Abstract
Background
The consumption of ultra-processed foods (UPF) has been associated with adverse health effects. The predictors of consumption, particularly in high-intake groups like children and adolescents, are scarcely understood.
Methods
We evaluated UPF consumption and associated factors among male and female Spanish children aged 8–12 years participating in the INMA birth cohort. We cross-sectionally analyzed data from 1565 mother–child pairs at age 8 and 314 pairs at age 12. Children’s diets were assessed using validated food frequency questionnaires. UPF consumption was defined using the NOVA classification and categorized into tertiles. The associations between maternal-child characteristics and children’s UPF consumption were analyzed using multinomial logistic regression to estimate relative risk ratios (RRR) and 95% confidence intervals (95%CI).
Results
At age 8, high UPF consumption was associated with low maternal social class, RRR = 2.22 (1.46–3.37), moderate maternal UPF consumption during pregnancy, RRR = 1.87 (1.35–2.58), high maternal UPF consumption during pregnancy, RRR = 3.47 (2.40–4.99), being male, RRR = 1.76 (1.34–2.30), and high children’s TV viewing, RRR = 1.76 (1.19–2.60). At age 12, high UPF consumption was associated with high maternal UPF consumption during pregnancy, RRR = 3.09 (1.27–7.50), and high children’s TV viewing, RRR = 2.34 (1.05–5.21), although it was inversely associated with maternal university education, RRR = 0.21 (0.06–0.69), and high children’s physical activity, RRR = 0.43 (0.20–0.93).
Conclusions
Maternal factors such as high UPF consumption during pregnancy, low social class and low education level, and low physical activity and TV watching during childhood and adolescence, were predictors of higher UPF consumption at the age of 8 and 12 years.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28008-6.
Keywords: Ultra-Processed Foods, Birth cohort study, Children, Adolescents, Sex-specific, Determinants
Introduction
Ultra-processed food (UPF) consumption is associated with several detrimental health effects in all vital life stages such as childhood, adolescence and adulthood [1]. A recent umbrella review and meta-analysis, which included 122 original articles, concluded that high UPF consumption was associated with renal function decline and wheezing in children and adolescents [2]. In addition, high UPF consumption in this population has been associated with elevated cardiometabolic risk indicators, such as higher BMI z-scores, waist circumference, and fasting plasma glucose [3, 4].
Despite their adverse health effects, the UPF consumption is increasing rapidly worldwide [5]. In Spain, UPF have been reported to account for around one third of children’s daily energy intake [6], a proportion that is even higher in other countries such as Canada, Jordan and USA, where UPF consumption may approach nearly half of the total dietary intake [4, 7, 8]. This high consumption may be related to the fact that they are industrial products designed to be time-saving (e.g., ready-to-eat or ready-to-heat), hyperpalatable, durable, and low-cost when compared with minimally processed options per 100 kcals [9, 10].
Although these factors may affect child’s and adolescent’s exposure to UPF, other individual and family factors could also play an important role in their consumption. In children, high UPF consumption was associated with older child age, greater number of hours watching television (TV) and a higher intake of soft drinks/fast food by parents [7]. In adolescents, high UPF consumption was associated with playing electronic games, using smartphones on weekdays and increased time spent engaged in sedentary activities, such as watching TV [11]. In line with these findings, we recently published a study in 4-year-old Spanish children and we found that greater UPF consumption was related to higher maternal UPF consumption during pregnancy, younger maternal age, and increased TV viewing in children [12].
However, evidence on the factors associated with UPF consumption beyond early childhood remains limited, particularly in Spain and when considering potential sex-specific patterns. Thus, the aim of this study was to describe UPF consumption according to the NOVA classification system and to evaluate the association with sociodemographic, behavioral and maternal factors associated, overall and by sex, with UPF consumption in Spanish children aged 8 and 12 years.
Methods
Study design and population
This cross-sectional study is based on data obtained from children at age 8 and 12-year in the follow-up visits within the framework of the INMA Project (INfancia y Medio Ambiente – Environment and Childhood), a Spanish prospective population-based birth cohort (https://www.proyectoinma.org/) [13].
Participant recruitment took place between November 2003 and February 2008, and targeted pregnant women attending their first prenatal consultation in public hospitals which were collaborating in the INMA project in four specific regions (Asturias, Gipuzkoa, Sabadell, and Valencia). Eligibility criteria were: to be over 16 years of age, to plan to give birth at a designated reference hospital, to have no language barriers, to have a singleton pregnancy, and to have conceived naturally (without assisted reproductive techniques). A total of 2,626 live births were registered between May 2004 and August 2008. The 8-year follow-up visit was conducted between February 2012 and January 2017 across the four study regions (Asturias, Gipuzkoa, Sabadell, and Valencia). A total of 1,565 mother–child pairs were analysed with complete data available after losses to follow-up (n = 870) and exclusion due to missing data on variables of interest (n = 191). The 12-year follow-up visit was conducted between June 2017 and December 2019 across the two regions (Asturias and Gipuzkoa) with dietary information, as Sabadell and Valencia did not collect dietary information at this visit. A total of 314 mother–child pairs were analyzed with complete data available after losses to follow-up (n = 1,150), lack of dietary information (n = 1,130), and exclusion due to missing data on variables of interest (n = 32). Detailed information on participant selection and flow is provided in Fig. 1. Given the differences in sample size and data availability between the two follow-up visits, analyses were conducted treating each visit as an independent cross-sectional sample.
Fig. 1.
Flowchart of participants in the INfancia y Medio Ambiente (INMA) Study
Dietary assessment and UPF classification
The dietary intake was assessed at the 8 and 12-year follow-up visits through structured, face-to-face interviews with parents or primary caregivers, using the same validated 46-item semi-quantitative food frequency questionnaire (FFQ) [14]. The FFQ collected habitual intake over the previous year and provided nine frequency response options, ranging from “Never or less than once per month” to “Six or more times per day”. Standard portion sizes appropriate for each age were specified for each item, and food consumption in grams was obtained by multiplying the portion size specified in the FFQ by the frequency of consumption reported by participants in the FFQ. Nutrient and energy intake estimates were then calculated using both the United States Department of Agriculture Nutrient Database for Standard Reference [15] and Spanish food composition Table [16].
Each food item listed in the FFQ was assigned to the corresponding categories of the NOVA classification system based on the nature, extent, and purpose of industrial processing [17]. A multidisciplinary expert panel consisting of dietitians and nutritional epidemiologists conducted the classification through consensus. The four NOVA groups were defined as follows: (1) Unprocessed or minimally processed foods: natural foods or those that have undergone minimal processing, such as cleaning or freezing (e.g., fruits, vegetables, legumes, fresh meat, eggs); (2) Processed culinary ingredients: foods used for preparations, seasoning and cooking, such as oils, sugar, salt, and butter; (3) Processed foods: products typically made by adding salt, sugar, or other culinary ingredients to unprocessed foods, which often involves preservation methods like fermentation or canning (e.g., bread, cheese, canned vegetables); (4) UPF: industrial products, which typically contain additives, flavourings, colourings, and other cosmetic or technological substances and often have little or no whole food content. This group includes products such as carbonated soft drinks, packaged snacks, processed meats, industrial pastries, ready-to-eat meals, and commercial sauces. A complete list of FFQ food items and their respective NOVA classification is available in Supporting Information (Additional file 1). Since the same FFQ was used for both the 8- and 12-year visits, the classification of food items using the NOVA system was consistent for both visits.
To quantify UPF consumption, we identified 18 food items classified within NOVA Group 4. These were further grouped into five predefined UPF subcategories: dairy desserts, processed meats, fast food, sweet foods, and beverages (Additional file 2). Total UPF consumption was calculated in grams per day by summing the intake in grams of all items in this group. To account for individual differences in total food consumption, UPF consumption was expressed as a percentage of total daily food intake (excluding water), using the following formula: 
. This relative measure of UPF consumption was then categorized into tertiles of UPF consumption for analysis. Energy contribution from UPF was not computed in this variable, as some items in this category (e.g., artificially sweetened beverages) are non-caloric, and therefore the use of energy-based proportions could result in an underestimation of total UPF consumption.
