Participants
This study used data from four waves collected in 2011 (wave 1), 2013 (wave 2), 2015 (wave 3) and 2018 (wave 4) of the CHARLS with wave 1 as the baseline. CHARLS is an ongoing longitudinal cohort study of a nationally representative sample of community-dwelling Chinese citizens 45 years of age or older and their spouses from 28 Chinese provinces. The protocol was approved by the Biomedical Ethics Committee of Peking University (approval number: IRB00001052-11,015), and all participants were required to sign informed consent. The detailed information about CHARLS had been published in previous literature(10).
The inclusion criteria for the present study were: 1) individuals aged at least 50 years old in CHARLS 2011; 2) having data regarding frailty status and CKD; 3) having covariates data. The cross-sectional analysis included 4231 participants. In the longitudinal analysis, we excluded 382 individuals with CKD in wave 1 and excluded 239 individuals with no CKD information in wave 2, wave 3 and wave4 which resulted in 3610 eligible individuals. The detailed flowchart of the sample selection process is shown in Figure 1.
Assessment of frailty
Frailty was measured by the FRAIL scale that included five elements: fatigue, resistance, ambulation, illnesses, and weight loss(11).
Fatigue: Fatigue was measured if participants answered: "I can't start" and "I feel like everything I do is effort" as "moderate amount of time; "3-4 days" or "most of the time" comes from the revised Center for Epidemiological Studies Depression (CES-D) Scale(12).
Resistance: When asked if you have trouble climbing several flights of stairs without taking a break? Participants who answered "I have difficulty but still can do it," "Yes, I have difficulty and need help," or "I can't do it," were deemed to have a declining resistance.
Ambulation: Participants met the criteria for ambulation if they self-reported that they did not walk continuously for 10 minutes or more during a typical week.
Illnesses: Participants reported whether they had been diagnosed with the following conditions: hypertension (subjects were diagnosed as hypertension when the systolic blood pressure was ≥140 mmHg or the diastolic pressure was ≥90 mmHg or self-reported hypertension), dyslipidemia, diabetes, cancer (excluding mild skin cancer), lung disease, liver disease, heart disease, stroke, stomach disease, depression, memory-related disease and arthritis/rheumatism. For this analysis, to investigate the role of the index independent of CKD, we removed CKD(9). Participants with ≥5 diseases were defined as having the disease.
Weight loss: Weight loss was defined as a self-reported weight loss of ≥5kg from the previous year or a BMI of ≤18.5kg/m2.
Participants who met at least three criteria were considered frail, those who met one or two criteria were considered prefrail, and those who didn't were defined as robust(13).
Assessment of CKD events
CKD was defined as an eGFR of < 60 mL/min/1.73m2, as calculated using the CKD epidemiology collaboration (CKD-EPI) equations, or self-reported CKD. The CKD-EPI equation was as follows(14):
eGFR (mL/min/1.73m2) =141min(SCr/κ, 1)αmax(SCr/κ, 1)-1.2090.993Age1.018[if female]
Where SCr is serum creatinine (mg/dL), κ is 0.7 for females and 0.9 for males, α is -0.329 for females and -0.411 for males, min indicates the minimum of SCr/κ or 1, and max indicates the maximum of SCr/κ or 1.
Covariates
Sociodemographic factors including age, sex, marital status (married and other, other includes separated, unmarried, divorced, and widowed), place of residence (rural, urban), level of education (elementary school and below, middle school or above), smoking (current smoking vs nonsmoking) and drinking (current drinking vs no drinking). Weight in kilograms (kg) divided by height in meters (m) is known as the body mass index (BMI). Underweight people (BMI<18.5 kg/m2), normal weight people (BMI 18.5-23.9 kg/m2) and overweight or obese people (BMI ≥24 kg/m2) were classified according to their BMI.
Statistical analysis
For continuous variables, data were presented as means with standard deviation (SD) and for categorical variables as percentages. First, baseline characteristics in the cross-sectional analytical samples were summarized based on sarcopenia status and compared between participants using the chi-squared test, analysis of variance, least-significant difference test as appropriate. Secondly, in the cross-sectional analysis, logistic regression analysis was performed to determine the relationships between CKD and frail status or frail components. Thirdly, we determined the incidence rates of CKD per 1000 person-years in the longitudinal analysis. In addition, we calculated the follow-up period as the period of time between the last interview and either the CKD diagnosis date. The hazard ratios (HRs) with 95% confidence intervals (CIs) using Cox proportional hazards models were used to determine the association between baseline frail status or frail components and incident CKD. All statistical analysis was performed retrospectively with SPSS 25.0. In all cases, P < 0.05 was considered statistically significant.