Visual Abstract
Keywords: progression of renal failure, frailty, older persons, geriatric nephrology, glomerular filtration rate, independent living, C-reactive protein, proportional hazards models, weight loss, follow-up studies, hand strength, chronic renal insufficiency, comorbidity, EGFR protein, human, receptor, epidermal growth factor, exercise, phenotype, gait
Abstract
Background and objectives
Low eGFR is known to be associated with frailty, but the association between the longitudinal decline of eGFR and incident frailty in older persons remains to be determined. The objective of this study was to investigate whether a fast decline on eGFR would be associated with incident frailty.
Design, setting, participants, & measurements
Community dwellers, aged ≥70, were included in this secondary analysis of the 5-year Multidomain Alzheimer Preventive Trial (MAPT). eGFR was calculated using CKD–Epidemiology Collaboration equation at baseline and at 6, 12, and 24 months. The lowest quartile of eGFR slope (−4.1 ml/min per 1.73 m2 per yr) defined a fast decline. The frailty phenotype (unintentional weight loss, exhaustion, low physical activity, slow gait, low handgrip strength assessed with a 0–5 score, where higher is worse; a score ≥3 defines frailty) was assessed at baseline, 6, 12, 24, 36, 48, and 60 months. Cox models were used to test the association between fast eGFR decline and incident frailty.
Results
A total of 833 participants were frail neither at baseline nor at 2 years and had appropriate follow-up data. Median (IQR) baseline eGFR was 73 (61–84) ml/min per 1.73 m2. Frailty occurred in 95 (11%) participants between 24 and 60 months. Among them, 31/207 (15%) had fast eGFR decline between baseline and 24 months, and 64/626 (10%) did not. In a Cox model adjusted for demographic variables, cardiovascular comorbidity, C-reactive protein, and baseline eGFR, a fast eGFR decline was associated with incident frailty (HR 1.67, 95% CI 1.03 to 2.71). Sensitivity analyses provided consistent findings.
Conclusions
In community-dwelling older adults with relatively preserved baseline eGFR, a fast eGFR decline is associated with incident frailty.
Introduction
Frailty is an age-associated state of reduced physiologic reserve and increased vulnerability to stressors and is associated with higher risks of subsequent falls, hospitalizations, disability, and death (1). Biomarkers associated with frailty incidence could be useful for prevention of functional decline and reduction of health costs (2). Low kidney function, as assessed by estimated glomerular filtration rate (eGFR), is associated with morbidity and mortality in older adults (3), but also with poorer physical performance, loss of independence (4,5), and prevalent (6,7) and incident frailty (8).
Nevertheless, a stable, moderately low eGFR (45–60 ml/min per 1.73 m2) is common in older persons due to physiologic senescence of the kidneys, and is not systematically associated with poor outcome (9). By contrast, fast decline of eGFR, a hallmark of an active pathologic process, has been associated with mortality (10), cardiovascular morbidity (11), and cognitive decline (12,13) in older persons. However, to the best of our knowledge, the association between a fast decline of eGFR and incident frailty has never been assessed.
We hypothesized a fast decline of eGFR would be associated with incident frailty among community-dwelling older adults. We also hypothesized the overtime decline on eGFR would be more strongly associated with incident frailty than baseline eGFR.
Materials and Methods
MAPT Study
This is a secondary analysis of the MAPT (NCT00672685), whose methods were published elsewhere (14,15). Briefly, MAPT was a randomized controlled trial assessing the effects of a 3-year multidomain lifestyle intervention, a 3-year omega-3 fatty acids supplementation, or their combination on cognition. The multidomain intervention consisted of nutritional and physical activity counseling plus cognitive training. Participants were recruited in 13 memory centers in France and Monaco between May 2008 and February 2011. A total of 1679 participants were randomized into four groups (the three above-mentioned interventions plus a placebo control group). The primary analysis showed no effect of the interventions on the 3-year evolution of the composite cognitive score (14). Two additional years of observational clinical follow-up were proposed to participants who completed the 3-year interventional study. The 5-years follow-up ended in April 2016. The protocol was approved by the ethical committee of Toulouse (Comité de Protection des Personnes Sud-Ouest et Outre Mer II) and each participant provided informed consent.
Participants
MAPT included community dwellers, aged ≥70 years, with spontaneous memory complaints and/or dependency in one or more instrumental activity of daily living (i.e., ability to use the telephone, buy groceries, prepare meals, do housework, do laundry, use transport, take medications, use money) (16) and/or usual gait speed <0.8 m/s. Exclusion criteria were a Mini Mental State Examination (17) score of <24, dementia diagnosis, dependency in any basic activities of daily living (18) (bathing, dressing, toileting, continence, transferring, feeding), and an ongoing omega-3 fatty acids supplementation at baseline.
Frailty
Frailty was assessed at baseline and at 6, 12, 24, 36, 48, and 60 months, according to the following five criteria of the phenotype model (19).
Unintentional weight loss: self-reported, >4.5 kg in the prior year.
Exhaustion: self-reported using two items (“I felt that everything I did was an effort,” “I could not get going”) from the Center for Epidemiologic Studies Depression Scale (20). The criterion was present if persons reported feeling this way (either of the two items) “a moderate amount of the time (3–4 days)” for “most of the time” during the prior week.
Low physical activity: self-reported using the 15-item Minnesota Leisure Time Physical Activity Questionnaire. The criterion was present with physical activity of <383 kcal/wk in men and <270 kcal/wk in women during the prior 2 weeks.
Slow gait: sex- and height-specific cutoffs (<0.65 m/s in men ≤173 cm and women ≤159 cm; <0.76 m/s in men >173 cm and women >159 cm) were applied on a 4-m usual walk test.