Other variables
Information on relevant covariates was collected through standardized questionnaires administered by trained personnel during both prenatal and childhood and adolescence follow-up visits. Maternal data collected during two prenatal visits (at 12 and 34 weeks of gestation) and included age (years), educational level (primary or less, secondary, or university), and social class (high, medium, or low) which was classify according to the Spanish adaptation of the ISCO88 classification system [18]. Maternal pre-pregnancy weight and height were self-reported and used to calculate body mass index (BMI), then BMI was categorized as normal weight (< 25 kg/m²), overweight (25–29.99 kg/m²), or obesity (≥ 30 kg/m²). Maternal dietary intake during pregnancy was collected using a validated 101-item FFQ at both 12 and 32 weeks of gestation visits [19]. All items of the FFQ were also grouped following the NOVA classification and maternal UPF consumption was calculated as the mean consumption across two time points during pregnancy (12 and 32 weeks of gestation) and categorized into tertiles (low, moderate, high). During the 8 and 12-year follow-up visits, information on the child’s sex (female, male), age (years), tertiles of TV viewing (hours/day), and tertiles of physical activity (hours/day) was also obtained. In addition, dietary intake at age 4 was also collected using a validated 105-item FFQ [20]. All items of the FFQ were also grouped following the NOVA classification and children’s UPF consumption at 4 years-old was categorized into tertiles (low, moderate, high). In the present study, we focused on the 8- and 12-year visits because the use of different dietary assessment tools at age 4 limits the direct comparability of UPF estimates across all three visits.
Statistical analyses
Descriptive statistics were used to summarize sociodemographic and lifestyle characteristics for both the total sample and by tertiles of UPF consumption at 8 and 12 years. Means and standard deviations (SD) were used to describe continuous variables while absolute and relative frequencies were used for categorical variables. Comparisons across UPF consumption tertiles were conducted using ANOVA for continuous variables and Chi-square tests for categorical variables.
We described the UPF consumption for the total sample and stratified it by child’s sex using both the median (interquartile range, IQR) and the mean (standard deviation, SD), since the Shapiro–Wilk test showed a non-normal distribution of UPF consumption. This approach ensures statistical appropriateness and consistency with previously published findings.
Multinomial logistic regression models were applied separately to the 8‑year and 12‑year visits, to explore the associations between sociodemographic and lifestyle characteristics and tertiles of UPF consumption (low, moderate, and high) at each visit. This approach was chosen to estimate category-specific associations for moderate and high UPF consumption, using the low tertile as the reference category, without imposing the proportional odds assumption required by ordinal logistic regression [21]. Results are presented as Relative Risk Ratios (RRR) with their corresponding 95% confidence intervals (95%CIs) and the low tertile of UPF consumption was used as the reference category. Multivariable models included variables identified as potential confounders in previous literature such as cohort (Asturias, Gipuzkoa, Sabadell, Valencia), and by mothers’ characteristics such as age (in years), educational level (primary or less, secondary, university), social class (low, medium, high), and pre-pregnancy weight status (normal weight, overweight, obesity), as well as by children’s characteristics such as sex (male, female), age (in years), TV viewing (hours/day), and physical activity (hours/day). In addition, given that previous research has documented sex-related differences in dietary behaviors [22, 23] we stratified analyses by sex to further investigate potential sex-specific patterns.
Sensitivity analyses were performed to evaluate the effect of UPF consumption at the previous visit. Thus, at 8 years of age, analyses are additionally adjusted by the consumption of UPF at 4 years old. At 12 years of age, analyses are additionally adjusted by the UPF consumption at 8 years old.
All analyses were conducted using STATA software (version 18, StataCorp LLC, College Station, TX, USA; http://www.stata.com). A two-sided p-value < 0.05 was used as a threshold for statistical significance.
Results
Main characteristics of sample
We analysed data from 1565 participants at age 8 and 314 at age 12 (Table 1). At both ages, mothers were mostly ≥ 30 years old, of low social class, and with normal weight before pregnancy. Maternal energy intake was similar at both timeframes (≈ 2,050 kcal/d), while the proportion of UPF decreased, from one visit to the other, from 14.3% to 11.3% of total intake. Children and adolescents showed a mean age of 7.6 and 11.9 years, respectively. Between 8 and 12 years, physical activity and TV viewing decreased, whereas energy intake increased from 1844 to 2114 kcal/d. Characteristics stratified by tertiles of UPF consumption at age 8 and 12 are shown in Additional file 3 and Additional file 4 respectively.
Table 1.
Sociodemographic characteristics of participants at 8 and 12 years-old visits from the INMA study
| 8-year-old visit (n = 1565) |
12-year-old visit (n = 314) |
|
|---|---|---|
| Mother characteristics | ||
| INMA Cohort, n(%) | ||
| Asturias | 297 (19.0) | 193 (61.5) |
| Gipuzkoa | 371 (23.7) | 121 (38.5) |
| Sabadella | 460 (29.4) | - |
| Valenciaa | 437 (27.9) | - |
| Age (years)b, n(%) | ||
| < 30 | 539 (34.4) | 78 (24.8) |
| ≥ 30 | 1026 (65.6) | 236 (75.2) |
| Educational levelb, n(%) | ||
| Primary or less | 299 (19.1) | 36 (11.5) |
| Secondary | 646 (41.3) | 135 (43.0) |
| University | 620 (39.6) | 143 (45.5) |
| Social classb,c, n(%) | ||
| High | 395 (25.2) | 89 (28.3) |
| Medium | 449 (28.7) | 75 (23.9) |
| Low | 721 (46.1) | 150 (47.8) |
| Pre-pregnancy weight status (kg/m2), n(%) | ||
| Normal weight (< 25.0) | 1149 (73.4) | 242 (77.1) |
| Overweight (25.0-29.9) | 298 (19.0) | 53 (16.9) |
| Obesity (≥ 30.0) | 118 (7.5) | 19 (6.1) |
| Energy intake (kcal/d)b, Mean (SD) | 2051 (446) | 2040 (404) |
| UPF consumption (% grams/d)b,d | ||
| Mean (SD) | 14.3 (8.1) | 11.3 (6.5) |
| Tertile 1 low (< 10.4), n(%) | 574 (36.7) | 170 (54.1) |
| Tertile 2 moderate (10.4–17.1), n(%) | 543 (34.7) | 90 (28.7) |
| Tertile 3 high (> 17.1), n(%) | 448 (28.6) | 54 (17.2) |
| Child characteristics | ||
| Sex, n(%) | ||
| Female | 760 (48.6) | 153 (48.7) |
| Male | 805 (51.4) | 161 (51.3) |
| Age (in years)e, Mean (SD) | 7.6 (0.6) | 11.9 (0.8) |
| Physical activity (hours/d)e | ||
| Mean (SD) | 1.9 (1.1) | 1.6 (0.8) |
| Tertile low, n(%) | 535 (34.2) | 99 (31.5) |
| Tertile moderate, n(%) | 495 (31.6) | 111 (35.4) |
| Tertile high, n(%) | 535 (34.2) | 104 (33.1) |
| Television viewing (hours/d)e,f | ||
| Mean (SD) | 1.8 (1.0) | 1.3 (1.1) |
| Tertile 1 low, n(%) | 436 (27.9) | 116 (36.9) |
| Tertile 2 moderate, n(%) | 342 (21.9) | 102 (32.5) |
| Tertile 3 high, n(%) | 327 (20.9) | 96 (30.6) |
| Total energy intake (kcal/d)e, Mean (SD) | 1844 (525) | 2114 (649) |
| UPF consumption at 4 years-old (% grams/d)d | ||
| Mean (SD) | 32.6 (14.9) | 25.0 (13.1) |
| Tertile 1 low, n(%) | 489 (33.8) | 169 (57.5) |
| Tertile 2 moderate, n(%) | 483 (33.4) | 80 (27.2) |
| Tertile 3 high, n(%) | 475 (32.8) | 45 (15.3) |
d day, UPF Ultra-processed foods
a Sabadell and Valencia were excluded at 12-year-old visit because dietary information was not available
b During pregnancy
c Social class according to the Spanish adaptation of the ISCO88 classification system
dThe consumption of UPF was expressed as a percentage of total daily dietary intake [intake of group (g/d)/ total dietary intake (g/d)*100]
e At the time of the visit
f Information of television viewing time was not collected in Sabadell cohort at the visit of 8 years (n = 460, 29.4%)
UPF consumption in children and adolescents
At age 8, the mean UPF consumption in the total sample was 282.2 (SD 193.8) g/day, which represented 25.4% of total daily dietary intake (Table 2). Within the total UPF consumed, the main contributors were the sweet foods subgroup (28.3%) and the beverages subgroup (27.6%). When exploring differences by sex, UPF consumption at 8 years was significantly higher in male children (305.3, SD 215.7 g/day) than in female children (257.8, SD 164.1 g/day; p < 0.001), with a percentage of UPF in total daily dietary intake of 26.8% and 23.8% (p < 0.001), respectively. Sweet foods subgroup was the main contributor for total UPF consumption among females (28.8%), while among males, it was the beverages subgroup (28.4%).
