1. Introduction
Diet and exercise are crucial for an individual’s health and well-being. Research consistently demonstrates that a balanced diet and regular physical activity significantly reduce the risk of chronic diseases, including obesity, type 2 diabetes, cardiovascular disorders, and certain cancers [
1]. In Portugal, childhood obesity has become a pressing public health issue, with recent studies revealing alarmingly high rates of overweight and obesity among younger populations [
2]. The rising prevalence of sedentary lifestyles among Portuguese youth exacerbates these concerns, reflecting a broader European trend towards physical inactivity and unhealthy eating behaviors.
Childhood obesity rates have increased dramatically across Europe, with approximately 340 million children and adolescents now classified as overweight or obese [
1]. The World Obesity Federation projects that by 2025, around 206 million children aged 5 to 19 will be living with obesity [
2,
3]. In Portugal, these trends present an urgent challenge, as high levels of obesity in childhood often require more complex and intensive medical interventions [
4]. Notably, Portugal’s historical adherence to the Mediterranean diet—a dietary pattern traditionally associated with positive health outcomes—has declined, particularly among younger generations [
5]. This shift raises concerns regarding the dietary quality of Portuguese youth and its potential impact on long-term health.
The decline in physical activity is a growing concern in many countries, especially in urban areas where spaces for recreational and sports activities are limited. Most adolescents worldwide, including Portugal, do not meet the recommended levels of physical activity, predisposing them to health issues such as obesity, diabetes, hypertension, and cardiovascular diseases [
1]. This trend is further complicated by the close relationship between physical and mental health in adolescents, affecting their overall quality of life and development [
4].
The Mediterranean diet, characterized by high consumption of fruits, vegetables, whole grains, and lean proteins, has been shown to promote healthy weight management, improve physical performance, and reduce the risk of chronic diseases [
1]. In particular, adherence to the Mediterranean diet is strongly inversely correlated with the incidence of non-communicable diseases such as cancer, metabolic syndrome, and cardiovascular diseases while also being associated with lower mortality rates [
5,
6,
7]. Among adolescents, higher adherence to the Mediterranean diet has been linked to better anti-inflammatory capacity, physical fitness, and improved quality of life-related to health [
5,
8]. Despite these benefits, recent systematic reviews indicate a troubling decline in adherence to the Mediterranean diet across Mediterranean countries, including Portugal, especially among younger populations [
5]. This decline underscores the need for targeted public health initiatives to revitalize and promote this dietary pattern within the Portuguese context.
Given the escalating trends of physical inactivity and unhealthy eating behaviors among Portuguese youth, understanding their dietary habits and physical activity levels is crucial. Such habits are not only linked to physical and mental health issues but also have the potential to impact academic performance and overall quality of life negatively [
1]. This study aims to evaluate the extent to which middle school students adhere to the Mediterranean diet and assess their physical activity habits and overall weight perceptions. By examining these factors, the study aims to identify areas where interventions can be implemented to promote healthier lifestyles among adolescents, ultimately contributing to improved public health outcomes in Portugal. Additionally, the findings are expected to inform policy and support the development of strategies that encourage a return to healthier, more traditional dietary practices and more active lifestyles among Portuguese youth.
2. Materials and Methods
2.1. Study Design and Sample Size
This study employed a quantitative and descriptive research design, focusing on middle school students in a small city in northern Portugal. The sample comprised students aged 12 to 18, enrolled in grades seven through nine. Despite convenience sampling, which may introduce biases in representativeness, this method was chosen due to logistical limitations (e.g., accessibility to the student population in a single school). Furthermore, this choice was made to ensure the feasibility of data collection within a limited timeframe while still aiming to capture a broad demographic of the school’s population.
2.2. Recruitment and Participants
Participants were recruited from the only school in a small town within the Braga district. The inclusion criteria specified that participants must be enrolled in grades 7 through 9 and aged between 12 and 18 years and have provided informed consent from both themselves and their legal guardians. Students who did not provide informed consent or could not complete the questionnaires were excluded from the study. To ensure voluntary participation, the objectives and procedures of the study were clearly explained to the participants and their guardians. Despite using a convenience sample, efforts were made to maximize the diversity of participants within the school’s population by including students from various academic tracks and socioeconomic backgrounds.
2.3. Data Collection and Measurement
Data were collected using a non-probabilistic convenience sampling method. The convenience sampling approach was chosen due to practical constraints, such as the study’s limited scope to one school and the need for easy access to participants. However, this method raises concerns about the generalizability of the findings. To mitigate these concerns, the sample was drawn from all eligible students within the school, ensuring a broad cross-section of the student population.