Low handgrip strength: measured with a handheld dynamometer. Cutoffs were ≤29, 30, and 32 kg in men with body mass index (BMI) ≤24, between 24.1 and 28, and >28 kg/m2, respectively; 17, 17.3, 18, and 21 kg in women with BMI ≤23, between 23.1 and 26, between 26.1 and 29, and >29 kg/m2, respectively.
Frailty condition was defined as meeting three or more criteria.
eGFR
Creatinine was measured in each center, with isotope dilution mass spectrometry–traceable colorimetric or enzymatic methods (according to local routine procedure) at baseline and at 6, 12, and 24 months. We calculated eGFR using age, sex, and serum creatinine in the CKD–Epidemiology Collaboration study equation (21). The equation provided eGFR expressed in ml/min per 1.73 m2.
In each participant with two to four available creatinine measurements, eGFR slope over time was calculated using a simple linear regression model and expressed in ml/min per 1.73 m2 per yr. The dependent variable in this model was eGFR, i.e., GFR values estimated at the different time points. The independent variable was time, i.e., the time points in years at which serum creatinine was measured. In this approach, the assumption is that there is an underlying eGFR trajectory for each patient, and the eGFR values vary randomly around this linear trajectory (22).
Because there is no consensus definition of a fast eGFR decline (23), the lowest quartile of eGFR slope among participants included in the primary analysis was defined as the fast decline group (i.e., more rapid than −4.1 ml/min per 1.73 m2 per yr) in this study. A binary variable was created: participants with fast eGFR decline versus others.
Confounders
Based on the literature (24) and available data, potential confounders for the association of a fast eGFR decline and incident frailty were age (categorized as 70–74, 75–79, and >79 years), sex, study intervention group, BMI (categorized as <21, 21–25, 25–30, and >30 kg/m2), history of diabetes, hypertension, heart failure, stroke, ischemic heart disease, C-reactive protein (CRP; categorized as <3, 3–10, and >10 mg/L), and baseline eGFR or mean eGFR. Death during follow-up was added as a confounder in a sensitivity analysis. Comorbidities, BMI, and CRP were considered time-varying covariates, using data measured concurrently with eGFR (i.e., between baseline and 24 months in the primary analysis and each sensitivity analyses except number 4; and between baseline and 12 months in sensitivity analysis 4). When examining the association of baseline eGFR and incident frailty, potential confounders were the same, except eGFR, and only baseline data were used.
Statistical Analysis
Descriptive statistics were presented as median and interquartile range (IQR) for continuous variables and absolute numbers and percentages for categoric variables. The Wilcoxon and the chi-squared tests were used to assess baseline differences between participants with a fast eGFR decline versus others.
Frailty incidence between 24 and 60 months was determined among nonfrail participants at baseline and at the 24-month visit; we excluded participants with a frailty condition or missing data on frailty at these two time points, as well as participants with a frailty condition at the 6- and/or 12-month visit. People with no data on frailty after 24 months were also excluded.
The association between fast eGFR decline and frailty incidence was tested using a Kaplan–Meier analysis and Cox models with time as a discrete variable and a time-to-first event approach. Participants were censored at the last clinical visit assessment (end of study, loss to follow-up, or death) if they did not become frail. Cox models included all of the potential confounders described above.
In addition, several sensitivity analyses were performed. In sensitivity analysis 1, we used an alternative cutoff to define a fast eGFR decline (−5 ml/min per 1.73 m2 per yr), as proposed by international guidelines, despite the lack of consensus on this point (23). In sensitivity analysis 2, we excluded participants with eGFR available at only two time points to minimize the creatinine random variation bias. In sensitivity analysis 3, we expressed eGFR slope as a percentage of baseline eGFR (eGFR slope divided by baseline eGFR) to take into account the wide distribution of baseline eGFR. The lowest quartile, which defined a fast eGFR decline, was −6%/yr. In sensitivity analysis 4, eGFR slope was calculated on the first three eGFR time points (i.e., baseline and 6 and 12 months) and incident frailty was assessed between 12- and 60-month visits, to assess whether eGFR decline in a shorter period of time was associated with subsequent frailty. In line with the primary analysis, participants with a frailty condition or without data regarding frailty at baseline or the 12-month visit, as well as participants with a frailty condition at the 6-month visit, were excluded from this sensitivity analysis. Because the study intervention could have an effect on the outcome (25), sensitivity analysis 5 was performed in the placebo group only. For the same reason, in sensitivity analysis 8, an interaction was added between study intervention group and a fast eGFR decline. Sensitivity analysis 6 was performed with adjustment on mean eGFR (calculated with the same eGFR time points used to calculate eGFR slope) instead of baseline eGFR. In sensitivity analysis 7, death during follow-up was added as a confounder in the Cox model. In sensitivity analysis 9, we used a model with three groups defined by baseline to 24-month eGFR slope quartiles: stable eGFR (between lowest and highest quartiles, considered as reference), eGFR fast decline (lowest quartile, more rapid than −4.1 ml/min per 1.73 m2 per yr), and eGFR improvement (highest quartile, >3.8 ml/min per 1.73 m2 per yr). In sensitivity analysis 10, eGFR slope was considered a continuous variable. The definition of a fast eGFR decline determined using the lowest quartile in the primary analysis (more rapid than −4.1 ml/min per 1.73 m2 per yr) was applied to each sensitivity analyses, except numbers 1 and 3 which used alternative cutoffs, and sensitivity analysis 10.
Finally, associations between baseline eGFR (as a continuous variable and as a binary variable: more rapid than or ≥60 ml/min per 1.73 m2) and incident frailty (between baseline and the 60-month visit) were tested using Cox models with the same approach as described above.
