University students represent a unique population at a vulnerable stage, characterised by significant new lifestyle habits, including shifts in diet, physical activity, screen time, eating out and more(Reference da-Silva-Domingues, Palomino-Moral and Gutiérrez-Sánchez1). The transition to university life often brings newfound independence, academic pressures, financial constraints and various social influences, all of which can significantly shape their food choices(Reference Pearcey and Zhan2,Reference Powell, Durham and Lawler3) .
Food choices are influenced by a complex, varied and interconnected set of factors that guide individuals’ decisions regarding what and how they eat(Reference Fernqvist, Spendrup and Tellström4). For many university students, food choices are driven by a combination of factors such as convenience (ease of preparation and consumption), price, sensory appeal (SA) (e.g. taste, appearance, smell and texture), ethical concerns (animal welfare, environmental impact and cultural values), mood and/or weight control(Reference Powell, Lawler and Durham5,Reference Lambert, Chivers and Farringdon6) . However, when conflicts arise between these motives (e.g. price, convenience or taste v. health) or when one motive prevails over another (e.g. price v. convenience), students are forced to prioritise them, highlighting their importance. Consequently, university students may opt for poor dietary choices, such as fast food or energy-dense and nutrient-poor foods(Reference Darmon and Drewnowski7–Reference Bouzas, Pastor and García9), increasing the risk of overweight and obesity (issues that are becoming increasingly concerning among university students) which are linked to cardiometabolic disorders like dyslipidaemia, elevated blood pressure, insulin resistance and systemic inflammation(Reference Darmon and Drewnowski7), further exacerbated by stress and sleep deprivation, commonly impaired at the university phase(Reference Sominsky and Spencer10).
Furthermore, an unhealthy cardiometabolic profile during university years can shorten life expectancy and increase the risk of future chronic cardiometabolic diseases such as type 2 diabetes, CVD and metabolic dysfunction-associated fatty liver disease, among others(Reference Yang, Liu and Zhang11). While the global prevalence of metabolic syndrome among university students varies significantly, reaching up to 19·2 %(Reference Campo-Arias, González-Guerrero and Peñaloza-Vázquez12), in Mexico, reported rates range from 7·0 % to 25·7 %(Reference Murguía-Romero, Jiménez-Flores and Sigrist-Flores13–Reference Vizmanos, Betancourt-Nuñez and Márquez-Sandoval16), highlighting an alarming public health concern in this population. This issue is further compounded by the fact that cardiometabolic disorders often develop slowly, gradually and silently in young adults, frequently remaining asymptomatic and therefore unnoticed. As these abnormalities accumulate over time, they substantially increase morbi-mortality in later life, as previously mentioned(Reference Higgins, Smith and Williams17). For this reason, early screening to detect latent cardiometabolic alterations in university students is critical to enable timely interventions and prevent long-term adverse health consequences(Reference Drozdz, Alvarez-Pitti and Wójcik18).
While genetic predisposition plays a role in these conditions(Reference Mahmoud, Kimonis and Butler19), modifiable lifestyle factors, particularly food choices, are central to both prevention and management(Reference Masood and Moorthy20). Despite increasing recognition of the relevance of assessing food choices, research on their impact on cardiometabolic health remains limited. Most existing studies have focused primarily on the general population and dietary intake (mainly the type and quantity of food consumed) and its association with metabolic syndrome(Reference Cruz-Rodríguez, González-Vázquez and Reyes-Castillo14,Reference Jung, Dan and Pang21–Reference Cano-Ibáñez, Gea and Ruiz-Canela27) , while largely overlooking the underlying motivations that guide these choices. This gap is especially relevant in Mexico and other Latin American countries, where the country’s unique cultural and socio-economic factors (such as the strong attachment to traditional foods) interact with the growing availability of ultra-processed products, shaping food preferences and behaviours in complex ways(Reference Gómez Delgado and Velázquez Rodríguez28).
Moreover, an important yet often overlooked factor influencing college students’ food choices is the marked difference in underlying motivations between men and women(Reference Boek, Bianco-Simeral and Chan29,Reference Salas-García, Bernal-Orozco and Díaz-López30) . These distinctions are particularly relevant given the well-documented sex disparities in the development and manifestation of cardiometabolic diseases(Reference Zhernakova, Sinha and Andreu-Sánchez31). Understanding how sex-specific food choice motives (FCM) relate to cardiometabolic risk is therefore crucial to designing more effective, tailored interventions. The study of motivations behind food choices in this vulnerable group is essential for (1) addressing these health challenges and (2) developing targeted interventions that promote healthier eating behaviours to prevent long-term cardiometabolic disorders among university students in both, males and females, looking for the existence of any sex-specific patterns.
In this context, the aim of this study was to examine the association between the importance towards diverse FCM and cardiometabolic risk among Mexican university students, with particular attention to potential sex-specific patterns.
Materials and methods
Study design
This cross-sectional study is a secondary analysis derived from a larger project which analysed food insecurity risk factors and their consequences on lifestyle and cardiovascular health among workers and students at a Mexican public university. Current analysis includes only university students’ data collected from April 2022 to November 2023(Reference Nava-Amante, Betancourt-Núñez and Díaz-López32).
Ethical approval
The study protocol adhered to the guidelines laid down in the Declaration of Helsinki and was approved by the Research, Research Ethics and Biosafety Committees from the University Center for Health Sciences (CUCS, by its acronym in Spanish) of the University of Guadalajara (code of approval CI-02323). All participants provided written informed consent.
