Association between creatinine-to-cystatin C ratio and handgrip strength in prediabetes patients: a nationwide cross-sectional study in China
Article information
Abstract
Background
This study aimed to investigate the relationship between the creatinine-to-cystatin C ratio (CCR) and handgrip strength in individuals with prediabetes, to identify patients with reduced handgrip strength.
Methods
This study used a nationally representative sample from the Chinese middle-aged and elderly population. The cross-sectional portion utilized data from the first wave of the 2011 China Health and Retirement Longitudinal Study (CHARLS), while the longitudinal portion used data from the fourth wave of the 2015 CHARLS. Data on CCR, handgrip strength, and other relevant variables were collected and analyzed using univariate and multivariate regression.
Results
A total of 2,704 participants were included, with 1,276 males (47.2%) and 1,428 females (52.8%), and the mean age was 60.5 ± 9.5 years. Univariate analysis showed a positive correlation between CCR and handgrip strength (β = 18.3; 95% confidence interval [CI], 16.46–20.14; p < 0.001). After adjusting for confounding variables, the β value was 3.52 (95% CI, 1.95–5.09; p < 0.001). Compared to the lowest CCR group (Q1, 0.27 to 0.67), the adjusted β values for Q2 (0.67 to 0.77), Q3 (0.77 to 0.89), and Q4 (0.89 to 2.39) were 0.37 (95% CI, –0.46 to 1.2; p = 0.38), 1.6 (95% CI, 0.73–2.47; p < 0.001), and 2.16 (95% CI, 1.24–3.09; p < 0.001), respectively. Subgroup and stratified analyses further supported these results.
Conclusion
This study suggests that in individuals with prediabetes, there is a positive correlation between the CCR and handgrip strength.
Introduction
According to the European Working Group on Sarcopenia in Older People (EWGSOP2) criteria, sarcopenia is diagnosed based on low muscle strength and low muscle mass or quality [1]. Officially classified as a muscle disease with an ICD-10-CM (International Classification of Diseases, 10th Revision, Clinical Modification) diagnosis code in 2016 [2], sarcopenia commonly affects older adults as part of the aging process. It is linked to adverse outcomes such as falls, functional decline, frailty, and mortality. Studies have shown that individuals with prediabetes have lower average handgrip strength compared to those with normal glucose levels [3,4], highlighting the need for careful monitoring of sarcopenia in prediabetic patients. Clinically, sarcopenia diagnosis begins with assessing muscle strength, typically measured through handgrip tests using standardized protocols [5]. Currently, no blood biomarkers are available for routine clinical use.
Serum creatinine and cystatin C are widely used to estimate glomerular filtration rate (eGFR) and assess renal function. Creatinine, a byproduct of muscle metabolism, is primarily influenced by skeletal muscle mass, whereas cystatin C, a small protein produced by all nucleated cells at a constant rate, is less affected by muscle mass. The creatinine-to-cystatin C ratio (CCR), calculated by dividing creatinine (mg/dL) by cystatin C (mg/L), has emerged as a promising marker for muscle mass independent of renal function [6–9]. However, no studies to date have examined the relationship between CCR and muscle strength in individuals with prediabetes.
Handgrip strength, a measure of the force exerted when gripping an object, is a key indicator of physical and muscle strength, overall health, and functional capacity. The EWGSOP2 identifies low muscle strength as a primary criterion for diagnosing sarcopenia. Handgrip strength testing is widely used due to its simplicity and reliability [10]. A review by Lian et al. [11] shows that handgrip strength is the most widely used physical function test (cited in 331 articles) and measurement formula (cited in 294 articles). Research by Blanquet et al. [12] suggests that it is an effective tool for early sarcopenia screening and muscle wasting detection in older adults.
This study aims to validate the relationship between CCR and handgrip strength in middle-aged and older individuals with prediabetes using data from the China Health and Retirement Longitudinal Study (CHARLS). We hypothesize a positive correlation between CCR and handgrip strength in individuals aged 45 years and older at the prediabetic stage.
Methods
Study population
This cross-sectional study analyzed CHARLS data from June 2011 to March 2012, using a nationally representative sample of middle-aged and elderly individuals in China. The study design, sampling methods, and data collection procedures have been detailed in previous literature [10]. Approximately 10,000 households and 17,708 individuals from 28 of China’s 31 provinces (excluding Tibet, Ningxia, and Hainan) were included. Trained personnel collected data in participants’ homes and local health stations or Chinese Centers for Disease Control and Prevention (China CDC) offices, following standardized protocols. Information gathered included sociodemographic characteristics (age, sex, education, marital status), health status (doctor-diagnosed chronic diseases, biomarkers), and community-level data (geographical area, public facilities), obtained through face-to-face computer-assisted interviews conducted by CHARLS staff trained at Peking University. Individuals under 45 years old, those with diabetes, or those missing data on creatinine, cystatin C, handgrip strength, or covariates were excluded.
