Associations between leisure time, non-leisure time physical activity, and kidney function in Japanese adults: a cross-sectional study

Abstract

Background

Chronic kidney disease (CKD) contributes to decreased life expectancy. We examined the association between leisure-time physical activity (LTPA), non-leisure-time physical activity (non-LTPA) and kidney function.

Methods

This was a cross-sectional study including 32 162 community-dwelling adults aged ≥ 20 years from the Tohoku Medical MegaBank community-based cohort study. Kidney function was evaluated using cystatin C-based estimated glomerular filtration rate (eGFR) as well as self-reported LTPA and non-LTPA. CKD was defined as either eGFR decline (≤ 60 mL/min/1.73 m²) or presence of albuminuria (albumin-creatinine ≥ 30 mg/g). The association between domain-specific physical activity and kidney function, and CKD prevalence was examined using multivariable-adjusted ordinary least squares and modified Poisson models.

Results

The mean eGFR was 98.1 (± 13.2) mL/min/1.73 m². 3 185 (9.9%) participants were classified as having CKD. The mean LTPA and non-LTPA levels were 2.9 (± 4.2) and 16.6 (± 14.2) METs-hour/day, respectively. For LTPA, in the adjusted model, the quartile groups with higher levels had a higher kidney function (β, 0.36; 95% confidence intervals [CI], [0.06, 0.66]; p = 0.019 for the 2nd quartile, β, 0.82; 95% CI, [0.51, 1.14]; p < 0.001 for the 3rd quartile, and β, 1.16; 95% CI, [0.83, 1.49]; p < 0.001 for the 4th quartile), whereas there were no apparent associations for prevalence of CKD. For non-LTPA, 4th quartile was associated with decreased eGFR (β, -0.42; 95% CI, [-0.72, -0.11]; p = 0.007) and higher prevalence of CKD prevalence (Prevalence ratio, 1.12; 95% CI, [1.02, 1.24]; p = 0.022). These associations with kidney function remained consistent in the subgroup analyses divided by demographic and biological variables.

Conclusions

We observed a positive association between higher LTPA levels and better kidney function, but not association with CKD prevalence. In contrast, higher non-LTPA was negatively associated with both kidney function and CKD prevalence. These findings suggest that promoting LTPA is beneficial for kidney function.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12882-024-03813-6.

Background

Chronic kidney disease (CKD) is a significant health issue and a major cause of decreased healthy life expectancy [1]. The condition causes adverse outcomes such as cardiovascular disease incidence, hospitalization, poor quality of life, and higher mortality [2–4]. The prevalence of CKD increases with age and has grown over the past decade in the general Japanese population [5]. Currently, no universally established definitive treatment for CKD exists; prevention of CKD is an urgent public health issue.

Greater physical activity is associated with not only reduced incidence of several risk factors for CKD, such as hypertension [6], cardiovascular disease [7], and diabetes [8], but also slower decline in kidney function [9–11] and reduced CKD incidence [11, 12]. The World Health Organization (WHO) recommended promoting regular physical activity, especially moderate-to-vigorous intensity of physical activity (MVPA), and replacing sedentary time with any intensity of physical activity to prevent adverse health outcomes, such as cardiovascular disease mortality, incident hypertension, and type-2 diabetes in adults [13]. Higher amount of physical activity is associated with higher kidney function [14, 15] and slower kidney function decline [16]. Therefore, promoting physical activity is considered a beneficial intervention to maintain kidney function and prevent CKD.

However, the association between domain-specific physical activity levels and kidney function remains unclear. While leisure-time physical activity (LTPA) has potential benefits for mortality and cardiorespiratory health, which are associated with kidney health, beneficial effects of non-leisure-time physical activity (non-LTPA), such as occupational physical activity, remain inconsistent [17–19]. LTPA generally includes walking, sports, and recreation in free time and is associated with reduced risk of cardiovascular disease and mortality [20–22]. LTPA generally includes MVPA and has short peaks of increased intensity (i.e., increased heart rate and blood pressure) with recovery. This activity pattern is expected to maintain or improve cardiovascular health. Higher occupational physical activity is associated with increased risk of major cardiovascular events and mortality, which is called the physical activity paradox [23, 24]. Occupational physical activity generally includes static and long-duration activities without an appropriate recovery time. This leads to sustained elevated heart rate and blood pressure, which ultimately adversely affects cardiovascular health [25]. Therefore, the association of kidney function can differ among physical activity domains; while LTPA is beneficial for kidney function, non-LTPA may not be. A UK biobank study revealed that engaging in LTPA is protective against incidences of CKD, while job-related activity does not demonstrate a protective relation [26]. This study emphasized the significance of LTPA in relation to CKD incidence, defined as an estimated glomerular filtration rate (eGFR) less than 60 mL/min/1.73 m² via medical record. However, the study did not provide sufficient evidence to support an association between domain-specific physical activity and kidney function itself. As kidney function generally declines from a younger age [27, 28], addressing the reduction in the slope of kidney function decline becomes a significant issue. Therefore, it is crucial to identify the association between domain-specific physical activity and kidney function (i.e., eGFR) to promote kidney health and CKD prevention. For example, promoting LTPA is recommended to maintain kidney function, even when individuals engage in non-LTPA.

The aim of this large-scale population-based cohort study was to identify whether LTPA and non-LTPA are associated with kidney function, assessed using a valid and reliable marker, in a general Japanese population.

Methods

Participants

This cross-sectional study included a general community-dwelling population aged ≥ 20 years living in Miyagi prefecture, Japan, who enrolled in a Tohoku Medical Megabank Project Community-Based Cohort study (TMM CommCohort Study), initiated in 2013 [29]. The survey was conducted between May 2013 and March 2016. The TMM CommCohort comprises three types of surveys. The Type 1 survey, which included basic data such as sociodemographic, blood, urine, questionnaire, and municipal health check-up data, was conducted along with the annual public community health examination provided by national health insurance. National health insurance covers people under the age 75 who are not enrolled in employees’ health insurance, such as self-employed or unemployed individuals. The Type 1 additional survey, conducted at different dates and locations from the annual public community health examination, included the same measurements of the Type 1 survey. The Type 2 survey, which additionally included detailed physical assessments such as muscle strength and body composition but not municipal health check-up data that was conducted at TMM community facilities. Participants in the type 2 survey voluntarily visited our facilities to undergo these assessments.

