Preprint
Article

This version is not peer-reviewed.

Associations Between Family Affluence and Adolescents’ Fruit and Sweet Consumption in 35 European and North American Countries

Submitted:

16 November 2026

Posted:

17 November 2026

You are already at the latest version

Abstract
Background: This study examined the association between family affluence and adolescents’ fruit and sweet consumption across 35 European and North American countries. Methods: Data from three cycles (2010, 2014, 2018) of the Health Behaviour in School-aged Children (HBSC) surveys were analysed. Cumulative (proportional-odds) mixed-effects models assessed associations between family affluence and fruit and sweet consumption, adjusting for BMI, age, sex, parental communication, and physical activity. Results: Adolescents from low-affluence and medium-affluence households had significantly lower odds of higher fruit consumption than high-affluence peers (OR = 0.63 and 0.79, respectively; p < 0.001), while low-affluence and medium-affluence adolescents also reported modestly lower odds of higher sweet intake (OR = 0.91–0.94, p < 0.001). Age and sex were significant determinants: older adolescents consumed less fruit but more sweets, and girls consumed more of both. High parental communication was associated with increased fruit intake and slightly reduced sweet consumption. Physical activity was positively linked to fruit consumption and showed a complex relationship with sweet intake. Regional factors explained 3% and 8% variance in fruit sweet intake, respectively. Conclusions: Adolescent dietary behaviours are shaped by an interplay of socioeconomic resources, developmental factors, family dynamics, and structural contexts, underscoring the need for multilevel strategies to improve nutritional equity. Interventions should address healthy and discretionary food consumption, engage parents, target age- and sex-specific needs, and consider broader regional and cultural contexts.
Keywords: 
;  ;  ;  ;  
Subject: 
Social Sciences  -   Other

Introduction

The World Health Organisation (WHO) (2024) defined adolescents as persons aged 10 and 19, with significant physical, cognitive, and metabolic, emotional and psychological changes. Adolescents’ dietary patterns and nutritional intake influence long-term health and directly affect their cognitive functioning and physical development. Nonetheless, globally, dietary risks persist among adolescents (Della Corte et al., 2021; Hargreaves et al., 2022; Keats et al., 2018), prompting stringent food regulation within this cohort. Dietary risks predispose adolescents to chronic conditions like obesity, diabetes, and cardiovascular diseases (Artanto, 2026; Muonde et al., 2024), which can result in long-term health complications. Therefore, nutritional mindfulness ensures optimum growth among adolescents, while reinforcing their immune system and cognitive performance (Neta et al., 2021; Sălcudean et al., 2025).
Fruit and sweet consumption, for instance, exerts influence on adolescents, providing essential vitamins, minerals, and requisite energy for growth and development (Hargreaves et al., 2022; Li, 2024). Fruits are sources of fibre, natural sugars, and antioxidants that support digestion and immune functioning while diminishing the risk of chronic diseases (Abobatta, 2021; Maqsood et al., 2025). Moderate sweet consumption from natural ingredients enhances energy levels and cognitive functioning (Panda & Mayrovitz, 2026), while disproportionate consumption of processed sweets engenders obesity and metabolic disorders (Emami et al., 2025). Thus, balanced fruit and sweets consumption helps maintain a healthy diet and proper nutrient absorption. Adolescents’ dietary habits are influenced by an intricate interplay of socio-demographic factors, family affluence, cultural norms, environmental factors, and food accessibility (Jayasinghe, Byrne & Hills, 2025; Liu et al., 2025; Neufeld et al., 2022). Among these, family affluence status is imperative, directly influencing adolescents’ access to nutritious food, healthcare, education, and overall well-being. Family affluence denotes households’ financial and material wealth, which affects education, healthcare, and nutrition (Corell et al., 2021). Often, higher affluence improves the quality of life (QoL), while lower affluence limits resource access. Indeed, family affluence status is important in adolescent dietary research: it directly affects food accessibility, nutritional choices, and food and nutrition-related variables (Agedew, Abebe & Ayelign, 2023). Therefore, understanding its nuanced influences is essential for developing interventions that promote healthier eating behaviours and mitigate nutritional inequalities.
Extant literature has associated adolescents from higher-affluent families with the consumption of more nutrient-rich foods, while those from lower-affluent backgrounds often face dietary challenges and overreliance on processed foods. Several systematic reviews, for example, have associated higher family affluence with lower consumption of sugar-sweetened beverages (SSB) and energy-dense foods (Desbouys et al., 2020; Louey et al., 2024; Purohit et al., 2023). For instance, Purohit et al. (2023) found that adolescents with higher family affluence had 48% lower odds of consuming SSB compared to adolescents from lower-affluent families. Additionally, Gautam et al. (2023) revealed a positive association between high SES, a proxy for family affluence status and unhealthy diets (sweet foods and soft drinks). Hinnig et al. (2018), however, reported no consistent effect of family affluence status on poor diets in medium human development countries. More so, Liu et al. (2021), in a qualitative systematic review, established that family affluence significantly influences adolescent dietary habits by shaping food accessibility, parental influence, nutrition education, and environmental factors. Again, cross-sectional studies have also established that lower SES correlates with higher consumption of unhealthy foods like sweets, snacks, and soft drinks (Keçi̇li̇, 2025). Nonetheless, evidence also suggests that adolescents with poorer family affluence scores (FAS) are less likely to consume sweets but more likely to consume soft drinks than their counterparts from affluent families (Zaborskis et al., 2021).
As established earlier, adolescent dietary patterns are influenced by family affluence, but this relationship is not unidirectional (Hinnig et al., 2018; Keçi̇li̇, 2025; Zaborskis et al., 2021). More so, other sociodemographic factors, cultural norms, and environmental factors interact with SES, shaping food choices and nutritional intake in complex ways (Daly, Kearney & O’Sullivan, 2024; Jayasinghe, Byrne & Hills, 2025; Ogundijo, Tas & Onarinde, 2022). While higher affluence generally improves access to nutritious food, variations in parenting styles, household dynamics, and external influences can modify this effect. Through reasonable control, gatekeeping, healthy role modelling, and support, parental relationships positively mould adolescents’ SSB consumption (Lwin et al., 2025; Van der Horst et al., 2007). Conversely, if parents consume sugary drinks, the adolescent feels sanctioned to do the same. Despite the significance of family affluence in dietary research, large-scale, nationally representative studies across different countries and continents remain scarce. Existing research often focuses on localised, small-scale studies, that limits generalisability and policy reach of the findings.
This paper employed Link and Phelan’s (1995) Fundamental Cause Theory (FCT) to explain the association between socioeconomic status (SES) and adolescents’ fruit and sweet consumption in 35 European and North American countries. In FCT, SES symbolises an array of resources: money, knowledge, prestige, power, and beneficial social connections that explain the emergence and persistence of health inequalities (Clouston & Link, 2021; Link & Phelan, 2025). FCT does not legitimise the extant hierarchical structure; instead, it focuses on “how the powerful secure health and avoid illness” while redistributing disease-causing and health-depleting exposures in vulnerable populations (Clouston & Link, 2021, p.150). Dietary choices and their consequential health inequalities are functions of differential exposures and vulnerabilities to health risks. By this, adolescents’ fruit and sweet consumption is determined by social conditions as resource access increases the scope for engaging in health-protective behaviours (Craveiro et al., 2021). In this study, FCT offered a unique longitudinal understanding of the causational relationships between certain social conditions and adolescents’ fruit and sweet consumption. This inspired evidence-based recommendations on how social policy can intervene to disrupt the social production of health inequality. This study examined the multifaceted relationship between family affluence and adolescent nutrition across diverse populations. By incorporating cross-nationally representative data from two continents, 35 countries, this research provides a broader understanding of how family affluence, sociodemographic factors and behavioural and physiological variables interact to shape adolescents’ dietary behaviours. The findings will contribute to global public health strategies that inform policies and interventions for healthy and equitable nutrition among adolescents.

