Physical mobility determinants among older adults: a scoping review of self-reported and performance-based measures

Abstract Objective To synthesise the available evidence on physical factors, such as muscle strength and power, body mass index and their association with older adults’ self-reported and performance-based mobility outcomes. Method This review followed the Askey and O'Malley framework. We systematically searched PubMed, EMBASE, PsychINFO, Web of Science, AgeLine, Allied and Complementary Medicine Database, and Cumulative Index to Nursing and Allied Health Literature databases, from Jan. 2000 to Jan. 2022. Teams of two reviewers independently conducted title, abstract, full-text screening, and data extraction using predefined inclusion and exclusion criteria. Result A total of 239 quantitative articles, mostly cross-sectional design, conducted in 32 countries were included in this review. We identified 18 physical factors significantly associated with mobility outcomes in the expected direction. Muscle strength, body composition, falls (number and history of), and chronic conditions (number of and type) were the most studied physical factors. Conclusion Older adults with muscle weakness, weight concerns, history of falls, and chronic conditions had poorer mobility outcomes, such as slower gait speed, poor balance, limited community mobility and poor driving outcomes compared to their counterparts. Studies exploring the role of physical factors on the use of an assisted device, transportation, or driving, are limited.


Introduction
Mobility has been described as a hallmark of healthy ageing [1,2].Mobility, defined as the ability to move from one place to another by one's self or with the use of assistive devices, or via transportation or driving, is fundamental for meaningful social interactions and community participation [3].Ageing-associated changes in sensory, cognitive and various physical structures (e.g.muscle) can pose threats to older adults' mobility [4].Mobility limitations have been described as self-reported difficulties in walking, performance deficits in objective mobility, and lack of access to assistive mobility devices, transportation, or driving [5].Mobility limitations increase the risk of disability, falls, hospitalisation, mortality and decreased quality of life [1].Older adults with mobility limitations often require assistance with their activities of daily living, leading to additional healthcare costs.For instance, Hardy et al. [6] reported that the total annual healthcare cost was $2773 higher in older adults reporting difficulty walking one-quarter a mile as compared to those with no difficulty.This increase in healthcare costs associated with mobility limitations warrants the need to explore the factors associated with mobility among the ageing population.
Factors influencing mobility are often multifactorial.Webber et al. [3], in the Conical Model of Theoretical Framework for Mobility in Older Adults (henceforth referred to as the Conical Model), described mobility determinants to include cognitive, environmental, financial, personal, physical, psychological, and social factors.The Conical Model provides a holistic perspective on the relationship between these mobility determinants and recognises that these determinants will have different levels of relevance depending on the older individual's living situation and capacities to navigate the seven mobility zones -the room where one sleeps, the home, the outdoor areas surrounding the home, the neighbourhood, the area in the community, within one's country and the world.For instance, mobility becomes more complex as one moves away from home and neighbourhood, making it difficult for older adults to go out and visit with friends and family or continue doing their activities independently.The Conical Model describes the interrelationship between mobility determinants and how they compound the complexity of mobility but does not indicate how each factor within each determinant influences mobility, making it difficult for researchers and clinicians to use this model to improve mobility in older adults as they move away from their home.To assess the usefulness and increase the practical use of this model, we conducted a series of scoping reviews for each determinant to identify factors within each determinant and describe their associations with mobility outcomes.This paper focuses on the physical mobility determinants in older adults.
Physical factors, such as muscle strength and power, proprioception, range of motion, balance, and comorbidity, have been found to influence older adults' mobility [7].Several studies have shown that age-related physical changes, such as reduced muscle strength and muscle power (defined as the ability of a body to perform an activity quickly and repeatedly by applying a given force -e.g.fast leg kicks [7]), impact mobility [7,8].For example, after age 40 years, people typically lose 8 percent or more of their muscle mass each decade, a process that accelerates significantly after age 70 [9].Sensory impairments, such as visual or hearing impairment and a sedentary lifestyle in older adults, can contribute to the decline in mobility [10].Further, the severity of the mobility decline is compounded when more than one physical determinant is impaired [11].Age-related physical changes are early indicators of mobility limitations.Although the age-related physical changes can be easily identified and defined by clinicians and researchers, how these physical factors interact individually or in combination to influence mobility is complex and multifactorial, highlighting the need to describe the associations between physical factors and mobility in older adults.
Reviews exploring the associations between physical factors and mobility outcomes among older adults are limited; most reviews report the risk factors for self-reported mobility limitations in older adults.While Yeom and colleagues [12] used the social-ecological model to describe risk factors for mobility limitations in older adults, Frieberger et al. [1] described physical age-related changes as risk factors for mobility decline.Both reviews reported some physical factors, including obesity, physical activity, muscle strength, and power, as risk factors for mobility decline, but their reviews are narrative rather than systematic.A systematic review of the literature has been published [13]; however, it focussed on the association between self-reported risk factors and self-reported mobility limitations.While these reviews have highlighted the risk factors for self-reported mobility decline in older adults, not all physical factors or mobility outcomes (e.g.gait speed, balance, use of assistive devices, driving and transportation) were included.
The combined use of self-reported and performancebased mobility assessment has been recommended because each measure assesses different aspects of older adults' physical functioning [14], providing unique and critical information, and ultimately complementing each other.For instance, while self-reported measures capture an individual's perceptions of mobility, performance-based measures capture an individual's real-time mobility ability [15].No systematic literature review has examined all physical factors on self-reported and performance-based mobility outcomes in older adults (60 years and older).Our review fills this gap by systematically and comprehensively describing the association between each physical factor and mobility outcomes in older adults.This review, alongside our previous reviews on the associations between self-reported and performancebased mobility outcomes, and environmental, personal, and financial factors [16]; and cognitive, psychological, and social factors [17], is needed to advance the use of the Conical Model [3] in clinical and research practice.This paper aims to synthesise the available evidence on physical determinants of the Conical Model and their association with self-reported and performance-based mobility outcomes in older adults.