Participants
Undergraduate and postgraduate students, primarily from health sciences fields (e.g. nutrition, medicine, psychology, nursing and dentistry) at CUCS, were recruited using non-probabilistic convenience sampling. Participants were invited to participate in the study through face-to-face invitation in classrooms. Eligibility criteria included being ≥18 years old, not taking medications that could affect cardiometabolic markers (e.g. corticosteroids, androgens, synthetic vitamin A derivates or antiretroviral drugs), no cancer diagnosis or treatment within the previous 6 months and not being pregnant or breast-feeding(Reference Nava-Amante, Betancourt-Núñez and Díaz-López32).
For the current study, sample size was calculated using the formula for estimating proportions due to the absence of previous studies examining the association between FCM and cardiometabolic variables. Based on a reported 0·9 % prevalence of elevated fasting glucose among Mexican university students(Reference Betancourt-Núñez, Márquez-Sandoval and Babio33), a minimum sample size of 383 participants was required, using 95 % CI and a 5 % margin of error. Participants with incomplete data regarding the study variables were excluded from this analysis.
Data collection
Exposure variable: food choice motives’ importance assessment
FCM’s importance was evaluated online using the Updated-Food Choice Questionnaire (U-FCQ)(Reference Salas-García, Bernal-Orozco and Díaz-López34), a seventy-five-item tool culturally adapted and validated for the Mexican population based on the original FCQ(Reference Steptoe, Pollard and Wardle35). Briefly, each item assesses motives underlying food choices and is preceded by the statement: ‘It is important to me that the food I eat on a typical day…’. Participants rated the importance of each item on a four-point Likert scale ranging from 1 (‘not at all important’) to ‘4 (‘very important’). The seventy-five items are grouped into eight FCM’s dimensions: (1) ‘Health and Natural Content’ (twenty-seven items; e.g. ‘Is nutritious’, ‘Keeps me healthy’, ‘Is low in sugar’); (2) ‘Environmental and Wildlife Awareness’ (E&WA) (twelve items; e.g. ‘Is produced respecting animal rights’, ‘Does not contribute to CO2 emissions’, ‘Has not been transported long distances’); (3) ‘Convenience’ (six items; e.g. ‘Can be cooked quickly’, ‘Can be consumed almost immediately’, ‘Is practical’); (4) ‘Price’ (five items; e.g. ‘Is cheap’, ‘Is not expensive’, ‘Is on sale’); (5) ‘SA’ (six items; e.g., ‘Smells nice’, ‘Tastes good’, ‘Is pleasant to me’); (6) ‘Mood’ (five items; e.g. ‘Cheers me up’, ‘Helps me reduce stress’, ‘Helps me get through life’); (7) ‘Image Management’ (IM) (eight items; e.g. ‘Is considered good by other people’, ‘Reflects a positive image of me’, ‘Is like the food I ate when I was a child’) and (8) ‘Food Identity’ (FI) (six items; e.g. ‘Are local and regional products’, ‘Is produced in Mexico’, ‘Is part of the traditional Mexican diet’). The complete questionnaire, including both in Spanish and English versions, is available as supplementary material in a previous publication(Reference Salas-García, Bernal-Orozco and Díaz-López34).
The absolute importance of each FCM was calculated by averaging the responses for all items within each dimension, yielding scores between 1 and 4, with higher scores reflecting greater importance. Next, the relative importance was calculated by dividing the absolute average score of each dimension by the average score of all seventy-five items on the questionnaire. Dimensions with scores ≥1 were considered more important than the overall average(Reference Konttinen, Halmesvaara and Fogelholm36,Reference Konttinen, Sarlio-Lähteenkorva and Silventoinen37) . In the current study, each FCM’s relative importance was used as main exposure variables.
Assessment of covariates
Through the same online survey, participants also provided sociodemographic information (sex, age and socio-economic status, living arrangement (living alone, with parents or others)), alongside information of the primary food preparer (self, parents or other), smoking habit (yes/no), depression/anxiety/stress levels, emotional eating behaviours and food consumption quality. Physical activity was also assessed through an interview, after responding the online surveys.
Socio-economic status
Socio-economic status was determined using the validated AMAI (Asociación Mexicana de Agencias de Inteligencia de Mercado y Opinión) questionnaire(38). This six-item tool allows the categorisation of participants into one of seven socio-economic levels, ranging from A/B (highest) to E (lowest). For current statistical analysis, these categories were dichotomised into ‘high’ (A/B and C+) and ‘middle-low’ (C, C–, D+, D– and E).
Depression, anxiety and stress levels
Depression, anxiety and stress levels were assessed using the twenty-one-item Depression, Anxiety, and Stress Scale (DASS-21), which has been validated for use in Hispanic populations (Cronbach’s α = 0·96)(Reference Daza, Novy and Stanley39). Each seven-item subscale utilised four Likert-type scale, ranging from 0 (‘did not apply to me at all’) to 3 (applied to me very much, or most of the time’). Based on the DASS-21 scores, participants were classified as normal, or with mild, moderate, severe or extremely severe symptoms according to the following cut-offs: depression, (normal: 0–4, mild: 5–6, moderate: 7–10, severe: 11–13, extremely severe: ≥14); anxiety (normal: 0–3, mild: 4–5, moderate: 6–7, severe: 8–9, extremely severe: ≥10) and stress (normal: 0–7, mild: 8–9, moderate: 10–12, severe: 13–16, extremely severe: ≥17)(Reference Román, Santibáñez and Vinet40).
For the statistical analyses, each subscale was dichotomised into normal or presence of symptoms (mild to extremely severe). Additionally, a composite covariable incorporating the three domains (stress/anxiety/depression) was created and categorised as: (1) no symptoms in any domains; (2) symptoms in one domain only or (3) symptoms in two or more domains.
