Association of physical activity levels and estimated glomerular filtration rate with all-cause mortality: a 10-year cohort study

The community-household-individual multistage sample method was used to select participants representative of the general population in Jiangsu Province. In the first stage, nine counties or districts were selected according to the geographic regions (Northern, Central and Southern Jiangsu Province). In the second stage, four townships were randomly selected from each county or district using random number tables. In the third stage, 50 households were randomly selected from each of the four residential communities or administrative villages, with sampling probabilities proportional to community size. Finally, one eligible individual (exclusion criteria for study participants at baseline, in accordance with the survey design protocol, were as the following: acute phase of myocardial infarction or stroke within 3 months; severe cognitive, hearing or physical impairment resulting in poor cooperation; presence of severe comorbidities with a life expectancy of less than 1 year; voluntary refusal to participate in the study) aged 35–70 years from each household was recruited for the study according to the Kish selection table. Finally, a total of 7200 participants were enrolled at baseline. Socio-demographic characteristics, medication history, medical history, lifestyle behaviors and baseline laboratory data were collected between January 2008 and December 2008. The participants were followed up annually from January 2009 to December 2019. A total of 268 participants were excluded due to missing outcome events or follow-up time; 253 were excluded for incomplete laboratory data; and 413 were excluded due to missing data on PA, age, education level, smoking status, alcohol consumption, height, weight, history of disease and blood pressure. Eventually, a total of 6266 participants were included in the final analysis. Participants were classified into six groups based on renal function (eGFR ≥60 or eGFR <60 ml/min/1.73 m²) and PA levels (low/moderate/high) [6,16]. The study flow chart is presented in Figure 1. Death cases were verified through the death surveillance system of the Chinese Center for Disease Control and Prevention.

All participants provided written informed consent prior to enrollment. The study protocol was approved by the Ethics Review Committee of Jiangsu Provincial Center for Disease Control and Prevention (approval number: SL2015-B004-01). No individual-level identifiable data, including personal details, images or videos, are presented in this manuscript.

Baseline information was collected via standardized questionnaires, including sociodemographic characteristics, lifestyle factors, medical history, and family history. Physical examination measurements comprised height, weight and blood pressure, while serum biochemical parameters were obtained through laboratory testing.

Height (without shoes) was measured to the nearest 0.1 cm using a stadiometer. Body weight was measured to the nearest 0.1 kg using an electronic scale with participants in light clothing and barefoot. BMI was calculated as weight in kilograms divided by height in meters squared. Participants were stratified into four groups based on Chinese national obesity guidelines [17]: underweight (BMI <18.5 kg/m²), normal weight (18.50 ≤ BMI <24.00 kg/m²), overweight (24.00 ≤ BMI <28.00 kg/m²) and obesity (BMI ≥28.00 kg/m²).

Blood pressure (mmHg) was measured in the left upper arm using an automated sphygmomanometer (Omron HEM-757; Omron, Kyoto, Japan). Each participant’s blood pressure was measured three-times at a 1 min interval, with the midpoint of the appropriate cuff at the same level as the heart. The final blood pressure was determined by averaging the three measurements. Hypertension was defined as a prior diagnosis of hypertension, current use of antihypertensive treatment, or an average measured systolic blood pressure of ≥140 mmHg and/or diastolic blood pressure of ≥90 mmHg.

Individuals were divided into three smoking groups: never smokers, former smokers (defined as those who had ceased smoking for more than 1 year), and current smokers (those who smoked tobacco products at any time in the past year, including those who had quit within the previous year) [18]. Regarding alcohol consumption, participants were considered drinkers if they had a history of drinking and consumed alcohol at least once a month in the past year [19].

At the baseline investigation, a 12 h fasting venous blood sample was obtained to measure creatinine, low density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), total cholesterol (TC), triglyceride (TG) and fasting blood glucose (FBG). The blood samples were centrifuged, aliquoted, and stored in a -80°C freezer or -180°C liquid nitrogen for subsequent testing within 1 month.

A certified laboratory used glucose oxidase or hexokinase methods to measure FBG within 12 h after sample collection. Diabetes was defined as a self-reported diagnosis diabetes, current use of hypoglycemic drugs, or FBG ≥7.0 mmol/l [20].

Creatinine, TG, TC, HDL-C and LDL-C were measured using an autoanalyzer (Abbott Laboratories, USA) in Jiangsu Province Center for Disease Control and Prevention, which is certificated by The China National Laboratory Accreditation (CNLA). Dyslipidemia was defined as current use of antidyslipidemia treatment, or TC ≥6.22 mmol/l, and/or TG ≥2.26 mmol/l and/or LDL-C ≥4.14 mmol/l and/or HDL-C ≤ 1.04 mmol/l [21].

The body surface area (BSA) effect was not considered while calculating the creatinine clearance rate (CrCl) using the Cockcroft-Gault (CG) equation [22]. A modified CG equation was used to estimate eGFR in this study [23]. Thus, the eGFR was calculated as follows: eGFR = (1.73/BSA) × [(140 age) × weight(kg)]/(72 × creatinine(mg/dl) × 0.85(for female). BSA was calculated using the formula: BSA = 0.007184 × weight^0.425 × height^0.725 [24]. Reduced renal function (abnormal renal function) or CKD was defined as eGFR <60 ml/min/1.73 m² [25]. Accordingly, participants were divided into two groups based on eGFR levels: eGFR ≥60 ml/min/1.73 m² and eGFR <60 ml/min/1.73 m² [25].

