To provide an overview of the status of monitoring and evaluation (M&E) of school feeding across sub-Saharan Africa and to reflect on the experience on strengthening M&E systems to influence policy making in low-income countries.

Literature review on the M&E of school feeding programmes as well as data from World Food Programme surveys.

Sub-Saharan Africa.

Countries implementing school feeding.

Only two randomized controlled impact evaluations have been implemented in sub-Saharan Africa. Where M&E data collection is underway, the focus is on process and service delivery and not on child outcomes. M&E systems generally operate under the Ministry of Education, with other Ministries represented within technical steering groups supporting implementation. There is no internationally accepted standardized framework for the M&E of school feeding. There have been examples where evidence of programme performance has influenced policy: considering the popularity of school feeding these cases though are anecdotal, highlighting the opportunity for systemic changes.

There is strong buy-in on school feeding from governments in sub-Saharan Africa. In response to this demand, development partners have been harmonizing their support to strengthen national programmes, with a focus on M&E. However, policy processes are complex and can be influenced by a number of factors. A comprehensive but simple approach is needed where the first step is to ensure a valid mandate to intervene, legitimizing the interaction with key stakeholders, involving them in the problem definition and problem solving. This process has been facilitated through the provision of technical assistance and exposure to successful experiences through South–South cooperation and knowledge exchange.

This exploratory study assessed how 8–13-year-old children categorised and labelled fruit and vegetables (FaV), and how these were influenced by child characteristics, to specify second-level categories in a hierarchical food search system for a computerised 24 h dietary recall (hdr).

Two sets of food cards, sixty-seven for fruit (F) and sixty-four for vegetables (V), with pictures and names of FaV from ten professionally defined food categories were sorted, separately, by each child into piles of similar foods. Demographic data, BMI and 6-n-propylthiouracil (PROP) taster status were obtained.

Participants attended the Children’s Nutrition Research Center in the summer of 2006.

In all, 152 8–13-year-old children, predominantly English-speaking, of whom sixteen were predominantly Spanish-speaking.

Children created an average of 8·5 (5·3) piles with 7·9 (11·4) cards per pile for the F, and an average of 10·1 (4·8) piles with 6·2 (7·9) cards per pile for the V. No substantial differences in Robinson clustering were detected across subcategories for each of the demographic characteristics, BMI or PROP sensitivity. Children provided clusters names that were mostly ‘Taxonomic – Professional’ labels, such as salads, berries, peppers, for both F (51·8 %) and V (52·1 %).

These categories should be tested to assess their ability to facilitate search of FaV items in a computerised 24 hdr for children in this age group.

To assess the burden of iodine deficiency in pregnancy in Africa using estimated pregnancy median urinary iodine concentration (pMUIC).

pMUIC for each African country was estimated using a regression equation derived by correlating the school-age children (SAC) median UIC (mUIC) and pMUIC from countries around the globe, and the SAC mUIC data for African countries obtained from the Iodine Global Network (IGN) 2017 and 2019 Score cards.

Iodine deficiency was endemic in many African countries before the introduction of iodine fortification, mainly through universal salt iodisation programmes about 25 years ago. There is a scarcity of data on the level of iodine nutrition in pregnancy in Africa. Women living in settings with pMUIC below 150 µg/l are at risk of iodine deficiency-related pregnancy complications.

Fifty of the fifty-five African countries that had data on iodine nutrition status.

A cut-off school age mUIC ≤ 175 µg/l is correlated with insufficient iodine intake in pregnancy (pregnancy mUIC ≤ 150 μg/l). Twenty-two African countries had SAC mUIC < 175 μg/l, which correlated with insufficient iodine intake during pregnancy (pMUIC < 150 μg/l). However, nine of these twenty-two countries had adequate iodine intake based on SAC mUIC.

There is likely a high prevalence of insufficient iodine intake in pregnancy, including in some African countries classified as having adequate iodine intake in the general population. A SAC mUIC ≤ 175 µg/l predicts insufficient iodine intake among pregnant women in these settings.

Traditionally, carbohydrate has been the largest contributor to energy intake among people with diabetes, yet different carbohydrate foods produce different glycaemic responses. Glycaemic load represents the total glycaemic effect of the diet and influences glycaemic control. Adequate self-efficacy and outcome expectations are needed to change carbohydrate intake and to evaluate relevant interventions. The purpose of this research was to develop and test instruments regarding self-efficacy and outcome expectations for the adoption of a lower glycaemic load diet.