Methods

Data Source

This study used data from the 2010, 2014, and 2018 Health Behaviour in School-aged Children (HBSC) surveys in high-income countries and regions across Europe and Canada during the 2017-2018 academic year (Boer et al., 2020). The 2010, 2014, and 2018 data were used for the present analysis because they consistently contained information on and measurements for the key variables of interest. The HBSC research team or network is an international alliance of researchers collaborating on a cross-national survey of school students. The HBSC collects data every four years on 11-, 13-, and 15- year-old adolescent boys’ and girls’ health and well-being, social environments, and health behaviours. The sampling procedure for the 2010, 2014, and 2018 surveys followed international guidelines (Currie et al., 2009; Moor et al., 2020). A systematic sampling method was used to identify a list of school classes in each region from the complete list of both public and private schools. Participants were recruited through a cluster sampling approach, using the school class as the primary sampling unit (Currie et al., 2009; Moor et al., 2020; Nardone et al., 2020). Some countries oversampled subpopulations (e.g., by geography and ethnicity), and standardised weights were created to ensure representativeness (Craig et al., 2020). Questionnaires were translated based on a standard procedure to allow comparability between the participating countries. Detailed methodological procedures or decisions about the study have been documented in earlier reports (Currie et al., 2009; Moor et al., 2020).
This study was premised on the analysis of complete, geographically diverse data on 388,187 adolescents across three waves (2010, 2014, and 2018) residing in 35 European and North American countries (see Table 1). It is important to note that the primary aim of this study was not to investigate “temporal” trends in adolescent fruit or sweet consumption, but rather to examine the overall association between family affluence and dietary behaviours across a large, diverse, multinational sample of adolescents from 35 European and North American countries/regions. Additionally, the three survey waves were selected because they consistently contained comparable measures for the key variables of interest and thus ensured methodological consistency across the pooled analyses. Pooling the three survey waves substantially increased the overall sample size and statistical power, improved the precision and stability of the proportional odds mixed-effects estimates, and enhanced the representativeness of the analyses across countries and regions.
As can be observed in Table 1, each country’s sample is broken down by survey year: 2010, 2014, and 2018. These survey waves contribute approximately equally to the overall dataset, with 34.4%, 32.2%, and 33.4% of the participants coming from 2010, 2014, and 2018, respectively. Canada was the single largest contributor to the analytic sample, with 25,893 participants (6.7% of the total sample). Other major contributors included Iceland (21,758 participants; 5.6%), the Czech Republic (18,118; 4.7%), Switzerland (17,671; 4.6%), and France (15,363; 4.0%). Smaller sample contributions came from countries such as England, Wales, and Ukraine.

Measures

Family Affluence

Family affluence is commonly applied in studies to examine the socioeconomic inequalities in health among children and adolescents (Corell et al., 2021; Currie et al., 2008). Family affluence was assessed with the third version of the family affluence scale (FAS III), which is a modified and validated 6-item scale developed by the HBSC network (Torsheim et al., 2016; Curie et al., 2008). The items referred to: having one’s own room, the number of computers in the family, the number of cars in the family, the number of bathrooms in the home, a dishwasher fitted at home and family vacations (Inchley et al., 2018; Moor et al., 2020). The FAS score was calculated by adding the points of the responses to these six items, and higher values indicated higher family affluence (Zaborskis et al., 2022). Further, FAS was coded into country-specific three categories, including those in the lowest 20 % (reference group), those in the medium 60 %, and those in the highest 20 % (Inchley et al., 2016).

Fruit and Sweet Consumption

Fruit and sweet consumption were the main outcome variables. To assess the frequency of fruit and sweet consumption, the following question was asked in the HBSC survey: How many times a week do you eat or drink…? A list of options including fruit, vegetables, sweets, and sugar-added soft drinks. In this study, we only focused on fruits and sweets as these were of our interest. Response options to the question included: ‘never’, ‘less than once a week’, ‘once a week’, ‘two to four days a week’, ‘five to six days a week’, ‘once a day, every day’ and ‘every day, more than once’. For our analysis, these options were recoded and categorised as: ~ ‘3’- High (once daily & more than once daily); ~ ‘2’-Medium (2-4 days a week & 5-6 days a week); and ~ ‘1’-Low(never & less than once a week).

Covariates

Informed by previous evidence about their association with either family affluence or dietary patterns (Queral et al., 2024; Bocci et al., 2024; Borraccino et al., 2016; Rouche et al., 2021; Voráčová et al., 2016), sex (male/female), age (11, 13 and 15 years old), BMI category (normal weight/obesity/overweight/severe thinness/thinness), parental communication (low/high) (Oduro et al., 2025), physical activity levels (daily active/active/somewhat active/inactive) (Oduro et al., 2025) were included in the analysis as covariates.

Statistical Methods and Analysis

In this study, both exploratory and inferential statistical methods were employed. Descriptive statistics on the analytic samples from the three-wave, multi-cohort data (2010, 2014, and 2018) on countries/regions were first generated to provide an overview of the data distribution. For the primary response variables, fruit consumption and sweet consumption, as well as for all independent, categorical variables, chi-square tests of independence were conducted to evaluate potential associations through cross-tabulation analyses. Furthermore, to examine the socioeconomic and behavioural determinants of adolescent fruit consumption across multiple regions, we employed cumulative (proportional-odds) mixed-effects models. This modelling framework accommodates the three-level ordered outcome: Low, Moderate, or High fruit/sweet consumption, while accounting for the clustering of adolescents within 35 countries/regions. In all models, country/region was modelled as a random intercept to capture unobserved regional influences on dietary patterns.
More specifically, let Y i j denote the fruit or sweet consumption category for adolescents i in country/region j , taking values 1 (Low), 2 (Moderate), or 3 (High). We define the cumulative probability as
π i j c = P ( Y i j ≤ c ) ,     c = 1 , 2
and model its logit via
l o g π i j c 1 − π i j c = θ c − ( X i j T ρ +   b j )   
Here,
  • θ 1 and θ 2 are threshold parameters that partition the underlying latent propensity for fruit or sweet consumption between Low vs. Moderate + High and Low + Moderate vs. High, respectively.
  • The linear predictor comprises two components: a vector of fixed effects, ( X i j T ρ ) and a country-specific random intercept, b j b j ∽   N ( 0 , D ) . D is the covariance matrix, and positive semi-definite.
More specifically, four models were fitted. In Model 1, the fixed-effect vector included only family affluence as the primary independent variable and adolescent fruit consumption as the ordered response variable. Model 2 extended this specification by adding factors for BMI category, age group, parental communication, sex and physical-activity level. Models 3 and 4 followed a similar specification but with adolescent sweet consumption status as the response variable. Model estimation proceeded via maximum likelihood with a Laplace approximation for the random-effects integral, implemented in R software using the clmm () function of the ordinal package. Standard errors and 95% confidence intervals were derived from the Hessian of the approximate marginal log-likelihood. We assessed the proportional-odds assumption that each predictor’s effect is constant across both cumulative logit equations through graphical diagnostics of category-specific coefficients and found no evidence of substantial violation. Intraclass correlation coefficients quantified the proportion of total latent-scale variance attributable to between-country differences.
Model performance was summarised by marginal R 2 (variance explained by fixed effects alone) and conditional R 2 (variance explained by both fixed and random effects) following the approach of Nakagawa and Schielzeth (2013). All analyses were conducted in R Software (v4.1.2; R Core Team 2021) with the ordinal, lme4, and MuMIn packages. Statistical inferences were made at the significance defined at a 5% significance level.