Emotional eating
Emotional eating was assessed using the ten-item Emotional Eater Questionnaire (EEQ) (Cronbach’s α: 0·77), with four-point Likert-type scale ranging from 0 (‘never’) to 3 (‘always’). Summed scores categorised participants as non-emotional eater (0–5 points), low emotional eater (6–10 points), emotional eater (11–20 points) and very emotional eater (21–30 points)(Reference Garaulet, Canteras and Morales41). For statistical effects, participants were classified as non-emotional eaters or emotional eaters.
Food consumption quality
Food consumption quality was measured using the second version of the Mini-Survey to Evaluate Food Intake Quality (by its acronym in Spanish: Mini-ECCA v.2)(Reference Bernal-Orozco, Salmeron-Curiel and Prado-Arriaga42). This fourteen-item survey evaluates the consumption of natural water, vegetables, legumes, sweetened beverages, processed foods and the types of cereals and meats consumed. Photographs are included in the survey to assist participants in estimating portion sizes. Response options (which could be between 1 and 4) were computed into three equations to categorise participants as having ‘healthy food intake’, ‘habits in need of improvement’ or ‘unhealthy food intake’.
Physical activity assessment
After completing the online survey, physical activity over the last 7 d was assessed through an in-person interview, at the CUCS’ Nutritional Status Assessment Laboratory. Participants were interviewed by a member of the research team, who applied the seven-item short version of the International Physical Activity Questionnaire (IPAQ)(Reference Mantilla-Toloza and Gómez-Conesa43). This tool provides information on the time spent walking, moderate- and vigorous-intensity activity, as well as sedentary activities. Physical activity was expressed in metabolic equivalent of task minutes per week (MET-min/week).
Outcome variables: measurement of cardiometabolic risk factors
During the in-person visit, participants underwent blood collection following a 12-h fast, an anthropometric evaluation and blood pressure measurement.
Biochemical analysis
Peripheral venous blood samples (average 12 mL) were collected into tubes, which were initially stored in a container with cooling gel at 4 °C. Tubes were subsequently transported to the laboratory and centrifuged (3500 rpm, 4 °C, 15 min) for serum extraction. All samples were stored at –80 °C until analysis and thawed and analysed simultaneously to minimise inter-batch variation. Fasting serum glucose (FSG), total cholesterol and TAG were measured using standard enzymatic automated methods(Reference Menezes, Neves and De Lima44), using the ERBA XL 180® automatic clinical wet-chemistry analyser (ERBA Diagnostics Mannheim)(45), and HDL-cholesterol via enzymatic colorimetric measures using the dry-chemistry VITROS 350® semi-automated bioanalyser (VITROS, Ortho Clinical Diagnostics)(46). LDL-cholesterol was calculated using the Friedewald equation: LDL-cholesterol = total cholesterol – HDL-cholesterol – (TAG/5), expressed in mg/dL(Reference Friedewald, Levy and Fredrickson47).
Anthropometric measurements
Trained researchers collected anthropometric measurements following the International Society for the Advancement of Kinanthropometry protocol(Reference Esparza-Ros, Vaquero-Cristóbal and Marfell-Jones48). Height (cm) was measured using a Seca® 213 stadiometer (Seca), weight (kg) was measured using a TANITA BC-568® weighing scale (TANITA) and waist circumference (WC) (cm) was measured using a Lufkin® tape measure (Lufkin). BMI was calculated by dividing weight (kg) by height (metres squared) (kg/m2). For descriptive purposes and according to the WHO(49), participants were categorised as having no excess body weight (BMI < 25 kg/m2) or excess body weight (BMI ≥ 25·0 kg/m2).
Blood pressure
Systolic blood pressure and diastolic blood pressure were measured in both arms using an automatic digital monitor (Omron model: HEM-721™, accuracy: ± 3 mmHg) after a 10-min rest. Three measurements were taken at 1-min intervals, and the average of the last two readings were used to determine systolic blood pressure and diastolic blood pressure. If the difference between measurements from the two arms exceeded 10 mmHg, the higher reading was used(Reference Muntner, Shimbo and Carey50).
Cardiometabolic risk score construction
A continuous clustered cardiometabolic risk score (CMRS) was constructed to capture both metabolic and haemodynamic alterations aligning with diagnostic criteria for metabolic syndrome (Reference DeBoer, Filipp and Gurka51,Reference Fernández-Aparicio, Perona and Schmidt-RioValle52) . This CMRS was calculated by summing sex- and age-specific standardised z-scores (z = [value – mean]/sd of the whole population)(Reference DeBoer, Filipp and Gurka51,Reference Fernández-Aparicio, Perona and Schmidt-RioValle52) for five variables including WC, FSG, TAG/FSG index(Reference Lopez-Jaramillo, Gomez-Arbelaez and Martinez-Bello53), TAG:HDL-cholesterol ratio(Reference Kundu, Biswas and Das55) and mean blood pressure (average of systolic blood pressure and diastolic blood pressure)(Reference Kundu, Biswas and Das55). A higher CMRS entails greater risk(Reference DeBoer, Filipp and Gurka51). In this study, individual cardiometabolic factors were the primary outcomes, while the CMRS served as a secondary outcome.
