Associations of sleeping patterns and isotemporal substitution of other behavior with the prevalence of CKD in Chinese adults

Abstract

Studies have found a U-shaped relationship between sleep duration and chronic kidney disease (CKD) risk, but limited research evaluated the association of reallocating excessive sleep to other behavior with CKD. We included 104 538 participants from the nationwide cohort of the Risk Evaluation of Cancers in Chinese Diabetic Individuals: A Longitudinal Study, with self-reported time of daily-life behavior. Using isotemporal substitution models, we found that substituting 1 h of sleeping with sitting, walking, or moderate-to-vigorous physical activity was associated with a lower CKD prevalence. Leisure-time physical activity displacement was associated with a greater prevalence reduction than occupational physical activity in working population. In stratified analysis, a lower CKD prevalence related to substitution toward physical activity was found in long sleepers. More pronounced correlations were observed in long sleepers with diabetes than in those with prediabetes, and they benefited from other behavior substitutions toward a more active way. The U-shaped association between sleep duration and CKD prevalence implied the potential effects of insufficient and excessive sleep on the kidneys, in which the pernicious link with oversleep could be reversed by time reallocation to physical activity. The divergence in the predicted effect on CKD following time reallocation to behavior of different domains and intensities and in subpopulations with diverse metabolic statuses underlined the importance of optimizing sleeping patterns and adjusting integral behavioral composition.

Keywords

sleep / physical activity / chronic kidney disease / isotemporal substitution / behavioral pattern

1 Introduction

As a major public health problem globally, chronic kidney disease (CKD) is a progressive disease with high mortality, and much of this substantial burden is expected to occur in Asia [1]. Given the interrelationship between CKD and diabetes, kidney function preservation can be attained in a huge population of patients through lifestyle modifications, which should be given priority in light of its long-term effect on improving cardiometabolic health without dialysis [2]. Specifically, as a behavioral component taking up nearly 1/3 of lifetime, sleeping has been increasingly recognized as a critical lifestyle factor associated with public health; thus, identifying an optimal sleeping pattern is imperative to ameliorate the CKD burden.

A U-shaped relationship is believed to exist between sleep duration and CKD risk [3–5]. In particular, long sleep durations were associated with CKD compared with ideal durations (7–8 h). Previous studies in the UK and East Asia associated long sleep durations with an increased incidence of declining estimated glomerular filtration rate (eGFR) [6–8], renal hyperfiltration [9,10], and urinary albumin/creatinine ratio (ACR) [11]. In addition, potential harmful effects of longer sleep duration on kidney function in a casual mode have been validated by Mendelian randomization analysis [12].

However, limited research has addressed the priority of optimizing sleeping patterns by adjusting the overall composition of daily behavioral subcomponents to achieve improved renal health. Isotemporal substitution models (ISMs) have enabled exploring the associations between the proportion spent in different activity patterns and health outcomes, providing insightful evidence for tailored behavioral adjustments.

A few studies have adopted ISMs to examine associations with CKD. As observed in the UK Biobank [13], Framingham Offspring Study [14], and Japanese older adults [15], replacing 1 h/day of sitting time with equivalent physical activities was inversely associated with decreased eGFR. However, the studies were confined largely to the effect of sedentary behavior substitution on CKD, but they hardly considered the replacement of excessive sleep. Given the vital role of sleep as a component of 24 h physical behavior, as well as the importance of 24 h behavior mode for metabolic health, issues concerning the modulation of sleep quantity, quality, and timing deserve further discussion, as is emphasized by recent international consensus [16].

Accordingly, by performing a comprehensive analysis of compositional behavior patterns, we aimed to (1) confirm the association between sleeping patterns and CKD prevalence and (2) investigate the effect of reallocating sleeping time to other behavior on the basis of our large-scale general population and among subgroups with different sleeping patterns and glycemic statuses. Findings will address key evidence gaps regarding the role of excessive sleeping patterns in CKD development, as well as the value of overall behavior mode adjustment to CKD prevention.

2 Materials and methods

2.1 Study population

The Risk Evaluation of Cancers in Chinese Diabetic Individuals: A Longitudinal Study has been set up as a community-based cohort study. The baseline phase was conducted from 2011 to 2012, and details of its design and protocol have been previously described elsewhere [17]. The current study was a cross-sectional analysis of the baseline data. Generally, a total of 259 657 people aged 40 years or older were recruited from 25 communities across the mainland of China. For the current study, we excluded participants with missing data on sleep, sedentary behavior or physical activity (n = 52 012), serum creatinine or urinary ACR (n = 85 827), and other metabolic measurements (n = 17 280). Ultimately, 104 538 people were included in the present analysis. The study was approved by the Medical Ethics Committee of Rui-jin Hospital, and all participants gave their written informed consent.

