positive weight gain perception is a motivation for unhealthy dietary behaviours. Other reported cultural beliefs included low social status accorded to fruits and vegetables, legumes and minimally refined cereals, and high social status accorded to unhealthy fast foods and eating out^{(Reference Everett-Murphy, De Villiers and Ketterer24,Reference Stern, Puoane and Tsolekile37–Reference Hiamey, Amuquandoh and Boison39)} .

Physical environment

Eleven studies reported food choices to be influenced by ubiquitous cheap high energy-dense nutrient-poor foods and food deserts^{(Reference Becquey, Savy and Danel17,Reference Dake, Thompson and Ng19–Reference Cockx, Colen and De Weerdt21,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Van Zyl, Steyn and Marais34,Reference Stern, Puoane and Tsolekile37–Reference Aryeetey, Oltmans and Owusu40)} . An abundance of ready-to-eat street food offers was reported^{(Reference Dake, Thompson and Ng19,Reference Everett-Murphy, De Villiers and Ketterer24,Reference Van ‘T Riet, Den Hartog and Mwangi26,Reference Stern, Puoane and Tsolekile37,Reference Hiamey, Amuquandoh and Boison39)} . These street offers are limited in fruits and vegetables but provide mostly refined carbohydrate-rich foods. Food purchases are from mainly three retail sources: traditional markets including farmers’ weekly market days, supermarkets and convenience retail stores that offer mainly processed food^{(Reference Dake, Thompson and Ng19,Reference Aryeetey, Oltmans and Owusu40)} . Studies reported that majority of the population still prefer traditional markets because they offer low-priced fresh food. It was also found that proximity to food retail stores determines purchasing behaviour^{(Reference Aryeetey, Oltmans and Owusu40)}. Participants in these studies reported a high dependence on convenience stores due to their ubiquity in the communities^{(Reference Dake, Thompson and Ng19,Reference Boatemaa, Badasu and De-Graft Aikins38,Reference Aryeetey, Oltmans and Owusu40)} . Some participants reported weekly food purchasing behaviour to counter accessibility difficulties to traditional markets^{(Reference Aryeetey, Oltmans and Owusu40)}.

Virtual/information environment

Three studies reported virtual environment (media) to influence food decisions^{(Reference Phillips, Comeau and Pisa23,Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36)} . A strong trust in credibility of nutrition information shared from mainstream media (television and radio) was reported^{(Reference Van Zyl, Steyn and Marais34,Reference Charlton, Brewitt and Bourne36)} .

Awareness

Seven studies reported key determinants of PA as being knowledge/awareness of PA recommendations, its health benefits and how one might engage in PA^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43,Reference Walter and du Rosa46,Reference Adeniyi, Ogwumike and Bamikefa47)} . Researchers found a high proportion of participants unaware of the health benefits of PA^{(Reference Kiawi, Edwards and Shu20,Reference Stern, Puoane and Tsolekile37,Reference Adeniyi, Ogwumike and Bamikefa47)} . Health benefits of PA were related to mainly body cleansing^{(Reference Kiawi, Edwards and Shu20,Reference Watson, Norris and Draper43)} . Knowledge on the role of PA in NCD prevention was lacking. Adeniyi et al. ^{(Reference Adeniyi, Ogwumike and Bamikefa47)} found that an estimated 61 % of Nigerian women were not aware of PA benefits towards enhancing postpartum health. Studying PA in pregnancy, Muzigaba et al. ^{(Reference Muzigaba, Kolbe-Alexander and Wong42)} reported that women were not aware of PA recommendations.

Intra-individual barriers

Barriers included lack of time, feeling tired (laziness), preference for sedentary activities (television viewing) and financial constraints (money to buy bike, to pay at gyms)^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43,Reference Hattingh, Le Roux and Nel45,Reference Adeniyi, Ogwumike and Bamikefa47,Reference Afolabi, Addo and Sonibare48)} . Long working hours and need to prioritise social obligations (family needs) were reported to be a major basis for insufficient time^{(Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43,Reference Adeniyi, Ogwumike and Bamikefa47)} . Physiological changes (increased stomach size) during pregnancy and psychological fear that increased PA may hurt the baby or cause miscarriage were reported to be barriers to PA^{(Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43)} .