Statistical analysis
All statistical analyses were performed using STATA® software version 15·0 (Stata Corp). Descriptive statistics are expressed as mean (standard deviation) or median (interquartile range) for continuous variables, as appropriate, and frequency (%) for categorical variables. Normality was assessed using the Shapiro–Wilk test and visual inspection of Q-Q plots. Comparisons according to sex were performed using the Student’s t-test or the Mann–Whitney U test for continuous variables and the χ 2 test for categorical variables. Given that sex is a well-established determinant of the cardiometabolic profile(Reference Zhernakova, Sinha and Andreu-Sánchez31) and may also influence FCM(Reference Salas-García, Bernal-Orozco and Díaz-López30,Reference Konttinen, Halmesvaara and Fogelholm36) , sex was considered a potential effect modifier. Therefore, descriptive and associations analyses were additionally conducted separately in men and women.
Linear regressions models were used to assess the associations between each FCM dimension and cardiometabolic risk factor (BMI, WC, FSG, TAG, TAG/FSG index, total cholesterol, LDL-cholesterol, HDL-cholesterol, TAG:HDL-cholesterol ratio, systolic blood pressure and diastolic blood pressure), as well as continuous CMRS, in the overall sample and stratified by sex. Each FCM’s importance dimension was analysed separately as the exposure variable of interest. Crude and multivariable-adjusted models were fitted. Multivariable models were adjusted for age (years), sex (except in sex-stratified analyses), BMI (kg/m2, except for BMI and WC outcomes), socio-economic status (middle-low, high), physical activity (MET-min/week), smoking habit (yes/no), stress/anxiety/depression (none, one domain or ≥2 domains) and emotional eating (emotional v. non-emotional eaters). Covariates were selected based on previous literature and a causal directed acyclic graph approach(Reference Textor, van der Zander and Gilthorpe56) (Supplementary Fig. 1). Although food consumption quality was conceptualised as a mediator in the hypothesised causal pathway, a formal mediation analysis was beyond the scope of the present study. Therefore, food consumption quality was examined across FCM dimensions for descriptive purposes only. Results are presented as β-coefficient with and 95 % CI. Statistical significance was set at P < 0·05.
Results
A total of 395 university students (71·4 % female) were enrolled in this study. The median age and BMI of the participants were 20·0 (3·0) years old and 22·8 (5·2) kg/m2, respectively. Approximately, one-third of the sample was classified as overweight/obese (BMI ≥ 25 kg/m2), and the majority lived with their parents. Most participants had a middle-low socio-economic status and were non-smokers (both 74·9 %, Table 1).
Descriptive characteristics of the participants according to sex
Table 1. Long description
A table comparing characteristics of participants by sex. The table has 39 rows and 12 columns. Column headers include Characteristics, Total, Women, and Men. Row labels include Age, Socio-economic status, Living arrangement, Primary food preparer, Smoking habit, Depression category, Anxiety category, Stress category, Food consumption quality frequency, Anthropometric variables, Biochemical and metabolic risk factors, Food choice motives’ relative importance. Each row provides median, IQR, mean, and standard deviation values for each category. Notable trends include differences in smoking habits, stress categories, and food choice motives between women and men.
IQR, interquartile range; PA, physical activity; MET-min/week, metabolic equivalent of task-minutes per week; WC, waist circumference; FSG, fasting serum glucose; TC, total cholesterol; SBP, systolic blood pressure; DBP, diastolic blood pressure; CMRS, cardiometabolic risk score.
Data are presented as mean (sd) or median (IQR) for quantitative variables, as appropriate, and frequency (%) for categorical variables.
* P values for between-groups differences were calculated using Student’s t-test, Mann–Whitney U test or χ 2 test, as appropriate. Bold indicates statistical significance.
† The category ‘others’ in living arrangement variable includes partner, children, friends and roommates.
‡ The category ‘others’ in the primary food preparer variable includes partner or spouse, people not included in the family and purchased food.
# n 340 (women n 249 and men n 91).
§ Presented in descendending order of importance.
Men exhibited significantly higher body weight, WC, BMI, FSG, SPB and CMRS values compared with women, while total cholesterol and HDL-cholesterol were lower (all P < 0·05, Table 1).
The most and the least important factors involved in food choices among this sample were ‘SA’ and ‘IM’, respectively. Men placed significantly greater relative importance on ‘IM’ and less relative importance on ‘E&WA’ than women (all P < 0·05). No significant differences were observed across sex for the remaining FCM.
‘Price’ had a significant greater relative importance in participants with emotional eating, high depression/anxiety/stress levels, unhealthy food consumption habits and low physical activity, both in total population and in women; in men, this motive had a greater importance in participants with emotional eating, anxiety symptoms and unhealthy food consumption habits. ‘Convenience’ had also greater relative importance in women in almost all variables (except for anxiety and physical activity) (Supplementary Table 1).
Linear regression analyses revealed several significant associations between FCM’s relative importance and cardiometabolic risk factors. Sex-stratified analyses showed notable differences, with women showing more apparent association patterns (Supplementary Tables 2 and 3). In the multivariable-adjusted model (Figure 1; Supplementary Tables 4 and 5 for anthropometric/clinical and biochemical cardiometabolic risk factors), total sample’s relative importance on both ‘Price’ and ‘Convenience’ was associated with greater values in adiposity-related parameters, that is, BMI for ‘Price’ (β = 2·05, P = 0·013) and WC (β = 7·09, P < 0·001 for BMI, and β = 5·91, P = 0·004 for WC, respectively), while ‘Mood’ and ‘FI’ showed a significant association with FSG levels (β = 5·21, P = 0·003 and β = −4·42, P = 0·037, respectively).
