The likelihood of self-reporting balance problems in those with advanced chronic kidney disease, slow gait speed, or low vitamin D

Abstract

Objective:

The purpose of this study was to determine the prevalence of vitamin D (25(OH)D) and balance deficits in persons with chronic kidney disease (CKD) and the likelihood of self-reported balance and falling problems, and measured gait speed in persons with kidney disease, and low levels of vitamin D and albumin.

Design:

Analysis of the NHANES 1999-2004 dataset.

Subjects:

The study included 8,554 subjects >40 years old divided into CKD stages based on Glomular Filtration Rate (normal kidney function and stages 1 and 2 served as the control group, stages 3 and 4/5 as the CKD groups).

Main outcome measures:

Measured 25(OH)D levels, timed 20-foot walk, Romberg standing balance task, and self-reported balance and falling issues

Results:

The prevalence of balance deficits was found to be high in this CKD sample, with fail rates increasing with kidney disease severity. Similarly, when examining the relationship between CKD stage and the measurement of balance, fail rates (impaired balance) increased and gait speed decreased with kidney disease severity. In addition, the likelihood of self-reporting a balance and falling problem in the past year was higher in persons who had advanced CKD, older age, female sex, former or current smoking status, lower 25(OH)D levels, and lower albumin levels. Similarly, the likelihood of having a 20-foot walk time of greater than eight seconds was associated with those who were older, had higher body mass index (BMI), and had lower levels of 25(OH)D and albumin.

Conclusion:

The unique finding of this study is that increased report of balance and falling issues (both perceived and measured), and slower gait was found in persons with increased CKD severity and lower 25(OH)D status.

Introduction

Chronic kidney disease (CKD) is a significant and expensive health problem that affects approximately 15% of the US population (1). In 2014, 118,000 people started dialysis treatment for end-stage renal disease (ESRD), and 662,000 people were living on chronic dialysis or with a kidney transplant (1). Contributing to the cost of care for this patient population is the treatment of fractures. The risk of fracture is largely increased in participants with ESRD (2). Participants with CKD have a high risk of fracture at every stage of their disease (2), but the risk is exacerbated as the disease progresses, with the highest incidence rates occurring for those with ESRD who are on dialysis (2,3). The mortality and morbidity after fracture is particularly grim in these participants (4) with one study finding a one-year mortality rate of 64% after hip fracture (5). Risk factors for fracture in this population include older age, female sex, Caucasian ethnicity, and lower body mass index (BMI) (6), yet low vitamin D status may also contribute to an increased risk of fracture through its effects on bone and balance.

Although fractures can occur spontaneously, they typically occur after a fall in this population, as the increased risk of falls correlates with the increased risk of fracture (7). Falls are a marker of functional disability and frailty (8,9) as well as a major source of injury and morbidity in participants with CKD (10). Increased falls have been well-documented in elderly participants with CKD both on dialysis (3,11) and not on dialysis (6,12), as well as in younger participants at all stages of CKD (6). The relative risk of falls when compared to the general population is four to five times higher (11,13). Falls are dangerous for patients even if they do not result in fractures, as fallers can suffer from head injury, wounds, and the fear of falling, which leads to subsequent inactivity (14). Falls may be the result of physical function deficits, such as impairments in balance and gait speed, found in this population (15,16). Low Vitamin D levels could potentially contribute to the high risk of fracture in participants with CKD. Low 25(OH)D levels in CKD are associated with poor physical activity (15), increased fractures (12), and fall risk (16), but it is unknown whether increasing 25(OH)D levels will counteract such problems (17).

Vitamin D deficiency, as measured by 25(OH)D (the best serum marker for vitamin D status), is reported to be as high as 70% to 90% in the dialysis population, depending on the cutoff that is used (18–21). Vitamin D deficiency is associated with increased mortality in participants with CKD (22–24), and as an extrarenal role for 25(OH)D is well accepted (i.e., vitamin D-mediated innate immunity) (25), it is becoming more common to treat patients with nutritional vitamin D supplements to improve 25(OH)D levels (26). Additional reasons for supplementation include improvement in 1,25(OH)₂D synthesis and the capacity to bind to the vitamin D receptor, prevention of secondary hypothyroidism, and decrease in mortality (27). Although it has yet to be established in participants with CKD, a recent meta-analysis suggests that vitamin D supplementation decreases the risk of falls by 14% in the general elderly population (28). It has also been found to improve gait speed in individuals older than age 70 (29).

