Skip to main content
. 2020 Mar 2;20:83. doi: 10.1186/s12877-020-1484-5

Table 2.

Results of dual-task walking tests across intervention groups and time points

EXCOG EX COG CON
baseline 6-week 12-week baseline 6-week 12-week baseline 6-week 12-week baseline 6-week 12-week p-val
Cognitive performance
 n 20 16 12 19 17 17 21 14 10 14 13 12
 number correct, mean (SD) 17.8 (6.83) 21.62 (9.69) 26.5 (8.48) 19.58 (13.55) 21.7 (13.83) 21.94 (13.64) 16.82 (6.71) 20.5 (7.78) 21.8 (8.52) 14.79 (9.49) 14.46 (9.4) 17.58 (10.15) 0.039
Gait variables (Dual-task cost, %)
 n 15 11 9 12 11 11 18 9 7 12 11 10
 Stride Length −4.36 (2.8) −1.87 (3.02) −2.35 (1.33) −4.26 (2.78) − 1.87 (4.54) − 1.85 (2.14) −5.67 (2.27) −3.96 (3.91) −3.23 (1.35) −5.37 (5.21) −5.07 (3.38) − 4.73 (4.59) 0.790
 Stride Length Variability 0.16 (0.4) −0.02 (0.09) 0.02 (0.2) 0.09 (0.29) −0.07 (0.32) −0.06 (0.19) 0.19 (0.26) 0 (0.22) 0.1 (0.39) 0.1 (0.31) 0.08 (0.16) 0.06 (0.3) 0.891
 Stride Duration 4.94 (6.54) 5.35 (7.35) 5.14 (7.5) 8.74 (5.89) 7.89 (6.63) 3.83 (4.75) 8.42 (10.32) 3.88 (3.1) 2.4 (3.22) 10.48 (13.32) 7.48 (9.09) 6.93 (8.98) 0.773
 Stride Duration Variability 45 (68.97) 34.32 (72.94) 8.57 (30.82) 43.28 (45.6) 27.16 (54.45) 19.12 (25.56) 40.50 (31.99) 4.68 (18.14) 35.92 (49.66) 40 (46.4) 40.29 (48.2) 37.55 (44.38) 0.439
 Stride Velocity −8.45 (6.86) −6.36 (7.39) −6.73 (6.6) −11.6 (5.75) −8.36 (6.95) −5.25 (4.62) − 12.3 (6.91) −7.42 (5.67) − 5.61 (3.48) −12.86 (12.89) − 10.8 (9.94) −10.06 (10.69) 0.925
 Stride Velocity Variability 17.91 (21.46) 5.53 (17.77) 0.58 (20.09) 17.96 (34.44) 1.17 (32.26) 0.65 (15.74) 22.92 (22.23) 4.9 (21.78) 4.52 (18.27) 15.82 (33.94) 14.47 (19.1) 13.94 (23.05) 0.739

Note: Data presented are mean (SD); variability is the coefficient of variation for a given variable. Number correct were determined during the dual-task condition only. Gait variables are the dual task cost calculated from both dual and single task trials (see text for further explanation). P-values are the interaction of group and time point for each variable from linear mixed effects models