Table 2.
Consumption of UPF in 8-year-old children from INMA study, overall sample and stratified by sex
| Total sample (n = 1565) |
Females (n = 760) |
Males (n = 805) |
||||||
|---|---|---|---|---|---|---|---|---|
| Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | p-value3 | p-value4 | |
| Total UPF consumption | ||||||||
| Grams/day | 282.2 (193.8) | 237.0 (173.1) | 257.8 (164.1) | 224.0 (155.6) | 305.3 (215.7) | 251.1 (199.9) | < 0.001 | < 0.001 |
| % of UPF in total daily dietary intake1 | 25.4 (11.6) | 23.7 (15.3) | 23.8 (10.5) | 22.4 (13.2) | 26.8 (12.4) | 24.8 (16.7) | < 0.001 | < 0.001 |
| Food groups within UPF | ||||||||
| Dairy desserts | ||||||||
| Grams/day | 34.4 (49.3) | 34.3 (29.0) | 30.7 (45.1) | 11.4 (29.0) | 38.0 (52.8) | 34.3 (29.0) | < 0.001 | 0.003 |
| % within total UPF2 | 11.8 (11.9) | 8.4 (14.6) | 11.5 (11.7) | 7.7 (13.7) | 12.2 (12.0) | 8.9 (15.2) | 0.166 | 0.222 |
| Processed meats | ||||||||
| Grams/day | 31.1 (20.4) | 28.1 (19.8) | 29.7 (17.8) | 28.1 (17.0) | 32.4 (22.5) | 28.1 (21.3) | 0.036 | 0.045 |
| % within total UPF2 | 13.4 (8.5) | 11.7 (9.6) | 13.8 (8.6) | 12.1 (9.7) | 13.0 (8.4) | 11.3 (9.4) | 0.113 | 0.018 |
| Fast food | ||||||||
| Grams/day | 46.4 (30.6) | 40.0 (31.5) | 43.0 (24.1) | 38.4 (30.0) | 49.5 (35.5) | 41.2 (33.7) | < 0.001 | 0.001 |
| % within total UPF2 | 18.9 (10.2) | 17.0 (13.1) | 19.2 (10.3) | 17.2 (13.2) | 18.7 (10.2) | 16.6 (13.1) | 0.470 | 0.331 |
| Sweet foods | ||||||||
| Grams/day | 69.3 (46.6) | 57.8 (46.9) | 65.4 (40.4) | 57.3 (40.9) | 73.0 (51.5) | 57.8 (51.5) | 0.036 | 0.070 |
| % within total UPF2 | 28.3 (15.6) | 26.0 (21.1) | 28.8 (15.3) | 26.4 (20.2) | 27.8 (15.9) | 25.4 (21.8) | 0.093 | 0.108 |
| Beverages | ||||||||
| Grams/day | 101.1 (145.6) | 55.4 (101.0) | 89.0 (128.3) | 42.0 (101.0) | 112.4 (159.5) | 57.2 (143.8) | 0.004 | 0.006 |
| % within total UPF2 | 27.6 (21.9) | 24.5 (34.5) | 26.8 (21.4) | 22.7 (32.5) | 28.4 (22.2) | 26.3 (35.9) | 0.162 | 0.191 |
Dairy desserts include custard, pudding or similar. Processed meats include ham (cured and cooked), sausages, dried sausage, chorizo or similar. Sweets include breakfast cereals, pastries, chocolate and derivates, sugar, honey, jams, candies, and sweets. Fast-foods includes fish derivates, French fries, margarine and butter, commercial mayonnaise, fried tomato sauce, ketchup, pizza. Beverages includes soft drinks (sugar- and artificial-sweetened) and packaged fruit juices
IQR Interquartile range, SD Standard Deviation, UPF Ultra-processed foods
1The UPF consumption was expressed as a percentage of total daily dietary intake in grams (g) (intake of UPF (g/d)/ total dietary intake (g/d)*100)
2The consumption of each food sub-group was expressed as a percentage of total UPF intake in g (intake of each food sub-group (g/d)/ total UPF intake (g/d)*100)
3P-value from ANOVA test
4P-value from Kruskal-Wallis’s test
At age 12 we observed a lower UPF consumption than the observed at age 8 (Table 3), both in daily grams (251.6 g/day, SD 178.2) and in the percentage of UPF from the total dietary intake (20.2%). Sweet foods subgroup was the major contributor (34.9%) of UPF in all children, and in females (35.4%) and males (34.5%) (Table 3).
Table 3.
Consumption of UPF in 12-year-old children from INMA study, overall sample and stratified by sex
| Total sample (n = 314) |
Females (n = 153) |
Males (n = 161) |
||||||
|---|---|---|---|---|---|---|---|---|
| Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | p-value3 | p-value4 | |
| Total UPF consumption | ||||||||
| Grams/day | 251.6 (178.2) | 203.0 (173.9) | 247.4 (190.3) | 206.6 (125.9) | 255.7 (166.4) | 200.7 (220.2) | 0.724 | 0.986 |
| % of UPF in total daily dietary intake1 | 20.2 (10.5) | 18.4 (13.8) | 20.4 (9.9) | 19.0 (12.9) | 20.0 (11.0) | 18.2 (15.8) | 0.842 | 0.458 |
| Food groups within UPF | ||||||||
| Dairy desserts | ||||||||
| Grams/day | 22.1 (32.3) | 6.7 (36.2) | 22.1 (32.5) | 6.7 (36.2) | 22.2 (32.2) | 6.7 (42.9) | 0.810 | 0.845 |
| % within total UPF2 | 8.3 (9.9) | 4.7 (10.4) | 8.4 (10.2) | 5.2 (9.7) | 8.1 (9.7) | 4.5 (11.3) | 0.666 | 0.744 |
| Processed meats | ||||||||
| Grams/day | 38.1 (24.7) | 33.8 (28.6) | 38.6 (23.0) | 33.8 (22.2) | 37.7 (26.2) | 33.8 (31.3) | 0.598 | 0.374 |
| % within total UPF2 | 18.3 (12.4) | 16.0 (14.8) | 18.9 (12.5) | 16.9 (13.4) | 17.7 (12.3) | 15.4 (15.4) | 0.357 | 0.288 |
| Fast food | ||||||||
| Grams/day | 42.1 (41.5) | 32.1 (26.3) | 42.0 (33.9) | 32.9 (24.0) | 42.1 (47.7) | 31.5 (27.9) | 0.889 | 0.387 |
| % within total UPF2 | 18.9 (11.7) | 16.2 (12.1) | 19.0 (11.1) | 16.2 (12.1) | 18.9 (12.3) | 16.5 (12.5) | 0.948 | 0.774 |
| Sweet foods | ||||||||
| Grams/day | 80.7 (65.1) | 69.2 (59.0) | 79.3 (60.0) | 70.1 (59.1) | 82.0 (69.7) | 67.6 (61.1) | 0.949 | 0.890 |
| % within total UPF2 | 34.9 (18.4) | 32.7 (27.9) | 35.4 (19.0) | 32.9 (29.2) | 34.5 (17.8) | 32.3 (27.0) | 0.379 | 0.719 |
| Beverages | ||||||||
| Grams/day | 68.6 (127.7) | 26.8 (85.8) | 65.5 (144.8) | 28.6 (85.8) | 71.6 (109.3) | 26.8 (72.4) | 0.515 | 0.457 |
| % within total UPF2 | 19.63 (19.7) | 14.4 (31.9) | 18.3 (18.8) | 13.7 (29.4) | 20.9 (20.5) | 14.7 (31.2) | 0.110 | 0.266 |
Dairy desserts include custard, pudding or similar. Processed meats include ham (cured and cooked), sausages, dried sausage, chorizo or similar. Sweets include breakfast cereals, pastries, chocolate and derivates, sugar, honey, jams, candies, and sweets. Fast-foods includes fish derivates, French fries, margarine and butter, commercial mayonnaise, fried tomato sauce, ketchup, pizza. Beverages includes soft drinks (sugar- and artificial-sweetened) and packaged juices
IQR Interquartile range, SD Standard Deviation, UPF Ultra-processed foods
1The UPF consumption was expressed as a percentage of total daily dietary intake in g (intake of UPF (g/d)/ total dietary intake (g/d)*100)
2The consumption of each food sub-group was expressed as a percentage of total UPF intake in g (intake of each food sub-group (g/d)/ total UPF intake (g/d)*100)
3P-value from ANOVA test
4P-value from Kruskal-Wallis’s test
Factors associated with UPF consumption at age 8
In Table 4 we present the factors associated with moderate and high UPF consumption at age 8, using low UPF consumption as the reference category. Factors significantly associated with moderate UPF consumption were low maternal social class, RRR = 1.87 (95%CI: 1.28–2.72) and high maternal UPF consumption during pregnancy, RRR = 1.81 (95%CI: 1.28–2.56).