All the participants provided written informed consent to participate in this study. This study was approved by the Institutional Review Board of the Tohoku Medical Megabank Organization (approval number: 2021–4–179).

The inclusion criterion was individuals who participated in the Type 1 survey, except for those who withdrew to participate and lacked self-reported questionnaires (n = 37 303). We selected these inclusion criteria because there were differences in characteristics between survey types, due to the variations in the recruitment method. Exclusion criteria were individuals with cystatin C estimated glomerular filtration rate (eGFR) < 30 (n = 82) or ≥ 120 mL/min/1.73 m² (n = 2 614), blood urea nitrogen ≥ 21 mg/dL (n = 2 334), with self-reported Kremezin intake (n = 5), self-reported dementia (n = 63), or Parkinson’s disease (n = 43). Finally, 32 162 participants were included in the analysis (Fig. 1).

Fig. 1.

Flow diagram showing the inclusion and exclusion criteria of the participants

Kidney function assessment

Kidney function was measured using eGFR calculated by levels of blood cystatin C based on colloidal-gold aggregation method. Cystatin C is a more reliable marker of kidney function than blood creatinine because it is not influenced by muscle mass. Furthermore, eGFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration Cystatin C equation with race adjustment, which predicts the measured GFR fairly well in Japanese patients [30, 31]. CKD was defined as either eGFR decline (eGFR < 60 mL/min/1.73 m²) or presence of albuminuria (albumin-to-creatinine ratio > 30 mg/g), according to the KDIGO guidelines [32]. As a continuous variable, eGFR was the primary outcome, and CKD (binary variable) was the secondary outcome. Blood samples were collected from non-fasting participants.

Physical activity assessment

Physical activity was assessed using the Japan Public Health Center-based prospective study Physical Activity Questionnaire [33]. Non-LTPA was assessed through the number of hours spent standing, walking, and strenuous work in non-leisure time including commuting, and household chores on a typical day in the last year, whereas LTPA was assessed through the frequency and number of hours spent walking slowly, such as when taking a walk; walking quickly; engaging in light to moderate exercise, such as golf or gardening; and strenuous exercise, such as tennis, jogging, aerobics, or swimming, during leisure time. Metabolic equivalents (METs) were assigned accordingly: 2.0 METs for standing and walking, and 4.5 METs for strenuous work in non-LTPA. For LTPA, 3.0 METs was assigned for walking slowly, 4.0 METS for walking quickly and light to moderate exercise, and 4.5 METs for strenuous exercise [34]. The daily total amounts of LTPA and non-LTPA were calculated as the sum of the mean duration each activity per day multiplied by the intensity of each physical activity.

Covariates

Data on age; sex; self-reported educational level; job (with or without income); smoking and alcohol intake status (never, past, or current); body mass index (BMI); history of comorbidities (with or without diabetes, hypertension, hyperlipidemia, heart disease, and knee osteoarthritis); depressive symptoms (Center for Epidemiologic Studies Depression Scale [CES-D] ≥ 16 points) [35]; and medication status of antihypertensive agents, including angiotensin-converting enzyme inhibitor (ACEI) and Angiotensin receptor blocker (ARB), anti-diabetic agents, and statin, were collected. Furthermore, self-reported data on the status of housing damage during the Great East Japan Earthquake (GEJE) was collected. The severity of housing damage was divided into complete housing damage or less because complete housing damage is associated with several health outcomes [36].

Blood and laboratory test results included hemoglobin A1c, triglyceride, creatinine, uric acid, blood urea nitrogen, and casual urine sodium and potassium levels.

Statistical analysis

Participant characteristics were described as mean (standard deviation [SD]) for continuous variables and numbers (%) for categorical variables. Missing data of variables were analyzed using multiple imputation (MI) by a fully conditional specification (M = 100). We divided both LTPA and non-LTPA values into quartile groups. The cut-off values of the quartile range for each physical activity domain were obtained from the mean cut-off values of the MI datasets. One-way analysis of variance (ANOVA) for continuous variables and the chi-square test for categorical variables were employed for group comparisons. To examine the association of kidney function and each physical activity domain, an ordinary least square (OLS) model was created with robust standard error to calculate beta coefficients (β) and 95 percent confidence intervals (95%CI) for the eGFR as a continuous variable according to the LTPA and non-LTPA quartile groups. The prevalence ratios (PR) and 95%CI for CKD (eGFR decline or albuminuria), eGFR decline, and albuminuria were calculated according to the LTPA and non-LTPA quartile groups using a modified Poisson model. The adjusted models included age, sex, smoking and alcohol intake status (current, past, or never), BMI, educational status (university graduate, higher, or neither), job, depressive symptoms, status of housing damage due to the GEJE, self-reported comorbidity (diabetes, hypertension, hyperlipidemia, heart disease, and knee osteoarthritis), medication status (ACEI and ARB, anti-diabetic agents, and statin), laboratory tests (hemoglobin A1c, triglyceride, creatinine, uric acid, blood urea nitrogen, and casual urine sodium and potassium levels), daily sitting time, and survey season. In addition, we conducted the complete case analyses using the main regression models. Because approximately 10% of participants responded to the questionnaire more than 30 days after blood collection, we also conducted sensitivity analyses by limiting the participants to those who responded within 30 days. To check the robustness of the results, a sub-group analysis was performed, stratified by age (< or > 65 years), sex, depressive symptoms, comorbidity (with or without hypertension and diabetes), and eGFR (< or > 90 mL/min/1.73 m²). A multiplicative model was established to investigate combined association between the LTPA and non-LTPA quartile groups for kidney function using the adjusted OLS model. Statistical significance was defined as P < 0.05. All analyses were performed using the R software (version 4.2.0; R Core Team, Vienna, Australia).