2.2 Sleeping, sitting, and physical activity (exposures)

Lifestyle habits were collected by face-to-face interviews with a structured questionnaire. Time spent on sleeping or sitting was self-reported in response to the following questions: “How much time did you spend on sitting on weekdays and on weekends in the recent 7 days?” and “How much sleep did you actually get at night on average in the recent 7 days?” The responses were recorded as hours per day.

Physical activity was estimated using the short form of the International Physical Activity Questionnaire, which includes questions on the intensity, duration, and frequency of a physical activity. In accordance with the reported intensity of physical activities, we classified physical activities into walking and moderate-to-vigorous physical activity (MVPA). For the working population, total MVPA was the sum of occupational physical activity (OPA) and leisure-time physical activity (LTPA).

The self-reported daily time spent on each behavior above was included in the analysis as continuous variables. If the total time reported exceeded 24 h per day, the excess time would be subtracted from each time module in accordance with the relative distribution. Otherwise, no special processing was performed for data with a total time < 24 h per day.

2.3 CKD outcome

All participants fasted overnight before blood samples were taken. Serum samples were shipped in dry ice to the central laboratory. Creatinine was measured with an autoanalyzer (Abbott Laboratories, IL, USA). eGFR was calculated using the CKD Epidemiology Collaboration equation for Asian individuals. As recommended in the Kidney Disease: Improving Global Outcomes (KDIGO) 2012 guidelines [18], we defined CKD with the following criteria: GFR < 60 mL/min per 1.73 m² or ACR ≥ 30 mg/g.

CKD severity was classified by the level of eGFR and ACR in accordance with the KDIGO 2012 recommendations. Specifically, GFR included a five-stage classification (–G5), while ACR was classified into three categories (A1–A3). Participants with G1–2 and A1 were defined to be at low risk for CKD; participants with G1–2 and A2 or G3a and A1 were defined to be at moderate risk for CKD; participants with G1–2 and A3, G3a and A2, or G3b and A1 were defined to be at high risk for CKD; participants with G4–5 and A1, G3b–5 and A2, or G3a–5 and A3 were defined to be at very high risk for CKD, as detailed in the Appendix in the Supplementary Material. We combined the moderate- and high-risk groups into a moderate-to-high-risk group in this analysis.

2.4 Classification of glycemic status

In accordance with the American Diabetes Association (ADA) criteria [16], prediabetes was diagnosed if the fasting plasma glucose was 5.6–6.9 mmol/L, 2 h-PG was 7.8–11.0 mmol, or HbA1c was 5.7% to 6.4%. Diabetes was diagnosed if the fasting plasma glucose ≥ 7 mmol/L, 2h-PG ≥ 11 mmol/L, or HbA1c ≥ 6.5%, or a prior diagnosis was made by physicians. Other participants were considered to have normal glucose regulation (NGR).

2.5 Covariates

Age, gender, education level, smoking, alcohol drinking, diet quality, and current medications were recorded by baseline questionnaire. Height and weight were taken by skilled nurses in accordance with standard protocols. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m²). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured 3 times after a rest of at least 5 min, and the average of readings was used. Detailed information on covariates and other metabolic factors is presented in the Appendix in the Supplementary Material.

2.6 Statistical analyses

Statistical analyses were performed with SAS version 9.4. and R version 4.2.1 software. The statistical significance level was a two-tailed P < 0.05. Baseline characteristics were demonstrated in accordance with sleep durations (< 6 h, 6–8 h, ≥ 8 h). Subject characteristics conforming to a normal distribution were presented as either means ± standard deviations (SDs) or as medians (interquartile ranges). All categorical variables were presented as numbers (proportions, %). Potential nonlinear associations between sleep duration and the odds ratios (ORs) of CKD were examined with a restricted cubic spline [19] model with three knots, namely, 25th, 50th, and 75th percentiles of sleep duration (7, 8, and 9 h, respectively). Analyses were adjusted for variables, including age, gender, education level, smoking, alcohol drinking, diet quality, BMI, HbA1c, SBP, DBP, and use of hypoglycemic and antihypertension drugs. Binary logistic regression models were utilized to assess the association between exposures (1 h unit of sleeping, sitting, walking, or MVPA) and outcomes (overall CKD, abnormal ACR, decreased eGFR). The adjusted ORs and 95% coefficient intervals (CIs) were calculated.