Methods
This scoping review was guided by the methodological framework proposed by Arksey and O'Malley [18] and Levac et al. [19].This methodology is appropriate because physical factors influencing mobility are complex and heterogeneous; exploring the extent, range, and nature of available research on the associations between physical factors and selfreported and performance-based mobility outcomes in older adults will enable us to identify research gaps in the literature [20].Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRISMA-Scr) guided the reporting of this scoping review [21].The protocol was registered with the Open Science Framework: https://doi.org/10.17605/OSF.IO/7Y5VG.

Step 1: Identifying the research questions
The two research questions of this scoping review are: (a) what is the available evidence for the physical factors related to mobility in older adults (60 years and older); and (b) what are the associations between physical factors and mobility self-reported and performance-based outcomes in older adults 60 years and older?We followed Webber et al.'s [3] definition of mobility: the ability to move from one place to another either independently or with assistance including mobility aids (e.g.cane, wheelchair, walkers) and use various modes of transportation (e.g.car, bus, bicycle).We defined an older adult as a person who is 60 years and older [22].Based on the literature, we defined physical mobility determinants as those physiological functions of the body systems, including the musculoskeletal system, respiratory system, cardiovascular system, nervous and sensory system, as well as other physical factors (e.g.physical activity and exercise levels), and body composition.

Step 2: Identification of relevant studies
We developed the search strategy in consultation with a health science librarian.We sent our search strategy to the health science librarian.She reviewed our search strategy and provided feedback and advice on using Medical Subject Heading (MeSH) for each database.We pilot-tested the search strategy with the librarian on PubMed.Broadly, our search concepts include physical factors (e.g.muscle strength, muscle power, number of falls, physical activity levels), mobility outcomes (self-reported -e.g.mobility limitations, life-space mobility and performance-based -e.g.gait speed, balance, lower limb function test) and older adults (e.g.ag � ing, older people).See Table 1 for a complete search strategy.Seven databases (PubMed, EMBASE, AMED, CINAHL, Psych INFO, Web of Science, and AgeLine) were searched from 2000 to Jan. 2022; the year 2000 was chosen because the impact of baby boomers on health outcomes would be most prominent across the developed countries [23].We hand searched the reference lists for additional articles that met our inclusion criteria were included.

