Probability of predicting dietary pattern adherence among primary schoolchildren based on socio-demographic data, anthropometric parameters, physical activity levels, and nutritional knowledge Authors: Małgorzata Ewa Drywień, Ewa Czarniecka-Skubina, Jadwiga Hamułka, Jerzy Gębski, Olga Januszko, Joanna Sadowska, Krystyna Gutkowska Published: 2026 ABSTRACT AbstractAdolescence is a critical period of rapid physical and psychosocial development during which eating behaviours are influenced by a variety of internal and external factors. Understanding these relationships is crucial for preventing unhealthy dietary habits and for effective nutrition education. This cross-sectional study included 7,763 schoolchildren aged 10–12 (49.2% boys, 50.8% girls) from Polish schools. The validated questionnaire was used to assess children’s nutritional knowledge, dietary habits, and physical activity. Standard anthropometric measurements (body weight, height, waist circumference) were taken, and BMI, WHtR, RFMp, and TMI were calculated. Descriptive statistics were used for sample characteristics. Principal component analysis (PCA) with varimax rotation (MSA = 0.796) was used to identify dietary patterns. Logistic regression modeled predictors of belonging to the highest tertile. Three dietary patterns were identified: Salty-Sweet, Plant-Dairy, and Protein. Children classified as overweight according to the RFMp index and children who were very physically active showed less adherence to the Salty-Sweet Pattern. Schoolchildren who were underweight according to the BMI index and those with low nutritional knowledge were more likely to follow the Salty-Sweet Pattern. Girls, children with an elevated waist-to-hip ratio (WHtR), and very physically active children were more likely to be in the third tertile of the Plant-Dairy Pattern. Schoolchildren with low nutritional knowledge showed less adherence to this pattern. Girls showed less adherence to the protein pattern, whereas underweight children and those who were very physically active showed greater adherence. The dietary patterns of schoolchildren are interrelated, and knowledge of nutrition and physical activity is the main modifiable determinant. These factors can be utilised when developing strategies for regularly assessing nutritional status and providing specialist advice to schoolchildrens, their families, and teachers.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-70753-6.Keywords: Schoolchildren, Dietary patterns, Socio-demographic factors, Anthropometric data, Physical activity, Nutritional knowledgeSubject terms: Diseases, Health care, Medical research, Risk factors Introduction Adolescence is a crucial stage of physical, psychosocial, and sexual development, during which individuals gain greater autonomy. During this period, energy and nutrient requirements increase, and growing independence in dietary decision-making can expose young people to the risk of over- or undernutrition1. Schoolchildrens eating habits are influenced by their peers, the media, and their perception of their own bodies2. Urbanisation and globalisation have led to a shift away from traditional eating habits towards a diet rich in processed foods (skipping meals, snacking, eating out, choosing fast food), which contributes to a decline in dietary quality and an increase in obesity3. The main risk factors are the excessive consumption of sweets and fast food, skipping breakfast4, and low levels of physical activity5. Knowledge of nutrition plays a key role in shaping healthy habits and weight control, and diet is a key modifiable risk factor in the development of chronic diseases6. Dietary patterns provide a more accurate reflection of actual behaviour than the analysis of individual products7. At the same time, links can be identified between dietary patterns and other characteristics, which in turn may be associated with different diseases7,8. Dietary behaviour, like other forms of consumer behavior, is shaped by the cultural environment in which an individual develops and grows up. Analyzing dietary habits allows us to understand deeper, automatic mechanisms and repetitive patterns of choices, because in addition to knowledge about the foods we consume, we gain a comprehensive picture of the causes behind specific behaviors. Neglecting dietary needs and habits during adolescence can have serious lifelong consequences for the health, well-being, and immunity of future generations9. This study aimed to determine whether gender, place