A feasibility study to prevent falls in older people who are sight impaired: the VIP2UK randomised controlled trial

Table 3 Adherence to exercise as measured by exercise calendars (self-report), Phone-FITT (self-report) and activPAL (activity monitors)

Mean (SD)	Control (n = 13)	Home safety only (n = 15)	Home safety + Home exercise (n = 15)
Exercise calendar	NA	NA	86.7 % (13/15)
Phone-FITT (TSS_FDI)
Baseline	36.86 (15.16)	35.23 (8.04)	31.66 (13.16)
3-month	47.93 (14.94)	35.63 (11.30)	42.56 (13.62)
6-month	51.05 (25.72)	47.37 (18.64)	51.97 (22.27)
GLM tests^c for (i) time p < 0.001, (ii) group-by-time p = 0.2, (iii) group p = 0.3
activPAL^a mean (SD) [median]
Walking time (min)
Baseline	127.1 (81.4) [96]	124.7 (101.2) [120]	95.0 (78.1) [69]
6-month follow-up	68.5 (22.9) [71]	70.2 (44.5) [59]	55.0 (24.8) [50]
GLM tests^b,c for (i) time p < 0.001, (i) group-by-time p = 0.8, (iii) group p = 0.4
Step count
Baseline	10,103 (7542) [6254]	10,339 (8797) [8956]	7426 (6588) [5110]
6-month follow-up	5000 (2192) [4962]	5321 (3892) [4128]	3927 (1815) [3446]
GLM tests^b,d for (i) time p < 0.001, (ii) group-by-time p = 0.8, (iii) group p = 0.5

It also measures these in relation to frequency, duration and intensity. Scores are summed to give a total. A higher score suggests a higher activity rate.
^aMedian also presented for skewed data
^bBased on natural log data at three time-points
^cSphericity assumed
^dGreenhouse-Geisser statistic Phone-FITT, measures physical activity in two main areas; recreational and household
GLM general linear model