Abstract
Rationale & Objective:
Physical activity is associated with lower risk of cardiovascular disease, diabetes, and hypertension, which have shared risk factor profiles with chronic kidney disease (CKD). However, there are conflicting findings regarding the relationship between physical activity and CKD. The objective was to evaluate the association between physical activity and CKD development over long-term follow-up using the Atherosclerosis Risk in Communities (ARIC) study.
Study Design:
Prospective cohort study.
Setting & Participants:
14,537 participants aged 45 to 64 years old.
Predictors:
Baseline physical activity status was assessed by the modified Baecke Physical Activity Questionnaire at visit 1 (1987–1989) and categorized according to the 2018 Physical Activity Guidelines for Americans to group participants as inactive, insufficiently active, active, and highly active.
Outcomes:
Incident CKD defined as estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 at follow up and ≥25% decline in eGFR relative to baseline, CKD-related hospitalization or death, or end stage renal disease.
Analytical Approach:
Cox proportional hazards regression.
Results:
At baseline, 37.8%, 24.2%, 22.7%, and 15.3% of participants were classified as inactive, insufficiently active, active, and highly active, respectively. During a median follow up of 24 years, 33.2% of participants developed CKD. After adjusting for age, sex, race-center, education, smoking status, diet quality, diabetes, coronary heart disease, hypertension, antihypertensive medication, body mass index, and baseline eGFR, higher categories of physical activity were associated with lower risk of CKD compared to the inactive group (HR for insufficiently active, 0.95 [95% CI, 0.88–1.02]; active, 0.93 [95% CI, 0.86–1.01]; highly active, 0.89 [95% CI, 0.81–0.97]; P for trend = 0.007).
Limitations:
Observational design and self-reported physical activity that was based on leisure time activity only. Due to low numbers, non black and white participants were excluded.
Conclusions:
Highly active participants had a lower risk of developing CKD compared to inactive participants.
Keywords: chronic kidney disease, physical activity, Atherosclerosis Risk in Communities Study, estimated glomerular filtration rate, cystatin C
Plain language summary:
Physical activity is associated with a decreased risk of chronic kidney disease
In this study, we evaluated the association between physical activity and risk of developing chronic kidney disease (CKD). We conducted a secondary analysis of the Atherosclerosis Risk in Communities Study, which is a community-based prospective multi-center cohort study of 15,792 middle-aged black and white men and women in the United States. During a median follow up of 24 years, 33.2% of participants developed CKD. After adjusting for confounding variables, the most physically active group had a statistically significant 11% reduction in risk of CKD when compared to the inactive group. Further research is needed to determine whether increasing physical activity can prevent the onset or progression of CKD.
INTRODUCTION
Chronic kidney disease (CKD) affects 11% to 13% of the population worldwide, and is expected to increase in prevalence.1 Physical activity is associated with reduced risk of developing chronic illnesses such as cardiovascular disease, diabetes, and hypertension, which have similar risk profiles and underlying pathophysiology as CKD.2 A better understanding of the relationship between physical activity and CKD may help guide CKD prevention efforts that target modifiable lifestyle factors.
Observational studies examining the relationship between physical activity and kidney function have been conflicting.3–12 Some cross-sectional studies suggested that higher physical activity was associated with higher estimated glomerular filtration rate (eGFR) or lower risk of eGFR <45 mL/min/1.73 m2;3–5 others showed no relationship between physical activity and eGFR <60 mL/min/1.73 m2 or presence of microalbuminuria.6–8 A prospective analysis of older adults in the Cardiovascular Health Study found that greater baseline physical activity was associated with lower risk of eGFR decline >3 mL/min/1.73 m2 per year over 7 years of follow-up.9 A study of the National Health and Nutritional Examination Survey (NHANES II) found that, compared to inactive people, highly active people had reduced risk of developing end stage kidney disease (ESKD) or dying from CKD over a mean of 13 years.10 However, a study of middle-aged participants in the Framingham Offspring study showed no relationship between physical activity and eGFR decline over 6.6 years, similar to a study of Dutch adults, even when multiple measures of physical activity were used.11,12
Using the Atherosclerosis Risk in Communities (ARIC) study, we evaluated the association of physical activity and the development of CKD over long-term follow-up using assessments of physical activity at baseline (midlife) and 6 and 25 years later in both black and white middle-aged adults in the U.S. Because estimates of GFR based on creatinine can be confounded by muscle mass, we also explored the association between physical activity levels at different time points and kidney function in old age, assessing both creatinine based and cystatin C based eGFR.