Step 3: Study selection
Retrieved citations were exported from each database and imported into Rayyan QCRI# [24], and duplicates were removed.We selected studies in two stages: title/abstract and full-text screening, using the inclusion and exclusion criteria in Box 1. Four raters independently performed a pilot of the title and abstract screening of the first 100 articles to determine inter-rater reliability.The Kappa agreement was 0.74, indicating moderate agreement; based on this, we conducted the title and abstract screening in pairs and resolved any conflicts during research meetings.Similarly, a pilot full-text screening was conducted independently by all four raters, and the Kappa agreement was 0.98, indicating a high magnitude of agreement [25].Therefore, full-text screening was divided evenly between the four raters.Disagreements were resolved in research meetings and discussions with the senior authors.

Step 4: Charting the data
We adapted a data extraction sheet from a previous review [26], pilot-tested and refined it based on the feedback during a research meeting.We extracted the following information: authors' name, the country in which the study was conducted, study aim(s)/research question(s)/hypothesis or hypotheses, the setting where participants were recruited, type of study (qualitative, quantitative, mixed-method), study design, study population (older adults with no defined conditions and those with a health condition, e.g.stroke), sample size, participants' mean age and sex distribution, physical factors being studied, type of mobility outcome used, and study findings related to our review questions.

Step 5: Collecting, summarising and reporting the result
We first listed and described all the physical factors being studied from the included studies.Second, we reported the associations between each physical factor and the mobility outcomes.Specifically, we presented the associations per analysis rather than per article because it enabled us to describe the association distinctly, as one article may report multiple associations between different physical factors and outcomes.For instance, one article reported the association between multiple physical factors b.The outcome or self-reported measure was related to functional decline in activities of daily living or instrumental activities of daily living without any specific measure of mobility.c.They described physical activity or exercise (except walking) as a form of mobility d.They were exercise intervention studies that showed the effect of exercise intervention on mobility outcomes listed above; for instance, older adults in the exercise improved in their gait speed but those in the non-exercise group did not improve.
(Body Mass Index, number of chronic conditions and muscle strength) and multiple mobility outcomes (walking distance, walking speed, self-reported inability to walk quarter a mile); thus, nine distinct findings were extracted.This type of reporting allowed us to understand the distinct associations between physical factors and mobility outcomes at a granular level.We considered whether a factor negatively or positive significantly associated with mobility outcome if the article reported such factor as significant based on their p-value.

Result
A total of 12,679 citations were identified through the database searches.After removing duplicates, 9786 underwent Reports not retrieved (n = 0) Reports assessed for eligibility (n = 567) Reports excluded: title and abstract screening, excluding 9219.The remaining 567 articles underwent full-text screening, and 239 articles were included in this review (Figure 1).

Association between physical factors and mobility outcomes
Eighteen physical factors were identified, and the most studied was muscle strength (n ¼ 84, 19.4%), followed by body composition (e.g.Body Mass Index; n ¼ 83, 19.2%), chronic conditions (number and type; n ¼ 45 (10.4%) and falls (number and history of falls; n ¼ 44, 10.2%) (see Table 3).Only significant associations are described in this section.The non-significant associations, including p-values, odds ratios, hazard ratios, prevalence ratios, and correlations for each included article, and specific physical factors are found in Supplementary Appendix 1.
Self-reported: Two studies reported that increased muscle power was associated with better LLFDI scores [33] and reduced risk of mobility disability [117]; one study reported that reduced muscle power predicted a greater likelihood of a decline in mobility function, as measured by LLFDI scores [96].

