Food advertising is an important determinant of unhealthy eating. However, analysing a large number of advertisements (ads) to distinguish between food and non-food content is a challenging task. This study aims to develop a machine learning-based method to automatically identify and classify food and non-food ad videos.

Methodological study to develop an algorithm model that prioritises both accuracy and efficiency in monitoring and classifying advertising videos.

From a collection of Brazilian television (TV) ads data, we created a database and split it into three sub-databases (i.e. training, validation and test) by extracting frames from ads. Subsequently, the training database was classified using the EfficientNet neural network. The best models and data-balancing strategies were investigated using the validation database. Finally, the test database was used to apply the best model and strategy, and results were verified with field experts.

The study used 2124 recorded Brazilian TV programming hours from 2018 to 2020. It included 703 food ads and over 20 000 non-food ads, following the protocol developed by the INFORMAS network for monitoring food marketing on TV.

The results showed that the EfficientNet neural network associated with the balanced batches strategy achieved an overall accuracy of 90·5 % on the test database, which represents a reduction of 99·9 % of the time spent on identifying and classifying ads.

The method studied represents a promising approach for differentiating food and non-food-related video within monitoring food marketing, which has significant practical implications for researchers, public health policymakers, and regulatory bodies.

To assess the prospective association of two diet quality scores based on the Nova food classification with BMI gain.

The NutriNet-Brasil cohort is an ongoing web-based prospective study with continuous recruitment of participants aged ≥ 18 years since January 2020. A short 24-h dietary recall screener including ‘yes/no’ questions about the consumption of whole plant foods (WPF) and ultra-processed foods (UPF) was completed by participants at baseline. The Nova-WPF and the Nova-UPF scores were computed by adding up positive responses regarding the consumption of thirty-three varieties of WPF and twenty-three varieties of UPF, respectively. Participants reported their height at baseline and their weight at both baseline and after approximately 15 months of follow-up. A 15-month BMI (kg/m2) increase of ≥5 % was coded as BMI gain.

Brazil.

9551 participants from the NutriNet-Brasil cohort.

Increasing quintiles of the Nova-UPF score were linearly associated with higher risk of BMI gain (relative risk Q5/Q1 = 1·34; 95 % CI 1·15, 1·56), whereas increasing quintiles of the Nova-WPF score were linearly associated with lower risk (relative risk Q5/Q1 = 0·80; 95 % CI 0·69, 0·94). We identified a moderate inverse correlation between the two scores (–0·33) and a partial mediating effect of the alternative score: 15 % for the total effect of the Nova-UPF score and 25 % for the total effect of the Nova-WPF score.

The Nova-UPF and Nova-WPF scores are independently associated with mid-term BMI gain further justifying their use in diet quality monitoring systems.

To investigate the advertising patterns on the posts of a fast-food chain in Brazil on three social media platforms in 2019.

An exploratory cross-sectional study.

Advertising strategies on the posts of a major fast-food chain on their official Facebook, Instagram and YouTube accounts. The strategies were investigated according to the INFORMAS protocol for food promotion monitoring. Principal component analysis (PCA) was employed to identify advertising patterns in each platform.

305 advertisements

Four advertising patterns were identified in the PCA of Facebook and Instagram. In both platforms, the components for kids and product exaltation were similar. On Facebook, a pattern corresponding to economic appeal was identified as price and discount, while on Instagram, this pattern also included a practical approach. On Facebook, the fourth component was named celebrity, while on Instagram it was celebrity/innovation since on this second social media the component also included the ‘new brand development’ variable. On YouTube, three advertising patterns were identified in PCA. Similar to the other platforms, the first and the second patterns were called for kids and price and discount, and the third component referred to both celebrity and commemorative dates.

The advertising patterns of the fast-food chain on three social media platforms were commonly directed to children and addressed price, discounts and the celebrities’ universe. The findings of this study corroborate other data in the literature regarding unhealthy food advertising on social media. This study discusses the urgency of regulating food advertising content on this medium.

