sensors

Systematic Review
Instrumented Timed Up and Go Test (iTUG)—More Than
Assessing Time to Predict Falls: A Systematic Review
Paulina Ortega-Bastidas 1,2 , Britam Gómez 3 , Pablo Aqueveque 4, * , Soledad Luarte-Martínez 2
and Roberto Cano-de-la-Cuerda 5

1 Health Sciences PhD Programme, International Doctoral School, Universidad Rey Juan Carlos,
28922 Madrid, Spain
2 Kinesiology Department, Faculty of Medicine, Universidad de Concepción, Concepción, 151 Janequeo St.,
Concepcion 4030000, Chile
3 Biomedical Engineering, Faculty of Engineering, Universidad de Santiago de Chile, Libertador Bernardo
O’Higgins Av., Santiago 9170022, Chile
4 Department of Electrical Engineering, Faculty of Engineering, Universidad de Concepción, 219 Edmundo
Larenas St., Concepción 4030000, Chile
5 Physiotherapy, Occupational Therapy, Rehabilitation and Physical Medicine Department,
Universidad Rey Juan Carlos, 28922 Madrid, Spain
* Correspondence: pablo.aqueveque@biomedica.udec.cl; Tel.: +56-41-2661262

Abstract: The Timed Up and Go (TUG) test is a widely used tool for assessing the risk of falls in
older adults. However, to increase the test’s predictive value, the instrumented Timed Up and Go
(iTUG) test has been developed, incorporating different technological approaches. This systematic
review aims to explore the evidence of the technological proposal for the segmentation and analysis
of iTUG in elderlies with or without pathologies. A search was conducted in five major databases,
following PRISMA guidelines. The review included 40 studies that met the eligibility criteria. The
most used technology was inertial sensors (75% of the studies), with healthy elderlies (35%) and
elderlies with Parkinson’s disease (32.5%) being the most analyzed participants. In total, 97.5% of
the studies applied automatic segmentation using rule-based algorithms. The iTUG test offers an
economical and accessible alternative to increase the predictive value of TUG, identifying different
Citation: Ortega-Bastidas, P.; Gómez,
variables, and can be used in clinical, community, and home settings.
B.; Aqueveque, P.; Luarte-Martínez,
S.; Cano-de-la-Cuerda, R.
Keywords: instrumented timed up and go; risk of falls; elderly
Instrumented Timed Up and Go Test
(iTUG)—More Than Assessing Time
to Predict Falls: A Systematic Review.
Sensors 2023, 23, 3426. https://
doi.org/10.3390/s23073426 1. Introduction
Falls are accidental events in which people lose control of their center of gravity,
Academic Editors: Laura Gastaldi
and Elisa Digo
where the effort to regain balance is insufficient [1]. In total, 25% of the elderly population
suffer at least one fall per year [2], increasing to two falls when the age is greater than
Received: 2 March 2023 70 years [3]. For this reason, the risk of falls (RoF) is a public health issue [4,5], being
Revised: 17 March 2023 considered one of the main causes of serious injuries in elderlies and the third cause of
Accepted: 20 March 2023 death due to unintentional injury [3–5], causing a sedentary life, loss of functional capacity
Published: 24 March 2023
and a decrease in the quality of life [6]. Thus, the clinical guidelines of “The American and
British Geriatric Societies” recommend asking elderlies over 65 years if they have suffered
two or more falls, if they have been injured during a fall or if they perceive any difficulty
Copyright: © 2023 by the authors.
walking or maintaining balance [3]. Therefore, it is important to detect risk factors of falls
Licensee MDPI, Basel, Switzerland. and balance early, to implement effective and specific preventive clinical strategies [7].
This article is an open access article The main causes of falls are multi-factorial, including extrinsic factors related to the
distributed under the terms and environment, intrinsic factors related to the person, and behavioral factors related to the
conditions of the Creative Commons activity [8]. The most common intrinsic factors are muscle weakness, balance deficits,
Attribution (CC BY) license (https:// and gait instability [7]. For this, proposals to identify people with RoF, and measure
creativecommons.org/licenses/by/ balance and gait [8]. In clinical practice, different tests and observational scales are used
4.0/). to measure static and dynamic balance, as well as gait in healthy subjects or with motor

Sensors 2023, 23, 3426. https://doi.org/10.3390/s23073426 https://www.mdpi.com/journal/sensors

Sensors 2023, 23, 3426 2 of 31

impairments [5]. One of the most widely used tests is the Timed Up and Go (TUG), which
measures dynamic balance and functional mobility [9–11].
The TUG test is a simple test that can be applied in several environments. It was
developed in 1991 by Podsiadlo et al. [12] as a timed modification of the ”Get up and Go”
test. It consists of a circuit in which the subject must get up from a chair, walk three meters,
turn around and walk back to the chair to sit on it again (Figure 1). The controlled variable
is the total test duration in seconds, which is then correlated with the RoF [8,12–14]. This
test presents a high inter-rater and intra-rater reliability, with values greater than 95% in the
prediction of RoF in elderlies, people with stroke [15,16] and Parkinson’s disease (PD) [17].
Other advantages of TUG are the simplicity and duration of its application. Additionally, it
requires minimal equipment and allows subjects with functional disabilities to perform
the evaluation. However, one of the limitations is that it cannot objectively determine the
risk in subjects with greater difficulty. Barry et al. [8] mentioned that a limitation in the
predictive value of the TUG test could be explained that it evaluates balance in a general
way, which could be improved with the addition of technological tools for movement
analysis [8,13].

Figure 1. Timed Up and Go test with the different sub-phases after the most complete segmentation.
(1) Standing. (2) Go Walking. (3) Three-meter turning. (4) Return Walking. (5) Pre-sitting turning.
(6) Sitting.

Nowadays, there are proposals in the literature that allow the instrumentation of
TUG through the use of sensors to capture and analyze movement. This variant of the
test is called Instrumented Timed Up and Go (iTUG) [13,18,19]. This instrumentation
makes it possible to identify the postural transitions performed during the test and to
segment TUG into different sub-phases for extraction of specific measures for each of the
identified sub-phases [20,21]. However, in the scientific literature, different technological
and algorithmic proposals have been presented for iTUG segmentation that differs in the
type of technology used, segmentation algorithms and extracted features for applications
of intervention, characterization and RoF prediction [13,18,19,22–24].
Current systematic reviews on the TUG test study the psychometric properties of
the test [25–27], incorporate information on RoF assessment instruments [5] or evaluate
it as an evaluation instrument for a specific intervention [28]. However, to the best of our
knowledge, there are no reviews where the applicability of iTUG in older adults is explored,
describing the technological elements of the different scientific proposals reported in the
literature and the segmentation and feature extraction strategies for its application.
Thus, we present a systematic review that explores the evidence of the technological
proposals for the segmentation and analysis of iTUG in elderlies with or without associated
pathologies, answering the following questions: What are the technological elements used,
the methodological variations, and the main variables extracted in the application of iTUG
in elderlies? What are the clinical applicability and predictive value of iTUG in elderlies?
Sensors 2023, 23, 3426 3 of 31

2. Methods
The present review follows the guidelines of the PRISMA guide for systematic re-
views [29]. The following databases were reviewed: National Library of Medicine, National
Center for Biotechnology Information (NIH), Pubmed.gov; IEEE Xplore Digital Library,
Scientific Electronic Library Online (SciELO), Elsevier, and Web of Science (WOS).
On 19 April 2022, a first search strategy was evaluated using the command:
(“itug” [All Fields] AND ((“accidental falls” [MeSH Terms] OR (“accidental” [All
Fields] AND “falls” [All Fields]) OR “accidental falls” [All Fields] OR “falling” [All Fields]
OR “falls” [All Fields] OR “fallings” [All Fields]) AND (“risk” [MeSH Terms] OR “risk”
[All Fields]))) AND (y_10[Filter]).
From this strategy, only 37 articles were obtained from the Pubmed database and no
results were obtained in the other databases. Thus, on 19 May 2022, the search command
was modified as follows:
((instrumented AND (y_10[Filter])) AND (((timed up and go AND (y_10[Filter])) OR
(Timed up & go AND (y_10[Filter]))) OR (TUG AND (y_10[Filter ])) AND (y_10[Filter])))
AND (elderly AND (y_10[Filter])) Filters: in the last 10 years.
This last command was used for the rest of the databases, obtaining the final search as
(instrumented) AND ((timed up and go) OR (Timed up & go) OR (TUG)) AND ( elderly),
including studies between the years 2012 and 2022. Three authors tested the last search
command in each database to determine the effectiveness of the search or any difference.
On 31 October 2022, a final search was performed to update the database and identify
new studies that could meet the eligibility criteria.
For the selection of the studies, a conceptual definition of iTUG was determined,
as well as the context in which the proposals were analyzed. Regarding the types of studies
included, no methodological limitations were applied to carry out the selection by level of
evidence. The target population of the studies consisted of elderlies with a mean age equal
to or greater than 65 years with or without associated pathologies. The description of the
eligibility criteria can be observed in detail in Tables 1 and 2.

Table 1. Eligibility criteria for study selection.

Criteria Description
iTUG Studies describe iTUG using technological support to enrich the
test (cameras, inertial sensors, environmental sensors, pressure
sensors, optoelectronic systems).
Index, parameters Proposals that, in addition to the total duration traditionally
or variables obtained in TUG, provide other variables or measurements that
allow the identification of motor alterations in the participants.
Context Proposals evaluated in community, clinical or academic settings.
Study methodology Descriptive, experimental, quasi-experimental and proof-of-
concept clinical studies were included that used validated com-
mercial technology or new technologies whose applications were
applied in the elderly population. No methodological limitations
were applied to carry out the screening by the level of evidence.
Participants Elderlies with a mean age equal or greater than 65 years, with or
without associated pathology.
Language Studies published in English or Spanish.
Study year Studies published between 2012 and 2022.
TUG = Timed Up and Go; iTUG = Instrumented Timed Up and Go.
Sensors 2023, 23, 3426 4 of 31

Table 2. Exclusion criteria for study selection.

Criteria Description
Index, parameters Proposals that only provided the total value of TUG without
or variables demonstrating any other characteristic extracted, independent of
declaring the use of any technology, were excluded.
Study methodology Other narratives, bibliographic, systematic or scoping reviews
were excluded, as well as “one-page” conference articles, abstracts,
posters, letters to the editor and studies of iTUG psychometric
properties validation.
TUG = Timed Up and Go; iTUG = Instrumented Timed Up and Go.

To avoid bias in the selection and analysis of the studies, the initial registration and
screening of the articles were carried out with COVIDENCE® (Melbourne, Australia). Initial
screening by title and abstract was performed by two authors using a blind methodology,
and differences were discussed in conjunction with a third author to resolve discrepan-
cies. During this selection, scientific studies were considered whose titles contained the
keywords: instrumented, timed up and go, falls or risk of falling elderly or older adults
and the conceptual definition of iTUG or that the description of the technology allowed
to identify an iTUG. If the studies are considered potentially eligible, even if they did not
meet the strategy described above, their extended reading was performed to corroborate
whether or not they met the eligibility criteria in Table 1.
The extended review was also conducted at COVIDENCE® (Melbourne, Australia) by
two authors, whose disagreements were resolved by a third author.
Finally, the information extraction of the final selected articles was performed by two
authors of the study, where the data were recorded and stored in a registration form made
with Excel (Microsoft 365® , Redmond, WA, USA), which considered the following data:
• Authors;
• Year/Country;
• Study methodology;
• Institutions;
• Inclusion criteria;
• Exclusion criteria;
• Participants;
• Age of participants;
• Number of participants;
• Number of analyzed participants;
• Gender distribution;
• Technology/Sensors used;
• iTUG implementation;
• TUG implementation;
• Raw data;
• Index, parameters, and variables extracted;
• Segmentation Algorithm;
• Main outcomes;
• Main results.
Any discrepancy was resolved with the participation of a third reviewer.