Therefore, the objectives of this study were to analyze data from the 1999-2004 National Health and Nutrition Examination Survey (NHANES) to: 1) determine the prevalence of balance deficits and 2) describe the relationships between CKD stage/severity and Vitamin D status, self-reported balance and falling problems, and measured gait speed in the CKD sample.

Methods

Data from the NHANES 1999-2004 datasets, consisting of cross-sectional data from multiple time points, were used for analyses. For more details regarding the NHANES sampling and methodology, please refer to https://www.cdc.gov/nchs/nhanes/index.htm. This study was exempt from IRB approval.

Selection Criteria

Participants over the age of 40 years were selected for analysis; in total, 8,554 subjects’ data were available for use. All subjects were classified by CKD severity using Glomerular Filtration Rate (GFR). GFR was calculated for each person using sex (variable RIAGENDR), race (variable RIDRETH1), age (variable RIDAGEYR), and serum creatinine (variable LBXSCR)(30) via a series of 4 formulas specific for gender and race (Non-Hispanic White / African American) published by Levey et al (31). By calculating the GFR, all subjects were initially classified into CKD stages 1-5 and then combined into three groups: Control group (n=7354), Stage 3A/B group (n=1090), and Stage 4/5 group (n=110). The Control group had a GFR ≥ 60 mL/min and included normal controls as well as stage 1 & 2 CKD, the Stage 3 A/B group had a GFR between 30-59 mL/min, and the Stage 4/5 group had a GFR ≤ 29 mL/min (Table 1).

Table 1.

Estimated U.S. population level proportions (%) of CKD stages and low 25(OH)D levels

ORIGINAL STAGES IN NHANES DATASET			NEW STAGES CREATED
Original Stage	N		New Stage	N
Stage 1 (normal/high GFR)	3383	→	Control	7354
Stage 2 Mild CKD	3971
Stage 3A Moderate CKD	813	→	Stage 3A/3B	1090
Stage 3B Moderate CKD	277
Stage 4 Severe CKD	75	→	Stage 4/5	110
Stage 5 End Stage CKD	35
		NHANES 1999-2000	NHANES 2001-2002	NHANES 2003-2004
OVERALL
CKD Stage
Stage 1 with Normal or High GFR (GFR > 90 mL/min)		80.30	48.34	47.30
Stage 2 Mild CKD (GFR = 60-89 mL/min)		17.02	44.46	45.81
Stage 3A Moderate CKD (GFR = 45-59 mL/min)		1.54	5.31	5.06
Stage 3B Moderate CKD (GFR = 30-44 mL/min)		0.79	1.47	1.35
Stage 4 Severe CKD (GFR = 15-29 mL/min)		0.11	0.26	0.41
Stage 5 End Stage CKD (GFR <15 mL/min)		0.23	0.16	0.07
Low 25(OH)D (< 32 ng/ml)			79.66	76.20
Low 25(OH)D (< 24 ng/ml)		N/A	45.96	44.66
Low 25(OH)D (< 12 ng/ml)			4.94	6.12
LIMITED TO PERSONS AGE 40+ YEARS
CKD Stage
Stage 1 with Normal or High GFR (GFR > 90 mL/min)		65.37	26.81	26.35
Stage 2 Mild CKD (GFR = 60-89 mL/min)		29.44	59.78	60.99
Stage 3A Moderate CKD (GFR = 45-59 mL/min)		3.03	9.90	9.24
Stage 3B Moderate CKD (GFR = 30-44 mL/min)		1.54	2.84	2.53
Stage 4 Severe CKD (GFR = 15-29 mL/min)		0.23	0.47	0.77
Stage 5 End Stage CKD (GFR <15 mL/min)		0.40	0.21	0.11
Low 25(OH)D (< 32 ng/ml)			81.29	77.84
Low 25(OH)D (< 24 ng/ml)		N/A	47.40	47.00
Low 25(OH)D (< 12 ng/ml)			5.28	7.26

NOTE: SURVEYFREQ procedure in SAS was used to estimate the prevalence of various CKD stages and low Vitamin D levels for all persons within each NHANES period and then separately limiting to those over 40 years old.