To analyse the extent and nature of food and beverage advertising on the three major Brazilian free-to-air television (TV) channels.

Cross-sectional study. A protocol developed for the International Network for Food and Obesity/Non-Communicable Diseases Research, Monitoring and Action Support was applied for data collection. A total of 432 h of TV programming was recorded from 06.00 to 24.00 hours, for eight non-consecutive and randomly selected days, in April 2018. All TV advertisements (ads) were analysed, and food-related ads were classified according to the NOVA classification system. Descriptive analyses were used to describe the number and type of ads, food categories and the distribution of ads throughout the day and time of the day.

The three most popular free-to-air channels on Brazilian TV.

The study did not involve human subjects.

In total, 14·2 % (n 1156 out of 7991) of ads were food related (858 were specific food items). Approximately 91 % of food items ads included ultra-processed food (UPF) products. The top three most promoted products were soft drinks, alcoholic beverages and fast-food meals. Alcoholic beverage ads were more frequently broadcast in the evening.

The high risk of exposure of the Brazilian population to UPF ads should be considered a public health concern given the impact of unhealthy food advertising on people’s food choices and health.

To measure change in price of food groups over time (1995–2030) in Brazil, considering the Brazilian Dietary Guidelines’ recommendations.

Data from the Household Budget Survey (2008–2009 HBS) and the National System of Consumer Price Indexes (NSCPI) were used to create a data set containing monthly prices for the foods and beverages most consumed in the country (n 102), from January 1995 to December 2017. Data on price of foods and beverages from 2008–2009 HBS (referring to January 2009) were used to calculate real price over time using the monthly variation in prices from NSCPI. All prices were deflated to December 2017. Foods and beverages were classified following the Brazilian Dietary Guidelines’ recommendations. The monthly price for each food group and subgroup was used to analyse changes in prices from 1995 to 2017 and to forecast prices up to 2030 using fractional polynomial models.

Brazil.

National estimates of foods and beverages purchased for Brazil.

In 1995, ultra-processed foods were the most expensive group (R$ 6·51/kg), followed by processed foods (R$ 6·44/kg), then unprocessed or minimally processed foods and culinary ingredients (R$ 3·45/kg). Since the early 2000s, the price of ultra-processed foods underwent successive reductions, becoming cheaper than processed foods and reducing the distance between it and the price of the other group. Forecasts indicate that unhealthy foods will become cheaper than healthy foods in 2026.

Food prices in Brazil have changed unfavourably considering the Brazilian Dietary Guidelines’ recommendations. This may imply a decrease in the quality of the population’s diet.

To describe malnutrition (undernutrition and excess weight) by income, education and race/ethnicity in the Brazilian population.

Cross-sectional study.

Brazil.

Children aged <5 years (n 14 580), adolescents aged 11–19 years (n 31 892) and adults aged 20–49 years (n 84 660).

Among children, prevalence of excess weight, wasting and stunting was 16·9, 2·8 and 6·0 %, respectively. Differences related to income, education and race/ethnicity were verified, except for prevalence of wasting by education level. Girls and boys presented 18·4 and 20·5 % of excess weight, 2·8 and 3·7 % of underweight and 5·5 and 7·3 % of stunting, respectively. Prevalence of excess weight was lower among poorer, lower-educated (only for boys) and white adolescents, while stunting was lower among higher-income, higher-educated and white adolescents. Over three-quarters of women and almost half of men presented excess weight. Among adults, 3·9 % of women and 1·7 % of men were underweight, and 5·7 % of women and 0·2 % of men presented short stature. Prevalence of excess weight for women was higher among lower education and black, while for men it was higher among higher income and education and white. Short stature was more prevalent among black and mixed-race, low-educated and low-income women. Underweight prevalence was higher among low-educated, black and mixed-race women.