Statistical analyses were performed using JMP (Cary, NC) Pro 12.0 and Stata (College Station, TX) version 14.0.
Results
Participants
A total of 2591 people were assessed for eligibility, and 1679 were included in the MAPT. Of these, 1286 and 939 completed the 3- and 5-year follow-up, respectively. The primary analysis of this study included 833 participants who were nonfrail both at baseline and 24 months, had appropriate follow-up data regarding frailty, and eGFR available at two time points, at least, within the first 24 months of the study. A total of 833, 619, and 538 participants completed the 3-, 4-, and 5-year follow-up (mean±SD follow-up time, 53±10 months), respectively. Death was the cause of end of follow-up in 11 of 833 (1%) participants. The baseline characteristics of the 833 participants included in the primary analysis are depicted in Table 1. Participants with fast eGFR decline had a history of ischemic heart disease more often than patients without fast eGFR decline.
Table 1.
Baseline characteristics of 833 participants of the MAPT according to 2-year slope of eGFR (primary analysis)
| Variable | Total (n=833) | 2-yr eGFR Slope | |
|---|---|---|---|
| No Fast Decline (n=626) | Fast Decline (n=207)a | ||
| Age (yr) | 74 (71–78) | 74 (71–78) | 73 (71–77) |
| Women | 546 (66) | 411 (66) | 135 (65) |
| Comorbidity | |||
| Hypertension | 414 (50) | 307 (49) | 107 (52) |
| Diabetes | 68 (8) | 46 (7) | 22 (11) |
| Stroke | 19 (2) | 14 (2) | 5 (2) |
| Heart failure | 12 (1) | 10 (2) | 2 (1) |
| Ischemic heart disease | 52 (6) | 33 (5) | 19 (9) |
| BMI (kg/m2) | |||
| <21 | 67 (8) | 53 (8) | 14 (7) |
| 21–25 | 288 (35) | 218 (35) | 70 (34) |
| 25–30 | 354 (42) | 267 (43) | 87 (42) |
| >30 | 124 (15) | 88 (14) | 36 (17) |
| eGFR (ml/min per 1.73 m2) | 73 (61–84) | 70 (58–82) | 79 (70–86) |
| eGFR <60 ml/min per 1.73 m2 | 184 (24) | 167 (29) | 17 (9) |
| Annual eGFR slope (ml/min per 1.73 m2 per yr) | −0.5 (−4.1 to 3.8) | 1.4 (−1.15 to 5.6) | −6.9 (−9.6 to −5.5) |
| CRP (mg/L) | |||
| <3 | 522 (70) | 394 (71) | 128 (68) |
| 3–10 | 195 (26) | 143 (26) | 52 (28) |
| >10 | 28 (4) | 21 (4) | 7 (4) |
| Frailty score | |||
| 0 | 516 (62) | 391 (62) | 125 (60) |
| 1 | 250 (30) | 184 (29) | 66 (32) |
| 2 | 67 (8) | 51 (8) | 16 (8) |
| Geriatric depression scale (0–15)b | 2 (1; 4) | 2 (1; 4) | 3 (1; 4) |
| Mini Mental State Examination (0–30)c | 29 (28–29) | 29 (28–29) | 29 (28–30) |
| Short physical performance battery (0–12)d | 11 (10–12) | 11 (10–12) | 11 (10–12) |
| Usual gait speed | 1.08 (0.95–1.26) | 1.09 (0.95–1.25) | 1.07 (0.94–1.32) |
| Study group | |||
| Placebo alone | 211 (25) | 165 (26) | 46 (22) |
| Multidomain intervention with placebo | 209 (25) | 155 (25) | 54 (26) |
| Omega-3 alone | 203 (24) | 150 (24) | 53 (26) |
| Multidomain intervention with omega-3 | 210 (25) | 156 (25) | 54 (26) |
Values are n (%) or median (IQR). Groups were compared using Wilcoxon test for quantitative variables and chi-squared test for categoric variables. eGFR was calculated using the CKD–Epidemiology Collaboration equation.
Fast decline was defined as more rapid than −4.1 ml/min per 1.73 m2 per yr.
Score of 15 represents worst performance.
Score of zero represents worst performance.
Score of zero represents worst performance.
Sample sizes were different in sensitivity analyses because of their design and exclusion criteria. For example, after exclusion of 84 participants with eGFR available at only two time points, sensitivity analysis 2 included 749 participants. In sensitivity analysis 4 (n=919), we did not exclude participants with incident frailty between 12 and 24 months.
eGFR
Median (IQR) baseline eGFR was 73 (61–84) ml/min per 1.73 m2, with higher values in participants with a 2-year fast eGFR decline than in patients without (Table 1).
During the first 2 years of the study, eGFR was available at two, three, and four time points for 13 (6%), 37 (18%) and 157 (76%) participants rated as fast eGFR decline, and for 37 (6%), 118 (19%) and 471 (75%) participants without fast eGFR decline, respectively. In the primary analysis population, the median (IQR) baseline to 24-month eGFR slope was −0.5 (−4.1 to 3.8) ml/min per 1.73 m2 per yr. The distribution of eGFR slope is depicted in Figure 1. We used the lowest quartile in this population to define fast eGFR decline (more rapid than −4.1 ml/min per 1.73 m2 per yr). When the eGFR slope was expressed as a percentage of baseline eGFR in a sensitivity analysis, the lowest quartile was −6%/yr.
Figure 1.
A fast eGFR decline was defined as the lowest quartile of eGFR slope (more rapid than −4.1 ml/min per 1.73 m2 per yr). Distribution of eGFR slope over 2 years in the MAPT study. GFR was estimated using the CKD–Epidemiology Collaboration equation using serum creatinine measured at baseline and at 6, 12, and 24 months. eGFR slope was calculated using linear regression. The box-and-whiskers plot above the histogram represents the slope of eGFR quartiles and the median (box) and 2.5 and 97.5 percentiles (whiskers). A fast eGFR decline was defined as the lowest quartile of eGFR slope (more rapid than −4.1 ml/min per 1.73 m2 per yr).
