;on behalf of the KNOW-CKD InvestigatorsIntroduction
Protein-energy wasting (PEW) is an adverse change in nutrition and body composition and is defined as a state of concurrent losses in protein and energy stores. PEW is not only a state of malnutrition, but the result of multiple mechanisms inherent in chronic kidney disease (CKD), including nutritional deficiency, comorbidities, systemic inflammation, hormonal derangement, the dialysis procedure itself, and uremic toxicity [1]. It is a prevalent complication in CKD patients, especially in those on dialysis. One meta-analysis revealed that 28% to 54% of the dialysis population and 11% to 18% of non-dialysis-dependent CKD (NDD-CKD) patients suffer from PEW [2].
Many observational studies have shown that PEW is a strong indicator of adverse outcomes, such as all-cause mortality and cardiovascular (CV) events, in dialysis patients [3–5]. Most PEW–mortality association has been documented in patients with dialysis, and there are only two reports on PEW–mortality association in NDD-CKD patients [6,7]. But, the results of both studies are inconsistent, either because of the small sized study or because the subject group was not representative of NDD-CKD patients.
The KoreaN cohort study for Outcome in patients With Chronic Kidney Disease (KNOW-CKD; ClinicalTrials.gov identifier NCT01630486) is a large-scale, longitudinal, prospective cohort study that enrolled 2,238 patients with NDD-CKD (CKD stages 1–5). Using datasets from the KNOW-CKD, the current study aimed to evaluate the association of PEW with adverse outcomes of all-cause death or composite of any CV event and all-cause death in patients with NDD-CKD. In addition, we evaluated whether the risk of adverse outcomes increased in a parameter number-dependent manner as the number of PEW parameters increased in patients with NDD-CKD.
Methods
Study design and participants
The KNOW-CKD study is a large-scale, longitudinal, prospective cohort study of patients with NDD-CKD in Korea. Informed consent was obtained from all patients voluntarily at the time of enrollment. The study was approved by the Institutional Review Board of each participating hospital: Seoul National University Hospital (No. 1104–089-359), Seoul National University Bundang Hospital (No. B-1106/129-008), Severance Hospital (No. 4-2011-0163), Kangbuk Samsung Medical Center (No. 2011-01-076), The Catholic University of Korea, Seoul St. Mary’s Hospital (No. KC11OIMI0441), Gachon University Gil Medical Center (No. GIRBA2553), Eulji Medical Center (No. 201105-01), Chonnam National University Hospital (No. CNUH-2011-092), and Pusan Paik Hospital (No. 11-091). This study was conducted in accordance with the Declaration of Helsinki. Further details on the design and method of the KNOW-CKD study have been published elsewhere [8].
A total of 2,238 adult patients with NDD-CKD (CKD stages 1–5) aged 20 to 75 years from nine centers were enrolled from 2011 to 2016. Of them, 391 whose PEW status could not be determined were excluded, including 247 with missing 24-hour urine creatinine values, 55 with missing 24-hour urine urea values, and five with urine volume <500 mL/24 hr. For the accuracy of urine collection, patients with a 24-hour urine volume <500 mL were excluded. In addition, 10 patients with missing serum albumin values, three with missing body weight or height, and 71 with missing 24-hour urine protein concentration were excluded. Therefore, 1,847 patients were included in the final analysis (Fig. 1).
Measurement
Clinical data, including demographic information and comorbidities, blood pressure, medication, baseline laboratory results, and 24-hour urine collection values, were collected at study entry according to the protocol. Serum samples were collected at baseline according to the standardized protocol and sent to a central laboratory for measurement of creatinine and cystatin C. Hypertension was defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or treatment with antihypertensive medications. Diabetes was defined as fasting glucose ≥126 mg/dL or treatment with insulin or oral hypoglycemic medication. Comorbidity of CV disease (CVD) was defined as any history of coronary artery disease, peripheral artery disease, arrhythmia, cerebrovascular disease, or congestive heart failure. Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation [9]. Estimated skeletal muscle mass (SMM) was calculated using 24-hour urine creatinine excretion, as follows: SMM (kg) = 18.9 × measured 24-hour urine creatinine (mg/day) × 0.001 + 4.1 [10]. Dietary protein intake (DPI) was calculated using 24-hour urinary urea nitrogen, as follows: estimated daily protein intake (g/day) = 6.25 × [24-hour urinary urea nitrogen (g/day) + 0.031 (g/kg/day) × ideal body weight (kg)] [11].