3. Results
From the initial search in the different databases, 497 studies were obtained, and
74 were removed because they corresponded to duplicates, leaving a total of 423 studies.
During the title and abstract screening stage, 344 articles were excluded, and 79 entered the
extended review stage to determine their eligibility, eliminating 34 studies. Finally, 5 studies
were excluded in the information extraction stage, and 40 were selected for analysis and
Sensors 2023, 23, 3426 5 of 31

discussion. The detail of the selection process can be seen in the PRISMA diagram from
Figure 2.
The presentation of results are presented according to five sections based on the
information extracted from the selected studies, which include characteristics of the partici-
pants and methodological design of the selected studies; types of technologies, procedures
and instrumentation used in the TUG test; algorithmic procedures for segmentation and
extraction of iTUG features; features extracted from iTUG; and main clinical results from
the selected studies (see Table 3).

Figure 2. Systematic review screening process performed in this studio. Flowchart template extracted
and modify from [30].
Sensors 2023, 23, 3426 6 of 31

Table 3. Extensive description of iTUG methodology of all the studies included in the review.

Author Methods iTUG Implementation Segmentation Features Main Results

Ayena et al., Experimental clinical Traditional 3-meter TUG procedure Automatic segmentation Gait index variability. A reduced set of FSR al-
2022 [22] study in people with PD. Segmentation in based on patterns of plan- lows the measurement of
four phases: Standing, Go and Return tar pressures and acceler- gait parameters for measur-
Walking and Sitting. Insole with 4 FSR, ations. The research team ing the RoF in young elderly
one in the heel, the other in the medial developed the algorithm. and PD.
and lateral foot, one in the first and the
other in the fifth metatarsal. The tri-axial
accelerometer is in the instep of the foot.
Ayena et al., Experimental clinical Traditional 3-meter TUG procedure Automatic segmentation 12 spatiotemporal gait pa- The risk of falls index calcu-
2020 [31] study in people with PD. Segmentation in based on patterns of plan- rameters. lated with a minimum num-
four phases: Standing, Go and Return tar pressures and acceler- ber of sensors is a promis-
Walking and Sitting. Insole with 4 FSR, ations. The research team ing tool for real-time analy-
one in the heel, the other in the medial developed the algorithm. sis, as it provides similar sta-
and lateral foot, one in the first and the tistical information in risk
other in the fifth metatarsal. The tri-axial analysis with a larger num-
accelerometer is in the instep of the foot. ber of sensors.
Bergquist et al., Cross-sectional study Traditional procedure of the 3-meter Segmentation using accel- 78 features: Total iTUG du- The iTUG model has a pre-
2020 [32] TUG procedure in communitary elder- eration and angular veloc- ration. iTUG sub-phase du- dictive ability similar to the
lies performing five repetitions. Segmen- ity signals with the algo- ration. Spatial gait param- scores of the CBMS and the
tation into five phases: Standing, Go rithm by Weiss et al. [33] eters in turning and walk- battery of clinical tests with
walking, 3-meter turning, Return walk- ing sub-phases. a significantly lower predic-
ing and Sitting. Smartphone is located tion error.
on the lower back.
Campillay et al., Descriptive study. Traditional procedure of the 3-meter Segmentation using Total iTUG duration. iTUG Duration of iTUG sub-
2017 [34] TUG procedure in elderlies, selecting acceleration and an- sub-phase duration. phases measured through
the best of three repetitions. Segmen- gular velocity signals the smartphone IMU are
tation into five phases: Standing, Go with the algorithm by highly reproducible.
walking, 3-meter turning, Return walk- Mellone et al. [35]
ing and Sitting. Smartphone is located
on the lower back.
Sensors 2023, 23, 3426 7 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Coni et al., Cohort study. Traditional procedure of the 3-meter Segmentation using accel- 38 features: Total du- A significant number of fea-
2015 [36] TUG procedure in communitary elder- eration and angular veloc- ration of iTUG; iTUG tures can be derived from
lies. Segmentation into five phases: ity signals with the algo- sub-phase duration. Sta- the sensor signals, which
Standing, Go walking, 3-meter turning, rithm by Weiss et al. [33] tistical descriptors of can be grouped into factors
Return walking and Sitting. Smartphone angular velocity in the with clear clinical value, al-
located on the waist. turning sub-phase: peak, lowing the measurement of
mean, Root Mean Square many mobility skills from
Value (RMS). Spatial gait one, and not just total time.
parameters: number of
steps on walking and
turning sub-phases.
David et al., Pilot study. Traditional procedure of the 3-meter Segmentation using Spatiotemporal gait pa- The system has proven to
2017 [37] TUG procedure in people with stroke. the spatial position rameters. be capable of measuring
Segmentation into three phases: Stand- of markers with Ma- the level of walking and
ing, Walking and Sitting. chine Learning (Neural TUG, providing details in
Networks) [38] the analysis of the move-
ment and parameters com-
pared with the current data,
which is the total time mea-
sured manually.
Delbroek et al., Randomized clinical Traditional procedure of the 3-meter APDM Mobility Lab pro- Total duration of iTUG. All iTUG measurements
2017 [39] study. TUG procedure in elderlies with visual prietary algorithm. Spatiotemporal gait pa- improve with l-dopa, ex-
tasks. IMU on ankles, wrists and chest. rameters. Statistical de- cept sit-to-stand transition
Segmentation into six phases: Stand- scriptors of angular veloc- (duration and AP accelera-
ing, Go walking, 3-meter turning, Return ity and accelerations. tion) and stand-to-sit transi-
walking, Pre-sitting turning, Sitting. tion (all parameters).
Dibilio et al., Does not specify. Traditional procedure of the 3-meter BTS G-Studio proprietary Freezing of gait. Total Low but significant cor-
2017 [40] TUG procedure in people with PD with algorithm. duration of iTUG. Du- relations were found be-
visual tasks. The sensor is on the lower ration per sub-phase of tween motor scores (iTUG
back at L4–L5. Segmentation into six iTUG. Spatiotemporal parameters) and global cog-
phases: Standing, Go walking, 3-meter gait parameters. nitive function.
turning, Return walking, Pre-sitting turn-
ing, Sitting.
Sensors 2023, 23, 3426 8 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Fastame et al., Exploratory study. Traditional procedure of the 3-meter BTS G-Studio proprietary Duration and average It is a robust, stable and sen-
2022 [41] TUG procedure in communitary elder- algorithm. speed of each sub-phase sitive measurement tool for
lies with visual tasks. The sensor is on of iTUG. “movement smoothness” in-
the lower back at L5–S1. Segmentation dependent of the speed
into six phases: Standing, Go walking, and duration of the move-
3-meter turning, Return walking, Pre- ment, capable of identify-
sitting turning, Sitting. ing RoF independent of the
day’s speed.
Figuiredo et al., case-control study. Traditional procedure of the 3-meter BTS G-Studio proprietary Total duration of iTUG. It is possible to identify
2020 [42] TUG procedure in elderlies with visual algorithm. Duration per sub-phase of significant differences in
tasks. The sensor is on the lower back iTUG. Statistical descrip- the temporal variables
at L5–S1. Segmentation into six phases: tors of angular velocity by of the sub-phases and
Standing, Go walking, 3-meter turning, sub-phase. Statistical de- their speed, mainly in the
Return walking, Pre-sitting turning, Sit- scriptors of degrees of ori- elderly population.
ting. entation and inclination by
sub-phase.
Mangano et al., Transversal study. Traditional procedure of the 3-meter BTS G-Studio proprietary Duration and average No differences were found
2020 [43] TUG procedure and an extended 7-meter algorithm. speed of each sub-phase between Parkinson’s
version in elderlies. The sensor is on the of iTUG. patients with RoF and
lower back at L1. Segmentation into six without RoF in stride length
phases: Standing, Go walking, 3-meter and speed parameters.
turning, Return walking, Pre-sitting turn- Conversely, there was a
ing, Sitting. difference in the time of
double support. In addition,
good correlations were
found between the ABC
Scale and iTUG scales.
Sensors 2023, 23, 3426 9 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Mancini et al., Cross-sectional study. Extended 7-meter TUG in people with APDM Mobility Lab pro- 52 features: Spatiotempo- An instrumented scale can
2012 [44] PD. The sensor is on the lower back and prietary algorithm. ral gait parameters. Statis- reveal deficits in the turns
on the leg shank segment. Segmentation tical descriptors of acceler- of patients with Parkinson’s
into six phases: Standing, Go walking, ation of the walking sub- (severe and mild), even
3-meter turning, Return walking, Pre- phase. Acceleration fre- though they have a normal
sitting turning, Sitting. quency analysis of the gait score and walk correctly.
sub-phase.
King et al., Exploratory study. Extended 7-meter TUG in people with APDM Mobility Lab pro- Total duration of iTUG; The system is useful for
2012 [45] PD. The sensor is on the lower back, leg prietary algorithm. Turning sub-phase dura- analyzing the gait patterns
shank segment, arms and chest. Seg- tion. Spatial gait param- during TUG, applicable to
mentation into six phases: Standing, Go eters: Step number dur- healthy subjects or with dif-
walking, 3-meter turning, Return walk- ing turning. Statistical de- ferent gait disorders by ana-
ing, Pre-sitting turning, Sitting. scriptors of angular veloc- lyzing the correlations of the
ity during turning. go and return signals.
Ishikawa et al., Exploratory study. Traditional procedure of the 3-meter Segmentation using Spatiotemporal gait param- It could be correctly identi-
2019 [46] TUG procedure in elderlies with iNPH. acceleration and an- eters: Gait speed, cadence, fied and with an error under
Smartphone on the abdomen. Segmenta- gular velocity signals stride time variability. the test transition elements
tion into six phases: Standing, Go walk- with the algorithm by in the predetermined order
ing, 3-meter turning, Return walking, Salarian et al. [47]. and in any other sequence of
Pre-sitting turning, Sitting. the movements or activities
measured during TUG.
Silva et al., Exploratory study. Traditional procedure of the 3-meter Automatic segmentation Total duration of iTUG; A strategy implemented in
2016 [48] TUG procedure in elderlies. The sensor algorithm using the inte- Duration per sub-phase of stages with and without
is in a pocket or attached to the thigh. gral of the angular veloc- iTUG; Spatial gait param- medication in patients with
Segmentation into three phases: Stand- ity signal developed by eters: Step number dur- Parkinson’s makes it possi-
ing, Go walking, 3-meter turning. the same research team. ing walking. Statistical ble to predict falls.
descriptors of acceleration
by sub-phase. Acceler-
ation frequency analysis
by sub-phase.
Sensors 2023, 23, 3426 10 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Hoskovcova et al., Prospective study. Traditional 3-meter TUG and extended Does not report segmenta- Spatiotemporal gait pa- Using principal component
2015 [49] 7-meter TUG in people with PD. Use of tion strategy. rameters: gait speed, analysis, it was observed
five inertial sensors located in the leg, cadence and stride that the characteristics that
wrist and sternum areas. time variability. contributed the most to the
discrimination with respect
to the control group were
in the phase of getting up,
the chest flexion, the maxi-
mum obliquity of the chest
and the maximum vertical
velocity of the chest.
Hellmers et al., Exploratory study. Traditional 3-meter TUG in elderlies. Segmentation with the al- Total duration of iTUG; There are some differences
2018 [50] A belt with sensors in L4–L5 and an in- gorithm of Nguyen et al., Turning sub-phase du- in the acceleration signals
strumented chair with four pressure sen- 2015 [11]. ration. Statistical de- between groups of frail
sors and a laser are used. Segmentation scriptors of acceleration adults compared to non-
into six phases: Standing, Go walking, by sub-phase. frail adults by means of
3-meter turning, Return walking, Pre- a smartphone placed on
sitting turning, Sitting. the chest.
Gasparutto et al., Prospective study. Traditional procedure of the 3-meter Automatic segmentation 33 features: joint mobility This simple test could be
2021 [23] TUG procedure in elderlies with hip with the algorithm by range descriptors by iTUG appropriate for quantifying
arthroplasty. In total, 35 reflective mark- Bayea et al. [51]. sub-phase: spine at thorax patient-specific deficits in
ers were used according to the conven- level and c7. The base of function, and hence, guid-
tional gait model. Segmentation into five support by sub-phase. Spa- ing and monitoring post-
phases: Standing, Go walking, 3-meter tial gait parameters: num- operative rehabilitation in
turning, Return walking and Sitting. ber of steps during walk- clinical settings.
ing and turning. Statistical
descriptors of angular ve-
locity by sub-phase of the
thorax and pelvis.
Sensors 2023, 23, 3426 11 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Galán-Mercant Cross-sectional study. Extended 10-meter TUG in elderlies with Segmentation using Duration per sub-phase of An inertial sensor from the
et al., 2015 [52] the frail syndrome. Smartphone on acceleration and an- iTUG. Statistical descrip- iPhone 4 can study and an-
