Assessing Lifestyle in a Large Cohort of Undergraduate Students: Significance of Stress, Exercise and Nutrition

Abstract

Background/objectives: Lifestyle (in particular, nutrition and exercise) determines present and future youths' health. The goal of the present study was to identify specific student groups who deserve precise lifestyle improvement interventions, tailored to their characteristics.

Methods: An anonymous web-based questionnaire to assess lifestyle was posted on the websites of two main Italian Academic Institutions, and 9423 students voluntarily participated. A personalised immediate report was provided to improve compliance/motivation. We assessed age, sex, affiliation, anthropometrics, lifestyle components (nutrition, exercise, sedentariness, stress perception, smoking, alcohol, sleep), and the desire to be helped with lifestyle improvement. Cluster analysis was performed to identify healthy lifestyle groups among the students.

Results: In total, 6976 subjects [age: 21 (20, 23) yrs; 3665 female, 3300 male] completed the questionnaire and were included. Of these students, 73.9% expressed the need for lifestyle improvement help, particularly for becoming physically active (66.7%), managing stress (58.7%), and improving nutrition (52.7%). We unveil three clusters of subjects, each corresponding to a distinct lifestyle pattern. The clusters are differentiated by exercise level and perceptions of stress/fatigue/somatic symptoms (cluster 1: 74.8% meet international exercise guidelines (IEGs), 67.4% have high stress perception, 49.1% drink 1-3 glasses of wine/beer per week, and 63.3% drink 0-1 glass of spirits per week; cluster 2: 75.6% meet IEGs, 75.7% have low/medium levels of stress perception, and 65.8% have low alcohol consumption; cluster 3: 72.5% do not meet IEGs, 77.6% have high stress perception, and 67.5% have low alcohol consumption). More active students present lower stress/somatic symptoms perception. Interestingly, the AHA diet score (nutrition quality) was not in the ideal range in any cluster (nevertheless, obesity was not of concern), being worst in cluster 3, characterized by higher stress perception (59.7% had poor nutrition quality). Those who were physically active but showed a high stress/fatigue perception were used to drinking alcohol.

Conclusions: Students desire help to improve their lifestyle, and this approach might help identify specific student groups to whom LIs in Academic Institutions can be tailored to foster well-being and promote health.

Keywords: lifestyle assessment; nutrition quality; physical activity; public health; stress management; tailored intervention; well-being.

Figure 1

Multiple Correspondence Analysis map. The plot shows the modalities of the investigated variables (see the legend below) in the space spanned by the axes determined by MCA. The grey segments connect the modalities of the variables related to stress perception, namely s4SQ, stress, and fatigue. WC = waist circumference; AHA_Score = AHA nutrition score.; BMI = body mass index; METsMV = moderate and vigorous physical activity volume (adequate if ≥600 MET·min/week, otherwise insufficient); _WD = sedentary behaviour during working days (active if <9 h/week, otherwise sedentary); _WE = sedentary behaviour during weekends (active if <9 h/week, otherwise sedentary); sleep_hours = hours of sleep (adequate if ≥7 h/day, insufficient otherwise); stress: LL (low: 0–2), L (moderate/low: 3–4 points), M (moderate: 5–6 points), H (moderate/high = 7–8 points), HH (high: 9–10 points); fatigue: LL (low: 0–2), L (moderate/low: 3–4 points), M (moderate: 5–6 points), H (moderate/high = 7–8 points), HH (high: 9–10 points); s4SQ = short questionnaire on subjective somatic-stress-related symptoms: LL (low: 0–3 points), L (moderate/low: 4–6 points), M (moderate: 7–10 points), H (moderate/high: 11–15 points), HH (high: 16–30 points); smoke = smoking habits: SM (smoker), FSM (former smoker), EC (electronic cigarettes), and NS (non-smoker); Wine_beer = wine and beer consumption: none (0 glass/week), low, medium, high (0, 0.1–1, 1.1–2.9, and >2.9 glasses/week); spirits = spirit consumption: none (0 glasses/week), low (0.1–1 glasses/week), high (>1 glasses/week). The modalities of the variables of interest (e.g., smoker, non-smoker) are represented by points, and the presence of points close to one another reveals that the corresponding modalities are tendentially observed together. On the right side of the plot (X-axis), the following variables are found: fatigue = LL (very low), L (moderate/low), and M (moderate); stress = LL (very low), L (moderate/low), and M (moderate); and s4SQ = LL (very low) and L (moderate/low). Higher levels of the same variables are observed in the left part. Note that the different classes for perceptions of stress, fatigue, and somatic symptoms are in progressive order, as evidenced, respectively, by the lines. In the top part of the figure (Y-axis) and near each other, the following variables are found: smokers (also smokers of electronic cigarettes, represented by the label EC), former smokers, and the highest levels of wine and beer (represented by the label Wine_Beer-High) and spirit (Spirits-High) consumption. The lowest levels of the same variables are observed in the bottom part.

Figure 2

Heatmap showing distributions of the student’s features for each cluster. The X-axis represents the three clusters. The Y-axis represents all the possible modalities of the 13 variables used for cluster analysis. The numbers in the cells express the percentages of students, showing the modalities of the variables for each cluster. See the legend of Figure 1 for abbreviations.

Figure 3

Characterization of students’ lifestyles according to the three clusters. AHA = American Heart Association.

Figure 4

PCoA plot. The figure shows students represented in two-dimensional spaces that preserve the highest possible number of differences (goodness—of—fit index equal to 18.8%). Distinct labels represent students belonging to distinct clusters: empty squares for cluster 1, grey crosses for cluster 2, and black triangles for cluster 3.