The concept of Mediterranean Diet (MeDi) was introduced in the late 1950s, describing the dietary habits found throughout the Mediterranean coast(Reference Keys and Grande1). Traditional MeDi is characterised by a high consumption of vegetables, fruits, legumes, nuts, unrefined cereals, fish and olive oil and a low consumption of dairy products and red meat(Reference Estruch, Ros and Salas-Salvadó2,Reference Willett, Sacks and Trichopoulou3) . The Mediterranean dietary pattern has been consistently shown to provide a grade of protection against CVD and main non-infectious chronic diseases, such as cancer or diabetes(Reference Estruch, Ros and Salas-Salvadó2–Reference Mentella, Scaldaferri and Ricci7). A wide range of studies support MeDi as an effective preventive tool to reduce morbidity and mortality in the general population(Reference Estruch, Ros and Salas-Salvadó2,Reference Sofi, Cesari and Abbate8,Reference Kastorini, Panagiotakos and Chrysohoou9) .
MeDi compliance has declined dramatically in recent years(Reference Bonaccio, Di Castelnuovo and Costanzo5,Reference Sofi, Cesari and Abbate8) . Particularly, young adults are strongly influenced by socio-cultural changes and tend to develop unhealthy eating habits and decrease their MeDi consumption(Reference Papadaki, Hondros and Scott10–Reference Guillem-Saiz, Wang and Guillem-Saiz13). There are multiple factors that affect the quality of diet: demographic characteristics, unhealthy behaviours – such as sedentary lifestyle – and family and social influence(Reference Lipsky, Nansel and Haynie14). In fact, a possible association has been established between the social context – eating more frequently outside home and accompanied – and changes in dietary habits(Reference Maugeri, San-Lio and Favara15). In addition, alcohol consumption is an important aspect in undergraduate students, a population prone to risky alcohol consumption(Reference Davoren, Demant and Shiely16,Reference Messina, D’Angelo and Ciccarelli17) .
Although measuring the level of adherence to a Mediterranean dietary pattern is not easy, it is of great importance for dietary advice in routine clinical practice(Reference Olmedo-Requena, González-Donquiles and Dávila-Batista18). For this purpose, diet quality indices have been proposed to assess the degree of adherence to MeDi and its health benefits(Reference Zaragoza-Martí, Cabañero-Martínez and Hurtado-Sánchez19,Reference Schröder, Fitó and Estruch20) .
On the other hand, it has been shown that increased nutritional knowledge is associated with improved dietary habits(Reference Bonaccio, Di Castelnuovo and Costanzo5,Reference Fiore, Ledda and Rapisarda21,Reference Rodrigo Vega, Ejeda Manzanera and Gonzalez Panero22) . Likewise, it is assumed that greater knowledge and better habits influence a better transmission of nutritional advice(Reference Saliba, Sammut and Vickers23,Reference Vickers, Kircher and Smith24) . The latter is increasingly required in health professionals due to the previously described. Adherence to MeDi by medical students has been little studied(Reference Fiore, Ledda and Rapisarda21,Reference Durá-Travé and Castroviejo-Gandarias25) . Their knowledge is of interest because, in addition to assessing the MeDi compliance of university students, they are future health professionals.
The purpose of this study is to evaluate the use of MeDi by medical undergraduate students and to analyse the influence of knowledge acquired over the years of study as well as other factors on dietary compliance.
Material and methods
Study design
This is a cross-sectional observational study to assess adherence to MeDi in medical students.
Population, study scope and recruitment
The study population included students enrolled in the medical degree of the Faculty of Health Sciences of the University of Las Palmas de Gran Canaria. It involved students from first to sixth year of the 2018–2019 academic year. Likewise, the candidates to medical resident interns of the province of Las Palmas of the same year. The only inclusion criteria was to belong to one of these courses at the time of the study. The only exclusion criteria was a refusal to participate. A survey was carried out and sent by e-mail and handed out on paper to all students. All students who responded to the survey (61·85 %) were included.