Participants completed each instrument at their convenience and mailed the instruments to the investigators.

A community sample of individuals 21–75 years of age with type 2 diabetes for ≥ 1 year (n = 108) was recruited.

Principal components analysis revealed three factors on the self-efficacy questionnaire: glycaemic index, negative food selection and self-regulation efficacy which accounted for 62% of the variance in these items. The outcome expectations instrument yielded three factors: barriers to dietary change and glycaemic control, and family support expectations which accounted for 48% of the variance. Coefficient α for each construct was >0.70 and coefficient H for each construct was ≥ 0.80.

The two instruments developed for this study can provide important insights about the self-efficacy and outcome expectations regarding the quantity and quality of carbohydrate consumed and self-monitoring performed for diabetes management. Future research is needed to evaluate the relationship among these constructs, dietary intake and glycaemic control.

There is limited understanding as to why children of low socio-economic position (SEP) consume poorer diets than children of high SEP. Evidence suggests that determinants of dietary intake may differ between SEP groups. The present study aimed to determine if SEP moderated associations of personal and environmental predictors with children’s non-core food and sweetened drink intakes and unhealthy dietary behaviours.

Children completed online questionnaires and parents completed computer-assisted telephone interviews to assess intrapersonal and environmental dietary predictors. Dietary intake was measured using an FFQ. Parents reported demographic information for maternal education, occupation and employment, and household income.

Twenty-six primary schools in South Australia, Australia.

Children aged 9–13 years and their parents (n 395).

Multiple personal and home environment factors predicted non-core food and sweetened drink intakes, and these associations were moderated by SEP. Maternal education moderated associations of girls’ sweetened drink intake with self-efficacy, cooking skills and pressure to eat, and boys’ non-core food intake with monitoring, parent’s self-efficacy and home environment. Maternal occupation and employment moderated associations of sweetened drink intake with attitudes, self-efficacy, pressure to eat and food availability, and non-core food intake with parents’ self-efficacy and monitoring. Income moderated associations with pressure to eat and home environment.

Identifying differences in dietary predictors between socio-economic groups informs understanding of why socio-economic gradients in dietary intake may occur. Tailoring interventions and health promotion to the particular needs of socio-economically disadvantaged children may produce more successful outcomes and reduce socio-economic disparities in dietary intake.

To describe women's views about aspects of infants' diets around the time of weaning, making comparisons with national guidelines.

A survey of women with a 9-month-old child.

Adelaide, South Australia.

Five hundred and five women who joined a longitudinal study during pregnancy.

Sources of information varied, with written material most commonly used (37%). Cows' milk was considered suitable as the main drink for weaned infants by 14% of women. There were divergent views about the suitability of eggs, with many women concerned about allergy. The majority of women (84%) viewed fruit juice as suitable although many qualified their response, often by stating that fruit juice should be diluted. Almost all women considered the amount of sugar mattered, primarily because of tooth decay, and that salt mattered although the reason was often uncertain. It was widely believed (77%) that additives in food could cause health problems, in particular hyperactivity and allergies, and half of the women reported avoiding specific foods because of concerns about allergies. Many women thought that giving their child food that was high in fat would encourage a liking for ‘junk’ food or lead to fatness in adulthood.

There is considerable diversity in the views women express about aspects of infant feeding that have been the subject of guidelines. Further health promotion efforts are needed to achieve greater consistency with recommendations and to address other concerns women have. This will entail greater engagement with parents and shared development of responses.

High frequency of low birth weight (LBW) is observed in rural compared with urban Indian women. Since maternal BMI is known to be associated with pregnancy outcomes, the present study aimed to investigate factors associated with BMI in early pregnancy of urban and rural South Indian women.

Prospective observational cohort.

A hospital-based study conducted at an urban and a rural health centre in Karnataka State.

Pregnant women (n 843) aged 18–40 years recruited in early pregnancy from whom detailed sociodemographic, environmental, anthropometric and dietary intake information was collected.