of residence, anthropometric parameters, physical activity, and nutritional knowledge could predict adherence to healthy or unhealthy eating patterns among schoolchildren. This could form the basis of educational and preventive public health activities aimed at preventing non-communicable chronic diseases. We hope that the experiences and conclusions from this study can provide a basis for educational and preventive measures in the field of public health for young adolescents. We trust that the findings of this study will fill gaps in both research and practice regarding adolescent nutrition, thereby enabling the development of more effective programmes that can contribute to improving schoolchildrens overall health and well-being and help them reach their full potential. Methods Study design and participants This cross-sectional study, part of the Junior-Edu-Nutrition (JEŻ) project, included 7,763 schoolchildrens aged 10–12 (49.2% boys, 50.8% girls) from Polish schools, public and non-public, located in all 16 voivodeships across Poland, taking into consideration three sizes of the locality (i.e., villages, towns with up to 100,000 inhabitants, and large urban agglomerations). Details regarding the selection of the research sample were presented in our previous article10. The study covered all schoolchildrens in the school. All of the participants, as well as their parents or legal guardians, had to provide informed consent to participate in the study. Excluded were schoolchildren who refused the procedure (they did not obtain informed consent from their parents or caregivers) and withdrew immediately before the measurements (e.g., they did not want to participate in the study despite their parents’ consent; we did not conduct the study in that case) or who were absent from school on the day of measurement (e.g., due to illness). Data concern only schoolchildrens for whom full data were obtained. It involved a questionnaire and anthropometric measurements. Data collection Assessment of food consumption, physical activity, and nutrition knowledge The validated SF-FFQ4PolishChildren® questionnaire was used to assess children’s nutritional knowledge, dietary habits, and physical activity11. Schoolchildren completed it independently during class after receiving clear instructions. Each anonymous questionnaire was checked for completeness and took about 30–45 min to complete. The tool included 20 closed-ended questions on dietary guidelines and physical activity. Nutritional knowledge was evaluated with five validated and 15 recommendation-based questions. Each correct answer was assigned one point. (maximum 20). The posteriori approach was based on tertile distributions of all three indexes, with the bottom, middle, and upper tertiles. Based on tertiles, nutrition knowledge was classified as low (0–7), moderate (8–14), or high (15–20). Physical activity was categorized as low, moderate, or high, depending on weekly duration and intensity. Low physical activity was defined as mostly sitting, watching TV, using a computer, reading, or a short walk, activities up to 2 h per week. Moderate physical activity was considered to be walking, cycling, gymnastics, or other light physical activity performed 2–3 h per week. Vigorous physical activity was defined as cycling, running, or other sporting activities that require physical effort and are performed for more than 3 h per week. Dietary habits were assessed by the frequency of consumption of 12 food groups over the previous year. Among the 12 food groups were: dairy products (e.g., milk, yogurt, cottage cheese, cheese), fish (e.g., baked, smoked, fried, canned), fast foods (e.g., chips, pizza, hamburgers), sugar-sweetened beverages (e.g., cola-type, tea-type, water with syrup), fruit or mixed fruit-veggie juices, energy drinks, vegetables (e.g., fresh, boiled, baked, stewed), fruit (fresh or frozen), sweets or confectionery (e.g., cookies, sweets, cake, chocolate bars, chocolate), whole grain products (e.g., bread, oats, brown rice, buckwheat groats), pulses (beans, lentils, chickpeas), and salty snacks (e.g., chips, coated peanuts). The frequency of consumption was defined as: never or almost never, less than once a week, once a week, 2–4 times/week, 5–6 times/week, every day, and several times a day. Detailed methods are described in Hamułka et al.10. Anthropometrics All measurements were in accordance with the International Standards for Anthropometric Assessment [ISAK] recommendations. Anthropometric measurements