Variables of interest and data collection
A survey was carried out (see Appendix 1), collecting data on age, sex, weight and height as self-referenced. BMI was estimated and categorised according to WHO criteria as: underweight (IMC < 18·5 kg/m2), normal weight (IMC ≥ 18·5 years < 25 kg/m2), overweight (IMC ≥ 25 years < 30 kg/m2) and obese (IMC ≥ 30 kg/m2)(26). In addition, personal medical history (hypertension, diabetes, dyslipidaemia…) and lifestyle-related variables such as physical activity and alcohol-tobacco consumption are detailed. In relation to physical activity, the fundamental components of the training load(Reference Verjoshanski27) are included: (1) the frequency with which physical activity is performed – none, 1–2 sessions per week or greater than or equal to 3 sessions; (2) the volume of time per session – less than 30 min, 30 min to one hour or more than 1 hour and (3) the subjective intensity of the exercise performed – light, moderate, intense or maximum. Alcohol consumption was also collected, including questions from the AUDIT-C questionnaire(Reference Bush, Kivlahan and McDonell28–Reference Reinert and Allen30), a short version of the AUDIT test consisting of the first three items of the AUDIT, including frequency and amount of alcohol consumption and frequency of binge drinking. The AUDIT-C cut-off point of ≥ 4 in men and ≥ 3 in women was used to identify hazardous alcohol consumption(Reference Reinert and Allen30). The 14-point Mediterranean Diet Adherence Score (MEDAS-14) questionnaire, widely validated for the Spanish population(Reference Schröder, Fitó and Estruch20) and in its English version(Reference Papadaki, Johnson and Toumpakari31) and simple to complete,(Reference Estruch, Ros and Salas-Salvadó2,Reference Schröder, Fitó and Estruch20) was used to record adherence to MeDi. The MEDAS-14 includes fourteen dichotomous response questions (yes/no) and the total adherence score ranges from a minimum of zero points to a maximum of fourteen and a score of nine points or more is considered good diet adherence(Reference Schröder, Fitó and Estruch20).
The questionnaires were completed through anonymous and voluntary surveys carried out in paper or electronic format. It was necessary to contact them in person or through institutional mail. If no initial response was obtained, up to three contacts were made.
Ethical considerations
This study was authorised by the Ethics and Clinical Research Committee of the University Hospital of Gran Canaria Doctor Negrín. Likewise, the approval of the Faculty of Health Sciences of the University of Las Palmas de Gran Canaria was also granted.
Statistical analysis
The data were analysed with the statistical package IBM SPSS software (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 26.0., IBM Corp.). Categorical variables are expressed as percentages and quantitative variables as mean and sd or median and interquartile range, depending on whether or not the distribution was normal. Normality of the quantitative variables was assessed using the Kolmogorov–Smirnov test.
To evaluate the relationship between qualitative variables, the Chi-square test or Fisher’s exact test was used, and for the association between quantitative variables and MeDi compliance, the Student’s t test or the Mann–Whitney U test was used, depending on whether or not the variables followed a normal distribution. To identify the variables independently associated with MeDi, a multivariable analysis was performed using logistic regression in which the variables that were significantly related in the univariate analysis were included. These variables were age, gender, academic year, physical activity, alcohol and tobacco consumption. Differences with a P value < 0·05 were considered significant.
Results
Of 589 medical students included in the study, 430 (73 %) were female and 159 (27 %) were male with a mean age of 22·1 years (sd: 3·1), range: 18 to 39, median 22 (interquartile range: 22–24).
From the total, 242 (41·1 %) were in the first to third year of their degree course and 347 (58·9 %) were in the second cycle, including resident opponents. The distribution according to academic year is detailed in Table 1.
Table 1 Distribution of students according to academic course
Abs freq, absolute frequency; IQR, interquartilic range.
According to BMI, 70·6 % (416) of the subjects included in the study were normal weight, 9 % (53) were underweight, 15·1 % (89) were overweight and 5·3 % (31) were obese, i.e. 20·4 % (120) were overweight or obese.