A high proportion of low BMI (32 v. 26 %, P<0·000) and anaemia (48 v. 23 %, P<0·000) was observed in the rural v. the urban cohort. Rural women were younger, had lower body weight, tended to be shorter and less educated. They lived in poor housing conditions, had less access to piped water and good sanitation, used unrefined fuel for cooking and had lower standard of living score. The age (β=0·21, 95 % CI 0·14, 0·29), education level of their spouse (β=1·36, 95 % CI 0·71, 2·71) and fat intake (β=1·24, 95 % CI 0·20, 2·28) were positively associated with BMI in urban women.

Our findings indicate that risk factors associated with BMI in early pregnancy are different in rural and urban settings. It is important to study population-specific risk factors in relation to perinatal health.

Prevalence of acute malnutrition is classically estimated by the proportion of children meeting a case definition in a representative population sample. In 1995 the WHO proposed the PROBIT method, based on converting parameters of a normally distributed variable to cumulative probability, as an alternative method requiring a smaller sample size. The present study compares classical and PROBIT methods for estimating the prevalence of global, moderate and severe acute malnutrition (GAM, MAM and SAM) defined by weight-for-height Z-score (WHZ) or mid-upper arm circumference (MUAC).

Bias and precision of classical and PROBIT methods were compared by simulating a total of 1·26 million surveys generated from 560 nutrition surveys.

Data used for simulation were derived from nutritional surveys of children aged 6–59 months carried out in thirty-one countries around the world.

Data of 459 036 children aged 6–59 months from representative samples were used to generate simulated populations.

The PROBIT method provided an estimate of GAM, MAM and SAM using WHZ or MUAC proportional to the true prevalence with a small systematic overestimation. The PROBIT method was more precise than the classical method for estimating the prevalence for GAM, MAM and SAM by WHZ or MUAC for small sample sizes (i.e. n<150 for SAM and GAM; n<300 for MAM), but lost this advantage when sample sizes increased.

The classical method is preferred for estimating acute malnutrition prevalence from large sample surveys. The PROBIT method may be useful in sentinel-site surveillance systems with small sample sizes.

Perceptions of social-contextual food environments and associated factors that influence food purchases are understudied in American Indian (AI) communities. The purpose of the present study was to: (i) understand the perceived local food environment; (ii) investigate social-contextual factors that influence family food-purchasing choices; and (iii) identify diet intervention strategies.

This qualitative study consisted of focus groups with primary household shoppers and key-informant interviews with food retailers, local government food assistance programme directors and a dietitian. An inductive, constant comparison approach was used to identify major themes.

A large AI reservation community in the north-central USA.

Four focus groups (n 31) and seven key-informant interviews were conducted in February and May 2016.

Perceptions of both the higher cost of healthy foods and limited access to these foods influenced the types of foods participants purchased. Dependence on government assistance programmes and the timing of benefits also contributed to the types of foods purchased. Participants described purchasing foods based on the dietary needs and preferences of their children. Suggestions for improving the purchase and consumption of healthy foods included: culturally relevant and family-centred cooking classes and workshops focused on monthly food budgeting. Participants also emphasized the importance of involving the entire community in healthy eating initiatives.

Cost and access were the major perceived barriers to healthy eating in this large rural AI community. Recommended interventions included: (i) family-friendly and culturally relevant cooking classes; (ii) healthy food-budgeting skills training; and (iii) approaches that engage the entire community.

The present study aimed to examine the effectiveness of two different implementation forms of the EU School Fruit and Vegetable Scheme (SFVS).

A quasi-experimental design was applied including a thrice as well as a twice weekly intervention group. Repeated 24 h dietary recalls were used to measure children’s fruit and vegetable (F&V) intake. Effects were analysed on days with and without F&V deliveries using hierarchical linear regression models.

Twelve primary schools in North Rhine-Westphalia, Germany.

Third and fourth graders (n 664).

Average daily F&V intake at pre-intervention was 0·84 frequencies in the thrice weekly intervention group, 0·90 frequencies in the twice weekly intervention group and 1·25 frequencies in the control group. Providing children thrice weekly with F&V increased children’s F&V intake on average by 0·96 (P<0·001) frequencies/d. The effects were higher on days with (1·07; P<0·001) than on days without (0·75; P<0·001) F&V deliveries. Distributing F&V twice weekly resulted in an increase of 0·75 (P<0·001) frequencies/d on average, again with higher effects on days with (1·30; P<0·001) than without (0·48; P<0·003) F&V deliveries. Subgroup analysis revealed some indications for differential effectiveness only in the twice weekly intervention group.

