Table 5 Data extraction fallers/ non-fallers

Auvinet et al., 2017 [74]

walking+ arithmetic (counting backwards from 50 subtracting serial 1 s)

auditory -verbal

3-D-acceleration sensor, Locometrix (electronic walkway), stopwatch, 30 (m) distance

walking speed (m/s)

stride frequency (Hz) and the stride regularity (dimensionless)

1. Non-Fallers: walking speed (m/s) ↓, stride frequency (Hz) ↓ and the stride regularity (dimensionless) ↓

2. Fallers: walking speed (m/s) ↓, stride frequency (Hz) ↓ and the stride regularity (dimensionless) ↓

Walking speed and stride regularity differed between subgroups under ST and DT (p < 0.01 and 0.05, respectively, for both conditions).

Beauchet et al., 2008 [19, 96]

walking+ arithmetic (counting backward (counting backwards from 50 subtracting serial 1 s)

auditory -verbal

Stopwatch, 10 (m)

walking speed (cm/s), cadence (steps/min)

1. Non-Fallers (0 falls): walking speed (cm/s) ↓, cadence (steps/min) ↓

2. Non-multiple Fallers (0 or 1 fall): walking speed (cm/s) ↓, cadence (steps/min) ↓

3. Fallers (1 or more falls): walking speed (cm/s) ↓, cadence (steps/min) ↓

4. Multiple Fallers (2 or more falls (walking speed (cm/s) ↓, cadence (steps/min) ↓

Single and recurrent fallers walked more slowly than non-fallers under ST and DT conditions

Bauer et al. 2010 [75]

walking+ arithmetic (counting backwards from 50 subtracting serial 1 s)

walking + arithmetic (counting backwards from 100 subtracting serial 3 s)

walking + verbal fluency (naming animals starting with a specific letter)

auditory verbal

GAITRite system

walking speed (cm/s)

Fallers walked slower than non-fallers under ST and DT conditions.

Both non-fallers and fallers showed comparable reduced DT performance for the task conditions. The verbal fluency task had a higher amount of DCT in comparison to the arithmetic task (cf. Fig. 5)

Bootsma et al., 2003 [76]

walking + verbal fluency (reciting names of animals or professions during a 30-s period)

auditory -verbal

Stopwatch, 12 (m)

walking time (s)

Number of steps (n), Number of complete stops (n)

1. Non-Fallers (0 falls): walking time (s) ↑, Number of steps (n) ↑

2. non-multiple Faller (0 or 1 fall): walking time (s) ↑, Number of steps (n) ↑

3. Fallers (1 or more falls): walking time (s) ↑, Number of steps (n) ↑

Freire Junior et al., 2017 [21]

1. walking + word task (naming animals starting with a specific letter)

2. walking + motor task (transferring a coin from one pocket to another)

1. auditory - verbal

2. manual

GAITRite, 8(m), 25–36 steps were collected to examine variability

gait speed (m/s), cadence (steps/min), step length (cm) stride time (s), single support time (as percentage of the gait cycle), and stride time variability (CoV (%))

1. Non-Fallers (0 falls) and Fallers (1 or more falls) + word task: gait speed (m/s) ↓, cadence (steps/min) ↓, stride time (s) ↑, step length (cm) ↓, single support time ↓, and stride time variability (CoV (%))↑

2. Non-Fallers (0 falls) and Fallers (1 or more falls) + motor task: gait speed (m/s) ↓, cadence (steps/min) ↓, stride time (s) ↑, step length (cm) ↓, single support time (↓, and stride time variability (CoV (%)) ↑

There were no significant main effects of group and interaction effects between group and task.

Howcroft et al., 2016 [77]

walking + word fluency task (recite words starting with A, F or S)

auditory – verbal

Stopwatch, Pressure-sensing insoles (F-Scan 3000E, Tekscan, Boston, MA), Accelerometers (X16-1C, Gulf Coast Data Concepts, Waveland, MS), 7,62 (m)

cadence (steps/min), double support time (%), speed (m/s) CoP path (s), Min CoP Vel (m/s), Mean Cop Vel (m/s), Median CoP Vel (m/s), Cadence (steps/min), stride time (s), stance time (s), swing time (s), stride time CoV, percent stance time (%), stride time symmetry index, Impulse (Foot-strike to first peak (Ns/kg), Min to second peak (Ns/kg), Second peak to foot-off (Ns/kg), Foot strike (Ns/kg)

1. Non-Fallers (0 falls): cadence (steps/min) ↓, Percent double support time (%) ↓, gait speed (m/s) ↓

2. Non-multiple fallers (0 or one falls): cadence (steps/min) ↓, Percent double support time (%) ➔ no data, gait speed (m/s) ↓

3. Fallers (1 or more falls): cadence (steps/min) ↓, Percent double support time (%) ↓, gait speed (m/s) ↓

4. Multiple Fallers (2 or more falls): cadence (steps/min) ↓, Percent double support time (%) ➔ no data, gait speed (m/s) ↓

Accelerometer: Differences were found between fallers and non-fallers for the head and posterior pelvis accelerometers.

