Influences of muscle mass loss and exercise habits and personality traits on lower limb motor function among university students

Secondary sarcopenia, a risk factor even for young people, has attracted attention because of the deterioration of physical activity and nutritional status due to lifestyle change among university students. However, studies on the factors affecting motor function and their involvement are lacking. This cross-sectional study aimed to examine the influences of muscle mass loss and exercise and sleep habits on lower limb motor function, as well as the involvement of personality traits, in 101 university students. Approximately 6% of the participants had low skeletal muscle mass index, similar to previous reports, and that only exercise habits in high school were responsible for muscle mass loss (direct effect =  − 0.493; p < 0.05), wherease low skeletal muscle mass (direct effect =  − 0.539; p < 0.01) and current exercise habits (direct effect = 0.410; p < 0.01) were responsible for lower limb motor function. Additionaly, only the personality trait of high intellectual curiosity was involved in the establishment of exercise habits in high school, but no other personality traits showed a significant effect. In the prevention of secondary sarcopenia, encouraging sustained exercise habits while considering the influence of different personality traits is expected to prevent the decline in muscle mass and motor function.


Instrumentation and outcomes
We used a body composition analyzer with an operating multifrequency bioelectrical impedance analyzer (BIA) (MC-780A-N, Tanita Co., Ltd., Tokyo) and a motor function analyzer (Zaritz BM-220, Tanita Co., Ltd., Tokyo) to measure and calculate the body composition and lower limb motor performance (Fig. 1).In both measurements, the participants were asked to stand barefoot on these devices, and the participant's clothing was subtracted from their weight as 1.0 kg.
The following values were used in the evaluation using the body composition analyzer.

SMI
SMI is calculated as follows: appendicular skeletal muscle mass (ASM) (kg) divided by the square of height (m), which is an index of muscle mass for an individual's body size.The equation for ASM measured by BIA used in this study has been published and demonstrated to be highly correlated with body composition measured using dual-energy X-ray absorptiometry 31 .Furthermore, referring to the Asian Working Group Sarcopenia (AWGS2014) criteria for SMI to determine sarcopenia 5 , male and female participants with SMI of < 7.0 and 5.7, respectively, were categorized as the significantly reduced skeletal muscle mass (Low SMI) group.BMI BMI, an index that expresses the balance between height and weight, is calculated as weight (kg)/height (m)/ height (m).The standard range for BMI is from ≥ 18.5 to < 25 kg/m 232 .

Total fat mass
Total fat mass is the amount of body fat estimated by using a body composition analyzer.BIA is a noninvasive examination technique used assess bone mass, fat mass, and fat-free mass.It involves passing weak currents with three different frequencies (5, 50, and 250 kHz) through the body using 8 electrodes: 2 for each sole and grip in a standing barefoot position.The difference in electric resistance is determined to derive the aforementioned mass measurements 31,33,34 .

Ratio of resistance to reactance (R/X)
Ratio of resistance to reactance using 50 kHz of BIA at both legs representing the balance between the intracellular and extracellular fluids.Resistance is strongly affected by extracellular fluid, whereas reactance is the resistance generated by the cell membrane and is also strongly affected by intracellular fluid.Therefore, the balance between the intracellular and extracellular fluids can be evaluated by looking at these ratios, with the lower the value, the more intracellular components there are, indicating that the intramuscular density is higher 34,35 .
For the motor function analyzer, participants sat in a sitting chair with a seat height of 40 cm from the floor and placed of their bare feet on the analyzer.In response to a cue given by the dedicated software, the participant quickly stood up from the chair with both hands crossed in front of the chest, stopped for 3 s, and then sat down in the chair again following the cue.Regarding the measurement method employed in this action, previous studies 36,37 have reported intraclass correlation coefficients exceeding the excellence criterion, i.e., ≥ 0.7, with no variation in performances observed during our procedure review phase.Following confirmation that participants had sufficient understanding of the operation by reviewing the instructional video in the dedicated software and simulating the action, they performed the action only once.This enabled us to measure the participant's maximum effort to stand up with his/her lower limbs, and to evaluate his/her lower limb muscle strength based on the load variation value (ground reaction force index) obtained at that time, elapsed time from standing up www.nature.com/scientificreports/ to stabilization, and balance ability based on the amount of displacement of the center of gravity on the left, right, and upper/lower sides 38,39 .The sampling frequency of the analyzer was 80 Hz, and data were acquired once every 0.0125 s.
The following values were used in the evaluation using the motor function analyzer.