were performed between 8:00 a.m. and 12:00 p.m. using standardized equipment: a portable stadiometer and electronic scale (TANITA MC-780- S MA, Tokyo, Japan, with European Certificate of approval No EU CE 0122) and a stretch-resistant tape (SECA 201, Hamburg, Germany). Height was measured to the nearest 0.5 cm without shoes, in an upright position with the head in the Frankfurt plane. Body weight was recorded to the nearest 0.1 kg in light indoor clothing. Waist circumference was measured to the nearest 0.1 cm at the midpoint between the iliac crest and lower rib during expiration. Measurements were compared with Polish reference values (< 90th percentile)12,13. Before measurements began at school, the device was calibrated. Before each measurement, the electrodes were thoroughly wiped with an appropriate disinfectant. The measurements were performed by the project’s research team, which was trained to conduct this type of measurement. Body composition was analysed in schoolchildrens, except those with arm or leg deformations that prevented proper contact with the electrodes. Measurements were taken twice. If the differences between two values were greater than 0.5 kg for body weight, 1 cm for height, 1.5 cm for waist or hip circumferences, and 2 kg for hand grip strength, a third measurement was taken, and the two closest values were selected. Detailed information on the anthropometric measurement procedure can be found in the previous article10. From these data, Body Mass Index (BMI), Waist-to-Height Ratio (WHtR), Pediatric Real Fat Mass Index (RFMp), and Tri-ponderal Mass Index (TMI) were calculated. BMI percentiles were used to classify nutritional status as underweight, normal weight, overweight, or obese, based on references14. WHtR was obtained as waist circumference divided by height (cm) and compared with national percentile charts15. RFMp, calculated using the pediatric equation for children aged 8–14 years16, estimated total body fat percentage: RFMp = 74 - (22 x height/waist) + (5 x gender); (where: gender = 0 for boys, = 1 for girls) RFMp values were classified according to Moya et al.17(Table 1). Table 1Relative Fat Mass pediatric references30.RFMp [%]underweightnormal weightoverweightobesityBoys≤20.720.8–25.725.8–31.231.3 and aboveGirls≤ 25.025.1–30.530.6–37.938.0 and aboveRFMp - Pediatric Real Fat Mass Index. Relative Fat Mass pediatric references30. RFMp - Pediatric Real Fat Mass Index. The Tri-ponderal Mass Index (TMI), calculated as weight (kg) divided by height (m³), was used to estimate body fat, as it better reflects body composition in children aged 8–17 than BMI18. A cut-off point of 14 kg/m³ (highest sensitivity, specificity, and 85.5% accuracy) was applied: ≤14 kg/m³ – normal; >14 kg/m³ – overweight/obesity19. Data analysis Descriptive statistics were used for sample characteristics. Principal component analysis (PCA) with varimax rotation (MSA = 0.796) was used to identify dietary patterns. The number of factors (patterns) was determined based on the scree plot and the eigenvalue criterion of ≥ 1. Variables with factor loadings of ≥ 0.45 were included in the respective factors. Logistic regression modeled predictors of belonging to the highest tertile. The model coefficients obtained are predictors of the dependent variable, and the significance of the variable in the model was assessed on the basis of Wald’s chi-square statistic. Three models were developed, in which the dependent variables were patterns. The patterns were divided into tertiles, and the predicted value of the dependent variable was highest in the third (highest) tertile. To obtain a binary dependent variable, tertiles 1 and 2 were combined. In addition to the patterns, the following were included as independent variables in the models: sex, place of residence, RFMp, WC, WHtR, TMI, physical activity, and nutrition knowledge. Translated with DeepL.com (free version). All questionnaire and anthropometric data obtained in the study were used for statistical analysis. Data were collected in direct contact with participants from the survey, and the researchers performed anthropometric measurements themselves, so missing data were updated immediately upon detection. Analyses were conducted in SAS 9.4 (α = 0.05). Ethics The study was approved by the Ethics Committee of the Faculty of Human Nutrition and Consumer Sciences, Warsaw University of Life Sciences (Resolution No. 18/2022) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from the parents and/or legal guardians