A cross-sectional study was conducted among 1305 female students who participated in a periodic health examination between April and May 2018 at a Japanese medical university. The students were from nine departments: radiological technology, clinical nutrition, physiotherapy, medical welfare, acupuncture, clinical engineering, medical information science, pharmaceutical sciences, and nursing. All participants provided written informed consent before participating in this study and understood that there was no penalty for choosing not to participate. The paper-based questionnaire was distributed only to students who had given written consent. Students who received the questionnaire completed it during their physical examination.
Body height and weight data was obtained from the periodic health examination. BMI was calculated as weight in kg/height in m2. Global SRH measures generally include a question asking “How would you rate your overall health?” with five response categories ranging from excellent to poor (i.e., excellent, very good, good, fair, and poor). These five categories are dichotomized into either good health (excellent, very good, or good) or poor/fair health (fair or poor) [21–23].
The seven independent variables for this study were grade (1–2 or ≥ 3), living status (living with others or living alone), smoking (none, past smoker, or current smoker), alcohol drinking (none, a few times per month, a few times per week, or daily), University of California Los Angeles (UCLA) activity score (< 5 points or ≥ 5 points) [24], Athens Insomnia Scale (AIS; non-insomnia or insomnia) [25, 26, 27], and frequency of breakfast consumption (< 6 days/week or ≥ 6 days/week).
Using Noh et al.’s method [28], we classified the participants into three groups by desired weight (underweight, overweight, or normal weight), as measured by the relationship between their current BMI and their desire to lose or gain weight. Physical activity was assessed using UCLA activity score [24], which we have used in previous research [29].
AIS is a self-assessment insomnia scale created by the World Health Organization for the World Project on Sleep and Health. This instrument’s reliability and validity have already been verified in other studies [25, 26, 27]. AIS items measure waking during the night, early morning awakening, total sleep duration, sleep quality, and daytime sleepiness. Each of the scale’s eight questions is answered from 0 (no problem) to 3 (a very serious problem). A total score of 4 or more on the selected items shows a suspicion of insomnia, and a total score 6 or more indicates insomnia.
We asked participants “Do you want to gain or lose weight in the future?” and defined their perceptions of their ideal weight into the three groups: underweight, overweight, and standard weight. If current BMI was < 18.5, but they wished to lose weight or stay the same, participants were defined as wanting to be underweight. Those with a current BMI < 18.5 who wanted to gain weight were defined as wanting to be on standard weight. Those with a current BMI of ≥ 25.0 who wanted to gain weight or stay the same weight were defined as wanting to be overweight, and if they wished to lose weight, they were defined as wanting to be on standard weight. If current BMI was 18.5–24.9, those who wanted to lose weight were in the wanting to be underweight group, those who thought they needed to gain weight were in the wanting to be overweight group, and those who wished to stay the same weight were in the wanting to be a standard weight group (Fig. 1).
Participants were classified as having either good SRH (excellent, very good, or good) or poor/fair SRH (fair or poor). Independent variables were compared between groups. Differences in categorical variables between those with good and poor/fair SRH were evaluated using the chi-square test.
Multivariate logistic regression analysis was used to examine the relationship between dietary habits, lifestyle, physical activity, recognition of physique, and SRH, after adjusting for grade and smoking. The independent variables included UCLA activity score, AIS scores, frequency of breakfast consumption, and recognition of physique. Adjusted odds ratios and 95% confidence intervals were calculated. All statistical analyses used JMP 9.0.2 (SAS Institute Inc., Cary, NC, USA).