ISM was conducted to study the potential association of reallocating time spent on one behavior to another by excluding the behavior of interest while keeping the total time and other behavior fixed. In this study, we examined the association of time reallocation with CKD outcomes (overall CKD, abnormal ACR, decreased eGFR, and the reduction in CKD severity) by using a generalized logit model. All exposure variables, including sleeping, sitting, walking, and MVPA, were modeled in 1 h unit. We considered three multivariable-adjusted logistic regression models. First, a single regression model examined the raw associations of each behavioral exposure and the CKD outcome with only covariates in the model. Second, a partition model examined the associations of each behavioral exposure variable after adjusting for all other exposure variables and covariates. Lastly, isotemporal substitution modeled the associations of reallocating one behavioral exposure variable to another (e.g., sleeping to walking) while adjusting for covariates and the time spent in other activities. This step was accomplished by including a total time variable in the model and dropping the exposure variable of interest. ISM is thus able to overcome the limitation of conventional regression models where the effects of activities were studied in isolation [20]. Adjustments for covariates were the same as previously mentioned. We conducted sensitivity analyses by rerunning the models in working or unemployed populations separately. In stratified analysis, participants were divided into groups in accordance with sleep durations (< 8 h, ≥ 8 h) and bedtime (≤ 23:00, > 23:00). Participants with abnormal sleep rhythms, such as night shift or daytime fragment sleep, were excluded. Analyses were further conducted in subgroups on the basis of glycemic status (NGR, prediabetes, and diabetes).

3 Results

The baseline characteristics of the participants based on the categories of sleeping duration are presented in Tab.1. Among the participants, 5241 (5.0%) had sleep durations less than 6 h, 30 358 (29.0%) had sleep durations between 6 and 8 h, and 68 939 (66.0%) slept for 8 h or more (Tab.1). The long sleepers tended to have lower BMI, higher BP, and a lower education level. They were also more likely to be male, drinkers, and currently working. A total of 12 042 (11.5%) participants were classified as CKD, of which long sleepers presented a significantly higher prevalence of CKD (12.1%).

Multiple-adjusted restricted cubic spline analyses indicated an evident U-shaped association between sleep duration and abnormal ACR (P for nonlinearity < 0.001), but such association was not apparent with decreased eGFR (P for nonlinearity = 0.619) (Fig.1).

OR estimates from fully adjusted single, partition, and substitution models are presented in Tab.2. In single and partition models, sleeping and sedentary time were significantly associated with CKD. In fully adjusted ISM, replacing 1 h/day of sleeping behavior with an equal amount of time in any other behavior, including sitting (OR: 0.98; 95% CI: 0.97–0.998), walking (OR: 0.95; 95% CI: 0.93–0.98), and MVPA (OR: 0.96; 95% CI: 0.95–0.98), was associated with a significantly lower prevalence of CKD. A statistically significant beneficial association of substituting 1 h/day of sedentary behavior with walking or MVPA was also observed. As for the two indicators of kidney damage, the replacement of sleep with any other behavior was associated with a lower prevalence of abnormal ACR, directionally consistent with that of overall CKD, but it was not associated with decreased eGFR, except for sleep-to-MVPA reallocation. Meanwhile, CKD featured in decreased eGFR largely benefited from other behavior substitution toward a more active way (sit-to-walk and sit-to-MVPA), whereas abnormal ACR did not (Table S1, Fig. S1). We next examined the correlation between behavior substitution and a decrease in CKD severity. Using the low-risk group as a reference, we identified an association between reallocating sleep to sitting, walking, or MVPA and a reduction in severity in the moderate-to-high-risk group. Such association also existed when replacing sitting with walking or MVPA. However, only the substitution of static behavior (sleep or sit) with MVPA supported a similar but even more pronounced correlation in the very-high-risk group (sleep-to-MVPA: OR: 0.82, 95% CI: 0.74–0.90; sit-to-MVPA: OR:0.84, 95% CI: 0.78–0.92) (Table S2, Fig. S2).

In stratified analysis, we separated the whole study population in accordance with working status. We observed a beneficial association between MVPA and CKD prevalence in both populations in single and partition models. Among the working population, MVPA was further categorized into LTPA and OPA. When 1 h/day of sleep was replaced, LTPA substitution displayed the most remarkable associations, with a 12% decrease in overall CKD prevalence (OR: 0.88; 95% CI: 0.83–0.94), compared with the 5%–8% reduction when sleep was replaced with other behavior. By contrast, replacing sitting and LTPA with OPA was associated with a higher CKD prevalence (Tab.2B). All these associations were similarly shown in abnormal ACR but were insignificant in decreased eGFR (Table S3, Fig. S3). Similarly, among those currently unemployed, substituting sleeping with LTPA was associated with the most pronounced association (OR: 0.90, 95% CI: 0.87–0.94) compared with replacing sleeping with other behavior, and any reallocation to a more active activity always showed a beneficial association (Tab.2C). For overall CKD and abnormal ACR, the magnitude of favorable association of LTPA replacement was more prominent in those unemployed than those who worked (Table S4, Fig. S4).