In relation to cardiovascular risk factors, 8 (1·4 %) reported hypertension, 3 (0·5 %) diabetes and 3 (0·5 %) dyslipidemia.
Regarding smoking habit, 559 (94·9 %) respondents reported not smoking, 9 (1·5 %) were smokers and 21 (3·6 %) defined themselves as ex-smokers (≥ 1 year without smoking).
In relation to alcohol consumption, the mean AUDIT-C score was 2·74 (± 2·01), being higher in males (3·4 ± 2·3) compared to females (2·5 ± 1·8) (P < 0·001). Seventy-nine percent of the students consumed alcohol at least once a month, and one-third (35·3 %) between 2 and 4 times a month, with 5 % consuming alcohol four times a week or more. The amount consumed per occasion was at least 3–4 drinks in one-third of the cases (32·4 %), with 8·5 % consuming more than five drinks. Half of the students (49·9 %) had risky alcohol consumption, with no differences according to sex (50·5 % in females and 48·4 % in males, P = 0·66), or age (22·0 v. 22·1 years; P = 0·84). However, a significant association was observed between risky alcohol consumption and tobacco consumption (80 % v. 48·3 %; P = 0·001; OR: 4·3, 95 % CI: 1·7, –10·6), and belonging to the highest medical school grades (5th or 6th grade v. 1st to 4th grades; 57·2 v. 45·3; P = 0·007; OR: 1·6, 95 % CI: 1·1, –2·3).
When it comes to physical activity, 71 % exercised at least once a week compared with 29 % who did not exercise regularly. The amount of time spent was less than 30 min per session in 10·7 %, from 30 min to 1 hour in 42·1 % and more than 1 hour in 23·8 %. Finally, exercise intensity was mild in 8·8 %, moderate in 39·4 % and intense maximum in 29·4 %.
The majority of the respondents (76·1 %) did not usually cook at home. There were a tendency to cook at home in second-cycle students compared to first-cycle students (28·7 % v. 20·2 %; P = 0·07).
Adherence to MeDi was acceptable in 58·9 % students. Table 2 shows the individualised analysis of the MEDAS-14 questions, globally and by sex. The mean MEDAS-14 score obtained was 8·9 (±1·9) points, with no differences according to sex (8·9 (±1·9) in women and 8·72 (± 2·0) in men (P = 0·46)). Fish or seafood consumption was significantly higher in men (42·1 % v. 25·1 %; P < 0·001). Consumption of white meat was higher in women (85·6 % v. 77·4 %; P = 0·017) and women consumed significantly less red meat (76·7 % women consumed less than once a day v. 68·6 % men; P = 0·043). Women also tended almost significantly to a higher consumption of vegetables (62·3 % v. 53·5 %; P = 0·051). There were no significant differences in the rest of the MEDAS-14 questions according to sex.
Table 2 Answer to each MEDAS-14 individual question and differences between women and men
* P < 0·05.
Relationship between the different variables and adherence to Mediterranean Diet
As shown in Table 3, adherence to MeDi was not related to sex; nevertheless, greater adherence to MeDi was observed at older age (P = 0·017).
Table 3 Relation between Mediterranean adherence and demographic characteristics, academic course, tobacco use and culinary habits
sd, standard deviation.
In the analysis by academic year (Fig. 1), MeDi consumption was higher in students in the second cycle (fourth to sixth year of studies and medical opponents) compared to students in the first three years (OR = 2·3; 95 % CI: 1·6, 3·1; P < 0·001).
Fig. 1 Differences in Mediterranean Diet adherence by academic years
No higher adherence to MeDi was observed in those who cooked at home compared to those who did not. Nor was it observed in normal-weight patients with respect to overweight or obese patients.
In relation to tobacco consumption, there was no statistically significant but close relationship (P = 0·06) in favour of a better dietary habit in non-smokers.
The frequency of alcohol consumption was associated with adherence to MeDi (P = 0·008): students who consumed alcohol two or more times per month showed a significantly higher adherence to MeDi (P = 0·002) (Table 4). However, MeDi compliance was not associated with the amount consumed per occasion (P = 0·350).