of all participating schoolchildren. Results Principal component analysis (PCA) identified three dietary patterns among the adolescents studied: Salty-Sweet, Plant-Dairy, and Protein, characterized by increased consumption of certain products (Table 2). Table 2Factor-loading matrix for the dietary patterns DPs identified by principal component analysis.Food GroupsSalty-Sweet PatternPlant-Dairy PatternProtein PatternSalty snacks(chips, pretzels, breaded nuts)0.79674--Sweetened drinks(cola, sprite, water with fruit syrup, Nestea-type drinks)0.75007--Fast food(French fries, pizza, hamburgers, toasted sandwiches)0.72496--Sweets(cakes, biscuits, candies, bars, chocolates)0.64706--Fruit(raw, frozen, added to dishes)-0.79035-Vegetables(raw, cooked, roasted, stewed)-0.78164-Dairy products(milk, yoghurts, cottage cheese, cheese)-0.45326-Whole grain products(brown bread, oatmeal, brown rice, groats, e.g. buckwheat, pearl barley)-0.45174-Fish--0.79838Legumes(lentils, soybeans, chickpeas, peas, beans)--0.64296 Factor-loading matrix for the dietary patterns DPs identified by principal component analysis. Salty snacks (chips, pretzels, breaded nuts) Sweetened drinks (cola, sprite, water with fruit syrup, Nestea-type drinks) Fast food (French fries, pizza, hamburgers, toasted sandwiches) Sweets (cakes, biscuits, candies, bars, chocolates) Fruit (raw, frozen, added to dishes) Vegetables (raw, cooked, roasted, stewed) Dairy products (milk, yoghurts, cottage cheese, cheese) Whole grain products (brown bread, oatmeal, brown rice, groats, e.g. buckwheat, pearl barley) Legumes (lentils, soybeans, chickpeas, peas, beans) Sample characteristic In towns with up to 100,000 residents, girls outnumbered boys in groups of children living there, especially in towns with up to 20,000 residents. In cities with over 100,000 inhabitants, however, boys constituted the larger group (Table S1). According to the BMI categories, the majority of children were of normal weight. WHtR ≥ 90prc was significantly more common in boys than in girls. Based on the RFMp index, the majority of respondents (N = 2,812) were underweight, with 60.53% of girls and 39.47% of boys falling into this category. However, more boys than girls were obese. In the case of the TMI index, girls dominated at values up to 14, while boys dominated at values above 14. More boys were characterised by a higher level of physical activity than girls. In the case of nutrition knowledge assessment, the opposite relationship was observed, i.e., more girls than boys demonstrated high nutrition knowledge. The three identified dietary patterns were divided into three tertiles. In the third tertile of both the Salty-Sweet and Protein Patterns, boys outnumbered girls. In the third tertile of the Plant-Dairy Pattern, however, girls predominated (Table S2). Girls were 19.2% less likely than boys to belong to the third tertile of the Salty-Sweet Pattern, and children from large cities were less likely than those from rural areas (Table 3). Belonging to the second or third tertile of the Plant-Dairy Pattern increased the likelihood of adherence to the Salty-Sweet Pattern by 36.2% and 34.4%, respectively. In contrast, being in the second tertile of the Protein Pattern reduced this likelihood by 24.4%. Underweight schoolchildren (BMI) were 19.6% more likely to adhere to the Salty-Sweet Pattern, while overweight and obese schoolchildren (RFMp) were 16.7% and 31.8% less likely, respectively. Moderate and high physical activity reduced adherence by 42.1% and 44.1%. Conversely, moderate and low nutritional knowledge significantly increased the likelihood of following the Salty-Sweet Pattern by 89.1% and 204.9%, respectively. Table 3Prediction of adherence to the 3rd tercile of the Salty-Sweet Pattern.ParameterLevelEstimatePoint estimate95% Wald Confidence LimitsWald Chi-SquarePr > ChiSqIntercept−0.69312.670.0004SexBoys (ref.)1Girls−0.2130.8080.730.9016.31< 0.0001Place of residenceVillage (ref.)1City < 20 k0.0371.0370.891.210.220.6384City 20–100 k0.1271.1350.981.322.820.0933City 101–500 k−0.1620.8510.740.984.800.0285City > 500 k−0.4110.6630.560.7824.73< 0.0001Plant-Dairy PatternT1 (ref.)1T20.3091.3621.211.5425.40< 0.0001T30.2961.3441.191.5221.67< 0.0001Protein PatternT1(ref.)1T2−0.2790.7560.670.8521.07< 0.0001T3−0.0990.9060.811.022.690.1009BMInormal weight (ref.)1underweight0.1791.1961.021.404.900.0269overweight−0.0100.9900.791.240.010.9294RFMpnormal weight (ref.)1underweight0.0251.0260.911.160.160.6861overweight−0.1820.8330.700.993.910.0479obesity−0.3820.6820.490.964.880.0272WCnormal