Definition of protein-energy wasting
PEW was defined following the International Society of Renal Nutrition and Metabolism (ISRNM) criteria [12]: 1) serum albumin, <3.8 g/dL, 2) body mass index (BMI), <23.0 kg/m2, 3) sex-specific estimated SMM, <19.7 kg in females or <26.9 kg in males, and 4) low DPI, <0.6 g/kg/day. If at least three of these criteria were met, the subject was defined as having PEW. In a previous Asian PEW study [5], BMI ≤18.5 kg/m2 was used [5]. However, in the present study, only 37 patients met this criterion, and the average BMI was 24.6 ± 3.4 kg/m2, with an average BMI of 22.5 ± 2.4 kg/m2 in the PEW group. Therefore, BMI ≤23 kg/m2 was used according to the ISRNM criteria.
Follow-up and definitions of outcomes
Patients were followed every year for renal and CV events. Information on the time and specific cause of mortality was obtained either from hospital records or from the Korean Statistical Information Service through March 31, 2021.
CV events included fatal and non-fatal events. Non-fatal CV events were any first experience of acute myocardial infarction; unstable angina; ischemic or hemorrhagic cerebral stroke; congestive heart failure; symptomatic arrhythmia; aggravated valvular heart disease; pericardial disease; abdominal aortic aneurysm; or severe peripheral arterial disease that required hospitalization, intervention, or therapy during the follow-up period. A fatal CV event was defined as death from a CV event. Composite outcomes included any CV event and all-cause death.
Statistical analysis
Baseline characteristics were compared using analysis of variance, chi-square test, or the Kruskal-Wallis test (for non-normal distribution). For incidence analysis of all-cause death and composite outcomes, Poisson regression analysis was used. To analyze the effect of the number of PEW parameters on adverse outcomes, three models were created according to the type of confounding factor and were analyzed with the Cox proportional hazards model. The variables that showed a statistical difference were selected as the confounding factors. Each model was configured as follows: model 1 was adjusted for age and sex; model 2 was adjusted for variables in model 1 plus diabetes mellitus, hypertension, and CVD comorbidities, as described above; and model 3 was adjusted for variables in model 2 plus baseline eGFR, hemoglobin, high- and low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and total CO2.
Results
A total of 1,847 patients (60% male) were included in this study (Table 1), with a mean age of 54.2 ± 12.0 years. The distribution of CKD stages at the time of enrollment was as follows: stage 1, 16.4%; stage 2, 18.7%; stage 3a, 16.5%; stage 3b, 21.1%; stage 4, 21.2%; and stage 5, 6.0%. Patients were classified according to the number of PEW parameters, and the characteristics of patients according to the classification were analyzed. Eighty-six patients (4.6%) had three or more PEW parameters, 381 patients (20.6%) had two PEW parameters, and 606 patients (32.8%) had one PEW parameter. Those who had a larger number of PEW parameters tended to have lower eGFR, lower total CO2, and higher urine protein excretion at baseline. Frequency of comorbidities such as CVD and diabetes mellitus also tended to increase as the number of PEW parameters increased. BMIs were 26.3 ± 2.6 kg/m2, 23.9 ± 3.3 kg/m2, 22.6 ± 3.5 kg/m2, and 22.5 ± 2.4 kg/m2 (p < 0.001) for groups with 0, 1, 2, and ≥3 PEW parameters, respectively, while SMMs were 26.3 ± 2.6 kg, 23.9 ± 3.3 kg, 22.6 ± 3.5 kg, and 22.5 ± 2.4 kg (p < 0.001). Patients with three or more PEW parameters had 8.5% lower BMI and 32.2% lower SMM values compared to non-PEW patients. Low-density lipoprotein cholesterol level was not significantly different.
The median follow-up duration was 6.9 years (interquartile range, 5.0–8.2 years). During follow-up, 129 deaths from any cause and 264 composite outcomes occurred. The incidence rates of all-cause death and composite outcomes in the total population were 11.1 and 23.7 per 1,000 person-years, respectively (Table 2).