The CHARLS protocol received ethical approval from Peking University’s Ethics Review Committee (IRB00001052-11015), and informed consent was obtained from all participants. Prediabetes was defined as a fasting glucose level of 100–125 mg/dL or a glycated hemoglobin (HbA1C) level of 5.7%–6.4% [13].
Handgrip strength
Participants with no history of surgical procedures, diseases, injuries, or severe hand pain in the past 6 months were included in the handgrip strength evaluation, measured in kilograms (kg). Trained technicians assessed handgrip strength using a handheld dynamometer (YuejianTMWL-1000; Nantong Yuejian Physical Measurement Instrument Co., Ltd.). Participants stood with their arms relaxed and squeezed the dynamometer with maximum effort. Each hand was tested twice, and the highest value from four tests (two per hand) was used for analysis.
Creatinine-to-cystatin C ratio
Blood samples were collected in the morning after an overnight fast and analyzed according to the “2011–2012 National Baseline Blood Data Users’ Guide” (http://charls.pku.edu.cn/). Serum creatinine (mg/dL) was measured by the rate-blanked, compensated Jaffe method, while cystatin C (mg/L) was determined using a particle-enhanced turbidimetric assay. The CCR was calculated by dividing creatinine by cystatin C.
Covariates
According to the literature [14–18], some potential confounding factors related to CCR were evaluated, including age, sex, marital status, work status, urban or rural residence, smoking status, drinking status, body height, body weight, waist circumference and hip fracture, hypertension, kidney disease, stroke, heart problems, hemoglobin, urea nitrogen, and C-reactive protein. These factors are also known to be associated with handgrip strength [19–23]. Suetta et al. [24] reported that men have significantly higher handgrip strength than women in all age groups, and this strength gradually decreases with age in both sexes. In cross-sectional and longitudinal analyses, there was a positive correlation between hemoglobin concentration and handgrip strength [25]. Patients were divided into two groups based on sex: male and female. Places of residence included rural and urban areas. Married and partnered are defined as having a companion, while separated, divorced, widowed, and never married are defined as being without a companion. Current smokers were defined by smoking habits. Current alcohol drinkers were defined by whether the respondent had consumed an alcoholic beverage in the last 12 months. Height, weight, and waist circumference were measured. Comorbidities (hypertension, kidney disease, heart problems, and stroke) were determined based on inquiries in the questionnaire regarding whether the participants had been informed of these conditions by a doctor in the past.
Covariate data, including age, sex, residence, marital status, work status, smoking and drinking status, hip fracture, body measurements, and comorbidities, were sourced from the Regular Waves section of the Harmonized CHARLS dataset. Blood test data, including blood urea nitrogen, hemoglobin, and C-reactive protein levels, were obtained from the 2011 Blood Test Data section.
The national survey was conducted in collaboration with the China CDC, which oversaw blood sample collection, developed the primary questionnaire, and established on-site protocols. CHARLS enumerators handled the collection of non-blood biomarkers. The China CDC maintained a network of trained personnel in every county, including basic laboratory facilities, even in remote areas.
Categorical variables were presented as proportions (%), while continuous variables were described as means with standard deviations or medians with interquartile ranges, as appropriate.
Statistical analysis
Differences between groups were assessed using one-way analysis of variance for normally distributed variables, the Kruskal-Wallis test for skewed variables, and either the chi-square test or Fisher exact test for categorical variables. Linear regression models estimated regression coefficients (β) and 95% confidence intervals (95% CIs) for the relationship between CCR and handgrip strength. Model I adjusted for sociodemographic factors (age, sex, companion status, residence, and work status), while Model II further accounted for smoking, alcohol use, height, weight, and waist circumference. Model III incorporated additional adjustments for blood urea nitrogen, hemoglobin, C-reactive protein, hip fracture, and comorbidities (hypertension, kidney disease, heart conditions, and stroke). CCR was analyzed as both a continuous and categorical variable (quartiles) in the regression model. A smooth curve fitting was generated, adjusting for covariates in Model III.
For reliability, individuals with self-reported cancer or with eGFR ≤30 mL/min/1.73 m2 were excluded, and the CCR-handgrip correlation was reevaluated.