Selection of Variables

The independent variable chosen for this analysis was severity of CKD stage per National Kidney Foundation Kidney Disease Outcomes Quality Initiative (32). The dependent variables were: the measurement of balance, the perception of balance and falling issues, measured gait speed, levels of 25(OH)D (LBDVIDMS), and levels of albumin (LBXSAL). Albumin was selected as an overall marker of nutritional status. Levels of 25(OH)D were recorded in nmol/L which we converted to ng/mL by multiplying the recorded values by a factor of 0.400641. Serum albumin was recorded in g/dL.

The measurement of balance was investigated using the Romberg test of standing balance on firm and compliant support surfaces (variables BAXPFC## and BAXFTC##, where ## = 11, 12, 21, 22, 31, 32, 41, and 42). The four conditions were: 1) firm surface with eyes open, 2) firm surface with eyes closed, 3) foam surface with eyes open, and 4) foam surface with eyes closed. Conditions one and two were conducted for 15 seconds and three and four were conducted for 30 seconds. Test failure was defined as: 1) opening eyes in an eyes-closed condition, 2) moving arms or feet to achieve stability, or 3) beginning to fall or requiring technician intervention.

The perception of balance and falling issues was assessed through answering ‘yes’, ‘no’, or ‘don’t know’ to the following questions: “During the past 12 months, have you had dizziness, difficulty with balance, or difficulty with falling?” (variable BAQ10), “Have you had difficulty with balance?” (variable BAQ20B), and, “Have you had difficulty with falling?” (variable BAQ20C). The latter two questions were only asked if subjects answered ‘yes’ to the first question.

To calculate gait speed, time in seconds to complete a 20-foot walk was used (variables MSXWTIME for 1999-2000 and MSXW20TM for 2001-2002). While subjects walked at their preferred pace, the 20-foot walk was timed using a hand-held stopwatch. The walk times were converted to walking speeds and an abnormally slow walking speed was determined as ≥ 0.8 m/s (27). Therefore, if subjects took longer than approximately eight seconds to complete the 20-foot walk, they were classified as having abnormal gait speed.

Age, sex, and BMI (variable BMXBMI), and smoking status (SMQ020, SMQ040, SMQ120, SMQ140, SMQ150, SMQ170) were selected as covariates for inclusion during the estimation of odds ratios.

Statistical Analysis

Prevalence of the various CKD stages and low 25(OH)D were estimated for the entire US population and for those 40+ years old at each of the three administrations of the NHANES examined in this study. For this analysis, we used SAS SURVEYFREQ, with strata (SDMVSTRA), cluster (SDMVPSU), and weight (WTMEC2YR) variables. For all other analyses, simple random samples were assumed and analyses were conducted without strata, cluster, or weight variables. Levels of 25(OH)D, serum albumin, Romberg test fail times, and 20-ft walk times were compared amongst CKD groups. Continuous outcomes were compared using ordinary least squares regression models for both unadjusted and age-adjusted analysis. Unadjusted models for categorical variables used were Pearson chi-square tests or Fisher’s exact tests. Age-adjusted models for categorical measures were done using logistic regression models. To evaluate the associations between CKD stage/severity and 25(OH)D levels, self-reported balance/falling problems, and measured gait speed, multivariable logistic regression models controlling for age, sex, BMI, and smoking status were used. Odds ratios and their 95% confidence intervals were calculated. P-values less than 0.05 were considered significant, and all statistical analyses were conducted in SAS (v9.4, SAS Institute, Inc., Cary, NC).

Results

After adjusting for age, when standing on a firm surface with eyes open, failure rates increased with kidney disease severity, with the Control, Stage 3A/B, and Stage 4/5 groups failing 0%, 1%, and 3% of the time, respectively (Table 2). The 3% failure rate in the Stage 4/5 group was significantly increased as compared to the Control group (p=0.02). In the most difficult condition (i.e. standing on a foam surface with eyes closed), failure rates again increased with kidney disease severity, with the Control, Stage 3A/B, and Stage 4/5 groups failing 54%, 76%, and 85% of the time, respectively. Similarly, the Stage 4/5 group had a significantly increased fail rate compared to the Control group (p=0.02).