The SFVS with thrice or twice weekly deliveries of F&V led to a significant increase in children’s F&V intake on days with and without deliveries. The latter might provide an indication of positive long-term effects of the scheme. The scheme shows equal efficiency for almost all subgroups.

To pilot-test a brief written prescription recommending lifestyle changes delivered by general practitioners (GPs) to their patients.

The Active Nutrition Script (ANS) included five nutrition messages and personalised exercise advice for a healthy lifestyle and/or the prevention of weight gain. GPs were asked to administer 10 scripts over 4 weeks to 10 adult patients with a body mass index (BMI) of between 23 and 30 kg m− 2. Information recorded on the script consisted of patients' weight, height, waist circumference, gender and date of birth, type and frequency of physical activity prescribed, and the selected nutrition messages. GPs also recorded reasons for administering the script. Interviews recorded GPs views on using the script.

General practices located across greater Melbourne.

Nineteen GPs (63% female) provided a median of nine scripts over 4 weeks. Scripts were administered to 145 patients (mean age: 54 ± 13.2 years, mean BMI: 31.7 ± 6.3 kg m− 2; 57% female), 52% of whom were classified as obese (BMI >30 kg m− 2). GPs cited ‘weight reduction’ as a reason for writing the script for 78% of patients. All interviewed GPs (90%, n = 17) indicated that the messages were clear and simple to deliver.

GPs found the ANS provided clear nutrition messages that were simple to deliver. However, GPs administered the script to obese patients for weight loss rather than to prevent weight gain among the target group. This has important implications for future health promotion interventions designed for general practice.

To examine urban–rural disparity in childhood stunting, wasting and malnutrition at national and subnational levels in Chinese primary-school children in 2010 and 2014.

Data were obtained from two nationwide cross-sectional surveys conducted in 2010 and 2014. Malnutrition was classified using the Chinese national ‘Screening Standard for Malnutrition of Children’.

All twenty-seven mainland provinces and four municipalities of mainland China.

Children aged 7–12 years (n 215 214; 107 741 in 2010 and 107 473 in 2014) from thirty-one provinces.

Stunting, wasting and malnutrition prevalence were 1·9, 12·3 and 13·7 % in 2010, but decreased to 1·0, 9·4 and 10·2 % in 2014, respectively. The prevalence of stunting, wasting and malnutrition in both urban and rural children was higher in western provinces, while lower in eastern provinces. Although the prevalence of wasting and malnutrition was higher in rural children than their urban counterparts, the urban–rural disparity in both wasting and malnutrition decreased from 2010 to 2014 (prevalence OR: wasting, 1·35 to 1·16; malnutrition, 1·50 to 1·27). A reversal occurred in 2014 in several eastern provinces where the prevalence of wasting and malnutrition in urban children surpassed their rural peers. The urban–rural disparity was larger in western provinces than eastern provinces.

The shrinking urban–rural disparity and the reversal in wasting and malnutrition suggest that the malnutrition situation has improved during the post-crisis period, especially in the western provinces. Region-specific policies and interventions can be useful to sustainably mitigate malnutrition in Chinese children, especially in rural areas and the western provinces.

To report on energy and nutrient intakes, as a risk factor for coronary heart disease in Indian South Africans.

Cross-sectional dietary study as part of a coronary heart disease survey.

Metropolitan area of Durban, South Africa.

Free-living Indian men (n = 406) and women (n = 370) 15–69 years of age.

Dietary data were collected by three interviewers using a 24-h dietary recall and expressed as median intakes of macronutrients.

Results reported a low energy intake and the percentage of energy derived from total fat varied between 32.3 and 34.9% in men and between 33.1 and 36.1% in women. The energy intake to basal metabolic rate (EI: BMR) ratios were low for all age groups suggesting potential under-reporting. Intake of polyunsaturated fatty acids was high, more than 10% of energy, with a median polyunsaturated to saturated fatty acid ratio (P: S ratio) that varied between 1.38 and 1.96 for the various age and sex groups. Dietary cholesterol varied between 66 and 117 mg per 4.2 MJ in men and between 76 and 109 mg per 4.2 MJ in women. Dietary fibre intakes were low and varied between 8.0 and 11.0 g per 4.2 MJ in men and between 7.6 and 9.6 g per 4.2 MJ in women.