METHODS
Study Population
The ARIC study is a community-based prospective cohort study of 15,792 middle-aged (45 – 64 years old at baseline) predominately black and white men and women.13 Study participants were recruited and enrolled in 1987 to 1989 (visit 1) from 4 U.S. communities: Forsyth County, North Carolina; Jackson, Mississippi; suburbs of Minneapolis, Minnesota; and Washington County, Maryland. Each community cohort was selected by probability sampling using drivers’ licenses, jury duty listing, voter registration cards, and/or county health census. Participants attended follow up visits in 1990 – 1992 (visit 2), 1993 – 1995 (visit 3), 1996 – 1998 (visit 4), 2011 – 2013 (visit 5), 2016 – 2017 (visit 6), and 2018 – 2019 (visit 7). An ethics committee at each site approved the study protocol, and study participants provided informed consent.
We considered all ARIC study participants, but excluded participants with eGFR <60 mL/min/1.73 m2 (n = 343), who were missing physical activity data at baseline (n = 26), or were missing key covariates (n = 786). To be consistent with previous work, we also excluded participants who were Asian or Indian (n = 46) and African Americans from Washington County, MD (n = 32) or Minneapolis, MN (n = 22). A total of 14,537 ARIC Study participants were included in the analysis (Figure 1).
Figure 1.
Flow chart of selection of ARIC study participants.
Assessment of Physical Activity
Physical activity was assessed only at visits 1, 3 and 5 by an interviewer-administered modified Baecke Physical Activity Questionnaire. Participants were asked about the leisure time physical activities they engaged in and the frequency and duration of participation in each. The format and content of the leisure time questionnaire was consistent across all three visits. The reliability and validity of the modified Baecke Questionnaire has been previously reported.14,15 Each activity was converted into a metabolic equivalent (MET) ranging from 1 to 12, which is a reflection of activity intensity.16 For example, a MET of 3 represents walking at 3 miles per hour and a MET of 8.8 represents jogging at 5.6 miles per hour. These METs were then classified into light (<3 METs), moderate (3 – 5.9 METs), or vigorous intensity (≥6 METs), or a composite of moderate and vigorous intensity (≥3 METs).16
For the categorical analyses, we followed the U.S. Department of Health and Human Services 2018 Physical Activity Guidelines for Americans, which incorporates intensity, duration and frequency of physical activity. Participants were categorized into inactive (0 min/week of moderate or vigorous intensity), insufficiently active (1 – 149 min/week of moderate or 1 – 74 min/week of vigorous or 1 – 149 min/week of moderate-vigorous intensity), active (150 – 300 min/week of moderate or 75 – 150 min/week of vigorous or 150 – 300 min/week of moderate-vigorous intensity), and highly active (>300 min/week of moderate or >150 min/week of vigorous or >300 min/week of moderate-vigorous intensity).17 For the continuous analyses, we generated a variable of MET*min/week to reflect total volume of physical activity. Our primary analysis incorporated baseline physical activity data collected at visit 1 only. In secondary analyses, we also evaluated time-updated physical activity (visits 3 and 5) as a risk factor.
Outcome Assessment
Serum creatinine was measured at visits 1 through 6 by the modified kinetic Jaffé method and used to calculate eGFR with the 2009 Chronic Kidney Disease Epidemiology Collaboration equation.18–20 Hospitalizations were identified by study participants through self-report and ongoing surveillance of community hospital discharge lists. For each hospitalization from visit 1 through December 31st 2017, International Classification of Disease, Ninth/Tenth Revision (ICD-9/10) codes were extracted. Incident CKD was defined by meeting at least one of the following criteria: (1) eGFR <60 mL/min/1.73 m2 accompanied by ≥25% eGFR decline relative to baseline, (2) kidney disease-related hospitalization or death based on ICD-9/10 codes (ICD-9 codes 581 – 583 or 585 – 589; ICD-10 codes N03, N04, N19, N25 – N27) identified through active surveillance and linkage to the National Death Index, or (3) ESKD identified by linkage to the U.S. Renal Data System (USRDS) registry.21 For the ICD-9/10 codes, any diagnosis (primary or secondary) was included.
Assessment of Covariates
Covariates were chosen a priori based on hypothesized associations to both physical activity and kidney disease. Data for sex, age, race, study center, education, smoking status, and alcohol use were assessed at baseline using interviewer-administered questionnaires.13 Race and study center were combined into a single variable (race-center) due to non-uniform distribution of racial groups across centers. Alcohol and smoking status were categorized into current, former, and never. Diet quality was captured using a Dietary Approaches to Stop Hypertension (DASH) diet score that rewards intake of fruits, vegetables, whole grains, nuts and legumes, and low-fat dairy while penalizing intake of red and processed meat, sweetened beverages, and sodium.22 For educational attainment, participants were categorized based on whether or not they graduated high school.