Two validated questionnaires were administered during class periods, with students given 45 min to complete them. Data were collected using a Google Form link sent directly to the students’ email addresses. The teacher introduced the questionnaire, explained its objectives, and clarified the students’ questions during the completion process. This ensured the students fully understood the questionnaire while maintaining its self-administered format. The first part of the questionnaire gathered sociodemographic information (age, gender, living arrangements, parents’ educational levels, and employment status). In contrast, the second part incorporated the Youth Risk Behavior Survey (YRBS) and the Mediterranean Diet Quality Index (KIDMED).
Youth Risk Behavior Survey (YRBS): The YRBS was employed to assess specific risk behaviors among adolescents. This instrument was validated for the Portuguese population and adapted to ensure cultural and linguistic appropriateness for Portuguese students [
9,
10]. This study evaluated three behavior categories relevant to diet and physical activity—body weight management, physical activity, and dietary behaviors—alongside two open-ended questions regarding the respondents’ weight and height. The selected domains were aligned with the study’s objectives, focusing on health behaviors that impact diet and physical activity in Portuguese adolescents.
Mediterranean Diet Quality Index (KIDMED): The KIDMED index is a widely used tool for assessing adherence to the Mediterranean diet among children and adolescents [
11]. It is based on a 16-item questionnaire, with responses scored as either +1 (for adherence-supporting behaviors) or −1 (for behaviors undermining Mediterranean diet principles). Total scores classify diet adherence into three categories: optimal (≥8), needs improvement (4–7), and poor (≤3). The KIDMED was previously translated and validated for Portuguese adolescents and widely used, ensuring its relevance to this population [
7,
8,
12]. The questionnaire was slightly adapted to reflect food consumption patterns specific to Portuguese youth, with changes made based on expert input from local dietitians familiar with adolescent nutrition in Portugal.
2.4. Statistical Methods
Data analysis was conducted using Excel 2023 and IBM SPSS statistical software version 29. The significance level was set at a
p-value less than or equal to 0.05. The Kolmogorov–Smirnov test was employed to assess the normality of data. Descriptive statistics summarize sociodemographic characteristics and dietary and physical activity behaviors. Nominal variables were described in terms of absolute values and percentages (reported in terms of column) [
13].
In contrast, continuous variables were characterized using measures of central tendency and dispersion (mean and standard deviation). The chi-square test of independence was utilized to analyze differences in probabilities between two groups, specifically differences in percentages. However, when any expected counts in the contingency tables fell below 5, indicating an inadequate sample size for the chi-square test of independence, Fisher’s exact test (2 × 2 contingency Table) was employed.
The necessary n for a power of 0.80 and to detect a medium degree of association in the population at α = 0.05 and one degree of freedom is n = 87; six degrees of freedom requires n = 151. The present study includes 232 participants, which is adequate for the analysis. For the chi-square test, the effect size index w is calculated by dividing the chi-square value by the number of scores and taking the square root, and it is considered small if w = 0.10, medium if w = 0.30, and large if w = 0.50.
An independent-sample
t-test was used to compare KIDMED scores between genders. Similarly, the necessary
n for the power of 0.80 to detect a medium difference between two independent sample means (d = 0.5) at α = 0.05 is
n = 64 in each group, and 232 participants are adequate for the analysis [
13]. Furthermore, the data had no outliers, as assessed by inspection of a boxplot. KIDMED scores for each level of gender were normally distributed, and variances were homogeneous, as assessed by Levene’s test for equality of variances.
Ordinal logistic regression analyzed the relationship between KIDMED scores and daily fruit consumption, and binary logistic regression assessed the likelihood of exercising to lose or avoid gaining weight based on KIDMED scores.
2.5. Ethics
Ethical approval was obtained from the University of Fernando Pessoa Ethics Committee (Approval Number: ESS/PI—205/21). All participants and their legal guardians were provided detailed information about the study, ensuring informed consent. Participants were assured of anonymity, and confidentiality was maintained throughout the study. Any questions or concerns raised by participants were addressed to ensure their complete understanding of the study’s procedures and the voluntary nature of their participation.
All subjects gave their informed consent for inclusion before participating in the study.
3. Results
Of the 232 enrolled students, 119 were females (51.3%) and 113 were males (48.7%).