Power (F/w)
This indicates the strength in the action of standing up, with a higher value indicating that the participants were able to stand up with greater strength in the lower limbs.It is the peak load point generated in standing up, defined as F, divided by the load at the stable point after standing up, defined as body weight (w), which means how many times the force of the body weight was applied to stand up.It is calculated as F/w (kgf/kg), which is the maximum load divided by the body weight 36,37 .

Speed (RFD/w)
This represents the quickness of standing up, with the higher the value, the quicker the participants could stand up.It is the maximum slope (rate of force development; RFD) of the graph between the point where the load begins to be applied and the point where the maximum load is recorded, calculated per kilogram of body weight.
It is calculated as RFD/w (kgf/sec/kg), which is the value obtained by converting the amount of increase during a total of seven sections (0.0875 s), including one section where the maximum increase in load (ground reaction force) is recorded and three sections before and after that, into 1.0 s, and dividing by the body weight 36,37 .

Balance (Vx/Vw)
This indicates the degree of wobble from side to side in the action of standing up and the stable time (the time from the point when the maximum ground reaction force is recorded to the point when the wobble subsides), with higher values indicating more unstable standing movements.It is the value of variation in the left-right direction, defined as Vx (the value obtained by calculating the movement speed as an absolute value from the distance traveled to the left and right per section, and dividing by the stable time), divided by the load variation value defined as Vw (the value obtained by calculating the movement speed as an absolute value from the speed traveled up and down per section, and dividing by the stable time).It is calculated as Vx/Vw(mm/kg) 36 .
We used a web-based questionnaire and asked participants to complete the self-administered survey form.

Japanese version of the Ten Item Personality Inventory (TIPI-J)
TIPI-J is an evaluation scale that can determine an individual's personality traits based on a combination of five factors.In response to 10 questions, the scale calculates the strength of extraversion (items 1 and 6), agreeableness (items 2 and 7), conscientiousness (items 3 and 8), neuroticism (items 4 and 9), and openness to experience (items 5 and 10) by self-rating the degree of agreement on a 7-point Likert scale ranging from 1 (completely disagree) to 7 (strongly agree).Items 2, 6, 8, 9, and 10 of TIPI-J were reverse-scored 40 .

Exercise and sleep habits
The participants were asked to respond to the number of times they exercised per week during their high school club activities (exercise_past), number of times they currently exercise per week (exercise_current), and average hours of sleep per week (sleep hours).Previous studies have reported that those who exercise regularly even once or twice a week are more effective in preventing the risk of developing serious diseases than those who do not exercise even once a week 41,42 .Therefore, with regard to past and current exercise habits, we converted the responses of those who answered that they exercised at least once per week into the following binary values: those who answered that they exercised at least once per week were considered to have a habit, and those who answered that they exercised 0 time per week were considered to have no habit.

Procedures
Because personal information, such as body weight and body fat mass were important for university students, we ensured the anonymization of personal information from the data acquisition stage for the participants by implementing a system of research procedures.The participants were given a research ID by lottery and entered their research ID, age, sex, and height into a personal computer on which a dedicated software program linked to the body composition analyzer was installed.Then, the motion of standing up from a sitting position was measured using a motor function analyzer that can be continuously run using the same dedicated software as the body composition analyzer.The participants also completed a self-administered survey form using Google Forms with their own research IDs and completed the TIPI-J and the exercise and sleep habit survey.The study protocol was implemented in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Ethics Review Committee of Fukuoka International University of Health and Welfare (approval number: 21-fiuhw-015).The participants were informed about this study and their written consent was obtained.
Informed consent was obtained from all individuals in the images of Fig. 1 for publication in an online open access publication.