Table 2.

25(OH)D, albumin, balance, and gait Results for each of the three CKD stages

	N	CKD GROUP			UNADJUSTED P-VALUE	AGE ADJUSTED P-VALUE
		Control	Stage 3A/3B	Stage 4/5
25(OH)D (ng/ml)	5863	22.9 (8.7)	24.0 (8.9)	20.5 (9.2)	<0.001
Serum Albumin (g/dL)	8554	4.3 (0.3)	4.1 (0.4)	4.0 (0.5)	<0.001
Balance Exam
Condition 1, Trial 1 Fail	5654	11 (0)	8 (1)	2 (3)	<0.001	0.04
Condition 2, Trial 1 Fail	5642	248 (5)	98 (14)	11 (17)	<0.001	0.18
Condition 3, Trial 1 Fail	5442	94 (2)	54 (8)	5 (8)	<0.001	0.09
Condition 4, Trial 1 Fail	5351	2523 (54)	462 (76)	47 (85)	<0.001	0.03
Condition 1, Trial 1 Fail Time (sec)	21	6.6 (4.7)	6.4 (4.0)	12.0 (1.4)	0.25	0.25
Condition 2, Trial 1 Fail Time (sec)	357	6.2 (3.6)	6.0 (3.6)	6.6 (4.5)	0.80	0.67
Condition 3, Trial 1 Fail Time (sec)	150	8.8 (7.9)	8.2 (8.2)	4.2 (6.2)	0.44	0.44
Condition 4, Trial 1 Fail Time (sec)	3031	8.0 (6.2)	6.7 (5.6)	7.0 (6.3)	<0.001	0.83
20’ Walk Time (sec)		6.8 (2.9)	8.3 (5.0)	9.8 (6.7)	<0.001	<0.001

NOTE: Values shown are Mean (SD) for continuous variables and n (%) for categorical. The p-values were against a null hypothesis that group means were equal for continuous variables and row and column variables were independent for categorical variables. Both unadjusted and age-adjusted p-values are presented.

The conditions are as follows: condition 1) firm surface with eyes open, 2) firm surface with eyes closed, 3) foam surface with eyes open, and 4) foam surface with eyes closed. Conditions one and two were conducted for 15 seconds and three and four were conducted for 30 seconds. Test failure was defined as: 1) opening eyes in an eyes-closed condition, 2) moving arms or feet to achieve stability, or 3) beginning to fall or requiring technician intervention.

The amount of time to complete the walking test increased (i.e. slower gait speed) with kidney disease severity. The mean time it took for subjects to complete the 20-foot walk was 6.8 seconds for the Control group, 8.3 seconds for the Stage 3A/B group, and 9.8 seconds for the Stage 4/5 group (Table 2). The Stage 4/5 mean time was significantly longer than the mean times for the Stage 3A/B group and the Control group (p<0.001 for both). In addition, the Stage 3A/B group took significantly longer than the Control group to complete the task (p=0.003).

Increased odds of self-reporting dizziness, difficulty with balance, or difficulty with falling in the past year were found in the Stage 3A/3B group (odds ratio and (95% CI) [OR]: 1.54 (1.29, 1.83) vs. Controls), Stage 4/5 group (OR: 2.07 (1.26, 3.40) vs. Controls), older subjects (OR: 1.02 (1.01, 1.03) vs. young subjects), females (OR: 1.73 (1.52, 1.98) vs. males), former smokers (OR: 1.35 (1.17, 1.56) vs. never smokers), current smokers (OR: 1.53 (1.28, 1.82) vs. never smokers), subjects with lower levels of 25(OH)D (OR: 0.99 (0.98, 1.00) vs. higher levels), and subjects with lower levels of albumin (OR: 0.79 (0.65, 0.98) vs. higher levels) (Table 3).

Table 3.

