Background
Methods
Data sources and search strategy
Eligibility criteria
Study selection
Data extraction
Quality assessment
Data synthesis and analysis
Results
Study characteristics
Study (first author and year) | Study Design | Study Characteristics (groups, number of participants, mean age) | MCI Diagnostic Criteria | Instrumented Functional Assessment | Instrument | Intervention | Data Collection (follow-up) | Main results in MCI |
---|---|---|---|---|---|---|---|---|
Doi et al. [56], 2013. | RCT. | Intervention Group: n = 25. 75.3 years old. Control Group: n = 25. 76.8 years old. | Petersen Criteria [11]. MMSE > 24 [49]. | Walking at preferred speed (11 m walkway). | Tri-axial accelerometer attached to the L3 spinous. | Intervention Group: Aerobic exercise (60% of aged predicted maximal heart rate), endurance walking, muscle strength training, postural balance retraining, and gait training (90 min, 2/ week, 6 months). Control Group: 2 Education classes about health promotion. | (T1) at baseline; (T2) six months. | ↑Gait speed ↓Stride time and ↑Stride length in both groups***. ↑HR in VT in the Intervention group***. ↑Gait speed, stride length and HR in VT in the Intervention group vs control group**. |
Donnezan et al. [57], 2018. | RCT. | PCT: n = 21. 75.2 years old. PT: n = 18. 77.1 years old. CT: n = 16. 76.3 years old. Control Group: n = 14. 79.2 years old. | Petersen Criteria [11]. | Walking speed at usual pace (6 m) in ST and DT conditions. WSC. TUG. | Electronic walkway GAITRite® (length: 4.3 m). | PT: Aerobic training on bikes (60% of aged predicted maximal heart rate). CT: Cognitive exercises (commercialized gaming software). PCT: Aerobic training on bikes (60% of aged predicted maximal heart rate) + cognitive exercises. Control Group: Maintaining their usual lifestyle. All groups: Two one-hour sessions/week,12 weeks. | (T1) at baseline; (T2) twelve weeks; (T3) six months. | ↓Time to perform the TUG***. ↑ Gait speed**. TUG improved after PT and PCT intervention***. Gait speed in ST and DT conditions improved after PCT training***. |
Schwenk et al. [58], 2016. | Pilot RCT. | Intervention: n = 12. 77.8 years old. Control: n = 10. 79.00 years old. | Petersen Criteria [11]. | Balance (to stand for 30 s with feet close together with EO and EC. Walking at usual pace and a fast pace (10 m). | Wearable sensors. | Intervention: Balance training (weight shifting and virtual obstacle crossing). Real-time visual/ audio lower-limb motion feedback provided from wearable sensors 2/week, 4 weeks). Control: No training. | (T1) at baseline; (T2) four weeks. | ↓CoM sway in both directions (AP, ML) in the intervention with EO**. |
Fogarty et al. [59], 2016. | RCT. | MIP + TTC: n = 22. 71.55 years old. MIP: n = 18. 72.61 years old. | Petersen Criteria [11]. MMSE > 24 [49]. MoCA < 26 [47]. | Walking at usual pace in ST and DT conditions. CTSIB with EO and EC. | GAITRite® Portable Walkway System. Digital Balance Platform. | TTC: Taoist Tai Chi (2/week, 90 min/session, 10 weeks). MIP: Education about lifestyle factors that impact memory and teaching of memory strategies (8 sessions). | (T1) at baseline; (T2) ten weeks; (T3) twenty-two weeks. | No significant change between groups in gait variables, the DT cost variables, or in the amount of sway on the balance measures. |
Bae et Al. [60], 2018. | RCT. | Intervention: n = 41. 75.5 years old. Control: n = 42. 76.4 years old. | Winblad Criteria [13]. MMSE > 24 [49]. | Maximum hand grip strength. Walking speed and physical activity (time spent in MVPA and step count). | Handheld dynamometer. Tri-axial accelerometer. | Intervention: Physical activities (walking, muscle strength training, stretching etc) + cognitive exercises (singing, playing a game, etc) + social activities (2/week, 90 min, 24 weeks). Control: 2 Health education classes (90 mins each, during the 24-week). | (T1) at baseline; (T2) six months. | ↓Time spent in MVPA after intervention in the control group**. ↓Step count after intervention in the control group**. Intervention Group kept baseline parameters. |