Associations between food choice motives’ relative importance (predictor variables) and cardiometabolic markers (outcome variable) in the total study population (n 395), women (n 282) and men (n 113). Multivariable linear regression models were used to calculate the β coefficient (circles for total study population, diamonds for women and squares for men) and 95 % CI. Models were adjusted for age (years), sex (except for stratified sex analyses), BMI (kg/m2, except for BMI and WC outcomes), socio‑economic status (middle-low, high), physical activity (MET-min/week), smoking habit (no, yes), stress/anxiety/depression (no symptoms, symptoms in one domain or symptoms in two or more domains) and emotional eating (emotional eaters and non-emotional eaters). WC, waist circumference; FSG, fasting serum glucose; TAG/FSG index, TAG-to-fasting serum glucose index; H&NC, Health and Natural Content; IM, image management; CO, convenience; SA, sensory appeal; FI, food identity; E&WA, Environmental and Wildlife Awareness. *Statistically significant results: P < 0.05, **P < 0.01, ***P < 0.001.
In women (Figure 1 and Supplementary Tables 4 and 5), ‘Price’, ‘Convenience’ and ‘SA’ were significantly and positively associated or had a trend with BMI and WC; however, ‘FI’ and ‘E&WA’ were inversely associated or had a trend with these variables. Additionally, ‘Price’ and ‘Mood’ showed positive associations with FSG levels (β = 3·84, P = 0·038; β = 5·28, P = 0·009, respectively), and ‘SA’ was positively associated with TAG/FSG index (β = 0·26, P = 0·041) and had a trend with TAG:HDL-cholesterol ratio (β = 0·61, P = 0·076) and TAG (β = 23·43, P = 0·081).
In men, relative importance placed on ‘Price’ had a trend in the association with WC (β = 6·56, P = 0·057). Additionally, a higher importance placed on ‘IM’ showed an inverse association or a trend with FSG (β = −11·67, P = 0·063), TAG (β = −88·34, P = 0·038), TAG/FSG index (β = −0·59, P = 0·090) and TAG:HDL-cholesterol ratio (β = −0·84, P = 0·017) and a positive association with HDL-cholesterol (β = −22·72, P = 0·074). In addition, ‘Mood’ showed a significant positive association or had a trend with TAG (β = 46·22, P = 0·078), TAG/FSG index (β = 0·46, P = 0·032) and TAG:HDL-cholesterol ratio (β = 1·49, P = 0·041).
For the continuous CMRS, ‘SA’ and ‘Convenience’ were significantly and positively associated with this variable or had a trend, both in total population (Figure 2A) and in women (Figure 2B). Conversely, among men, only greater concern about the ‘IM’ motive consistently showed a negative association with it (Figure 2C).
Associations between relative importance to food choice motives (predictor variables) and cardiometabolic risk score (outcome variable) in the total study population (n 340) (A), women (n 249) (B) and men (C) (n 91). Multivariable linear regression models were used to calculate the β coefficient (in circles) and 95 % CI (presented both in the whisker plots and in parentheses). Model 1: unadjusted. Model 2: adjusted for age (years), sex (except for stratified sex analyses), BMI (kg/m2, except for BMI and waist circumference outcomes), socio‑economic status (middle-low, high), physical activity (MET-min/week), smoking habit (no, yes), stress/anxiety/depression (no symptoms, symptoms in one domain or symptoms in two or more domains) and emotional eating (emotional eaters, non-emotional eaters). *Bold indicates statistical significance (P < 0·05).
Discussion
Our study is the first to assess the associations between the relative importance of various dimensions of FCM and cardiometabolic risk among Mexican university students. Overall, we observed that greater relative importance placed on ‘Price’, ‘Convenience’, ‘SA’ and ‘Mood’ motives was associated with a poorer cardiometabolic risk profile, with females showing more pronounced patterns of association. In contrast, in our study, motives such as ‘IM’, ‘FI’ and ‘E&WA’ were associated with more favourable cardiometabolic outcomes. The ‘Health and Natural Content’ motive, however, appeared to be less influential, showing no significant associations.
University students often report limited financial resources(Reference Wang, Wang and Zhao57). In our study, the U-FCQ assesses ‘Price’ as a general motive(Reference Salas-García, Bernal-Orozco and Díaz-López34), capturing participants’ subjective perceptions of affordability rather than objective cost calculations. However, these price-related and concerns, combined with academic pressure, tight schedules, time scarcity and allure of palatable food, may exacerbate the knowledge–practice gap and may influence students to prioritise convenience over nutritional quality or health considerations(Reference Pearcey and Zhan2,Reference Powell, Lawler and Durham5) , selecting more readily available, shelf-stable processed options, possibly perceived as a budget-friendly solution that maximise satiety and meal planning efficiency(Reference Darmon and Drewnowski7,Reference Salas-García, Bernal-Orozco and Díaz-López30,Reference Bøe, Hysing and Lønning58) such as ready-to-eat meals and snacks, avoiding the preparation of healthier home-cooked meals(Reference Drewnowski, Monsivais, Marriott, Birt, Stallings and Yates8) and illustrating the difficulty of maintaining healthy eating habits amid academic demands(Reference Shu, Zhang and Zhou25). This is how ‘Price’, ‘Convenience’ and ‘SA’ may become key factors shaping dietary behaviours in university students(Reference Robinson, Jones and Marty59,Reference Hruschka60) , even among health science students, typically considered more health-conscious(Reference Merhout and Doyle61).
While the perception of energy-dense foods being cheaper depends on the cost metric employed (e.g. ultra-processed foods may appear more economical when assessed per calorie, but fresh food can be less expensive when evaluated per gram or per portion)(Reference Robinson, Jones and Marty59), and evidence suggests that healthier dietary patterns are not necessarily more expensive than less healthy ones(Reference Vandevijvere, Pedroni and De Ridder62–Reference Batis, Rivera and Popkin64), perceived price barriers may stem from issues of food access, availability and purchasing habits than from actual cost differences(Reference Unar-Munguía, Cervantes-Armenta and Rodríguez-Ramírez65).