Statistical analyses
After testing for normality, descriptive statistics were calculated for all participants and for each sex.Using the Welch test, we examined whether there were significant differences in body composition, motor function, personality traits, and sleep duration between the sex groups and between the groups of those who fell below the criteria for sarcopenia using SMI (low muscle mass group) and the others (normal group).The variables exercise_past and exercise_current were examined using either the chi-square or Fisher's exact probability test.A correlation analysis (Pearson, polyserial, and polychoric) was used to calculate the correlation coefficient between each variable.The values of the correlation coefficients were presented as limited (0.00-0.25), fair (0.25-0.50), moderate (0.50-0.75), and excellent (0.75-1.0) 43 .To eliminate the possibility of multicollinearity in SEM analysis, we also checked for the relationships between variables with the correlation coefficients exceeding 0.9 44 .Additionally, we used SEM to check the goodness of fit between the hypothesized model (Fig. 2) and the data collected in this study and to examine the structural relationships among the variables by obtaining the standardized path coefficients for the convergent analytical model.Our hypothesized model centered on a causal relationship with regular exercise (exercise_past and exercise_current) and sleep (sleep hours) habits as independent variables and Low SMI as the dependent variable, and predicted a negative effect of Low SMI and a positive effect of regular exercise and sleep habits on motor function (latent variable consisting of power, speed, and balance) in the null point.Furthermore, we predicted that the personality traits would act as a confounding factor for regular exercise and sleep habits and Low SMI, but that each personality factor would differ in its characteristics.In other words, conscientiousness was predicted to positively affect regular exercise and sleep habits, extraversion and openness were predicted to positively affect exercise habits, and neuroticism and agreeableness were predicted to negatively affect sleep habits, with the direct effects on Low SMI being partial or minimal [26][27][28] .The estimation method of SEM was performed using the modified weighted least squares method (WLSMV) with missing data, and we referred to the following fit indices: comparative fit index (CFI), Tucker Lewis Index (TLI), and root mean square error of approximation (RMSEA) with 90% confidence interval (CI) 45 .For CFI and TLI, a value of > 0.9 is the best model fit 46 .For RMSEA, a value of < 0.05 indicated a close fit, < 0.08 indicated a reasonable fit, and > 0.1 indicated a poor fit 47 .If the standardized path coefficient between the direct variables was significant and if the standardized path coefficient to the next latent variable was also significant, an indirect effect was also estimated by multiplying the two standardized path coefficients.Furthermore, we checked the goodness of fit of the SEM analysis for the hypothesized model for each sex.HAD17 48 and Mplus version8 45 were used in the analysis.

Characteristics of the participants
The study participants were 101 university students (47 men and 54 women; mean age 20.59 ± 0.66 years, range: 20-24 years).The other descriptive statistics are shown in the Table 1.Two men and four women were categorized into the Low SMI group, representing 5.9% of all participants.In the normality test, the Kolmogorov-Smirnov test confirmed the normality for all variables, except for the agreeableness of personality traits and sleep hours.Additionally, descriptive statistics by sex are shown in the Table 2.

Comparison between the normal and low SMI groups
Comparing the Low SMI group with the normal group (Table 3), BMI (p < 0.01), balance (p = 0.027), and openness tended to be significantly higher in the normal group (p = 0.046).Additionally, the normal group was more likely to be habitual in exercise_past (p = 0.021).The R/X value tended to be suggestively higher (p = 0.074) and the power tended to be suggestively lower (p = 0.081) in the Low SMI group.www.nature.com/scientificreports/

Relationship between the variables
The correlation analysis revealed no relationship between variables with high correlation coefficients (Table 4).