Results

Overall, the analyses demonstrated consistent socioeconomic gradients in adolescents’ dietary behaviours across the 35 countries/regions included in the study. Adolescents from low-affluence households had substantially lower odds of higher fruit consumption compared with adolescents from high-affluence households (OR = 0.63, 95% CI: 0.61–0.64), while those from medium-affluence households also showed reduced odds (OR = 0.79, 95% CI: 0.77–0.80). In contrast, the association between family affluence and sweet consumption was comparatively modest, with low- and medium-affluence adolescents showing slightly lower odds of higher sweet intake (ORs ranging from 0.91 to 0.94). Across both dietary outcomes, demographic and behavioural characteristics were important correlates. Older adolescents consumed less fruit but more sweets, females reported higher consumption of both fruits and sweets, and high parental communication was associated with healthier dietary patterns, including increased fruit intake and modestly reduced sweet consumption. Detailed results from the descriptive analyses and proportional odds mixed-effects models are presented in the subsequent sections.
Sample characteristics and Exploratory Analysis of fruit and sweet consumption and family affluence status
To characterise dietary behaviours (sweet and fruit consumption) and their relationship with demographic and health characteristics, an extensive descriptive analysis was performed. This included stratification of participants by both their fruit and sweet consumption levels alongside bivariate comparisons across affluence status, BMI category, age, parental communication, sex, and physical activity status as observed in Table 2 and Table 3. These tables also included Chi-square tests conducted to assess the bivariate relationships between adolescent sweet and fruit consumption and the categorical, independent variables and covariates.
Observing the contingency analysis results (in tables 2 and 3), the distribution of affluence levels varied significantly across sweet and fruit consumption categories. Also, according to the chi-square tests conducted, significant bivariate associations were observed between sweet consumption levels and BMI category (p = 0.039), affluence status (p< 0.001), sex (p<0.001), physical activity (p<0.001) and parental communication status (p< 0.001). Across all sweet consumption groups, the majority of participants were classified as having a normal weight, with 73.5% of low sweet consumers falling into this category. Also, distributional differences in affluence status were evident across the fruit consumption categories. For example, individuals with high fruit consumption were more likely to report high-affluence, whereas lower fruit consumption was relatively more common among individuals from low affluent backgrounds. Furthermore, a Chi-Square association test explored bivariate relationships between fruit consumption and the categorical, independent variables and covariates (see Table 3). Results also demonstrated a significant bivariate association between fruit consumption levels and affluence status (p< 0.001), sex (p<0.001), physical activity (p<0.001), and parental communication status (p< 0.001).
Mixed effects proportional odds modelling for the association between Fruit Consumption and Family Affluence Status
To investigate how family affluence shapes adolescents’ fruit-eating behaviours across diverse regions, we fitted two proportional odds mixed-effects logit models with fruit consumption categorised as Low, Moderate, or High as the ordered response. In Model 1(Table 4), family affluence, coded in three levels (High, Medium, Low), entered the model as the sole predictor of interest, while country served as a random intercept to account for clustering of adolescents within each of the 35 countries/regions. Inferring from this base model, adolescents from low-affluence families had 44% lower odds of consuming more fruit than their high-affluence peers (OR=0.56; 95% CI: 0.55–0.58; p<0.001). Those in the medium-affluence group also displayed significantly diminished fruit-eating tendencies, with 26 percent lower odds relative to the High-affluence reference (OR=0.74; 95% CI: 0.73–0.75; p<0.001).
In the second model, the initial model was extended by adjusting for body mass index (BMI) category, age, sex, parental communication, and physical activity. In this adjusted model, the adjusted odds ratios (AORs) confirmed the persistence of a socioeconomic gradient in fruit consumption even after controlling for relevant covariates. Adolescents from low-affluence households had 37% lower odds of reporting higher fruit consumption compared to their high-affluence counterparts (OR = 0.63, 95% CI: 0.61–0.64, p < 0.001). Similarly, those from medium-affluence families had 21% lower odds (OR = 0.79, 95% CI: 0.77–0.80, p < 0.001). These estimates were attenuated compared to the unadjusted model, indicating that part of the association between affluence and fruit consumption is explained by other demographic and behavioural factors, yet a significant disparity remains. Regarding BMI, only adolescents classified as severely thin were significantly less likely to report higher fruit consumption than those with normal weight (OR = 0.92, 95% CI: 0.86–0.98, p = 0.008). No significant associations were found for overweight (OR = 1.00, 95% CI: 0.98–1.01, p= 0.858) or obesity (OR = 1.00, 95% CI: 0.98–1.03, p= 0.886).
Age also played a statistically significant role in shaping fruit consumption behaviours. Compared to 11-year-olds, 13-year-olds were 22% less likely to consume fruit at higher levels (OR = 0.78, 95% CI: 0.77–0.80, p < 0.001), while 15-year-olds were 32% less likely (OR = 0.68, 95% CI: 0.67–0.69, p < 0.001). Furthermore, adolescents who reported high levels of parental communication had 33% greater odds of consuming more fruit compared to those with low parental engagement (OR = 1.33, 95% CI: 1.31–1.35, p < 0.001). Sex differences were also observed. Adolescent girls were 53% more likely than boys to consume higher levels of fruit (OR = 1.53, 95% CI: 1.51–1.54, p < 0.001). Physical activity exhibited a strong positive association with fruit consumption. Compared to adolescents who were active daily, those who were active less frequently had substantially lower odds of consuming more fruit: 29% lower for those who were active but not daily (OR = 0.71, 95% CI: 0.70–0.72, p < 0.001), 48% lower for the somewhat active group (OR = 0.52, 95% CI: 0.51–0.53, p < 0.001), and 66% lower for inactive adolescents (OR = 0.34, 95% CI: 0.34–0.35, l p < 0.001). With respect to model variance, the intra-class correlation coefficient (ICC) was 0.03, indicating that only 3% of the variance in fruit consumption was attributable to differences between countries. The marginal R 2 value, reflecting variance explained by fixed effects alone, was 0.066, suggesting that the covariates included in the model accounted for 6.6% of the variance in fruit consumption. When the random effects were included, the conditional R 2 increased to 0.094, underscoring the modest yet meaningful improvement in explanatory power gained by accounting for country/region level differences.
Table 4. Model 1: Proportional odds mixed effects modelling results with only Affluence Category as predictor and Fruit Consumption as Response. 
Table 4. Model 1: Proportional odds mixed effects modelling results with only Affluence Category as predictor and Fruit Consumption as Response. 
Response: Adolescent Fruit Consumption
Predictors Odds Ratios CI P-value
Affluence Category [Ref: High]
Low 0.56 0.55 – 0.58 <0.001
Medium 0.74 0.73 – 0.75 <0.001
Intercepts
Low | Moderate 0.16 0.14 – 0.18 <0.001
Moderate | High 1.2 1.06 – 1.36 0.004
Random Effects
Within Region Variance 3.29
Between-Region Variance 0.1
Intra-Class Correlation 0.03
No of Clusters 35
Marginal R2 / Conditional R2 0.010 / 0.039
Table 5. Model 2: Proportional odds mixed effects modelling results with Affluence Category adjusted with multiple covariates and Fruit Consumption as Response. 
Table 5. Model 2: Proportional odds mixed effects modelling results with Affluence Category adjusted with multiple covariates and Fruit Consumption as Response. 
Response: Adolescent Fruit Consumption
Predictors Odds Ratios CI P-value
Affluence Category [Ref: High]
Low 0.63 0.61 – 0.64 <0.001
Medium 0.79 0.77 – 0.80 <0.001
BMI Category [Ref: Normal weight]
Obesity 1 0.98 – 1.03 0.886
Overweight 1 0.98 – 1.01 0.858
Severe Thinness 0.92 0.86 – 0.98 0.008
Thinness 0.99 0.96 – 1.03 0.681
Age Category (Ref: 11-year-olds)
13-year-olds 0.78 0.77 – 0.80 <0.001
15-year-olds 0.68 0.67 – 0.69 <0.001
Parental Comm. (Ref: Low)
High 1.33 1.31 – 1.35 <0.001
Sex (Ref: Male)
Female 1.53 1.51 – 1.54 <0.001
Physical Activity (Ref: Daily Active)
Active 0.71 0.70 – 0.72 <0.001
Somewhat active 0.52 0.51 – 0.53 <0.001
Inactive 0.34 0.34 – 0.35 <0.001
Intercepts
Low | Moderate 0.11 0.10 – 0.13 <0.001
Moderate | High 0.95 0.85 – 1.06 0.334
Random Effects
Within Region Variance 3.29
Between-Region Variance 0.1
Intra-Class Correlation 0.03
No of Clusters 35
Marginal R2 / Conditional R2 0.066 / 0.094
Mixed effects proportional odds modelling for the association between Sweet Consumption and Family Affluence Status
Similar to the fruit consumption analysis, we specified a baseline model (Model 3 (Table 6)) in which family affluence was the sole predictor of adolescents’ sweet consumption. Here, both low and medium-affluence adolescents showed modest (about 6%) but significant reductions in the odds of reporting higher sweet-consumption categories compared with their high-affluence peers (OR = 0.94 for both low and medium, 95% CI: 0.92–0.95 and 0.93–0.96, respectively, p < 0.001). Model 4 introduced the same set of demographic and behavioural covariates used previously for fruit consumption. After adjustment, low-affluence adolescents remained slightly less likely (9%) to report higher sweet intake than high-affluence adolescents (OR = 0.91, 95% CI: 0.89–0.93, p < 0.001), as did those of medium affluence (OR = 0.94, 95% CI: 0.92–0.95, p < 0.001). BMI category showed no significant associations with sweet consumption—obese and overweight adolescents had ORs indistinguishable from 1, and only thinness trended upward (OR = 1.03, p = 0.135).
Age was a statistically significant determinant: 13-year-olds had 41% higher odds (OR = 1.41, 95% CI: 1.39–1.43, p < 0.001) and 15-year-olds 49% higher odds (OR = 1.49, 95% CI: 1.46–1.51, p < 0.001) of greater sweet consumption than 11-year-olds. High parental communication was associated with a modest reduction in sweet intake (OR = 0.93, 95% CI: 0.91–0.95, p < 0.001). Female adolescents consumed more sweets than males (OR = 1.14, 95% CI: 1.13–1.15, p < 0.001), and lower levels of physical activity were linked to slightly lower sweet consumption—active (but not daily) teens had OR = 0.91 (95% CI: 0.89–0.92), somewhat active OR = 0.89 (95% CI: 0.88–0.91), while fully inactive adolescents did not differ significantly (OR = 1.00, p = 0.869). Country/region variance and ICC remained at 0.30 and 0.08, respectively. The inclusion of all covariates modestly improved explained variance: fixed effects now accounted for 1.1% of variance (marginal R² = 0.011), and total variance explained rose to 9.4% (conditional R² = 0.094) when country/region clustering was included.
Table 7. Model 4: Proportional odds mixed effects modelling results with Affluence Category adjusted with multiple covariates and Sweet Consumption as Response. 
Table 7. Model 4: Proportional odds mixed effects modelling results with Affluence Category adjusted with multiple covariates and Sweet Consumption as Response. 
Sweet Consumption
Predictors Odds Ratios CI P-value
Affluence Category [Ref: High]
Low 0.91 0.89 – 0.93 <0.001
Medium 0.94 0.92 – 0.95 <0.001
BMI Category [Ref: Normal weight]
BMI category [Obesity] 1.01 0.98 – 1.04 0.461
BMI category [Overweight] 1 0.98 – 1.01 0.573
BMI category [Severe 1 0.94 – 1.07 0.968
BMI category [Thinness] 1.03 0.99 – 1.07 0.135
Age Category (Ref: 11-year-olds)
13-year-olds 1.41 1.39 – 1.43 <0.001
15-year-olds 1.49 1.46 – 1.51 <0.001
Parental Comm. (Ref: Low)
High 0.93 0.91 – 0.95 <0.001
Sex (Ref: Male)
Female 1.14 1.13 – 1.15 <0.001
Physical Activity (Ref: Daily Active)
Active 0.91 0.89 – 0.92 <0.001
Somewhat active 0.89 0.88 – 0.91 <0.001
Inactive 1 0.98 – 1.02 0.869
Intercepts
Low | Moderate 0.51 0.42 – 0.61 <0.001
Moderate | High 3.58 2.95 – 4.34 <0.001
Random Effects
Within Region Variance 3.29
Between-Region Variance 0.3
Intra-Class Correlation 0.08
No of Clusters 35
Marginal R2 / Conditional R2 0.011 / 0.094