In Brazil, the prevalence of excess weight was at least threefold higher than that of undernutrition for children and adolescents and at least sevenfold higher for adults. Social inequalities were observed in the distribution of malnutrition across the lifespan and by gender.

To verify differences in the availability, variety, quality and price of unprocessed and ultra-processed foods in supermarkets and similar establishments in neighbourhoods with different social deprivation levels at Juiz de Fora, Minas Gerais, Brazil.

Cross-sectional study.

The Obesogenic Environment Study in São Paulo’s Food Store Observation Tool (ESAO-S) was applied in thirty-three supermarket chains, wholesale and retail supermarkets.

Fruits, vegetables and ultra-processed foods were available in almost all establishments, without differences according to Health Vulnerability Index (HVI; which varies from 0 to 1 point and the higher the worse; P > 0·05). Most establishments were concentrated in low vulnerability areas and offered healthy foods with greater variety and quality, despite higher prices. The Healthy Food Store Index (HFSI; which varies from 0 to 16 points and the higher the best) was calculated from the ESAO-S and the mean score was 8·91 (sd 1·51). The presence and variety of unprocessed foods count as positive points, as do the absence of ultra-processed products. When HFSI was stratified by HVI, low HVI neighbourhoods presented higher HFSI scores, compared with medium, high and very high HVI neighbourhoods (P = 0·001).

Supermarkets and similar establishments are less dense in areas of greater social deprivation and have lower prices of healthy foods, but the variety and quality of those foods are worse, compared with areas of low vulnerability. We found worse HFSI for supermarkets located in areas with greater vulnerability. Those findings can guide specific public policies improving the urban food environment.

To analyse trends in sweetened beverages consumption among adults in Brazil between 2007 and 2016.

A time-series analysis, with data from the Surveillance System of Risk and Protection Factors for Chronic Diseases by Telephone Survey (VIGITEL). The prevalence of regular consumption (≥5 d/week), the average daily consumption (millilitres) and the prevalence of non-consumption of these beverages were analysed. The temporal variations of the indicators were calculated by linear regression. The analyses were performed for the complete set of the evaluated population and stratified by sociodemographic characteristics.

Brazilian capitals and Federal District.

Brazilian adults aged ≥18 years (n 519 641).

There was a reduction in both regular and average daily consumption of sugar- and artificially sweetened beverages (−1·28 percentage points (pp)/year, P=0·001 and −9·63 ml/year, P=0·001, respectively). The same result regarding regular consumption was found when only sugar-sweetened beverages were analysed (−1·11 pp/year, P=0·011). Similar trends were identified in the stratified analyses, with a greater magnitude of reduction among males, young adults, those with higher schooling and residents of more developed regions. Coincidentally, there was an increase in the prevalence of adults who did not consume sweetened beverages (1·54 pp/year, P=0·005).

The consumption of sweetened beverages decreased during the period. However, a significant portion of the population still referred a daily consumption of these beverages.

To analyse the association between food store type and the consumption of ultra-processed products in Brazil.

Data from the 2008–2009 Household Budget Survey involving a probabilistic sample of 55 970 Brazilian households. Food stores were grouped into nine categories. Foods and drinks were grouped according to characteristics of food processing. The contribution of each food store type to the total energy acquired from each food processing group, and according to quintiles of consumption of ultra-processed products, was estimated. Exploratory factor analysis was conducted to identify a pattern of food store usage. Linear regression models were performed to estimate the relationship between the purchase pattern and the consumption of ultra-processed products.

In line with their larger market share, supermarkets accounted for 59 % of total energy and participated most in acquisition for three food groups, with emphasis on ultra-processed products (60·4 % of energy). The participation of supermarkets in total purchase tended to increase in populations with higher consumption of ultra-processed products, while the participation of small markets and small producers tended to decrease. The purchase pattern characterized by use of traditional retail (street fairs and vendors, small markets, small farmers, butcheries) was associated with a smaller consumption of ultra-processed products.