Muscle endurance
Performance-based: Studies reported that increased muscle endurance (e.g.Truck Muscle Endurance Test [118]) was associated with faster gait speed [59], better balance scores [57,59], higher SPPB scores [57], walking longer distance on a 6MWT [66], completing the CST [59] and the TUG test [77] in less time.Reduced muscle endurance was associated with taking longer to complete the TUG test [116].
Self-reported: Studies reported that increased muscle endurance was associated with better LLFDI scores [33,92] and better lower extremity function [59].One study reported that reduced muscle endurance predicted a decline in mobility function as measured by LLFDI scores [116].

Muscle coordination
Performance-based: Studies reported that better muscle coordination test scores (e.g.Heel-to-Shin Test [59]) were associated with faster gait speed [119] and completing the CST [59] in less time.Poor muscle coordinator test scores were associated with slower gait speed, lower SPPB scores, poor balance scores, and taking longer to complete the CST [120].We found no study reporting the association between muscle coordination and self-reported mobility outcomes.

Range of motion (ROM)
Performance
Self-reported: Studies reported that increased pain was associated with greater mobility difficulty [59,136,139-142], lower LSA scores [94], and incident use of assistive walking devices [104,143].

Hearing
No studies examined the association between hearing and performance-based mobility outcomes.Self-reported: Five studies reported that poor hearing was associated with greater mobility difficulty [98, 140,147,155] and reduced driving [156].

Proprioception
Performance-based: Studies reported that poor proprioception (defined as the inability to identify the lower limb in position) was associated with poor balance scores [157] and taking longer to complete the TUG [74,157].One study reported that the ability to identify lower limb in position was n: number of articles that explored the association between the physical factors and mobility outcomes.
associated with completing the TUG in less time [27].We found no studies reporting the association between proprioception and self-reported mobility outcomes.

Dizziness
Performance-based: Two studies reported that dizziness was associated with slower gait speed [158] and poor balance scores [157].No studies reported the association between dizziness and self-reported mobility outcomes.

Respiratory parameters
Performance  [166], slower gait speed [167,168], poor balance scores [167], lower SPPB scores [166] and taking longer to complete the TUG [116].One study reported that dyspnoea was associated with slower gait speed, and respiratory muscle weakness was associated with lower SPPB scores [168].Another study reported that low breathing reserve, defined as the difference between the maximal voluntary ventilation and the maximum ventilation measured during the exercise test, was associated with completing a 400meter walk in a less time [169].
Self-reported: One study reported that a decrease in FEV1 was associated with mobility difficulty [166], and another study reported dyspnoea severity was associated with lower LSA scores [93].One study reported that older adults using a ventilator were more likely to use a walker compared to those not using ventilators [29].

Cardiovascular parameters/biomarkers
Performance-based: Studies reported that heart rate within the expected range was associated with faster gait speed, better balance scores, and completing the TUG test and the CST in less time [170].One study reported that a high anklebrachial index score was associated with faster gait speed and better balance scores [171].Studies reported that a low ankle-brachial index score (indicative of peripheral arterial disease) was associated with slower gait speed, poor balance scores, taking longer time to complete the CST [172] and walking shorter distance on a 6MWT [173].High blood pressure [174,175] and high levels of calcium in the coronary [176] were associated with slower gait speed.
Self-reported: Studies reported that a high ankle-brachial index score was associated with reduced mobility limitations [171] and a faster self-reported walking pace [177].One study reported that high blood pressure and abnormal heart rate were associated with self-reported walking speed [177].

Exercise and physical activity (type and frequency)
Performance-based: Studies reported that increased physical activity was associated with an increased number of daily steps [178], greater likelihood of being able to walk 10 metres independently [179], faster gait speed [46,180], higher cadence [46,58], longer stride [46] and step length [181], and walking longer distance on a 6MWT [182], better balance scores [46,58,183], completing a walking test [58,178] in less time.Low levels of exercise were associated with slower gait speed [38,177].
Self-reported: Studies reported that being inactive or having lower levels of physical activity were associated with reduced mobility limitation [102,140,184], lower LSA scores [93], and incident use of assistive walking devices [104].One study reported that increased physical activity levels were associated with high LSA scores [185].
Self-reported: Eight studies reported that an increased number of falls were associated with mobility limitation [30,137,184,209], lower LSA scores [144], use of walking aid [203,206] and reduced use of public transit [215].

Frailty
This concept is defined as unintentional weight loss (10 or more pounds within the past year), muscle loss and weakness, a feeling of fatigue, slow walking speed and low levels of physical activity [102].Performance-based: Higher frailty score was associated with lower balance scores [183].
Self-reported: Studies reported that frailty was associated with lower LSA scores [216] and increased incidents of mobility limitations [102].

Non-chronic conditions (e.g. fracture) & non-specific symptoms (e.g. fatigue)
Performance-based: Studies reported that having a fracture was associated with a slower gait speed [53,236], increased double support time [53], poor balance score, taking longer to complete the CST [236].One study reported that having a foot deformity was associated with poor balance scores [127].Two studies reported that fatigue was associated with slower gait speed [237] or lower LSA scores [94].
Self-reported: One study reported that a history of fracture was associated with incident use of walking aid [104], and another reported that fatigue was associated with mobility limitations [102].

Discussion
Our review aimed to identify physical factors and their association with mobility outcomes in older adults.We identified 18 factors grouped into musculoskeletal system-related (e.g.muscle strength), sensory and nervous system-related (e.g.pain), respiratory (e.g.FEV1) and cardiovascular (e.g.cardiovascular biomarkers) system-related, and other factors including type and frequency of exercise, the number and types of chronic and non-chronic conditions, falls and body composition.Compared to other factors, including environmental, cognitive, psychological, social, personal, and financial, physical factors are consistently associated with older adults' mobility [32,94,184,185] and are easily identified by clinicians, older adults, and family members, highlighting the critical role that physical factors play in explaining the complexity associated with older adults' mobility.
Although the InCHIANTI study found that low power was associated with a 2-3-fold increase in mobility limitations than low strength [255], more studies in our review focussed on muscle strength rather than muscle power.Arguments explaining why muscle strength has been consistently studied more muscle power include but are not limited to the lack of validation across popular muscle power tests [256] as well as frequent discrepancies across testing protocols between studies [257].Even though studies validating these different muscle power tests are needed to encourage researchers further in exploring the role of muscle power on older adults' mobility, investigating the additive role of muscle strength and muscle power on mobility outcomes is pertinent.Could exploring the additive role of muscle power and strength provide more insight into understanding the complexity of mobility and possible focussed assessment and intervention?Future research should explore this since each independently predict mobility outcomes among older adults.
Poor physical health indicative of chronic conditions is a strong predictor of community mobility [184].Approximately 85% of older adults have at least one chronic health condition, and 60% have at least two chronic conditions [258].Our review revealed that chronic conditions are often not examined independently; they are often examined in association with other factors, indicating that multiple risk factors may significantly impact mobility.However, studies exploring the additive effect of multiple factors, including chronic conditions, was scarce, indicating the need for future research to explore these possible relationships.Future studies should highlight the additive-predictive power of chronic conditions relating to mobility outcomes.
Our review also highlighted the complexity in association between physical factors and mobility.For instance, obese or overweight older adults are at higher risk of developing chronic conditions such as hypertension and diabetes.Subsequently, diabetes could lead to vision and sensory impairment, causing pain which further limits older adult mobility.Reciprocally, limited mobility may lead to muscle weakness, further limiting participation in physical and social activities, which could accelerate age-related changes in cognitive and affective domains -thus facilitating the continuous cycle mobility decline in older adults.Since older adults often report multiple physical factors concurrently, combinations of multiple risk factors may have a more significant impact than the sum of their individual effects, and future studies should explore these compounding effects [13].Future interventions targeting mobility maintenance should be multidimensional, focussing on targeting modifiable risk factors concurrently.
Our review noted some gaps or areas to address regarding physical factors and mobility.First, the associations between physical factors and mobility focussed mainly on community-dwelling older adults, with limited studies conducted among hospitalised older adults or those residing in nursing homes.Older adults experience a decline in several physical factors, including muscle mass, strength, function, and pulmonary functions during hospitalisation, leading to difficulty in performing activities of daily living (ADL) and limitations in community mobility [259,260].Reduced community mobility has been associated with other adverse health outcomes for older people after discharge, including loneliness, depression, and mortality [261], highlighting the need for studies to explore the association between physical factors and mobility, both on admission and discharge.
Studies on associations between physical factors and mobility among hospitalised older adults are limited.This finding is troubling, as mobility is an independent predictor of hospital readmission [26].Other possible reasons could be that mobility loss is not recognised as an essential outcome in hospital care, and the average hospital stay is too short to assess mobility [90].Hospital cultures that do not encourage the active involvement of physiotherapists during admission could highlight the limited number of studies among hospitalised older adults.Therefore, stand-alone programs focussing on routine mobility assessment among hospitalised older adults are recommended, as they promise to reduce hospital costs caused by readmission.
Understandably, assessing nursing home residents' mobility can be challenging as most residents might not be able to complete performance-based or self-reported mobility measures, especially nursing home residents with cognitive impairment or dementia.However, recent technology advancements, including wearable sensors (such as smartwatches and Fitbits), although currently used to explore activity levels of nursing home residents and track falls [262], measure several mobility outcomes, including gait parameters (e.g.gait speed) and balance [263].Therefore, future studies should use routinely collected mobility data in nursing homes to explore the association between several mobility outcomes and physical factors; this is promising to inform care plans and interventions that could improve mobility for nursing home residents.
Second, despite the impact of the cardiovascular system on mobility in older adults, only eight included studies explored the association between the cardiovascular system and mobility among older adults.While five studies crosssectionally examined the association between high blood pressure [264] and cardiovascular biomarkers, such as anklebrachial index [171,172,175,176] and mobility outcomes, three longitudinal studies [173,174,177] highlighted the impact of cardiovascular functions on older adults' mobility across the life course.These longitudinal studies provided cardiovascular cut-offs for clinicians to ascertain which older adults are at higher risk of mobility decline based on their cardiovascular parameters allowing early intervention to reduce the age-related cardiovascular decline [265].However, these studies are from a USA sample population (primarily Caucasian), limiting the application of these findings in populations different from the USA population.Cardiovascular parameters, including blood pressure, have been correlated to genetic variations found within continental regions.For instance, compared to Europeans and white Americans, people of Sub-Saharan Africa and African descent in America and Europe have higher systolic and diastolic blood pressure, whereas South Asians have lower systolic blood pressure but similar diastolic blood pressure with Europeans and White Americans [266].Therefore, it is plausible that the influence of cardiovascular parameters on mobility could differ across regions globally.Hence, studies exploring the longitudinal association between cardiovascular parameters and mobility outcomes among older adults in other regions, including Africa and Asia, are needed.
The findings from this review and our previous reviews on the associations between self-reported and performancebased mobility outcomes, and environmental, personal, and financial factors [16]; and cognitive, psychological, and social factors [17] provided information for advancing the use of Conical Model.Our study provided a comprehensive list of factors that can guide further development of core factors within each determinant that influences mobility in a different context.For instance, this may take the form of exploring which factors within each determinant are critical to assess when older adults are being discharged from hospital-tohome.With the associations between each factor within each determinant synthesised, it can create a foundation for transdisciplinary collaborations to explore further the complexity of mobility and more effective ways to actively incorporate and assess the interrelationship effect of each determinant or combined effect on mobility across different settings.