Results

Tables 1 and 2 presents the characteristics of all study participants (n = 32 162), and each group was defined by LTPA and non-LTPA quartiles. In the overall participant population, the mean age was 60.9 (± 10.1) years, and 20 247 (63.0%) participants were women. The mean LTPA and non-LTPA values were 2.9 (± 4.2) and 16.6 (± 14.2) METs-hour/day, respectively. The mean eGFR value was 98.1 (± 13.2) mL/min/1.73 m², and 3 185 (9.9%) participants were classified as having CKD. Cut-off values for each LTPA and non-LTPA quartile group were 0.15, 7.25 for the 1st to 2nd, 1.30, 12.0 for 2nd to 3rd, and 3.82, 22.2 3rd to 4th quartile, respectively. The components of the LTPA and non-LTPA are listed in Additional file: Table S1. Moderate exercise accounted for the largest proportion of LTPA, and standing and strenuous work comprised the major proportion of non-LTPA. Several characteristics differed between the LTPA and non-LTPA quartile groups. In particular, mean age and higher educational level increased as LTPA quartiles increased but decreased as non-LTPA quartiles increased. In contrast, the proportion of individuals with jobs and income decreased as LTPA quartiles decreased but increased as non-LTPA quartiles increased.

Table 1.

Characteristics of the study participants divided into leisure physical activity quartile groups

Variables	Number missing (%)	Overall	LTPA
			Quartile 1	Quartile 2	Quartile 3	Quartile 4	P-value
		N = 32,162	N = 6,416	N = 6,864	N = 6,457	N = 6,296
Age, years	0 (0.0%)	60.9(10.1)	56.4(10.9)	58.2(11.0)	61.6(9.3)	63.8(8.1)	< 0.001
Women, n (%)	0 (0.0%)	20,247 (63.0%)	2,640 (41.1%)	2,243 (32.7%)	2,042 (31.6%)	2,321 (36.9%)	< 0.001
Education level	532 (1.7%)						< 0.001
Less than university graduate, n (%)		26,639 (84.2%)	5,443 (86.4%)	5,561 (82.0%)	5,174 (81.3%)	5,157 (83.2%)
University graduate or more, n (%)		4,991 (15.8%)	856 (13.6%)	1,221 (18.0%)	1,189 (18.7%)	1,045 (16.8%)
Job with income, n (%)	790 (2.5%)	9,298 (29.6%)	2,963 (47.6%)	2,054 (30.4%)	1,415 (22.3%)	1,225 (19.9%)	< 0.001
Complete housing damage by GEJE, n (%)	1,745 (5.4%)	2,442 (8%)	525 (8.7%)	532 (8.1%)	484 (7.8%)	460 (7.7%)	0.183
BMI, kg/m2	339 (1.1%)	23.4(3.4)	23.6(3.6)	23.4(3.5)	23.2(3.2)	23.2(3.1)	< 0.001
BMI (category)							< 0.001
< 18.5 kg/m2, n (%)		1,617 (5.1%)	349 (5.5%)	390 (5.7%)	351 (5.5%)	297 (4.7%)
≥ 18.5 and < 25 kg/m2, n (%)		21,154 (66.5%)	4,077 (64.3%)	4,455 (65.4%)	4,390 (68.4%)	4,333 (69.2%)
≥ 25 and < 30 kg/m2, n (%)		7,760 (24.4%)	1,572 (24.8%)	1,661 (24.4%)	1,466 (22.8%)	1,461 (23.3%)
≥ 30 kg/m2, n (%)		1,292 (4.1%)	342 (5.4%)	310 (4.5%)	211 (3.3%)	171 (2.7%)
Smoking	813 (2.5%)						< 0.001
Current, n (%)		4,541 (14.5%)	1,369 (21.8%)	1,044 (15.4%)	707 (11.1%)	625 (10.1%)
Past, n (%)		7,597 (24.2%)	1,489 (23.7%)	1,519 (22.5%)	1,502 (23.6%)	1,513 (24.4%)
Never, n (%)		19,211 (61.3%)	3,429 (54.5%)	4,202 (62.1%)	4,158 (65.3%)	4,068 (65.5%)
Alcohol status	295 (0.9%)						0.844
Current, n (%)		14,570 (45.7%)	2,884 (45.2%)	3,164 (46.2%)	2,984 (46.4%)	2,857 (45.5%)
