Targeting Fall Risk in CKD

High rates of falls and serious fall injuries are documented in multiple studies of patients with CKD, dialysis-dependent and non-dialysis-dependent (1–3). A serious fall injury can precipitate functional decline, loss of independence, nursing home placement, and death. Is there a window of opportunity in CKD care for effectively screening patients and intervening to influence the likelihood of incurring falls and related injuries?

Low muscle mass, muscle weakness, and mobility limitation are common findings in patients with CKD. Measures of physical performance reflect activities of multiple organ systems (neurologic, musculoskeletal, and cardiopulmonary) and are therefore highly relevant in the CKD clinical setting (4). Gait speed is a prominent performance measure in studies that seek to quantify the impact of aging and comorbid conditions on function, and slower gait speed has been shown to be strongly associated with falls. In addition to slowing of gait speed, however, gait disorders encompass issues such as loss of smoothness, symmetry, or synchrony of body movement (5). The stance phase typically makes up about 60%, and the swing phase about 40%, of the gait cycle, and variability in gait patterns has been shown to be associated with incident fall risk in the general population (6). The research reported by Tran et al. (7) in this issue of the CJASN extends this work to include incident fall risk in CKD.

Based on their investigation of gait speed and markers that further define spatiotemporal parameters of the gait cycle, Tran et al. concluded that CKD in older adults is associated with specific quantitative gait abnormalities that manifest in clinical signs, which they term a “gait phenotype,” associated with fall risk. Their research was conducted in a cohort of 330 community-dwelling adults aged ≥65 years, among whom CKD (eGFR<60 ml/min per 1.73 m²) was identified for 134 study participants. Gait cycle parameters were first assessed via the computerized GAITrite system. Consistent with prior research, this quantitative gait measurement indicated that, after adjustment for age and sex, CKD participants had slower gait speed. CKD participants also had shorter stride length, reduced time in the swing phase and increased time in the double support stance phase of the gait cycle, and progressively lower eGFR was linearly associated with increasing abnormality in these parameters. Study participants’ gait was also evaluated by neurologists who classified observed gait patterns as neurologic or non-neurologic and recorded gait signs such as short steps and sway while walking. Both (a) short steps and (b) marked sway or loss of balance with straight or tandem walking were more common among CKD adults. Compared with individuals without CKD or these clinical signs (“gait phenotype”), those with both CKD and the gait phenotype had the most pronounced abnormalities in gait speed and in stride length, swing time, and double support time.

Finally, the investigators estimated participants’ fall risk over a median follow-up of almost 2 years as a function of presence/absence of CKD and presence/absence of gait phenotype signs. CKD participants with any gait phenotype sign had increased risk of falling compared with non-CKD participants without these signs, and the phenotype was associated with risk of injurious falls among CKD participants. The authors also provided a statistical analysis showing that shorter stride length, reduced time in the swing phase, and increased time in the double support phase of the gait cycle mediated the observed association of eGFR with gait speed.

As study limitations, Tran et al. note that they defined CKD based on only one measure of serum creatinine and that no information on albuminuria was collected; comorbidities were self-reported rather than obtained from medical records; and their study cohort was relatively healthy, as indicated by low number of comorbidities and medications, potentially reducing the generalizability of their findings. It is also relevant to ask whether the gait abnormalities identified in this and in similar studies are specifically and causally related to the presence of CKD, rather than representing consequences of commonly co-occurring CKD risk factors (vascular disease, diabetes).

Additional questions that may be raised concern the high correlation that often exists among single gait parameters (6) and the role of muscle strength and cognitive status in gait abnormalities. For example, increased time in double stance reduces the time the swing leg has to advance and shortens step length (5), reflecting the interrelation of these gait domains. With regard to muscle strength and cognitive status, when Tran et al. included these variables in multivariable linear regression models, the association of eGFR with quantitative gait abnormalities was no longer statistically significant. In addition, shortened steps were associated with reduced fall risk among non-CKD participants, consistent with the general population finding that shortened steps may be protective against falling (5). Ability to attend to step patterns with the goal of purposely avoiding falls, reflecting cognitive function, may have contributed to this finding for non-CKD participants, while increased fall risk was observed for participants who were classified as having CKD and had shortened steps.

The work by Tran et al. contributes to an important research agenda, i.e., gait abnormalities, mobility impairment and fall risk in persons with CKD. Consistent with this agenda, a group in The Netherlands recently reported their investigation of CKD, gait patterns, and falls incurred in the past year in a large general population cohort (8). Investigators assessed kidney function both by eGFR and by albumin-to-creatinine ratio (ACR) in 1430 participants, along with gait velocity and seven gait domains termed Rhythm, Phases, Variability, Pace, Tandem, Turning and Base of Support (8,9). Lower eGFR and higher ACR were associated with slower gait speed and with worse scores in the Variability gait domain. Higher ACR was also associated with slower scores in the Pace domain and worse scores in the Phase domain (8). Interestingly, in the Reasons for Geographic and Racial Differences in Stroke Study, Bowling et al. (3) observed an increased risk of serious fall injuries with elevated ACR but not with lower eGFR. The Netherlands group found an association between impaired kidney function (assessed by ACR or eGFR) and history of fall only in participants with worse gait scores (8). The Netherlands study (8) and the work by Tran et al. (7) differ in the size and nature of their study cohorts and in retrospective versus prospective examination of falls, but both found that not only slower gait speed but also other specific gait abnormalities were associated with lower kidney function.

Gait is a highly complex concept that can be studied using many different variables, which in turn vary in level of complexity (9). Although gait speed provides a simple and quick option for measuring fall risk (6), many gait domains are difficult to assess visually (8). A thorough computerized gait assessment may be needed to identify people most in need of fall prevention programs (8), which may not be available or practical for most patients. However, Tran et al. described “easily identifiable individual clinical signs” of spatiotemporal gait abnormalities that they found to be consistent with the quantitatively measured gait abnormalities they identified. These signs—specifically, short steps, and marked swaying or loss of balance—were more common in the CKD patients they studied, and the signs were associated with risk for incident falls. The researchers point out that patients with these signs did not necessarily have slow gait speed and would not be flagged for fall risk using gait speed alone as a screening tool. Tran et al. (7) advocate further testing and validation of these clinical signs in relation to fall risk in patients with CKD. More work is also needed to determine if providers who care for CKD patients (primary care physicians, nurses, nephrologists, and others) could be trained to accurately identify these clinical signs.

Not only fall risk but also mobility limitation in general are concerns in the CKD population. Exercise trials have shown that gait speed and walking capacity are modifiable across the spectrum of kidney disease (4). Gait abnormalities identified by Tran et al. are potentially modifiable, e.g., gait variability and double-support phase have been reported to improve with treadmill training (6). In many ways CKD is a model for accelerated aging (4), and many resources from the geriatric literature are relevant to guide evaluation in the clinical setting and recommendations for treatment. Strength training, balance training and assistive devices can lead to improvement. Although identifying abnormal gait is important, Judge (5) cautions that interventions to alter gait are not always indicated. For example, a slowed, aesthetically abnormal gait may enable a person to walk safely and without assistance. Hearing from the patient, at least annually, is the first step: Is the individual having any difficulty with walking, balance, or both? What are the details of the onset of the difficulty? Has the person fallen, or do they fear they might fall? Maintaining independence has been rated the top health outcome priority of older adults with advanced CKD (10). Optimal mobility, e.g., being able to cross a street before the light changes, as well as avoiding falls, are crucial for this objective.

Disclosures

Dr. Kutner and Dr. Bowling have nothing to disclose.