When stratified by sleeping duration, the patterns of association differed between long and short sleepers. For long sleepers (those with sleep duration ≥ 8 h, n = 68 439), replacing sleep with other behavior was associated with a comparable reduction in CKD prevalence (sit: OR: 0.92, 95% CI: 0.89–0.94; walk: OR: 0.89, 95% CI: 0.86–0.92; MVPA: OR: 0.91, 95% CI: 0.88–0.93). Conversely, in short sleepers (sleep duration < 8 h, n = 35 599), such reallocations were associated with an elevated CKD prevalence, and only sit-to-MVPA replacement was protective (Fig.2). No apparent association was observed during the mutual substitution between other behaviors. In further analysis, the association on abnormal ACR was generally coincident with that on overall CKD in long and short sleepers. As for the risk of decreased eGFR, any reallocation to a more active activity was associated with a lower prevalence in long sleepers, except for sleep-to-sit replacement. In short sleepers, only reallocating sitting to MVPA was related to a beneficial association (Fig. S5). When long sleepers were further classified by bedtime, the beneficial association was significant only in early sleepers (before 23:00, n = 47 494); in late sleepers (after 23:00, n = 14 794), sleep reallocation was no longer favorable. A protective relationship was also observed when replacing sitting or walking with MVPA in late sleepers, but not in early sleepers (Fig.2).

When further stratified by glucose condition, participants with diabetes exhibited a stronger inverse association with overall CKD prevalence by replacing sleeping with other behavior compared with the corresponding association among those with prediabetes. For these participants with abnormal glucose regulation, the most substantial reduction in CKD prevalence was identified when sleep was replaced with walking (prediabetes: OR: 0.91, 95% CI: 0.86–0.95; diabetes: OR: 0.87, 95% CI: 0.83–0.92). Moreover, when sitting was substituted with walking or MVPA, the modest but apparent reduction in CKD prevalence was found among diabetic long sleepers, but not among nondiabetic ones (Fig.3).

4 Discussion

In line with previous studies, our findings confirmed a U-shaped relationship between sleep duration and CKD prevalence and emphasized the potential disadvantage of long sleep duration (≥ 8 h). Possible mechanisms in support of such adverse associations encompassed socioeconomic and biological factors, including low employment, depression [21,22], sleep fragmentation, circadian rhythm disturbances [23], physical inactivity [24,25], and dysregulated BP [26–28]. However, how to adjust behavioral composition specifically to obtain renal benefits remains unclear; studies that examined the associations of activities with CKD mainly focused on the beneficial associations with eGFR for replacing sedentary time with equivalent physical activities. Therefore, our study, for the first time, applied ISM analysis to investigate the effect of replacing excessive sleep on CKD, finding that exchanging 1 h unit of time spent on excessive sleep with sitting, walking, or MVPA was beneficially associated with a lower prevalence of CKD. Our data added to broader evidence on behavioral mode improvement by performing a more comprehensive analysis of compositional daily activities.

As for physical activities of different intensities, we found that any shifts toward a more active way of activity—not only sit-to-MVPA reallocation, but also sit-to-walk and walk-to-MVPA, were associated with eGFR preservation. The most significant association was realized through MVPA replacement. Such results provided support to promote a more active and intense behavioral pattern in those with excessive sleep to optimize renal function. Unexpectedly, we found that a similar yet stronger benefit was achieved when excessive sleeping time was reallocated to walking instead of MVPA for albuminuria. A possible explanation is that exercise can induce a transient increase in urinary albumin excretion, which makes the association between regular physical activity and albuminuria difficult to study, suggesting that patients with albuminuria should be cautioned against extremely high-intensity exercise.

Another novel finding is that the greatest protective associations were observed when sleep was replaced with LTPA rather than OPA in the working population. When sitting or LTPA was displaced, OPA was associated with a higher CKD prevalence, indicating that reallocation to physical activity was probably favorable only in leisure time, but not in working time. Evidence on the relationship between domain-specific physical activity and CKD is scarce. Our results are in support of the “physical activity paradox,” which advocates that LTPA promotes health, whereas OPA impairs it. Cardiovascular outcomes have demonstrated that LTPA or non-OPA was inversely associated with all-cause mortality [29,30], CVD mortality [31], BP [32], and cardiovascular events [30,33,34], whereas a weaker or adverse effect was observed in OPA [35]. We supported the application of PA paradox to CKD. Several hypotheses have been proposed for potential underlying mechanisms for the paradox; e.g., OPA has an extremely low intensity, excessively long duration, and insufficient recovery time for maintaining or improving cardiorespiratory fitness [36]. Furthermore, OPA may elevate 24 h heart rate and blood pressure when including heavy lifting (involving excessive anaerobic exercise) or constrained static postures, which is deleterious to renal health. It is also associated with increased levels of inflammation (e.g., C reactive protein [37]). In sum, OPA does not provide benefits for CKD as LTPA does.

