| Hall et al. 19 (2019) |
Randomized clinical trial |
Adults from the United States aged between 18 to 50 years old in 2018 (n = 20) (NIH Study). |
Ad libitum offer for two weeks of diets made with, on average, 83% of calories from ultra-processed foods or diets without ultra-processed foods. |
Daily energy intake and change in weight and body fat in two weeks. |
The ultra-processed and non-ultra-processed diets were paired for calories, energy density, macronutrients, sugar, sodium and fiber. |
When exposed to an ultra-processed diet, participants consumed, on average, 508 ± 106Kcal more per day than when exposed to diets without ultra-processed foods. At the end of two weeks, the participants increased 0.9kg ± 0.3kg in weight and 0.4kg ± 0.1kg in body fat consuming the ultra-processed diet and decreased 0.9kg ± 0.3kg in weight and 0.3kg ± 0.1kg of body fat consuming the non-ultra-processed diet (p-value < 0.001). |
High |
| Canhada et al. 20 (2019) |
Cohort |
Brazilian adults aged 35 years old or older with a mean follow-up of 3.8 years between 2008/2010-2012/2014 (n = 11,827) (ELSA-Brasil). |
Participation of ultra-processed foods in the food’s total energy. |
Incidence of overweight and obesity and weight gain and waist circumference. |
Age, sex, race/color, family income, education, physical activity and smoking status. |
Participants in the largest quartile of consumption of ultra-processed foods had a higher risk of weight gain (adjusted RR = 1.27; 95%CI: 1.07; 1.50) and waist circumference (adjusted RR = 1.33; 95%CI: 1.12; 1.58), excessive and higher incidence of overweight and obesity (adjusted RR = 1.20; 95%CI 1.03; 1.40) when compared to those in the lowest quartile. |
High |
| Mendonça et al. 21 (2016) |
Cohort |
Spanish middle-aged adults with a mean follow-up of 8.9 years between 1999-2012 (n = 8,541) (The Sun). |
Consumption of ultra-processed foods servings/day. |
Incidence of overweight and obesity. |
Age, sex, education, marital status, physical activity, smoking status, hours watching television, nap times, BMI at baseline, following a special diet at baseline, snacking between meals and consumption of fruits and vegetables. |
Participants in the largest quartile of consumption of ultra-processed foods had a higher risk of overweight/obesity (adjusted HR = 1.26; 95%CI: 1.10; 1.45) when compared to those in the lowest quartile. |
High |
| Rauber et al. 22 (2020) |
Cohort |
British adults aged between 40 to 69 years old with a mean follow-up of 5 years between 2006-2019 (n = 22,659) (UK Biobank). |
Participation of ultra-processed foods in the food’s total energy. |
Incidence of obesity and abdominal obesity and gain of BMI, waist circumference and body fat. |
Sex, socioeconomic deprivation index, physical activity, smoking status and hours of sleep. |
Participants in the highest quartile of consumption of ultra-processed foods had a higher risk of obesity (adjusted HR = 1.58; 95%CI: 1.32; 1.90) and high waist circumference (adjusted HR = 1.38; 95%CI: 1.21; 1.57) and to experience an increase of ≥ 5% in BMI (adjusted HR = 1.30; 95%CI: 1.19; 1.42), in waist circumference (adjusted HR = 1.30; 95%CI: 1.21; 1.40) and in the percentage of body fat (adjusted HR = 1.14; 95%CI: 1.04; 1.26) when compared to those with the lowest consumption quartile. |
High |
| Beslay et al. 23 (2020) |
Cohort |
French adults aged 18 years old or over and a mean follow-up of 4.1 years (n = 110,260) |
Participation of ultra-processed foods in the total of food grams. |
Incidence of overweight and obesity and BMI gain. |
Sex, age, marital status, education, physical activity, smoking status, number of dietary records, alcohol intake, energy intake. |
The consumption of ultra-processed foods was positively associated with a higher risk of overweight (adjusted HR associated with a 10% increase in the consumption of ultra-processed foods = 1.11; 95%CI: 1.08; 1.14), obesity (adjusted HR associated with a 10% increase in the consumption of ultra-processed foods = 1.09; 95%CI: 1.05; 1.13) and BMI gain (adjusted B associated with a 10% increase in the consumption of ultra-processed foods: 0.02; 95%CI: 0.01; 0.02) |
