【Objective】Objective: To elucidate the causes of falls among terminal cancer patients that occurred during a one - month period.

【Methods】We conducted a retrospective study of medical records and incident reports during a one - month period in 397 patients who died from cancer at a Komatsu Municipal Hospital between January 1, 2013 and December 31, 2014. The following were investigated: their basic attributes and psychosomatic states; treatments and procedures they received; medications; Barthel Index; the circumstances of falls; and the level of injury. The relationships between these parameters and falls were analyzed.

【Results】In the falling group (80 patients［20.2%］, 100 falls), odds ratios［ORs］were 4.092 (95% confidence interval［CI］: 2.296 - 7.294) for cognitive impairment, 2.281 (95%CI: 1.287 - 4.045) for lower extremity weakness, and 1.963 (95%CI: 1.134 - 3.396) for sleep-medication use. Among the 80 patients, 17 (21.2%，all aged ≥65years) who fell multiple times (37 falls) were included in the multiple-falling group. Large ORs were observed in the multiple-falling group: 3.575 (95%CI: 1.050-12.168, p = 0.041) for metastasis, 2.849 (95%CI: 1.045 - 7.766, p = 0.019) for cognitive impairment, 3.809 (95%CI: 1.342-10.81, p = 0.011) for sleep-medication use, and 3.348 (95%CI: 1.126-9.953, p =0.025) for poor appetite. The most common location of falls was at bedside (66.0%), and 59.0% of falls occurred when they went to the bathroom. The rate of falls at night was 52.0%, and April was the month when they fell most frequently (16.0%).

【Conclusions】Factors associated with falls among terminal cancer patients that occurred during a one-month period were cognitive impairment, lower extremity weakness, and sleep-medication use, and factors associated with falls among patients with multiple falls were metastasis, cognitive impairment, sleep-medication use, and poor appetite.

【Objective】The plantar surface of the foot is the only area that makes contact with the ground and is responsible for supporting gait movement and posture control. This study examines the relationship between plantar pressure while walking and a history of falling.

【Methods】Experiments involving 10-meter walks were conducted with 30 females (since females are more prone to falling than males) using the plantar pressure system (F-scan II). Eight participants (average age: 61.88 ± 8.29 years) had a history of falling and 22 (average age: 59.64 ± 8.89 years) had no such history. We analyzed the entire waveform of the plantar pressure while walking by using principle component analysis (PCA). This was conducted using a 300 × 357 matrix of time-normalized plantar data for seven areas (five gait trials of 30 participants × 2 ((datasets for both feet)) × 50 equal parts × 7 plantar areas). T-tests between 2 groups were conducted on the principal component scores of the principal component vectors (PCVs) 1 ‒ 36.

【Results】The PCA clarified that the 1st, 8th, and 10th PCVs were related to fall history (p ＜ 0.05). These PCVs indicated a significant relationship between plantar pressure and history of falling. For the group with a history of falling, reduced plantar pressure on the heel and toe areas was seen in the reconstructed waveform based on the PVCs.

【Conclusions】Our results indicated the possibility of differences in foot contact strategy during walking between the group with no history of falling and the group with such a history. This study highlights not only singularity parameters such as peak pressure values but also the characteristics of the entire waveform in plantar pressure. These findings indicate a possibility of identifying those people with a higher risk of falling.