SEM of the hypothesized model
The results of the SEM analysis of the hypothesized model created in this study showed that the fit indices were CFI of 0.935, TLI of 0.860, and RMSEA of 0.048 (90% CI: 0.000-0.097),CFI was the best model fit and RMSEA was a close fit criterion.However, the standardized path coefficient from low SMI to motor function was − 1.259 www.nature.com/scientificreports/(p < 0.05).Given that the standardized path coefficient ranged between − 1 and 1, this standardized path coefficient was considered a result indicating an unsuitable solution (Supplementary information file 2).Therefore, a modified model was created by removing the paths of exercise_past and sleep hours that were not significant among the paths for the motor function as objective variables and was analyzed again with SEM.As a result, the fit indices were CFI of 0.921, TLI of 0.842, and RMSEA of 0.051 (90% CI: 0.000-0.097).CFI was the best model fit, RMSEA was an almost close fit criterion, and no path coefficient indicating an inappropriate solution was found (Table 5).There was a significant effect of Low SMI (direct effect = − 0.539; p < 0.01) and exercise_current (direct effect = 0.410; p < 0.01) on motor function.Only exercise_past (direct effect = − 0.493; p < 0.05) had a significant effect on Low SMI.Regarding the personality traits, a significant effect was observed on openness on exercise_past (direct effect = 0.387; p < 0.05).There was a suggestive effect from agreeableness www.nature.com/scientificreports/ on exercise_past (direct effect = − 0.451; p = 0.075), and from conscientiousness on exercise_current (direct effect = 0.228; p = 0.058) and sleep hours.No personality traits had a direct effect on Low SMI (Table 5).The indirect effect of exercise_past on motor function via Low SMI (indirect effect = 0.266; p = 0.058) was at a suggestive level.The indirect effect from openness to Low SMI via exercise_ past (indirect effect = − 0.191 (p = 0.107)) was not significant.We conducted SEM analysis on separate models for males and females.However, we found that the goodness of fit of the original hypothesis was poor in both cases (Supplementary information files 3 and 4).Specifically, the fit indices of the hypothesized model with male participants were as follows: CFI of 0.789; TLI of 0.545; RMSEA of 0.083 (90% CI: 0.000-0.151).The fit indices of the hypothesized model with female participants were as follows: CFI of 0.665; TLI of 0.279; RMSEA of 0.083 (90% CI: 0.000-0.145).

