Purpose The study purpose was to establish the number (and type) of days needed to estimate mean pedometer-determined steps/day among Japanese adults.
Methods During June and August of 2008, seven days of data were collected from 223 participants in S Center between 20 and 64 years old and living in Shizuoka Prefecture (63 males, age=50.3±10.4 years, BMI=25.8±3.7 kg/m
2; 160 females, age=49.0±10.6 years, BMI=23.5±3.8 kg/m
2)). Mean steps/day were computed for all 7 days, each single day, and combinations of days. Analyses used in this study were repeated measures analysis of variance, intra-class correlations (ICC) and regression analyses.
Results The study participants averaged 8,854±3,356 steps/day over the 7-day monitoring frame. There were no gender differences in steps/day for any days of the week or for the weekly average of steps/day. There was a significant difference (F=3.91, df=5.21,
P<0.001) between days for the total sample, Tuesday and Friday showing significantly more steps than Saturday and Sunday (
P<0.001). ICC analyses found a minimum of 3 days is necessary to achieve a reliability of 0.80. The adjusted R2 of 35 models for 3 days were all 0.80 or more.
Conclusion This study suggests any 3 days at random can provide a sufficient estimate tor predict weekly physical activity among Japanese adults. Further studies on sample size and survey period are now required.
View full abstract