weight (ref.)1overweight−0.1700.8440.671.012.110.1466WHtR< 90prc0≥ 90prc−0.0540.9480.781.250.140.7062TMI≤ 14 (ref.)1> 140.1291.1380.931.391.630.2023Physical activitylow (ref.)1moderate−0.5470.5790.500.6847.57< 0.0001high−0.5820.5590.480.6552.36< 0.0001Nutrition knowledgehigh (ref.)1moderate0.6371.8911.602.2454.85< 0.0001low1.1153.0492.543.66142.03< 0.0001BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Prediction of adherence to the 3rd tercile of the Salty-Sweet Pattern. BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Girls were 44% more likely than boys to be in the third Plant-Dairy Pattern tertile, and children living in cities with over 500,000 inhabitants were 53% more likely to be in the third Plant-Dairy Pattern tertile (Table 4). Belonging to the third tertile of the Salty-Sweet Pattern increased the odds of belonging to the third Plant-Dairy Pattern tertile by 59%. Having a WHtR above 0.5 increased the likelihood by 41%. Moderate and high levels of physical activity increased the odds by 40% and 105%, respectively. Conversely, having low or moderate nutritional knowledge reduced the odds of belonging to the third Plant-Dairy Pattern tertile by 71% and 50%, respectively. Table 4Prediction of adherence to the 3rd tercile of the Plant-Dairy Pattern.ParameterLevelEstimatePoint estimate95% Wald Confidence LimitsWald Chi-SquarePr > ChiSqIntercept−1.05628.30< 0.0001SexBoys (ref.)1Girls0.3671.4441.31.6047.39< 0.0001Place of residenceVillage (ref.)1City < 20 k−0.0480.9540.811.120.340.5593City 20–100 k−0.0480.9530.821.110.360.5467City 101–500 k0.1261.1340.981.312.800.0944City > 500 k0.4261.5311.311.7928.64< 0.0001Salty-Sweet PatternT1 (ref.)1T20.0691.0720.951.211.280.2581T30.1721.1881.051.347.530.0061Protein PatternT1(ref.)1T2−0.0070.9930.881.120.010.9149T3−0.0830.920.821.041.810.1784BMInormal weight (ref.)1underweight−0.0250.9750.831.150.090.7609overweight0.1831.2010.951.512.420.1194RFMpnormal weight (ref.)1underweight−0.0940.9110.811.032.220.1365overweight−0.0400.9610.801.150.180.6693obesity0.0971.1010.781.560.290.5872WCnormal weight (ref.)1overweight, obesity0.0601.0620.851.330.270.6039WHtR< 90prc (ref.)1≥ 90prc0.3461.4131.061.885.630.0176TMI≤ 14 (ref.)1> 14−0.1160.8910.731.091.220.2697Physical activitylow (ref.)1moderate0.3381.4021.171.6813.520.0002high0.7202.0531.722.4661.56< 0.0001Nutrition knowledgehigh (ref.)1moderate−0.7000.4960.430.5795.63< 0.0001low−1.2400.2890.250.34212.79< 0.0001BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Prediction of adherence to the 3rd tercile of the Plant-Dairy Pattern. BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Compared to boys, girls were 12% less likely to belong to the third tertile of the Protein Pattern (Table 5). Belonging to the second tertile of the Salty-Sweet Pattern reduced the likelihood of reaching the third Protein Pattern tertile by 14%. Underweight schoolchildren were 26% more likely than children of normal weight to belong to this tertile. Compared to low physical activity levels, moderate and high activity levels increased the likelihood of belonging to the third Protein Pattern tertile by 36% and 102%, respectively. Table 5Prediction of adherence to the 3rd tercile of the Protein pattern.ParameterLevelEstimatePoint estimate95% Wald Confidence LimitsWald Chi-SquarePr > ChiSqIntercept−0.90322.35< 0.0001SexBoys (ref.)1Girls−0.1240.8830.800.985.660.0173Place of residenceVillage (ref.)1City < 20 k−0.0230.9780.841.140.080.7719City 20–100 k−0.0910.9130.791.061.390.2378City 101–500 k0.0121.0120.881.170.030.8672City > 500 k0.1191.1260.971.312.270.1320Salty-Sweet PatternT1 (ref.)1T2−0.1490.8620.770.976.160.0131T3−0.0370.9640.861.090.370.5452Plant-Dairy PatternT1(ref.)1T20.00010.891.130.000.9989T3−0.0820.9210.821.041.720.1895BMInormal weight (ref.)1underweight0.2291.2571.081.478.200.0042overweight0.0381.0390.841.290.120.7301RFMpnormal weight (ref.)1underweight−0.0900.9140.811.032.080.1488overweight−0.0460.9560.801.140.250.6151obesity−0.1120.8940.641.240.440.5053WCnormal weight (ref.)1overweight, obesity−0.0930.9110.731.140.670.4121WHtR< 90prc (ref.)1 ≥ 90prc−0.0950.9100.701.190.470.4923TMI≤ 14 (ref.)1> 140.1451.1560.951.412.150.1429Physical activitylow (ref.)1moderate0.3071.3601.141.6212.210.0005high0.7072.0271.712.4164.82< 0.0001Nutrition knowledgehigh (ref.)1moderate−0.0030.9970.861.150.000.9707low−0.0790.9240.781.090.860.3541BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Prediction of adherence to the 3rd tercile of the Protein