Johansson et al., 2016 [78]

walking+ arithmetic task (counting backwards from 10 by subtracting serial 1 s)

auditory – verbal

GAITRite, 8.6 (m)

Double support time CV(%), gait speed (m/s) at baseline (normal speed) -- > All gait parameters present as Coefficient of Variations (CVs; SD/M × 100)

For Non-Fallers (0 falls) and Fallers (1 or more falls): step width CV (%) ↑, stride width CV (%) ↑, step length CV (%) ↑, stride length CV (%) ↑, step time CV (%) ↑, stance time CV (%) ↑, stride time CV (%) ↑, stride velocity CV (%) ↑, swing time CV(%) ↑, DST CV(%) ↑

Mirelman et al., 2012 [79]

walking + arithmetic task (counting backwards from a predefined 3 digit number by subtracting serial 3 s)

auditory - verbal

force-sensitive insoles, stopwatch, 25 (m)

gait speed (m/s)

gait variability (%)

1. Non-fallers (0 falls), Non-multiple fallers (1 fall), Multiple fallers (2 or more falls): gait speed (m/s) ↓

Muhaidat et al., 2013 [20]

1. walking + arithmetic task (counting backwards by subtracting serial 3 s/ 7 s)

2. walking + Stroop task (identify incongruency between the words high/low in different pitches)

3. walking + word task (generating words starting with I, N, O or naming animals)

4. walking + visuospatial task (clock task)

5. walking + motor task (carrying a cup)

1. auditory-verbal

2. auditory -verbal

3. auditory -verbal

4. visual-spatial (manual)

5. manual

stopwatch, 10 (m)

Gait speed (m/s)

Gait speed (m/s) ↓ for all subgroup form ST to DT

A trend of a difference in complexity of secondary and absolute values of task performance was shown between community-dwelling fallers and non-fallers (p ≤ 0.05).

Nordin et al., 2010 [80]

1. walking + motor task (carrying a cup)

2. walking + motor task (carrying a tray)

3. walking + word/verbal fluency task (animals naming)

4. walking + arithmetic task (counting backwards from 50 by subtracting serial 3 s)

1. manual

2. auditory -verbal

3. auditory -verbal

GAITRite, 10 (m)

walking speed (m/s), step length (cm), double support time (ms)

step width (mm), step time (ms)

1a. Non-fallers (0 falls) and Fallers (1 or more falls) + motor task (cup): walking speed (m/s) ↑

1b. Non-fallers (0 falls) and Fallers (1 or more falls) + motor task (tray): walking speed (m/s) ↑

1c. Non-fallers (0 falls) and Fallers (1 or more falls) + word task: walking speed (m/s) ↓

1d. Non-fallers (0 falls) and Fallers (1 or more falls) + arithmetic task: walking speed (m/s) ↓

Reelick et al., 2011 [81]

1. walking + arithmetic task (counting backwards from 100 by subtracting serial 7 s)

2. walking + word task (naming words starting with a given letter)

1. auditory -verbal

2. auditory -verbal

GAITRite, 6.1 (m)

gait velocity (cm/s), stride length (cm)

Number of strides (n), stride length CV (%), stride time (cm), stride time CV, stride width (cm), stride with CV (%)

1. Non-multiple Fallers (0 or 1 fall) and Multiple Fallers (zer0o or 1 fall) + arithmetic task: gait velocity (cm/s) ↓, Number of strides (n) ↑, stride length (cm) ↓, stride length CV (%) ↑

2. Non-multiple Fallers (0 or 1 fall) and Multiple Fallers (0 or 1 fall) + word task: gait velocity (cm/s) ↓, Number of strides (n) ↑, stride length (cm) ↓, stride length CV (%) ↑

Stride-length CV was higher when participants performed a DT, and higher in recurrent fallers compared with non-recurrent fallers, although this difference was only significant during performance of the verbal fluency task

Springer et al., 2006 [82]

1. walking + listening

2. walking + listening plus answering questions3. walking + arithmetic task (counting backwards from 500 by subtracting serials 7 s)

1. auditory -verbal2. auditory -verbal

force-sensitive insoles, 25 (m)

gait speed (m/sec) ➔ Gait speed was normalized with height, average swing time (%)

swing time variability CV (%)

In swing time variability fallers and elderly non-fallers differed (p = 0.003)

Toulotte et al., 2006a [83]

walking + motor task (carrying a glass)

manual

VICON 370 system (Oxford Metrics), Three AMTI force plates (250 Hz), 10 (m)

cadence (steps/min), walking speed (m/s), stride length (m)

stride time (s), step time (s), single-support time (s)

1. Non-Fallers (0 Falls): cadence (steps/min) ↓, walking speed (m/s) →, stride length (m) ↑, stride time (s), single-support time (s)

2. Fallers (1 or more falls): cadence (steps/min) ↓, walking speed (m/s) ↓, stride length (m) ↓

Toulotte et al., 2006b [84]

walking + motor task (carrying a glass)

manual

VICON 370 system (Oxford Metrics), Three AMTI force plates (250 Hz), 10 (m)

cadence (steps/min), walking speed (m/s), stride length (cm), step length (cm), single support time (s)