Multivariable logistic regressions of any (N=5592), balance (N=5591) and falling (N=5590) problems in the past year

Variable	SELF-REPORT A DIZZINESS, BALANCE AND/OR FALL PROBLEM			SELF-REPORT A BALANCE PROBLEM			SELF-REPORT A FALLING PROBLEM
	Odds Ratio	OR 95% CI	p-value	Odds Ratio	OR 95% CI	p-value	Odds Ratio	OR 95% CI	p-value
CKD Stage
Normal/Controls	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Stage 3A/3B	1.54	1.29, 1.83	<0.001	1.44	1.18, 1.75	<0.001	1.39	1.07, 1.83	0.02
Stage 4/5	2.07	1.26, 3.40	0.004	1.71	1.01, 2.90	0.05	1.29	0.63, 2.62	0.49
Age (Years)	1.02	1.01, 1.03	<0.001	1.03	1.02, 1.03	<0.001	1.04	1.03, 1.05	<0.001
Sex
Male	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Female	1.73	1.52, 1.98	<0.001	1.51	1.29, 1.76	<0.001	1.65	1.30, 2.08	<0.001
Body Mass Index	1.00	0.99, 1.01	0.51	1.00	0.99, 1.01	0.88	1.01	0.99, 1.03	0.32
Smoking Status
Never	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Former	1.35	1.17, 1.56	<0.001	1.37	1.16, 1.63	<0.001	1.34	1.04, 1.72	0.02
Current	1.53	1.28, 1.82	<0.001	1.69	1.37, 2.08	<0.001	1.66	1.21, 2.28	0.002
Vitamin D (ng/ml)	0.99	0.98. 1.00	0.01	0.99	0.98, 1.00	0.006	0.97	0.96, 0.99	<0.001
Albumin (g/dL)	0.79	0.65, 0.98	0.03	0.65	0.51, 0.83	0.001	0.55	0.39, 0.78	0.001

NOTE: The p-values presented are comparison against the reference.

The first question asks whether the subject has had a dizziness, balance, and/or falling problem in the past year. If the subject answers yes to the first question (columns 2-4), two subsequent questions are asked: whether the subject had a balance problem in the past year (columns 5-7) or whether the subject had a falling problem in the past year (columns 8-10). Data is included from 2001-2004.

Of those who reported a problem, two additional questions were asked to determine whether it was a balance or a falling problem (Table 3). Increased odds of reporting a balance problem in the past year were found for the Stage 3A/3B group (OR: 1.44 (1.18, 1.75) vs. Controls), Stage 4/5 group (OR: 1.71 (1.01, 2.90) vs. Controls), older subjects (OR: 1.03 (1.02, 1.03) vs. younger subjects), females (OR: 1.51 (1.29, 1.76) vs. males), former smokers (OR: 1.37 (1.16, 1.63) vs. never smokers), current smokers (OR: 1.69 (1.37, 2.08) vs. never smokers), lower levels of 25(OH)D (OR: 0.99 (0.98, 1.00) vs. higher levels), and lower levels of albumin (OR: 0.65 (0.51, 0.83) vs. higher levels). Increased odds of reporting a falling problem within the past year were found for the Stage 3A/3B group (OR: 1.39 (1.07, 1.83) vs. Controls), older subjects (OR: 1.04 (1.03, 1.05) vs. younger subjects), females (OR: 1.65 (1.30, 2.08) vs. males), former smokers (OR: 1.34 (1.04, 1.72) vs. never smokers), current smokers (OR: 1.66 (1.21, 2.28) vs. never smokers), lower levels of 25(OH)D (OR: 0.97 (0.96, 0.99) vs. higher levels) and lower levels of albumin (OR: 0.55 (0.39, 0.78) vs. higher levels).

Increased odds of taking longer than eight seconds to complete the 20-foot-walk were found for the Stage 4/5 group (OR: 3.39 vs. Controls), older subjects (OR: 1.12 vs. younger subjects), females (OR: 1.35 vs. males), subjects with higher BMI (OR: 1.06 vs. lower BMI), subjects with lower levels of 25(OH)D (OR: 0.97 vs. higher levels), and subjects with lower levels of albumin (OR: 0.38 vs. higher levels) (Table 4).

Table 4.