Clinical factors included diabetes, coronary heart disease (CHD), hypertension, use of antihypertensive medication, body mass index (BMI), and baseline eGFR. Diabetes was defined at baseline as fasting blood glucose ≥126 mg/dL, non-fasting glucose ≥200 mg/dL, self-reported history of physician-diagnosed diabetes, or use of diabetes medication in the preceding 2 weeks. CHD was defined by self-reported diagnosis of myocardial infarction or coronary revascularization, or silent myocardial infarction as seen on ARIC electrocardiogram. Hypertension was defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or use of antihypertensive medication in the preceding 2 weeks. To calculate BMI, weight in kilograms was divided by height in meters squared.
Statistical Analysis
Descriptive statistics were used to characterize the study population stratified by physical activity categories. P-values for trend across physical activity as an ordinal variable were tested using logistic regression for categorical variables and linear regression for continuous variables.
We estimated hazard ratios (HR) and 95% confidence intervals (CI) for the association between physical activity measured at visit 1 and incident CKD using Cox proportional hazards regression models. Time at risk began at visit 1 (1987 – 1989) with follow up through December 31st 2017. Three distinct models are presented in order to better understand the impact of covariates on the relationship between physical activity and development of CKD. The minimally adjusted model (model 1) adjusted for age, sex, and race-center. In model 2, we additionally adjusted for education, smoking status, and DASH diet score. In model 3, the fully adjusted model, we additionally adjusted for diabetes, CHD, hypertension, antihypertensive medication, BMI, and baseline eGFR. We repeated these analyses using time-updated assessments of physical activity at visits 3 and 5 as well as accounting for the competing event of death using the method of Fine and Gray.23 Models were run using complete case analysis. In sensitivity analyses, missing data was included using multiple imputation. To demonstrate consistent direction of associations, we also evaluated the association between physical activity and mortality. To test for changes in the association between physical activity and CKD over time, we tested an interaction term with a knot at visit 3 (6 years after baseline).
To explore how the association between physical activity and kidney function varied with time of physical activity assessment and method of eGFR estimation, we evaluated the association between physical activity as a continuous variable at visit 1, visit 3, and visit 5 and eGFR at visit 5 among participants who attended visit 5 with data on kidney function (N = 6,050), using linear regressions adjusted for age, sex, and race-center. Because creatinine is influenced by muscle mass as a non-GFR determinant, we evaluated associations using creatinine based eGFR (eGFRcr) and cystatin C based eGFR (eGFRcys) separately. The 2012 Chronic Kidney Disease Epidemiology Collaboration cystatin C equation was used to calculate the eGFRcys from measured serum cystatin C.24 Lastly, we compared physical activity level at visit 1, 3, and 5 between participants at visit 5 who had an eGFR <60 mL/min/1.73 m2 and those who had an eGFR ≥60 mL/min/1.73 m2 when estimating GFR by creatinine and by cystatin C.
All P values were 2-tailed, and statistical significance was set a priori at P < 0.05. All analyses were completed using Stata software (version 14.0; StataCorp, College Station, TX).
RESULTS
Baseline Characteristics of Participants
Of the 14,537 study participants included in our study, 37.8% were classified as inactive, 24.2% were insufficiently active, 22.7% were active, and 15.3% were highly active at baseline. Participants who were active or highly active were more often male and white, and more likely to consume alcohol, smoke less, and have graduated high school (Table 1). African Americans comprised 39.1% of inactive participants and 13.6% of highly active participants. Inactive participants had on average higher BMI and lower diet quality score. Additionally, inactive participants had a higher frequency of diabetes, hypertension, and antihypertensive medication use. Baseline eGFR was higher among inactive participants compared to highly active participants.
Table 1.
Characteristics of Study Cohort at Visit 1 (1987 – 1989) According to 2018 Physical Activity Guideline Categories.
| Inactive | Insufficiently Active | Active | Highly Active | P | |
|---|---|---|---|---|---|
| N | 5499 (37.8) | 3515 (24.2) | 3301 (22.7) | 2222 (15.3) | |
| MET*min/week | 7 (50) | 361 (203) | 972 (255) | 2011 (716) | <0.001 |
| Female | 3265 (59.4) | 2121 (60.3) | 1733 (52.5%) | 896 (40.3) | <0.001 |
| African American | 2148 (39.1) | 734 (20.9) | 562 (17.0) | 303 (13.6) | <0.001 |
| Age, y | 54.1 (5.7) | 53.9 (5.7) | 54.4 (5.8) | 54.2 (5.8) | 0.01 |
| BMI, kg/m2 | 28.6 (5.9) | 27.4 (5.2) | 27.1 (4.6) | 26.7 (4.4) | <0.001 |
| Fasting Glucose, mg/dL | 112 (45) | 107 (36) | 107 (36) | 106 (34) | <0.001 |
| DASH Diet score | 22.6 (4.8) | 24.3 (4.9) | 25.1 (4.8) | 25.7 (4.9) | <0.001 |
| Total Energy Intake, kcal/d | 1638 (628) | 1580 (586) | 1595 (582) | 1641 (591) | 0.3 |
| eGFRcr, ml/min/1.73 m2 | 106 (15) | 103 (14) | 101 (13) | 100 (13) | <0.001 |
| Alcohol intake status | <0.001 | ||||
| Current Drinker | 2603 (47.5) | 2032 (57.9) | 2084 (63.2) | 1491 (67.2) | |
| Former Drinker | 1154 (21.1) | 639 (18.2) | 536 (16.3) | 374 (16.9) | |
| Never Drank | 1725 (31.4) | 836 (23.8) | 675 (20.4) | 352 (15.8) | |
| Smoking status | <0.001 | ||||
| Current Smoker | 1782 (32.4) | 840 (23.9) | 715 (21.7) | 452 (20.3) | |
| Former Smoker | 1493 (27.2) | 1095 (31.2) | 1196 (36.2) | 920 (41.4) | |
| Never Smoked | 2224 (40.4) | 1580 (45.0) | 1390 (42.1) | 850 (38.3) | |
| CHD | 216 (3.9) | 149 (4.2) | 177 (5.4) | 157 (7.1) | <0.001 |
| Prevalent Heart Failure | 342 (6.2) | 118 (3.4) | 132 (4.0) | 57 (2.6) | <0.001 |
| Prevalent MI | 195 (3.5) | 124 (3.5) | 150 (4.5) | 123 (5.5) | <0.001 |
| Hypertension | 1867 (34.0) | 854 (24.3) | 840 (25.4) | 475 (21.4) | <0.001 |
| Antihypertensive Medication | 1648 (30.0) | 766 (21.8) | 760 (23.0) | 428 (19.3) | <0.001 |
| Diabetes | 673 (12.2) | 292 (8.3) | 259 (7.8) | 144 (6.5) | <0.001 |
| High school graduate | 3671 (66.8) | 2827 (80.4) | 2741 (83.0) | 1940 (87.3) | <0.001 |
| Study site | <0.001 | ||||
| Forsyth County, NC | 1195 (21.7) | 970 (27.6) | 927 (28.1) | 667 (30.0) | |
| Jackson, MS | 1988 (36.2) | 601 (17.1) | 476 (14.4) | 261 (11.7) | |
| Minneapolis, MN | 980 (17.8) | 1027 (29.2) | 1094 (33.1) | 715 (32.2) | |
| Washington County, MD | 1336 (24.3) | 917 (26.1) | 804 (24.4) | 579 (26.1) |
Abbreviations: MET, metabolic equivalent of task; BMI, body mass index; DASH, Dietary Approaches to Stop Hypertension; eGFRcr, creatinine based estimated glomerular filtration rate; CHD, coronary heart disease; MI, myocardial infarction.
Values for categorical variables are given as number (percentage); for continuous variables, as mean (standard deviation).
Categories defined as inactive (0 min/week of moderate or vigorous intensity), insufficiently active (1 – 149 min/week of moderate or 1 – 74 min/week of vigorous or 1 – 149 min/week of moderate-vigorous intensity), active (150 – 300 min/week of moderate or 75 – 150 min/week of vigorous or 150 – 300 min/week of moderate-vigorous intensity), and highly active (>300 min/week of moderate or >150 min/week of vigorous or >300 min/week of moderate-vigorous intensity).
Association Between Physical Activity and Incident CKD
There were 4,820 (33.2%) participants who developed CKD during a median follow up of 24 years. In model 1 and 2, any level of physical activity was significantly associated with a lower risk for incident CKD compared to being inactive. In model 3, the fully adjusted model, the trend was consistent but only the highly active group remained significantly associated with a lower risk for incident CKD (Table 2). Those who were insufficiently active, active, and highly active had, respectively, a 5% (HR: 0.95, 95% CI: 0.88 – 1.02), 7% (HR: 0.93, 95% CI: 0.86 – 1.01), and 11% (HR: 0.89, 95% CI: 0.81 – 0.97) lower risk of CKD compared to those who were inactive (P for trend = 0.007). There were no statistically significant differences in the association between level of physical activity and risk for incident CKD before or after visit 3.
Table 2.
Risk for CKD from 1987 – 1989 to 2017 Associated with Physical Activity Category.
| 2018 Physical Activity Guidelines Category | ||||||
|---|---|---|---|---|---|---|
| Inactive (N=5,499) |
Insufficiently Active (N=3,515) |
Active (N=3,301) |
Highly Active (N=2,222) |
P for Trend | ||
| No. of Events | 1,905 | 1,129 | 1,078 | 708 | ||
| Association between CKD and physical activity at baseline | ||||||
| Model 1 | 1 (reference) | 0.87 (0.80–0.93) | 0.87 (0.80–0.94) | 0.81 (0.74–0.89) | <0.001 | |
| Model 2 | 1 (reference) | 0.90 (0.84–0.97) | 0.92 (0.85–0.99) | 0.87 (0.80–0.96) | 0.003 | |
| Model 3 | 1 (reference) | 0.95 (0.88–1.02) | 0.93 (0.86–1.01) | 0.89 (0.81–0.97) | 0.007 | |
| Associations incorporating time updated measures of physical activity | ||||||
| Model 1 | 1 (reference) | 0.78 (0.71–0.85) | 0.75 (0.68–0.82) | 0.73 (0.66–0.81) | <0.001 | |
| Model 2 | 1 (reference) | 0.81 (0.74–0.90) | 0.79 (0.72–0.87) | 0.79 (0.71–0.88) | <0.001 | |
| Model 3 | 1 (reference) | 0.88 (0.80–0.96) | 0.85 (0.78–0.94) | 0.83 (0.74–0.92) | <0.001 | |
| Associations accounting for competing risk of death | ||||||
| Model 1 | 1 (reference) | 0.94 (0.87–1.01) | 0.95 (0.88–1.03) | 0.93 (0.85–1.01) | 0.09 | |
| Model 2 | 1 (reference) | 0.94 (0.88–1.02) | 0.96 (0.89–1.04) | 0.94 (0.85–1.03) | 0.2 | |
| Model 3 | 1 (reference) | 0.97 (0.90–1.05) | 0.98 (0.90–1.06) | 0.95 (0.87–1.05) | 0.4 | |
Values are given as hazard ratio (95% confidence interval).
Model 1: Adjusted for age, sex, and race-center.
Model 2: Model 1 + education, smoking status, and DASH diet score.
Model 3: Model 2 + diabetes, CHD, hypertension, antihypertensive medication, BMI, and baseline eGFR.
Categories defined as inactive (0 min/week of moderate or vigorous intensity), insufficiently active (1 – 149 min/week of moderate or 1 – 74 min/week of vigorous or 1 – 149 min/week of moderate-vigorous intensity), active (150 – 300 min/week of moderate or 75 – 150 min/week of vigorous or 150 – 300 min/week of moderate-vigorous intensity), and highly active (>300 min/week of moderate or >150 min/week of vigorous or >300 min/week of moderate-vigorous intensity).
Associations were stronger when incorporating time-updated measures of physical activity (Table 2). In the fully adjusted model, those who were insufficiently active, active, and highly active, had, respectively, a 12% (HR: 0.88, 95% CI: 0.80 – 0.96), 15% (HR: 0.85, 95% CI: 0.78 – 0.94), and 17% (HR: 0.83, 95% CI: 0.74 – 0.92) lower risk of CKD compared to those who were inactive (P for trend < 0.001).
There were 7,109 deaths over the study period, and physical activity category was strongly associated with mortality risk. In the fully adjusted model, those who were insufficiently active, active, and highly active, had, respectively, a 8% (HR: 0.92, 95% CI: 0.86 – 0.98), 9% (HR: 0.91, 95% CI: 0.85 – 0.97), and 14% (HR: 0.86, 95% CI: 0.80 – 0.93) lower risk of mortality compared to those who were inactive (P for trend < 0.001). Associations with CKD were thus attenuated and no longer significant when taking into account the competing risk of death (Table 2). In the fully adjusted model, those who were insufficiently active, active, and highly active at baseline had, respectively, a 3% (subHR: 0.97, 95% CI: 0.90 – 1.05), 2% (subHR: 0.98, 95% CI: 0.90 – 1.06), and 5% (subHR: 0.95, 95% CI: 0.87 – 1.05) lower risk of CKD compared to those who were inactive (P for trend = 0.4).
There were no meaningful changes to the results when missing data was included using multiple imputation (Table S1).
Association between Physical Activity at Visits 1, 3, and 5 and GFR estimated using serum creatinine or cystatin C
There was a significant association between physical activity at visit 1 and eGFRcr at visit 5 (ß: 0.104, 95% CI: 0.050 – 0.157 for every 100 MET*min/week) (Figure 2, top row). This association was similar to a more proximal measurement of physical activity at visit 3 (ß: 0.102, 95% CI: 0.048 – 0.155) and stronger when measured at visit 5 (ß: 0.152, 95% CI: 0.098 – 0.206). All associations between physical activity and eGFRcys were stronger (Figure 2, bottom row).
Figure 2. Physical activity measured at visit 1 (1987 – 1989), visit 3 (1993 – 1995), and visit 5 (2011 – 2013) is associated with higher creatinine and cystatin C based eGFR at visit 5 (2011 – 2013) among attendees at visit 5 (N=6,050).
Association between creatinine based (eGFRcr; top row) and cystatin C based (eGFRcys; bottom row) estimated glomerular filtration rate at visit 5 and log scaled physical activity levels measured at visits 1, 3, and 5. Dots represent individual subjects and the line is a fitted line. The beta is adjusted for age, sex, and race-center. There were stronger associations between physical activity at all visits and eGFRcys at visit 5 compared to eGFRcr at visit 5.
Participants with an eGFRcr <60 mL/min/1.73 m2 had lower average physical activity level at visit 1 compared to participants with an eGFRcr ≥60 mL/min/1.73 m2 (643 MET*min/week vs. 699 MET*min/week, p = 0.006), as well as at visit 3 (689 MET*min/week vs. 728 MET*min/week, p = 0.05) and at visit 5 (654 MET*min/week vs. 815 MET*min/week, p < 0.001). Differences were greater when eGFRcys was used (visit 1: 628 MET*min/week vs. 737 MET*min/week, p < 0.001; visit 3: 664 MET*min/week vs. 769 MET*min/week, p < 0.001; visit 5: 617 MET*min/week vs. 911 MET*min/week, p < 0.001).
DISCUSSION
In this study of 14,537 middle-aged adults in the United States, we found that nearly 38% of participants were physically inactive at baseline. Those who were highly active had a lower risk of developing CKD compared to those who were physically inactive. Moreover, greater physical activity was associated with higher eGFR in older age, even when physical activity was assessed 25 years prior. Associations between physical activity and eGFR were uniformly stronger using cystatin C compared to creatinine, indicating that the associations with incident CKD assessed using serum creatinine may be conservative.
Our study adds to the existing literature by demonstrating an association between physical activity and incident CKD over a median of 24 years of follow up in a population of middle-aged black and white adults. The Cardiovascular Health Study of adults aged 65 and older also found that greater physical activity was associated with a lower risk of kidney function decline.9 Together, these results suggest that the benefit of physical activity in adults may span both middle and older adulthood. Physical inactivity assessed at a single time point in NHANES II was also associated with ESKD or death from CKD (RR: 2.2, 95% CI: 1.3 – 3.8 for inactive vs. very active).10
Studies demonstrating no association between physical activity and CKD have generally evaluated more active, homogeneous populations or had shorter follow-up. The Doetinchem Cohort Study of people from rural Netherlands found that physical activity was not associated with a change in eGFRcys assessed 3 times over 15 years (β: 0.57 mL/min/1.73 m2, 95% CI: −1.70 – 0.56 for inactive vs. active).11 However, the study enrolled predominately white participants and only 4% were physically inactive. An analysis of the Framingham Offspring Cohort (average age = 59 years) did not find a relationship between physical activity levels and incident eGFR <60 mL/min/1.73 m2 (OR: 1.19, 95% CI: 0.71 – 1.99 for highest activity vs. lowest activity) or rapid eGFR decline ≥3 mL/min/1.73 m2 per year (OR: 0.93, 95% CI: 0.61 – 1.42 for highest activity vs. lowest activity) over 6.6 years.12
There is also clinical-trial evidence of short-term benefits of physical activity in patients with CKD. Patients with CKD stage G3–4 randomized to a 3-month aerobic training program had improved aerobic capacity compared to those who did not.25 Patients with CKD, obesity, and type 2 diabetes randomized to a diet management and exercise program had improved exercise capacity compared to the diet management only group, although there were no significant differences in renal function.26 A small study (N=18) of patients with CKD stage G3–4 randomized to a 1-year resistance and aerobic training program showed a slower rate of eGFR decline compared to usual care (mean difference = 7.8 [95% CI, 1.1 – 13.5] mL/min/1.73m2 per year; p = 0.02).27
A significant proportion of the ARIC cohort was classified as physically inactive at baseline. One practical intervention to improve physical activity levels may be targeted counseling by primary care physicians. Many physicians do not ask about the physical activity habits of their patients.28–30 Physical activity counseling by primary care physicians improves physical activity levels in patients, as evidenced by interventions such as the Green Prescription and the Patient-centered Assessment and Counseling on Exercise plus Nutrition (PACE+), and regular physical activity may protect against many chronic illnesses.2,31–34
Our study had several strengths. The ARIC study is a large prospective cohort with a median of 24 years of follow-up and representation from four different U.S. communities. The population has black and white men and women, enabling fairly broad generalizability. The long study duration and well-measured outcomes allowed for robust characterization of the association between physical activity and CKD. We provide both baseline and time-updated associations of physical activity and outcomes. We show that the association between physical activity and incident CKD was attenuated when taking into account the competing risk of death. This is due to the fact that low physical activity was also a strong risk factor for death, and in this long-term study of middle aged adults, nearly 50% of the study population died. The consistent protective association of greater physical activity and CKD provides a degree of reassurance as to the safety of a potential intervention to increase physical activity.
