Although the physical activity reference value for older adults (10 METs*hour/week) has been promoted by Japan Ministry of Health, Welfare, and Labour since 2013, little is known about how many steps/day cut-off values that optimally identify meeting the reference value according to the differences of age, sex, medical history, and joint pain. The purpose of this study were 1) to determine the steps/day that optimally identify meeting the reference value, and 2) to identify the differences by the effects of age, sex, medical history, and joint pain on cut-off values. This study included 583 community-dwelling older Japanese adults (aged 73.2 ± 5.4 years; 153 men, 430 women). A uniaxial accelerometer survey was conducted to estimate the total physical activity volume and steps/day. Receiver operating characteristic (ROC) analyses were used to detect steps/day cut-off values for meeting the reference value among all participants and stratified by age, sex, medical history, and joint pain. The optimal cut-off value (AUC (area under the ROC curve), sensitivity, and specificity) for the reference value was 4376 steps/day (0.99, 95.2%, and 97.2%) in all participants. Differences among the cut-off values according to age, sex, medical history, and joint pain ranged from 39 to 169 steps/day. These results suggest that step counts has satisfactory validity to represent the reference value in older adults, and the effects of age, sex, medical history, and joint pain on cut-off values were considerably small. Therefore, this step-count level may be a useful indicator for modifying the daily-life activities of older adults.

Habitual moderate- to vigorous-intensity physical activity positively impacts lower extremely performance in older adults. However, little is known whether habitual sedentary behaviour time independently impacts lower extremity performance. The purpose of this cross-sectional study was to identify whether sedentary behaviour time is associated with lower extremity performance independent of moderate- to vigorous-intensity physical activity time in older adults. Eight-hundred-and-two community-dwelling older Japanese adults (72.5 ± 5.9 years) participated in this study. Sedentary behavioir time and moderate- to vigorous-intensity physical activity time were assessed using a uniaxial accelerometer. Lower extremity performance was assessed by repeated chair stand, single leg stance, and timed up and go tests. Jonckheere terpstra trend test and Spearman rank correlation coefficient were used to identify the single relation between sedentary behaviour time and lower extremity performance. A hierarchical multiple regression analysis was used to identify whether sedentary behaviour time is associated with lower extremity performance independent of moderate- to vigorous-intensity physical activity time. Results of single relation analysis showed there were negative associations between sedentary behaviour time and all lower extremity performance tests. Multivariable analysis indicated that sedentary behaviour time was not significantly associated with any lower extremity performances but moderate- to vigorous-intensity physical activity time was positively associated with all lower extremity performance tests. In conclusion, sedentary behaviour time would be less useful than moderate- to vigorous-intensity physical activity time as an indicator for modifying habitual behavior to maintain good lower extremity performance in community-dwelling older adults.

The purpose of this study was to retrospectively examine the association of habitual exercise with “single fall (= 1)”, “multiple falls (≥ 2)”, and “injurious falls (≥ 1)” among community-dwelling older adults. A total of 1,683 community-dwelling older adults, aged 60-97 years (72.6 +/- 6.6 yr, 512 men and 1,171 women) were included in this study. Habitual exercises continued one year or longer (6.4 +/- 9.5 yr) were classified into twelve types. Exercise components (time, quantity, period of continuity, and number of exercises) were divided with median or tertiles. To assess the association between habitual exercises and fall status, multivariable logistic regression analyses with stepwise selection method, were applied. The multivariable logistic regression analyses showed that dance (odds ratio (OR): 0.30, 95% confidence interval (CI): 0.09-0.96) was negatively associated with “single fall”. Bicycling (OR: 3.72, 95% CI: 1.32-10.77) was positively associated with “multiple falls”, and the period of continuity (OR: 0.74, 95% CI: 0.60-0.91) was negatively associated with “multiple falls”. None of the exercise components were selected with regard to “injurious falls”. Results indicate that dance may be an effective type of exercise for fall prevention among community-dwelling older adults. However, caution about falling is warranted toward bicycling as an exercise. Moreover, a longer period of continuity (≥ 4 years) appears to be a positive factor of habitual exercise for fall prevention.