Past, n (%)		817 (2.6%)	159 (2.5%)	181 (2.6%)	171 (2.7%)	161 (2.6%)
Never, n (%)		16,480 (51.7%)	3,341 (52.3%)	3,499 (51.1%)	3,279 (51%)	3,255 (51.9%)
CES-D ≥ 16 points, n (%)	1,846 (5.7%)	7,750 (25.6%)	1,884 (31.2%)	1,843 (27.9%)	1,419 (22.9%)	1,203 (19.9%)	< 0.001
Heart disease, n (%)	5,917 (18.4%)	1,259 (4.8%)	183 (3.6%)	250 (4.4%)	261 (4.9%)	270 (5.1%)	< 0.001
Hypertension, n (%)	2,737 (8.5%)	8,159 (27.7%)	1,316 (22.8%)	1,605 (25.3%)	1,742 (29.1%)	1,809 (30.7%)	< 0.001
Hyperlipidemia, n (%)	5,252 (16.3%)	3,842 (14.3%)	546 (10.4%)	806 (13.9%)	975 (17.6%)	899 (16.5%)	< 0.001
Diabetes, n (%)	5,203 (16.2%)	1,415 (5.2%)	188 (3.6%)	270 (4.6%)	299 (5.4%)	325 (6.0%)	< 0.001
Stroke, n (%)	5,875 (18.3%)	500 (1.9%)	80 (1.6%)	99 (1.7%)	99 (1.8%)	121 (2.3%)	0.089
Knee Osteoarthritis, n (%)	7,280 (22.6%)	1,240 (5.0%)	161 (3.3%)	252 (4.7%)	286 (5.6%)	305 (6.1%)	< 0.001
Cancer, n (%)	5,720 (17.8%)	2,271 (8.6%)	327 (6.3%)	439 (7.7%)	502 (9.3%)	534 (10.0%)	< 0.001
Anti-hypertensive agents	581 (1.8%)
ACEI, n (%)		633 (2%)	88 (1.4%)	114 (1.7%)	141 (2.2%)	146 (2.4%)	< 0.001
ARB, n (%)		4,244 (13.4%)	649 (10.3%)	853 (12.6%)	889 (14.0%)	959 (15.5%)	< 0.001
other agents, n (%)		5,929 (18.8%)	938 (14.9%)	1,191 (17.6%)	1,290 (20.3%)	1,330 (21.4%)	< 0.001
Anti-diabetic agents, n (%)	581 (1.8%)	1,520 (4.8%)	212 (3.4%)	282 (4.2%)	313 (4.9%)	345 (5.6%)	< 0.001
Statin, n (%)	581 (1.8%)	4,981 (15.8%)	691 (11.0%)	1,022 (15.1%)	1,169 (18.4%)	1,094 (17.6%)	< 0.001
TG, mg/dL	32 (0.1%)	117.5 (81.7)	122.2 (96.4)	117.1 (83.6)	114.2 (74.3)	112.2 (70.4)	< 0.001
UA, mg/dL	0 (0.0%)	5.1(1.3)	5.2(1.4)	5.1(1.3)	5.0(1.3)	5.1(1.3)	< 0.001
HbA1c, %	3 (0.0%)	5.6(0.6)	5.5(0.6)	5.6(0.6)	5.6(0.5)	5.6(0.5)	< 0.001
BUN, mg/dL	0 (0.0%)	14.3(3.1)	13.8(3.2)	13.8(3.1)	14.3(3.0)	14.7(2.9)	< 0.001
Serum creatinine, mg/dL	0 (0.0%)	0.7(0.2)	0.7(0.1)	0.7(0.1)	0.7(0.2)	0.7(0.2)	< 0.001
Urine sodium, mEq/L	665 (2.1%)	127.8(50.6)	132.7(51.8)	126.4(51.7)	125.0(49.5)	127.1(49.7)	< 0.001
Urine potassium, mEq/L	545 (1.7%)	3.9(2.5)	3.8(2.5)	3.9(2.5)	3.9(2.5)	4.0(2.5)	< 0.001
Urine ACR ≥ 30 mg/g, n (%)	130 (0.4%)	2,902 (9.1%)	530 (8.3%)	602 (8.8%)	530 (8.2%)	597 (9.5%)	0.036
eGFR-cys, mL/min/1.73m2	0 (0.0%)	98.1(13.2)	100.7(13.3)	99.9(13.1)	97.8(12.9)	96.6(12.6)	< 0.001
eGFR-cys < 60 mL/min/1.73m2, n (%)	0 (0.0%)	428 (1.3%)	78 (1.2%)	75 (1.1%)	99 (1.5%)	83 (1.3%)	< 0.001
LTPA, METs-hour/day	6,129 (19.1%)	2.9(4.2)	0.0(0.0)	0.6(0.4)	2.5(0.7)	8.6(5.0)	< 0.001
Non-LTPA, METs-hour/day	3,950 (12.3%)	16.6(14.2)	19.8(17.3)	14.6(13.0)	14.2(12.0)	17.5(12.8)	< 0.001
Sitting time, hour/day	2,804 (8.7%)	225.4(143.9)	216.7(156.6)	237.3(150.5)	235.2(138.8)	220.1(130.5)	0.438
Survey season	0 (0.0%)						< 0.001
Spring, n (%)		4,848 (15.1%)	1,099 (17.1%)	1,122 (16.3%)	938 (14.5%)	783 (12.4%)
Summer, n (%)		6,689 (20.8%)	1,334 (20.8%)	1,363 (19.9%)	1,299 (20.1%)	1,398 (22.2%)
Autumn, n (%)		20,459 (63.6%)	3,947 (61.5%)	4,339 (63.2%)	4,187 (64.8%)	4,089 (64.9%)
Winter, n (%)		166 (0.5%)	36 (0.6%)	40 (0.6%)	33 (0.5%)	26 (0.4%)