The incidence rate of all-cause death increased (6.2, 9.2, 20.5, and 33.7 per 1,000 person-years; p < 0.001) with the number of PEW parameters (0, 1, 2, or ≥3, respectively). Also, the incidence rate of composite outcomes showed a similar ascending trend (16.6, 21.9, 38.0, and 47.1 per 1,000 person-years; p < 0.001) with an increased number of PEW parameters, respectively (Table 2).
The risks of PEW parameters for all-cause death and composite outcomes were analyzed using the Cox proportional hazards model (Table 3, Fig. 2).
To investigate the impact of PEW on adverse outcomes, multivariate adjustment analysis with models 1, 2, and 3 was performed. In the unadjusted model, all-cause death showed a significant increase in patients with two PEW parameters (hazard ratio [HR], 3.36; 95% confidence interval [CI], 2.13–5.23; p < 0.001) and those with three or more PEW parameters (HR, 5.72; 95% CI, 3.09–10.61; p < 0.001). Composite outcomes also showed a significant increase in patients with two (HR, 2.29; 95% CI, 1.68–3.12, p < 0.001) and three or more PEW parameters (HR, 2.84; 95% CI, 1.74–4.63; p = 0.004). In the fully adjusted analysis using model 3, this trend was maintained among patients with two (HR, 2.78; 95% CI, 1.61–4.08; p < 0.001) and three or more PEW parameters (HR, 3.78; 95% CI, 1.81–7.89; p < 0.001). Composite outcomes also showed a significantly increased risk among patients with two (HR, 2.16; 95% CI, 1.51–3.11; p < 0.001) and three or more PEW parameters (HR, 2.30; 95% CI, 1.30–4.07; p = 0.004),
We additionally conducted a Cox regression analysis to explore the effect of each parameter of PEW on all-cause death and composite outcomes (Table 4). Among the four parameters, hypoalbuminemia and low SMM adversely affected the outcomes. On the other hand, DPI and BMI did not have a statistically significant effect on the outcomes.
For subgroup analysis, we divided NDD-CKD into two subgroups: early CKD (CKD stages 1–3a) and advanced CKD (CKD stages 3b–5) and then analyzed the effects of the PEW parameters on mortality in each subgroup using Cox regression analysis adjusted for the variables in model 3 (Fig. 3). In the advanced CKD subgroup, mortality was increased among patients with two PEW parameters (HR, 2.65; 95% CI, 1.37–5.11; p = 0.004) and three or more PEW parameters (HR, 2.96; 95% CI, 1.19–7.38; p = 0.02). Similarly, in the early CKD subgroup, mortality was increased in patients with two PEW parameters (HR, 4.29; 95% CI, 1.38–13.32; p = 0.01) and three or more PEW parameters (HR, 6.64; 95% CI, 1.45–30.45; p = 0.02).
Discussion
In this prospective CKD cohort study including patients with all stages of NDD-CKD, we investigated whether PEW was associated with all-cause death and composite outcomes among NDD-CKD patients. In this study, the prevalence of PEW in patients with NDD-CKD was 4.6%. As seen in patients on dialysis, PEW was a strong indicator of adverse outcomes like all-cause death and composite outcomes; it increased all-cause death (HR, 3.78) and composite outcomes (HR, 2.16) in our cohort of NDD-CKD patients. Moreover, as the number of PEW parameters increased, the incidence of all-cause death and composite outcomes showed a tendency to increase in a parameter number-dependent manner. A significantly increased risk of all-cause death and composite outcomes was observed among patients with two or more PEW parameters, which is below the threshold of diagnostic criteria for PEW. In our subgroup analysis, such an observation was consistent in both the early (CKD stages 1–3a) and advanced CKD subgroups (CKD stages 3b–5).
Most PEW–mortality association has been documented in patients with dialysis, and there are only two reports on PEW–mortality association in NDD-CKD patients [6,7]. One study by Franco et al. [6], which enrolled 137 patients with CKD stages 3–5, failed to reveal an association between PEW and mortality over both 5- and 10-year follow-up, and a study by Beddhu et al. [7], which analyzed 1,156 patients with CKD as a subgroup of a general population study, revealed that PEW increased mortality. However, that study had limitations in that 90% of the patients had CKD stage 3 disease, and the PEW prevalence rate was only 1.64% (19 patients) [7].