Multivariable logistic regression of 20ft walk time longer than 8 seconds (N = 1823)

Variable	MUTIVARIABLE MODEL
	Odds Ratio	OR 95% CI	p-value
CKD Stage
Normal/Controls	Ref.	Ref.	Ref.
Stage 3A/3B	1.13	0.82, 1.57	0.46
Stage 4/5	3.39	1.29, 8.95	0.01
Age (Years)	1.12	1.10, 1.14	<0.001
Sex
Male	Ref.	Ref.	Ref.
Female	1.35	1.00, 1.81	0.05
Body Mass Index	1.06	1.03, 1.09	<0.001
Smoking Status
Never	Ref.	Ref.	Ref.
Former	0.87	0.64, 1.19	0.38
Current	1.40	0.91, 2.16	0.13
Vitamin D (ng/ml)	0.97	0.95, 0.99	<0.001
Albumin (g/dL)	0.38	0.23, 0.63	<0.001

NOTE: The p-values presented are comparison against the reference.

Data is included from 2001-2002, as Vitamin D data was only available for 2001-2004 and 20ft walk time data was only available from 1999-2002.

Discussion

The purpose of this study was to use the NHANES 1999-2004 dataset to determine the prevalence of balance deficits, and to describe the relationships between CKD severity and 25(OH)D status, self-reported balance and falling problems, and measured gait speed in a CKD sample. The prevalence of balance deficits and abnormal gait speed were found to be higher than Controls in this CKD sample, with fail rates increasing with increasing kidney disease severity. In addition, the likelihood, as defined by odds ratios, of participants self-reporting a balance and falling problem in the past year was higher in participants with advanced stages of CKD, older subjects, females, former and current smokers, and subjects with lower levels of 25(OH)D and albumin. Similarly, the likelihood of having an abnormally slow gait speed was found in participants with advanced stages of CKD, older subjects, females, subjects with higher BMI, and subjects with lower levels of 25(OH)D and albumin.

One potential explanation for the relationships between kidney disease severity and impaired balance, slower gait, and increased falls could be the result of decreased physical activity in the CKD population. Approximately 28% of the CKD sample in the NHANES III (1999-2004) dataset were found to be inactive (33), and potentially increasing physical activity might be a survival mechanism in this population. Physical inactivity in participants with CKD has been associated with increased mortality rates and reduced strength (17,21). Decreased lower-extremity muscle strength negatively impacts balance and gait in nursing home residents and community-dwelling older adults, leading to an increased risk of falls (34). Decreased physical activity can also be an indicator of underlying fragility, and fragility fractures are common in participants with CKD, leading to increased morbidity, mortality, and cost for this population (33,35).

Furthermore, slower gait speed increases the risk of impaired balance and therefore exacerbates the occurrence of falls. Gait speed slows with age (36), but slower gait speed has also been associated with the fear of falling (9). Specifically, slower gait speed, shorter stride length, increased stride width, and longer double limb support time were associated with the fear of falling in community-dwelling seniors (37). Community-dwelling seniors with a higher risk of fractures also had a heightened fear of falling and impaired balance compared to age-matched controls (38).

Based on our results, it was determined that vitamin D insufficiency (defined as 25(OH)D <32 ng/ml) was prevalent in the CKD sample in 81.29% from 2001-2002 and in 77.84% from 2003-2004. There was a strong negative association between vitamin D status and balance and falls. Vitamin D deficiency has been found to be associated with low muscle mass, poor functional performance, and disability (39–43), so improving vitamin D status in this sample could potentially improve muscle strength and balance and lead to a reduced risk of falls and fracture. Discovery of a vitamin D receptor in skeletal muscle makes this a potential treatment target (44). Vitamin D supplementation in various forms of CKD (i.e., nondialysis-dependent CKD, hemodialysis, peritoneal dialysis, and renal transplant recipients) led to a small, but significant, effect on lower limbs muscle strength, especially in those persons with very low 25(OH)D (< 12ng/ml) (18). In older adults with stable health, vitamin D supplementation >700 IU had a reduction in falls by approximately 20%, while lower doses did not show a fall reduction (42). These findings suggest there is reason to believe that sufficient vitamin D supplementation to raise 25(OH)D levels above a threshold may have the potential effects of increasing muscle strength and improving balance in CKD participants and ultimately reducing falls and fractures.