This tendency is reinforced by obesogenic university environments, where unhealthy foods options that emphasise sensory-driven food choices, including cafeterias, vending machines and social settings where indulgent foods are common(Reference Hernández-F, Figueroa and Colchero66,Reference Berthoud, Morrison and Ackroff67) , are more accessible than healthy alternatives(Reference Berthoud, Morrison and Ackroff67,Reference Roy, Soo and Conroy68) . Coupled with lifestyle changes common during this life stage – such as decreased physical activity and increased ultra-processed food consumption –, these conditions contribute to an excessive energy intake, abdominal fat accumulation and adiposity(Reference Bøe, Hysing and Lønning58,Reference Vilaro, Colby and Riggsbee69) , to an average 3·4 kg weight gain during students’ first year of college(Reference Selvakumaran, Lin and Hadgraft70), with an increase in BMI(Reference Bøe, Hysing and Lønning58,Reference Vadeboncoeur, Townsend and Foster71) and cardiometabolic risk(Reference Selvakumaran, Lin and Hadgraft70).
With regard to the above, food designed to maximise SA – such as sweets, fried snacks and sugary beverages – are typically energy-dense and have a high glycaemic load(Reference Vadeboncoeur, Townsend and Foster71). Frequent consumption of these foods can trigger postprandial spikes in blood glucose and TAG levels(Reference Shu, Zhang and Zhou25,Reference Monteiro, Cannon and Levy72) , which over time may impair insulin sensitivity and increase the risk of metabolic syndrome, even in young adults(Reference Monteiro, Cannon and Levy72). In this sense, while convenience-oriented food choices often contribute to energy imbalance and weight gain(Reference Almarshad, Algonaiman and Alharbi73), sensory-driven choices may exert a more direct impact on metabolic health by influencing diet quality and composition(Reference Gordon-Larsen74), which is why ‘SA’ may have had a positive association with TAG/FSG index, a marker of insulin resistance and metabolic dysfunction, which reflects the interplay between lipid and glucose metabolism(Reference Lopez-Jaramillo, Gomez-Arbelaez and Martinez-Bello53).
Besides ‘Price’, ‘Convenience’ and ‘SA’, greater importance to ‘Mood’ had a positive association with cardiometabolic factors such as FSG. Notably, individuals characterised as emotional eaters (76 %) placed significantly more importance to the ‘Mood’ dimension compared with non-emotional eaters (Supplementary Table 1). Emotional eating has been linked to the consumption of comfort foods high in sugars(Reference Bauer, van-der-Laan and de-Bruijn75), which can lead to rapid spikes in blood glucose levels(Reference Godet, Fortier and Bannier76). Furthermore, stress and negative emotions – such as anxiety, sadness or boredom – are well-documented triggers for emotional eating, as individuals may turn to food as a coping mechanism(Reference Dakanalis, Mentzelou and Papadopoulou77). Among university students, this behaviour may be further exacerbated by easy access to highly palatable, energy-dense foods on or near campus(Reference Berthoud, Morrison and Ackroff67,Reference Roy, Soo and Conroy68) , potentially contributing to impaired glucose regulation.
In contrast, there was a negative association between importance to ‘FI’ and FSG. In our questionnaire(Reference Salas-García, Bernal-Orozco and Díaz-López34), this dimension refers to the cultural significance of food and its role in social interactions. These results suggest that aligning food choices with cultural values may promote healthier dietary patterns, thereby supporting better glucose regulation. In fact, those students classified as having ‘unhealthy habits’ placed significantly less importance to this dimension compared with the other categories (Supplementary Table 1). For these Mexican students, FI may reflect a preference for traditional foods such as corn, beans, vegetables and fruits, as well as home-cooked meals(Reference Evers, Dingemans and Junghans78,Reference Sánchez-Velázquez, Luna-Vital and Morales-Hernandez79) . Notably, the majority of students in our study lived with their parents, who also were primarily responsible for meal preparation, which may facilitate sustained access to these culturally rooted dietary practices(Reference Betancourt-Núñez, Nava-Amante and Bernal-Orozco80,Reference Mills, Brown and Wrieden81) , potentially explaining the protective association observed. These whole-food options are typically rich in dietary fibre and antioxidants, which exert anti-inflammatory effects and may help regulate blood glucose levels(Reference Evers, Dingemans and Junghans78). This association is particularly relevant for college students, who are often navigating the transition to independent living and developing lifelong dietary habits(Reference Soederberg-Miller, Falbe and Chodur82). For these students, aligning their food choices to cultural or family values might serve as a protective factor against metabolic dysfunction. Moreover, a strong FI provides a sense of belonging and consistency towards dietary behaviours(Reference Nice and Joseph83), reducing the likelihood of impulsive or emotionally driven eating.
When performing sex-stratified analysis, we observed notable differences between men and women. Among men, assigning greater importance on ‘IM’ was negatively associated or had a trend with FSG, TAG, the TAG/FSG index and the TAG:HDL-cholesterol ratio, suggesting a potential protective effect on metabolic health. On the other hand, among women, placing importance on ‘E&WA’ was negatively associated or had a trend with BMI and WC. These findings suggest that, beyond the well-established hormonal differences influencing lipid profile, fat distribution and cardiometabolic risk patterns between sexes(Reference Zhernakova, Sinha and Andreu-Sánchez31) – which were also evident in our study – sociocultural gender roles shape food choices in distinct ways. These roles can influence dietary behaviours and lifestyle practices(Reference Saltzgiver, Nielson and Costello24), thereby contributing further to sex-specific cardiometabolic outcomes.