Discussion

This study examined the association between family affluence and adolescents’ consumption of sweet and fruit products in European and North American countries. Specifically, it sheds light on the interplay between family affluence, sociodemographic characteristics, and behavioural and physiological factors in shaping adolescents’ dietary behaviour. Findings show that adolescents from low- and medium-affluence households had 37% and 21% lower odds, respectively, of reporting higher fruit consumption compared to their peers from high-affluence households, even after adjusting for BMI, age, sex, parental communication, and physical activity. In contrast, for sweet consumption, adolescents from low- and medium-affluence families reported modestly lower odds (6 - 9%) of higher sweet intake than their peers from high-affluence families. Although adolescents from higher-affluence households consumed more fruits and sweets than their peers from low- and middle-affluence households, the likelihood of intake varied significantly by food type. The patterns suggest that greater affluence promotes healthier dietary behaviours through increased fruit consumption, and it may also facilitate higher access to sweets. These patterns demonstrate that higher affluence supports healthier dietary behaviours in the case of fruit intake but may also facilitate greater access to sweet foods, highlighting that economic resources influence food choices differently depending on food type.
Our findings align with the Fundamental Cause Theory, which posits that socioeconomic resources such as money act as flexible assets that can simultaneously enable health-promoting behaviours, like fruit consumption, and risk-enhancing behaviours, such as greater access to sweet foods (Link and Phelan 2025). In other words, individuals with greater access to flexible resources, such as income, knowledge, social capital, and power, are better able to avoid risks and pursue health-promoting opportunities, even as the specific mechanisms linking socioeconomic status (SES) and health evolve (Link & Phelan, 2025). In this study, family affluence functioned as such a flexible resource, shaping adolescents’ dietary opportunities by expanding their capacity to access and select from a wider range of foods. In this context, affluence does not operate in a unidirectional manner towards better health, but rather expands the range of available choices, including both nutritious and unhealthy options. This underlines the paradox that while higher income can reduce barriers to healthy diets, it can also facilitate greater consumption of energy-dense, processed foods, thereby reinforcing the complexity of socioeconomic influences on diet-related health outcomes (Fismen et al., 2012; Lou et al., 2024). This dual pattern is consistent with FCT’s central proposition that flexible resources can be mobilised in ways that reflect prevailing market conditions (Clouston & Link, 2021). In contemporary food systems characterised by aggressive marketing and ubiquitous availability of ultra-processed products, greater purchasing power can facilitate both healthier and less healthy consumption (Baker et al., 2020). Partly in line with our results, the ‘SHOPPER study’ in the United States found that lower-income families tended to purchase fewer healthful foods compared with their higher-income counterparts (French, Tangney et al. 2019). Interestingly, findings from the 2002 and 2014 Czech HBSC surveys indicate that adolescents with greater family affluence consumed fewer sweets (Voráčová et al. 2016). The difference in findings may reflect cultural variations, higher health awareness, differences in sample characteristics and study methodologies, as well as evolving social norms that discourage sugary food consumption among higher socioeconomic groups. Nevertheless, our findings suggest that affluence amplifies dietary opportunities rather than uniformly improving diet quality, thereby explaining why higher SES supports increased fruit intake but may also facilitate modestly higher sweet consumption.
Demographic characteristics were consistently associated with both dietary behaviours (consumption of sweet and fruit products). Firstly, we found that age was inversely associated with fruit consumption, with older adolescents less likely to consume fruits than younger adolescents. On the other hand, age was positively associated with sweet consumption, with the results showing that 13- and 15-year-olds had 41% and 49% higher odds of consuming more sweets than 11-year-olds, respectively. The inverse association between age and fruit consumption, alongside the positive association with sweet intake, may reflect developmental and social factors in adolescence. Older adolescents often have greater autonomy over their food choices and increased exposure to peers, school environments, and marketing that promote energy-dense, sugary foods (Neufeld et al., 2022). Consistent with FCT (Link & Phelan, 2025), these behaviours are embedded within broader systems, including peer norms, marketing exposure, school environments, and time constraints that mediate the relationship between SES and diet. Additionally, busy schedules, academic pressures, and convenience-oriented eating may reduce fruit consumption, while sweets remain readily accessible and socially reinforced (Middha et al., 2021). These patterns suggest that as adolescents grow, their dietary behaviours can shift toward less healthful options, highlighting the importance of targeted interventions for older age groups.
Sex differences in food consumption were also notable. Girls consumed more fruit and more sweets than boys, which may reflect gendered preferences, social norms, or differential exposure to marketing strategies. Possibly, girls’ higher consumption of both fruit and sweets may be associated with a greater health awareness and adherence to social norms promoting fruit intake, stronger preferences for sweet flavours, and parental or household practices that encourage healthy eating while allowing occasional treats (Lommi et al., 2021; Barrea et al., 2024). Peer influence, targeted marketing, and psychosocial factors such as emotional eating may also contribute, leading to higher intake of both nutritious and indulgent foods compared with boys (Khoshghadam & Rajabi, 2024). This is consistent with a study conducted at an Endocrinology, Diabetology, and Andrology Unit in Italy, which analysed medical records of 968 females and 680 males and found that females consumed more fruits and sweets than males (Barrea et al., 2024). Surprisingly, BMI was largely unrelated to either dietary outcome, except for thin adolescents, indicating that current body weight does not strongly predict either healthy or discretionary food intake in this population. These findings suggest that adolescent food choices are influenced more by age-related independence, gendered socialisation, and environmental contexts than by just weight status, emphasising the importance of targeting behavioural and contextual factors in interventions aimed at promoting healthy eating.
Moreover, behavioural factors contributed significantly to dietary patterns in the sample. High parental communication was associated with higher fruit intake and modestly reduced sweet consumption, emphasising the protective role of engaged parenting. The associations involving parental communication underscore the importance of non-material resources within FCT (Link & Phelan, 2025). Effective communication likely reflects the transmission of health knowledge, behavioural norms, and emotional support forms of social capital that influence adolescents’ food practices (Lee & Kam, 2015). Socioeconomically advantaged households may therefore confer both material and relational resources that jointly support healthier dietary regulation (Halkier & Holm, 2021). Physical activity was positively associated with fruit consumption, with less active adolescents consuming substantially less fruit. For sweet consumption, however, physical activity displayed a more complex pattern. Somewhat active and active adolescents consumed slightly fewer sweets than fully active peers, while fully inactive adolescents did not differ significantly. This may indicate nuanced interactions among energy expenditure, reward-seeking behaviours, and social or age-related influences. Thus, behavioural factors appear to play a crucial role in shaping adolescent dietary patterns, reflecting the influence of family and lifestyle behaviours on food choices. High parental communication likely promotes greater fruit intake and modestly reduces sweet consumption by nurturing health-conscious behaviours, setting expectations, and modelling balanced eating, consistent with evidence that engaged parenting supports healthier adolescent diets (Grey et al., 2022; Norton et al., 2023). The positive response association between physical activity and fruit intake may reflect broader health-oriented lifestyles, where more active adolescents are also more likely to consume nutrient-rich foods as part of a holistic approach to wellbeing. Conversely, the more complex relationship between physical activity and sweet consumption could indicate that energy expenditure and reward-seeking behaviours interact in nuanced ways: somewhat active and active adolescents may balance caloric intake with activity levels, consciously or unconsciously limiting sweets, whereas fully active adolescents may have higher energy needs or perceive indulgence as acceptable, and fully inactive adolescents may not differ due to limited opportunities or motivation for dietary regulation. This aligns with a systematic review and meta-analysis showing that parent-based interventions can effectively influence adolescents’ physical activity, screen time, and dietary behaviours (Champion et al., 2022). These patterns highlight that behavioural factors influence dietary choices not only directly but also through interactions with social, psychological, and developmental processes, emphasising the need for integrated interventions that consider both nutrition and physical activity in tandem.
The study also highlights the importance of contextual, regional factors in dietary behaviour. For fruit consumption, the intraclass correlation coefficient (ICC) was 0.03, indicating that only 3% of variance was attributable to country-level differences. In contrast, regional influences on sweet consumption were somewhat larger (ICC = 0.08), suggesting that broader cultural norms, national policies, food marketing, and availability may play a more substantial role in ‘discretionary’ food intake. Marginal and conditional R² values further reinforce that while individual and familial characteristics explain a modest proportion of variance, accounting for regional clustering meaningfully improves model fit and understanding of dietary behaviours. This finding suggests that, although family affluence, age, sex, and parental communication are important predictors, contextual factors specific to each country or region, such as cultural dietary norms, national food policies, availability of sweet products, and marketing practices, also influence adolescents’ sweet consumption.
The relatively low marginal R² indicates that individual-level fixed effects alone explain very little of the variance, whereas the conditional R² shows that including regional clustering substantially increases the explained variance. This shows the importance of accounting for hierarchical structure in cross-national dietary research and suggests that interventions aimed at reducing sugar consumption may need to be tailored not only to individual behaviours but also to regional or country-level contexts. For example, public health strategies may differ in effectiveness depending on the regulatory environment, cultural acceptability of sweets, or school-based nutrition policies in different countries. Overall, the findings highlight that adolescent dietary behaviours are affected by a combination of personal, familial, and broader societal factors, with regional influences playing a modest but non-negligible role.