Fast eGFR Decline and Frailty Incidence
In the primary analysis, frailty occurred in 95 of 833 (11%) participants between 24 and 60 months. Among them, 31/207 (15%) had fast eGFR decline between baseline and 24 months, and 64/626 (10%) did not (log-rank P=0.05; see Figure 2 for Kaplan–Meier graph). A fast eGFR decline was significantly associated with a higher hazard of incident frailty in Cox adjusted model (hazard ratio [HR], 1.67; 95% CI, 1.03 to 2.71; P=0.04; Table 2).
Figure 2.
A fast loss of eGFR is associated with incident frailty. Kaplan–Meier curves of the cumulative probability of frailty according to fast eGFR decline in the MAPT study. A fast eGFR decline was defined as the lowest quartile of eGFR slope (more rapid than −4.1 ml/min per 1.73 m2 per yr) over the first 2 years of the MAPT study. Participants with a fast eGFR decline (solid line) were compared with participants without (dashed line). In nonfrail participants both at baseline and at 2 years, frailty incidence was assessed over the three subsequent years. According to Fried phenotype model (19), frailty was defined by the presence of at least three of the following items: weight loss, exhaustion, slow gait, low grip strength, and low physical activity.
Table 2.
Associations of fast eGFR decline with subsequent frailty in the MAPT
| Definition of eGFR Fast Decline | N at Risk/N Events | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | ||
| Primary analysis | |||||
| More rapid than −4.1 ml/min per 1.73 m2 per yr | 833/95 | 1.52 (0.99 to 2.34) | 0.05 | 1.67 (1.03 to 2.71) | 0.04 |
| Sensitivity analyses | |||||
| 1: More rapid than −5 ml/min per 1.73 m2 per yr | 833/95 | 1.44 (0.9 to 2.3) | 0.12 | 1.46 (0.87 to 2.44) | 0.15 |
| 2: More rapid than −4.1 ml/min per 1.73 m2 per yr (slope calculated with three or four eGFR time points) | 749/79 | 1.35 (0.85 to 2.14) | 0.20 | 1.51 (0.91 to 2.49) | 0.11 |
| 3: More rapid than −6%/yr (slope expressed as a percentage of baseline eGFR) | 774/82 | 1.50 (0.96 to 2.33) | 0.07 | 1.66 (1.03 to 2.66) | 0.04 |
| 4: More rapid than −4.1 ml/min per 1.73 m2 per yr (baseline to 12 mo) | 919/112 | 1.07 (0.72 to 1.60) | 0.70 | 1.12 (0.73 to 1.73) | 0.60 |
eGFR was determined using the CKD–Epidemiology Collaboration equation. eGFR slope was calculated using eGFR time points between baseline and 24 mo, except in sensitivity analysis 4. Incident frailty was assessed between 24 and 60 mo, except in sensitivity analyses 4 (between 12 and 60 mo). Cox models were adjusted for age, sex, study intervention group, baseline eGFR, CRP, BMI, and history of diabetes, hypertension, ischemic heart disease, heart failure, and stroke. Comorbidities, BMI, and CRP were considered as time-varying covariates, using data measured concurrently with eGFR (i.e., between baseline and 24 mo in the primary analysis and each sensitivity analyses except number 4; and between baseline and 12 mo in sensitivity analysis 4).
This association was confirmed in several sensitivity analyses (Supplemental Table 1, Table 2); i.e., with adjustment on mean eGFR instead of baseline eGFR (sensitivity analysis 6: HR, 1.56; 95% CI, 1.01 to 2.41), when death was added as a confounder (sensitivity analysis 7: HR, 1.67; 95% CI, 1.03 to 2.71), and when eGFR decline was expressed as a percentage of baseline eGFR (sensitivity analysis 3: HR, 1.66; 95% CI, 1.03 to 2.66). In sensitivity analysis 1 (which used an alternative cutoff of −5 ml/min per 1.73 m2 per yr to define fast eGFR decline), sensitivity analysis 2 (which was restricted to participants with eGFR available at three or four time points), sensitivity analysis 8 (which had an interaction term between the intervention group and fast eGFR decline), and sensitivity analysis 9 (which had three groups based on eGFR slope quartiles), we found nonsignificant trends to associations between fast eGFR decline and frailty incidence. In sensitivity analysis 5 (which was restricted to participants in the placebo-alone group) and sensitivity analysis 4 (where eGFR slope was calculated using the first year data; three time points of data collection), fast eGFR decline was not significantly associated with frailty incidence. eGFR slope, as a continuous variable, was not associated with the outcome (sensitivity analysis 10).
Baseline eGFR and Frailty Incidence
A total of 1201 participants were nonfrail at baseline, had appropriate follow-up data regarding frailty, and available baseline eGFR data. Frailty occurred in 165 (14%) participants during the 5-year study. Baseline eGFR (both as a continuous and a binary variable) was not significantly associated with the frailty incidence in Cox-adjusted models (Table 3).
Table 3.
Associations of baseline eGFR with subsequent frailty in the MAPT
| Variable | Unadjusted | Adjusted | ||
|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | |
| Baseline eGFR <60 ml/min per 1.73 m2a | 1.45 (1.14 to 1.85) | 0.002 | 1.19 (0.84 to 1.68) | 0.33 |
| Baseline eGFR (per 10-U decrease) | 1.25 (1.13 to 1.38) | <0.001 | 1.10 (0.98 to 1.23) | 0.11 |
eGFR was determined using the CKD–Epidemiology Collaboration equation. Cox models were adjusted for age, sex, study intervention group, BMI, CRP, and history of diabetes, hypertension, heart failure, stroke, and ischemic heart disease (baseline data only).
Baseline eGFR ≥60 ml/min per 1.73 m2 was considered as reference.
Discussion
In a population of community-dwelling older persons, we showed that a fast eGFR decline (more rapid than −4.1 ml/min per 1.73 m2 per yr) was independently associated with incident frailty, whereas baseline eGFR was not. Consistent results were found when eGFR decline was expressed as a percentage of baseline eGFR.
There is no consensus definition of a fast eGFR decline, which we therefore chose to define by the lowest quartile. This value (−4.1 ml/min per 1.73 m2 per yr) matched the definition used in the Three-City study, in which an eGFR decline more rapid than −4 ml/min per 1.73 m2 per yr was associated with cognitive decline (12). In previous studies with different older populations, the lowest quartile of eGFR slope was slightly different (−3 ml/min per 1.73 m2 per yr) and a fast eGFR decline was associated with mortality (10) and cardiovascular morbidity (11). Despite this lack of consensus, international guidelines suggested −5 ml/min per 1.73 m2 per yr as an alternative cutoff (23), regardless of age. In a sensitivity analysis using this definition, a fast eGFR decline tended to be associated with incident frailty, although this was not statistically significant.
The preferred statistical method to describe eGFR slope has also been discussed (22). The difference between eGFR at two time points were used by others (12,13), but this may misclassify people with creatinine fluctuations as people with fast eGFR decline. Indeed, international guidelines have emphasized that the confidence in assessing kidney function decline increases with the number of measurements and the duration of follow-up (23). We performed a sensitivity analysis after exclusion of participants with only two available eGFR time points. In this analysis, a fast eGFR decline tended to be associated with incident frailty. The loss of significance could confirm the risk of bias in our linear regression approach, or could be explained by a lack of power in a small population.
The association between a fast loss of eGFR and incident frailty may be mediated by several mechanisms. Firstly, even if our adjusted model ruled out cardiovascular diseases as confounding factors, vascular damage (especially in the brain) is associated with both eGFR decline (26) and frailty (27) and could mediate their association. Secondly, inflammation is associated with physical and psychologic components of the frailty phenotype (28,29), and with a fast eGFR decline (30,31). Even if the association we found here is independent of CRP, it could still be mediated by chronic low-grade inflammation, the pathways of which are not accounted for by CRP alone. Finally, metabolic changes such as vitamin-D deficiency and metabolic acidosis, frequent complications of CKD, have been associated with frailty (32) or functional loss (33), but the MAPT data set was not appropriate for studying those parameters.
Interestingly, we showed that a fast eGFR loss was associated with frailty incidence, despite relatively preserved eGFR in our population. Indeed, frailty has mostly been studied in patients with advanced CKD (34,35). Those patients frequently suffer from cachexia, a syndrome that shares features with frailty such as weight loss, fatigue, decreased muscle strength, and low-grade inflammation (36). Thus, European guidelines state that frailty assessment should be included in the management of older patients with eGFR <45 ml/min per 1.73 m2 to individualize care (37). Our results suggest that people with more preserved kidney function, but exhibiting a hallmark of a pathologic process such as fast eGFR decline, are also at risk for frailty incidence.
Besides its originality, because this is the first study looking at the association between change in eGFR over time and the incidence of frailty, this study has several strengths. First, our outcome was measured using the original criteria of the frailty phenotype (19), a model that has been widely validated in different clinical contexts (1), including in the nephrology field (34). Furthermore, the multiple eGFR time points available for slope calculation probably limited overestimation of eGFR variation over time due to the random fluctuations of serum creatinine which are often observed in clinical practice (23). The linear regression approach that we used remains a simple way to describe eGFR evolution, which can be used in clinical practice. Finally, results of several sensitivity analyses using different methods or definitions of a fast eGFR decline were consistent with our primary analysis.
This study also has limitations. First, this is a secondary analysis of a randomized controlled trial designed to slow cognitive decline. Interventions that were tested in this trial (i.e., a multidomain lifestyle intervention containing nutritional and physical activity advice and an omega-3 supplementation) may have influenced frailty incidence (25). However, the association we found between a fast eGFR decline and incident frailty was independent of the intervention study group. The fact that the sensitivity analysis performed in the placebo-alone group found no association was probably related with a reduced statistical power (n=211 participants and 28 cases of incident frailty). Secondly, we estimated GFR using serum levels of creatinine, a substance produced by muscles. Therefore, variations of serum creatinine may account for variations of kidney function or muscle mass (possibly influenced by the multidomain intervention). This may explain the absence of association between baseline eGFR and incident frailty in our study. Indeed, previous studies using creatinine and cystatin C (a kidney function marker that is independent of muscle mass [38]) reported an association between frailty and eGFR only when estimated using cystatin C (7,8). Our hypothesis should thus be tested in a population with repeated cystatin-C measurements. Finally, by design and compared with other cohorts (8), our older population was rather healthy, with few comorbidities, preserved physical and cognitive functions, and few mortality events during follow-up. In particular, participants had a relatively preserved eGFR, so our results cannot be directly translated to patients with more advanced CKD.
Our findings have important clinical implications for older persons and their physicians. Firstly, the fact that eGFR slope, but not eGFR alone, is associated with incident frailty emphasizes the principle that eGFR must be analyzed as a dynamic parameter, as already suggested for other health outcomes (10–13). Secondly, identification of a fast loss of eGFR should motivate the screening for frailty, in addition to measures to preserve kidney function, regardless of the stage of CKD.
To conclude, a fast loss of eGFR is associated with incident frailty in community-dwelling older adults, whereas baseline eGFR is not. A fast eGFR decline should lead to screening for frailty. Our results need to be confirmed using GFR estimated from a marker independent of muscle mass (e.g., cystatin C) and in patients with more advanced CKD.
Disclosures
Dr. Bourdel-Marchasson, Dr. de Souto Barreto, Dr. Guerville, Dr. Rolland, Dr. Taton, and Dr. Vellas have nothing to disclose.
Funding
The MAPT study was supported by grants from the French Ministry of Health (PHRC 2008, 2009), University Hospital Center of Toulouse/Gérontopôle, Pierre Fabre Research Institute (manufacturer of the omega-3 supplement), Exhonit Therapeutics (biologic sample collection), and Avid Radiopharmaceuticals (amyloid positron emission tomography measurement). The promotion of this study was supported by the University Hospital Center of Toulouse. The data sharing activity was supported by the Association Monegasque pour la Recherche sur la maladie d’Alzheimer (AMPA) and the UMR 1027 Unit INSERM–University of Toulouse III.
Supplementary Material
Acknowledgments
We would like to thank the investigators from Toulouse University Hospital (CHU Toulouse), Lyon-Sud Hospital Center, Tarbes Hospital, Foix Hospital, Castres Hospital, CHU Limoges, CHU Bordeaux, Lavaur Hospital, CHU Montpellier, Princess Grace Hospital, Montauban Hospital, CHU Nice, and CHU Dijon for their participation in this study.
No sponsor placed any restriction on this work or had any role in the design of the study, data collection, data analyses or interpretation, or in the preparation, review, or approval of the manuscript.
The MAPT Study Group:
Principal investigator: Dr. Vellas (Toulouse). Coordination: Sophie Guyonnet. Project leader: Isabelle Carrié. Clinical Research Assistant: Lauréane Brigitte. Investigators: Catherine Faisant, Françoise Lala, Julien Delrieu, and Hélène Villars. Psychologists: Emeline Combrouze, Carole Badufle, Audrey Zueras. Methodology, statistical analysis and data management: Sandrine Andrieu, Christelle Cantet, and Christophe Morin. Multidomain group: Gabor Abellan Van Kan, Charlotte Dupuy, and Dr. Rolland (physical and nutritional components); Céline Caillaud and Pierre-Jean Ousset (cognitive component); and Françoise Lala (preventive consultation). The cognitive component was designed in collaboration with Sherry Willis from the University of Seattle, and Sylvie Belleville, Brigitte Gilbert, and Francine Fontaine from the University of Montreal.
Coinvestigators in associated centers: Jean-François Dartigues, Isabelle Marcet, Fleur Delva, Alexandra Foubert, and Sandrine Cerda (Bordeaux); Marie-Noëlle-Cuffi and Corinne Costes (Castres); Olivier Rouaud, Patrick Manckoundia, Valérie Quipourt, Sophie Marilier, and Evelyne Franon (Dijon); Lawrence Bories, Marie-Laure Pader, Marie-France Basset, Bruno Lapoujade, Valérie Faure, Michael Li Yung Tong, Christine Malick-Loiseau, and Evelyne Cazaban-Campistron (Foix); Françoise Desclaux and Colette Blatge (Lavaur); Thierry Dantoine, Cécile Laubarie-Mouret, Isabelle Saulnier, Jean-Pierre Clément, Marie-Agnès Picat, Laurence Bernard-Bourzeix, Stéphanie Willebois, Iléana Désormais, and Noëlle Cardinaud (Limoges); Marc Bonnefoy, Pierre Livet, Pascale Rebaudet, Claire Gédéon, Catherine Burdet, and Flavien Terracol (Lyon); Alain Pesce, Stéphanie Roth, Sylvie Chaillou, and Sandrine Louchart (Monaco); Kristelle Sudres, Nicolas Lebrun, and Nadège Barro-Belaygues (Montauban); Jacques Touchon, Karim Bennys, Audrey Gabelle, Aurélia Romano, Lynda Touati, Cécilia Marelli, and Cécile Pays (Montpellier); Philippe Robert, Franck Le Duff, Claire Gervais, and Sébastien Gonfrier (Nice); Yannick Gasnier, Serge Bordes, Danièle Begorre, Christian Carpuat, Khaled Khales, Jean-François Lefebvre, Samira Misbah El Idrissi, Pierre Skolil, and Jean-Pierre Salles (Tarbes).
Magnetic resonance imaging group: Carole Dufouil (Bordeaux); Stéphane Lehéricy, Marie Chupin, Jean-François Mangin, and Ali Bouhayia (Paris); Michèle Allard (Bordeaux); Frédéric Ricolfi (Dijon); Dominique Dubois (Foix); Marie Paule Bonceour Martel (Limoges); François Cotton (Lyon); Alain Bonafé (Montpellier); Stéphane Chanalet (Nice); Françoise Hugon (Tarbes); Fabrice Bonneville, Christophe Cognard, and François Chollet (Toulouse).
Positrom emission tomography scans group: Pierre Payoux, Thierry Voisin, Julien Delrieu, Sophie Peiffer, and Anne Hitzel (Toulouse); Michèle Allard (Bordeaux); Michel Zanca (Montpellier); Jacques Monteil (Limoges); Jacques Darcourt (Nice).
Medicoeconomics group: Laurent Molinier, Hélène Derumeaux, and Nadège Costa (Toulouse).
Biologic sample collection group: Bertrand Perret, Claire Vinel, and Sylvie Caspar-Bauguil (Toulouse).
Safety management: Pascale Olivier-Abbal.
Data Sharing Alzheimer Group: Sandrine Andrieu, Christelle Cantet, and Nicola Coley.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
See related editorial, “Frailty and CKD: Chicken or the Egg?” on pages 1554–1556.
Supplemental Material
This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.03750319/-/DCSupplemental.
Supplemental Table 1. Associations of fast eGFR decline with subsequent frailty in the Multidomain Alzheimer Preventive Trial (sensitivity analyses #5–#10).
References
- 1.Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K: Frailty in elderly people [published correction appears in Lancet 382: 1328, 2013]. Lancet 381: 752–762, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Partridge L, Deelen J, Slagboom PE: Facing up to the global challenges of ageing. Nature 561: 45–56, 2018 [DOI] [PubMed] [Google Scholar]
- 3.Matsushita K, van der Velde M, Astor BC, Woodward M, Levey AS, de Jong PE, Coresh J, Gansevoort RT; Chronic Kidney Disease Prognosis Consortium : Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: A collaborative meta-analysis. Lancet 375: 2073–2081, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fried LF, Lee JS, Shlipak M, Chertow GM, Green C, Ding J, Harris T, Newman AB: Chronic kidney disease and functional limitation in older people: Health, aging and body composition study. J Am Geriatr Soc 54: 750–756, 2006 [DOI] [PubMed] [Google Scholar]
- 5.Liu CK, Lyass A, Massaro JM, D’Agostino RB Sr., Fox CS, Murabito JM: Chronic kidney disease defined by cystatin C predicts mobility disability and changes in gait speed: The Framingham Offspring Study. J Gerontol A Biol Sci Med Sci 69: 301–307, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Reese PP, Cappola AR, Shults J, Townsend RR, Gadegbeku CA, Anderson C, Baker JF, Carlow D, Sulik MJ, Lo JC, Go AS, Ky B, Mariani L, Feldman HI, Leonard MB; CRIC Study Investigators : Physical performance and frailty in chronic kidney disease. Am J Nephrol 38: 307–315, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ballew SH, Chen Y, Daya NR, Godino JG, Windham BG, McAdams-DeMarco M, Coresh J, Selvin E, Grams ME: Frailty, kidney function, and polypharmacy: The Atherosclerosis Risk in Communities (ARIC) study. Am J Kidney Dis 69: 228–236, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dalrymple LS, Katz R, Rifkin DE, Siscovick D, Newman AB, Fried LF, Sarnak MJ, Odden MC, Shlipak MG: Kidney function and prevalent and incident frailty. Clin J Am Soc Nephrol 8: 2091–2099, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Roderick PJ, Atkins RJ, Smeeth L, Nitsch DM, Hubbard RB, Fletcher AE, Bulpitt CJ: Detecting chronic kidney disease in older people; what are the implications? [published correaction appears in Age Ageing 38: 638, 2009]. Age Ageing 37: 179–186, 2008 [DOI] [PubMed] [Google Scholar]
- 10.Rifkin DE, Shlipak MG, Katz R, Fried LF, Siscovick D, Chonchol M, Newman AB, Sarnak MJ: Rapid kidney function decline and mortality risk in older adults. Arch Intern Med 168: 2212–2218, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shlipak MG, Katz R, Kestenbaum B, Siscovick D, Fried L, Newman A, Rifkin D, Sarnak MJ: Rapid decline of kidney function increases cardiovascular risk in the elderly. J Am Soc Nephrol 20: 2625–2630, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Helmer C, Stengel B, Metzger M, Froissart M, Massy ZA, Tzourio C, Berr C, Dartigues JF: Chronic kidney disease, cognitive decline, and incident dementia: The 3C study. Neurology 77: 2043–2051, 2011 [DOI] [PubMed] [Google Scholar]
- 13.Chen YC, Weng SC, Liu JS, Chuang HL, Hsu CC, Tarng DC: Severe decline of estimated glomerular filtration rate associates with progressive cognitive deterioration in the elderly: A community-based cohort study. Sci Rep 7: 42690, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Andrieu S, Guyonnet S, Coley N, Cantet C, Bonnefoy M, Bordes S, Bories L, Cufi MN, Dantoine T, Dartigues JF, Desclaux F, Gabelle A, Gasnier Y, Pesce A, Sudres K, Touchon J, Robert P, Rouaud O, Legrand P, Payoux P, Caubere JP, Weiner M, Carrié I, Ousset PJ, Vellas B; MAPT Study Group : Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): A randomised, placebo-controlled trial. Lancet Neurol 16: 377–389, 2017 [DOI] [PubMed] [Google Scholar]
- 15.Vellas B, Carrie I, Gillette-Guyonnet S, Touchon J, Dantoine T, Dartigues JF, Cuffi MN, Bordes S, Gasnier Y, Robert P, Bories L, Rouaud O, Desclaux F, Sudres K, Bonnefoy M, Pesce A, Dufouil C, Lehericy S, Chupin M, Mangin JF, Payoux P, Adel D, Legrand P, Catheline D, Kanony C, Zaim M, Molinier L, Costa N, Delrieu J, Voisin T, Faisant C, Lala F, Nourhashémi F, Rolland Y, Van Kan GA, Dupuy C, Cantet C, Cestac P, Belleville S, Willis S, Cesari M, Weiner MW, Soto ME, Ousset PJ, Andrieu S: Mapt study: A multidomain approach for preventing Alzheimer’s disease: Design and baseline data. J Prev Alzheimers Dis 1: 13–22, 2014 [PMC free article] [PubMed] [Google Scholar]
- 16.Lawton MP: Scales to measure competence in everyday activities. Psychopharmacol Bull 24: 609–614, 1988 [PubMed] [Google Scholar]
- 17.Folstein MF, Folstein SE, McHugh PR: “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12: 189–198, 1975 [DOI] [PubMed] [Google Scholar]