pattern. BMI - Body Mass Index; WC - Waist Circumference; WHtR - Waist to Height Ratio, RFMp - Pediatric Real Fat Mass Index, TMI - Tri-pounderal Mass Index. Wald’s chi-square test showed that nutritional knowledge had the greatest influence on the characteristics of the Salty-Sweet and Plant-Dairy Patterns, followed by physical activity (Fig. 1). Place of residence ranked third. In the Protein Pattern, physical activity had the greatest influence, whilst the other variables accounted for only a few per cent of the variation. The proportion of variables in Dietary Patterns. DP – Dietary Patterns; NK – Nutrition Knowledge; PhA – Physical Activity; PR – Place of Residence. Discussion Three dietary patterns were identified among the adolescents studied: Salty-Sweet, Plant-Dairy, and Protein. These results depended primarily on gender, physical activity, place of residence, anthropometric factors, and nutritional knowledge. Predicting adherence to these patterns revealed some counterintuitive tendencies in the group of schoolchildrens studied. Schoolchildrens from large cities were less likely to follow the unhealthy pattern. While some studies have produced mixed results20, this finding is consistent with earlier Polish research, which suggests that urban environments may encourage healthier food choices due to the greater availability of nutritious products, health education, and socio-economic status21,22. Interestingly, participants who frequently adhered to the Plant-Dairy Pattern also tended to follow the Salty-Sweet Pattern. This apparent contradiction may result from the ‘health halo’ effect, whereby foods are marketed as healthy despite being nutritionally poor23,24. This can mislead consumers, especially children, into believing their diet is correct. Another possible explanation is moral licensing: after making a ‘good’ dietary choice, individuals may feel entitled to indulge later on, thus maintaining a positive self-image while overlooking less healthy decisions25. This empowers the selection of a more self-indulgent option and is also relevant for younger children26. Similarly, some adolescents may adopt the ‘cheat meal’ mentality, assuming that the occasional consumption of unhealthy food will not affect their overall diet quality. However, this attitude often leads to frequent indulgence and reinforces unhealthy habits27. Schoolchildrens in the third tertile of the Protein Pattern showed lower adherence to the Salty-Sweet Pattern. This is likely due to the satiating effect of protein, which helps control appetite and reduces the desire for processed snacks28. Protein intake stimulates hormones such as GLP-1, CCK, and PYY, which contribute to a longer-lasting feeling of fullness29. Consequently, schoolchildrens consuming more protein may naturally avoid high-sugar and/or high-fat foods. Individuals in the upper tertiles of both the Salty-Sweet and Protein Patterns were more likely to be underweight. Underweight schoolchildrens may try to gain weight by increasing their calorie intake through processed foods, as they perceive these foods to be effective for achieving that goal30. At the same time, many underweight teenagers — especially boys — seek to build muscle by increasing their protein intake, thus adhering to the Protein Pattern31. This aligns with research showing that schoolchildrens focused on weight control consume less energy but more protein32. Constitutional thinness, which is characterised by an innate resistance to weight gain despite normal or high caloric intake and is recognised as the first human model of physiological resistance to weight gain, may also help to explain these results33. These individuals can consume large quantities of food, including high-calorie items, without gaining weight. Nevertheless, being underweight can lead to hormonal imbalances that affect hunger and satiety, and stress and anxiety can further distort eating habits. For instance, high cortisol reactivity has been linked to increased intake of fatty and sugary snacks34. Using the RFMp classification, overweight and obese schoolchildrens were less likely to follow the unhealthy pattern. This association was not evident when BMI alone was considered, which highlights the importance of body fat rather than body weight. Furthermore, schoolchildrens with a high waist-to-height ratio (WHtR), a marker of central obesity, were more likely to adhere to the Plant-Dairy Pattern35. These findings suggest that, when exploring diet–health relationships, body