Discussion
The present study tested a hypothesized model in which significant muscle mass loss affects the lower limb motor function, and exercise and sleep habits and personality traits also affect significant muscle mass loss, in university students aged ≥ 20 years.The results of the examination of the hypothetical model by performing a SEM analysis showed that the modified model in which the paths between variables that were non-significant were removed from the hypothetical model generated in this study showed little change when compared with the original hypothesis, and both models had a level of fit that could be adopted as a criterion for close fit.Therefore, it can be concluded that the modified model was finally adopted in the analysis using the data obtained in this study.The structural relationships in the modified model showed that only exercise habits in high school had a significant effect on significant muscle mass loss.University students who were previously habituated to exercise through participation in athletic clubs during high school have been reported to exhibit higher mean SMI values 49 .These findings underscore the importance of physical activity starting from a young age, as emphasized by WHO 19 , which may have implications not only for the prevention of obesity and hypertension [50][51][52] but also for averting secondary sarcopenia.Additionally, muscle mass loss and lack of current exercise habits had a significant effect on lower limb motor function, given that the detrimental effect of muscle mass loss on motor function may outweigh the impact of lacking current exercise habits.Previous studies examining age-related changes in muscle mass among Japanese individuals have highlighted the importance of lower limb muscle mass.Although upper limb muscles show minimal age-related changes, the decline in lower limbs among Japanese people begins in their 20s, with the most substantial age-related decline in observed in the lower limb muscle mass compared with all other body parts 53,54 .Consequently, even younger individuals who do not currently experience major life-related issues may faces challenges in maintaining mobility as they age, especially due to age-related declines in lower limb muscle mass, despite having exercise habits in the past 49 .No paths were found for the personality traits to directly influence muscle mass loss, and the only significant relationship was the positive influence of openness to experience on exercise habits in high school.The other findings assume that a high level of conscientiousness tends to form the current exercise and sleep habits more, whereas a high level of agreeableness may make it difficult to form exercise habits in high school.This result supports the relationship reported in previous studies 26 , suggesting that conscientiousness may be important in developing adherence behavior in people with a high sense of purpose and self-control, whereas agreeabeleness may result from the fact that individuals are easily influenced by the temptations and enticements found in their surroundings, making it difficult to establish their own regular behavioral patterns.These findings suggest that personality traits have partial effects on the past and current exercise habits, but do not have a strong causal effect on the body components such as muscle mass and fat, or even motor performance.In other words, the lack of regular exercise habits in high school may lead to the current marked decline of muscle mass.Furthermore, the marked decline of muscle mass and the lack of current exercise habits are predicted to cause a decline in motor function of the lower limbs when standing up.Additionally, it was inferred that those with intellectually curious personalities are more likely to form exercise habits through club activities in high school, and previous studies supported the tendency for high openness to experiementation, preferring various activities and engaging in physical activity 27,28 .
In the present study population, approximately 6% of healthy adults in their 20s met the criteria for presarcopenia, confirming a risk associated with a decline in lower limb motor function.Approximately 90% of the participants had a habit of exercising in high school, indicating that the majority had engaged in regular exercise in the past.However, approximately 60% of the participants reported having a current exercise habit, suggesting that approximately 30% may have reduced their exercise habits.This trend aligns with reports indicating that students are more likely to experience a decline in physical activity during their university years than during high school 22 .Regarding sleep duration, in international reports, Japanese people are recognized for sleeping shorter hours, as evidenced by an organization for Economic Co-operation and Development survey showing that the average sleep duration of Japanese individuals was approximately 440 min, the shortest period among the 33 countries surveyed 55 .Additionally, a survey conducted by Japan's Ministry of Health, Labour and Welfare on sleep hours revealed that approximately 70-80% of Japanese individuals in their 20 s sleep for < 7 h 56 .Although a previous study 57 involving Japanese university students reported an average sleep duration of < 7 h, participants in the present study tended to sleep for fewer hours than those reported previously.
The comparison between the Low SMI and normal groups showed significant difference in terms of lower BMI, balance of motor function, openness personality traits, and absence of an exercise habit in high school, suggesting a trend toward higher R/X values and lower power of motor function.BMI and SMI differ only in that they use total body weight or ASM, respectively, in their calculation formulas, resulting in similar trends in the BMI values.However, BMI and SMI have been used in predicting musculoskeletal injuries 58 and determining the prognosis of patients with cancer 59,60 .A higher R/X value indicates a smaller percentage of intracellular muscle fibers 34 .Therefore, muscles density was lower in the group with Low SMI.In a study concerning muscle density and motor function, Goodpaster et al. 61 suggested that a significant reduction in leg strength may be

Figure 1 .
Figure 1.Body composition analyzer with an operating multi-frequency BIA and motor function analyzer.(A) Main body of the device (MC-780A-N, Tanita Co., Ltd., Tokyo).(B) Reach for and grasp the grip handle by oneself while on the measuring table.(C) Remain standing on the measuring table until the measurement is completed.(D) Main body of the device (Zaritz BM-220, Tanita Co., Ltd., Tokyo).(E) Image of the posture at the start of measurement.(F) Image during measurement holding standing position after standing up.

Table 2 .
Descriptive statistics by sex.SD standard deviation, SE standard error.

Table 3 .
Comparison between normal group and low SMI group.SD standard deviation, SE standard error.