e334 Y.G. Yi et al. / Annals of Physical and Rehabilitation Medicine 61S (2018) e309–e433 equipped with a balloon probe. Tongue volume was evaluated with T1-weighted MR imaging. Tongue lean muscle mass was evaluated with Dixon MRI. The Pearson analysis was used to calculate coef- ficients between tongue pressure and tongue volume/lean tongue muscle mass. Results The mean tongue pressure, tongue volume and lean tongue muscle mass were 29.9 ± 8.1 kPa, 66.8 ± 12.3 cm 3 and 56.2 ± 10.4 cm 3 . Tongue pressure was significantly correlated with tongue volume and lean tongue muscle mass (r = 0.60, P = 0.049 and r = 0.64, P = 0.035). Conclusion To our knowledge, this is the first study to investigate tongue muscle mass, lean tongue muscle mass and tongue pressure using MRI images. Tongue muscle mass and lean tongue muscle mass was significantly associated with tongue pressure. Compared with previous studies for healthy young adults (Utanohara, 2008; Fabrice, 2010), tongue pressure was decrease in spite of tongue volume was similar. Bassler (1987) suggested that the tongue volume after muscular atrophy is compensated by metaplasia of fatty tissue. We speculated that adipose might be an important factor for reducing tongue pressure and dysphagia. Keywords Sarcopenia of tongue; MRI Disclosure of interest The authors declare that they have no com- peting interest. https://doi.org/10.1016/j.rehab.2018.05.780 ISPR8-2154 The relationship between sleep duration and falls, muscle mass and mortality: A cohort study in Chinese, elderly population L. Fu 1,2,∗ , Q. Guo 1,2 1 TEDA International Cardiovascular Hospital, Cardiovascular Clinical College of Ti, Department of Rehabilitation Medicine, Tianjin, China 2 Tianjin Medical University, Department of Rehabilitation Medicine, Tianjin, China ∗ Corresponding author. E-mail address: 121439728@qq.com (L. Fu) Introduction/Background Epidemiological studies report that more than half of people over the age of 65 years suffer from vari- able sleep problems. In this study, we conducted a cohort study to investigate the effects of sleep duration on muscle mass and func- tion within a Chinese, community-dwelling elderly population. Material and method Our study population consisted of residents living in the township central hospital of suburban Tianjin, China. We measured muscle strength and walk speed. We divided sleep duration into the following four groups: < 7 h, 7–8 h, > 8–9 h, > 9 h. Results A total of 902 participants completed the 3-year follow- up. We observed a U-shaped relationship between sleep duration and fall risk. Compared to the 7–8 h group, the fall risk within the < 7 h group was 3.58 (1.99, 5.24) times higher, and the fall risk within the > 9 h group was 2.23 (1.20, 3.13) times higher. After adjustment, muscle mass declined by −6.82% (−11.27%, −3.83%) in the < 7 h group. The > 9 h group exhibited a mortality risk that was 3.46 (1.37, 6.23) times higher the normal sleep group (7–8 h). Conclusion In summary, we observed a U-shaped relationship between sleep duration and falls. Short sleep duration have neg- ative effect on muscle mass decline, and long sleepers have higher mortality within a Chinese, community-dwelling, elderly popula- tion. Keywords Muscle mass; Muscle function; Sleep Disclosure of interest The authors declare that they have no com- peting interest. https://doi.org/10.1016/j.rehab.2018.05.781 A6.04 Geriatrics – Risk of falls in the elderly ISPR8-1126 Can side predict the fall after discharge for the patients after surgery for proximal femoral fracture? I. Kondo ∗ , K. Ito , E. Takano , N. Morioka , K. Sato , Y. Mizuno , N. Itoh National Center for Geriatrics and Gerontology, Rehabilitation medicine, Obu, Japan ∗ Corresponding author. E-mail address: ik7710@ncgg.go.jp (I. Kondo) Introduction/Background Standing balance test for imbalance and disequilibrium (SIDE) was developed as a predictive measure for the fall prevention of the patients in hospital. Excellent predictive validity of SIDE was defined with the patients who admitted to reminiscence rehabilitation ward. On the other hand, the preven- tion of fall after discharge for the patients with fracture is essential to avoid recurrence of fracture. In this study, we use SIDE to predict fall event after the discharge for the patients after the surgery for proximal femoral fracture. Material and method Participants of this study were the patients who discharged and returned to home from reminiscence rehabil- itation ward after the surgery for proximal femoral fracture during the period from April, 2014 to September 2016. They were 10 male and 35 female patients and their average age was 78.3 (SD: 8.9). We sent questionnaire about the fall event and recurrence of fracture. Multiple logistic regression analysis was performed with age, dura- tion of admission, the result of Mini Mental State Examination, FIM, SIDE and comfortable gait velocity at discharge as the independent factor according to the fall event occurrence. Results The fall event was occurred for seven patients (15.6%). Two of them (4.4%) experienced recurrence of fracture at femoral neck and vertebral body. SIDE (P = 0.049) and comfortable gait velocity (P = 0.027) had the significant effect on fall event and their odds ratio and 90% confidential interval were 0.32 (0.11 − 0.99) and 1119.47 (2.18 − 574866.87), respectively. Conclusion In addition to comfortable gait velocity, SIDE was the significant factor to predict the fall event for the one year after discharge. Cutting point of comfortable gait velocity would be war- ranted to use it as predictive measure. Discreet selection of walking aids and modification of home environment are necessary for the patients with high fall risk detected by SIDE. Keywords Fall prevention; Predictive measure; SIDE Disclosure of interest The authors declare that they have no com- peting interest. https://doi.org/10.1016/j.rehab.2018.05.782 ISPR8-1717 Posturographic detection index of fear of falling in elderly – A limits of stability study – Pilot study J. Michalska ∗ , A. Kamieniarz , M. Pawłowski , W. Marszałek , M. Sowa , G. Juras , K. Słomka The J. Kukuczka Academy of Physical Education, Human Motor Behavior, Katowice, Poland ∗ Corresponding author. E-mail address: j.michalska@awf.katowice.pl (J. Michalska) Introduction/Background Elderly people often use a smaller per- centage of their base of support during maximal weight shifting, which leads to balance instability, that can lead to frequent falls and serious injuries. Even the most active people cannot reach the- oretical limits of stability, therefore the need to investigate the