Table 4 Relation between Mediterranean adherence and physical activity and alcohol consumption
AUDIT-C, Alcohol Use Disorder Identification Test–Consumption.
AUDIT-C cut-off = low-risk drinking (score < 3 in women and < 4 in men) and hazardous drinking (score ≥ 3 in women and ≥ 4 in men).
Adherence to MeDi was highly significantly associated (P < 0·001) with the performance of physical activity (Table 4), both with frequency (P < 0·001), intensity (P = 0·003) and volume of exercise performed (P = 0·001). Figure 2 represents a summary of the results described above in a more visual form.
Fig. 2 Most relevant variables associated to Mediterranean Diet
Multivariable analysis
The significant association between greater adherence to MeDi in second-cycle students compared to first-cycle students was maintained (OR = 2·1; 95 % CI = 1·3, 3·2; P = 0·001). Similarly, MeDi consumption was higher in those who consumed alcohol two or more times a month (OR = 1·5; 95 % CI = 1·0, 2·1; P = 0·039) and those who were physically active (OR = 1·5; 95 % CI = 1·2, 1·9; P < 0·001). The remaining variables were not independently associated with MeDi compliance although tobacco use showed a trend toward statistical significance in its negative association with MeDi consumption (OR = 0·5; 95 % CI: 0·2, 1·01; P = 0·055) (Table 5).
Table 5 Multivariable analysis to evaluate the factors independently associated with adherence to the Mediterranean Diet
Discussion
Adherence to MeDi was acceptable in 59 % of undergraduate medical students. This degree of adherence, despite not being optimal, is notably better than that obtained in other studies in Spanish undergraduates in general, which show good adherence in only 36 %(Reference López-Moreno, Garcés-Rimón and Miguel32) and 34 %(Reference Cobo-Cuenca, Garrido-Miguel and Soriano-Cano33), using the same MEDAS-14 questionnaire. Other studies using different questionnaires have also observed a low percentage of university students with good compliance(Reference García-Meseguer, Cervera-Burriel and Vico-García11).
This difference could be justified by the evidence that future physicians, compared to the overall university population, probably show greater concern for their health. Indeed, it is worth noting that MeDi compliance was significantly higher in students with higher grades. In this case, since these are medical students, it could be considered that nutrition knowledge may be increased. This could support that higher MeDi knowledge is related to higher MeDi adherence. In fact, there are studies that support that teaching in nutrition-food subjects implies greater adherence to MeDi(Reference Rodrigo Vega, Ejeda Manzanera and Gonzalez Panero22). An Italian study that evaluated nutritional knowledge and its association with adherence to MeDi showed a significant association between both: the greater the nutritional knowledge, the greater the adherence to MeDi(Reference Aureli and Rossi34). Furthermore, a study that compared differences in terms of adherence to MeDi in health science students during their first academic year with respect to the second found greater adherence at the end of the second academic year, which could be justified by greater knowledge about dietary habits(Reference Sanchez-Fideli, Gutiérrez-Hervás and Rizo-Baeza35). Nevertheless, another study that analysed adherence to MeDi in medical students in Italy did not observe that being in the first or last courses influenced the results(Reference Fiore, Ledda and Rapisarda21).
Likewise, in our study, adherence to MeDi increases significantly with age. This association has also been described recently in children and adolescents in Italy(Reference Bonaccorsi, Furlan and Scocuzza36), although contrary results, lower adherence at older age, are described in most studies included in a meta-analysis that also assessed adherence to MeDi in children and adolescents(Reference Iaccarino-Idelson, Scalfi and Valerio37).
It is worth emphasising that lifestyle habits are developed from childhood and become entrenched in adolescence. Diet of young people and especially of university students poses an important challenge, as it may involve major changes in their lifestyle. In fact, another systematic review confirmed that the diet of adolescents tends to be characterised by an unsatisfactory dietary intake(Reference Moreno, Gottrand and Huybrechts38). The importance that nutritional education can have on medical students should be emphasised, given that physicians have a very important role and opportunity to advise their patients on diet.