Specifically, for male students, the importance placed on ‘IM’ may reflect internalised societal ideals of masculinity, which often emphasise physical fitness, muscularity and a lean body composition(Reference Nordström, Coff and Jönsson84). It is possible that men prioritising ‘IM’ engage in specific eating behaviours, although we did not observe significant differences in the importance to this motive across food consumption quality categories (Supplementary Table 1). They may include portion control, carefully timing meals around physical activity, deliberately selecting foods with lower glycaemic load and heightened awareness of macronutrient composition of their meals(Reference Wardle, Haase and Steptoe85). All these practices could favourably impact glucose homeostasis and lipid metabolism(Reference Lefkowich, Oliffe and Hurd-Clarke86), contributing to the protective associations observed.
On the other hand, women in our study were more inclined towards pro-environmental behaviours, consistent with previous research(Reference Salas-García, Bernal-Orozco and Díaz-López30,Reference Peres, Costa and Silva87) , the inverse association with BMI and WC may reflect healthier eating practices aligned with environmental concerns, although we did not observe significant differences in the importance placed on ‘E&WA’ across food consumption quality categories (Supplementary Table 1). For example, these women may follow more mindful food consumption – such as controlling portion sizes and reducing food waste(Reference Acar-Tek, Karaçil-Ermumcu and Erdoğan-Gövez88) – and prefer locally produced minimally processed, whole foods(Reference Betancourt-Núñez, Nava-Amante and Bernal-Orozco80). Additionally, environmentally motivated eating often involves reducing the consumption of animal-based foods, which have been linked to higher BMI and obesity(Reference Serban, Ursoniu and Moleriu89,Reference Herrmann, Gonnet and Millogo90) . Our findings also align with evidence showing that adherence to environmentally sustainable dietary patterns, such as the Planetary Health Diet, is associated with reduced CVD risk in women compared with low adherence(Reference Reger, Leitzmann and Rohrmann91). Collectively, these environmentally conscious behaviours may underline the observed protective associations.
This study has several strengths. First, it is one of the few investigations – and the first conducted in Mexico – to examine the associations between FCM’s relative importance and cardiometabolic risk(Reference Sawicki, Ramesh and Bui92–Reference McCurley, Levy and Dashti95). By focusing on underlying FCM rather than dietary intake, as done in most previous research(Reference Jung, Dan and Pang21–Reference Shu, Zhang and Zhou25), this study adds a novel perspective to the field of nutritional epidemiology. It provides a more comprehensive understanding of the drivers behind eating behaviours and their potential implications for public health, policy-making and intervention design. Furthermore, the use of relative importance scores represents a methodological strength. When evaluating motives that are potentially important across the board, absolute scores fail to reflect how individuals prioritise among them(Reference Christensen, Harding and Voigt96). Analysing individuals’ motive priorities through relative importance may better capture the complexity of the motive structure and provide deeper insight into highlighting trade-offs and how priorities are balanced in the decision-making process(Reference Salas-García, Bernal-Orozco and Díaz-López30,Reference Konttinen, Halmesvaara and Fogelholm36) .
Moreover, this research offers valuable insights into how food choices influence metabolic health in a population often presumed to possess higher health literacy – health science university students – who nonetheless may face substantial barriers to healthy eating(Reference Drzal, Kerver and Strakovsky97). Importantly, we employed an updated and validated questionnaire tailored for the Mexican population to assess FCM(Reference Salas-García, Bernal-Orozco and Díaz-López34). Using a culturally relevant tool strengthens the validity of our findings, particularly in a context like Mexico, where cultural norms and socio-economic conditions strongly influence dietary decisions(Reference Gómez Delgado and Velázquez Rodríguez28).
Finally, the sex-stratified analysis highlights key differences between men and women, offering insights into how societal expectations and cultural norms shape food choices and associated health outcomes. By examining these relationships separately, this study provides valuable evidence on how sex-specific factors – such as societal expectation, cultural norms and sustainability concerns – may shape food choices and their health consequences.
Nevertheless, several limitations should be acknowledged. First, the cross-sectional design precludes establishing causal relationships. Second, data on FCM were self-reported, which introduces potential bias, due to the subjective nature of the responses and the influence of social desirability(Reference Anvari, Efendić and Olsen98). Participants may have provided answers they deemed socially acceptable rather than reflecting their true motivations. Moreover, because food choices are often habitual or unconscious(Reference Fernqvist, Spendrup and Tellström4), university students may find it difficult to accurately state the underlying motives for their selections. Third, the study sample was limited to health science university students, who may differ from the general population in terms of health knowledge and behaviours, limiting the generalisability of the results to other contexts. Fourth, the smaller sample size of male participants (n 113), compared with females, might have reduced statistical power to detect additional associations in sex-stratified analyses. Notably, some associations observed in men showed similar directional trends as those found in women but did not reach statistical significance, which may reflect insufficient statistical power rather than true absence of associations. For instance, the positive association between ‘Mood’ and FSG showed a consistent direction in both sexes, though only reaching significance in women, with similar patterns observed for ‘SA’ and the TAG/FSG index. Therefore, these non-significant findings in men should be interpreted cautiously and warrant further studies with larger male samples are needed to adequately assess sex-specific associations. Finally, longitudinal studies would help clarify the temporal relationships between food choices and cardiometabolic health outcomes.