Data are presented as mean (SD) or n (%). One way ANOVA for continuous variables and chi-square test for categorical variable were performed

Cut-off values of quartile range for each physical activity domain are employed from mean cut-off values of multiply imputed datasets (M = 100)

Participants with missing values in each physical activity domain were not listed in each quartile group

BMI Body mass index, CES-D The Center for Epidemiologic Studies Depression scale, ACEI Angiotensin converting enzyme inhibitor, ARB Angiotensin receptor blocker, TG Triglyceride, UA Uric acid, BUN Blood urea nitrogen, eGFR-cys Cystatin C estimated glomerular filtration rate, ACR Albumin-creatinine ratio, LTPA Leisure time physical activity

Table 2.

Characteristics of the study participants divided non-leisure time physical activity quartile groups

Variables	Number missing (%)	Overall	Non-LTPA
			Quartile 1	Quratile2	Quartile 3	Quartile 4	P-value
		N = 32,162	N = 6,266	N = 7,291	N = 7,602	N = 7,053
Age, years	0 (0.0%)	60.9(10.1)	60.7(10.6)	61(10.1)	60.2(10.2)	59.5(10.1)	< 0.001
Women, n (%)	0 (0.0%)	20,247 (63.0%)	2,671 (42.6%)	2,433 (33.4%)	2,347 (30.9%)	2,944 (41.7%)	< 0.001
Education level	532 (1.7%)						< 0.001
Less than university graduate, n (%)		26,639 (84.2%)	4,892 (79.3%)	5,774 (80.1%)	6,254 (83.6%)	6,261 (90.0%)
University graduate or more, n (%)		4,991 (15.8%)	1,279 (20.7%)	1,433 (19.9%)	1,224 (16.4%)	695 (10.0%)
Job with income, n (%)	790 (2.5%)	9,298 (29.6%)	1,250 (20.4%)	1,605 (22.4%)	2,166 (29.1%)	3,259 (47.4%)	< 0.001
Complete housing damage by GEJE, n (%)	1,745 (5.4%)	2,442 (8%)	447 (7.5%)	547 (7.8%)	568 (7.9%)	560 (8.4%)	0.300
BMI, kg/m2	339 (1.1%)	23.4(3.4)	23.7(3.5)	23.3(3.3)	23.2(3.3)	23.4(3.4)	< 0.001
BMI (category)							< 0.001
< 18.5 kg/m2, n (%)		1,617 (5.1%)	303 (4.9%)	383 (5.3%)	417 (5.5%)	350 (5%)
≥ 18.5 and < 25 kg/m2, n (%)		21,154 (66.5%)	3,936 (63.4%)	4,933 (68.1%)	5,181 (68.7%)	4,639 (66.3%)
≥ 25 and < 30 kg/m2, n (%)		7,760 (24.4%)	1,672 (26.9%)	1,642 (22.7%)	1,697 (22.5%)	1,738 (24.8%)
≥ 30 kg/m2, n (%)		1,292 (4.1%)	300 (4.8%)	284 (3.9%)	250 (3.3%)	267 (3.8%)
Smoking	813 (2.5%)						< 0.001
Current, n (%)		4,541 (14.5%)	960 (15.6%)	839 (11.7%)	988 (13.2%)	1,270 (18.4%)
Past, n (%)		7,597 (24.2%)	1,679 (27.3%)	1,752 (24.5%)	1,629 (21.8%)	1,585 (23.0%)
Never, n (%)		19,211 (61.3%)	3,518 (57.1%)	4,567 (63.8%)	4,848 (64.9%)	4,030 (58.5%)
Alcohol status	295 (0.9%)						< 0.001
Current, n (%)		14,570 (45.7%)	2,795 (44.9%)	3,347 (46.1%)	3,524 (46.6%)	3,085 (44.0%)
Past, n (%)		817 (2.6%)	212 (3.4%)	180 (2.5%)	172 (2.3%)	159 (2.3%)
Never, n (%)		16,480 (51.7%)	3,224 (51.7%)	3,734 (51.4%)	3,873 (51.2%)	3,769 (53.7%)
CES-D ≥ 16 points, n (%)	1,846 (5.7%)	7,750 (25.6%)	1,658 (27.8%)	1,687 (24.1%)	1,704 (23.6%)	1,789 (26.8%)	< 0.001
Heart disease, n (%)	5,917 (18.4%)	1,259 (4.8%)	320 (6.4%)	289 (4.8%)	254 (4.0%)	224 (3.8%)	< 0.001
Hypertension, n (%)	2,737 (8.5%)	8,159 (27.7%)	1,738 (30.4%)	1,940 (28.8%)	1,816 (25.8%)	1,604 (24.8%)	< 0.001
Hyperlipidemia, n (%)	5,252 (16.3%)	3,842 (14.3%)	843 (16.2%)	1,062 (17.1%)	957 (14.8%)	608 (10.2%)	< 0.001
Diabetes, n (%)	5,203 (16.2%)	1,415 (5.2%)	354 (6.8%)	348 (5.6%)	279 (4.3%)	219 (3.7%)	< 0.001
Stroke, n (%)	5,875 (18.3%)	500 (1.9%)	139 (2.8%)	114 (1.9%)	84 (1.3%)	92 (1.6%)	< 0.001
Knee Osteoarthritis, n (%)	7,280 (22.6%)	1,240 (5.0%)	226 (4.7%)	304 (5.3%)	309 (5.2%)	233 (4.3%)	0.059
Cancer, n (%)	5,720 (17.8%)	2,271 (8.6%)	503 (9.9%)	585 (9.6%)	493 (7.8%)	383 (6.5%)	< 0.001
Anti-hypertensive agents	581 (1.8%)
ACEI, n (%)		633 (2%)	138 (2.2%)	151 (2.1%)	125 (1.7%)	125 (1.8%)	0.061
ARB, n (%)		4,244 (13.4%)	908 (14.7%)	1,050 (14.6%)	932 (12.4%)	819 (11.8%)	< 0.001
other agents, n (%)		5,929 (18.8%)	1,304 (21.1%)	1,413 (19.6%)	1,264 (16.9%)	1,164 (16.8%)	< 0.001
Anti-diabetic agents, n (%)	581 (1.8%)	1,520 (4.8%)	368 (6.0%)	363 (5.0%)	312 (4.2%)	245 (3.5%)	< 0.001
Statin, n (%)	581 (1.8%)	4,981 (15.8%)	1,069 (17.3%)	1,233 (17.1%)	1,161 (15.5%)	859 (12.4%)	< 0.001
TG, mg/dL	32 (0.1%)	117.5 (81.7)	126.4(90.4)	115.2(78.0)	112.0(75.4)	115.4(84.3)	< 0.001
UA, mg/dL	0 (0.0%)	5.1(1.3)	5.3(1.4)	5.1(1.3)	5(1.3)	5.1(1.3)	< 0.001
HbA1c, %	3 (0.0%)	5.6(0.6)	5.6(0.6)	5.6(0.5)	5.6(0.5)	5.6(0.6)	< 0.001
BUN, mg/dL	0 (0.0%)	14.3(3.1)	14(3.1)	14.2(3.1)	14.2(3.1)	14.5(3.1)	< 0.001
Serum creatinine, mg/dL	0 (0.0%)	0.7(0.2)	0.7(0.2)	0.7(0.2)	0.7(0.1)	0.7(0.1)	< 0.001
Urine sodium, mEq/L	665 (2.1%)	127.8(50.6)	125.5(50.3)	125.1(50.6)	126.3(50.5)	134.1(51.1)	< 0.001
Urine potassium, mEq/L	545 (1.7%)	3.9(2.5)	3.9(2.4)	4(2.5)	4(2.5)	3.9(2.5)	0.01
Urine ACR ≥ 30 mg/g, n (%)	130 (0.4%)	2,902 (9.1%)	565 (9.1%)	581 (8.0%)	679 (9.0%)	631 (9.0%)	0.075
eGFR-cys, mL/min/1.73m2	0 (0.0%)	98.1(13.2)	97.3(14.2)	98.1(13.1)	99.0(12.5)	99.7(12.6)	< 0.001
eGFR-cys < 60 mL/min/1.73m2, n (%)	0 (0.0%)	428 (1.3%)	122 (1.9%)	97 (1.3%)	70 (0.9%)	75 (1.1%)	< 0.001
LTPA, METs-hour/day	6,129 (19.1%)	2.9(4.2)	1.8(2.4)	2.9(3.5)	3.3(4.2)	3.4(5.6)	< 0.001
Non-LTPA, METs-hour/day	3,950 (12.3%)	16.6(14.2)	3.6(1.4)	9(1.5)	16.2(2.8)	36.6(13.4)	< 0.001
Sitting time, hour/day	2,804 (8.7%)	225.4(143.9)	266.0(174.3)	240.2(139.4)	219.5(127.8)	178.8(119.4)	< 0.001
Survey season	0 (0.0%)						0.006
Spring, n (%)		4,848 (15.1%)	1,036 (16.5%)	1,046 (14.3%)	1,119 (14.7%)	1,020 (14.5%)
Summer, n (%)		6,689 (20.8%)	1,307 (20.9%)	1,514 (20.8%)	1,555 (20.5%)	1,475 (20.9%)
Autumn, n (%)		20,459 (63.6%)	3,879 (61.9%)	4,698 (64.4%)	4,886 (64.3%)	4,525 (64.2%)
Winter, n (%)		166 (0.5%)	44 (0.7%)	33 (0.5%)	42 (0.6%)	33 (0.5%)