- 18.Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW: Studies of illness in the aged. the index of adl: A standardized measure of biological and psychosocial function. JAMA 185: 914–919, 1963 [DOI] [PubMed] [Google Scholar]
- 19.Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G, McBurnie MA; Cardiovascular Health Study Collaborative Research Group : Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci 56: M146–M156, 2001 [DOI] [PubMed] [Google Scholar]
- 20.Radloff LS: The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas 1: 385–401, 1977 [Google Scholar]
- 21.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) : A new equation to estimate glomerular filtration rate [published correction appears in Ann Intern Med 155: 408, 2011]. Ann Intern Med 150: 604–612, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Leffondre K, Boucquemont J, Tripepi G, Stel VS, Heinze G, Dunkler D: Analysis of risk factors associated with renal function trajectory over time: A comparison of different statistical approaches. Nephrol Dial Transplant 30: 1237–1243, 2015 [DOI] [PubMed] [Google Scholar]
- 23.Andrassy KM: Comments on ‘KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease’. Kidney Int 84: 622–623, 2013 [DOI] [PubMed] [Google Scholar]
- 24.Feng Z, Lugtenberg M, Franse C, Fang X, Hu S, Jin C, Raat H: Risk factors and protective factors associated with incident or increase of frailty among community-dwelling older adults: A systematic review of longitudinal studies. PLoS One 12: e0178383, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.de Souto Barreto P, Rolland Y, Maltais M, Vellas B; MAPT Study Group : Associations of multidomain lifestyle intervention with frailty: Secondary analysis of a randomized controlled trial. Am J Med 131: 1382.e7–1382.e13, 2018 [DOI] [PubMed] [Google Scholar]
- 26.Bugnicourt JM, Godefroy O, Chillon JM, Choukroun G, Massy ZA: Cognitive disorders and dementia in CKD: The neglected kidney-brain axis. J Am Soc Nephrol 24: 353–363, 2013 [DOI] [PubMed] [Google Scholar]
- 27.Maltais M, de Souto Barreto P, Moon SY, Rolland Y, Vellas B; MAPT/DSA Study Group : Prospective association of white matter hyperintensity volume and frailty in older adults. Exp Gerontol 118: 51–54, 2019 [DOI] [PubMed] [Google Scholar]
- 28.Schaap LA, Pluijm SM, Deeg DJ, Visser M: Inflammatory markers and loss of muscle mass (sarcopenia) and strength. Am J Med 119: 526.e9–526.e17, 2006 [DOI] [PubMed] [Google Scholar]
- 29.Howren MB, Lamkin DM, Suls J: Associations of depression with C-reactive protein, IL-1, and IL-6: A meta-analysis. Psychosom Med 71: 171–186, 2009 [DOI] [PubMed] [Google Scholar]
- 30.Amdur RL, Feldman HI, Gupta J, Yang W, Kanetsky P, Shlipak M, Rahman M, Lash JP, Townsend RR, Ojo A, Roy-Chaudhury A, Go AS, Joffe M, He J, Balakrishnan VS, Kimmel PL, Kusek JW, Raj DS; CRIC Study Investigators : Inflammation and progression of CKD: The CRIC study. Clin J Am Soc Nephrol 11: 1546–1556, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Salimi S, Shardell MD, Seliger SL, Bandinelli S, Guralnik JM, Ferrucci L: Inflammation and trajectory of renal function in community-dwelling older adults. J Am Geriatr Soc 66: 804–811, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ju SY, Lee JY, Kim DH: Low 25-hydroxyvitamin D levels and the risk of frailty syndrome: A systematic review and dose-response meta-analysis. BMC Geriatr 18: 206, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Yenchek R, Ix JH, Rifkin DE, Shlipak MG, Sarnak MJ, Garcia M, Patel KV, Satterfield S, Harris TB, Newman AB, Fried LF; Health, Aging, and Body Composition Study : Association of serum bicarbonate with incident functional limitation in older adults. Clin J Am Soc Nephrol 9: 2111–2116, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Nixon AC, Bampouras TM, Pendleton N, Woywodt A, Mitra S, Dhaygude A: Frailty and chronic kidney disease: Current evidence and continuing uncertainties. Clin Kidney J 11: 236–245, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.van Loon IN, Goto NA, Boereboom FTJ, Bots ML, Verhaar MC, Hamaker ME: Frailty screening tools for elderly patients incident to dialysis. Clin J Am Soc Nephrol 12: 1480–1488, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Reid J, Noble HR, Adamson G, Davenport A, Farrington K, Fouque D, Kalantar-Zadeh K, Mallett J, McKeaveney C, Porter S, Seres DS, Shields J, Slee A, Witham MD, Maxwell AP: Establishing a clinical phenotype for cachexia in end stage kidney disease - study protocol. BMC Nephrol 19: 38, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Farrington K, Covic A, Aucella F, Clyne N, de Vos L, Findlay A, Fouque D, Grodzicki T, Iyasere O, Jager KJ, Joosten H, Macias JF, Mooney A, Nitsch D, Stryckers M, Taal M, Tattersall J, Van Asselt D, Van den Noortgate N, Nistor I, Van Biesen W; ERBP guideline development group : Clinical practice guideline on management of older patients with chronic kidney disease stage 3b or higher (eGFR <45 mL/min/1.73 m2). Nephrol Dial Transplant 31[suppl 2]: ii1–ii66, 2016 [DOI] [PubMed] [Google Scholar]
- 38.Vinge E, Lindergård B, Nilsson-Ehle P, Grubb A: Relationships among serum cystatin C, serum creatinine, lean tissue mass and glomerular filtration rate in healthy adults. Scand J Clin Lab Invest 59: 587–592, 1999 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.