The prevalence of PEW in the present study was lower than that reported in previous studies, where the prevalence ranged from 11% to 18% in NDD-CKD patients and 28% to 54% in dialysis patients [2]. This is because, unlike previous studies that targeted patients with advanced CKD stages [13–16], the present study included both early and advanced CKD patients. Within the subgroup of advanced CKD patients, the prevalence rates of PEW were 7.5% and 10.5% in CKD stage 3–5 and stage 4–5 patients, respectively. Our observed prevalence of PEW was slightly lower than rates from previous studies in Europe and Latin America [13–16].
Recently, although PEW study enrollees have been expanded from dialysis patients to NDD-CKD patients, only a few PEW studies on NDD-CKD patients have been completed [13–17]. Therefore, prevalence comparisons between these studies may not be meaningful.
Comparing the KNOW-CKD study [18], the Chronic Kidney Disease-Japan Cohort (CKD-JAC) study [19], and the Chronic Renal Insufficiency Cohort (CRIC) study [20,21], which are large-scale prospective cohort studies of NDD-CKD patients, the respective mortality rates were 9.6, 7.2, and 31 per 1,000 person-years and were lower in Asian CKD patients compared to American CKD patients. After adjusting for age, sex, comorbidities, and laboratory markers, the respective adjusted incidence rates in the KNOW-CKD, CKD-JAC, and CRIC studies were 7, 9, and 43 per 1,000 person-years [22]. Considering that PEW is closely related to mortality, the lower PEW prevalence in Asians may be related to their lower mortality rate. Additional research is warranted to explain the low prevalence of PEW in the present study.
To meet the diagnostic criteria for PEW proposed by ISRNM, at least three of the four criteria must be met [12]. The four PEW parameters include biochemical indicators (low serum albumin, prealbumin, and cholesterol), body weight and fat (low BMI, weight loss), muscle mass (low mid-arm circumference and creatinine concentration), and low protein intake. This study revealed an increase in all-cause death and composite outcomes in the patient group with three or more PEW parameters, which would be classified as PEW according to the ISRNM criteria. However, significant increases in all-cause death and composite outcomes were also observed in the patient group with two PEW parameters. In addition, as the number of PEW parameters increased, the risk of both all-cause death and composite outcomes increased in a parameter number-dependent manner. A previous 3-year prospective study by Foucan et al. [4] that enrolled 216 Afro-Caribbean hemodialysis patients revealed increased mortality in patients with two PEW parameters (HR, 3.43; p = 0.021) and those with three or more PEW parameters (HR, 6.59; p = 0.001). These results are similar to those of our study.
Our study is meaningful since it revealed that the presence of two PEW parameters also significantly influenced disease outcomes, and the number of PEW parameters correlated with disease outcome parameter number-dependently. Therefore, clinicians should be cautious about CKD patients who do not meet the current PEW criteria but who are at risk for PEW, as well as patients with existing PEW.
CKD is unique in that organ wasting occurs before frank cachexia begins, and neither anorexia nor nutritional deficiency can account for these adverse changes in nutrition and body composition [1]. To explain the complex nature of this nutrition-related wasting syndrome, ISRNM defined this condition as PEW and proposed diagnostic criteria [12]. Many contributing factors are associated with PEW, such as uremic toxins, metabolic acidosis, nutritional deficiency, comorbidities, impairment of insulin, chronic inflammation, frailty, depression, and the dialysis procedure itself [7]. As renal function decreases, the deterioration of metabolic derangement induced by the orchestral effect of these factors is accompanied by increased protein catabolism, which depletes body protein stores and reduces muscle mass [3,23].
Patients with CKD show high rates of mortality and CV events [18,24]. In the general population, excess energy intake, hypercholesterolemia, and obesity are traditional risk factors for CV disease; in dialysis patients, various indices that indicate nutritional insufficiency, such as low BMI [25,26], low DPI, hypoalbuminemia, hypocholesterolemia [27,28], and low muscle mass, are more closely related to the increase in CV events and mortality [29]. This “reverse epidemiology” emphasizes the effect of PEW on the long-term effects of traditional CV risk factors [23].