thorax. Segmentation into five phases: gular velocity signals tors of acceleration by sub- alyze the kinematics of the
Standing, Go walking, 3-meter turning, with algorithms from phase. Statistical descrip- different sub-phases of the
Return walking and Sitting. Weiss et al. [33] and tors of angular velocity by Extended TUG test in frail
Salarian et al. [47]. sub-phase. Statistical de- and non-frail elderly people.
scriptors of degrees of ori-
entation and inclination by
sub-phase.
Galán-Mercant Cross-sectional study. Extended 10-meter TUG in elderlies with Segmentation using Statistical descriptors of The inertial sensor found
et al., 2013 [53] frail syndrome. Smartphone on thorax. acceleration and an- acceleration by sub-phase. in the iPhone 4 is able to
Segmentation into five phases: Stand- gular velocity signals Statistical descriptors of study and analyze the kine-
ing, Go walking, 3-meter turning, Return with algorithms from angular velocity by sub- matics of the turning tran-
walking and Sitting. Weiss et al. [33] and phase. Statistical descrip- sitions in frail and physi-
Salarian et al. [47]. tors of degrees of orienta- cally active elderly persons.
tion and inclination by sub- The accelerometry values
phase. Total duration of for the frail elderly are lower
iTUG. Duration per sub- than the physically active
phase of iTUG. elderly, while variability in
the readings for the frail el-
derly is also lower than the
control group.
Galán-Mercant Cross-sectional study. Extended 10-meter TUG in elderlies with Segmentation using Statistical descriptors of For the Extended TUG test,
et al., 2014 [54] frail syndrome. Smartphone on thorax. acceleration and an- acceleration by sub-phase. this device allows more
Segmentation into five phases: Stand- gular velocity signals Statistical descriptors of sensitive differentiation be-
ing, Go walking, 3-meter turning, Return with algorithms from angular velocity by sub- tween population groups
walking and Sitting. Weiss et al. [33] and phase. Total duration of than the traditionally used
Salarian et al. [47]. iTUG. Duration per sub- variable, namely time.
phase of iTUG. Joint mobil-
ity ranges by sub-phase of
iTUG (Thorax).
Sensors 2023, 23, 3426 12 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Van Uem et al., Cross-sectional study. Traditional 3-meter TUG in people with Segmentation using Statistical descriptors of Spontaneous physical ac-
2016 [55] PD. A belt with sensors in L4–L5. Seg- angular velocity sig- acceleration by sub-phase. tivity at home and instru-
mentation into six phases: Standing, Go nals and accelerations Statistical descriptors of mented assessments in the
walking, 3-meter turning, Return walk- with the algorithm of angular velocity by sub- clinic demonstrated strong
ing, Pre-sitting turning, Sitting. Walgaard et al. [56]. phase. Total duration of discriminatory power in de-
iTUG. Duration per sub- tecting impaired motor func-
phase of iTUG. Joint mobil- tion in Parkinson’s disease.
ity ranges by sub-phase of
iTUG (Lumbar).
Van Lum- Exploratory study. Traditional 3-meter TUG in people with Segmentation using Total duration of iTUG. Intra-rater, inter-rater and
mel et al., PD. A belt with sensors in L4–L5. Seg- angular velocity sig- Duration per sub-phase test-retest reliability of the
2016 [18] mentation into six phases: Standing, Go nals and accelerations of iTUG. individual components of
walking, 3-meter turning, Return walk- with the algorithm of iTUG was excellent to good
ing, Pre-sitting turning, Sitting. Walgaard et al. [56]. for total duration and turn-
ing durations and good to
poor for sitting durations
and kinematics.
Toosizadeh Cohort study. Traditional 3-meter TUG in people with Segmentation using Statistical descriptors of iTUG measurements ob-
et al., 2015 [53] PD. Five units of measurement, one on acceleration and an- acceleration by sub-phase. tained from trunk angular
each leg, one on each thigh, and one on gular velocity signals Statistical descriptors of velocity during the turn-
the lower back. Segmentation into five with the algorithm by angular velocity by sub- ing and standing phases
phases: Standing, Go walking, 3-meter Salarian et al. [47]. phase. Total duration of are adequate measures
turning, Return walking, Sitting. iTUG. Duration per sub- of dynamic balance in
phase of iTUG. Joint mo- Parkinson’s disease. These
bility ranges by sub-phase measurements respond to
of iTUG. rehabilitation, being able
to detect improvement in
patients after treatment.
Sensors 2023, 23, 3426 13 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Tan et al., Does not specify. Reduced version at 2-meter TUG in peo- Algorithm developed by Spatial gait parameters: The only Kinect-derived
2018 [57] ple with PD. The Kinect camera is located the same research team. Length of the first step. To- variable significantly associ-
1.4 m from the test circuit. tal duration of iTUG. Turn- ated with the motor analysis
ing sub-phase duration. scale was first step length.
Sheehan et al., Longitudinal study. Traditional 3-meter TUG in elderlies. Does not specify a seg- Spatiotemporal gait pa- The quantitative values of
2014 [58] Sensors are located on each leg. mentation algorithm. rameters. Spatial gait pa- iTUG and linear regression
rameters. Statistical de- models allow us to iden-
scriptors of angular veloc- tify decreases in the bal-
ity by sub-phase. ance of highly functional
older adults.
Seo et al., Transversal study. Traditional procedure of the 3-meter APDM Mobility Lab pro- 36 features: Spatiotempo- It was possible to predict
2019 [59] TUG procedure in elderlies with visual prietary algorithm. ral gait parameters. Sta- the falls with 70.2% accu-
tasks. IMU on ankles, wrists, lower back tistical descriptors of angu- racy using the characteris-
and chest. Segmentation into five phases: lar velocity by sub-phase. tics of the total duration of
Standing, Go walking, 3-meter turning, Total duration of iTUG. the test, standing character-
Return walking, Sitting. Duration per sub-phase istics, the maximum speed
of iTUG. Joint mobility of the trunk in the sagit-
ranges by sub-phase of tal plane and its range of
iTUG (thorax). movement in the horizon-
tal plane during walking
and the speed angular max-
ima during sitting.
Picardi et al., Longitudinal study. Traditional 3-meter TUG in people with Segmentation using Spatiotemporal gait pa- iTUG measurements ob-
2020 [60] PD. The sensor on the lower back. Seg- the morphology of rameters. Statistical de- tained from trunk angular
mentation into five phases: Standing, Go the acceleration signal scriptors of angular veloc- velocity during the turning
walking, 3-meter turning, Return walk- with the algorithm of ity by sub-phase. Total du- and standing phases are
ing and Sitting. Mellone et al. [35]. ration of iTUG. Duration adequate measures of
per sub-phase of iTUG. dynamic balance in Parkin-
son’s disease, and these
measurements respond to
rehabilitation, being able
to detect improvement in
patients after treatment.
Sensors 2023, 23, 3426 14 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Nguyen et al., Exploratory study. Extended 5-meter TUG and 10-meter Algorithm developed by 36 features: Spatiotempo- The proposed algorithm al-
2015 [61] TUG in elderlies. 17 IMUs in each the same research team. ral gait parameters. Statis- lows detecting and segment-
body segment. Segmentation into six tical descriptors of angular ing typical TUG activities
phases: Standing, Go walking, 3-meter velocity by sub-phase. To- using inertial sensors.
turning, Return walking, Pre-sitting tal duration of iTUG. Dura-
turning, Sitting. tion per sub-phase of iTUG.
Joint mobility ranges by
iTUG sub-phase.
Nguyen et al., Exploratory study. Extended 5-meter TUG and 10-meter Algorithm developed by Spatial gait parameters: The transferability of the
2017 [11] TUG in eldelries. 17 IMUs in each the same research team. number of steps during segmentation methodology
body segment. Segmentation into six walking and turning. Sta- to Parkinson’s patients
phases: Standing, Go walking, 3-meter tistical descriptors of accel- is demonstrated, using
turning, Return walking, Pre-sitting eration by sub-phase. Sta- only 4 of the 17 initially
turning, Sitting. tistical descriptors of angu- predisposed sensors.
lar velocity by sub-phase.
Joint mobility ranges by
sub-phase of iTUG.
Najafi et al., Exploratory study. Traditional procedure of the 3-meter Evaluation of postural tran- Duration per sub-phase The proposed system can
2013 [62] TUG procedure in elderlies with periph- sitions through accelerom- of iTUG. identify and monitor pos-
eral neuropathy and diabetes with vi- eter analysis. The algo- tural transitions, accurately
sual tasks. Sensor on thorax. Segmen- rithm was developed by identifying subjects with a
tation into five phases: Standing, Go the same research team. high RoF with potential use
walking, 3-meter turning, Return walk- in monitoring older adults
ing and Sitting. with diabetes.
Mirelman et al., Cross-sectional study. Traditional 3-meter TUG in elderlies. Segmentation using the Duration per sub-phase Using a single sensor on the
2014 [63] Sensor on the lower back. Segmentation morphology of the acceler- of iTUG. back during TUG can quan-
into five phases: Standing, Go walking, ation and angular velocity tify mobility, facilitating the
3-meter turning, Return walking, Sitting. signal with the algorithm understanding of problems
of Weiss et al. [33]. related to cognitive decline.
Sensors 2023, 23, 3426 15 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Mariani et al., Exploratory study. Traditional 3-meter TUG in people with Segmentation using angu- Duration per sub-phase of All the sub-phases of the
2013 [64] PD. IMU on the instep of the foot. Seg- lar velocity signals and iTUG. Spatiotemporal gait test could be detected in the
mentation into three phases: Go walking, accelerations. The algo- parameters: step number control groups, with medi-
3-meter turning, Return walking. rithm was developed by and gait velocity. cation and without medica-
the same research team. tion. Variations of the mea-
sured gait characteristics are
observed, where the tempo-
ral variables are the most im-
portant within the TUG test.
Frenken et al., Experimental study. Traditional 3-meter TUG in elderlies. No Algorithm developed by Statistical descriptors of de- The proposed method is
2013 [24] sensors on the participants. aTUG (ambi- the same research team. grees of orientation and able to accurately compute
ent TUG) chair integrated with environ- inclination by sub-phase. the duration of the TUG
mental sensors, four force sensors and Duration per sub-phase components using only the
an optical laser. Segmentation into fivex of iTUG. force sensors and the laser
phases: Standing, Go walking, 3-meter range scanner.
turning, Return walking, Sitting.
Gerhardy et al., Cross-sectional study. Traditional procedure of the 3-meter Segmentation using Total duration of iTUG. Total TUG time was strongly
2019 [65] TUG procedure in elderlies. The sensor the morphology of Duration per sub-phase associated with vestibular
is on lower back. Segmentation into five the acceleration signal of iTUG. and somatosensory system
phases: Standing, Go walking, 3-meter with the algorithm of performance.
turning, Return walking, Sitting. Mellone et al. [35].
Weiss et al., 2013 Cohort study. Traditional 3-meter TUG in elderlies. Segmentation using accel- Spatio temporal gait pa- People with mobility impair-
[33] Sensor on the lower back. Segmentation eration and angular ve- rameters: Stride speed, ments have problems in the
into five phases: Standing, Go walking, locity signals with algo- stride length, swing width, five sub-phases of TUG.
3-meter turning, Return walking, Sitting. rithms developed by the stride length. Statistical
same team [66]. descriptors of degrees of
orientation and inclination
by sub-phase of rotation.
Duration per sub-phase
of iTUG.
Sensors 2023, 23, 3426 16 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Weiss et al., 2016 Cohort study. Traditional 3-meter TUG in elderlies Segmentation using accel- Total duration of iTUG. Longer separations between
[67] with dementia. The sensor on the eration and angular ve- Duration per sub-phase the movements of the sub-
lower back. Segmentation into six locity signals with algo- of iTUG. phases and a longer over-
phases: Standing, Go walking, 3-meter rithms developed by the lap between turning and
turning, Return walking, Pre-sitting same team [66]. the stand-to-sit sub-phase
turning, Sitting. are related to poorer cogni-
tive and motor function and
greater disability.
Williams et al., Cross-sectional study. Traditional 3-meter TUG in elderlies Segmentation using Statistical descriptors The sensor and its algo-
2021 [21] with dementia. The sensor is on the angular velocity sig- of angular velocity rithms were able to quantify
lower back. Do not specify level nals and accelerations by sub-phase. the sub-phases of iTUG and
of segmentation. with the algorithm of demonstrate that age mod-
Walgaard et al. [56]. erates differences in the per-
formance of iTUG and its in-
formal caregivers.
Yu et al., Descriptive study. Traditional procedure of the 3-meter Segmentation using Statistical descriptors of The proposal is a reliable op-
2021 [68] TUG procedure in communitary elder- acceleration and an- acceleration by sub-phase. tion for objective, unsuper-
lies. Sensor on the lower back. Seg- gular velocity signals Statistical descriptors of vised and unobtrusive bal-
mentation into five phases: Standing, with algorithms from angular velocity by sub- ance measurement in a clini-
Go walking, 3-meter turning, Return Weiss et al. [33]. phase. Statistical descrip- cal or home setting.
walking, Sitting. tors of degrees of ori-
entation and inclination
by sub-phase.
Sensors 2023, 23, 3426 17 of 31