1. Non-Fallers (0 Falls): cadence (steps/min) ↓, walking speed (m/s) ↓, stride length (cm) ↓, step length (cm) ↑, stride time (s) ↑, step time (s) ↓, single-support time (s) ↓

2. Fallers (1 or more falls): cadence (steps/min) ↓, walking speed (m/s) ↓, stride length (cm) ↓, step length (cm) ↓, stride time (s) ↑, step time (s) ↓, single-support time (s) ↑

Significant difference (P < 0.05) between the fallers and non-fallers under DT-conditions for cadence, walking speed, stride time, step time and single-support time.

Verghese et al., 2017 [85]

walking + word task (reciting alternate letters of the alphabet)

auditory -verbal

electronic walkway (Zenometrics LLC, Peekskill, NY), 14 (ft)

cadence (steps/min), walking speed (m/s), stride length (cm), step length (cm), double support time (%) stride time (s), step time (s)

1. Non-fallers (0 falls) and Non-multiple fallers (0 or 1 fall) Fallers (1 or more falls), Multiple fallers (2 or more falls):cadence (steps/min) ↓, walking speed (m/s) ↓, stride length (cm) ↓, step length (cm) ↓, stride time (s), step time (s), double support time (%) ↑

Asai et al., 2014 [87]

1. walking + arithmetic task (counting backwards from 100 by subtracting serials 1 s)

2. walking + motor task (carrying a ball on a tray)

1. auditory- verbal

2. auditory-

motor

triaxial accelerometer,

20(m)

walking speed (m/s)

STV (%), RMS in the ML direction (m/s²), RMS in the AP direction (m/s²), Standardized RMS in the ML direction (%), Standardized RMS in the AP direction (%)

1. no-FOF and FOF and arithmetic task: walking speed (m/s) ↓, STV (%) ↑, Standardized RMS in the ML direction (%) ↑, Standardized RMS in the AP direction (%) ↑

1b.no-FOF and FOF and motor task: walking speed (m/s) ↓, STV (%) ↑, Standardized RMS in the ML direction (%) ↓, Standardized RMS in the AP direction (%) ↓

Subjects with FoF walked slower during cognitive-task gait than subjects without FoF and walked with greater STV during single-task gait than subjects without.

Donoghue et al., 2013 [88]

walking + verbal fluency task (recite alternate letters of the alphabet (A-C-E, etc.))

auditory- verbal

GAITRite, 4.88 (m)

gait speed (m/s), stride length (m), Double support phase (%),

stride length CV(%), stride time CV (%), step width (cm)

1. no-FOF, FOF-NAR, FOF-AR: gait speed (m/s) ↓, stride length (m) ↓, Double support phase (%) ↑, stride length CV (%) ↑, stride time CV (%) ↑, step width (cm)↑

FOF-NAR and FOF-AR groups significantly different to no-FOF group in gait speed (p > 0.001), stride length (p > 0.001), an DSP (p > 0.001).

Reelick et al., 2009 [89]

1. walking +arithmetic task (counting backwards by subtracting serial 7 s)

2. walking +verbal fluency task

1. auditory- verbal

2. auditory-

motor

GAITRite, Balance during walking(SwayStar), 10 (m)

gait velocity (cm/s)

Stride-length variability (% CV), Stride-time variability (% CV), Mediolateral angular displacement (deg.), Mediolateral angular velocity (deg./s

1. no-FOF and FOF and arithmetic task: gait velocity (cm/s) ↓, Stride-length variability (% CV) ↑, Stride-time variabilit (% CV) ↑, Mediolateral angular displacement (deg.) ↑, Mediolateral angular velocity (deg./s) ↑

2. no-FOF and FOF and verbal fluency task: gait velocity (cm/s) ↓, Stride-length variability (% CV) ↑, Stride-time variability (% CV) ↑, Mediolateral angular displacement (deg.) ↑, Mediolateral angular velocity (deg./s) ↑

Significantly lower gait velocity for walking at the preferred velocity and during the performance of both dual tasks in the FoF group compared to the no-FoF group. Stride-length and stride time variability were significantly higher in the FoF group dur.ing DT Stride-time variability was also significantly higher in the FoF group when walking at the preferred gait velocity and while performing the arithmetic task.

Wollesen et al., 2017 [90]

walking + visual Stroop task (recite colours, not words)

auditory- verbal

treadmill (h/p/cosmos, Zebris; Isny, Germany: FDM-T)

step length (cm)

step width (cm), and gait line (mm)

Intervention with FES-I < 20: and FES-I > 20 step length (cm) ↑, step width (cm) ↓, and gait line (mm) ↑

Control group with FES-I < 20: step length (cm) ↓, step width (cm) ↑, and gait line (mm) ↑

Control group with FES-I > 20: step length (cm) ↑, step width (cm) ↑, and gait line (mm) ↑: