Assessment of Function Limitations in People with Chronic Kidney Disease for Implementation in Clinical Practice

Chronic Kidney Disease (CKD) is a global health problem and a significant contributor to mortality, morbidity and disability from non-communicable diseases (NCD). The current consensus amongst researchers in the field of renal rehabilitation and healthcare practitioners involved in the management of people with CKD, is that physical dysfunction and inactivity are severe and prevalent in all ages and CKD stages compared to normative data. The negative impact of CKD on independence is cumulative, and health interventions and policies should target prevention of deterioration of kidney function and its consequences for physical frailty, disability and ultimately, quality of life. Accurate and feasible assessment of physical function is key for supporting the clinical implementation of current clinical care management guidelines. This overview, therefore, presents the rationale and some key scientific evidence to support the recommendation of physical function measurement tools that reflect function limitations for immediate implementation in clinical practice. Measurement property characteristics of the proposed measurement tools are also summarised in order to support health and exercise professionals in selecting the right tool and in interpreting and evaluating the measured outcomes.


Introduction
Chronic Kidney Disease (CKD) is a global health problem and a significant contributor to mortality, morbidity and disability from non-communicable diseases (NCD). Although age-standardised mortality and disability-adjusted life years have seen small increases in the last 20 years at a global level, regional disparities do exist. Higher mortality and CKD burden are reported in world regions of lower sociodemographic index, and these disparities have been mainly attributed to limited access to renal replacement therapies (RRT) and increases in the prevalence of diabetes and hypertension [1]. CKD is also considered a significant risk factor for cardiovascular disease (CVD), independent from other conventional risk factors (e.g., hypertension, diabetes, smoking, physical inactivity). The current global estimated prevalence of CKD sits at 9.1%, with the prevalence of people in CKD stage 5 and on some form of RRT standing at 0.1%, indicating that the largest proportion of people with CKD are in stages 2-4 [1]. In the UK, about 7% of the adult population above 35 years presents with CKD 3-5 [2]. Global all-age CKD prevalence has increased by 29.3% in the period of 1990-2017, which is mainly attributed to larger increases in incidence of RRT resulting from population ageing [1]. In the UK, about 67,000 people are currently on RRT, with the number of people requiring RRT increasing by about 2% every year since 2018 [3]. Furthermore, although the most prevalent cases of CKD are in stages 2-4, life years lost to disability are estimated to be the largest for CKD stage 5 and dialysis (40%) vs. 22% for CKD 2-4 [1]. This statistic indicates that the negative impact of CKD on independence may be cumulative. It makes sense, therefore, that greater emphasis of health interventions and policies is placed on targeting prevention of deterioration of kidney function and associated consequences in order to slow or reverse loss of physical independence and, ultimately, quality of life.
A case already exists for the regular assessment and promotion of physically active lifestyles to enhance clinical management of the CKD patient and alleviate disability symptoms [28,29]. Accurate and feasible assessment of physical function is key to supporting the clinical implementation of care management guidance, not only to characterise prognosis and adverse clinical risk, but also to monitor progress and response to interventions towards optimised levels of physical function for a given individual. Routine monitoring of physical function and sharing this information with the people concerned can also be used to motivate patients to actively engage with physical activity (PA) to maintain/enhance physical function [18].

Function Limitations Measurement Instruments for Clinical Implementation
The execution of comprehensive incremental exercise testing, which includes measures of gas exchange, cardiac function, systemic blood pressure monitoring and patients' subjective responses to general and specific discomfort (via ratings of perceived exertion, angina and breathlessness scales) is considered to be the "gold standard" of integrated assessment of exercise capacity (physiological level of impairment). However, this approach is not feasible in all patients, environments and settings [4]. Physical performance tests that imitate and reflect the ability to undertake physical tasks of daily living offer a clinically feasible alternative and/or complementary method of testing incremental and upper limits of physiological exercise capacity.
All recommended measurement tools for clinical implementation should ideally satisfy minimum standards of good measurement properties (validity, reliability, responsiveness), clinical utility (interpretability), patient safety and feasibility [30]. Table 1 provides a summary of available clinimetric properties for the perusal of interested readers when trying to decide what tool to use and how to interpret results at the level of the individual. It is beyond the scope of this paper to expand on clinimetric evaluation of outcome measures, but interested readers are directed to the resources and guidance by the COSMIN Taxonomy of Measurement Properties [31]. HGS and early admission to hospital post KTx (not significant) [51] HGS predicts mortality [54,55] better than measures of muscle mass and other measures of sarcopenia [49,56,57] and hospitalisations [9,57] HGS and fractures (not significant [58] Who In all adult people with CKD in any stage In all adult people with CKD in any stage In all adult people with CKD in any stage

Impact/Value
Cut off points for sarcopenia classification [59] Cut-off points for frailty phenotype [23] Cut off points for sarcopenia [59] Normative data from UK population exists [60] Expertise No specialist training required but standardised assessment protocol needs to be followed [5] No specialist training required but standardised assessment protocol needs to be followed [ MCID-minimum clinically important difference; MNA-mini nutritional assessment; NDKEF-non-dominant knee-extension force; RM-repetition maximum; STS-sit-to-stand; SEM-standard error of measurement defined as test-retest variability in measurement score with 68% confidence intervals; TUG-3 m timed up-and-go test; VO 2 peak-peak oxygen uptake. Validity indices reflect the extent to which the measurement tool reflects the construct it is supposed to measure. If validity is poor, then the tool is not a good substitute for the construct of interest. Reliability indices reflect the amount of variation (error) associated with a measurement tool, and therefore change scores need to exceed these indices in order to count as true changes. Interpretability indices reflect the clinical or practical importance of scores or change of scores, and this information can be used to support appropriate risk stratification and monitoring of progress at patient level.

Function Limitations-Which Measurement Instruments?
Batteries of physical performance tests such as the Short Physical Performance Battery (SPPB) have been reported in the CKD literature as a prognostic tool in survival studies in kidney transplant recipients [62], but also as an outcome in exercise-based interventions [63]. The SPPB is composed of objectively tested performance components of standing balance, five chair rises and gait speed, that are aggregated to produce a scoring range from 0 (worst) to 12 (best). The SPPB has also been suggested as one of the tools to assess physical performance to confirm the presence of sarcopenia [59]. Physical frailty, predominantly assessed using Fried's phenotype [23], has also emerged in the last decade as an indicator of function limitations, disability and other adverse clinical outcomes [6,13,19]. Although Fried's phenotype is a relatively expedient assessment, it requires a combination of both objectively measured physical function (handgrip strength and gait speed) and patientreported outcomes (fatigue and physical activity). There is also a considerable amount of variation and accuracy in the estimation of prevalence and responsiveness of this index to various interventions in the CKD population [17], mainly due to the different approaches used to measure the constituent components of physical frailty [48]. Therefore, many clinicians still find these procedures for assessing physical-function-related outcomes more time consuming and impractical for universal clinical implementation. Moreover, given that changes in the total scores and classifications produced by these indices are driven mostly by changes in single-constituent components (e.g., gait speed and/or STS performance), ref. [63] and that single tests are usually simpler, more time efficient and carry superior prognostic value [62], it makes sense to propose the following stand-alone tests and outcomes for routine use in clinical practice, but also for research purposes.

Walking Mobility
Several investigators have evaluated walking ability via measures of distance walked during a fixed time period (e.g., 6 min walk test) and time and gait speed over a set distance (e.g., Timed-Up and Go test over 3 m or gait speed over 4 m). Maintaining walking ability is important for active living and life participation, considering that walking is the most accessible form of PA and mobility. In a systematic review of falls in CKD, Lopez-Soto (2015) [64], identified that the majority of reported falls occurred during ambulation and rising from a seated and/or lying position. Thus, assessing and improving walking and body transfer abilities may be important for immediate patient benefit in the short term. Table 1 summarises measurement properties of the most commonly reported walking tests in the CKD literature.
Gait speed stands out as the most consistent and discriminatory outcome for a number of health-related outcomes [6]. Nixon et al., (2020) [20], identified gait speed as the best single screening outcome for physical frailty in a cohort of 90 participants in CKD4-5. Gait speed had an area under the curve (AUC) value of 0.97 [95%CI 0.93-1.00] with sensitivity and specificity of 0.84 and 0.96, respectively, against the Fried's frailty phenotype. Similarly, gait speed was associated with excellent diagnostic accuracy for physical frailty in young (<65 years) people on HD with an AUC of 0.98 (95%CI 0.96-1.00) and in older (>65 years) patients (AUC: 0.79; 95%CI: 0.64-0.94). Notably, there was a significant difference in the AUC between the age groups, indicating that gait speed performs better as a frailty screening tool in the younger HD subgroup [21]. A cut-off point of <0.85 m/s for physical frailty screening was derived based on the best balance between sensitivity (79%) and specificity (93%) achieved against the Fried's frailty phenotype [21]. Moreover, a study including 800 people on HD suggested that gait speed may be more sensitive in discriminating function limitations in women than handgrip strength, or other measures of sarcopenia, as the prevalence of slowness amongst women (48.3%) was twice as common than in men (28.7%) [49]. From the same study, gait speed also significantly improved the accuracy of predicting death over 2 years, with 1SD of faster gait speed being associated with a significant (36%) reduction in death (HR: 0.64; 95%CI: 0.48-0.84). Gait speed <0.6 m/s as opposed to >0.6 m/s was associated with a HR of death of 2.17 (95%CI: 1.19-3.98) over a period of 12 months, in a cohort of about 700 people on HD of 65 years of age [50]. From the same study, people who could not complete a walk of 4.57 m were almost seven times more likely to die over a year (HR: 6.93; 95%CI: 4.01-11.96). The risk of hospitalisation over a year was similar, but still significant between subgroups with gait speed <0.8 m/s and <0.6 m/s (OR; 2.04; 95%CI: 1.19 to 3.49), but difficulties with activities of daily living (ADL) were more prevalent in the subgroup with gait speed <0.6 m/s. Each 0.1 m/s decrement in gait speed was associated with a 17% increased risk of death in the same group of HD patients over a period of 2 years [50] and with 26% increased risk of death in CKD 2-4 over a 3-year period [15].
Distance walked over 6 min (6MWT) of >350 m has also been associated with almost a three times higher risk of death over 3 years in pre-dialysis patients, in models controlling for multiple confounders [15]. Torino et al., (2014) reported that a 20 m increase in the 6MWT was associated with an 11% reduced risk of all-cause death, 7% of all fatal and non-fatal cardiovascular events and 4% of all-cause hospitalisations in 182 patients on HD [65]. Similar observations regarding the clinical interpretability of the timed up-andgo performance (TUG) have been reported in the literature, with 1 s slower TUG being associated with 8% increased risk of death [15], and scores > 10 s (or unable to complete) being associated with a seven-fold increased likelihood of ADL disability in elderly CKD5 patients [66]. More recently, Zanotto et al. (2021) reported that the 3 m TUG score was an excellent screening tool for physical frailty (AUC: 0.90; 95%CI: 0.80-0.99). A score >10.8 s was associated with sensitivity and specificity of 0.89 and 0.85, respectively, against Fried's Frailty phenotype in a cohort of 76 people on HD, and this test performed equally well as a screening tool in younger and older subgroups [21].
Gait speed and dynamic mobility are highly relevant for ADLs and strong predictors of patient relevant outcomes and might also outperform other measures of sarcopenia and muscle strength for discriminating function limitations between men and women on HD [49]. However, they may require some preparatory work (set up and marking of distances) and access to sufficient and safe walking indoor surfaces and spaces. In addition, assessment of walking mobility would be feasible only in ambulatory people (Table 1).

Body Posture Transfers
Sit-to-Stand (STS) performance tests involve rising unassisted from a standard height chair (42-46 cm) and sitting back on the chair, as fast as possible. Sit-to-Stand body transfers are commonly used as outcomes in exercise-based interventions in the CKD population, and several variations of the test exist, such as STS-5 and STS-10, which reflect the fastest time (in seconds) people can complete 5 or 10 STS cycles. These tests have been interpreted as proxy indicators of muscle power [5,67]. Sit-to-Stand-5-derived muscle power has been shown to significantly discriminate varied abilities in performing ADLs and walking, and even outperformed a muscle-mass index in a group of 100 individuals in pre-dialysis CKD [68]. STS-30 and STS-60, on the other hand, have been used as indicators of muscular endurance and fatiguability, as they require patients to perform as many STSs as they can in 30 and 60 s respectively [5]. The clinical value of the STS performance is supported by observations that for every 1 mL/min/1.73 m 2 drop in GFR, the odds were 1.5 times higher that the patient would not be able to rise unassisted from a chair once (STS1), and when the diagnosis was diabetes, the odds ranged from 1.7 to 21 times higher [69]. In a more recent study in 162 CKD 2-5 patients, lower STS30 was significantly associated with a 35% reduced risk of major adverse cardiovascular events (HR: 0.65; 95%CI: 0.47-0.89) and 16% reduced risk of all-cause hospitalisation (HR: 0.84; 95%CI: 0.74-0.95) over a period of 29 months [53]. Zanotto et al. (2021), also reported excellent diagnostic accuracy of STS-5 against Fried's frailty phenotype in 76 people on HD (AUC: 0.86; 95%CI: 0.75-0.96). A cut off point of >15.6 s was associated with sensitivity of 87% and specificity of 77% and performed equally well in younger and older patients [21]. The responsiveness characteristics of the STS-5 test were reported in a group of 26 individuals with CKD2-4, following a 12-week exercise intervention [47]. Change in QoL, as assessed using the SF36 questionnaire, was used as the criterion against which STS5 change scores were validated. Out of five physical function outcomes tested (STS-60, shuttle walk test, VO 2 peak, one repetition max), STS-5 change was the only outcome that significantly discriminated between those who reported better QoL vs. same or worse QoL following the exercise programme (refer to Table 1 for measurement properties characteristics).

Handgrip Strength (HGS)
In the last decade, we have seen a significant volume of evidence on the predictive value of HGS as an indicator of muscle function and mortality and morbidity outcomes in the CKD population [24,70]. Indicatively, Isoyama et al., (2014) reported that HGS only, remained a significant predictor of death over a period of 29 months (HR: 1.79; 95%CI: 1.09 to 2.94) vs. muscle mass (HR: 1.17; 95%CI: 0.73 to 1.87), in a cohort of 330 people on HD [56]. Furthermore, HGS seemed to identify a higher proportion of people at risk of death, particularly during the early months of the follow-up period, as opposed to the number of people identified as at risk based on measures of muscle mass alone or by sarcopenia classification [49,56]. Similarly, in a cohort of 128 KTx recipients (49 ± 15 year) who were followed for up to 64 months, only decreased HGS remained significantly associated with higher risk of combined mortality and hospitalisations and it even outperformed indices of muscle mass, sarcopenia and sarcopenic obesity in fully adjusted models [57]. However, a recent report by Zanotto et al. (2021), which included data from 76 people on HD, suggested that HGS might not be the best screening tool for frailty in comparison with other tools such as STS-5 and TUG, as it had the lowest AUC: 0.71 (95%CI: 0.59-0.83) and specificity (46%) characteristics [21]. As mentioned earlier, HGS alone may underestimate the prevalence of function limitations in women on HD [49], the predictive power of HGS for survival in women [71] and the extent of lower-limb weakness in all genders [15]. Nevertheless, assessment of HGS is highly common practice for evaluating frailty and sarcopenia in a wide range of population groups with established cut-off criteria [23,59], and existing UK normative data for comparison purposes [60]. Reliability characteristics of the HGS assessment are scarce in the CKD population, but given the highly desirable feasibility features of this assessment tool (Table 1), it could be a tool appropriate for universal implementation in clinical practice.

Summary and Conclusions
Adequate physical function is essential for life-participation activities and perceived satisfaction with overall quality of life. Maintenance of functional independence and prevention of disability is a priority target in all clinical practice guidelines for the management of people in all stages of CKD, and accurate and feasible assessment of physical function, is key to supporting the effective implementation of these guidelines. The most commonly reported tests in the CKD literature, which objectively assess function limitations at the level of the individual, are walking tests to determine gait speed and dynamic walking mobility, sit-to-strand transfers and isometric hand grip strength. All three of these measurement tools can be easily implemented in routine clinical practice and in research settings alike for monitoring and evaluation of physical health. Based on the currently available knowledge, all three of these measurement tools are independent and strong determinants of a range of adverse health outcomes such as falls, frailty, hospitalisations and mortality. Interpretation and evaluation of test results is supported by research evidenced clinical anchors and other clinimetric characteristics, as summarised in Table 1.