In the individualised analysis of the MEDAS-14 questions, a low overall consumption of fish was observed, a fact previously noted in university students(Reference López-Moreno, Garcés-Rimón and Miguel32,Reference Cobo-Cuenca, Garrido-Miguel and Soriano-Cano33) , but it was significantly higher in men, contrary to what was described by Cobo-Cuenca(Reference Cobo-Cuenca, Garrido-Miguel and Soriano-Cano33). It was also observed that women consume more white meat compared to men who consume significantly more red meat, hamburgers or sausages. A higher consumption of red meat has been previously described in men, as well as a higher consumption of vegetables in women(Reference Predieri, Sinesio and Monteleone39), an aspect that in this study was also found to be close to statistical significance. It has been described in the literature that in Western societies women tend to show better dietary habits and give more importance to body weight than men(Reference Arganini, Saba and Comitato40). These aspects contributing to the better adherence to MeDi observed in several studies in young women(Reference Bonaccorsi, Furlan and Scocuzza36,Reference La Fauci, Alessi and Assefa41) , as well as in medical students(Reference Fiore, Ledda and Rapisarda21). However, other studies show no differences according to sex(Reference Iaccarino-Idelson, Scalfi and Valerio37), as in our results. Similar findings, that gender does not influence adherence to MeDi have been observed in studies in university students in Spain(Reference López-Moreno, Garcés-Rimón and Miguel32) and Lebanon(Reference Karam, Bibiloni and Serhan42).
It is worth mentioning that the percentage of overweight obesity in our sample is 20·9 %, similar to other studies(Reference Cobo-Cuenca, Garrido-Miguel and Soriano-Cano33). Even though this series did not find a relationship between overweight obesity and adherence to MeDi, another study in Italian adolescents found that good adherence to MeDi significantly reduced the likelihood of overweight obesity(Reference Mistretta, Marventano and Antoci43). Other studies have reported weight gain in patients with increased adherence to MeDi mainly at the expense of lean mass(Reference Cobo-Cuenca, Garrido-Miguel and Soriano-Cano33).
On the other hand, cooking at home was not related to adherence to MeDi. Some studies(Reference Papadaki, Hondros and Scott10,Reference Durá-Travé and Castroviejo-Gandarias25) report that students who live away from home develop worse habits than those who live at home, associating this with a decrease in the intake of home-cooked meals. In a similar vein, a study showed that eating away from home was associated with a lower trend to consume vegetables, fruits and legumes and a higher predisposition to consume processed meat, salty snacks and carbonated beverages(Reference Maugeri, San-Lio and Favara15).
About smoking, a trend toward greater adherence to MeDi was observed in non-smokers, probably in relation to a healthier lifestyle. Some authors describe the same association between a lack of adherence to a healthy diet and tobacco consumption(Reference Roig Grau, Rodríguez Roig and Delgado Juncadella44,Reference Elizondo, Guillén and Aguinaga45) .
In relation to alcohol consumption, approximately one over three students (35·3 %) consume alcohol 2 to 4 times a month and 5 % 2–3 times a week. In this study, a greater adherence to MeDi was found in students with a higher frequency of alcohol consumption. This is somewhat expected since MeDi includes regular moderate ethanol consumption, mainly in the form of wine. A Spanish study of health science students found a strong association between adherence to MeDi and those who consumed alcohol such as wine or beer compared to both abstainers and consumers of other distilled beverages(Reference Scholz, Navarrete-Muñoz and García de la Hera46). Regarding the benefit of moderate alcohol consumption, specifically wine, several authors report that despite discretely increasing the risk of some types of cancer, it is consistently associated with a reduction in cardiovascular risk and, therefore, has an overall protective effect on total mortality(Reference Arredondo Bruce and del Risco Morales47). Nevertheless, other studies suggest that the protective associations between alcohol consumption and mortality may be attributable in part to inappropriate selection of the reference group and weak adjustment for confounding factors(Reference Knott, Coombs and Stamatakis48). Therefore, there is controversy as to whether or not moderate alcohol consumption should be recommended as part of a balanced and healthy diet.