Data are presented as mean (SD) or n (%). One way ANOVA for continuous variables and chi-square test for categorical variable were performed

Cut-off values of quartile range for each physical activity domain are employed from mean cut-off values of multiply imputed datasets (M = 100)

Participants with missing values in each physical activity domain were not listed in each quartile group

BMI Body mass index, CES-D The Center for Epidemiologic Studies Depression scale, ACEI Angiotensin converting enzyme inhibitor, ARB Angiotensin receptor blocker, TG Triglyceride, UA Uric acid, BUN Blood urea nitrogen, eGFR-cys Cystatin C estimated glomerular filtration rate, ACR Albumin-creatinine ratio, LTPA Leisure time physical activity

Mean eGFR values decreased as LTPA quartiles increased (100.7, 99.9, 97.8, and 96.6 mL/min/1.73 m² for the 1st, 2nd, 3rd, and 4th quartile groups, respectively), whereas they increased as non-LTPA quartiles increased (97.3, 98.1, 99.0, and 99.7 mL/min/1.73 m² for the 1st, 2nd, 3rd, and 4th quartile groups, respectively).

In the crude OLS model, eGFR decreased with higher LTPA quartiles (β, -0.89; 95% CI, [-1.35,-0.44]; p < 0.001 for the 2nd quartile, β, -2.72; 95% CI, [-3.16, -2.27]; p < 0.001 for the 3rd quartile, and β, -3.87; 95% CI, [-4.31, -3.42]; p < 0.001 for the 4th quartile group) and increased with higher non-LTPA quartiles (β, 0.82; 95% CI, [0.37, 1.27]; p < 0.001 for the 2nd quartile, β, 1.76; 95% CI, [1.32, 2.20]; p < 0.001 for the 3rd quartile, and β, 2.30; 95% CI, [1.86, 2.74]; p < 0.001 for the 4th quartile groups). For the adjusted OLS model, higher LTPA quartiles were associated with increased eGFR (β, 0.36; 95% CI, [0.06, 0.66]; p = 0.019 for the 2nd quartile, β, 0.82; 95% CI, [0.51, 1.14]; p < 0.001 for the 3rd quartile, and β, 1.16; 95% CI, [0.83, 1.49]; p < 0.001 for the 4th quartile groups, compared with the 1st quartile group), and 4th non-LTPA quartile was associated with decreased eGFR (β, -0.42; 95% CI, [-0.72, -0.11]; p = 0.007) but there were no apparent association for 2nd (β, 0.03; 95% CI, [-0.27, 0.33]; p = 0.842) and 3rd (β, 0.02; 95% CI, [-0.28, 031]; p = 0.912) non-LTPA quartile (Table 3). These associations were consistent in complete case analysis (Additional file: Tables S2) and the sensitivity analysis by limiting the participants to those who responded to the questionnaire within 30 days after blood collection (Additional file: Tables S4).

Table 3.

Association of leisure time and non-leisure time physical activity with kidney function

Variables	Crude model				Adjusted model
	β	95%CI Lower	95%CI Upper	P-value	β	95%CI Lower	95%CI Upper	P-value
LTPA
Quartile 1	ref	ref	ref	ref	ref	ref	ref	ref
Quartile 2	-0.89	-1.35	-0.44	< .001	0.36	0.06	0.66	= .019
Quartile 3	-2.72	-3.16	-2.27	< .001	0.82	0.51	1.14	< .001
Quartile 4	-3.87	-4.31	-3.42	< .001	1.16	0.83	1.49	< .001
Non-LTPA
Quartile 1	ref	ref	ref	ref	ref	ref	ref	ref
Quartile 2	0.82	0.37	1.27	< .001	0.03	-0.27	0.33	= .842
Quartile 3	1.76	1.32	2.20	< .001	0.02	-0.28	0.31	= .912
Quartile 4	2.30	1.86	2.74	< .001	-0.42	-0.72	-0.11	= .007

Adjusted models included age, sex, smoking and alcohol intake status, body mass index, educational status, job, depressive symptom, housing damage, self-reported comorbidity, medication status, and laboratory tests