Furthermore, we conducted subgroup analyses based on sex (male vs. female), current smoker (no vs. yes), current alcohol (no vs. yes), and the presence of comorbidities, including hypertension (no vs. yes) and kidney disease (no vs. yes).
Several sensitivity analyses were conducted to evaluate the robustness of our findings. First, a longitudinal cohort study was performed using data from 2015. Second, participants with an eGFR ≤60 mL/min/1.73 m2 were excluded from the analysis to mitigate potential bias. Third, for individuals with missing covariate data, multiple imputation using five iterations was employed to address missingness and enhance data completeness.
Finally, acknowledging the inherent sex differences in handgrip strength, we conducted separate analyses for males and females. Given the potential age-related associations between CCR and handgrip strength, we stratified the participants based on the World Health Organization classification of older adults, dividing them into two age groups: <60 years (1,362 individuals) and ≥60 years (1,342 individuals), and performed subgroup analyses accordingly.
The sample size was based on available data, without prior power calculation. All analyses were conducted using R (R Foundation for Statistical Computing) and Free Statistics software (version 1.8). A two-tailed p-value of <0.05 was considered statistically significant.
Results
Study population selection
Due to missing data on age, questionnaires, and fasting blood samples, 8,530 individuals were excluded, leaving 9,178 participants, of whom 4,470 were diagnosed with prediabetes. Additionally, individuals with an eGFR ≤30 mL/min/1.73 m2 (n = 5), a history of cancer (n = 51), absent handgrip strength data (n = 709), missing creatinine and cystatin C measurements (n = 870), or incomplete covariate data (n = 131) were excluded from the analysis. This resulted in a final study population of 2,704 individuals. Fig. 1 outlines the selection and exclusion criteria.
Flow chart of participant selection.
This study includes approximately 10,000 households and 17,708 individuals from 28 out of 31 provinces in mainland China (excluding Tibet, Ningxia, and Hainan). Due to missing age information, age less than 45 years, missing questionnaires and fasting blood samples, a total of 8,530 individuals were excluded, leaving 9,178 individuals, among whom 4,470 were diagnosed with prediabetes. Participants with estimate glomerular filtration rate (eGFR) ≤30 mL/min/1.73 m2 (n = 5), a history of cancer (n = 51), missing handgrip strength data (n = 709), missing creatinine-to-cystatin C ratio blood samples (n = 870), and missing covariate data (n = 131) were excluded. As a result, the final study population for this cross-sectional study consisted of 2,704 individuals.
Study population characteristics
Detailed characteristics by CCR quartiles are shown in Table 1. The mean age of the 2,704 participants was 60.5 ± 9.5 years, with 52.8% female and 35.1% from urban areas. In the high CCR group, body weight, blood urea nitrogen, and hemoglobin levels were significantly higher (p < 0.05 for all). The key characteristics of both excluded and included participants are provided in Supplementary Table 1 (available online).
Baseline characteristics of participants
Creatinine-to-cystatin C ratio and handgrip strength relationship
CCR, as a continuous variable, was positively associated with handgrip strength (β = 18.3; 95% CI, 16.46–20.14; p < 0.001). There is a linear relationship between CCR and handgrip strength (p = 0.69) (Fig. 2). In adjusted models, CCR remained positively correlated with handgrip strength (Table 2). In Model III, each unit increase in CCR was associated with a 3.52 kg increase in handgrip strength (β = 3.52; 95% CI, 1.95–5.09; p < 0.001). Stratified by quartiles, Q2, Q3, and Q4 exhibited progressively stronger associations compared to Q1, with significant adjusted β values observed in Q3 and Q4 (p < 0.001), whereas Q2 did not show significance.
Curve fitting of CCR and handgrip strength in prediabetic population.
They were adjusted for age at visit, sex, place of residence, companion, work status, current smoker, current alcohol drinker, body height, body weight, waist circumference, blood urea nitrogen, hemoglobin, and C-reactive protein levels, hip fracture and comorbidities (including hypertension, kidney disease, heart problems, and stroke). Only 99% of the data is shown.
CCR, creatinine-to-cystatin C ratio.
Association between CCR and handgrip strength in prediabetic population in different models
Subgroup and sensitivity analyses
The stratified analysis indicated that the results were consistent across subgroups based on hypertension status (p = 0.97), smoking status (p = 0.06), and alcohol status (p = 0.26). However, after adjusting for multiple comparisons, the p-value for the interaction between sex and kidney disease, although less than 0.05, may not reach statistical significance. These findings should be interpreted with caution (Fig. 3).
Subgroup analyses of the association between CCR and handgrip strength in the prediabetic population.