Implications for Public Health, Practice, and Research

The findings of this study have several important implications. From a public health perspective, the contrasting associations of family affluence with fruit and sweet consumption underscore the need to address socioeconomic disparities in both undernutrition and overconsumption of discretionary foods. While higher affluence supports healthier fruit intake, it also appears to facilitate access to sweets, highlighting the dual influence of economic resources on adolescent diet. Interventions aimed at reducing sugar intake should consider the accessibility and affordability of sugary foods, particularly in high-affluence households, while simultaneously promoting equitable access to fruits across all socioeconomic groups. Age- and gender-specific patterns suggest that older adolescents and female adolescents may benefit from tailored strategies, such as school-based nutrition education, awareness campaigns, or parental guidance programs emphasising balanced eating habits and mindful consumption.
For practice, the protective role of parental communication in both fruit and sweet consumption indicates that family engagement is critical. Health practitioners, school counsellors, and community organisations should actively involve parents in dietary interventions, promoting open discussions, modelling healthy behaviours, and providing practical strategies for moderating sweet intake while encouraging fruit consumption. The observed associations between physical activity and dietary behaviours further suggest that interventions should adopt a holistic approach, integrating nutrition and physical activity rather than addressing these behaviours in isolation (Harner, 2024).
In terms of research implications, the relatively low variance explained by individual-level covariates for sweet consumption and moderate variance for fruit consumption indicate that unmeasured contextual and environmental factors, possibly including cultural norms, peer influences, school policies, neighbourhood food environments, and marketing exposure, may substantially affect adolescent dietary patterns. Future studies should employ multi-level or longitudinal designs to explore causal pathways, the interplay between dietary patterns and physical activity, and interactions with socio-demographic characteristics. Cross-country analyses may also help identify policy approaches and best practices for promoting healthy eating and reducing excessive sugar intake among adolescents in diverse cultural and economic contexts.

Strengths and Limitations

This study has several notable strengths. First, it employed a large, multi-country, multi-cohort dataset spanning 35 European and North American countries, enhancing the generalisability of findings and allowing robust cross-national comparisons. Second, the use of cumulative (proportional-odds) mixed-effects models enabled the analysis of ordered categorical outcomes (Low, Moderate, High consumption) while accounting for the clustering of adolescents within countries/regions, providing more precise estimates of individual- and contextual-level effects. Third, both unadjusted and adjusted models were employed, allowing us to assess the independent association of family affluence with dietary behaviours while controlling for key demographic and behavioural covariates, including age, sex, BMI, parental communication, and physical activity. Fourth, the study examined both healthy (fruit) and discretionary (sweet) food consumption, providing nuanced insights into how SES may differentially influence dietary patterns. Finally, the proportional-odds assumption was checked and satisfied, ensuring appropriate model specification and valid inference.
Despite these strengths, several limitations should be acknowledged. First, the cross-sectional nature of the data limits causal inference, as observed associations cannot confirm temporal relationships between family affluence and dietary behaviours. Second, dietary intake and behavioural measures were self-reported, which may introduce recall bias or social desirability bias. Third, although the models included several key covariates, unmeasured factors, such as peer influence, school or neighbourhood food environments, cultural norms, and exposure to food marketing, may have contributed to residual confounding. Fourth, while regional clustering was accounted for, the relatively low intraclass correlation for fruit consumption and modest explained variance indicate that other contextual determinants remain unexplored. This contextual nuance is now noted to aid the interpretation of the findings. Further consideration relates to our use of pooled data from the 2010, 2014, and 2018 HBSC survey waves. This approach was intentional and aligned with the study’s primary aim of examining socioeconomic differences in adolescent dietary behaviours across a large and diverse multinational sample. Pooling the three waves increased statistical power, enhanced the precision of mixed-effects estimates, and ensured broad representativeness across 35 countries and regions. These survey years were selected specifically because they contain fully comparable measures for all key variables, allowing for methodologically consistent analyses. We acknowledge, however, that dietary behaviours and socioeconomic inequalities can shift over time, as illustrated by Nicolas et al. (2023). Although our study was not designed to assess temporal trends, combining survey waves may mask country-specific changes. Lastly, variations in data collection methods across countries and waves could introduce heterogeneity, although multi-level modelling partially mitigates this concern.

Conclusions

This study provides important evidence that family affluence is a key determinant of adolescent dietary behaviours, but its effects vary by food type. Higher affluence was associated with substantially greater fruit consumption, reflecting healthier dietary patterns, whereas sweet consumption showed only modestly higher intake among high-affluence adolescents. This contrast highlights that economic resources can support both healthy and discretionary food choices, underscoring the complexity of socioeconomic influences on diet. Demographic and behavioural factors, including age, sex, parental communication, and physical activity, also significantly influenced dietary patterns. Older adolescents and females were more likely to consume sweets and less likely to maintain high fruit intake, indicating the need for age- and gender-specific interventions. Strong parental engagement emerged as a protective factor, promoting greater fruit intake and modestly reducing sweet consumption, while physical activity was positively associated with fruit intake and displayed a complex relationship with sweet consumption. Regional differences accounted for a modest proportion of variance, particularly for sweet intake, suggesting that cultural norms, food availability, marketing, and policy contexts at the country or regional level can influence adolescent dietary behaviours, albeit to a smaller extent than individual- and family-level factors. The low variance explained by fixed effects alone indicates that unmeasured environmental, social, and cultural determinants remain important areas for investigation. Thus, addressing adolescent diet requires a holistic, multi-level approach that considers economic, familial, behavioural, and environmental factors simultaneously, ensuring interventions are both effective and equitable across diverse populations.

Author Contributions

MSO: conceptualisation, statistical methodology and analysis and writing; PP: conceptualisation, methodology, and writing; AKM: writing; MAP: writing; IYA: conceptualisation, analysis and writing. All the authors reviewed and contributed to the publication of this paper.

Funding

This research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

The data supporting the findings of this study are available from the HBSC, but restrictions apply; they were used under license for the current study and are not publicly available. Data are, however, available from the corresponding author (24049064r@connect.polyu.hk) upon reasonable request and with permission of the data custodians.

Acknowledgments

We thank the 2010, 2014, and 2018 HBSC survey team/network, the coordinator and the Data Bank Manager for granting us access to the datasets. We duly acknowledge all school children who participated in the surveys.

Disclosure statement

The authors declare no competing interests.