LTPA Leisure time physical activity, CI Confidence interval

In the crude Poisson model, there was no apparent association between LTPA quartiles and CKD (PR, 1.04; 95% CI, [0.93, 1.15]; p = 0.520 for the 2nd quartile, PR, 1.02; 95% CI, [0.92, 1.13]; p = 0.757 for the 3rd quartile, and PR 1.12; 95% CI, [1.01, 1.24]; p = 0.035 for the 4th quartile groups), whereas higher non-LTPA quartiles were associated with a lower CKD prevalence (PR, 0.87; 95% CI, [0.79, 0.96]; p = 0.007 for the 2nd quartile, PR, 0.89; 95% CI, [0.81, 0.98]; p = 0.023 for the 3rd quartile, and PR, 0.94; 95% CI, [0.85, 1.03]; p = 0.194 for the 4th quartile groups). In the adjusted Poisson model, there was no apparent association between LTPA quartiles and CKD prevalence (PR, 1.00; 95% CI, [0.90, 1.11]; p = 0.959 for the 2nd quartile, PR, 0.91; 95% CI, [0.82, 1.01]; p = 0.079 for the 3rd quartile, and PR 0.93; 95% CI, [0.84, 1.04]; p = 0.210 for the 4th quartile groups), and the 4th non-LTPA quartile was associated with a higher CKD prevalence (PR, 1.12; 95% CI, [1.02, 1.24]; p = 0.022), whereas there were no apparent association in the 2nd (PR, 0.98; 95% CI, [0.89, 1.08]; p = 0.727) and 3rd (PR, 1.06; 95% CI, [0.96, 1.17]; p = 0.265) non-LTPA quartiles, compared with the 1st quartile (Table 4). These associations were not clearly shown in complete case analysis (Additional file: Tables S3) but consistent in the sensitivity analysis by limiting the participants to those who responded to the questionnaire within 30 days after blood collection (Additional file: Tables S5). Higher LTPA was associated with a lower prevalence of eGFR decline, whereas non-LTPA showed no apparent associations in the adjusted model (Additional file: Table S6). For the albuminuria, LTPA showed no apparent associations, whereas 4th non-LTPA quartile was associated with higher prevalence in the adjusted model (Additional file: Table S7).

Table 4.

Association of leisure time and non-leisure time physical activity with chronic kidney disease prevalence

Variables	Crude model				Adjusted model
	PR	95%CI Lower	95%CI Upper	P-value	PR	95%CI Lower	95%CI Upper	P-value
LTPA
Quartile 1	ref	ref	ref	ref	ref	ref	ref	ref
Quartile 2	1.04	0.93	1.15	= .520	1.00	0.90	1.11	= .959
Quartile 3	1.02	0.92	1.13	= .757	0.91	0.82	1.01	= .079
Quartile 4	1.12	1.01	1.24	= .035	0.93	0.84	1.04	= .210
Non-LTPA
Quartile 1	ref	ref	ref	ref	ref	ref	ref	ref
Quartile 2	0.87	0.79	0.96	= .007	0.98	0.89	1.08	= .727
Quartile 3	0.89	0.81	0.98	= .023	1.06	0.96	1.17	= .265
Quartile 4	0.94	0.85	1.03	= .194	1.12	1.02	1.24	= .022

Adjusted models included age, sex, smoking and alcohol intake status, body mass index, educational status, job, depressive symptom, housing damage, self-reported comorbidity, medication status, and laboratory tests

LTPA Leisure time physical activity, PR Prevalence ratio, CI Confidence interval

In the subgroup analyses, higher LTPA quartiles were consistently associated with increased eGFR, whereas higher non-LTPA quartiles showed no apparent associations, similar to the findings of the primary analysis (Fig. 2 for sociodemographic subgroups, and Additional file: Figure S1 for biological subgroups). For the multiplicative model, beta eGFR coefficients for LTPA increased with higher non-LTPA quartiles; however, inconsistent associations of non-LTPA among LTPA quartiles were observed (Additional file: Figure S2).

Fig. 2.

Association between leisure time (A) and non-leisure time (B) physical activity and kidney function in subgroups defined by sociodemographic factors

Discussion

In this cross-sectional study of community-dwelling adults, a positive association was found between higher LTPA levels and better kidney function, whereas no apparent association in prevalence of CKD. The eGFR difference between the 1st and 4th LTPA quartiles was 1.16 mL/min/1.73 m², indicating a change beyond mean for one year [27, 37]. Conversely, highest non-LTPA quartile showed negative association for kidney function, and prevalence of CKD. These associations for kidney function remained robust in subgroup analysis. Moreover, beneficial associations with LTPA were consistent across all non-LTPA quartile strata, whereas non-LTPA showed inconsistent associations. These results reveal that higher LTPA level is particularly beneficial for kidney function, compared with non-LTPA.

Several pathways connect LTPA to kidney function. Engaging in regular physical activity, particularly LTPA, reduces the risk factors associated with CKD, such as hypertension and diabetes [38]. In a study conducted by Kristensen et al., LTPA was linked to a reduced risk of nephropathy in participants with type 2 diabetes from the UK Biobank study [39]. Physical activity also has positive impact on kidney function through the endocrine system. Studies have shown that physical activity is associated with a higher irisin concentration, which beneficially affects kidney function [40–42]. Irisin is known as a myokine which improves kidney energy metabolism and prevent kidney damage [43], and physical activity may help maintain it by preserving muscle mass. Additionally, it suppresses the renin–angiotensin–aldosterone system [44] and enhances nitric oxide concentration, which protects kidney function [45–47]. Moreover, LTPA is associated with exercise tolerance (peak VO2max) [48] which in turn relates to kidney function [49].