Table 4.
Factors associated with UPF consumption by sex at age 8 in INMA study participantsa
| Total sample (n = 1565) | Females (n = 760) | Males (n = 805) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Moderatec | Highc | Moderatec | Highc | Moderatec | Highc | |||||||
| RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | |
| Maternal characteristics | ||||||||||||
| Age (years)b. Ref. <30 | ||||||||||||
| ≥ 30 | 0.91 | 0.69–1.20 | 0.86 | 0.64–1.14 | 0.89 | 0.61–1.29 | 0.83 | 0.55–1.27 | 0.93 | 0.62–1.40 | 0.87 | 0.58–1.30 |
| Educational levelb. Ref. Primary or less | ||||||||||||
| Secondary | 1.22 | 0.83–1.81 | 0.75 | 0.52–1.10 | 0.86 | 0.49–1.50 | 0.52 | 0.30–0.92 | 1.82 | 1.03–3.24 | 1.06 | 0.63–1.77 |
| University | 1.48 | 0.94–2.33 | 0.70 | 0.44–1.10 | 1.11 | 0.59–2.09 | 0.63 | 0.32–1.22 | 2.02 | 1.04–3.92 | 0.77 | 0.41–1.45 |
| Social classb. Ref. High | ||||||||||||
| Medium | 1.33 | 0.95–1.87 | 1.37 | 0.92–2.03 | 1.21 | 0.76–1.93 | 1.61 | 0.88–2.95 | 1.49 | 0.90–2.45 | 1.22 | 0.71–2.10 |
| Low | 1.87 | 1.28–2.72 | 2.22 | 1.46–3.37 | 1.76 | 1.03-3.00 | 2.66 | 1.38–5.11 | 1.92 | 1.12–3.31 | 1.93 | 1.11–3.37 |
| Pre-pregnancy weight status. Ref. Normal weight | ||||||||||||
| Overweight | 1.28 | 0.93–1.76 | 1.10 | 0.78–1.55 | 1.31 | 0.85–2.02 | 1.11 | 0.67–1.84 | 1.26 | 0.77–2.04 | 1.07 | 0.66–1.75 |
| Obesity | 0.98 | 0.58–1.64 | 1.26 | 0.77–2.08 | 1.54 | 0.69–3.43 | 1.95 | 0.85–4.49 | 0.74 | 0.36–1.49 | 1.01 | 0.54–1.89 |
| UPF consumption (% grams/d)b. Ref. Low (0.6–10.3) | ||||||||||||
| Moderate (10.4–17.2) | 1.32 | 0.96–1.76 | 1.87 | 1.35–2.58 | 1.44 | 0.96–2.16 | 2.13 | 1.29–3.50 | 1.19 | 0.78–1.80 | 1.68 | 1.09–2.58 |
| High (17.2–62.4) | 1.81 | 1.28–2.56 | 3.47 | 2.40–4.99 | 1.72 | 1.06–2.82 | 3.62 | 2.08–6.28 | 1.80 | 1.09–2.98 | 3.18 | 1.94–5.22 |
| Child characteristics | ||||||||||||
| Age (years) | 0.86 | 0.56–1.33 | 0.89 | 0.57–1.40 | 0.85 | 0.45–1.59 | 0.71 | 0.35–1.46 | 0.83 | 0.46–1.52 | 1.01 | 0.56–1.82 |
| Sex. Ref. Female | ||||||||||||
| Male | 1.10 | 0.86–1.41 | 1.76 | 1.34–2.30 | - | - | - | - | - | - | - | - |
| Physical activity (h/day). Ref. Tertile 1 low (0-1.2) | ||||||||||||
| Tertile 2 moderate (1.2-2.0) | 1.00 | 0.74–1.35 | 0.78 | 0.56–1.10 | 1.04 | 0.68–1.59 | 0.85 | 0.51–1.42 | 0.97 | 0.62–1.52 | 0.75 | 0.47–1.21 |
| Tertile 3 high (2.1–6.8) | 0.77 | 0.51–1.15 | 0.94 | 0.60–1.47 | 1.11 | 0.64–1.95 | 1.21 | 0.62–2.37 | 0.51 | 0.28–0.92 | 0.75 | 0.41–1.36 |
| Television viewing (h/day). Ref.Tertile 1 low (0-1.2) | ||||||||||||
| Tertile 2 moderate (1.3-2.0) | 1.24 | 0.88–1.76 | 1.39 | 0.95–2.02 | 1.63 | 1.01–2.63 | 1.34 | 0.76–2.36 | 0.88 | 0.52–1.50 | 1.37 | 0.82–2.31 |
| Tertile 3 high (2.1–5.6) | 1.35 | 0.93–1.95 | 1.76 | 1.19–2.60 | 1.19 | 0.70–2.01 | 1.58 | 0.88–2.82 | 1.55 | 0.91–2.64 | 2.00 | 1.17–3.44 |
CI Confidence Intervals, h hours, RRR Relative risk Ratios, UPF Ultra-processed foods
aUPF consumption consistent with NOVA classification, expressed as % of total daily intake in grams
bDuring pregnancy
cLow consumption of UPF as reference
Relative risk ratios from multinomial logistic regression adjusted by all variables in the table, plus cohort study
Bold style indicates statistically significant results (p < 0.05)
Factors associated with high UPF consumption, compared with low UPF consumption, were low maternal social class, compared with high maternal social class, RRR = 2.22 (95%CI: 1.46–3.37), moderate and high maternal UPF consumption during pregnancy, compared with low maternal UPF consumption, RRR = 1.87 (95%CI: 1.35–2.58) and RRR = 3.47 (95%CI: 2.40–4.99), respectively, being male, RRR = 1.76 (95%CI: 1.34–2.30), and high children’s TV viewing (2.1–5.6 h/d), compared with low children’s TV viewing (0–1.2 h/d), RRR = 1.76 (95%CI: 1.19–2.60).