Food policies and interventions aiming to reduce the consumption of ultra-processed products should consider the influence of supermarkets on the consumption of these products. A purchase pattern based on traditional retail constitutes an important tool for promoting healthy eating in Brazil.

To present and discuss the dietary guidelines issued by the Brazilian government in 2014.

The present paper describes the aims of the guidelines, their shaping principles and the approach used in the development of recommendations. The main recommendations are outlined, their significance for the cultural, socio-economic and environmental aspects of sustainability is discussed, and their application to other countries is considered.

Brazil in the twenty-first century.

All people in Brazil, now and in future.

The food- and meal-based Brazilian Dietary Guidelines address dietary patterns as a whole and so are different from nutrient-based guidelines, even those with some recommendations on specific foods or food groups. The guidelines are based on explicit principles. They take mental and emotional well-being into account, as well as physical health and disease prevention. They identify diet as having cultural, socio-economic and environmental as well as biological and behavioural dimensions. They emphasize the benefits of dietary patterns based on a variety of natural or minimally processed foods, mostly plants, and freshly prepared meals eaten in company, for health, well-being and all relevant aspects of sustainability, as well as the multiple negative effects of ready-to-consume ultra-processed food and drink products.

The guidelines’ recommendations are designed to be sustainable personally, culturally, socially, economically and environmentally, and thus fit to face this century. They are for foods, meals and dietary patterns of types that are already established in Brazil, which can be adapted to suit the climate, terrain and customs of all countries.

To analyse the dietary intake of Brazilian adolescents and investigate its association with sociodemographic factors as well as health-risk and health-protective behaviours.

Cross-sectional study.

The study was based on data supplied by the National Survey of Schoolchildren’s Health (2012) on sociodemographic factors, dietary intake and health-risk and health-protective behaviours of schoolchildren in Brazil. A nutritional scale was elaborated combining markers of healthy and unhealthy diets. Poisson regression analysis was applied to investigate the association between the sociodemographic factors and regular intake (≥5 times/week) of selected foods; linear regression analysis was applied to investigate the association of sociodemographic and behavioural factors with nutritional scale score.

A total of 109 104 adolescents attending the ninth year of education at 2842 schools in Brazil.

Fewer than 30 % of the adolescents consumed raw or cooked vegetables on a regular basis, whereas more than one-third reported regular intake of sweets, soft drinks and sweet biscuits. Adolescents from the southern area and the oldest ones were those most exposed to inadequate dietary intake. The nutritional scale average score was higher in the students attending public school and exhibited a positive correlation with protective behaviours, such as being physically active, having meals with parents and eating breakfast, and a negative correlation with risk behaviours such as eating while studying or watching television and having smoked, drunk alcohol or used other drugs in the previous 30 d.

The results indicate an association between undesirable nutritional habits and other risk behaviours among Brazilian adolescents.

To identify behavioural patterns of protective and risk factors for non-communicable diseases (NCD) and to explore the association between these patterns and sociodemographic characteristics.

Principal component analysis was used to identify behavioural patterns from a list of twelve protective and risk factors for NCD. Linear regression was used to explore the association between the patterns and sociodemographic characteristics.

Participants from the Brazilian Surveillance System of Risk and Protective Factors for NCD through Telephone Interviews (VIGITEL) from the years 2009 and 2010 were included.

A sample of 108 706 adults was included in the analysis.

Two behavioural patterns were identified in the analysis, a ‘prudent pattern’ and a ‘risky pattern’. The first involved mostly protective behaviours, while the second one involved essentially the risky ones. Both the less prudent and the more risky behaviour patterns were concentrated in younger men, with lower education, from the more developed region.

Public policies to decrease NCD should be aware of the possible tendency towards behavioural patterns in order to be more effective.

To investigate consumption of ultra-processed products in Canada and to assess their association with dietary quality.