However, our study also had limitations. The primary outcome in our study was assessed by eGFRcr. Results may have been stronger had kidney function been routinely assessed using cystatin C. Cystatin C and creatinine have distinct non-GFR determinants, with cystatin C less affected by muscle mass compared to creatinine.35 People with higher levels of physical activity may have higher creatinine levels due to muscle mass, which could result in an underestimation of GFR when using eGFRcr. The physical activity data was self-reported, which is subject to error and potentially systematic bias, and was based on leisure time physical activity data only. We may have underestimated the baseline activity level of participants in the study, particularly those with physically demanding jobs. Due to low numbers, we excluded participants whose self-reported race was not black or white, which limits the generalizability of the study findings to these populations. Moreover, while we controlled for behaviors that may be associated with physical activity such as diet and smoking status, residual confounding is possible. It is likely that physical activity correlates with other healthy lifestyle behaviors, which may influence the observed associations with CKD. Finally, the study was observational, thus causality could not be determined.
In summary, our large prospective cohort study of 14,537 black and white men and women in the United States found that physical inactivity was associated with a higher risk of incident CKD over a median of 24 years of follow-up. CKD is one of the most common comorbidities among older adults and represents a significant public health burden. Additional studies are needed to determine whether increasing physical activity levels may prevent the onset or progression of CKD.
Supplementary Material
Acknowledgements:
The authors thank the staff and participants of the ARIC study for their important contributions.
Support: The Atherosclerosis Risk in Communities study has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services (contract numbers HHSN268201700001I, HHSN268201700002I, HHSN268201700003I, HHSN268201700004I and HHSN268201700005I). KP was supported by the Johns Hopkins School of Medicine Dean’s Summer Research Funding.
CMR was supported by funding from the National Institute of Diabetes and Digestive and Kidney Diseases (K01 DK107782) and the National Heart, Lung, and Blood Institute (R21 HL143089).
PP was supported by grant R00 AG052830 from the National Institute of Aging. The funders had no role in study design, data collection, analysis, reporting, or the decision to submit for publication.
Footnotes
Financial Disclosure: The authors declare that they have no relevant financial interests.
Publisher's Disclaimer: Disclaimer: Some of the data reported here have been supplied by the US Renal Data System (USRDS). The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US government.
REFERENCES
- 1.Hill NR, Fatoba ST, Oke JL, et al. Global Prevalence of Chronic Kidney Disease - A Systematic Review and Meta-Analysis. PLoS One. 2016;11(7):e0158765. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Warburton DE, Nicol CW, Bredin SS. Health benefits of physical activity: the evidence. CMAJ. 2006;174(6):801–809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Martens RJH, van der Berg JD, Stehouwer CDA, et al. Amount and pattern of physical activity and sedentary behavior are associated with kidney function and kidney damage: The Maastricht Study. PLoS One. 2018;13(4):e0195306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hallan S, de Mutsert R, Carlsen S, Dekker FW, Aasarod K, Holmen J. Obesity, smoking, and physical inactivity as risk factors for CKD: are men more vulnerable? Am J Kidney Dis. 2006;47(3):396–405. [DOI] [PubMed] [Google Scholar]
- 5.Finkelstein J, Joshi A, Hise MK. Association of physical activity and renal function in subjects with and without metabolic syndrome: a review of the Third National Health and Nutrition Examination Survey (NHANES III). Am J Kidney Dis. 2006;48(3):372–382. [DOI] [PubMed] [Google Scholar]
- 6.Bharakhada N, Yates T, Davies MJ, et al. Association of sitting time and physical activity with CKD: a cross-sectional study in family practices. Am J Kidney Dis. 2012;60(4):583–590. [DOI] [PubMed] [Google Scholar]
- 7.Hawkins MS, Sevick MA, Richardson CR, Fried LF, Arena VC, Kriska AM. Association between physical activity and kidney function: National Health and Nutrition Examination Survey. Med Sci Sports Exerc. 2011;43(8):1457–1464. [DOI] [PubMed] [Google Scholar]
- 8.White SL, Dunstan DW, Polkinghorne KR, Atkins RC, Cass A, Chadban SJ. Physical inactivity and chronic kidney disease in Australian adults: the AusDiab study. Nutr Metab Cardiovasc Dis. 2011;21(2):104–112. [DOI] [PubMed] [Google Scholar]