References

  1. Abobatta, W. F. Nutritional and healthy benefits of fruits. Diabetes 2021, 40(2), 31979–31983. [Google Scholar] [CrossRef]
  2. Agedew, E.; Abebe, Z.; Ayelign, A. Dietary patterns in relation with nutritional outcomes and associated factors among adolescents: Implications for context-specific dietary intervention for the Agrarian Community, Northwest Ethiopia. Front. Nutr. 2023, 10, 1–10. [Google Scholar] [CrossRef] [PubMed]
  3. Artanto, K. Ultra-Processed Food Consumption Patterns as an Early Risk Factor for Obesity and Cardiovascular Disease in Today’s Adolescents: A Literature Review. J. Divers. Med. Res. Medicosphere 2026, 3(2), 97–110. [Google Scholar] [CrossRef]
  4. Baker, P.; Machado, P.; Santos, T.; Sievert, K.; Backholer, K.; Hadjikakou, M.; Lawrence, 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(12), 1–51. [Google Scholar] [CrossRef] [PubMed]
  5. Barrea, L.; Verde, L.; Suárez, R.; Frias-Toral, E.; Vásquez, C. A.; Colao, A.; Savastano, S.; Muscogiuri, G. Sex-differences in Mediterranean diet: a key piece to explain sex-related cardiovascular risk in obesity? A cross-sectional study. J. Transl. Med. 2024, 22(1), 44–60. [Google Scholar] [CrossRef] [PubMed]
  6. Bocci, B. M.; Lipari, D.; Pammolli, A.; Simi, R.; Frongillo, E.; Miserendino, A.; Lazzeri, G.; et al. How family structure can influence adolescents’ eating habits: An Italian Health Behaviour in School-Aged Children (HBSC) 2022 sample. Children 2024, 11(11), 1–9. [Google Scholar] [CrossRef] [PubMed]
  7. Boer, M.; van den Eijnden, R. J. J. M.; Boniel-Nissim, M.; Wong, S. L.; Inchley, J. C.; Badura, P.; Craig, W. M.; Gobina, I.; Kleszczewska, D.; Klanšček, H. J.; Stevens, G. W. J. M. Adolescents’ intense and problematic social media use and their well-being in 29 countries. J. Adolesc. Health 2020, 66(6), 89–99. [Google Scholar] [CrossRef] [PubMed]
  8. Borraccino, A.; Lemma, P.; Berchialla, P.; Cappello, N.; Inchley, J.; Dalmasso, P. & Italian HBSC 2010 Group Unhealthy food consumption in adolescence: role of sedentary behaviours and modifiers in 11-, 13-and 15-year-old Italians. Eur. J. Public Health 2016, 26(4), 650–656. [Google Scholar] [PubMed]
  9. Craig, W.; Boniel-Nissim, M.; King, N.; Walsh, S. D.; Boer, M.; Donnelly, P. D.; Pickett, W.; et al. Social media use and cyber-bullying: A cross-national analysis of young people in 42 countries. J. Adolesc. Health 2020, 66(6), 100–108. [Google Scholar] [CrossRef] [PubMed]
  10. Craveiro, D.; Marques, S.; Zvěřinová, I.; Máca, V.; Ščasný, M.; Chiabai, A.; Taylor, T.; et al. Explaining inequalities in fruit and vegetable intake in Europe: the role of capabilities, opportunities and motivations. Appetite 2021, 165, 1–11. [Google Scholar] [CrossRef] [PubMed]
  11. Champion, K. E.; Gardner, L. A.; McCann, K.; Hunter, E.; Parmenter, B.; Aitken, T.; Chapman, C.; Spring, B.; Thornton, L.; Slade, T.; Teesson, M.; Newton, N. C. Parent-based interventions to improve multiple lifestyle risk behaviours among adolescents: A systematic review and meta-analysis. Prev. Med. 164 2022, 1–43. [Google Scholar]
  12. Clouston, S. A.; Link, B. G. A retrospective on fundamental cause theory: State of the literature and goals for the future. Annu. Rev. Sociol. 2021, 47(1), 131–156. [Google Scholar] [CrossRef] [PubMed]
  13. Corell, M.; Chen, Y.; Friberg, P.; Petzold, M.; Löfstedt, P. Does the family affluence scale reflect actual parental earned income, level of education and occupational status? A validation study using register data in Sweden. BMC Public Health 2021, 21, 1–11. [Google Scholar] [CrossRef] [PubMed]
  14. Currie, C.; Nic Gabhainn, S.; Godeau, E.; International HBSC Network Coordinating Committee. The Health Behaviour in School-aged Children: WHO Collaborative Cross-National (HBSC) study: origins, concept, history and development 1982–2008. Int. J. Public Health 2009, 54 (Suppl 2), 131–139. [Google Scholar] [CrossRef] [PubMed]
  15. Daly, A. N.; Kearney, J. M.; O’Sullivan, E. J. The underlying role of food guilt in adolescent food choice: A potential conceptual model for adolescent food choice negotiations under circumstances of conscious internal conflict. Appetite 2024, 192, 1–14. [Google Scholar] [CrossRef] [PubMed]
  16. Della Corte, K.; Fife, J.; Gardner, A.; Murphy, B. L.; Kleis, L.; Della Corte, D.; Buyken, A. E.; et al. World trends in sugar-sweetened beverage and dietary sugar intakes in children and adolescents: a systematic review. Nutr. Rev. 2021, 79(3), 274–288. [Google Scholar] [CrossRef] [PubMed]
  17. Desbouys, L.; Méjean, C.; De Henauw, S.; Castetbon, K. Socio-economic and cultural disparities in diet among adolescents and young adults: a systematic review. Public Health Nutr. 2020, 23(5), 843–860. [Google Scholar] [CrossRef] [PubMed]
  18. Emami, O.; Nikparast, A.; Sepehrinia, M.; Hadi, S.; Mirzay Razzaz, J.; Homayounfar, R. Ultra-processed food consumption and the risk of developing metabolic dysfunction-associated steatotic liver disease (MASLD): a five-year prospective cohort study in Iranian adults. J. Health Popul. Nutr. 2025, 44(1), 1–14. [Google Scholar] [CrossRef] [PubMed]
  19. Fismen, A.-S.; Samdal, O.; Torsheim, T. Family affluence and cultural capital as indicators of social inequalities in adolescent’s eating behaviours: a population-based survey. BMC Public Health 2012, 12(1), 1–9. [Google Scholar] [CrossRef] [PubMed]
  20. French, S. A.; Tangney, C. C.; Crane, M. M.; Wang, Y.; Appelhans, B. M. Nutrition quality of food purchases varies by household income: the SHoPPER study. BMC Public Health 2019, 19(1), 1–17. [Google Scholar] [CrossRef] [PubMed]
  21. Gautam, N.; Dessie, G.; Rahman, M. M.; Khanam, R. Socioeconomic status and health behaviour in children and adolescents: a systematic literature review. Front. Public Health 2023, 11, 1–23. [Google Scholar] [CrossRef] [PubMed]
  22. Grey, E. B.; Atkinson, L.; Chater, A.; Gahagan, A.; Tran, A.; Gillison, F. B. A systematic review of the evidence on the effect of parental communication about health and health behaviours on children’s health and wellbeing. Prev. Med. 2022, 159, 1–11. [Google Scholar] [CrossRef] [PubMed]
  23. Halkier, B.; Holm, L. Linking socioeconomic disadvantage to healthiness of food practices: Can a practice-theoretical perspective sharpen everyday life analysis? Sociol. Health Illn. 2021, 43(3), 750–763. [Google Scholar] [CrossRef] [PubMed]
  24. Hargreaves, D.; Mates, E.; Menon, P.; Alderman, H.; Devakumar, D.; Fawzi, W.; Patton, G. C.; et al. Strategies and interventions for healthy adolescent growth, nutrition, and development. The Lancet 2022, 399(10320), 198–210. [Google Scholar] [CrossRef] [PubMed]
  25. Harner, T. The Holistic Approach Integrating Health Physical Education and Recreation. Health Sci. J. 2024, 18(4), 1–3. [Google Scholar]
  26. Hinnig, P. D. F.; Monteiro, J. S.; De Assis, M. A. A.; Levy, R. B.; Peres, M. A.; Perazi, F. M.; Canto, G. D. L.; et al. Dietary patterns of children and adolescents from high, medium and low human development countries and associated socioeconomic factors: a systematic review. Nutrients 2018, 10(4), 1–25. [Google Scholar] [CrossRef] [PubMed]
  27. Inchley, J.; Currie, D.; Young, T. Growing up unequal: gender and socioeconomic differences in young people’s and well-being. Health Behaviour in School-aged Children (HBSC) study: international report from the 2013/2014 survey. WHO Regional Office for Europe. 2016. Available online: http://www.euro.who.int/en/healthtopics/Life-stages/child-and-adolescent-health/health-behaviourin-school-aged-children-hbsc. (accessed on 19 May 2026).
  28. Inchley, J.; Currie, D.; Cosma, A.; Samdal, O. Health Behaviour in School-aged Children (HBSC) study protocol: background, methodology and mandatory items for the 2017/18 survey; CAHRU: St Andrews, 2018; Available online: https://assets-eu-01.kc-usercontent.com/4624460d-da56-01fc-8b6f-a2cc597f19c1/c0979633-d6ad-4a5a-a939-18c51c9ebb0f/Allegato%20A%20Protocollo%20HBSC%20internazionale%202017-18.pdf (accessed on 19 May 2026).
  29. Jayasinghe, S.; Byrne, N. M.; Hills, A. P. Cultural influences on dietary choices. Prog. Cardiovasc. Dis. 2025, 90, 1–5. [Google Scholar]
  30. Keats, E. C.; Rappaport, A. I.; Shah, S.; Oh, C.; Jain, R.; Bhutta, Z. A. The dietary intake and practices of adolescent girls in low-and middle-income countries: a systematic review. Nutrients 2018, 10(12), 1–14. [Google Scholar] [CrossRef] [PubMed]
  31. Keçi̇li̇, M. Ç. Effect of Socioeconomic Status of Parents on Nutrition Habits Among Children: Türkïye Child Survey Application. Child Indic. Res. 2025, 18(2), 719–734. [Google Scholar] [CrossRef]
  32. Khoshghadam, L.; Rajabi, R. The role of emotions in food consumption choice: Systematic review and directions for future studies. Int. J. Consum. Stud. 2024, 48(1), 1–16. [Google Scholar] [CrossRef]
  33. Lee, C. J.; Kam, J. A. Why does social capital matter in health communication campaigns? Commun. Res. 2015, 42(4), 459–481. [Google Scholar] [CrossRef]
  34. Li, S. Nutrition and dietary balance on growth and development of children and adolescents. In Biological Sciences and Environmental Health; CRC Press, 2024; pp. 115–121. [Google Scholar]
  35. Link, B. G.; Phelan, J. Social conditions as fundamental causes of disease. J. Health Soc. Behav. 1995, 35, 80–94. [Google Scholar] [CrossRef]
  36. Link, B. G.; Phelan, J. C. Fundamental-Cause Theory. In Medical Sociology on the Move: Revised Edition including New Directions in Theory; Springer Nature Switzerland: Cham, 2025; pp. 65–82. [Google Scholar] [CrossRef]
  37. Liu, K. S.; Chen, J. Y.; Ng, M. Y.; Yeung, M. H.; Bedford, L. E.; Lam, C. L. How does the family influence adolescent eating habits in terms of knowledge, attitudes and practices? A global systematic review of qualitative studies. Nutrients 2021, 13(11), 1–29. [Google Scholar] [CrossRef] [PubMed]
  38. Liu, K. S.; Chen, J. Y.; Kang, Q.; Sun, K. S.; Tsang, J. P.; Ip, P.; Lam, C. L.; et al. Assessing family influence on adolescent healthy eating: insights from knowledge, attitudes, and practices in a cross-sectional survey in Hong Kong. BMC Public Health 2025, 25(1), 1–9. [Google Scholar] [CrossRef] [PubMed]
  39. Lommi, S.; Engberg, E.; Tuorila, H.; Kolho, K.-L.; Viljakainen, H. Sex-and weight-specific changes in the frequency of sweet treat consumption during early adolescence: a longitudinal study. Br. J. Nutr. 2021, 126(10), 1592–1600. [Google Scholar] [CrossRef] [PubMed]
  40. Louey, J.; He, J.; Partridge, S. R.; Allman-Farinelli, M. Facilitators and barriers to healthful eating among adolescents in high-income countries: a mixed-methods systematic review. Obes. Rev. 2024, 25(11), 1–24. [Google Scholar] [CrossRef] [PubMed]
  41. Lwin, M. O.; Seet, A.; Leo, S.; Schulz, P. J. Parent-Child Mutual Influences on Sugar-Sweetened Beverage Consumption Behaviors: Actor-Partner Analysis. JMIR Pediatr. Parent. 2025, 8(1), 1–12. [Google Scholar] [CrossRef] [PubMed]
  42. Maqsood, S.; Arshad, M. T.; Ikram, A.; Gnedeka, K. T. Fruit-Based Diet and Gut Health: A Review. Food Sci. Nutr. 2025, 13(5), 1–22. [Google Scholar] [CrossRef] [PubMed]
  43. Middha, B.; Strengers, Y.; Lewis, T.; Horne, R. Spatio-temporalities of convenience eating for sustainability outcomes at an inner-urban university. Geogr. Res. 2021, 59(3), 407–423. [Google Scholar] [CrossRef]
  44. Moor, I.; Winter, K.; Bilz, L.; Bucksch, J.; Finne, E.; John, N.; Richter, M.; et al. The 2017/18 Health Behaviour in School-aged Children (HBSC) study–Methodology of the World Health Organization’s child and adolescent health study. J. Health Monit. 2020, 5(3), 88–102. [Google Scholar] [CrossRef] [PubMed]
  45. Muonde, M.; Olorunsogo, T. O.; Ogugua, J. O.; Maduka, C. P.; Omotayo, O. Global nutrition challenges: A public health review of dietary risks and interventions. World J. Adv. Res. Rev. 2024, 21(1), 1467–1478. [Google Scholar] [CrossRef]
  46. Nakagawa, S.; Schielzeth, H. A general and simple method for obtaining R2 from generalised linear mixed-effects models. Methods Ecol. Evol. 2013, 4(2), 133–142. [Google Scholar] [CrossRef]
  47. Nardone, P.; Pierannunzio, D.; Ciardullo, S.; Lazzeri, G.; Cappello, N.; Spinelli, A.; Rita, S.; et al. Dietary habits among Italian adolescents and their relation to socio-demographic characteristics. Ann. dell’Istituto Super. Di Sanita 2020, 56(4), 504–513. [Google Scholar] [CrossRef] [PubMed]
  48. Neufeld, L. M.; Andrade, E. B.; Suleiman, A. B.; Barker, M.; Beal, T.; Blum, L. S.; Zou, Z.; et al. Food choice in transition: adolescent autonomy, agency, and the food environment. The Lancet 2022, 399(10320), 185–197. [Google Scholar] [CrossRef] [PubMed]
  49. Neta, A.; Steluti, J.; Ferreira, F.; Farias Junior, J. C. D.; Marchioni, D. M. L. Dietary patterns among adolescents and associated factors: longitudinal study on sedentary behaviour, physical activity, diet and adolescent health. Ciência Saúde Coletiva 2021, 26, 3839–3851. [Google Scholar] [PubMed]
  50. Nicolas, C.; Rouche, M.; Dierckens, M.; Kelly, C.; Fismen, A. S.; Nardone, P.; Chatelan, A.; et al. Sixteen-year trends in fruit consumption and related socioeconomic inequalities among adolescents in Western European countries. Eur. J. Nutr. 2023, 62(8), 3287–3296. [Google Scholar] [CrossRef] [PubMed]
  51. Norton, L.; Parkinson, J.; Harris, N.; Darcy, M.; Hart, L. Parental food communication and child eating behaviours: A systematic literature review. Health Promot. J. Aust. 2023, 34(2), 366–378. [Google Scholar] [CrossRef] [PubMed]
  52. Oduro, M. S.; Adeleye, K. K.; Agyemang-Duah, W.; Peprah, P. Substance Use and Risky Sexual Behavior Among Adolescents: A Cross-National Clustered Analysis of 35 European and North American Countries. Int. J. Sex. Health 2025, 37(2), 284–296. [Google Scholar] [CrossRef] [PubMed]
  53. Oduro, M. S.; Fasola, E.; Peprah, P.; Addo, I. Y. Interplay of physical activity, self-rated health, and life satisfaction among large-bodied adolescents in Europe and North America: analysis using clustered binary mixed effects logit modelling. BMC Public Health 2025, 25(1), 2437. [Google Scholar] [CrossRef] [PubMed]
  54. Ogundijo, D. A.; Tas, A. A.; Onarinde, B. A. Age, an important sociodemographic determinant of factors influencing consumers’ food choices and purchasing habits: an English university setting. Front. Nutr. 2022, 9, 1–10. [Google Scholar] [CrossRef] [PubMed]
  55. Panda, R.; Mayrovitz, H. N. Synergistic Actions of Natural Compounds for Enhancing Cognitive and Physical Performance: A Narrative Review. Cureus 2026, 18(1), 1–7. [Google Scholar] [CrossRef] [PubMed]