Physical activity attitude

Perceived severity, benefits and positive experience were reported to have an influence on PA attitude^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43,Reference Walter and du Rosa46,Reference Adeniyi, Ogwumike and Bamikefa47)} . Perceived PA benefits included positive pregnancy outcomes, body freshening, keeping in shape and feel-good mood^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43)} . Likewise, health concerns (adverse health risks to an individual or loved one) were discussed to increase likelihood to engage in PA^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Walter and du Rosa46)} . Positive experience, for example, blood pressure normalisation resulting from a past PA programme, was reported as a motivator to continue with PA^{(Reference Phillips, Comeau and Pisa23)}. The PA perceived to be simple, and socially engaging was described to contribute to a positive attitude^{(Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Adeniyi, Ogwumike and Bamikefa47)} .

Physical activity self-efficacy

Two studies reported self-efficacy to be a determinant of PA behaviour^{(Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Adeniyi, Ogwumike and Bamikefa47)} . Several factors were described to determine PA self-efficacy. Social support, having a PA buddy or belonging to a group, increased PA self-efficacy^{(Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Adeniyi, Ogwumike and Bamikefa47)} .

Socio-economic status

Four descriptive cross-sectional studies found socio-economic indices, asset ownership and occupation to be associated with PA behaviours^{(Reference Gradidge, Crowther and Chirwa44,Reference Hattingh, Le Roux and Nel45,Reference Adeniyi, Ogwumike and Bamikefa47,Reference Afolabi, Addo and Sonibare48)} . A study among Nigerian women working in informal markets reported that of the total 16 h spent on daily activities, 11·5 h are spent on predominantly sedentary market vending activities^{(Reference Afolabi, Addo and Sonibare48)}.

Social environment

Eight studies found family, friends and community groups to have an influence on PA^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37,Reference Muzigaba, Kolbe-Alexander and Wong42–Reference Gradidge, Crowther and Chirwa44,Reference Walter and du Rosa46,Reference Adeniyi, Ogwumike and Bamikefa47)} . Direction of social environment influence was inconclusive. Family members were reported to discourage their relatives from PA they do not approve^{(Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Walter and du Rosa46)} . Membership in a community health club/group positively influenced PA^{(Reference Stern, Puoane and Tsolekile37,Reference Watson, Norris and Draper43)} , while lack of access to sports teams/community groups had a negative influence^{(Reference Phillips, Comeau and Pisa23,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Watson, Norris and Draper43,Reference Adeniyi, Ogwumike and Bamikefa47)} . Community groups were reported to offer practical solutions to overcome PA barriers, emotional support and an opportunity to participate in group PA.

Cultural beliefs

Six studies reported cultural misconceptions to be determinants of PA^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37,Reference Watson, Norris and Draper43,Reference Walter and du Rosa46,Reference Adeniyi, Ogwumike and Bamikefa47)} . PA is viewed as unfeminine and in conflict with the perceived weight gain prestige^{(Reference Kiawi, Edwards and Shu20,Reference Phillips, Comeau and Pisa23,Reference Stern, Puoane and Tsolekile37,Reference Walter and du Rosa46)} . Walking is considered a sign of poverty and idleness^{(Reference Kiawi, Edwards and Shu20,Reference Stern, Puoane and Tsolekile37,Reference Watson, Norris and Draper43)} . Two studies reported a gender stereotype that women by tradition should not engage in outdoor-related PA^{(Reference Phillips, Comeau and Pisa23,Reference Walter and du Rosa46)} . Dress code also had an influence on PA, and exercise attires (tight and short) are perceived negatively^{(Reference Walter and du Rosa46)}.

Physical environment

Insufficient infrastructure (playgrounds, parks, pavements for pedestrians and safe roads for cyclists) and safety concerns (violence against women) were the major impediments to outdoor PA^{(Reference Kiawi, Edwards and Shu20,Reference Stern, Puoane and Tsolekile37,Reference Muzigaba, Kolbe-Alexander and Wong42,Reference Hattingh, Le Roux and Nel45)} . Three studies reported cheap motorised transportation (motor bikes) to influence PA^{(Reference Kiawi, Edwards and Shu20,Reference Stern, Puoane and Tsolekile37,Reference Hattingh, Le Roux and Nel45)} .

Fig. 3 gives a schematic overview of determinants of PA among WRA in urban SSA arranged according to the socio-ecological framework.