composition indicators may be more precise than BMI. Physical activity was clearly linked to dietary behaviour. Those reporting moderate or high activity levels were less likely to exhibit Salty-Sweet Pattern and more likely to follow Plant-Dairy or Protein Patterns. This reflects the natural interconnection between active lifestyles and healthy diets36,37. Regular exercise improves physical fitness and encourages healthy eating. Therefore, promoting physical activity remains a cornerstone of health promotion for schoolchildrens. Unhealthy dietary patterns – in our case, the Salty-Sweet Pattern – and insufficient physical activity are recognised, modifiable risk factors for obesity, type 2 diabetes, cardiovascular diseases and other non-communicable diseases. According to the WHO, a lack of physical activity is also associated with an approximately 20–30 per cent higher risk of death compared with sufficient physical activity38. Nutritional knowledge also plays a significant role. Teenagers with low or average knowledge were more likely to follow an Salty-Sweet Pattern and less likely to follow a Plant-Dairy one. Similar findings have been reported in Polish studies, showing that better-informed schoolchildrens consume more whole grains and fewer processed foods and energy drinks39. The low levels of knowledge observed here suggest that many schoolchildrens lack the understanding necessary to make informed food choices40. This highlights deficiencies in current nutrition education at school and at home, despite widespread access to various types of media, which, if used skilfully, could become a good carrier of nutritional knowledge. Although some sources point to the beneficial influence of the home environment on healthier eating habits and BMI41, this may apply to only a few. To summarise, knowledge of predictors of dietary behaviour, and thus dietary patterns, makes it possible to identify at-risk groups and tailor interventions to their individual, social and environmental characteristics42. It is important to emphasize that health promotion programs, especially consistent and long-term interventions, enable schoolchildren to become familiar with a variety of nutrition and physical activity. This allows young people to find personal meaning in activity by combining healthy eating with physical activity, which may motivate them to continue as they get older. This will also improve their overall attitude towards health and well-being43. Strengths and limitations The study’s main strength lies in its large, nationally representative sample of 10–12-year-old schoolchildrens, which was collected recently (2022–2023). The questionnaires were completed under the supervision of researchers and teachers in the classrooms, and any queries were clarified. Anthropometric data were obtained objectively by trained researchers using standardised equipment, thereby reducing measurement error. These factors increase confidence that the results accurately represent dietary patterns among Polish youth. However, the cross-sectional design limits causal interpretation, we acknowledge that we are not able to draw causal conclusions but only observations. Further prospective and longitudinal research would clarify the direction of associations and causal relationships, between diet, physical activity, and anthropometric indicators. Another limitation may be the fact that the information on physical activity was based on the participants’ own self-reports; although this was based on the time taken to perform specific activities, it may not accurately reflect their actual behaviour. The self‑reported survey tool relies on subjects remembering what they had to eat the day before and how they exercise in any given week. Other authors44,45 recommend the use of accelerometry as a measurement tool is key, providing objective, reliable, and valid data on children’s physical activity. This will avoid biases inherent in other methods, such as self-reported questionnaires. Accelerometry with machine learning-based processing may be a future way to identify vigorous activities in studies of habitual physical activity based on the proportion of VPA performed in the HE-PECs. Additionally, dietary patterns were identified based on frequency rather than quality or meal regularity, which may obscure some nuances of schoolchildrens’ eating habits. These data indicate reasonable reproducibility and validity of the major dietary patterns defined by factor analysis with data from an FFQ46. Another