Statistically significant associations for low social class, compared with high social class, and maternal UPF consumption remained consistent in sex-specific analyses (Table 4). In female children, moderate UPF consumption, compared with low UPF consumption, was associated with moderate TV viewing (1.3–2.0 h/d), RRR = 1.63 (95% CI: 1.01–2.63). High UPF consumption, compared with low UPF consumption, was associated with secondary and university mothers’ educational level, compared with primary or less, RRR secondary = 1.82 (95%CI:1.03–3.24) and RRR university = 2.02 (95%CI: 1.04–3.92), respectively, and high TV viewing (2.1–5.6 h/d), RRR = 2.00 (95%CI: 1.17–3.44).
In male children, moderate UPF consumption, compared with low UPF consumption, was associated with maternal secondary educational level, RRR = 1.82 (95% CI: 1.03–3.24), and maternal university educational level, RRR = 2.02 (95% CI: 1.04–3.92). High UPF consumption, compared with low UPF consumption, was associated with high TV viewing (2.1–5.6 h/d), compared with low TV viewing, RRR = 2.00 (95% CI: 1.17–3.44).
All statistically significant associations remained consistent when we performed sensitivity analyses by adjusting for UPF consumption at 4 years visit (Additional file 5), and UPF consumption at age 4 emerged as a strong predictor of UPF consumption at age 8.
Factors associated with UPF consumption at age 12
Table 5 shows the factors associated with moderate and high UPF consumption at age 12, using low UPF consumption as the reference category. Moderate tertile of UPF consumption was inversely associated with maternal university educational level RRR = 0.28 (95%CI: 0.09–0.92) and low maternal social class, RRR = 0.45 (95%CI: 0.45–0.99).
Table 5.
Factors associated with UPF consumption by sex at age 12 in INMA study participantsa
| Total sample (n = 314) | Females (n = 153) | Males (n = 161) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Moderatec | Highc | Moderatec | Highc | Moderatec | Highc | |||||||
| RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | RRR | 95% CI | |
| Maternal characteristics | ||||||||||||
| Age (years)b. Ref. <30 | ||||||||||||
| ≥ 30 | 0.72 | 0.35–1.47 | 0.84 | 0.41–1.75 | 0.83 | 0.28–2.43 | 0.96 | 0.32–2.82 | 0.57 | 0.20–1.65 | 0.82 | 0.28–2.41 |
| Educational level2. Ref. Primary or less | ||||||||||||
| Secondary | 0.59 | 0.20–1.78 | 0.65 | 0.22–1.89 | 0.22 | 0.05–1.03 | 0.92 | 0.18–4.72 | 2.35 | 0.32-17.00 | 0.43 | 0.09–1.99 |
| University | 0.28 | 0.09–0.92 | 0.21 | 0.06–0.69 | 0.15 | 0.03–0.80 | 0.34 | 0.06–1.99 | 0.57 | 0.07–4.93 | 0.09 | 0.02–0.54 |
| Social classb. Ref. High | ||||||||||||
| Medium | 1.03 | 0.47–2.29 | 2.05 | 0.87–4.88 | 0.96 | 0.26–3.50 | 2.06 | 0.49–8.70 | 0.80 | 0.26–2.53 | 1.89 | 0.61–5.92 |
| Low | 0.45 | 0.20–0.99 | 0.85 | 0.36–2.02 | 0.25 | 0.07–0.84 | 0.82 | 0.22–3.04 | 0.57 | 0.17–1.87 | 0.71 | 0.21–2.42 |
| Pre-pregnancy weight status. Ref. Normal weight | ||||||||||||
| Overweight | 0.59 | 0.27–1.30 | 0.57 | 0.25–1.28 | 0.43 | 0.13–1.39 | 0.26 | 0.07–0.94 | 0.39 | 0.10–1.48 | 0.85 | 0.27–2.70 |
| Obesity | 1.51 | 0.46–4.99 | 0.87 | 0.22–3.37 | 1.16 | 0.23–5.97 | 0.53 | 0.09–3.25 | 3.99 | 0.56–28.24 | 1.18 | 0.13–10.86 |
| UPF consumption (% grams/d)b. Ref. Low (1.4–10.3) | ||||||||||||
| Moderate (10.4–17.2) | 1.33 | 0.67–2.62 | 1.93 | 0.96–3.87 | 1.32 | 0.48–3.65 | 2.14 | 0.75–6.07 | 1.39 | 0.49–3.93 | 1.89 | 0.67–5.30 |
| High (17.3–45.1) | 1.82 | 0.75–4.46 | 3.09 | 1.27–7.50 | 3.28 | 0.70-15.34 | 6.39 | 1.33–30.80 | 1.13 | 0.33–3.85 | 1.98 | 0.60–6.50 |
| Children characteristics | ||||||||||||
| Age (years) | 0.51 | 0.13–1.97 | 0.60 | 0.15–2.42 | 0.11 | 0.01–1.03 | 0.23 | 0.02–2.35 | 1.61 | 0.24–10.81 | 0.74 | 0.11–5.23 |
| Sex. Ref. Female | ||||||||||||
| Male | 0.70 | 0.39–1.25 | 0.83 | 0.45–1.54 | - | - | - | - | - | - | - | - |
| Physical activity (h/day). Ref.Tertile 1 low (0.2–1.2) | ||||||||||||
| Tertile 2 moderate (1.3–1.7) | 1.11 | 0.52–2.35 | 1.16 | 0.54–2.50 | 0.59 | 0.19–1.84 | 0.78 | 0.24–2.53 | 3.24 | 0.99–10.62 | 1.72 | 0.56–5.33 |
| Tertile 3 high (1.8–6.5) | 0.52 | 0.25–1.09 | 0.43 | 0.20–0.93 | 0.20 | 0.06–0.63 | 0.21 | 0.06–0.72 | 1.50 | 0.47–4.80 | 0.77 | 0.25–2.32 |
| Television viewing (h/day). Ref. Tertile 1 low (0.1–0.8) | ||||||||||||
| Tertile 2 moderate (0.9–1.3) | 0.90 | 0.46–1.76 | 1.40 | 0.68–2.87 | 1.05 | 0.39–2.84 | 1.89 | 0.67–5.35 | 0.79 | 0.28–2.17 | 1.06 | 0.37–3.05 |
| Tertile 3 high (1.4–5.5) | 1.26 | 0.59–2.70 | 2.34 | 1.05–5.21 | 2.52 | 0.65–9.69 | 6.70 | 1.61–27.92 | 1.00 | 0.34–2.91 | 1.31 | 0.44–3.90 |
CI Confidence Intervals, RRR Relative risk Ratios, UPF Ultra-processed foods
aUltra-processed food consumption consistent with NOVA classification, expressed as % of total daily intake in grams
b During pregnancy
cLow consumption of ultra-processed food as reference
Relative risk ratios from multinomial logistic regression adjusted by all variables in the table and cohort study
Bold style indicates statistically significant results
High UPF consumption, compared with low UPF consumption, was inversely associated with maternal university educational level, compared with primary or less, RRR = 0.21 (95% CI: 0.06–0.69), and high children’s physical activity (1.8–6.5 h/d), compared with low physical activity, RRR = 0.43 (95% CI: 0.20–0.93). In contrast, high UPF consumption, compared with low UPF consumption, was positively associated with high maternal UPF consumption during pregnancy, compared with low maternal UPF consumption, RRR = 3.09 (95% CI: 1.27–7.50), and high TV viewing (1.4–5.5 h/d), compared with low TV viewing, RRR = 2.34 (95% CI: 1.05–5.21).
Part of this associations were consistent for females but not for males in sex-specific analyses (Table 5). In female adolescents, UPF moderate consumption, compared with low UPF consumption, was associated with maternal university educational level, RRR = 0.15 (95%CI: 0.03–0.80), low maternal social class, RRR = 0.25 (95% CI: 0.07–0.84) and high physical activity (1.8–6.5 h/d), RRR = 0.20 (95%CI: 0.06–0.63). Females’ high UPF consumption was associated with maternal pre-pregnancy overweight, RRR = 0.26 (95%CI: 0.07–0.94), high maternal UPF consumption during pregnancy, RRR = 6.39 (95%CI: 1.33–30.80) and high TV viewing (1.4–5.5 h/d), RRR = 6.70 (95%CI: 1.61–27.92). In male adolescents, we found no statistically significant associations for moderate UPF consumption, but we found an inverse association between high UPF consumption, compared with low UPF consumption, and maternal university educational level, compared with primary or less, RRR = 0.09 (95%CI: 0.02–0.54) (Table 5).