  56. Purohit, B. M.; Dawar, A.; Bansal, K.; Nilima; Malhotra, S.; Mathur, V. P.; Duggal, R. Sugar-sweetened beverage consumption and socioeconomic status: a systematic review and meta-analysis. Nutr. Health 2023, 29(3), 465–477. [Google Scholar] [CrossRef] [PubMed]
  57. Queral, J.; Wargers, A.; Llaurado, E.; Jansen, W.; Sola, R.; Tarro, L.; Mölenberg, F. J. Effectiveness of participatory research interventions on obesity and obesity-related behaviors in adolescents: A systematic review and meta-analysis. Nutr. Rev. 2025, 83(3), 904–930. [Google Scholar] [CrossRef] [PubMed]
  58. Rouche, M.; Lebacq, T.; Dzielska, A.; Kelly, C.; Nic Gabhainn, S.; Mertens, C.; Castetbon, K. Cross-National Study on the Convergent, Discriminant, and Concurrent Validity of the “Body Size Perception” Item in the Health Behaviour in School-aged Children Survey. Curr. Dev. Nutr. 2024, 8(9), 1–11. [Google Scholar] [CrossRef] [PubMed]
  59. Sălcudean, A.; Cîmpian, D. M.; Popovici, R. A.; Forna, N.; Corodan-Comiati, D. M.; Sasu, A. B.; Muntean, D. L.; et al. Dietary Habits and Their Influence on the Microbiome and Mental Health in Adolescents. Nutrients 2025, 17(9), 1–17. [Google Scholar] [CrossRef] [PubMed]
  60. Van der Horst, K.; Kremers, S.; Ferreira, I.; Singh, A.; Oenema, A.; Brug, J. Perceived parenting style and practices and the consumption of sugar-sweetened beverages by adolescents. Health Educ. Res. 2007, 22(2), 295–304. [Google Scholar] [CrossRef] [PubMed]
  61. Voráčová, J.; Sigmund, E.; Sigmundová, D.; Kalman, M. Family affluence and the eating habits of 11-to 15-year-old Czech adolescents: HBSC 2002 and 2014. Int. J. Environ. Res. Public Health 2016, 13(10), 1–11. [Google Scholar] [CrossRef] [PubMed]
  62. World Health Organization (WHO). The adolescent health indicators recommended by the Global Action for Measurement of Adolescent Health. 2024. Available online: https://www.who.int/publications/i/item/9789240092198 (accessed on 14 June 2025).
  63. Zaborskis, A.; Grincaitė, M.; Kavaliauskienė, A.; Tesler, R. Family structure and affluence in adolescent eating behaviour: a cross-national study in forty-one countries. Public Health Nutr. 2021, 24(9), 2521–2532. [Google Scholar] [CrossRef] [PubMed]
Table 1. Country/region-specific sample descriptives. 
Table 1. Country/region-specific sample descriptives. 
Country/Region 2010 2014 2018 Region Totals
Armenia 1257 1719 2805 5781
Austria 4192 2778 3494 10464
Belgium (Flemish) 3390 3394 3403 10187
Belgium (French) 2379 3479 3018 8876
Canada 10656 8520 6717 25893
Croatia 5693 4403 4403 14499
Czech Republic 3909 4402 9807 18118
Denmark 3128 3121 2530 8779
England 1321 1991 285 3597
Estonia 3225 3392 3891 10508
France 4324 4152 6887 15363
Germany 3804 4363 3421 11588
Greece 4385 3757 3491 11633
Greenland 453 421 408 1282
Hungary 4071 3147 3132 10350
Iceland 8882 7611 5265 21758
Ireland 1210 991 772 2973
Israel 3547 1963 4505 10015
Italy 3961 3259 3566 10786
Latvia 3451 4829 3955 12235
Luxembourg 3074 2259 2943 8276
Macedonia 2825 3040 3804 9669
Netherlands 3526 3192 3251 9969
Norway 3162 2553 1976 7691
Poland 3725 3818 4511 12054
Portugal 3412 4322 5203 12937
Romania 4113 2074 2671 8858
Russia 4179 3527 3654 11360
Scotland 2427 1591 1279 5297
Slovakia 3842 4481 2755 11078
Spain 4400 5648 3747 13795
Sweden 4817 5657 3073 13547
Switzerland 5768 5377 6526 17671
Ukraine 4847 3839 5058 13744
Wales 2196 1953 3407 7556
Wave Totals 133551 125023 129613 388187
Table 2. Contingency analysis on Sweet Consumption by study variables. 
Table 2. Contingency analysis on Sweet Consumption by study variables. 
Variables Sweet Consumption
Low (N=130423) Moderate (N=163687) High
(N=94077)
Total
(N=388187)
Chi-Square Test
Affluence Status
High 25431 (19.5%) 33123 (20.2%) 17406 (18.5%) 75960 (19.6%) < 0.001
Low 24438 (18.7%) 31530 (19.3%) 22927 (24.4%) 78895 (20.3%)
Medium 80554 (61.8%) 99034 (60.5%) 53744 (57.1%) 233332 (60.1%)
BMI Category
Normal weight 95900 (73.5%) 119820 (73.2%) 69116 (73.5%) 284836 (73.4%) 0.039
Obesity 7248 (5.6%) 9408 (5.7%) 5330 (5.7%) 21986 (5.7%)
Overweight 22527 (17.3%) 28657 (17.5%) 16160 (17.2%) 67344 (17.3%)
Severe thinness 1147 (0.9%) 1448 (0.9%) 822 (0.9%) 3417 (0.9%)
Thinness 3601 (2.8%) 4354 (2.7%) 2649 (2.8%) 10604 (2.7%)
Age Category
11-year-olds 46800 (35.9%) 42716 (26.1%) 25254 (26.8%) 114770 (29.6%) < 0.001
13-year-olds 42852 (32.9%) 57335 (35.0%) 33666 (35.8%) 133853 (34.5%)
15-year-olds 40771 (31.3%) 63636 (38.9%) 35157 (37.4%) 139564 (36.0%)
Parental Communication
Low 15994 (12.3%) 21870 (13.4%) 13275 (14.1%) 51139 (13.2%) < 0.001
High 114429 (87.7%) 141817 (86.6%) 80802 (85.9%) 337048 (86.8%)
Sex
Male 65771 (50.4%) 81002 (49.5%) 42836 (45.5%) 189609 (48.8%) < 0.001
Female 64652 (49.6%) 82685 (50.5%) 51241 (54.5%) 198578 (51.2%)
Physical Activity Status
Daily active 25137 (19.3%) 26641 (16.3%) 19622 (20.9%) 71400 (18.4%) < 0.001
Active 33544 (25.7%) 43458 (26.5%) 20952 (22.3%) 97954 (25.2%)
Somewhat active 42179 (32.3%) 55319 (33.8%) 28499 (30.3%) 125997 (32.5%)
Inactive 29563 (22.7%) 38269 (23.4%) 25004 (26.6%) 92836 (23.9%)
Table 3. Contingency analysis on Fruit Consumption by study variables. 
Table 3. Contingency analysis on Fruit Consumption by study variables. 
Variables Fruit Consumption
Low (N=68361) Moderate (N=168675) High (N=151151) Total (N=388187) Chi-Square Test
Affluence Status
High 10495 (15.4%) 31040 (18.4%) 34425 (22.8%) 75960 (19.6%) < 0.001
Low 17105 (25.0%) 34897 (20.7%) 26893 (17.8%) 78895 (20.3%)
Medium 40761 (59.6%) 102738 (60.9%) 89833 (59.4%) 233332 (60.1%)
BMI Category
Normal weight 50187 (73.4%) 123796 (73.4%) 110853 (73.3%) 284836 (73.4%) 0.06
Obesity 3807 (5.6%) 9569 (5.7%) 8610 (5.7%) 21986 (5.7%)
Overweight 11876 (17.4%) 29083 (17.2%) 26385 (17.5%) 67344 (17.3%)
Severe thinness 629 (0.9%) 1542 (0.9%) 1246 (0.8%) 3417 (0.9%)
Thinness 1862 (2.7%) 4685 (2.8%) 4057 (2.7%) 10604 (2.7%)
Age Category
11-year-olds 16128 (23.6%) 46360 (27.5%) 52282 (34.6%) 114770 (29.6%) < 0.001
13-year-olds 23294 (34.1%) 58805 (34.9%) 51754 (34.2%) 133853 (34.5%)
15-year-olds 28939 (42.3%) 63510 (37.7%) 47115 (31.2%) 139564 (36.0%)
Parental Communication
Low 12325 (18.0%) 21856 (13.0%) 16958 (11.2%) 51139 (13.2%) < 0.001
High 56036 (82.0%) 146819 (87.0%) 134193 (88.8%) 337048 (86.8%)
Sex
Male 36475 (53.4%) 86248 (51.1%) 66886 (44.3%) 189609 (48.8%) < 0.001
Female 31886 (46.6%) 82427 (48.9%) 84265 (55.7%) 198578 (51.2%)
Physical Activity Status
Daily active 8672 (12.7%) 25104 (14.9%) 37624 (24.9%) 71400 (18.4%) < 0.001
Active 12542 (18.3%) 43680 (25.9%) 41732 (27.6%) 97954 (25.2%)
Somewhat active 21489 (31.4%) 59723 (35.4%) 44785 (29.6%) 125997 (32.5%)
Inactive 25658 (37.5%) 40168 (23.8%) 27010 (17.9%) 92836 (23.9%)
Table 6. Model 3: Proportional odds mixed effects modelling results with only Affluence Category as predictor and Sweet Consumption as Response. 
Table 6. Model 3: Proportional odds mixed effects modelling results with only Affluence Category as predictor and Sweet Consumption as Response. 
Adolescent Sweet Consumption
Predictors Odds Ratios CI P-value
Affluence Category [Ref: High]
Low 0.94 0.92 – 0.95 <0.001
Medium 0.94 0.93 – 0.96 <0.001
Intercepts
Low | Moderate 0.42 0.35 – 0.50 <0.001
Moderate | High 2.91 2.43 – 3.48 <0.001
Random Effects
Within Region Variance 3.29
Between-Region Variance 0.3
Intra-Class Correlation 0.08
No of Clusters 35
Marginal R2 / Conditional R2 0.000 / 0.083
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.