High |
| Adams & White 24 (2015) |
Cross-sectional |
British adults aged 18 years old or older in 2008-2012 (n = 2,174). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, overweight and obesity. |
Age, sex, social class and alcohol consumption. |
The consumption of ultra-processed foods was not significantly associated with the outcomes. |
Mean |
| Louzada et al. 25 (2015) |
Cross-sectional |
Brazilians aged 10 years old or older in 2008/2009 (n = 32,898). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, overweight and obesity. |
Age, sex, race/color, income, education, region, urbanity, physical activity, smoking status, intake of fruits, vegetables and beans. |
Participants in the largest quintile of consumption of ultra-processed foods had a higher mean BMI (adjusted coefficient: 0.94; 95%CI: 0.42; 1.42) and a greater chance of being overweight (adjusted OR = 1.26; 95%CI: 0.95; 1.69) and obesity (adjusted OR = 1.98; 95%CI: 1.26; 3.12) when compared to those in the lowest quintile. |
Mean |
| Juul et al. 26 (2018) |
Cross-sectional |
Adults from the United States aged between 20 to 64 years old in 2005-2014 (n = 15,977). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, waist circumference, overweight and obesity. |
Age, sex, ethnicity, socioeconomic status, education, marital status, physical activity and smoking status. |
Participants in the largest quintile of consumption of ultra-processed foods had a higher mean BMI (adjusted coefficient: 1.61; 95%CI: 1.11; 2.10) and waist circumference (adjusted coefficient: 4.07; 95%CI: 2.94; 5.19), greater chance of having obesity (adjusted OR = 1.53; 95%CI 1.29; 1.81), overweight (adjusted OR = 1.48; 95%CI: 1.25; 1.76) and abdominal obesity (adjusted OR = 1.62; 95%CI: 1.39; 1.89) when compared to those in the lowest quintile. |
Mean |
| Machado et al. 27 (2020) |
Cross-sectional |
Australian adults aged 20 years old or older between 2011-2012 (n = 7,411). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, waist circumference, obesity and abdominal obesity. |
Age, sex, income, education, area of residence, country of birth, physical activity and smoking status. |
Participants in the largest quintile of consumption of ultra-processed foods had a higher mean BMI (adjusted coefficient: 0.97; 95%CI: 0.42; 1.51), and waist circumference (adjusted coefficient: 1.92cm; 95%CI: 0.57; 3.27) and greater chance of having obesity (adjusted OR = 1.61; 95%CI: 1.27; 2.04) and abdominal obesity (adjusted OR = 1.38; 95%CI: 1.10; 1.72) when compared to those in the lowest quintile. |
Mean |
| Rauber et al. 28 (2020) |
Cross-sectional |
British adults aged 19 years old or older in 2008-2016 (n = 6,143). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, waist circumference, obesity and abdominal obesity. |
Age, sex, ethnicity, region, occupation, physical activity, smoking status, hours of sleep, year of research and whether they were on a diet for weight loss. |
Participants in the largest quartile of consumption of ultra-processed foods had a higher mean BMI (adjusted coefficient: 1.66; 95%CI: 0.96; 2.36) and waist circumference (adjusted coefficient: 3.56; 95%CI: 1.79; 5.33), and a greater chance of having obesity (adjusted O =: 1.90; 95%CI: 1.39; 2.61) when compared to those in the lowest quartile. |
Mean |
| Nardocci et al. 29 (2018) |
Cross-sectional |
Canadian adults aged 18 years old or older in 2004 (n = 19,363). |
Participation of ultra-processed foods in the food’s total energy. |
Overweight and obesity. |
Age, sex, income, education, immigration, area of residence, physical activity, smoking status, group's total energy intake and type of weight and height measurement (self-reported or directly measured). |
Participants in the largest quintile of consumption of ultra-processed foods were more likely to have obesity (adjusted OR = 1.32; 95%CI: 1.05; 1.57) when compared to those in the lowest quintile. |
Low |
| Seale et al. 30 (2020) |
Cross-sectional |