Delbroek et al. [61], 2017. | RCT. | Intervention: n = 10. 86.9 years old. Control: n = 10. 87.5 years old. | MoCa < 26 [47]. | TUG in ST and DT conditions. | Inertial measurement units on the ankles, wrists and sternum. | Intervention: Virtual reality dual-task training using the BioRescue (2/week, 18–30 min, 6 weeks). Control: No training. | (T1) at baseline; (T2) six weeks. | ↓Total time to perform the TUG in the intervention group during ST condition**. |
Liao et al. [62], 2019. | RCT. | Intervention: n = 18. 75.5 years old. Control: n = 16. 73.1 years old. | MoCa < 26 [47]. | Walking at preferred Speed in ST and DT conditions. | GAIT Up System. | Intervention: VR-based physical and cognitive training (60 min, 3/week, 12 weeks). Control: Combined physical (resistance, aerobic [50–75% heart rate] and balance exercises) and cognitive exercises. | (T1) at baseline; (T2) three months. | ↑ Gait speed and stride length in ST and DT conditions in VR group** ↑ Gait speed and cadence only in ST in Control Group** No differences between groups*. |
Study (first author and year) | Study Design | Study Characteristics (groups, number of participants, mean age) | MCI Diagnostic Criteria | Instrumented Functional Assessment | Instrument | Data Collection (follow-up) | Main results in MCI |
---|---|---|---|---|---|---|---|
Gillain et al. [63], 2015. | Pilot Cohort Study. | - MCI +: n = 9. 74.44 years old. - MCI -: n = 4. 70.00 years old. | Petersen Criteria [15]. CDR = 0.5 [46]. MMSE > 24 [49]. | Walking at preferred speed (40 m) in ST and DT conditions. | Tri-axial accelerometric (Locometrix®) attached to the L3. | (T1) at baseline; (T2) one year; (T3) four years. | ↑Gait speed in ST and in DT in MCI- than in MCI+ **. ↑Symmetry in DT in MCI- than in MCI+ **. ↓Gait performances in DT compared to ST. |
Hayes et al. [64], 2008. | Transversal and longitudinal study (paired comparison and repeated measure ANOVA) | - Healthy Group: n = 7. 90 years old. - MCI: n = 7. 88.44 years old. | All: MMSE ≥24 [49]. Control: CDR = 0 [46]. MCI: CDR = 0.5 [46]. | Activity in the home, amount of variance in activity, tracking visitors, absences from the home, and walking speed. | Motion sensors and magnetic contact sensors placed in home, and wireless contact switches. | (T1) mean of 315 days. | ↑COV in the median gait speed in MCI compared with Healthy group **. ↑24-h wavelet variance in MCI Group than Healthy Group (↑variance in the day-to-day pattern of activity)**. |
Ansai et al. [65]. 2018. | Longitudinal prospective study. | - AD: n = 37. 78.5 years old. - MCI: n = 38. 74.75 years old. | MCI Group: CDR = 0.5 [46]. MMSE > 24 [49]. Pfeffer [66]. | TUG | Qualisys ProReflex motion analysis system with seven cameras. | (T1) at baseline; (T2) six months. | ↓Total time to perform the TUG in MCI vs AD**. ↑ Gait speed in the walking forward subtask in MCI vs AD**. ↓ Time in the turn subtask in MCI vs AD**. ↑ Gait speed in the walking back subtask in MCI vs AD**. ↓ Time in the turn-to-sit subtask in MCI vs AD**. |
Dodge et al. [67], 2012. | Longitudinal (Latent trajectory model). Part of cohort study. | - aMCI: n = 8. 84,5 years old. - naMCI: n = 31. 83.8 years old. - Healthy Group: n = 54. 84.9 years old. | ALL: CDR ≤ 0.5 [46]. MMSE > 24 [49]. MCI: Petersen Criteria [11]. | Walking speed and its variability; total daily activity, visitors and time out of home. | Motion sensors and contact sensors fixed in the homes, and wireless contact switches. | (T1) at baseline; (T2) mean of 2.6 ± 1.0 years. | Slow gait speed in naMCI**. ↑or↓ baseline COV of gait speed groups in naMCI. ↓Gait speed in MCI than in Healthy Group**. ↑COV of gait speed in MCI than in Healthy Group**. |