Table 1 describes the distribution of students according to their age and school year. In total, 76.7% of students live with their parents. In comparison, 16.4% live only with their mother and 6.9% live in other housing solutions; 74.6% of the participants reported having married parents, while 19% said that their parents were divorced, and 6.4% reported other marital status.
Table 1 represents the distribution of students based on their sociodemographic and school-year characteristics.
There was no statistically significant association between gender and how the respondents described their weight (p = 0.112). There was no statistically significant association between gender and weight goal (p = 0.642).
Supplementary Table S1 describes the results from the Portuguese version of the risk behavior questionnaire (body weight, eating habits and physical activity domain).
Table 2 describes the proportion of participants who exercised, aiming to lose or avoid gaining weight, divided by gender.
A statistically significant association exists between having practiced exercise to lose weight or avoid gaining weight and gender (χ
2(1) = 6.138,
p = 0.019,
w = 0.23 (
Table 2)). Most male respondents report exercising to lose or avoid gaining weight, while the difference is not so evident for the female counterparts.
There was no association between gender and decreasing calorie intake to lose weight or avoid gaining weight (p = 0.284), the use of fasting techniques to lose weight (p = 0.317), or the use of unprescribed medication to aid weight control (p = 0.236). Furthermore, using laxatives or throw-up techniques to lose weight was not associated with gender (p = 0.744).
Concerning eating habits, there was no association between gender and the description of the eating habits (p = 0.903) or the report of fruit intake (p = 0.084). There was also no association between gender and the report of vegetable intake (p = 0.303), the report of eating and drinking before school (p = 0.119), or the habit of drinking milk (p = 0.643).
The physical activities that produced sweat and faster breathing, divided by gender, are described in
Table 3.
There was a statistically significant association between gender and days spent participating in physical activity producing sweat (Fisher’s exact test (7) = 22.34,
p = 0.002,
w = 0.43 (
Table 3)). Most female participants practised 2 to 3 days a week, while male participants reported significantly more days of practice (5, 7 and 2) days a week.
Gender and days spent participating in physical activity that did not produce sweat did not demonstrate a statistically significant association (p = 0.575).
Table 4 describes the number of strength exercises the respondents engaged in over the last seven days, divided by gender.
There was a statistically significant association between gender and days spent participating in strengthening exercises (Fisher’s exact test (7) = 28.73,
p < 0.001,
w = 0.49 (
Table 4)). Most participants tend not to perform strengthening exercises and those who do tend to choose to do so 1 to 2 days a week.
The following variables did not present a significant association with gender: time spent watching TV (p = 0.317); days spent in physical education classes (p = 0.224); and the number of minutes they spent exercising or playing a sport (p = 0.099).
Table 5 describes the number of sports teams the respondents attended, divided by gender.
There was a statistically significant association between gender and the number of sports teams attended (Fisher’s exact test = 8.991, p = 0.027, w = 0.28). Most female participants did not participate in team sports, and male participants were almost equally distributed between not participating in and participating in one team sport.
Gender and need to see a healthcare provider did not demonstrate a statistically significant association (p = 0.417).
Supplementary Table S2 provides a comprehensive overview of the results obtained from the KIDMED questionnaire, which assesses the study participants’ adherence to the Mediterranean diet. There was no association between eating one piece of fruit daily and gender (
p = 0.503), eating two or more pieces of fruit daily (
p = 0.515), or having regular fish meals (
p = 0.293). Furthermore, no association was found between gender and the habit of consuming fast food regularly (
p = 0.885),the inclusion of beans in meals (
p = 0.376), the inclusion of rice or pasta regularly (
p = 0.380), the use of cereals and derivatives at breakfast (
p = 0.579), or the inclusion of nuts in the diet (
p = 0.128). There was no association between gender and the use of olive oil at home (
p = 0.716), the habit of eating sweets (
p = 0.866), the use of dairy and derivatives daily (
p = 0.163), the use of cakes for breakfast (
p = 0.847), or the use of dairy and derivatives at breakfast (
p = 0.768).
The proportion of participants who reported eating vegetables regularly, divided by gender, is described in
Table 6.
A chi-square homogeneity test was run, and there is a statistically significant association between the two variables (χ2(1) = 6.212, p = 0.016, w = 0.23).
The proportion of participants who fast-forwarded or skipped breakfast, divided by gender, is described in
Table 7.
A chi-square homogeneity test was run, and there is a statistically significant association between the two variables (χ2(1) = 4.670, p = 0.033, w = 20).
Generally, there was moderate adhesion to the Mediterranean dietary pattern, as shown in
Table 8.
An independent-sample t-test was run to determine whether scores differed between males and females, and no differences (p = 0.639) were found.
The effects of the KIDMED scoring on the likelihood that participants had exercised in the last 30 days to lose or avoid gaining weight were statistically significant (χ2(4) = 5.08, p = 0.024, w = 0.21). The model explained 3.0% (Nagelkerke R2) of the variance in exercising to lose weight and correctly classified 65.5% of the cases. The variable adhesion to the Mediterranean diet (Kidmed score) was statistically significant, with higher adherence to the Mediterranean diet associated with an increased likelihood of tending to exercise to lose weight (p = 0.026). Those who did not exercise, aiming to lose or avoid gaining weight, were less likely to adhere to a Mediterranean diet. They presented 0.76 times higher odds of not exhibiting an adequate diet.
Cumulative odds ordinal logistic regression with proportional odds was run to determine the effect of the KIDMED scoring on the self-reported daily dose of fruit consumption. There were proportional odds, as assessed by a full likelihood ratio test comparing the fitted model to a model with varying location parameters (χ2(3) = 6.47, p = 0.091). The deviance goodness-of-fit test indicated that the model fit was good (χ2(47) = 45.820 p = 0.521), with zero frequencies in 30.8% of cells. The final model statistically significantly predicted the dependent variable over and above the intercept-only model (χ2(1) = 51.85, p < 0.001). An increase in the KIDMED scoring (towards higher adherence to the Mediterranean diet) was associated with an increase in the odds of a higher daily dose of fruit consumption, with an odds ratio of 1.48 (95% CI [1.33 to 1.66], χ2(1) = 4654, p < 0.001).
5. Conclusions
This study provides a comprehensive overview of adolescents’ dietary habits, physical activity levels, and weight perceptions in the third cycle of education in Northern Portugal. Our findings revealed moderate adherence to the Mediterranean diet and significant gender differences in exercise intensity. These results underscore the need for holistic, gender-sensitive approaches addressing this age group’s dietary and physical activity behaviors.
In light of these findings, there is a clear and pressing need for targeted, evidence-based interventions at the family, school, and community levels. School-based health promotion programs, such as the “National Program for the Promotion of Healthy Eating” (PNPAS) and “National Program for the Promotion of Physical Activity” (PNPAF) in Portugal, in Portugal, can provide a strong foundation for enhancing student health and well-being. Studies can provide a strong foundation for enhancing student health and well-being.
Policy Implications and Recommendations: Our findings should inform the development of future health promotion programs that are culturally relevant and tailored to adolescents. Given the critical role that schools play in shaping student behaviours, school-based interventions should be expanded to include the following:
Nutritional Education: Integrating curriculum elements that promote the Mediterranean diet, emphasizing the consumption of whole grains, legumes, nuts, and olive oil. These lessons should be reinforced through school meal programs, ensuring alignment between education and practice.
Physical Activity Promotion: Schools should implement gender-sensitive physical activity programs that cater to the preferences and needs of both boys and girls. Introducing a wider variety of activities, including those that appeal to girls (such as dance or group fitness), could help close the gender gap in physical activity.
Parental Involvement: Family-centered interventions should be encouraged, as research consistently highlights the role of parental dietary habits and physical activity in shaping adolescent behaviour. Programs that involve parents in promoting healthy eating and physical activity at home are likely to yield more sustainable outcomes.
At the policy level, collaboration between schools, healthcare professionals, and policymakers is essential to implement and sustain these initiatives. In particular, local health authorities and municipalities should actively support school-based programs by providing resources and funding and creating environments conducive to physical activity (e.g., safe spaces for exercise). Community engagement through partnerships with sports clubs and local businesses could also enhance these efforts.
Future Research Directions: To deepen the understanding of adolescent health behaviors, future research should employ longitudinal or experimental designs that explore the causal relationships between dietary habits, physical activity, and health outcomes. Additionally, studies should examine a broader range of determinants, including socioeconomic status, cultural influences, and peer dynamics, which may impact adolescent behaviors.
Importantly, future research should also focus on developing and assessing the efficacy of targeted interventions. These interventions should be evaluated for their ability to improve dietary and physical activity habits, mental health, and body image, which are closely linked to overall well-being. Integrating these findings into national health strategies could further contribute to Portuguese adolescents’ long-term health and development.
By taking proactive and coordinated action at the family, school, community, and policy levels, we can foster healthier behaviors among adolescents, ultimately contributing to improved health outcomes and a brighter future for all.