Application of a classification of foodstuffs based on the nature, extent and purpose of food processing to data from a national household food budget survey. Foods are classified as unprocessed/minimally processed foods (Group 1), processed culinary ingredients (Group 2) or ultra-processed products (Group 3).

All provinces and territories of Canada, 2001.

Households (n 5643).

Food purchases provided a mean per capita energy availability of 8908 (se 81) kJ/d (2129 (se 19) kcal/d). Over 61·7 % of dietary energy came from ultra-processed products (Group 3), 25·6 % from Group 1 and 12·7 % from Group 2. The overall diet exceeded WHO upper limits for fat, saturated fat, free sugars and Na density, with less fibre than recommended. It also exceeded the average energy density target of the World Cancer Research Fund/American Institute for Cancer Research. Group 3 products taken together are more fatty, sugary, salty and energy-dense than a combination of Group 1 and Group 2 items. Only the 20 % lowest consumers of ultra-processed products (who consumed 33·2 % of energy from these products) were anywhere near reaching all nutrient goals for the prevention of obesity and chronic non-communicable diseases.

The 2001 Canadian diet was dominated by ultra-processed products. As a group, these products are unhealthy. The present analysis indicates that any substantial improvement of the diet would involve much lower consumption of ultra-processed products and much higher consumption of meals and dishes prepared from minimally processed foods and processed culinary ingredients.

To assess time trends in the contribution of processed foods to food purchases made by Brazilian households and to explore the potential impact on the overall quality of the diet.

Application of a new classification of foodstuffs based on extent and purpose of food processing to data collected by comparable probabilistic household budget surveys. The classification assigns foodstuffs to the following groups: unprocessed/minimally processed foods (Group 1); processed culinary ingredients (Group 2); or ultra-processed ready-to-eat or ready-to-heat food products (Group 3).

Eleven metropolitan areas of Brazil.

Households; n 13 611 in 1987–8, n 16 014 in 1995–5 and n 13 848 in 2002–3.

Over the last three decades, the household consumption of Group 1 and Group 2 foods has been steadily replaced by consumption of Group 3 ultra-processed food products, both overall and in lower- and upper-income groups. In the 2002–3 survey, Group 3 items represented more than one-quarter of total energy (more than one-third for higher-income households). The overall nutrient profile of Group 3 items, compared with that of Group 1 and Group 2 items, revealed more added sugar, more saturated fat, more sodium, less fibre and much higher energy density.

The high energy density and the unfavourable nutrition profiling of Group 3 food products, and also their potential harmful effects on eating and drinking behaviours, indicate that governments and health authorities should use all possible methods, including legislation and statutory regulation, to halt and reverse the replacement of minimally processed foods and processed culinary ingredients by ultra-processed food products.

To study the role of energy derived from sugar (both table sugar and sugar added to processed foods) in the total energy content of food purchases in Brazil.

Food purchase data were collected during a national household budget survey carried out between June 2002 and July 2003 on a probabilistic sample representative of all households in the country. The amount of food purchased in this 12-month period was transformed into energy and energy from sugar using food composition tables. Multiple linear regression models were used to study the association between amount of energy from sugar and total energy content of food purchases, controlling for sociodemographic variables and potential interactions between these variables and sugar purchases.

There was a positive and significant association between energy from sugar and total household energy purchases. A 1 kJ increase in sugar purchase corresponded to a 3·637 kJ increase in total energy. In the absence of expenditure on meals outside the home, i.e. when household food purchases tend to approximate actual food consumption by household members, sugar purchase of 1926·35 kJ/d (the 90th percentile of the distribution of sugar purchases in Brazil) was associated, depending on income strata, with total energy purchase over 40–60 % of the recommended daily value for energy intake in Brazil.

The present results corroborate the recommendations of the WHO and the Brazilian Ministry of Health regarding limiting the consumption of sugar.