Pieruccini- Faria et al. [68], 2018. | Part of a prospective cohort study. | - MCI: n = 52. 73.7 years old. - Healthy Group: n = 27. 71.7 years old. | Control: - CDR = 0 [46]. - MoCA ≥27 [47]. MCI: - CDR = 0.5 [46]. - MoCA < 26 [47]. | Walking speed in ST and DT conditions. | Electronic walkway (lenght: 6 m) embedded with sensors. | (T1) at baseline; (T2) two years; (T3) four years; (T4) five years. | ↓ Gait speed in DT conditions in MCI**. ↓Step length adjustments in DT conditions in MCI**. ↓ Gait speed in MCI**. |
Montero-Odasso et al. [69], 2009. | Reliability study. | - MCI: n = 11. 76.6 years old. | Petersen Criteria [14]. CDR = 0.5 [46]. MoCA < 26 [47]. MMSE > 24 [49]. | Gait performance under ST and DT conditions. | Electronic walkway (GAITRite® System. Lenght: 6 m). | (T1) at baseline; (T2) one week. | ↓Mean gait speed under DT conditions**. ↑Gait variability on stride time, step time, and double support time under DT conditions**. |
Study (first author and year) | Study Design | Study Characteristics (groups, number of participants, mean age) | MCI Diagnostic Criteria | Instrumented Functional Assessment | Instrument | Data Collection (follow-up) | Main results in MCI |
---|---|---|---|---|---|---|---|
Byun et al. [70], 2018. | Prospective cohort study. | Healty: n = 91. 67.3 years old. | Not diagnosis MCI at baseline: CDR = 0 [46]. MMSE > 24 [49]. Winblad Criteria [13] for diagnosis of MCI. | Walking at usual pace (20 m). | Tri-axial accelerometer (FITMETER®) at the level of the 3rd–4th lumbar vertebra. | (T1) at baseline; (T2) 2 years; (T3) median duration was 47.1 months. | ↑Gait variability was a significant predictor of MCI (HR = 11.97, 95% CI = 1.29–111.37)***. Gait speed was slightly associated with incident MCI risk (HR = 5.04, 95% CI = 0.53–48.18) **. |
Akl et al. [71], 2015. | Longitudinal study (trajectory with time window vector machines and random forests). | Older adults: n = 97. NS, 70 years old and + . | Cognitively Healthy: - CDR < 0.5 [46]. - MMSE > 24 [49]. MCI: - CDR = 0.5 [46]. - MMSE > 24 [49]. | Walking speed and general activity in the home. Visitors and absences from the home. | Motion sensors and wireless contact switches placed in the home. | (T1) at baseline; (T2) one year; (T3) two years; (T4) three years. | Trajectories of weekly gait speed, COV of the gait speed, COV of the morning and evening gait speeds could detect MCI in older adults. |
Akl et al. [72], 2015. | Longitudinal study (linear regression). | Older adults: n = 15. NS, 70 years old and + . | Cognitively Healthy: -CDR < 0.5 [46]. MCI: -CDR = 0.5 [46]. | Walking speed in home. | Motion sensors on the ceiling in areas such as a hallway or a corridor. | (T1) at baseline; (T2) one year; (T3) two years; (T4) three years. | Gait speed distributions was different in the subjects when cognitively intact and when having MCI. Transitioning to MCI, daily activities were less distinguishable and often occurred later. |
Buchman et al. [73], 2019. | Longitudinal cohort study. | Older adults: n = 1249. 80.0 years old. | MCI: - MMSE > 24 [49]. | Walking at their self-selected Speed (10 m). TUG. Standing Posture with closed eyes. | Wearable sensor on the lower back. | (T1) at baseline; (T2) during 3.6 years. | ↓ Cadence and regularity were associated with incident MCI **. Gait speed and gait variability were not associated with incident MCI *. |
Instrument | Papers n, % | References |
---|---|---|
Tri-axial accelerometer (e.g. Locometrix®, etc.) | 4, 23.5% | |
Electronic walkway (e.g. GAITRite®, etc.) | 4, 23.5% | |
Wearable sensors | 2, 12% | |
Digital Balance Platform | 1, 6% | [59] |
Inertial measurement units (IMUs) | 1, 6% | [61] |
Motion and contact sensors | 4, 23.5% | |
Qualisys ProReflex motion analysis System (cameras) | 1, 6% | [65] |
GAIT Up System. | 1, 6% | [62] |
Criteria | Papers n, % | References |
---|---|---|
7, 41% | ||
Winblad et al. [13] | 2, 12% | |
CDR [41] | 9, 53% | |
MoCA [42] | 5, 29% | |
MMSE [44] | 11, 65% | |
Pfeiffer [60] | 1, 6% | [65] |