Canadian adults aged 18 years old or older in 2014/2015 (n = 10,942). |
Number of different types of ultra-processed foods consumed in the previous seven days. |
BMI. |
Age, sex, income and region. |
The consumption of ultra-processed foods was positively associated with BMI (adjusted coefficient: 0.04; 95%CI: 0.02; 0.07). |
Low |
| Julia et al. 31 (2018) |
Cross-sectional |
French adults aged 18 years old or older between 2009-2014 (n = 74,470). |
Participation of ultra-processed foods in the total of food grams. |
Overweight and obesity. |
Age, sex, income, education, marital status, smoking status, BMI and energy intake. |
Higher consumption of ultra-processed foods was significantly associated with overweight and obesity (p-value < 0.0001). |
Mean |
| Silva et al. 32 (2018) |
Cross-sectional |
Brazilian adults aged between 35 and 74 years in 2008-2010 (n = 8,977). |
Participation of ultra-processed foods in the food’s total energy. |
BMI, waist circumference, overweight, obesity, high waist circumference and significantly high waist circumference. |
Age, sex, race/color, family income per capita, physical activity, smoking status, hypertension, diabetes, consumption of fresh and minimally processed foods added to culinary ingredients and energy intake. |
Participants in the largest quartile of consumption of ultra-processed foods had a higher mean BMI (adjusted coefficient: 0.64; 95%CI: 0.33; 0.95) and waist circumference (adjusted coefficient: 0.95; 95%CI: 0.17; 1.74) and greater chance of overweight (adjusted OR = 1.32; 95%CI: 1.15; 1.53), obesity (adjusted OR: 1.43, 95%CI: 1.20; 1.72) and high waist circumference (OR = 1.21; 95%CI: 1.01; 1.46) when compared to those in the lowest quartile. |
Mean |
| Djupegot et al. 33 (2017) |
Cross-sectional |
Norwegian parents of 2-year-old children in 2014/2015 (n = 497). |
Frequency score of the consumption of ultra-processed foods. |
Overweight and obesity. |
Age, sex, education, shortage of time and number of children in the household. |
Overweight/obese people were more likely to have a high consumption of ultra-processed foods at dinner (adjusted OR = 1.54; 95%CI: 1.04; 2.30) when compared to those with normal weight. |
Low |
| Pan-American Health Organization 34 (2015) |
Ecological |
Latin American countries between 2000-2013 (n = 13). |
Total volume, in kg/per capita, of sales of ultra-processed foods. |
BMI trajectories. |
Population size, urbanization and gross national income. |
The increase in the per capita volume of sales of ultra-processed foods was significantly and positively associated with the average increase in the countries’ BMI (p-value < 0.001). |
Mean |
| Vandevijvere et al. 35 (2019) |
Ecological |
Countries of the Euromonitor food sales database between 2002-2014 (n = 80). |
Total volume, in kg/per capita, of sales of ultra-processed foods. |
BMI trajectories. |
National income per capita, education, urbanization, average consumption of fruits and vegetables in 2005, GINI index and, indirectly, physical activity. |
Increases in the per capita volume of sales of ultra-processed foods were significantly and positively associated with the BMI trajectories in the population (p-value < 0.001). |
Mean |
| Monteiro et al. 36 (2018) |
Ecological |
European countries between 1991-2008 (n =19). |
Participation of ultra-processed foods in the total energy available for consumption in households (national per capita). |
Prevalence of obesity in adults. |
GDP per capita, GDP2 per capita, years of difference between the estimates of obesity and availability of ultra-processed foods, obesity measurement method (self-reported or direct measurement), prevalence of physical inactivity and smoking status. |
The national household availability of ultra-processed foods was positive and significantly associated with the national prevalence of obesity among adults. The increase of a percentage in the availability of ultra-processed foods at home was associated with an increase of 0.25 percentage points in the prevalence of obesity. |
Low |
| Mendonça et al. 37 (2017) |
Cohort |
Spanish middle-aged adults with an average follow-up of 9.1 years between 1999-2012 (n = 14,790) (The Sun). |
Consumption of ultra-processed foods servings/day. |
Hypertension incidence. |
Age, sex, physical activity, smoking status, time watching TV, use of analgesics, family history of hypertension and hypercholesterolemia, following a special baseline diet, baseline BMI, energy intake, alcohol consumption, oil and fruit and vegetable intake. |
Participants in the highest tertile of consumption of ultra-processed foods had a higher risk of hypertension (adjusted HR = 1.21; 95%CI: 1.06; 1.37) when compared to those in the lowest tertile. |
High |
| Martínez Steele et al. 38 (2019) |
Cross-sectional |
American adults aged 20 years old or older between 2009-2014 (n = 6,385). |
Participation of ultra-processed foods in the food’s total energy. |
Metabolic syndrome. |
Age, sex, race/ethnicity, socioeconomic status, education, physical activity and smoking status. |
Participants in the largest quintile of consumption of ultra-processed foods had a higher prevalence of metabolic syndrome (adjusted PR = 1.28; 95%CI: 1.09; 1.50) when compared to those in the lowest quintile. |
Mean |
| Lavigne-Robichaud et al. 38 (2018) |
Cross-sectional |
Canadian indigenous people aged 18 years old or older (n = 811). |
Participation of ultra-processed foods in the food’s total energy, sodium and added sugar. |
Metabolic syndrome. |
Age, sex, area of residence, smoking status, energy intake and alcohol consumption. |
Participants in the largest quintile of consumption of ultra-processed foods were more likely to have metabolic syndrome (adjusted OR = 1.90; 95%CI: 1.14; 3.17) when compared to those in the lowest quintile. |
Low |
| Nasreddine et al. 40 (2018) |
Cross-sectional |
Lebanese adults aged 18 years old or older (n = 302). |
Participation of ultra-processed foods in the food’s total energy. |
Metabolic syndrome. |
Age, sex, income, education, marital status, area of residence, physical activity, smoking status, BMI and energy intake. |
The consumption of ultra-processed foods was not significantly associated with metabolic syndrome. |
Low |
| Lopes et al. 41 (2019) |
Cross-sectional |
Brazilian adults aged between 35 to 74 years old in 2008-2010 (n = 8,468). |
Participation of ultra-processed foods in the food’s total energy. |
Serum levels of C-reactive protein. |
Age, race/color, education, physical activity, smoking status and BMI. |
Women in the highest tertile of consumption of ultra-processed foods had higher serum levels of C-reactive protein (adjusted coefficient: 1.14; 95%CI: 1.04; 1.24) when compared to those in the lowest tertile in the model adjusted for age, race/color, education, physical activity, smoking status. The association lost significance when adjusted in the model additionally adjusted for BMI (adjusted coefficient: 1.00; 95%CI: 0.92; 1.08). No association was found between men. |
Mean |
| Montero-Salazar et al. 42 (2020) |
Cross-sectional |
Spanish adult men aged between 40 to 60 years old (n = 1,876) |
Grams of ultra-processed food. |
Coronary calcium score. |
Age, marital status, education, smoking status, physical activity, sleep duration, serum cholesterol, blood pressure, diabetes, BMI, alcohol intake, fiber intake, cholesterol and total energy. |
Men in the highest quartile of consumption of ultra-processed foods had a greater chance of high coronary calcium score (≥ 100) (adjusted OR = 2.0; 95%CI: 1.26; 3.16) when compared to those in the first quartile. |
Mean |
| Alonso-Pedrero et al. 43 (2020) |
Cross-sectional |
Spanish adults aged between 57 and 91 years old (n = 886) |
Ultra-processed food portions. |
Short telomeres. |
Age, sex, education, smoking status, physical activity, television time, family history of diabetes and cardiovascular diseases, prevalence of cancer, diabetes and dyslipidemia, BMI, energy intake. |
Participants in the largest quartile of consumption of ultra-processed foods were more likely to have short telomeres (adjusted OR = 1.82; 95%CI: 1.05; 3.22) when compared to those in the lowest quartile. |
Low |