Background: The elderly population is increasing rapidly worldwide. Muscle mass, usual walking speed (UWS), knee extension strength (KES), hand grip strength (HGS), peak expiratory flow (PEF), and depression is used for sarcopenia diagnosis. All four of these factors (KES, HGS, PEF, and depression) correlated with UWS and also to muscle mass. But, many studies have suggested that no correlation exists between muscle mass and UWS.
Objects: This study aimed: 1) to investigate whether muscle mass reduction affected UWS, as mediated by KES, HGS, PEF and depression, and 2) to explored whether significant changes in these mediators varied by the body segment in which muscle mass evaluated in elderly female aged 65–80 years.
Methods: A total of 100 female aged 65–80 years were surveyed. Muscle mass was measured by body segment (upper and lower segment), and KES, HGS, PEF, depression, and UWS were also assessed. Median analyses were progressed in IBM SPSS software (ver. 23.0, IBM Co.) using a downloaded INDIRECT macro.
Results: The direct effect of the KES and PEF were significant, and the indirect effect of KES and PEF were not significant. Thus, KES and PEF served as full mediators of the effect of muscle mass on UWS. Regardless of bodily region, KES and PEF combined with muscle mass were significant mediators of UWS, with similar indirect effect sizes.
Conclusion: KES and PEF are the only mediators regardless of body part. Therefore, mediating the KES and PEF may prevent sarcopenia progression in elderly female. Also, sarcopenia can be readily assessed by evaluating either the upper or lower body; it is not necessary to measure total muscle mass.

INTRODUCTION
MATERIALS AND METHODS
    1. Subjects
    2. Outcome Measures
    3. Statistical Analysis
RESULTS
    1. Results of Pearson Correlations BetweenIndependent Variables
    2. Relationship Between ULMM and UWS
    3. Relationship Between UMM and UWS
    4. Relationship Between LMM and UWS
DISCUSSION
CONCLUSIONS
FUNDING
ACKNOWLEDGEMENTS
CONFLICTS OF INTEREST
AUTHOR CONTRIBUTIONS
ORCID
REFERENCES

• Yun-jeong Baek(Department of Rehabilitation Medicine, Seoul National University Bundang Hospital, Seongnam)
• Chung-hwi Yi(Department of Physical Therapy, College of Health Sciences, Yonsei University, Wonju, Korea) Corresponding author
• Oh-yun Kwon(Department of Physical Therapy, College of Health Sciences, Yonsei University, Wonju, Korea)
• Sang-hyun Cho(Department of Physical Therapy, College of Health Sciences, Yonsei University, Wonju, Korea)