A stratified analysis was conducted to assess the potential effect modification on the relationship between CCR and handgrip strength in several subgroups. The association between CCR and handgrip strength varied across subgroups as follows: sex (male vs. female, p for interaction = 0.004), current smoker (no vs. yes, p for interaction = 0.06), current alcohol drinker (no vs. yes, p for interaction = 0.26), kidney disease (no vs. yes, p for interaction = 0.02), and hypertension (no vs. yes, p for interaction = 0.97). Considering multiple testing, a p-value less than 0.05 for the interaction of sex and kidney disease may not have statistical significance.
CCR, creatinine-to-cystatin C ratio; CI, confidence interval.
The results of the sensitivity analysis are presented in the Supplementary Tables 2–4 (available online). In the 2015 longitudinal cohort study, 617 individuals with missing handgrip strength data were excluded, resulting in a final sample of 2,087 participants. The association between CCR and handgrip strength remained consistent, with each unit increase in CCR corresponding to a 4.11 kg increase in handgrip strength (β = 4.11; 95% CI, 2.35–5.87; p < 0.001), as outlined in Supplementary Table 2. Analysis of CCR stratified into quartiles revealed progressively stronger associations in Q2, Q3, and Q4 compared to Q1, with significant adjusted β values observed in Q3 and Q4 (p < 0.001), whereas no significant association was found in Q2. After excluding individuals with an eGFR <60 mL/min/1.73 m2, the relationship between CCR and handgrip strength persisted, with each unit increase in CCR leading to a 3.32 kg increase in handgrip strength (β = 3.32; 95% CI, 1.73–4.92; p < 0.001), as shown in Supplementary Table 3 (available online). Among the 4,470 participants, after excluding those with a history of cancer and eGFR ≤30 mL/min/1.73 m2, 1,579 individuals had missing data on both CCR and handgrip strength. Multiple imputation was performed for the missing covariates in the remaining 2,835 participants, yielding consistent results. Specifically, each unit increase in CCR was associated with a 3.63 kg increase in handgrip strength (β = 3.63; 95% CI, 2.09–5.16; p < 0.001), as presented in Supplementary Table 4 (available online).
Subsequent analyses conducted separately by age and sex further confirmed the robustness of the results. Detailed information can be found in the Supplementary Figs. 1, 2 (available online).
Discussion
In this large cross-sectional study of Chinese adults aged 45 years and older with prediabetes, a positive linear correlation was observed between the CCR and handgrip strength, even after adjusting for sociodemographic factors (age, sex, marital status, urban/rural residence, and work status), smoking status, drinking status, body height, body weight, waist circumference, hip fracture, hypertension, kidney disease, stroke, heart problems, hemoglobin, urea nitrogen, and C-reactive protein. A one-unit increase in CCR was associated with a 3.52 kg increase in handgrip strength (β = 3.52; 95% CI, 1.95–5.09, p < 0.001). Stratified and sensitivity analyses confirmed the robustness of these findings. In the longitudinal cohort studies, when CCR was treated as a continuous variable in the models, the effect size was 4.11, with corresponding confidence intervals of 2.35–5.87, which was statistically significant (p < 0.001). Both cross-sectional and cohort studies have demonstrated a linear relationship between CCR and handgrip strength. When CCR was categorized as a categorical variable in cross-sectional and cohort studies, although the difference between the Q2 and Q1 groups did not reach statistical significance, the direction of the effect sizes was consistent with the primary effect. This lack of significance may be attributed to the reduced sample size following categorization, which likely resulted in decreased statistical power. Nonetheless, the overall trend suggests a positive correlation.
Previous studies also support the relationship between CCR and handgrip strength. Tabara et al. [26] found CCR to be an independent marker of muscle quality, significantly correlated with handgrip strength in 1,329 elderly participants. Hirai et al. [27] observed a positive correlation between CCR and muscle strength in 234 chronic obstructive pulmonary disease (COPD) outpatients, while Tang et al. [28] found a similar correlation in patients with late-stage non-small-cell lung cancer (NSCLC). Studies by Lin et al. [9] and Yamada et al. [29] further reported positive associations between CCR and handgrip strength in patients with chronic kidney disease and memory clinic outpatients, respectively. Osaka et al. [30] identified a link between CCR and sarcopenia in type 2 diabetes patients. Qiu et al. [15] suggested that higher CCR may improve insulin levels and reduce inflammation, which is consistent with research indicating an inverse relationship between inflammatory markers (e.g., interleukin 6 and tumor necrosis factor alpha α) and handgrip strength [1,31]. Sarcopenia is commonly, but not exclusively, associated with aging and is observed across many species including humans. Early intervention in sarcopenia has the potential to prolong healthy life expectancy, reduce healthcare and medical costs, delay disease progression, and even improve the condition [32]. Studies have identified prediabetes as an independent risk factor for sarcopenia [33]. As a result, individuals with prediabetes are considered to be at high risk for developing sarcopenia. Screening this population using widely available, simplified, or existing indicators, followed by timely intervention, holds significant clinical value for disease prevention and early detection. However, research specifically on prediabetic populations is limited. The CHARLS dataset allowed for an in-depth assessment of the CCR-handgrip strength association, accounting for numerous confounding factors and stratified analyses.