When we performed sensitivity analyses by adjusting UPF consumption at 8 years, the associations found with moderate UPF consumption remained similar while the associations between high UPF consumption and high maternal UPF consumption during pregnancy and children’s physical activity levels were no longer statistically significant (Additional file 5). However, UPF consumption at age 8 emerged as a strong predictor of UPF consumption at age 12.
Discussion
We showed that UPF consumption among Spanish children was consistently associated with higher maternal UPF consumption during pregnancy and greater TV viewing in children and adolescents. At age 8, high UPF consumption was also associated with lower maternal social class and being male. At age 12, maternal university education and higher levels of physical activity in children were protective factors against high UPF consumption.
Higher maternal UPF consumption during pregnancy emerged as a strong predictor of higher UPF consumption in both at 8 and 12 years of age. Our findings align with a previous cross-sectional study which showed that children whose parents reported higher UPF consumption had 54% higher UPF consumption than children whose parents reported lower UPF consumption [24]. In this sense, a recent review concluded that high UPF consumption during pregnancy had negative effects on child growth and development, adiposity gain, mental disorders and poor overall diet quality and nutrition [25]. These results partially explain how early dietary exposures during the prenatal period could influence the development of food preferences. Another possible explanation for this finding is that higher maternal UPF consumption may be acting as an indicator of social disadvantage.
In this sense, we have also found that participants whose mothers had low social class, had a higher consumption of UPF. These products are typically cheaper than minimally processed foods and require less preparation time [26], which increases their availability in all populations. Traditionally, parents with lower social class and lower educational level have been described to be less critical about UPF and to offer this type of products more often to their children [27]. This fact may explain our results which are in line with the results found in a cohort study with children aged 3–5 years [7].
Television viewing was the only child-related characteristic found to be associated with higher UPF consumption at both ages, consistent with those results reported in children [28] and adolescents [29]. In Brazil, children who watched TV during meals had 87% higher risk of having higher consumption of UPF [30]. In Spain, children who watched TV during meals ≥ 4 times per week consumed a mean of 4.67% more energy from UPF than those who watched TV < 3 times per month [28]. Spanish children received an average of 10 impacts per day from TV spots for unhealthy foods and drinks and, importantly, this exposure is almost double in low class children [31]. In this sense, TV viewing during meals has been associated with poorer diet quality, including a higher consumption of sugar-sweetened beverages and high-fat, high-sugar foods and fewer fruits and vegetables [32].
In contrast, we found that physical activity was a protective factor against high UPF consumption, particularly at age 12. In line with this, a recently published review including 19 observational studies from several countries concluded that engaging in physical activity is strongly associated with lower UPF consumption in children and adolescents [33]. However, some studies have reported a positive association between physical activity and high UPF consumption in these age groups, contrary to expectations [34, 35]. This mixed evidence suggests that the relationship between physical activity and UPF consumption in children and adolescents is not yet fully understood. In fact, in our results, the protective effect of higher physical activity levels against a high UPF consumption was only significant at age 12, but not at age 8, which may reflect this complexity.
In our study, at age 8, male children consumed more UPF than female but at age 12, UPF consumption did not differ significantly between males and females. We also found a sex-specific variation in the factors associated with high UPF consumption in children at 8 and 12 years. For instance, TV viewing was positively associated with high UPF consumption in males at age 8 and in females at age 12. We also found sex-specific variation in the associations between maternal educational levels and categories of UPF consumption. At age 8, inverse associations were observed for high UPF consumption in females, whereas in males, positive associations were observed for moderate consumption. At age 12, inverse associations were observed for moderate UPF consumption in females and for high consumption in males. These sex-specific findings can reinforce the idea that males and females may respond differently to environmental or familial influences [36, 37], which underscore the importance of considering sex-specific pathways in the development of dietary behaviors and the need for tailored public health strategies.
All the evidence collected in this discussion, alongside our findings, provides robust evidence to guide public health policies aimed at reducing UPF consumption during childhood and adolescence. Early-life interventions focused on maternal health, prenatal, and childhood nutrition should implement a two-fold approach: employing equity-based public health strategies to address socioeconomic disadvantage, such as accessible nutritional education and marketing regulations; and promoting parental strategies to reduce TV viewing, thereby limiting exposure to food marketing and reducing sedentary behaviors.
This study presents some limitations. First, the cross-sectional nature of the analysis prevents the establishment of causal relationships. Nonetheless, we repeated data collection at well-defined time points, which enhances data validity and reduces the potential for recall bias. Second, dietary intake was reported by caregivers using an FFQ, which may have introduced measurement error, particularly for foods consumed outside the home. In addition, the FFQ relies on caregivers’ ability to recall the child’s dietary intake over the previous year, which may introduce both random and systematic error. Furthermore, social desirability bias cannot be ruled out. Respondents may have overreported the consumption of health-promoting foods and underreported less favourable dietary choices [38]. However, the FFQ employed in both visits was previously validated [14] and administered by trained personnel, improving the reliability and standardization of the dietary data. Third, some degree of misclassification in the categorization of foods according to the NOVA classification system may have occurred. We used a short FFQ to assess dietary intake, which, due to its reduced length, required grouping many foods into the same item. Although these groupings were carried out by a specialized panel of dietitians and nutritional epidemiologists, based on food type, similarity, and nutrient composition, they sometimes prevented distinguishing the level of industrial processing. The most evident example is the item of whole dairy products, which grouped whole milk, whole yogurts, and milkshakes into a single item. We decided not to include this item in any NOVA group, as it mixed foods from group 1 (milk and natural yogurts) and group 4 (milkshakes and yogurts with added sugar). Fourth, although the analyses were adjusted for a wide range of covariates, the possibility of residual confounding due to unmeasured or unknown variables cannot be entirely discarded. Fifth, the smaller sample size at the 12-year follow-up and in the sex-stratified analyses may have limited statistical precision, resulting in wider confidence intervals; therefore, these findings should be interpreted with caution. Sixth, although we compared participants with and without dietary data at 12 years and observed no differences in sex, television viewing, or total energy intake at age 8, significant differences were found for physical activity and UPF consumption. Specifically, children with complete dietary data at 12 years had lower physical activity and lower UPF consumption at age 8 than those without complete data. This selective retention may have led to underestimation of the associations observed at the 12-year follow-up, and the reduced sample size, particularly in sex-stratified analyses, may have limited statistical precision. Therefore, some degree of selection bias cannot be ruled out.
This study also presents several strengths. It draws on a large, population-based sample from a well-characterized prospective cohort, with data collected across multiple geographical regions in Spain, which enhance the external validity of the findings. The inclusion of maternal variables collected prospectively during pregnancy allows for a more accurate assessment of early-life exposures and their potential influence on dietary outcomes in childhood and adolescence. In addition, stratified analyses by sex were conducted to assess the consistency of the associations and explore possible sex-specific effects. Finally, this study addresses a relevant public health concern by focusing on the consumption of UPFs during late childhood and early adolescence, which are critical developmental periods for the establishment of long-term dietary patterns.
Conclusion
This study shows that UPF consumption is high among Spanish children and adolescents, especially at age 8, and is influenced by maternal and lifestyle factors. Higher maternal UPF consumption during pregnancy and greater time of TV viewing were consistently associated with a higher UPF consumption in children and adolescents. We also observed sex-specific associations that highlight the need for targeted approaches. These findings underscore the importance of early interventions focused on family dietary habits and lifestyle to help reduce UPF consumption in childhood and adolescence. Further longitudinal studies are needed to confirm these associations and to better understand the long-term health implications of early UPF consumption in female and male children and adolescents.