One of the major reasons for the different patterns of association between LTPA and non-LTPA lies in the cardiovascular response [23]. Occupational physical activity often involves prolonged, static, and repetitive tasks without adequate recovery time [23]. This sustained cardiovascular response, leading to elevated blood pressure and heart rate, is detrimental to cardiovascular health. The components of non-LTPA, shown in Table S1, indicate that standing and strenuous work accounted for the most time in our study, which aligns with the mechanism discussed. In contrast, leisure activities typically include moderate-to-vigorous intensity, dynamic activities with shorter durations, and adequate recovery times. These activities result in fluctuations in the cardiovascular responses and promote cardiovascular health. Particularly intense occupational activities are linked to vascular stiffness, whereas leisure-time activities show a negative association [50]. Mundwiler et al. demonstrated that LTPA is positively associated with cardiorespiratory fitness, whereas occupational physical activity (i.e., non-LTPA) is not [51]. Considering these physiological pathways, it can be concluded that LTPA has greater benefits on kidney function, compared with non-LTPA. There are no apparent association between LTPA and prevalence of CKD or albuminuria but LTPA was associated with a decreased prevalence of eGFR decline. This result consistently shows a positive association between LTPA and kidney function, though no clear association was found with kidney damage. Xiong et al. reported that higher LTPA is associated with lower risk of albuminuria in patients with diabetes or prediabetes, while occupational physical activity is not [52]. This inconsistence may be due to difference in participant characteristics. In the present study, only 5.2% of participants had diabetes, and the mean HbA1c value was 5.6%. In addition, prevalence of albuminuria in this study (9.1%) was lower than that reported in the previous study, which shows a 14.5% prevalence in general Japanese population [53]. Therefore, participants in this study may have had a lower kidney damage risk, which could explain the lack of a clear association with LTPA. Conversely, higher non-LTPA levels were not associated kidney function but were associated with an increased risk of kidney damage. This associations may be due to the difference in cardiovascular response between non-LTPA and LTPA [23]. Hypertension is associated with albuminuria [54], and sustained elevated blood pressure due to occupational activities, which are included in non-LTPA may contribute to this risk.

This study demonstrated a positive association between LTPA and kidney function, whereas no apparent association was observed with non-LTPA. Holtermann et al. reported that LTPA reduces risk of major adverse cardiovascular events and all-cause mortality. Conversely, intense physical activity at work is associated with increased risk [24]. LTPA is associated with reduced risk of mortality, cardiovascular events, and type 2 diabetes [55]. Zhang et al. showed that engaging in LTPA reduces the risk of CKD incidence [26]. The results of the present study agree with those of previous studies. Existing guidelines by the WHO and American Heart Association recommend promoting MVPA but lack specificity regarding different domains of physical activity (e.g., leisure time and work) [13, 56]. Findings from the present study underscore the importance of physical activity during leisure time, irrespective of activity during non-leisure time; however, many people do not allocate physical activity during their free time [57]. The level of physical activity is influenced not only by individual factors but also by environmental factors such as neighborhood environment and social participation [58]. In this context, it is important to consider how these factors change throughout the life course. Our data indicated that LTPA levels increase with age, while non-LTPA decrease. This trend may be due to full or part-time retirement from work. People who retire, whether at statutory and part-time, increase their physical activity after retirement [59]. LTPA is beneficial even at an early age; engaging in physical activity throughout life is associated with lower subsequent mortality [60], and active behavior is linked to continued activity after retirement [61]. Therefore, developing effective strategies to promote physical activity during leisure time, beginning early in life before retirement from job is important for maintaining kidney function.

The major strengths of this study are the large sample size of community-dwelling general adults and the reliable measurement of kidney function using serum cystatin C. However, this study has some limitations. First, due to the cross-sectional design, the possibility of reverse causation could not be eliminated. This concern was mitigated by excluding individuals with eGFR < 30 mL/min/1.73 m², a group prone to remarkable lifestyle restrictions. However, this issue has not yet been entirely resolved. Second, there is a potential measurement bias in exposure and several covariates because these variables were assessed using a self-report questionnaire. Third, unobserved factors such as health consciousness and socioeconomic status can influence LTPA and kidney function. Especially, socioeconomic status can affect both kidney function and healthy lifestyle, including exercise behavior but we could not measure overall socioeconomic status. However, these factors may not be critical to the results, as we have considered several comorbidities, lifestyle factors (e.g. smoking and alcohol intake), and educational level. Finally, participation bias may have influenced the results, as the participants in this study were recruited from municipal health checkups. People who participate in health checkups generally have better overall health conditions than non-participants [62]. Consequently, there is a possibility of underestimating the proportion of individuals with low kidney function, although this may not significantly impact the results. Furthermore, it is unclear whether these findings can be generalized to other populations, as our study participants were limited to the Japanese population not enrolled in employees’ health insurance. This may particularly affect non-LTPA status due to occupation difference. Further studies are needed to address these limitations and examine longitudinal relationships.

Conclusions

LTPA was found to be beneficially associated with kidney function in community-dwelling adults. In addition, an association between higher LTPA and a lower risk of CKD was observed. Notably, these beneficial associations of LTPA with kidney function were consistent, irrespective of non-LTPAs, which did not show a similar beneficial association. These findings underscore the significance of engaging in physical activity during leisure time in maintaining kidney function.

Abbreviations

CKD

Chronic kidney disease

LTPA

Leisure-time physical activity

eGFR

Estimated glomerular filtration rate

PR

Prevalence ratio

CI

Confidence Interval

MVPA

Moderate-to-vigorous intensity of physical activity

OLS

Ordinary least square

BMI

Body mass index

CES-D

Center for Epidemiologic Studies Depression

ACEI

Angiotensin-converting enzyme inhibitor

ARB

Angiotensin receptor blocker

GEJE

Great East Japan Earthquake

SD

Standard deviation

MI

Multiple imputation

ANOVA

Analysis of variance

WHO

World Health Organization

Authors’ contributions

I.C. contributed to the study design and statistical analysis and wrote the manuscript. N.N. and A.H. contributed to the study design, supported data analysis, and drafted the manuscript. N.N., M.K., R.H., K.N., S.T., T.N., S.N., S.O., T.O., N.F., Y.I., S.K., and A.H. collected the data. T.S. critically revised the manuscript. All the authors contributed intellectually to the manuscript and approved its final version.

Funding

This work was supported by the Tohoku Medical Megabank Project of the Ministry of Education, Culture, Sports, Science, and Technology, and the Japan Agency for Medical Research and Development [AMED; JP22tm0124005]. This study used a supercomputer system provided by the Tohoku Medical Megabank Project (funded by AMED under Grant Number JP21tm0424601).

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All the participants provided written informed consent. This study was approved by the Institutional Review Board of the Tohoku Medical Megabank Organization (approval number: 2022–4–160).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.