The International PA Questionnaire (IPAQ) was used to assess PA [26]. This scale covers four domains: household, transportation, recreational and occupational activity. Participants recalled all activities performed over the past week, including the number of active days, daily duration and activity intensity. Total PA, expressed as metabolic equivalent (MET)-min/week, was summed across the four domains. Weekly PA energy expenditure was calculated as: MET value of each activity × weekly frequency (d/w) × daily duration (min/d) [26]. Participants were further categorized into three PA level based on total MET-min/week: low (≤600 MET-min/week), moderate (600–3000 MET-min/week) and high (≥3000 MET-min/week) [6].

Trained professional interviewer collected annual follow-up data on all-cause mortality for each participant via telephone or face-to-face interviews. Information regarding deaths was ascertained through household inquiry, medical records and official death certificates.

For categorical variables, baseline characteristics of participants were presented as percentages and frequencies. The Chi-square test was used to compare differences between groups. Renal function and all-cause mortality risk were analyzed using the Cox proportional hazard model. The proportional hazards assumption for the Cox regression model was evaluated using Schoenfeld residual plots and global tests. The assumption was satisfied for all variables included in the model (global test p > 0.05). Meanwhile, the multiplicative interaction between renal function and PA were assessed in the Cox model to investigate whether PA levels modify the associations between renal function and all-cause mortality. When examining the association of the combination of PA level and renal function with all-cause mortality, the group with high PA and eGFR ≥60 ml/min/1.73 m² was used as the reference. Hazard ratio (HRs) and 95% CIs were reported. For all analyses, two nested models were constructed. Model 1 was adjusted for age, sex, region; Model 2 was further adjusted for BMI, alcohol consumption, smoking status, educational level, hypertension, dyslipidemia, diabetes, chronic respiratory disease, CVD, cancer and other covariates, while retaining adjustment for age, sex and region. The proportional hazards assumption was tested and satisfied for all variables in the Cox models. All statistical analyses were performed using SPSS 25.0. A 2-sided p-value < 0.05 was considered statistically significant.

Of the 7200 participants recruited from Jiangsu Province, 6266 eligible participants were included in our analysis, of whom 1034(16.5%) had an eGFR <60 ml/min/1.73 m². A total of 2756 (44.0%) participants were male, and 3095 (49.4%) participants were from urban areas. Over a median follow-up of 9.1 years (IQR: 7.7–9.9), 196 (3.1%) deaths occurred. Table 1 presents the baseline characteristics of participants across groups. Age, sex, region, alcohol consumption, smoking status, education level, hypertension, cancer, chronic respiratory diseases, dyslipidemia, CVDs and BMI were all significantly different across the groups. Regardless of renal function level, the low PA group had the highest proportion of obese participants, whereas the high PA group had the lowest proportion of individuals with diabetes. In both the eGFR <60 ml/min/1.73 m² and the eGFR ≥60 ml/min/1.73 m² subgroups, the prevalence of dyslipidemia, hypertension and CVD was lowest in high PA group, followed by the moderate and low PA groups.

Table 2 shows that the HR for all-cause mortality (HR: 1.97; 95% CI: 1.44–2.70) was substantially higher in participants with eGFR <60 ml/min/1.73 m² than in those with eGFR ≥ 60 ml/min/1.73 m² in the models adjusted for age, sex and region. This association remained significant even after further adjustment for additional confounding variables. Individuals with moderate PA levels had a significantly higher risk of all-cause death (HR: 1.50; 95% CI: 1.07–2.10) compared with those with high PA levels.

Significant association was observed between renal function and PA in relation to all-cause mortality (p = 0.004) (Table 3). In the fully adjusted model, among participants with eGFR ≥60 ml/min/1.73 m², individuals with moderate PA presented a higher hazard of all-cause mortality (HR: 1.84; 95% CI: 1.20–2.81) compared with those in the high PA group. For participants with reduced renal function (eGFR <60 ml/min/1.73 m²), low PA was independently associated with an increased risk of all-cause mortality (HR: 1.88; 95% CI: 1.03–3.44), relative to the high PA reference group.

Figure 2 illustrated the all-cause mortality risks according to the combined categories of PA and renal function. In model 2, compared with participants with eGFR ≥60 ml/min/1.73 m² and high PA, those with eGFR <60 ml/min/1.73 m² and low PA exhibited the highest risk of all-cause mortality (HR: 4.06; 95% CI: 2.34–7.03). This was followed by individuals with eGFR <60 ml/min/1.73 m² and moderate PA (HR: 2.29; 95% CI: 1.36–3.86) and those with eGFR <60 ml/min/1.73 m² and high PA (HR: 2.06; 95% CI: 1.18–3.62). Conversely, among participants with eGFR ≥60 ml/min/1.73 m², low PA was not associated with a statistically significant all-cause mortality risk relative to high PA (HR: 0.86, 95% CI: 0.47–1.59).