In the first adjusted model of the multiple regression analysis, controlling for sociodemographic factors significantly reduced the effect size from 18.3 to 4.53. This is consistent with findings by He et al. [34], who reported that handgrip strength in China peaks in the 20s for men and 30s for women, and declines after age 50, with the lowest values observed in the 70–80 age group. This suggests that age and sex may have significantly influenced the effect size in our study. Consequently, we conducted separate analyses by sex and performed stratified analysis based on age groups, with the results remaining consistent and robust.
Contradictory findings have been reported by other studies. For example, He et al. [35] evaluated the sarcopenia index (serum CCR × 100) using four diagnostic criteria (EWGSOP, Asian Working Group for Sarcopenia, International Working Group on Sarcopenia, Foundation for the National Institutes of Health) in 371 older adults with normal kidney function and found that its area under the curve ranged from 0.505 to 0.618, indicating low reliability for detecting sarcopenia. This discrepancy may be due to differing muscle mass measurement methods, as this study used bioelectrical impedance, while prior studies relied on computed tomography scans. Handgrip strength, used in our study, offers advantages such as simplicity, reliability, and ease of implementation in clinical and community settings [36–38]. Many studies suggest a link between cerebrovascular disease and sarcopenia; our study found no significant association between stroke and handgrip strength. One potential explanation for this result is that respondents who had undergone surgery or experienced hand swelling, inflammation, severe pain, or injury within the past 6 months, or those who did not understand the instructions, expressed safety concerns, or attempted but were unable to complete the handgrip strength measurement, were assigned a handgrip strength score of 0. These data were excluded from the analysis, which likely led to the removal of stroke patients who may have experienced substantial impairments in handgrip strength. As a result, the remaining sample consisted of individuals with milder symptoms and less pronounced handgrip strength deficits, potentially accounting for the lack of a significant association. Another contributing factor could be the relatively small sample size, suggesting the need for further investigation in a larger cohort in the future.
This study is subject to several limitations. Although we utilized multiple regression, subgroup analysis, and sensitivity analysis to mitigate potential confounding biases, some potential confounders—such as medication use, nutritional supplementation, dietary patterns, bone density, skeletal muscle mass, and physical activity levels—may not have been fully addressed due to limitations in the available database variables and sample size. Nevertheless, a review of the existing literature [39,40] reveals that the majority of the variables considered in prior studies align with those incorporated in our analysis, thereby bolstering the robustness and reliability of our findings. Additionally, significant differences were observed between participants included and excluded from the analysis (p < 0.01), suggesting that non-random missing data may introduce bias into the results. However, sensitivity analyses, including longitudinal cohort studies and multiple imputation techniques, demonstrated that the association between CCR and handgrip strength remained stable and consistent. It is also important to acknowledge that this study specifically focuses on individuals aged 45 years and older with prediabetes in China, which may limit the generalizability of the findings to other populations. Finally, due to considerable missing data regarding the year of death in the CHARLS database, we were unable to investigate the relationship between CCR, handgrip strength, and long-term mortality.
In summary, in a prediabetic population comprised of individuals aged >45 years, a positive correlation was observed between CCR and handgrip strength. These findings provide a basis for further large-scale prospective studies in order to elucidate the precise causal mechanisms underlying this relationship.
Supplementary Materials
Supplementary data are available at Kidney Research and Clinical Practice online (https://doi.org/10.23876/j.krcp.24.259).
Notes
Conflicts of interest
All authors have no conflicts of interest to declare.
Data sharing statement
The data supporting the findings of this study are available on the CHARLS website (http://charls.pku.edu.cn/).
Authors’ contributions
Conceptualization, Supervision: KZ
Data curation: WZ
Formal analysis: WZ, XH
Methodology: WZ, KZ
Project administration: KZ
Writing–original draft: WZ
Writing–review & editing: WZ
All authors read and approved the final manuscript.