Table 3. Cont.

Author Methods iTUG Implementation Segmentation Features Main Results

Zarzeczny et al., Cross-sectional study. Traditional procedure of the 3-meter BTS G-Studio proprietary Statistical descriptors of The performance of the
2017 [69] TUG procedure in elderlies with visual algorithm. acceleration by sub-phase. functional tests is more
tasks. Sensor on the lower back at L4–L5. Statistical descriptors of dependent on the range
Segmentation into six phases: Stand- angular velocity by sub- of force developed than
ing, Go walking, 3-meter turning, Return phase. Statistical descrip- the maximum isometric
walking, Pre-sitting turning, Sitting. tors of degrees of ori- force of the muscles of the
entation and inclination lower extremities during
by sub-phase. acceleration when sitting
is measured.
PD = Parkinson’s Disease; TUG = Timed Up and Go; FSR = Force Resistive Sensor; iTUG = Instrumented Timed Up and Go; CBMS = Community Balance and Mobility Scale;
IMU = Inertial Measurement Unit; PD = Parkinson’s Disease; RMS = Root Mean Square; AP = Antero-posterior; RoF = Risk of Falls; ABC Scale = Activities-specific Balance Confidence
Scale; iNPH = Idiopathic normal pressure hydrocephalus.
Sensors 2023, 23, 3426 18 of 31

3.1. Characteristics of the Participants

From the selected studies, it can be observed that most participants were elderlies
without health issues [24,33,34,39,42,43,48,50,58,59,61,63,65,67]. In only four studies, they
were described as community elderlies [32,36,41,68], and in only one of them, they were de-
scribed as residence elderlies [69]. Thirteen studies included participants with a diagnosis of
PD [11,18,31,40,44,45,49,55,57,60,64,70] and only one of them with a diagnosis of stroke [13].
In addition, seven studies included participants with other pathologies, such as hip
arthroplasty [23], possible “idiopathic normal pressure hydrocephalus (iNPH) [46], frailty
index greater than 3.9 according to the Fried scale [52–54], neuropathy peripheral [62]
and dementia [21]. Table 4 shows the studies according to the participant characteristics,
diagnosis and the total number of participants.

Table 4. Participants of the selected studies.

Subjects Studies Participants

Elderlies [24,33,34,39,42,43,48,50,58,59,61,63,65,67] 14 2448
Community elderlies [32,36,41,68] 4 521
Residence elderlies [69] 1 26
PD [11,18,31,40,44,45,49,55,57,60,64,70] 13 426
People with stroke [13] 1 35
Elderlies with other pathology [21,23,46,52–54,62] 7 376
Total 40 3822
PD = Parkinson’s disease.

According to the eligibility criteria, the age of the participants had to be greater than or
equal to 65 years (see Figure 3), and the distribution by gender mostly showed a tendency
to a greater number of women than men. Figure 4 illustrates these tendencies.

Figure 3. Distribution of the selected study participant’s ages, indicating the participant’s condition
and the predominant general age range. In the participant’s condition plot, the “Other” category
means elderlies with dementia, hip arthroplasty, diabetes and frail syndrome.

3.2. Methodological Design of the Selected Studies

The selected studies vary in the type of study design; although the level of evi-
dence declared no methodological limitations, it can be seen in Table 5 that the pre-
dominant methodological designs were cross-sectional [21,32,44,52–55,63,65,69] and ex-
ploratory studies [11,18,41,45,46,48,50,61,62,64], with ten articles each, followed by cohort
studies [33,36,67,70], with four articles. A smaller number of studies were experimen-
tal [22,24,31], with three articles, and descriptive [34,68], cross-sectional [43,59], prospec-
tive [23,49] and longitudinal [58,60], with two articles each. Finally, there were observed
Sensors 2023, 23, 3426 19 of 31

studies with case-control [42], clinical-randomized [39], and pilot [37] methodologies, with
one article each. Finally, two articles did not mention the type of study [40,57].

Figure 4. Proportion in percentages of the participants by sex from the studies selected.

Table 5. Type of studies reviewed.

Type of Study Studies

Cross-sectional [21,32,44,52–55,63,65,69] 10
Exploratory [11,18,41,45,46,48,50,61,62,64] 10
Cohort [33,36,67,70] 4
Clinical-Experimental [22,24,31] 3
Descriptive [34,68] 2
Transversal [43,59] 2
Prospective [23,49] 2
Longitudinal [58,60] 2
case-control [42] 1
Clinical-Randomized [39] 1
Pilot [37] 1
Not mentioned [40,57] 2
Total 40

3.3. Types of Technology, Procedure and Instrumentation Used in the Timed Up and Go Tests
The iTUG test has been defined as the use of inertial sensors to achieve the segmenta-
tion of the test and extract characteristics from the identified sub-phases [20,45]. However,
in this review, different technological proposals were found. Table 6 shows the number of
studies by type of technology used to implement the TUG.
Sensors 2023, 23, 3426 20 of 31

Table 6. Technologies and sensors used for iTUG.

Technology Studies
Insoles [22,31,37] 3
Smartphone [32,34,36,46,48,52–54,65] 9
Inertial Sensors [11,18,21,33,39–45,49,55,58–64,67–70] 24
Opto–electronic System [23,55] 2
Xbox Kinect [57] 1
Instrumented Chair [24,50] 2

Several proposals use inertial sensors for the instrumentation of the test [11,18,21,
33,39–45,49,55,58–64,67–70]. Most of the studies use commercial sensors such as the G-
Sensor, BTS G Walk® (BTS Bioengineering, Lombardia, Italy) [39–43,69], whose inertial
unit consists of a sensor with a tri-axial accelerometer and a tri-axial gyroscope, with a
maximum sampling frequency of 1000 Hz. Other commercial sensors used include the Opal
Sensor (APDM wearable technologies, Portland, OR, USA) [39,59], MTX XSens sensors
(49A33G15, Xsens, Enschede, the Netherlands) [44,45,49], Dynaport sensor (McRoberts
technologies, the Hague, the Netherlands) [18,33,55,63,67], which integrates a tri-axial
accelerometer and gyroscope with a sampling frequency of 100 Hz, Shimmer inertial
sensor (Shimmer technologies, Dublin, Ireland) [58], LEGSys and BalanSens (BioSensics,
Boston, MA, USA) [70], mHT (mHealth Technologies, Bologna, Italy) [60,65], PAMSys
inertial sensor (Biosensics, Newton, MA, USA) [62] and the tri-axial inertial sensor (Balance
THETAmetrix, Portsmouth, UK) [21].
On the other hand, some studies mention the use of up to 17 measurement units [11,61],
being able to acquire data from each body segment during iTUG. Other studies propose
the use of six sensors, one in the sternum, one in L3, two in both hips and two in both
thighs [59], or located in L5, two in the front part of the leg below the knee, two in the
lateral part of the arm and one in the sternum [45]. However, most studies propose the
use of a single sensor that is generally located between the lower back between L4 and S1,
depending on the protocol used by the investigators [18,21,33,40–43,55,63,67–69].
Regarding the use of smartphone inertial sensors as technology for iTUG, four studies
mention the use of the iPhone 4 smartphone (Apple Inc., Cupertino, CA, USA), located at
the lower back [34] and on the sternum [52–54]. Two studies used Samsung Galaxy smart-
phones [46,48] located on the lower back. Another study used a Huawei P8 smartphone
(Huawei, Shenzhen, China) positioned on the lower back [32].
On the other hand, three studies included the use of insoles, two of which considered
an insole with 4 FSR (force-sensing resistors), which were positioned to measure the
distribution of force in the foot. Two FSRs were positioned on the heel, one medial and
one lateral, and the other two were located on the first and fifth metatarsals approximately,
in conjunction with a 3D accelerometer attached to the foot [22,31]. A single study included
the eSHOE insole system, which consists of a pair of orthopedic insoles that includes
tri-axial accelerometers, tri-axial magnetometers, and a tri-axial gyroscope, as well as a
pressure sensor on the greater toe, first and fifth metatarsal heads [37]. Two studies have
incorporated the use of a sensorized chair called aTUG (ambiental Timed Up and Go),
which considers an integrated chair with environmental sensors, four force sensors and a
laser bar [24,50].
As mentioned above, the iTUG allows the segmentation of the TUG test into different
sub-phases related to the activities that the participants must perform when executing
the test. In the included literature, differences have been found regarding the number of
sub-phases described in the segmentation. Some proposals have included the segmentation
into three sub-phases, which consider the activities sit to stand, walk and stand to sit [31]
or standing, forward walking and turning [39]. On the other hand, segmentation proposals
Sensors 2023, 23, 3426 21 of 31

have been found in four sub-phases that, in general, analyze the phases sit to stand, walk,
180º turn, and stand to sit [23,33,34,41,63,64,70]. However, segmentation proposals differ at
the moment when a transition or transfer is initiated, for example, sub-phases have been
described as sit-to-walk, walk, first turn [32,36] and turn to sit [32,33,36], or considering the
last sub-phase directly from walking to sitting [55]. Likewise, five sub-phase segmentation
proposals have been found, which describe sit-to-stand, walk-to-stand, turn, walk-to-sit
and sit [24,52–54,68]. In the other five sub-phase proposals, differences have been found
in the last phase of TUG, also describing the turn-to-sit phase [34,62,65]. On the other
hand, the segmentation into six sub-phases of TUG considers sit-to-stand, forward gait,
180° turn, backward gait, turn, and stand-to-sit [11,18,40,43,46,50,61,69]. A single study
proposes segmentation into the following phases: sit-to-stand, gait, turn, stand-to-sit,
the full duration of the last turn to sit, the interval between the end of the last turn and the
start of the stand-to-sit sub-phase [67].
Podsiadlo et al. [12] indicated that the standard procedure of TUG is that the user
must stay sit in a chair without armrests, stand up from the chair, walk forward a dis-
tance of 3 m, turn around a mark or cone in the three meters, walk back and sit down
on the chair. Most of the studies used the conventional procedure described previ-
ously [18,21–24,31–33,36,37,40,41,44,46,48,50,55,58,59,62–65,67–70]. However, some pro-
posals used the extended versions of TUG with a distance of 5 m [11,61], 7 m [43–45,49] and
10 m [11,52–54,61], maintaining the same activities requested in the 3-meter TUG. In addi-
tion, proposals include traditional TUG plus dual tasks during its execution [39,42]. Lastly,
only one study used a 2-meter TUG proposal [57].