Study implications
We did not provide a synthesised mean or SD or other clinical parameters to enable clinicians to use the findings in their clinic mainly because of the heterogeneous nature of the included study aligning with the scoping review aims.However, our study provided a comprehensive list of physical factors influencing mobility and clinicians interested in any of the factors could review the appendices for more details.For instance, the factors identified and their association with mobility described in this paper could guide clinicians in deciding which factor to intervene in amid competing demand for hospital resources.
The findings from this review and our previous reviews on the associations between self-reported and performancebased mobility outcomes, and environmental, personal, and financial factors [16]; and cognitive, psychological, and social factors [17] provided information for advancing the use of Conical Model.The review provided a comprehensive list of physical factors influencing mobility for researchers, clinicians, policymakers, organisations, older adults, and family members to develop core outcome sets for a specific context, population, or mobility forms, including transportation, driving, and use of assistive devices.A core outcome set is a recommended minimum set of outcomes or outcome measures for a particular health construct, condition, or population, which should be reported for all trials on that issue [267].This comprehensive list of mobility factors represents an essential first step towards greater standardisation in assessing and measuring mobility.Currently, there is no core outcome set for mobility factors in the older adult population.Given the importance of mobility and its associated role in predicting several health outcomes in older adults, there is a need for greater consistency in measuring factors influencing older adults' mobility across studies, which a core outcome set would establish.Indeed, a core outcome set of mobility factors that can be widely used across a range of contexts and settings can facilitate easy comparison and interpretation of result findings, ultimately informing clinical and research practice.
With the associations between each factor within each determinant synthesised in this review and previous reviews [16,17] it can create a foundation for transdisciplinary collaborations to explore further the complexity of mobility and more effective ways to actively incorporate and assess the interrelationship effect of each determinant or combined effect on mobility across different settings.Currently, mobility assessment and interventions are heavily conducted by physiotherapists and often occupational therapists.Most mobility assessments, for example, Activity Measure for Post-Acute Care 6-Clicks [268] used by physiotherapists or other rehabilitation professionals, are heavily focussed on body positions, transfers, personal care, home skills, and applied cognition, such as speaking and understanding, with no components on psychological, and financial factors influencing mobility.Creating a comprehensive set of mobility assessments comprising physical, environmental, personal, economic, cognition, psychological and social factors could create an opportunity for other professionals, such as nurses or psychologists, who typically are not involved in mobility assessment.This proposed comprehensive mobility tool can guide as a screening to determine which healthcare professionals should be actively engaged as the transdisciplinary mobility assessment team enhancing interdisciplinary approach to older adults' mobility care.

Study strengths and limitations
The strength of this review lies in its comprehensiveness and inclusion of studies from 32 countries, demonstrating the applicability of the review findings.The a priori publication of the review protocol reduced publication bias and improved the study findings' reproducibility.The use of health science librarians strengthened the search strategy enabling the retrieval of possible articles.
Despite our effort to develop a comprehensive search strategy, some articles may have been missed, especially of the article keywords were not in the MESH terms in our search strategy.Also, we may have missed some articles published in other languages than English.In addition, we included only articles published after 2000, which may have missed some factors that could influence older adults' mobility published before 2000.Although we did not limit our search by country, we found no studies conducted in Africa.Our study defined older adults as individuals 60 years and older, which could explain why studies in Africa did not meet our inclusion criteria.Most gerontological studies in Africa define older adults as individuals 50 years and above [269].We argue that the association between most physical factors and mobility outcomes may not differ across regions, except for some physical factors, such as blood pressure cutoffs and chronic conditions, which have regional variations.About two-thirds of the studies included in this review were cross-sectional and were unable to determine cause-effect relationships between physical factors and mobility outcomes.There is a need for more longitudinal studies to allow a systematic analysis of either independent or additive physical factors in order to produce predictive models of mobility in older adults.
Grouping some physical factors into one group, for instance, types of chronic conditions, did not account for the heterogeneous nature of chronic diseases in the older adult population.The pathways in which diabetes influences mobility might differ from how cancer influences mobility.Therefore, readers should bear this in mind when interpreting the study findings.This grouping further illuminates the complexity associated with mobility and highlights that it may not be possible to simply or saturate factors influencing mobility, as some aspects would be left out or misrepresented.