Although the research on vitamin D reducing falls and fractures is compelling, some recent studies have found that high dose vitamin D is associated with an increased risk of falls and fractures. Older adult women with vitamin D insufficiency given vitamin D supplementation for 12-months reported an increased rate of falls in higher doses of 25(OH)D (4000 and 4800 IU daily) compared to medium doses (1600, 2400, 3200 IU), but not compared to the placebo group (45). Higher monthly doses (60,000 IU) of vitamin D₃ in ambulatory older adults aged 70 years and older did not improve lower extremity function and was associated with an increased risk of falls compared to the low-dose control group (24,000 IU) (46). Community-dwelling women over the age of 69 years given one annual dose of 500,000 IU of cholecalciferol experienced 15% more falls and 26% more fractures compared to the placebo group (47). Based on these mixed results regarding the association between vitamin D supplementation and fall/fracture risk, more research is needed to determine whether vitamin D supplementation can be a potential solution for preventing falls and fractures and if dosing in large infrequent doses is detrimental compared to smaller daily doses. Furthermore, due to the unique pathophysiology of renal disease, supplementation in the CKD population may result in different outcomes compared to other populations.

The potential role for vitamin D in the improvement of muscle strength or balance has not been extensively studied in participants with CKD. In persons with end-stage renal disease, decreased muscle strength and increased falls risk were associated with 25(OH)D levels (11). Most interventional vitamin D trials in hemodialysis patients have not had functional outcomes as a primary outcome and studies that have reported functional tests as a secondary outcome have been disappointing. Vitamin D supplementation has no effect on grip strength, six-minute walks, repeated chair stands, or postural stability after six months in hemodialysis patients (48). Blair et al. (2008) treated 344 participants receiving hemodialysis with Vitamin D deficiency (< 40 ng/ml) with 50,000 IU of Vitamin D₂ weekly for six months and found no effect on quality of life surveys (reported drop-out rate of 97%) (19).

There are limitations of the current study. First, examining the NHANES cross-sectional dataset allows for a limited scope in examining the relationships between CKD severity and impairments in 25(OH)D levels, balance, and gait speed. A single creatinine level may be imprecise of a person’s kidney function as it is influenced by extremes of muscle mass, medications, unusual diets and conditions that influence renal excretion of creatinine. Further, albumin may not be a reliable indicator of nutritional status in this population (49–51). Balance and gait measurements are not biomechanically defined and include the overall observations of balance errors and gait speed. Future studies should examine the CKD population using a quantitative biomechanical analysis to better understand the factors contributing to impaired posture and gait. Second, the variables used to test perceived balance and falling problems were subjective to opinion. Future research could have subjects record this data in a diary. Third, it is possible that participants receiving dialysis and/or transplant were included in the analysis. NHANES data was not available regarding transplant. Upon inspection, it was found that 12 of the 110 subjects used for Stage 4/5 in the current study reported having had dialysis in the past 12 months. However, this information was not available for all years and was not specific in the chronicity of disease. Therefore, it is possible that this 11% of the Stage 4/5 cohort may have confounded the analysis. Fourth, it is important to consider frailty, inflammation and poor intake may lead to low 25(OH)D and albumin levels independent of kidney disease. Frailty was not examined in this current analysis and should be considered for future investigations. Fifth, we used the Modified of Diet in Renal Disease (MRDR) study equation to calculate GFR. While this performs well in GFR <60 mL/min/1.73m² populations which was the focus of this paper, it wouldn’t differentiate those who fell in our control group with GFR >60 mL/min/1.73m². Lastly, we combined normal kidney function and Stage 1 and 2 CKD and used these participants as the control group. It is possible that deficits in balance and gait are present very early on in the kidney disease stages which our study would not have elicited. This warrants an additional study to determine the onset of balance and gait problems in terms of disease progression.

Practical Application

The unique finding of this study is that the likelihood of reporting more balance and falling issues (both perceived and measured) and having slower gait is higher in persons with increased CKD severity and those with lower 25(OH)D levels. Patients with moderate to severe CKD may benefit from screening and treatment of both balance and vitamin D status.