3.4. Algorithmic Procedures for Segmentation and Extraction of iTUG Features

One objective of the instrumentation of clinical trials is to obtain augmented informa-
tion that allows automating the evaluation, complementing the evaluation results with
quantifiable information. In general, the iTUG allows to extract information from different
sub-phases or postural transfers, through a process called segmentation.
In this review, different segmentation proposals have been found using rule based
algorithms—that used the patterns of the acceleration and angular velocity signals—and,
in some cases, machine learning tools for the identification of postural transitions. Table 7
presents a summary of the algorithms used in the included studies and their characteristics.
A large number of algorithms have been proposed for the analysis of iTUG, corre-
sponding to 27 studies that use rule-based algorithms for the automatic segmentation of
iTUG, 1 algorithm that is based on Machine Learning, 9 that use segmentation strategies
developed by companies such as BTS G-Studio and APDM Mobility Lab and two stud-
ies that did not propose segmentation strategies. This could be explained byanalyzing
pre-established movement patterns for the iTUG sub-phases, where rule-based algorithms
perform well. All these algorithms implement the segmentation after the execution of the
test and not in real-time.
Seven studies used the algorithm of Weiss et al. [33], being the most applied algorithm
for the segmentation of iTUG using inertial sensors on the lower back. Five studies used
the BTS G-Studio platform, which also uses a segmentation strategy with a single sensor
located in the lower back. Four studies use the APDM Mobility Lab system. It is unclear
from the studies which segmentation strategy is used, since inertial sensors can be located
in up to six different body segments in combination to perform the segmentation process.
Three studies used the Mellone et al. algorithm [35], being the preferred algorithm for
iTUG segmentation when the technological base consists of inertial sensors of smartphones
located on the upper thorax. Three studies used the algorithm of Walgaard et al. [56], which,
in addition to using the information on the acceleration and angular velocity of the back,
uses spatial orientation data (inclination and rotation) to determine postural transitions.
Three studies used the Salarian et al. algorithm [47] to determine postural transitions from
waist accelerometry and waist angular velocity.
Sensors 2023, 23, 3426 22 of 31

Table 7. Algorithms for iTUG segmentation identified from the selected studies.

Author Proposal Inputs Transitions Studies

Ayena et al. [31] Rule-based. Tri-axial foot accelerations. S2W, W2S. 2
Plantar pressures.
Weiss et al. [33] Rule-based. Tri-axial lower back accelerations and tri-axial lower back angular velocities St, S2W, W2T, T2W, W2T, T2S, Si. 7
Mellone et al. [35] Rule-based. Tri-axial chest accelerations. St, S2W, W2T, T2W, W2T, Si. 3
Holzreiter et al. [38] Machine Learning. On-body reflective infrared marker’s coordinates. St, S2W, W2T, T2W, W2T, Si. 1
APDM Mobility Lab. Propietary. Tri-axial accelerations and tri-axial angular velocities from both wrists, St, S2W, W2T, T2W, W2T, T2S, Si. 4
shanks, lower back and/or chest.
BTS G-Studio. Propietary. Tri-axial lower back accelerations and tri-axial lower back angular velocities. St, S2W, W2T, T2W, W2T, T2S, Si. 5
Silva et al. [48] Rule-based. Tri-axial thigh angular velocities. St, S2W, W2T. 1
Salarian et al. [19] Rule-based. Tri-axial waist accelerations and tri-axial waist angular velocities. St, S2W, W2T, T2W, W2T, T2S, Si. 3
Nguyen et al. [11] Rule-based. Tri-axial lower back accelerations and tri-axial lower back angular velocities. St, S2W, W2T, T2W, W2T, T2S, Si. 3
Bayea et al. [51] Rule-based. Tri-axial upper body accelerations. St, S2W, W2T, T2W, W2T, T2S, Si. 1
Tri-axial upper body angular velocities.
Walgaard et al. [56] Rule-based. Tri-axial lower back accelerations. St, S2W, W2T, T2W, W2T, T2S, Si. 3
Tri-axial lower back angular velocities.
Displacements and angles of lower back and feet.
Tan et al. [57] Rule-based. Kinect video. St, S2W, W2T, T2W, W2T, T2S, Si. 1
Najafi et al. [62] Rule-based. Tri-axial chest accelerations. St, S2W, W2T, T2W, W2T, T2S, Si. 1
Tri-axial chest angular velocities.
Mariani et al. [64] Rule-based. Tri-axial foot accelerations. W2T, T2W. 1
Tri-axial foot angular velocities.
Frenken et al. [24] Rule-based. Pressure sensors on chair. St, S2W, W2T, T2W, W2T, Si. 1
Light laser distance.
St = Standing, S2W = Sit-to-Walk, W2T = Walk-to-Turn, T2W = Turn-to-Walk, W2T = Walk-to-Turn, T2S = Turn-to-Sit, and Si = Sitting.
Sensors 2023, 23, 3426 23 of 31

Other studies that use information from inertial sensors to automatically identify
postural transitions differ from previous proposals with respect to the sensor’s location—
see Silva et al. for the thigh [48], Beyea et al. for the upper body [51], Najafi et al. for the
chest [62] and Mariani et al. for the foot [64].
Regarding studies that use other technologies for the instrumentation of TUG, they
use algorithms created by the same authors, such as Ayena et al. [31] for acceleration and
foot pressure sensors, Tan et al. [57] for Kinect video, Frenken et al. [24] for environmental
sensors (chair with pressure sensors and laser) and Holzreiter et al. [38] for coordinates
of infrared markers for motion capture, the latter being the only one that uses Machine
Learning strategies for the automatic identification of postural transitions.
Finally, it can be observed that nine segmentation strategies automatically identify
seven postural transitions, which allowed twenty-eight studies to segment iTUG into the
six main sub-phases (standing, go walking, first turn, return walking, pre-sitting turn
and sitting). Three segmentation strategies only identified six postural transitions, which
allowed five studies to segment iTUG into five sub-phases, combining the pre-sitting turn
and sitting stage in a single sub-phase. Ayena et al. [31] segmented TUG into two postural
transitions, allowing the identification of the sub-phases of standing, walking and sitting of
the TUG test. Silva et al. [48] only used their segmentation strategy to identify the postural
transitions of the first half of the test (before the return march), which, however, can be
replicated for the return stage. Mariani et al. [64] only identified the turning sub-phases of
the TUG.

3.5. Features Extracted from iTUG

Although the first stage of the iTUG analysis begins with the segmentation of the
sub-phases or identification of the postural transitions, different methodologies can then
extract characteristics, indices or parameters that characterize them quantitatively and
objectively. Table 8 summarizes the characteristics used in the reviewed articles.

Table 8. Main features extracted from the selected studies.

Feature Relevant Parameters Studies

Total iTUG duration. Clinical Score. 19
Duration per sub-phase of Standing, walking (go and return), turning 23
iTUG. (3-meter and pre-sitting), sitting.
Spatiotemporal gait parameters. Cadence, stride time, step time, stride 20
length, step length, gait velocity, number
of steps, double support time, stride time
variability, gait index variability.
Statistical descriptors in the Maximum value, minimum value, mean, 9
acceleration time domain standard deviation, kurtosis, root mean
by sub-phase. square value, entropy.
Statistical descriptors in the Maximum value, minimum value, mean, 17
time domain of angular velocity standard deviation, kurtosis, root mean
by sub-phase. square value, entropy.
Statistical descriptors in time do- Main joints measured: thorax (chest or ster- 14
main of mobility ranges in de- num), cervical (C7), lumbar (L4–L5), pelvis.
grees per sub-phase.
Acceleration frequency domain Fast Fourier transform. 2
descriptors per sub-phase.
iTUG = Instrumented Timed Up and Go.
Sensors 2023, 23, 3426 24 of 31

As indicated in Table 8, 23 studies use the duration parameters of each sub-phase of

the segmented TUG, being the most used characteristic in studies for the characterization
of elderly populations, intervention strategies and application of risk prediction tools.
The gait sub-phase is the most analyzed, with 20 studies that extract and use gait cycle
spatiotemporal parameters. Moreover, 19 studies used the total time of the test as a
predictor of the risk of falls, 17 studies used statistical characteristics in the time domain
of the angular velocity and 9 studies used statistical characteristics in the time domain
of the acceleration, since the latter is more used to establish the patterns of postural
transitions in segmentation algorithms and estimation of spatiotemporal gait parameters.
Fourteen studies used statistical descriptors of mobility ranges or spatial orientation of
body segments, where the thoracic, cervical, lumbar and pelvic segments were the most
used. Two studies used statistical properties of acceleration from its Fourier transform,
being the least used.

3.6. Main Clinical Results from the Selected Studies

In this section, the main results provided by the articles included in this review will
be presented; the information will be organized by the type of participant included in the
study, described in Table 4.
In elderlies without associated pathologies, it has been shown that iTUG, through
the different technological proposals, allows extracting characteristics automatically, pro-
viding significant information on the temporal and velocity variables of each sub-phase,
allowing the identification of groups with big or low RoF [34,43,48]. In addition, it al-
lows to evaluate responses to the different treatments for improving balance, cognition
and performance in dual tasks [39]. Furthermore, it can be combined with rating scales and
allows to identification sub-clinical gait impairments [33,42]. On the other hand, it allows
objectively and quickly identifying postural transitions during iTUG, analyzing sensory
deficits and assessing the performance of the vestibular and somatosensory system [50,65].
iTUG and its segmentation allow characterizing the main sub-phases, identifying
alterations in its execution in elderlies, such as the specific performance of people in the
pre-sitting turning and sitting sub-phases, and relating it to other variables, such as cog-
nitive and motor function [21,63,67]. Furthermore, iTUG combined machine learning
algorithms, such as linear regression models [58,59] and other algorithms [24,61], identify-
ing the decrease in the balance of highly functional elderlies, with a precision of 70% and a
relationship of 80% for the identification of poor balance in tests performed one year later
using one or more arrays of inertial sensors [58,59,61].
In community elderlies, it was shown that iTUG has adequate precision when com-
pared with tests and community clinical assessment scales, suggesting that iTUG is a fast,
economical tool, easy to administer in person or remotely [32], allowing it to be a reliable
option for objective, unsupervised and unobtrusive measurement of balance in the clinical
setting or at home [68]. In addition, it has been used to determine the functional decline as-
sociated with aging and determine specific differences in sub-phases related to gender [36],
in the relationship between motor functioning and global cognitive function [41] and in
residence contexts, where a study was able to correlate features of some sub-phases with
the age of the participants [69].
In elderlies with PD, some studies have shown that an instrumented insole with FRS
located on the heel and toes with a reduced number of sensors may be sufficient to estimate
the risk index for falls during walking, being able to calculate the variation in gait and
balance parameters [22,31]. Furthermore, it has been shown that it is possible to reduce the
number of inertial sensors from 17 to 4 to achieve test segmentation [11]. On the other hand,
in elderlies with stroke, the use of instrumented insoles in rehabilitation was evaluated,
being capable of measuring the level of gait and TUG, providing details in the movement
analysis [37].
Sensors 2023, 23, 3426 25 of 31