- 9.Robinson-Cohen C, Katz R, Mozaffarian D, et al. Physical activity and rapid decline in kidney function among older adults. Arch Intern Med. 2009;169(22):2116–2123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Stengel B, Tarver-Carr ME, Powe NR, Eberhardt MS, Brancati FL. Lifestyle factors, obesity and the risk of chronic kidney disease. Epidemiology. 2003;14(4):479–487. [DOI] [PubMed] [Google Scholar]
- 11.Herber-Gast GC, Hulsegge G, Hartman L, et al. Physical Activity Is not Associated with Estimated Glomerular Filtration Rate among Young and Middle-Aged Adults: Results from the Population-Based Longitudinal Doetinchem Study. PLoS One. 2015;10(10):e0133864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Foster MC, Hwang SJ, Massaro JM, Jacques PF, Fox CS, Chu AY. Lifestyle factors and indices of kidney function in the Framingham Heart Study. Am J Nephrol. 2015;41(4–5):267–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129(4):687–702. [PubMed] [Google Scholar]
- 14.Baecke JA, Burema J, Frijters JE. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr. 1982;36(5):936–942. [DOI] [PubMed] [Google Scholar]
- 15.Hertogh EM, Monninkhof EM, Schouten EG, Peeters PH, Schuit AJ. Validity of the modified Baecke questionnaire: comparison with energy expenditure according to the doubly labeled water method. Int J Behav Nutr Phys Act. 2008;5:30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575–1581. [DOI] [PubMed] [Google Scholar]
- 17.Piercy KL, Troiano RP, Ballard RM, et al. The Physical Activity Guidelines for Americans. JAMA. 2018;320(19):2020–2028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lustgarten JA, Wenk RE. Simple, rapid, kinetic method for serum creatinine measurement. Clin Chem. 1972;18(11):1419–1422. [PubMed] [Google Scholar]
- 19.Parrinello CM, Grams ME, Couper D, et al. Recalibration of blood analytes over 25 years in the atherosclerosis risk in communities study: impact of recalibration on chronic kidney disease prevalence and incidence. Clin Chem. 2015;61(7):938–947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Grams ME, Rebholz CM, McMahon B, et al. Identification of incident CKD stage 3 in research studies. Am J Kidney Dis. 2014;64(2):214–221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rebholz CM, Crews DC, Grams ME, et al. DASH (Dietary Approaches to Stop Hypertension) Diet and Risk of Subsequent Kidney Disease. Am J Kidney Dis. 2016;68(6):853–861. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496–509. [Google Scholar]
- 24.Inker LA, Eckfeldt J, Levey AS, et al. Expressing the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) cystatin C equations for estimating GFR with standardized serum cystatin C values. Am J Kidney Dis. 2011;58(4):682–684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Van Craenenbroeck AH, Van Craenenbroeck EM, Van Ackeren K, et al. Effect of Moderate Aerobic Exercise Training on Endothelial Function and Arterial Stiffness in CKD Stages 3–4: A Randomized Controlled Trial. Am J Kidney Dis. 2015;66(2):285–296. [DOI] [PubMed] [Google Scholar]
- 26.Leehey DJ, Collins E, Kramer HJ, et al. Structured Exercise in Obese Diabetic Patients with Chronic Kidney Disease: A Randomized Controlled Trial. Am J Nephrol. 2016;44(1):54–62. [DOI] [PubMed] [Google Scholar]
- 27.Greenwood SA, Koufaki P, Mercer TH, et al. Effect of exercise training on estimated GFR, vascular health, and cardiorespiratory fitness in patients with CKD: a pilot randomized controlled trial. Am J Kidney Dis. 2015;65(3):425–434. [DOI] [PubMed] [Google Scholar]
- 28.Petrella RJ, Lattanzio CN, Overend TJ. Physical activity counseling and prescription among canadian primary care physicians. Arch Intern Med. 2007;167(16):1774–1781. [DOI] [PubMed] [Google Scholar]
- 29.Belfrage ASV, Grotmol KS, Tyssen R, et al. Factors influencing doctors’ counselling on patients’ lifestyle habits: a cohort study. BJGP Open. 2018;2(3):bjgpopen18X101607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Carroll JK, Fiscella K, Meldrum SC, et al. Clinician-patient communication about physical activity in an underserved population. J Am Board Fam Med. 2008;21(2):118–127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Calfas KJ, Sallis JF, Zabinski MF, et al. Preliminary evaluation of a multicomponent program for nutrition and physical activity change in primary care: PACE+ for adults. Prev Med. 2002;34(2):153–161. [DOI] [PubMed] [Google Scholar]
- 32.Elley CR, Kerse N, Arroll B, Robinson E. Effectiveness of counselling patients on physical activity in general practice: cluster randomised controlled trial. BMJ. 2003;326(7393):793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kruk J Physical activity in the prevention of the most frequent chronic diseases: an analysis of the recent evidence. Asian Pac J Cancer Prev. 2007;8(3):325–338. [PubMed] [Google Scholar]
- 34.Friedenreich CM, Orenstein MR. Physical activity and cancer prevention: etiologic evidence and biological mechanisms. J Nutr. 2002;132(11 Suppl):3456S–3464S. [DOI] [PubMed] [Google Scholar]
- 35.Inker LA, Schmid CH, Tighiouart H, et al. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med. 2012;367(1):20–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.