Supplementary Information
Acknowledgements
We specially thank to all the families who participated in this study, the field technicians of the INMA project, the hospital technicians, midwives, and health workers for recruitment, and the whole INMA team for their commitment and their role in the success of the study. We also thank Jessica Gorlin for the English revision.
Abbreviations
- BMI
Body Mass Index
- CI
Confidence Interval
- FFQ
Food Frequency Questionnaire
- INMA
INfancia y Medio Ambiente
- IQR
Interquartile range
- RRR
Relative Risk Ratios
- SD
Standard Deviation
- TV
Television
- UPF
Ultra-processed foods
Authors’ contributions
LMCG and SGP designed the research study and developed the methodology. JV contributed to the conceptualization, methodology, validation, and supervision of the study. GFT contributed to the conceptualization of the research. GFT, AT, LSMR, ZB, MV, SW, SL, RSB, MGH and JV conducted the investigation. AOC, LTC and COB curated the data. LMCG and SGP performed the formal analysis. LMCG and SGP wrote the original draft of the manuscript. All authors contributed to the review and editing of the manuscript. All authors read and approved the final manuscript.
Funding
This work was supported by ESP24PI02/2024 CIBER -Consorcio Centro de Investigación Biomédica en Red (Convocatoria Intramural Proyectos CIBERESP 2024), Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación and Unión Europea – European Regional Development Fun”. Project PI23/01568, funded by Instituto de Salud Carlos III (ISCIII) and co-funded by the European Union. ISGlobal acknowledges support from the grant CEX2023-0001290-S funded by MCIN/AEI/ https://doi.org/10.13039/501100011033, and support from the Generalitat de Catalunya through the CERCA Program. Sarah Warkentin acknowledges receiving the funding by the Agency for Management of University and Research Grants with a Beatriu de Pinós post-doctoral fellowship (Ref: 2021 BP 00058). Raquel Soler-Blasco acknowledges receiving the funding by the Generalitat Valenciana (CIGE2023/142) and General Council of Official Nursing Boards of Spain and the Spanish Institute of Nursing Research (grant code PN22_CGE45). Carolina Ojeda is recipient of a fellowship “Personal Investigador en Formación” from the “Universidad Miguel Hernández de Elche” (04-541-6-2024-0153-N). This study was funded by Grants from UE (FP7-ENV-2011 cod 282957, HEALTH.2010.2.4.5-1, cod 874583, and cod 101136566), Spain: ISCIII (G03/176; FIS-FEDER: PI03/1615, PI04/1509, PI06/1213, PI11/01007, PI11/02591, PI11/02038, PI12/00610, PI13/1944, PI13/2032, PI14/00891, PI14/01687, PI16/1288, PI17/00663, PI19/1338; PI23/1578), Generalitat Valenciana (CIAICO/2021/132, BEST/2020/059, AICO 2020/285, AICO/2021/182 and CIDEGENT/2019/064, PROMETEO-CIPROM/2024/66), Consejo General de Enfermería (PNI22_CGE45).This study was also funded by grants from Instituto de Salud Carlos III (FIS-PI06/0867, FIS-PI09/00090, FIS-PI13/02187, FIS-PI18/01142 include FEDER funds, FIS-PI18/01237 incl. FEDER funds CIBERESP, Department of Health of the Basque Government (2005111093, 2009111069, 2013111089, 2015111065 and 2018111086), and the Provincial Government of Gipuzkoa (DFG06/002, DFG08/001 and DFG15/221, DFG 89/17 and DFG22-cien-000006-01 and annual agreements with the municipalities of the study area (Zumarraga, Urretxu, Legazpi, Azkoitia y Azpeitia y Beasain).
Data availability
The datasets generated and/or analysed during the current study are not publicly available due participant consent and ethics approvals did not permit open sharing but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
All pregnant women provided written informed consent prior to their inclusion in the study. In the case of their offspring, written informed consent was obtained from their parents or legal guardians. The study was conducted in accordance with the principles of the Declaration of Helsinki and its later amendments. The study protocol was approved by the following Ethics Committees: the Clinical Research Ethics Committee of Hospital Universitario La Fe (Valencia, Spain); the Clinical Research Ethics Committee of Área Sanitaria de Gipuzkoa (Gipuzkoa, Spain); the Clinical Research Ethics Committee of Parc de Salut del Mar (Barcelona, Spain); and the Clinical Research Ethics Committee of the Principado de Asturias (Spain). The present work was additionally approved by the Clinical Research Ethics Committee of the Alicante Health Department – Hospital General (Ref.: PI2024-087, June 19, 2024).
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Monteiro CA, Louzada ML, Steele-Martinez E, Cannon G, Andrade GC, Baker P et al. Ultra-processed foods and human health: the main thesis and the evidence. The Lancet. 2025;:S014067362501565X. 10.1016/S0140-6736(25)01565-X [DOI] [PubMed]
- 2.Dai S, Wellens J, Yang N, Li D, Wang J, Wang L, et al. Ultra-processed foods and human health: An umbrella review and updated meta-analyses of observational evidence. Clin Nutr. 2024;43:1386–94. 10.1016/j.clnu.2024.04.016. [DOI] [PubMed] [Google Scholar]
- 3.Khoury N, Martínez MÁ, Garcidueñas-Fimbres TE, Pastor-Villaescusa B, Leis R, de Las Heras-Delgado S, et al. Ultraprocessed Food Consumption and Cardiometabolic Risk Factors in Children. JAMA Netw Open. 2024;7:e2411852. 10.1001/jamanetworkopen.2024.11852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen ZH, Mousavi S, Mandhane PJ, Simons E, Turvey SE, Moraes TJ, et al. Ultraprocessed Food Consumption and Obesity Development in Canadian Children. JAMA Netw Open. 2025;8:e2457341. 10.1001/jamanetworkopen.2024.57341. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Baker P, Machado P, Santos T, Sievert K, Backholer K, Hadjikakou M, et al. Ultra-processed foods and the nutrition transition: Global, regional and national trends, food systems transformations and political economy drivers. Obes Rev. 2020;21:e13126. 10.1111/obr.13126. [DOI] [PubMed] [Google Scholar]
- 6.García-Blanco L, de la Pascual O, Berasaluce V, Moreno-Galarraga A, Martínez-González L, Martín-Calvo MÁ. Individual and family predictors of ultra-processed food consumption in Spanish children: The SENDO project. Public Health Nutr. 2023;26:437–45. 10.1017/S136898002200132X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Carroll JE, Sturgeon SR, Bertone-Johnson E, VanKim N, Longacre MR, Dalton MA, et al. Factors Correlated With Ultra-Processed Food (UPF) Intake in Preschool-Aged Children and Association With Weight. J Nutr Educ Behav. 2024;56:196–208. 10.1016/j.jneb.2023.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Al Hourani H, Shhadeh HA, Al-Jawaldeh A. Association between consumption of ultra processed foods and obesity among Jordanian children and adolescents. Sci Rep. 2025;15:9326. 10.1038/s41598-025-93506-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vandevijvere S, Pedroni C, De Ridder K, Castetbon K. The Cost of Diets According to Their Caloric Share of Ultraprocessed and Minimally Processed Foods in Belgium. Nutrients. 2020;12:2787. 10.3390/nu12092787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Poti JM, Mendez MA, Ng SW, Popkin BM. Highly Processed and Ready-to-Eat Packaged Food and Beverage Purchases Differ by Race/Ethnicity among US Households. J Nutr. 2016;146:1722–30. 10.3945/jn.116.230441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Machado-Rodrigues AM, Padez C, Rodrigues D, Dos Santos EA, Baptista LC, Liz Martins M, et al. Ultra-Processed Food Consumption and Its Association with Risk of Obesity, Sedentary Behaviors, and Well-Being in Adolescents. Nutrients. 2024;16:3827. 10.3390/nu16223827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Compañ-Gabucio L-M, Torres-Collado L, Oncina-Cánovas A, Fernández-Tardón G, Tardón A, Santa-Marina Rodríguez L, et al. Predictors of ultra-processed foods consumption in 4-year-old Spanish children. Appetite. 2025;216:108268. 10.1016/j.appet.2025.108268. [DOI] [PubMed] [Google Scholar]
- 13.Guxens M, Ballester F, Espada M, Fernández MF, Grimalt JO, Ibarluzea J, et al. Cohort Profile: the INMA–INfancia y Medio Ambiente–(Environment and Childhood) Project. Int J Epidemiol. 2012;41:930–40. 10.1093/ije/dyr054. [DOI] [PubMed] [Google Scholar]
- 14.Vioque J, Garcia-de-la-Hera M, Gonzalez-Palacios S, Torres-Collado L, Notario-Barandiaran L, Oncina-Canovas A, et al. Reproducibility and Validity of a Short Food Frequency Questionnaire for Dietary Assessment in Children Aged 7–9 Years in Spain. Nutrients. 2019;11:933. 10.3390/nu11040933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.U.S. Department of Agriculture (USDA). Agriculture Research Service 2010 USDA National Nutrient Database for Standard Reference; Release 23. 2010.