However, MeDi compliance was not associated with the beverages consumed per occasion. It is important to highlight a pattern of consumption characterised by drinking large amounts in short periods of time, or binge drinking, which is frequently observed in young people. This alcohol consumption deviates from MeDi drinking pattern and can have significant negative repercussions on health. In our study, an 8·5 % of the respondents were at risk for alcohol consumption. Two recent Spanish studies, one of them involving first-year university students(Reference Romero-Rodríguez, Amezcua-Prieto and Morales-Suárez-Varela49) and the other one all-year science students(Reference López-Moreno, Garcés-Rimón and Miguel32), concluded that 16·9 % and 26·2 %, respectively, had a risky alcohol consumption. These percentages are higher than in our sample of medical students and correlate with that described in other studies on university students. In deed, in a systematic review assessing alcohol consumption in Irish and UK university students,(Reference Davoren, Demant and Shiely16) over 20 % exceeded sensible limits each week, and a high frequency of at-risk consumption was observed. Another recent study conducted in Italy(Reference Messina, D’Angelo and Ciccarelli17) also shows a high percentage of high-risk (53·3 %) drinkers.
Furthermore, in our series, risk alcohol consumption is associated with smoking and with higher medical courses, which could point to poor stress management among our students. All this suggests the need to carry out an intervention program with the aim of reducing alcohol consumption and making future health professionals aware of the risks of alcohol and other toxic habits such as smoking. It is important to know about alcohol use in students, and to make them aware of its risks, especially in future health professionals.
Regarding physical activity, a highly significant association with MeDi was observed in our study. Several studies have shown an association between adherence to MeDi and physical activity in children and adolescents(Reference Iaccarino-Idelson, Scalfi and Valerio37). In addition, several studies support that good physical fitness and high adherence to MeDi are associated in isolation and in combination with a higher quality of life and lower morbidity and mortality(Reference Evaristo, Moreira and Lopes50).
Limitations and strengths
This study has several limitations. The cross-sectional nature of the study precludes establishing causality. The fact that the sample consisted of medical students and the voluntary nature of the survey may have influenced those included in the study were more concerned about a healthy diet. Furthermore, the use of questionnaires as a method of dietary assessment is limited by the fact that veracity of the data depends on the correct understanding of the questions and the accuracy of the information provided. Finally, no analytical determinations nor information on additional determinants of dietary choices such as economic or social determinants were collected. In relation to alcohol consumption, not having established the type of alcohol consumed (fermented or distilled beverage) limits the interpretation of the data.
Despite these limitations, the large sample size of the study and the non-exclusion of any student from participating allows us to obtain representative data on adherence to MeDi by medical students. Another strong point is the participation of students from all courses, which enables us to analyse the differences between them.
Conclusions
MeDi compliance by medical students was 58·9 %, being significantly higher at older age and in higher academic years. It was also significantly related to greater physical activity.
This could support the greater knowledge about nutrition and diseases, the greater dietary compliance. However, it would be useful to quantify the dietary knowledge of medical students and see its direct relationship with a better dietary habit.
It should be noted that almost one in two medical students have a risky alcohol consumption and one-third did not engage in regular physical exercise. This added to the above suggests the need to implement early nutritional and healthy lifestyle educational programs for university students in general and even more so for future health professionals.
Acknowledgements
Acknowledgements: not applicable. Authorship: The first author contributed with design of the study, data collection, analysis and drafting of the report. The second and last authors contributed to the conception, design and analysis of data. All authors contributed to revising the manuscript critically. Ethics pf human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Ethics and Clinical Research Committee of the University Hospital of Gran Canaria Doctor Negrín. Verbal informed consent was obtained from all subjects. Verbal consent was witnessed and formally recorded.
Financial support:
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest:
The authors declare that they have no conflict of interest.
Open access