The dietary P:S ratios were high and the effect of such a high P:S ratio on the oxidation of low density lipoprotein in this population, with a high prevalence of coronary heart disease (CHD), should be investigated as a possible risk factor for CHD.

Height-for-age shorter than 3th percentile or 2 sd below the median of the National Center for Health Statistics/WHO growth reference is known as stunting. The present study aimed to measure the association of several factors with stunting in a cohort born from 2009 to 2010 in Shiraz, Iran.

Nested case–control study conducted on pre-school children in 2016. Participants were 200 children suffering from stunting (case group) and 200 children with normal height (control group).

Results of multivariable logistic regression analysis suggested a significant relationship (adjusted OR; 95 % CI) between stunting and mother’s occupation (employed v. housewife: 3·58; 1·73, 7·39) and chronic diseases (yes v. no: 2·93; 1·25, 6·88). In addition, significant associations were found between family income (<10 million v. >30 million Rials: 5·63; 2·46, 12·87), diary consumption (very little v. high: 5·93; 1·74, 20·18) and animal protein in diet (low v. very high: 2·42; 1·13, 5·19) and stunting among children. Also, chance of stunting was inversely associated with birth interval (<2 years v. first child: 3·13; 1·45, 6·76) and duration of exclusive breast-feeding (18–24 months v. formula only or breast-feeding for <18 months: 0·53; 0·30, 0·92).

Mothers being housewife, possibly due to having more time to take care of their children, childhood diseases and consumption of dairy foods are strong and modifiable factors which can positively affect Iranian children’s stature. Higher family income is another important factor in stunting but under less parental control.

To compare federally reimbursable school meals served when competitive foods are removed and when marketing and nudging strategies are used in school cafeterias operating the National School Lunch Program (NSLP). The second objective was to determine how marketing and nudging strategies influence competitive food sales.

In the Healthy Choices School, all competitive foods were removed; the Healthy Nudging School retained competitive foods and promoted the school meal programme using marketing and nudging strategies; a third school made no changes. Cafeteria register data were collected from the beginning of the 2013–2014 school year through the four-week intervention. Outcome measures included daily entrées served; share of entrées served with vegetables, fruit and milk; and total competitive food sales. Difference-in-difference models were used to examine outcome measure changes.

Three high schools in a diverse, Northeast US urban district with universally free meals.

High-school students participating in the NSLP.

During the intervention weeks, the average number of entrées served daily was significantly higher in the Healthy Choices School (82·1 (se 33·9)) and the Healthy Nudging School (107·4 (se 28·2)) compared with the control school. The only significant change in meal component selection was a 6 % (se 0·02) higher rate of vegetable servings in the Healthy Choices School compared with the control school. Healthy Nudging School competitive food sales did not change.

Both strategies – removing competitive foods and marketing and nudging – may increase school meal participation. There was no evidence that promoting school meals decreased competitive food sales.

To estimate and compare dietary energy intake (DEI) and total energy expenditure (TEE) among adults, using questionnaires.

Comparative, cross-sectional study.

Community-based, at the demographic surveillance site (DSS) in Butajira District of Ethiopia.

A total of 619 adults, 18–64 years of age, were randomly selected from among the urban and rural population of Butajira using the DSS sampling frame. Habitual dietary intake and physical activity were assessed using questionnaires. BMR was estimated using a regression equation, and TEE was calculated from BMR and the metabolic energy equivalent task (MET) and duration of reported activities. Physical activity level (PAL) was calculated as TEE/BMR, while food intake level (FIL) was calculated as DEI/BMR. The mean DEI:TEE ratio was used to evaluate reported DEI at the population level, while individual misreporters were identified by applying the Goldberg cut-off points at three levels of PAL.

Based on the Goldberg method, 57 % of the study participants were identified as acceptable reporters of DEI, among whom mean TEE was 8·21 (95 % CI 8·01, 8·42) MJ (1963 (95 % CI 1914, 2012) kcal), mean DEI was 8·13 (95 % CI 7·93, 8·34) MJ (1944 (95 % CI 1895, 1993) kcal) and mean DEI:TEE was 1·01 (95 % CI 0·99, 1·04).