- 16.Palma Linares I, Farran A, Cantós D. Tablas de composición de alimentos por medidas caseras de consumo habitual en España = Taules de composició d’aliments per mesures casolanes de consum habitual a Espanya. Barcelona, Madrid: Universitat de Barcelona; McGraw-Hill-Interamericana; 2008. [Google Scholar]
- 17.Monteiro CA, Cannon G, Moubarac J-C, Levy RB, Louzada MLC, Jaime PC. The UN Decade of Nutrition, the NOVA food classification and the trouble with ultra-processing. Public Health Nutr. 2018;21:5–17. 10.1017/S1368980017000234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Domingo-Salvany A, Regidor E, Alonso J, Alvarez-Dardet C. [Proposal for a social class measure. Working Group of the Spanish Society of Epidemiology and the Spanish Society of Family and Community Medicine]. Aten Primaria. 2000;25:350–63. [PubMed] [Google Scholar]
- 19.Vioque J, Navarrete-Muñoz E-M, Gimenez-Monzó D, García-de-la-Hera M, Granado F, Young IS, et al. Reproducibility and validity of a food frequency questionnaire among pregnant women in a Mediterranean area. Nutr J. 2013;12:26. 10.1186/1475-2891-12-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Vioque J, Gimenez-Monzo D, Navarrete-Muñoz EM, Garcia-de-la-Hera M, Gonzalez-Palacios S, Rebagliato M, et al. Reproducibility and Validity of a Food Frequency Questionnaire Designed to Assess Diet in Children Aged 4–5 Years. PLoS ONE. 2016;11:e0167338. 10.1371/journal.pone.0167338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Agresti A. Analysis of Ordinal Categorical Data. 1st edition. Wiley; 2010. 10.1002/9780470594001. [DOI]
- 22.Hilsen M, Te Velde SJ, Bere E, Brug J. Predictors and mediators of differences in soft drinks consumption according to gender and plans of further education among Norwegian secondary-school children. Public Health Nutr. 2013;16:1250–6. 10.1017/S1368980011002953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Magalhães V, Severo M, Correia D, Torres D, Costa De Miranda R, Rauber F, et al. Associated factors to the consumption of ultra-processed foods and its relation with dietary sources in Portugal. J Nutr Sci. 2021;10:e89. 10.1017/jns.2021.61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Oliveira GAL, Gonçalves VSS, Nakano EY, Toral N. The effect of parents’ lifestyle on Schoolchildren’s consumption of ultra-processed food. Public Health. 2024;237:443–8. 10.1016/j.puhe.2024.11.005. [DOI] [PubMed] [Google Scholar]
- 25.Morales-Suarez-Varela M, Rocha-Velasco OA. Impact of ultra-processed food consumption during pregnancy on maternal and child health outcomes: A comprehensive narrative review of the past five years. Clin Nutr ESPEN. 2025;65:288–304. 10.1016/j.clnesp.2024.12.006. [DOI] [PubMed] [Google Scholar]
- 26.Valicente VM, Peng C-H, Pacheco KN, Lin L, Kielb EI, Dawoodani E, et al. Ultraprocessed Foods and Obesity Risk: A Critical Review of Reported Mechanisms. Adv Nutr. 2023;14:718–38. 10.1016/j.advnut.2023.04.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Vidal L, Antúnez L, Ares G, Gugliucci V, Vitola A, Girona A. Factors underlying mothers’ decision to offer ultra-processed meals during early childhood. Food Qual Prefer. 2024;121:105279. 10.1016/j.foodqual.2024.105279. [DOI] [Google Scholar]
- 28.Martín-Calvo N, Usechi A, Fabios E, Gómez SF, López-Gil JF. Television watching during meals is associated with higher ultra-processed food consumption and higher free sugar intake in childhood. Pediatr Obes. 2024;19:e13130. 10.1111/ijpo.13130. [DOI] [PubMed] [Google Scholar]
- 29.López-Gil JF, Quiroz-Cárdenas F, Montenegro-Espinosa JA, Cisneros-Vásquez E, Miño C, Castillo-Miñaca ME, et al. Eating distractions as predictors of ultra-processed food consumption and Mediterranean diet adherence in adolescents. Sci Rep. 2025;15:7579. 10.1038/s41598-025-91754-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.de Lacerda AT, Carmo AS do, de Sousa TM, Santos LCD, Participation of ultra-processed, foods in brazilian school children’s diet and associated factors. Rev Paul Pediatr. 2020;38:e2019034. 10.1590/1984-0462/2020/38/2019034 [DOI] [PMC free article] [PubMed]
- 31.Royo-Bordonada MÁ, Cavero-Esponera C, Romero-Fernández MM, González-Díaz C, Ordaz Castillo E. Differences in children’s exposure to television advertising of unhealthy foods and beverages in Spain by socio-economic level. BMC Public Health. 2024;24:739. 10.1186/s12889-023-17410-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Avery A, Anderson C, McCullough F. Associations between children’s diet quality and watching television during meal or snack consumption: A systematic review. Matern Child Nutr. 2017;13:e12428. 10.1111/mcn.12428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Mescoloto SB, Pongiluppi G, Domene SMÁ. Ultra-processed food consumption and children and adolescents’ health. J Pediatr (Rio J). 2024;100(1 Suppl 1):S18–30. 10.1016/j.jped.2023.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Gonçalves Silva C, Saunders C, Da Cruz Rangel LF, Alegre De Matos A, Martins Teixeira F, Gonçalves Ribeiro B. Impact of ultra-processed foods on excess weight and dyslipidemia in schoolchildren. Nutrition. 2025;131:112638. 10.1016/j.nut.2024.112638. [DOI] [PubMed] [Google Scholar]
- 35.Li M, Shi Z. Ultra-processed food consumption and obesity among children and adolescents in China-Findings from China Health and Nutrition Survey. Pediatr Obes. 2025;20:e70012. 10.1111/ijpo.70012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Monge-Rojas R, Colón-Ramos U, Chinnock A, Smith-Castro V, Reyes-Fernández B. Gender-based eating norms, the family environment and food intake among Costa Rican adolescents. Public Health Nutr. 2021;24:4840–50. 10.1017/S1368980021000835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Faith MS, Rhea SA, Corley RP, Hewitt JK. Genetic and shared environmental influences on children’s 24-h food and beverage intake: sex differences at age 7 y. Am J Clin Nutr. 2008;87:903–11. 10.1093/ajcn/87.4.903. [DOI] [PubMed] [Google Scholar]
- 38.Hebert JR, Clemow L, Pbert L, Ockene IS, Ockene JK. Social Desirability Bias in Dietary Self-Report May Compromise the Validity of Dietary Intake Measures. Int J Epidemiol. 1995;24:389–98. 10.1093/ije/24.2.389. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analysed during the current study are not publicly available due participant consent and ethics approvals did not permit open sharing but are available from the corresponding author on reasonable request.