The iTUG test can also be used in clinical practice to assess the effects of pharmacology
and physical therapy in people with PD, such as the effects that L-Dopa may have on gait
parameters and freezing on gait (FOG), as well as quantify and measure FOG [40,44].
On the other hand, studies indicated that iTUG could be correlated with scores from
different balance and gait scales, demonstrating that an instrumented scale can reveal
deficits in turnings of people with PD (severe and mild), as well as instrumentation in
stages with or without medication making it possible to predict falls [45,49].
Another study determined the intra-rater, inter-rater, and test-retest reliability of
iTUG in people with PD, proving excellent to good for the total duration and turning
durations [18].
Regarding variables identified for people with PD through the instrumentation of
TUG, studies explored gait speed in the clinic and home and the execution of iTUG at a
fast speed, demonstrating that the related parameters during walking and turning showed
strong correlations with the stage of the disease and that the application of the iTUG
procedure at a fast speed allows the identification of movement deficits in mild to moderate
stages, while the correlation in the parameters of the standing and sitting phases could
determine the level of automation of the movements and the kinematic parameters of iTUG
can have the potential to reflect functioning in movement execution [55,70].
On the other hand, iTUG measurements obtained from trunk angular velocity during
the turning and standing phases adequately reflect dynamic balance in people with PD [60].
In the study carried out with the mTUG through the Kinectic system for Xbox One, it was
possible to determine that the length of the first step can be significantly associated with
the motor analysis scale implemented [57].
In addition, another study detected four phases of the test in groups of people with PD
with and without medication through the use of sensors in the shoe during the execution
of iTUG, where temporal variables proved to be the most relevant ones [64].
Finally, it has been found that iTUG is useful to analyze the gait patterns during
the execution of the test applicable to healthy elderlies or with different gait disorders
analyzing the correlations of the go and return sub-phases [23,46]. In elderlies with a frailty
index greater than 3.95 according to the Fried scale, it was possible to observe that there
are differences in acceleration signals and angular velocities of the trunk, allowing a more
sensitive differentiation between frail and non-frail groups than only the TUG duration
variable, which is traditionally used [52–54]. In elderlies with peripheral neuropathy and
diabetes, the iTUG test allows them to identify and monitor postural transitions [62].

4. Discussion
The TUG test is a common tool for assessing mobility and fall risk in older adults
that uses the time to identify from a global perspective the RoF; however, it is not able,
in general, to incorporate the information of each sub-task performed in the evaluation,
such as standing, walking, turning and sitting. The iTUG is a modified version of TUG
that incorporates wearable sensors to capture additional objective data about gait, balance
and other factors that may contribute to RoF.
This systematic review aimed to examine the currently existing evidence on the
segmentation and analysis proposals of iTUG, the type of technologies used, the variables
acquired and how these measures allow the specific detection of impairments in older
adults with or without associated pathologies.
Previous systematic reviews related to the traditional TUG have been found that
account for the different uses of the test alone. On the one hand, there are reviews in
which TUG has been included as an assessment tool to identify changes in therapeutic
interventions [28] and as an instrument to measure RoF [2,5], showing the applicability
of the test in different clinical contexts. On the other hand, reviews were found that were
related to different psychometric properties of the traditional TUG in different populations
and with good reliability and validity values [25–27], showing that TUG has stable relative
sensitivity when applied to older people in community settings [5], excellent intra-rater
Sensors 2023, 23, 3426 26 of 31

and inter-rater reliability and good construct validity, and is sufficiently sensitive to detect
small changes in basic functional mobility after stroke [16] and adequate reliability and
validity in people with PD [25].
However, it has been shown to have limited ability to predict falls in older people at
high risk [8], as well as inconclusive results in its ability to predict falls after stroke [16].
It has been recommended that its application with other RoF measurement tools could
increase its predictive value [5]. It has also been mentioned that its predictive value
increases when instrumented (iTUG) [8,25]. In addition, we found a systematic review
that addresses the instrumentation through the use of inertial sensors of different scales
used to measure the risk of falls, including TUG, whose results account for the different
characteristics extracted, the positioning of the sensors and the predictive value of each of
them on the risk of falls [71]. However, to our knowledge, no similar systematic reviews
have been found addressing the specific outcomes described in the present review using
iTUG. These results are consistent with what was found in the present review, pointing
out the high number of characteristics and parameters that can be obtained through the
use of technologies. It is important to highlight that although TUG has shown different
conclusions regarding its psychometric properties [4,16,25] both in older people and in
people with stroke and PD, it is a tool that is still considered in clinical guidelines, and is
commonly used, probably because of its easy implementation and because it requires little
equipment [4,8].
The results of this review show the wide variety of devices used for TUG instrumenta-
tion, the type of procedure used, the characteristics of the population and the parameters
that can be obtained from it. It demonstrates that all iTUG proposals, to a greater or lesser
extent, allow the extraction of characteristics and variables of the subject’s performance
during the procedure, increasing the test’s objectivity and providing additional values to
the total time.
During the review, several technological proposals for TUG instrumentation were iden-
tified. However, the largest number of proposals include the use of inertial
sensors [11,18,21,32,33,37,39,46,52–55,59–61,63–65,67–70]. Most proposals used a single
sensor on the lower back, which has been shown to identify specific spatiotemporal charac-
teristics, biomechanical elements of the pelvis and gait events during the execution of TUG,
since it is close to the center of gravity [72–74]. This is important since the reduced use of
sensors simplifies the implementation of iTUG in clinical settings.
On the other hand, it is important to highlight that extended versions of TUG were
found at 5 [11,61], 7 [43–45,49], and 10 meters [11,52–54,61], delivering more information
about gait characteristics together with the features extracted from the different transitions.
It has been reported that iTUG using inertial sensors in its extended 7-meter version for
people with stroke has shown excellent test-retest reliability [75]. However, it is important
to consider new RoF labels for different populations, since most studies have analyzed the
psychometric properties of the 3-meter version.
The different segmentation proposals allow researchers and clinicians to select TUG
sub-phases according to their desired measurement objectives, which have provided a
better analysis of TUG performance, increasing its predictive value. It is interesting to note
that the segmentation of six sub-phases allows the identification of the last turn and the
final stand phase before sitting. With this number of phases, it is possible to identify specific
problems in elderlies with balance or sensory impairments, or the use of compensatory
strategies [11,18,40,43,46,50,61,67,69].
The variables obtained by iTUG can be analyzed from a clinical point of view and
how this information contributes to therapeutic decisions within the rehabilitation process.
For example, in people with PD, it has been shown that they present alterations in the
turning sub-phases, mainly associated with freezing [44]. The analysis of turning during
iTUG segmentation allows healthy subjects to be differentiated from people with different
pathologies so that the variables that can be analyzed during the turn not only provide
information on the effects of aging [13]. This is of great clinical relevance, since iTUG could
Sensors 2023, 23, 3426 27 of 31

diagnose certain impairments or dysfunctions that lead to a loss of functional capacity

and early implementation of rehabilitation strategies. Furthermore, people with advanced
Parkinson’s disease have longer performance times in the non-instrumented TUG [76].
Still, the use of iTUG has made it possible to identify at different stages of the disease the
variables in each of the sub-phases that are most altered, thus increasing its sensitivity
in early stages [19]. Even the turning variables during iTUG execution are much more
important in the case of pathologies that consider asymmetry during gait, as in people with
stroke and PD [13].
The variables obtained with iTUG related to trunk movement could be clinically
important for people with stroke since, in people with hemiparesis, it has been reported
that those subjects who have experienced one or more falls present a greater sway of the
center of mass in mediolateral and anteroposterior directions than those subjects who have
not experienced falls. In addition, they present a longer duration in performing the sitting
sub-phase when standing during the non-instrumented TUG. This time has been reported
as an indicator of RoF [77].
The instrumentation of TUG not only allows observing more variables than just the
total time of execution of the test, but also allows identifying the existing problems in each
sub-task, thus optimizing the planning of rehabilitation plans and their follow-up, both in
clinical and community contexts.
However, due to a large number of studies and their variability in designs, as well
as the diverse implementation strategies of iTUG used and the varied segmentation or
sub-stages identified, it is difficult to compare studies and extrapolate their findings to
clinical scenarios. In future studies, it may be considered to standardize an implementation
strategy based on the original protocol described by Podsiadlo for iTUG [12], as well as
analyzing the highest amount of segmentation (see Table 7). This would allow for identify-
ing more specific characteristics of the phases of sitting to standing, turning and standing
to sitting, or an extended version of iTUG that allows for more gait characteristics to
be extracted [43–45,52–54]. On the other hand, this review did not identify any studies
on the predictive validity of iTUG, which should be considered, since one of the major
problems presented by the traditional TUG is its low capacity to predict falls in some
populations [8,16]. Studies comparing and correlating iTUG with other clinical scales with
greater predictive value should also be conducted.
On the other hand, some methodological limitations of the articles included in this
systematic review are the diversity of study designs. However, differences are found
between analyzed populations, not only in terms of their features, but also in terms of
sample sizes, which could interfere with the extrapolation or generalization of the results.
The reviewed studies focus on the RoF analysis with iTUG in populations between 65 and
75 years old (see Figure 3). Future studies should include balanced samples of older adults
above 75, recognizing that mobility and static and dynamic balance change with aging.
This systematic review has some limitations. In the first place, only five databases
were searched (Pubmed, SciELO, IEEE, WOS and Elsevier), so we do not know if there is
evidence in other databases. Furthermore, the search only considered articles in English
and Spanish, and did not include other languages. Moreover, this review considers iTUG
in subjects over 65 years of age; therefore, evidence of the application of this test in other
age ranges or with other clinical tests that measure RoF was omitted. Furthermore, from a
methodological point of view, the wide variety of study designs included did not allow a
meta-analysis process to be performed; however, it was possible to incorporate the greatest
number of technological proposals used for TUG instrumentation, which allows us to meet
the aims of this review.
Finally, the results of this review show the possibility of increasing the predictive value
of TUG through instrumentation by providing a more significant number of characteristics
and parameters related to the subjects’ performance, allowing for objectively identifying
alterations associated with static and dynamic balance disorders. Furthermore, this instru-
mentation allows for the increased applicability of the test not only as a measurement tool,
Sensors 2023, 23, 3426 28 of 31

but also as a possible diagnostic or predictive tool for RoF, as well as continuous moni-
toring of users during the rehabilitation process and their reintegration into community
settings. The incorporation of technology or instrumentation could not only be applied to
TUG, but could also be incorporated into other assessment scales or tests, which, along
with the development of mobile applications and telemonitoring, could expand access
for users residing in remote areas and therapeutic teams that do not have all the tools
for implementing rehabilitation processes. This opens up the possibility of future lines of
research in the clinical validation processes of technological proposals in different contexts
and populations, the description of their psychometric properties and lines of technological
development and data processing.

5. Conclusions
The iTUG test provides objective evaluations and guides treatment, making it a
valuable tool for assessing the risk of falls in older adults. In this systematic review, the most
used technology was inertial sensors, and healthy elderlies and elderlies with Parkinson’s
disease were the most analyzed participants. The algorithm proposed by Weiss et al.
was the most used for automatic segmentation. The iTUG test offers an economical and
accessible alternative to increase the predictive value of the TUG test, identifying different
variables, and can be used in clinical, community and home settings. The review’s findings
highlight the potential benefits of incorporating technological approaches to increase the
predictive value of TUG and improve RoF assessments.