limitation was the fact that parents whose children have problems with overweight and obesity didn’t want to take part in the study because their children have negative nutrition behaviours. These results may be similar to the dietary behaviours of schoolchildren in European countries. However, cultural differences on other continents may limit the generalisability of these results, for example, in America, Africa, or Asia. Conclusions Schoolchildren exhibit both healthy (Plant-Dairy) and unhealthy (Salty-Sweet) dietary patterns, the explanation of which requires an analysis of numerous variables that appear to be unrelated. Adherence to one pattern does not preclude adherence to another, which highlights the complexity of dietary behaviours in this age group. The level of nutritional knowledge, whether low or high, significantly determined adherence to the Salty-Sweet or Plant-Dairy Patterns, respectively, whilst physical activity proved to be the most significant factor in the Protein Pattern. Furthermore, when analysing the relationship between diet and health, body composition indices (RFMp, WHtR) may be more accurate than BMI. These findings emphasise the need for comprehensive nutrition education and health promotion strategies targeting schoolchildrens. Future research should also explore the psychological, social, and cultural influences on dietary behaviour, such as peer pressure, family habits, and social media. This will help us to understand better how healthy and unhealthy patterns coexist and intermingle among young people. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Abbreviations Short-Form, Multicomponent Dietary Questionnaire to Assess Food Frequency Consumption, Nutrition Knowledge, and Lifestyle in Polish Children and Adolescents Body Mass Index Waist-to-Height Ratio Pediatric Real Fat Mass Index Tri-ponderal Mass Index Principal Component Analysis Measures of Sampling Adequacy Glucagon-like Peptide-1 Cholecystokinine Peptide YY Author contributions Author Contributions/Author Statement Małgorzata Ewa Drywień (MED): Writing—original draft and review and editing, Validation, Visualization, Investigation, Methodology, Data curation, and Formal analysis. Ewa Czarniecka-Skubina (EC-S): Conceptualization, Data curation, Investigation, Methodology, Project administration, Validation, and Writing—review and editing. Jadwiga Hamulka (JH): Conceptualization, Data curation, Investigation, Methodology, Project administration, Validation, and Writing—review and editing. Jerzy Gębski (JG): Data curation, Formal analysis, Software.Olga Januszko (OJ): Writing – original draft, Data curation, Visualization Joanna Sadowska (JS): Data curation, Writing—review and editing. Krystyna Gutkowska (KG): Conceptualization, Methodology, Data curation, Investigation, Writing – review and editing, Project administration, Funding acquisition. Funding This research was financially supported by the Polish Ministry of Education and Science (Agreement No: MEiN/2022/DPI/96 of 7 March 2022). The task entitled ‘Conducting scientific research in the field of nutrition of children and adolescents, developing and implementing a nutritional education program for pupils of grades 1–6 of primary schools. Acronym: ‘Junior-Edu-Żywienie (JEŻ)’. Data availability The datasets used and analysed during the current study are available from project managers, Krystyna Gutkowska (krystyna_gutkowska@sggw.edu.pl), Ewa Czarniecka-Skubina (ewa_czarniecka-skubina@sggw.edu.pl), and Jadwiga Hamulka (jadwiga_hamulka@sggw.edu.pl), on reasonable request. The data cannot be shared publicly because of ethical concerns. Declarations Competing interests Ewa Czarniecka-Skubina - Editor Board Member. Implications and contribution This study has implications for educational and public health policies targeting young people. Adolescents follow distinct dietary patterns shaped by modifiable factors. These findings may inform strategies to prevent nutritional deficiencies through regular dietary assessment and targeted guidance for schoolchildrenadolescents, families, and educators. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Data Availability Statement The datasets used and analysed during the current study are available from project managers, Krystyna Gutkowska (krystyna_gutkowska@sggw.edu.pl), Ewa Czarniecka-Skubina (ewa_czarniecka-skubina@sggw.edu.pl), and Jadwiga Hamulka (jadwiga_hamulka@sggw.edu.pl), on reasonable request. The data cannot be shared publicly because of ethical concerns.