The findings of this study have important implications for public health strategies and interventions aimed at improving dietary behaviours and metabolic health, particularly among university students. By identifying those FCM most strongly associated with adverse or protective health outcomes, this study provides policymakers, educators and healthcare providers a foundation for designing culturally relevant and sex-specific nutrition programmes. For instance, campus-based initiatives could promote sustainable eating practices among women, while fitness and nutrition programmes tailored to men could emphasise the link between diet, physical appearance and metabolic health. Addressing structural barriers, such as the affordability and accessibility of healthy foods, may help mitigate the negative impact of price-driven food choices.
Therefore, this study provides novel evidence on how diverse FCM, such as ‘Price’, ‘Convenience’, ‘SA’, ‘Mood’, ‘IM’, ‘FI’ and ‘E&WA’, are associated with key on cardiometabolic markers, including BMI, WC, FSG, TAG, TAG/FSG index, TAG:HDL-cholesterol ratio and overall cardiometabolic health, in Mexican university students. From a public health perspective, these findings highlight the importance of fostering healthier dietary behaviours on university campus. Interventions should focus on increasing the availability of affordable, nutrient-dense and convenient food options that are also appealing, to support cardiometabolic health. Educational programmes should emphasise the long-term metabolic risks of poor food choices while equipping students with practical tools to manage stress and time constraints. These strategies can empower students to make healthier food choices, even in the face of time and resources constraints, as well as competing academic and social demands. Importantly, the sex-specific patterns observed suggest that interventions may be more effective if tailored accordingly – for instance, by leveraging environmentally driven motives among women and image-related motives among men to promote healthier eating habits and prevent future cardiometabolic complications.
Conclusions
This study highlights the significant influence of FCM on cardiometabolic health in Mexican university students. Prioritising ‘Price’, ‘Convenience’ and ‘SA’ were consistently associated or had a trend with higher adiposity markers such as BMI and WC. Moreover, less healthy dietary behaviours suggest a tendency towards higher CMRS. Notably, motives linked to ‘E&WA’ were associated with more favourable anthropometric outcomes among women, while ‘IM’ showed protective associations in men.
Supplementary material
For supplementary material/s referred to in this article, please visit https://doi.org/10.1017/S0007114526108344
Data availability statement
Data are available upon request.
Acknowledgements
The authors wish to thank Karla Araceli Solórzano-Castanedo, Melissa Karina Navarro-Flores, José Gerardo Mora-Almanza, Paola Lizbeth Pérez-Malta, Beatriz Vázquez-Herrera, Nancy Yamile Márquez-Mayorga, David Vázquez-Candelario, Jonathan Ulises Martínez-Escobar, Valeria Estefanía Buenrostro-Velasco, Lizeth Estefanía Vargas-Díaz, Francisco Trujillo-Sánchez, Sarai Citlalic Rodríguez Reyes, Érika Martínez-López and Susan Andrea Gutiérrez-Rubio, who provided valuable support for the development of this project. The authors also thank Yolanda Fabiola Márquez-Sandoval, Rosa Ivette Vargas-de la Cruz and Elisa Vargas García, for their important contributions and feedback during the development and implementation of this research. Additionally, the authors extend their gratitude to the all the participants of this study.
Miguel Amaury Salas-García: Conceptualisation-Equal, Data curation-Lead, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Software-Equal, Validation-Equal, Visualisation-Equal, Writing – original draft-Lead, Writing – review and editing-Equal; María Fernanda Bernal-Orozco: Conceptualisation-Equal, Formal analysis-Supporting, Funding acquisition-Equal, Investigation-Equal, Methodology-Lead, Project administration-Lead, Resources-Equal, Supervision-Lead, Validation-Equal, Visualisation-Equal, Writing – original draft-Equal, Writing – review and editing-Equal; Andrés Díaz-López: Conceptualisation-Lead, Data curation-Lead, Formal analysis-Lead, Investigation-Equal, Methodology-Lead, Project administration-Equal, Software-Equal, Supervision-Lead, Validation-Lead, Visualisation-Lead, Writing – original draft-Lead, Writing – review and editing-Lead; Alejandra Betancourt-Núñez: Funding acquisition-Lead, Methodology-Equal, Project administration-Equal, Resources-Equal, Writing – review and editing-Supporting; Pablo Alejandro Nava-Amante: Data curation-Equal, Writing – review and editing-Supporting; Ulises De la Cruz-Mosso: Methodology-Supporting, Supervision-Supporting, Writing – review and editing-Supporting; Barbara Vizmanos: Conceptualisation-Equal, Formal analysis-Supporting, Funding acquisition-Equal, Investigation-Equal, Methodology-Lead, Project administration-Lead, Resources-Equal, Supervision-Lead, Validation-Equal, Visualisation-Equal, Writing – original draft-Supporting, Writing – review and editing-Equal.
This research was funded by the Programa para el Desarrollo Profesional Docente (PRODEP) from the Secretaría de Educación Pública for the resources that allowed the development of the macroproject from which the present study is derived (A.B.-N., grant number: UDG-PTC-1511). Additional support was provided by the Consejo Nacional de Humanidades, Ciencias y Tecnologías (CONAHCYT), now called Secretaría de Ciencia, Humanidades, Tecnología e Innovación (SECIHTI), which granted doctoral scholarships to M.A.S.-G. (863882) and P.A.N.-A. (1035497). Furthermore, researchers M.F.B.-O. (176919), A.B.-N. (484393), U.D.-M (439490) and B.V. (22064) were granted funds by the Sistema Nacional de Investigadoras e Investigadores (SNII), also by CONAHCYT (now SECIHTI).
The authors declare none. The funders had no role in the design of the study, in the collection, analyses or interpretation of data, in the writing of the manuscript or the decisions to publish the results.