Author Contributions: Conceptualization, P.O.-B., B.G., P.A. and R.C.-d.-l.-C.; methodology, P.O.-B.,
B.G., S.L.-M. and R.C.-d.-l.-C.; software, P.O.-B., B.G. and S.L.-M.; validation, P.A., S.L.-M. and R.C.-d.-
l.-C.; formal analysis, P.O.-B., B.G. and S.L.-M.; investigation, P.O.-B. and B.G.; resources, P.O.-B., P.A.
and R.C.-d.-l.-C.; data curation, P.O.-B., B.G. and S.L.-M.; writing—original draft preparation, P.O.-B.
and B.G.; writing—review and editing, P.A., S.L.-M. and R.C.-d.-l.-C.; visualization, P.O.-B., B.G., P.A.
and R.C.-d.-l.-C.; supervision, P.A. and R.C.-d.-l.-C.; project administration, P.A. and R.C.-d.-l.-C.;
funding acquisition, P.O.-B., P.A. and R.C.-d.-l.-C. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by the Vice-rectory for Research and Development (VRID) of the
Universidad de Concepción, titled “Clinical Validation of a Human Movement Analysis System Based
on Inertial Measurement Sensors in the Chilean population at risk of falls.”, Code: 220.092.003-M.
Universidad de Concepción, Concepción, Chile.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Sharif, S.I.; Al-Harbi, A.B.; Al-Shihabi, A.M.; Al-Daour, D.S.; Sharif, R.S. Falls in the elderly: Assessment of prevalence and risk
factors. Pharm. Pract. (Granada) 2018, 16, 1206. [CrossRef] [PubMed]
2. Mak, T.C.; Wong, T.W.; Ng, S.S. Visual-related training to improve balance and walking ability in older adults: A systematic
review. Exp. Gerontol. 2021, 156, 111612. [CrossRef]
3. Phelan, E.A.; Ritchey, K. Fall prevention in community-dwelling older adults. Ann. Intern. Med. 2018, 169, ITC81–ITC96.
[CrossRef]
4. Ambrose, A.F.; Paul, G.; Hausdorff, J.M. Risk factors for falls among older adults: A review of the literature. Maturitas 2013,
75, 51–61. [CrossRef] [PubMed]
5. Park, S.H. Tools for assessing fall risk in the elderly: A systematic review and meta-analysis. Aging Clin. Exp. Res. 2018, 30, 1–16.
[CrossRef]
6. Marques, N.R.; Spinoso, D.H.; Cardoso, B.C.; Moreno, V.C.; Kuroda, M.H.; Navega, M.T. Is it possible to predict falls in older
adults using gait kinematics? Clin. Biomech. 2018, 59, 15–18. [CrossRef]
7. Gschwind, Y.J.; Kressig, R.W.; Lacroix, A.; Muehlbauer, T.; Pfenninger, B.; Granacher, U. A best practice fall prevention exercise
program to improve balance, strength/power, and psychosocial health in older adults: Study protocol for a randomized controlled
trial. BMC Geriatr. 2013, 13, 1–13. [CrossRef] [PubMed]
Sensors 2023, 23, 3426 29 of 31

8. Barry, E.; Galvin, R.; Keogh, C.; Horgan, F.; Fahey, T. Is the Timed Up and Go test a useful predictor of risk of falls in community
dwelling older adults: A systematic review and meta-analysis. BMC Geriatr. 2014, 14, 1–14. [CrossRef]
9. Zasadzka, E.; Borowicz, A.M.; Roszak, M.; Pawlaczyk, M. Assessment of the risk of falling with the use of timed up and go test in
the elderly with lower extremity osteoarthritis. Clin. Interv. Aging 2015, 10, 1289. [CrossRef]
10. Persson, C.U.; Danielsson, A.; Sunnerhagen, K.S.; Grimby-Ekman, A.; Hansson, P.O. Timed Up & Go as a measure for longitudinal
change in mobility after stroke–Postural Stroke Study in Gothenburg (POSTGOT). J. Neuroeng. Rehabil. 2014, 11, 1–7.
11. Nguyen, H.; Lebel, K.; Boissy, P.; Bogard, S.; Goubault, E.; Duval, C. Auto detection and segmentation of daily living activities
during a Timed Up and Go task in people with Parkinson’s disease using multiple inertial sensors. J. Neuroeng. Rehabil. 2017,
14, 1–13. [CrossRef]
12. Podsiadlo, D.; Richardson, S. The timed “Up & Go”: A test of basic functional mobility for frail elderly persons. J. Am. Geriatr.
Soc. 1991, 39, 142–148. [PubMed]
13. Vervoort, D.; Vuillerme, N.; Kosse, N.; Hortobágyi, T.; Lamoth, C.J. Multivariate analyses and classification of inertial sensor data
to identify aging effects on the Timed-Up-and-Go Test. PLoS ONE 2016, 11, e0155984. [CrossRef] [PubMed]
14. Mancilla, E.; Valenzuela, J.; Escobar, M. Rendimiento en las pruebas “Timed Up and Go” y “Estación Unipodal” en adultos
mayores chilenos entre 60 y 89 años. Rev. Méd. Chile 2015, 143, 39–46. [CrossRef]
15. Flansbjer, U.B.; Holmbäck, A.M.; Downham, D.; Patten, C.; Lexell, J. Reliability of gait performance tests in men and women with
hemiparesis after stroke. J. Rehabil. Med. 2005, 37, 75–82. [PubMed]
16. Hafsteinsdottir, T.B.; Rensink, M.; Schuurmans, M. Clinimetric properties of the Timed Up and Go Test for patients with stroke: A
systematic review. Top. Stroke Rehabil. 2014, 21, 197–210. [CrossRef] [PubMed]
17. Dal Bello-Haas, V.; Klassen, L.; Sheppard, M.S.; Metcalfe, A. Psychometric properties of activity, self-efficacy, and quality-of-life
measures in individuals with Parkinson disease. Physiother. Can. 2011, 63, 47–57. [CrossRef]
18. Van Lummel, R.C.; Walgaard, S.; Hobert, M.A.; Maetzler, W.; Van Dieën, J.H.; Galindo-Garre, F.; Terwee, C.B. Intra-rater,
inter-rater and test-retest reliability of an instrumented timed up and go (iTUG) test in patients with Parkinson’s disease. PLoS
ONE 2016, 11, e0151881. [CrossRef]
19. Salarian, A.; Horak, F.B.; Zampieri, C.; Carlson-Kuhta, P.; Nutt, J.G.; Aminian, K. iTUG, a sensitive and reliable measure of
mobility. IEEE Trans. Neural Syst. Rehabil. Eng. 2010, 18, 303–310. [CrossRef]
20. Caronni, A.; Sterpi, I.; Antoniotti, P.; Aristidou, E.; Nicolaci, F.; Picardi, M.; Pintavalle, G.; Redaelli, V.; Achille, G.; Sciumè, L.;
et al. Criterion validity of the instrumented Timed Up and Go test: A partial least square regression study. Gait Posture 2018,
61, 287–293. [CrossRef]
21. Williams, J.M.; Nyman, S.R. Age moderates differences in performance on the instrumented timed up and go test between people
with dementia and their informal caregivers. J. Geriatr. Phys. Ther. 2021, 44, E150–E157. [CrossRef] [PubMed]
22. Ayena, J.C.; Otis, M.J.D. Dimensional reduction of balance parameters in risk of falling evaluation using a minimal number of
force-sensitive resistors. Int. J. Occup. Saf. Ergon. 2022, 28, 507–518. [CrossRef] [PubMed]
23. Gasparutto, X.; Gueugnon, M.; Laroche, D.; Martz, P.; Hannouche, D.; Armand, S. Which functional tasks present the largest
deficits for patients with total hip arthroplasty before and six months after surgery? A study of the timed up-and-go test phases.
PLoS ONE 2021, 16, e0255037. [CrossRef] [PubMed]
24. Frenken, T.; Brell, M.; Gövercin, M.; Wegel, S.; Hein, A. aTUG: Technical apparatus for gait and balance analysis within
component-based Timed Up & Go using mutual ambient sensors. J. Ambient Intell. Humaniz. Comput. 2013, 4, 759–778.
25. Mollinedo, I.; Cancela, J.M. Evaluation of the psychometric properties and clinical applications of the Timed Up and Go test in
Parkinson disease: A systematic review. J. Exerc. Rehabil. 2020, 16, 302. [CrossRef]
26. Gafner, S.C.; Allet, L.; Hilfiker, R.; Bastiaenen, C.H.G. Reliability and diagnostic accuracy of commonly used performance tests
relative to fall history in older persons: A systematic review. Clin. Interv. Aging 2021, 16, 1591. [CrossRef]
27. Christopher, A.; Kraft, E.; Olenick, H.; Kiesling, R.; Doty, A. The reliability and validity of the Timed Up and Go as a clinical tool
in individuals with and without disabilities across a lifespan: A systematic review: Psychometric properties of the Timed Up and
Go. Disabil. Rehabil. 2021, 43, 1799–1813. [CrossRef]
28. Wiener, J.; McIntyre, A.; Janssen, S.; Chow, J.T.; Batey, C.; Teasell, R. Effectiveness of high-intensity interval training for fitness
and mobility post stroke: A systematic review. PM&R 2019, 11, 868–878.
29. Page, M.J.; Moher, D.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan,
S.E.; et al. PRISMA 2020 explanation and elaboration: Updated guidance and exemplars for reporting systematic reviews. BMJ
2021, 372, n160. [CrossRef]
30. Page, M.J.; McKenzie, J.E.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.;
Brennan, S.E.; et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. Syst. Rev. 2021,
10, 1–11. [CrossRef]
31. Ayena, J.C.; Otis, M.J.D. Validation of minimal number of force sensitive resistors to predict risk of falling during a timed up and
go test. J. Med Biol. Eng. 2020, 40, 348–355. [CrossRef]
32. Bergquist, R.; Nerz, C.; Taraldsen, K.; Mellone, S.; Ihlen, E.A.; Vereijken, B.; Helbostad, J.L.; Becker, C.; Mikolaizak, A.S. Predicting
advanced balance ability and mobility with an instrumented timed up and go test. Sensors 2020, 20, 4987. [CrossRef] [PubMed]
Sensors 2023, 23, 3426 30 of 31