Several observational studies have shown that PEW increases mortality and CV disease rates in CKD patients, particularly dialyzed patients. However, there is no clear explanation of the associated mechanisms. In CKD progression, uremia compromises the immune system [30] and PEW by virtue of uremia-exacerbated anorexia and malnutrition components and compromises the immune system and host resistance to infection, increasing vulnerability to various inflammatory diseases. In addition, chronic inflammation, closely related to PEW, promotes atherosclerotic plaque formation by promoting endothelial cell damage and endothelial dysfunction [23]. Hypoalbuminemia also triggers a hypercoagulable state and increases the frequency of CV events [31]. A decrease in muscle mass leads to a decrease in skeletal, respiratory, and cardiac muscle mass and impairs the associated functions, which reduces the muscle-based antioxidant defense [23,32].
In dialysis patients, most PEW parameters are associated with mortality. Lower BMI [25,33] and lower cholesterol level increased the mortality rate [28]. However, a protein intake level up to 1.4 g/kg/day increased survival [34]. Low serum albumin serves as the strongest predictor of mortality [35]. Regarding the effect of each parameter on mortality, patients with NDD-CKD exhibit slightly different patterns. In this study, only SMM and albumin were significantly associated with all-cause death and composite outcomes in NDD-CKD patients (Table 4). Previous studies involving NDD-CKD also trigger debate about the association between BMI and mortality [36–39]. Other studies have shown that serum albumin and SMM are strongly associated with mortality in NDD-CKD [23], but BMI and DPI failed to show consistent results when considering adverse outcomes [39]. The effect of each parameter of PEW on adverse outcomes was slightly different between dialysis and NDD-CKD patients. In our analysis of NDD-CKD patients, only serum albumin and SMM were significant factors for adverse outcomes.
Notably, both serum albumin and SMM showed stronger associations with all-cause mortality than with the composite outcome. Hypoalbuminemia and low muscle mass are closely linked to systemic conditions such as malnutrition, inflammation, impaired immune function, and physical frailty [1,12]. These underlying conditions may contribute to adverse outcomes beyond CV events, which may help explain their stronger association with all-cause mortality.
In our subgroup analysis, the increase in HRs with a greater number of PEW parameters appeared less pronounced in advanced CKD than in early CKD. This may be attributable to the higher burden of comorbidities and complications in advanced CKD, which could attenuate the relative impact of PEW. However, as the p-value for interaction was not statistically significant (p = 0.49), this difference should be interpreted with caution.
Our study showed that the detrimental effects of PEW on all-cause death and composite outcomes were also significant in NDD-CKD patients. In addition, a significant association with adverse outcomes was evident in the subgroup with two PEW parameters. This indicates that the adverse outcomes were induced by the orchestrated effect of multiple PEW parameters rather than a single parameter.
Our study has several limitations. First, as this is an observational study, there remains a possibility of a hidden confounder, even after multivariable adjustment. Second, approximately 17% of the initial cohort was excluded due to missing or inadequate urine data. Baseline characteristics were largely comparable between those included in the analysis and those excluded, and the observed differences were accounted for as covariates in the analysis. Thus, the impact on external validity and generalizability is expected to be limited. Third, DPI and SMM were estimated indirectly by 24-hour urine collection instead of using a diet diary or direct measurement of body composition. Although 24-hour urine collection is a practical method supported by several guidelines and studies [10,11,40], results may be influenced by the patient’s transient metabolic state, recent dietary intake, or sampling errors, potentially leading to under- or overestimation of DPI or SMM.
However, this study is a large prospective cohort of 1,847 patients with NDD-CKD (stages 1–5) with a median follow-up of 6.94 years, which could provide statistical power. The strength of this study is that it demonstrated that PEW also increases all-cause death and composite outcomes in NDD-CKD patients. Additionally, the increased risk of adverse outcomes was parameter number-dependently related to an increased number of PEW parameters. Such a risk was evident both in patients with two PEW parameters and in those with three or more PEW parameters. This study is one of the few to examine the effect of PEW on disease outcomes in patients with NDD-CKD. For these reasons, it provides valuable and important messages despite its limitations.
In conclusion, PEW is a strong indicator of adverse outcomes, such as all-cause death and composite outcomes, in NDD-CKD patients. As the number of PEW parameters increases, adverse outcomes parameter number-dependently increase. Significant increases in all-cause death and composite outcomes were observed in both the subgroup of patients with two PEW parameters and those who met traditional (≥3 parameters) PEW diagnostic criteria. More attention should be paid to the nutritional status of NDD-CKD patients. Further studies of nutritional interventions to improve the outcomes in NDD-CKD patients are warranted.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
