Background: The concept of health has undergone profound changes. Lifestyle Medicine (LSM) consists of therapeutic approaches that focus on the prevention and treatment of diseases. It follows that the quality of life of university students directly affects their health and educational progress.

Experimental Methodology: Socioeconomic, lifestyle (LS), and sense of coherence (SOC) questionnaires were administered to college students from three different areas. The results were analyzed for normality and homogeneity, followed by ANOVA variance analysis and Dunn and Tukey post hoc test for multiple comparisons. Spearman's correlation coefficient evaluated the correlation between lifestyle and sense of coherence; p values < 0.05 were considered statistically significant. 

Results: The correlation between LS and SOC was higher among males and higher among Medical and Human sciences students compared to Exact sciences. Medical students' scores were higher than Applied sciences and Human sciences students on the LS questionnaire. Exact science students' scores on the SOC questionnaire were higher than Human sciences students. In the LS areas related to alcohol intake, sleeping quality, and behavior, there were no differences between the areas. However, women scored better in the nutrition domain and alcohol intake. The SOC was also higher in men compared to women. 

Conclusion: The results obtained demonstrate in an unprecedented way in the literature that the correlation between the LS and SOC of college students varies according to gender and areas of knowledge, reflecting the importance of actions on improving students' quality of life and enabling better academic performance.

Data gathering

The researchers invited the students to answer an online form - through Google Forms virtual platform - containing the questionnaires: sociodemographic information, FANTASTIC questionnaire on Lifestyle, and a questionnaire on Sense of Coherence. The researchers clearly explained the research objectives and collection procedures on the home page, and the participants were given the Free and Informed Consent Form.

The data gathered in the online form were transferred to a spreadsheet in Microsoft Excel. The results were filtered, classified, and treated in order to be in line with the desired statistical analysis and could feed the statistical programs used.

Statistical analysis

The statistical analyses were performed by the JASP statistical software, and part of the graphics by the SPSS software. First, the researchers submitted the results to normality (Shapiro Wilk) and homogeneity (Levene test) analysis. Next, the normal homogeneous data were submitted to the ANOVA analysis of variance and Kruskal Wallis non-parametric test, followed by Dunn's post hoc test of multiple comparisons and Tukey's correction.

Spearman's correlation coefficient evaluated the correlation between Lifestyle and Sense of Coherence by determining the value of R. Values of p < 0.05 were considered statistically significant.

The normality of data was checked by the Shapiro-Wilk test, and since the distribution was not normal, analyses were performed as described below:

1. Descriptive results are presented by the median and interquartile ranges. 

2. The comparisons between the study variables in HA, ESA and HM (age, BMI, lifestyle, sense of coherence and domains of the questionnaires) and by gender (Lifestyle and Sense of Coherence) were performed by the Mann-Whitney test. 

3. The domains of the questionnaires in each group (HA, ESA, and HM) were compared by analyzing repeated measures, Friedman test, and the Post Hoc by Dunn's multiple comparisons test. 

4. The comparisons between lifestyle and Sense of Coherence among students in each of the selected courses were performed by analysis of variance, Kruskal-Wallis non-parametric test and Post Hoc by Dunn's multiple comparisons test. 

5. The correlations between the profile of lifestyle and sense of coherence of students in each area of knowledge and by gender were performed by Spearman's correlation coefficient. 

The significance index adopted in all analyses was 5% (p ≤ 0.05).