Conclusion
This review found that physical factors such as muscle strength and power are consistently associated with performance-based and self-reported mobility outcomes in older adults.Our study provided physical factors influencing mobility that could guide the development of a core set for physical factors influencing mobility in a different context, allowing for efficient and compelling comparison of physical factors influencing mobility.Longitudinal studies exploring the additive association of physical factors with mobility outcomes are recommended to highlight the complexity and enhanced intervention and prevent mobility decline.

Table 1 .
Search strategy.Elderly OR geriatric OR ageing OR older people � OR senior � OR retirees OR Aged OR older persons OR gerontology OR ageing OR adult � Muscle strength OR previous injury OR proprioception OR vertigo OR vestibular diseases OR dizziness OR body composition OR visual acuity OR vision disorders OR physical endurance OR muscle power OR comorbidity OR chronic conditions OR coordination OR sensation disorders OR sensation OR range of motion mobility limitations, OR life space measures OR mobility OR walking OR ability level OR physical mobility OR movement OR gait OR time up and Go OR physical functioning OR, six minutes' walk test OR berg Scale OR short physical performance battery OR Transportation OR travel OR driving OR driving safety OR crashes OR accident OR road test OR Walking aid OR ambulation aids OR assistive devices OR wheelchair � OR scoote � OR cane � OR crutche � OR prosthetic devices OR orthotic devices OR walker Population terms AND Exposure terms AND Outcome terms was adapted and search in seven databases including PubMed, EMBASE, AMED, CINAHL, Psych INFO, Web of Science, AgeLine.MeSH and keywords -heading were adapted for each database using Boolean/phrases e.g.� .
Table 2 presents the main characteristics of the included studies.Studies were conducted in 32 countries on five continents, including Asia, Australia, Europe, North and South America.Close to half of the articles (n ¼ 108, 45.2%) were conducted in North America, primarily in the United States of America (n ¼ 97, 40.6%) and Canada (n ¼ 11, 4.6%).All were quantitative studies; close to three-quarters (n ¼ 175) were cross-sectional studies, and 207 (86.7%) recruited participants from the community.The mean age of the participants in the included articles ranged from 60 [27] to 93.1 [28] years.Sample sizes varied considerably, ranging from 13 [29] to 164,597 [30].Of the 239 articles, 175 (73.2%) articles assessed mobility using performance-based measures only, 42 (17.6%)articles assessed mobility using self-reported measures, and 22 (9.2%) articles assessed mobility using both (see Table 2 for details).
Self-reported mobility outcomes: Self-reported walking included the distance walked, number of times walked per day/ week/month, amount of time walked (second or minutes), self-reported walking speed, self-reported walking capabilities, number of days outdoors; Other self-reported mobility questionnaire included Rivermead Mobility Index, Independent Mobility Questionnaire (IMQ), RAND-36 Physical Functioning Questionnaire, Walk 12-G Questionnaire, Functional Mobility Scale, EuroQol -five-Dimension Scale -mobility domain, and Late Life Function and Disability Instrument; Driving related outcomes included driving performance (ability or inability), access to car, driving duration & frequency, driving distance, preference to be driver versus passenger; Mobility limitation defined as self-reported inability on all or any of the following: walking up and down a flight of stairs (10 steps) or several flights of stairs, walking a mile (1600 meter) or half a mile (800 meter) or a quarter mile or a block (400meter) or 100-300meter, or across the room and running/jogging for 20-30 min; Mobility assistive devices included scooter, powered and manual wheelchairs, walking aids (cane, walker, crutches).
Studies reported that joint ROM within the normal range was associated with better LLFDI scores [33,92].Abnormal joint ROM was associated with increased mobility difficulties [96,124,126].

Table 3 .
Physical factors from the included studies and their description (n ¼ 433).
Medical condition of reduced function and health in older individuals.People who are frail usually have 3 out of the following five symptoms: muscle loss, weakness a feeling of fatigue, slow walking speed, and low levels of physical activity.