The dietary history and physical activity questionnaires provide comparable estimates of mean energy intake and expenditure at a population level. Acceptable reporters have to be identified in order to obtain better estimates. Questionnaire-based estimates of energy intake should not be interpreted without an inherent system of comparison or validation.

To evaluate the effectiveness of a pilot study of a parental school-based intervention to promote healthy eating behaviours in young children.

A quasi-experimental longitudinal design with three conditions (complete intervention (CIG), minimal intervention (MIG), control (CG)), with repeated measures at baseline, immediately after the intervention, 6 months and 1 year after intervention.

Fourteen public and state-funded kindergartens near Lisbon, Portugal.

Parents (n 349) of 3- to 6-year-old children assigned to the three conditions completed the baseline protocol. The ‘Red Apple’ intervention included four parental group sessions about young children’s growth, nutritional guidelines and positive parental feeding strategies, which was combined with adult–child activities at home and in the classroom, and newsletters (CIG). MIG included only a single nutritional counselling session, whereas the CG had no intervention. At the end, thirty-eight, twenty-six and fifty-four parents in the CIG, MIG and CG, respectively, had completed all evaluation components. Data regarding parental perception of children’s weight, self-efficacy, nutritional knowledge, feeding strategies, eating behaviours and BMI were collected at the four assessment moments.

The CIG showed improvements in children’s healthy food intake, compared with the MIG and CG. Parental self-efficacy regarding the regulation of children’s eating behaviours decreased in the CG but not in both intervention groups.

Considering the low dosage of the intervention, the results obtained were positive. Future studies should offer additional solutions to overcome barriers to parents’ participation.

We aimed to examine the association between red meat consumption and stroke in a group of Iranian adults.

A hospital-based case–control study.

The study included stroke patients and hospital-based controls. Usual dietary intakes of participants were assessed by means of a validated 168-item semi-quantitative FFQ. Total red meat consumption was calculated by summing up the consumption of red, processed and visceral meats.

One hundred and ninety-five cases were stroke patients hospitalized in the neurology ward and 195 controls were recruited from patients hospitalized in other wards with no history of cerebrovascular diseases or neurological disorders.

Participants with stroke were older, more likely to be male and less likely to be obese. Individuals in the highest tertile of red meat intake were 119 % more likely to have stroke (OR=2·19; 95 % CI 1·33, 3·60) compared with those in the lowest tertile. After controlling for age, sex and total energy intake, the association between red meat consumption and stroke was strengthened (OR=2·72; 95 % CI 1·53, 4·83). This association remained significant even after further controlling for physical activity and smoking as well as dietary intakes. Additional adjustments for BMI, diabetes, hypertension and hyperlipidaemia did not influence the association significantly (OR=2·51; 95 % CI 1·19, 5·09).

Consumption of red meat was associated with greater odds of having stroke in a group of Iranian adults.

To compare the contribution of different food consumption places to the energy and nutrient intake among Brazilian adolescents.

We analysed data from the Study of Cardiovascular Risk in Adolescents – ERICA, carried out in 2013–2014. Foods were categorised into thirty-three food groups. Energy, nutrients and food groups were analysed according to home, public and private schools, and other places of foods consumption. Linear regression models were used to test the association between eating away from home and nutrient intake.

Brazilian public and private schools.

All adolescents who had undergone anthropometric evaluation and provided information through the questionnaires, including the 24-h recall (n 71 740).

The main portion of energy intake per day was consumed within the adolescent households (8112·776 kJ (1939 kcal), 95 % CI 1892, 1985). Away-from-home eating was reported by 52 % of students in a given day, but it contributed to only 15 % of total energy intake. This contribution was mainly due to high energy-dense food intake. The percentage contribution of foods consumed at public school and other places was very similar with respect to nutrients. However, food consumption at other places meant less intake of protein, fibre and Fe, in addition to increasing the intake of added sugar and fats.

The frequency of food consumption outside the home by adolescents is high, although the main contribution to energy intake occurs at home, and despite the similarity of nutrients consumed in school and restaurants, the latter tends to worsen the nutritional quality of meals.