33. Weiss, A.; Mirelman, A.; Buchman, A.S.; Bennett, D.A.; Hausdorff, J.M. Using a body-fixed sensor to identify subclinical gait
difficulties in older adults with IADL disability: Maximizing the output of the timed up and go. PLoS ONE 2013, 8, e68885.
[CrossRef] [PubMed]
34. Guzmán, J.C.; Silva, R.G.; Guzmán-Venegas, R. Reproducibilidad de los tiempos de ejecución de la prueba de Timed Up and Go,
medidos con acelerómetros de smartphones en personas mayores residentes en la comunidad. Rev. Esp. Geriatr. Gerontol. 2017,
52, 249–252. [CrossRef]
35. Mellone, S.; Tacconi, C.; Chiari, L. Validity of a Smartphone-based instrumented Timed Up and Go. Gait Posture 2012, 36, 163–165.
[CrossRef] [PubMed]
36. Coni, A.; Mellone, S.; Colpo, M.; Bandinelli, S.; Chiari, L. Influence of age and gender on sensor-based functional measures: A
factor analysis approach. In Proceedings of the 2015 37th Annual International Conference of the IEEE Engineering in Medicine
and Biology Society (EMBC), Milan, Italy, 25–29 August 2015; pp. 5054–5057.
37. David, V.; Forjan, M.; Martinek, J.; Kotzian, S.; Jagos, H.; Rafolt, D. Evaluating wearable multimodal sensor insoles for
motion-pattern measurements in stroke rehabilitation—A pilot study. In Proceedings of the 2017 International Conference on
Rehabilitation Robotics (ICORR), London, UK, 17–20 July 2017; pp. 1543–1548.
38. Holzreiter, S.; Jennings, S. Programming language for motion analysis. Hum. Mov. Sci. 1996, 15, 497–508. [CrossRef]
39. Delbroek, T.; Vermeylen, W.; Spildooren, J. The effect of cognitive-motor dual task training with the biorescue force platform on
cognition, balance and dual task performance in institutionalized older adults: A randomized controlled trial. J. Phys. Ther. Sci.
2017, 29, 1137–1143. [CrossRef]
40. Dibilio, V.; Nicoletti, A.; Mostile, G.; Toscano, S.; Luca, A.; Raciti, L.; Sciacca, G.; Vasta, R.; Cicero, C.E.; Contrafatto, D.; et al.
Dopaminergic and non-dopaminergic gait components assessed by instrumented timed up and go test in Parkinson’s disease. J.
Neural Transm. 2017, 124, 1539–1546. [CrossRef]
41. Fastame, M.C.; Mulas, I.; Putzu, V.; Asoni, G.; Viale, D.; Mameli, I.; Pau, M. Motor proficiency as a correlate of coping in late adult
lifespan. An exploratory study. Anxiety Stress Coping 2022, 35, 687–700. [CrossRef] [PubMed]
42. Figueiredo, A.I.; Balbinot, G.; Brauner, F.O.; Schiavo, A.; Baptista, R.R.; Pagnussat, A.S.; Hollands, K.; Mestriner, R.G. SPARC
metrics provide mobility smoothness assessment in oldest-old with and without a history of falls: A case control study. Front.
Physiol. 2020, 11, 540. [CrossRef]
43. Mangano, G.R.; Valle, M.S.; Casabona, A.; Vagnini, A.; Cioni, M. Age-related changes in mobility evaluated by the timed up and
go test instrumented through a single sensor. Sensors 2020, 20, 719. [CrossRef] [PubMed]
44. Mancini, M.; Priest, K.C.; Nutt, J.G.; Horak, F.B. Quantifying freezing of gait in Parkinson’s disease during the instrumented
timed up and go test. In Proceedings of the 2012 Annual International Conference of the IEEE Engineering in Medicine and
Biology Society, San Diego, CA, USA, 28 August–1 September 2012; pp. 1198–1201.
45. King, L.A.; Mancini, M.; Priest, K.; Salarian, A.; Rodrigues-de Paula, F.; Horak, F. Do clinical scales of balance reflect turning
abnormalities in people with Parkinson’s disease? J. Neurol. Phys. Ther. 2012, 36, 25. [CrossRef]
46. Ishikawa, M.; Yamada, S.; Yamamoto, K.; Aoyagi, Y. Gait analysis in a component timed-up-and-go test using a smartphone
application. J. Neurol. Sci. 2019, 398, 45–49. [CrossRef]
47. Salarian, A.; Zampieri, C.; Horak, F.B.; Carlson-Kuhta, P.; Nutt, J.G.; Aminian, K. Analyzing 180° turns using an inertial system
reveals early signs of progression of parkinson’s disease. In Proceedings of the 2009 Annual International Conference of the IEEE
Engineering in Medicine and Biology Society , Minneapolis/St. Paul, MN, USA, 3–6 September, 2009; pp. 224–227.
48. Silva, J.; Sousa, I. Instrumented timed up and go: Fall risk assessment based on inertial wearable sensors. In Proceedings of the
2016 IEEE International Symposium on Medical Measurements and Applications (MeMeA), Benevento, Italy, 15–18 May 2016;
pp. 1–6.
49. Hoskovcová, M.; Dušek, P.; Sieger, T.; Brožová, H.; Zárubová, K.; Bezdíček, O.; Šprdlík, O.; Jech, R.; Štochl, J.; Roth, J.; et al.
Predicting falls in Parkinson disease: What is the value of instrumented testing in off medication state? PLoS ONE 2015,
10, e0139849. [CrossRef] [PubMed]
50. Hellmers, S.; Izadpanah, B.; Dasenbrock, L.; Diekmann, R.; Bauer, J.M.; Hein, A.; Fudickar, S. Towards an automated unsupervised
mobility assessment for older people based on inertial TUG measurements. Sensors 2018, 18, 3310. [CrossRef]
51. Beyea, J.; McGibbon, C.A.; Sexton, A.; Noble, J.; O’Connell, C. Convergent validity of a wearable sensor system for measuring
sub-task performance during the timed up-and-go test. Sensors 2017, 17, 934. [CrossRef]
52. Galan-Mercant, A.; Cuesta-Vargas, A. Clinical frailty syndrome assessment using inertial sensors embedded in smartphones.
Physiol. Meas. 2015, 36, 1929. [CrossRef] [PubMed]
53. Galán-Mercant, A.; Cuesta-Vargas, A.I. Differences in trunk kinematic between frail and nonfrail elderly persons during turn
transition based on a smartphone inertial sensor. BioMed Res. Int. 2013, 2013, 279197. [CrossRef]
54. Galán-Mercant, A.; Cuesta-Vargas, A.I. Differences in trunk accelerometry between frail and non-frail elderly persons in functional
tasks. BMC Res. Notes 2014, 7, 1–9. [CrossRef]
55. Van Uem, J.M.; Walgaard, S.; Ainsworth, E.; Hasmann, S.E.; Heger, T.; Nussbaum, S.; Hobert, M.A.; Mico-Amigo, E.M.;
Van Lummel, R.C.; Berg, D.; et al. Quantitative timed-up-and-go parameters in relation to cognitive parameters and health-
related quality of life in mild-to-moderate Parkinson’s disease. PLoS ONE 2016, 11, e0151997. [CrossRef]
Sensors 2023, 23, 3426 31 of 31

56. Walgaard, S.; Faber, G.S.; van Lummel, R.C.; van Dieën, J.H.; Kingma, I. The validity of assessing temporal events, sub-phases
and trunk kinematics of the sit-to-walk movement in older adults using a single inertial sensor. J. Biomech. 2016, 49, 1933–1937.
[CrossRef] [PubMed]
57. Tan, D.; Pua, Y.H.; Balakrishnan, S.; Scully, A.; Bower, K.J.; Prakash, K.M.; Tan, E.K.; Chew, J.S.; Poh, E.; Tan, S.B.; et al. Automated
analysis of gait and modified timed up and go using the Microsoft Kinect in people with Parkinson’s disease: Associations with
physical outcome measures. Med Biol. Eng. Comput. 2019, 57, 369–377. [CrossRef]
58. Sheehan, K.; Greene, B.; Cunningham, C.; Crosby, L.; Kenny, R. Early identification of declining balance in higher functioning
older adults, an inertial sensor based method. Gait Posture 2014, 39, 1034–1039. [CrossRef] [PubMed]
59. Seo, J.; Kim, T.; Lee, J.; Kim, J.; Choi, J.; Tack, G. Fall prediction of the elderly with a logistic regression model based on
instrumented timed up & go. J. Mech. Sci. Technol. 2019, 33, 3813–3818.
60. Picardi, M.; Redaelli, V.; Antoniotti, P.; Pintavalle, G.; Aristidou, E.; Sterpi, I.; Meloni, M.; Corbo, M.; Caronni, A. Turning and
sit-to-walk measures from the instrumented Timed Up and Go test return valid and responsive measures of dynamic balance in
Parkinson’s disease. Clin. Biomech. 2020, 80, 105177. [CrossRef] [PubMed]
61. Nguyen, H.P.; Ayachi, F.; Lavigne-Pelletier, C.; Blamoutier, M.; Rahimi, F.; Boissy, P.; Jog, M.; Duval, C. Auto detection and
segmentation of physical activities during a Timed-Up-and-Go (TUG) task in healthy older adults using multiple inertial sensors.
J. Neuroeng. Rehabil. 2015, 12, 1–12. [CrossRef] [PubMed]
62. Najafi, B.; Armstrong, D.G.; Mohler, J. Novel wearable technology for assessing spontaneous daily physical activity and risk of
falling in older adults with diabetes. J. Diabetes Sci. Technol. 2013, 7, 1147–1160. [CrossRef]
63. Mirelman, A.; Weiss, A.; Buchman, A.S.; Bennett, D.A.; Giladi, N.; Hausdorff, J.M. Association between performance on Timed
Up and Go subtasks and mild cognitive impairment: Further insights into the links between cognitive and motor function. J. Am.
Geriatr. Soc. 2014, 62, 673–678. [CrossRef] [PubMed]
64. Mariani, B.; Jiménez, M.C.; Vingerhoets, F.J.; Aminian, K. On-shoe wearable sensors for gait and turning assessment of patients
with Parkinson’s disease. IEEE Trans. Biomed. Eng. 2012, 60, 155–158. [CrossRef]
65. Gerhardy, T.; Gordt, K.; Jansen, C.P.; Schwenk, M. Towards using the instrumented timed up-and-go test for screening of sensory
system performance for balance control in older adults. Sensors 2019, 19, 622. [CrossRef]
66. Weiss, A.; Herman, T.; Plotnik, M.; Brozgol, M.; Maidan, I.; Giladi, N.; Gurevich, T.; Hausdorff, J.M. Can an accelerometer enhance
the utility of the Timed Up & Go Test when evaluating patients with Parkinson’s disease? Med. Eng. Phys. 2010, 32, 119–125.
[PubMed]
67. Weiss, A.; Mirelman, A.; Giladi, N.; Barnes, L.L.; Bennett, D.A.; Buchman, A.S.; Hausdorff, J.M. Transition between the Timed Up
and Go turn to sit subtasks: Is timing everything? J. Am. Med Dir. Assoc. 2016, 17, 864.e9–864.e15. [CrossRef]
68. Yu, L.; Zhao, Y.; Wang, H.; Sun, T.L.; Murphy, T.E.; Tsui, K.L. Assessing elderly’s functional balance and mobility via analyzing
data from waist-mounted tri-axial wearable accelerometers in timed up and go tests. BMC Med Inform. Decis. Mak. 2021, 21, 1–14.
[CrossRef]
69. Zarzeczny, R.; Nawrat-Szołtysik, A.; Polak, A.; Maliszewski, J.; Kiełtyka, A.; Matyja, B.; Dudek, M.; Zborowska, J.; Wajdman, A.
Aging effect on the instrumented Timed-Up-and-Go test variables in nursing home women aged 80–93 years. Biogerontology 2017,
18, 651–663. [CrossRef] [PubMed]
70. Toosizadeh, N.; Mohler, J.; Lei, H.; Parvaneh, S.; Sherman, S.; Najafi, B. Motor performance assessment in Parkinson’s disease:
Association between objective in-clinic, objective in-home, and subjective/semi-objective measures. PLoS ONE 2015, 10, e0124763.
[CrossRef]
71. Montesinos, L.; Castaldo, R.; Pecchia, L. Wearable inertial sensors for fall risk assessment and prediction in older adults:
A systematic review and meta-analysis. IEEE Trans. Neural Syst. Rehabil. Eng. 2018, 26, 573–582. [CrossRef]
72. Kose, A.; Peruzzi, A.; Cereatti, A.; Laudani, L.; Della Croce, U. Detection of heel strikes and toe-offs during gait using a single
inertial measurement unit attached to the waist. In Proceedings of the Second National Congress of Biongineering, Turin, Italy,
8–10 July 2010; Volume 233.
73. Weinberg, H. Using the ADXL202 in pedometer and personal navigation applications. Analog Devices AN-602 Appl. Note 2002,
2, 1–6.
74. Sejdić, E.; Lowry, K.A.; Bellanca, J.; Perera, S.; Redfern, M.S.; Brach, J.S. Extraction of stride events from gait accelerometry during
treadmill walking. IEEE J. Transl. Eng. Health Med. 2015, 4, 1–11. [CrossRef]
75. Wüest, S.; Massé, F.; Aminian, K.; Gonzenbach, R.; De Bruin, E.D. Reliability and validity of the inertial sensor-based Timed “Up
and Go” test in individuals affected by stroke. J. Rehabil. Res. Dev. 2016, 53, 599–610. [CrossRef]
76. Pelicioni, P.H.; Menant, J.C.; Latt, M.D.; Lord, S.R. Falls in Parkinson’s disease subtypes: Risk factors, locations and circumstances.
Int. J. Environ. Res. Public Health 2019, 16, 2216. [CrossRef]
77. Janssen, W.G.; Bussmann, H.B.; Stam, H.J. Determinants of the sit-to-stand movement: A review. Phys. Ther. 2002, 82, 866–879.
[CrossRef] [PubMed]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.