Benefits induced by substitution of sleeping were inconsistent across different sleeping modes. Such benefits were mostly focused on those with longer sleep and early sleep onset, similar to a previous study [38] reporting a significantly higher CKD incidence in the early bedtime group. Impaired circadian rhythm may act as an underlying mechanism linking sleep onset time and CKD. Moreover, our middle-aged and elder study populations were more likely to go to bed earlier and sleep longer owing to reduced participation in physical/social activities [39]. A U-shaped relationship was confirmed between sleep duration and CKD in our study and many previous studies, but individuals with minimal sleep (< 6 h) were relatively few in number in our population (only 5%). Our data only indicated a significantly beneficial association of reallocating excessive sleep to other domains of behavior and thereby were insufficient to provide reliable evidence on the behavioral adjustment for those short sleepers. These data suggested that reallocating sleeping to sedentary behavior—which was commonly considered a high-risk behavior—was associated with a lower CKD prevalence. A reverse causality relation due to our cross-sectional data could not be ruled out; hence, such results should be interpreted with particular caution. The benefit of sleep-to-sit substitution could be partly explained by several possible mechanisms; one lies in that the accelerometer (e.g., activPAL) used to record daily behavior did not discriminate between time being inactive (sitting or lying) in some studies. In comparison with that for other metabolic abnormalities, for CKD, sleeping is probably a more “inactive” mode of behavior than sitting. The reduced muscle wasting and cardiorespiratory fitness, inflammation, endothelial injury, arterial stiffness, anemia [40], and increased hemodynamic reactivity during bed rest-mediated inactivity could be even more deleterious to CKD. The unique pathophysiology processes, such as sympathetic hyperactivity, hormonal abnormalities, and disrupted renal circadian rhythm during oversleeping, but not sitting, may also contribute to increased CKD risk. According to previous ISM studies that discussed time reallocation between sedentary behavior and sleep, sit-to-sleep substitution was associated with an improved cardiometabolic profile [20], a better quality of life, and a lower mortality risk [41]. Hence, our findings filled the blank regarding kidney health and specially supported that the advice of reducing excessive sleep and increasing active time should be more applied to early long sleepers to improve the renal risk profile.

When glycemic condition was considered, we found that the sleeping time reallocation was always associated with more pronounced effects in participants with diabetes than in those with prediabetes. In addition, diabetic long sleepers benefited more from nearly all behavior substitution toward a more active way. Growing evidence showed that habitual exercise was associated with a lower incidence of CKD [42] and lower odds and mortality risk of DM/CKD [43]. Our results indicated that people with worse glucose conditions might need to substitute static behavior (including sleeping and sitting) with more active activities to lower CKD risk.

Our study has essential strengths, including large-scale nationwide representative sample, unified detection of metabolic metrics, and detailed behavioral information. Most importantly, the application of ISM, which can simulate the reallocation of time between different types of behavior, captured more benefits of changes in behavioral patterns than previous “static” analysis methods. In terms of limitations, first, owing to the cross-sectional design of the study, neither could we analyze the time-varying effects of lifestyle factors nor could we identify incident CKD cases; consequently, the causal relationship between sleep duration and CKD could not be confirmed. Nevertheless, the results remained robust even when we excluded patients with kidney disease history, thus avoiding reverse causation to some extent. Our sensitivity analyses, as well as the difference between LTPA and OPA, proved the health effect attributed to the spontaneous behavioral patterns of individuals, rather than the passive physical frailty caused by CKD. Second, the time spent on daily behavior was self-reported instead of being recorded using objective measuring devices, inducing a potential recall bias. Further studies using bracelets or other wearable devices to monitor behavior objectively are warranted. In consideration of our large sample, an acceptable statistical conclusion would also be drawn. The lack of comprehensive information on sleep complaints, snoring, and other disorders also limited further investigation. Third, ISM, as a statistical hypothesis, would only provide theoretical effects of time substitution. Although we found that the hypothetical replacement of sleep with sitting, walking, or MVPA was associated with 8%–11% reduction in CKD prevalence risk in long sleepers, the interpretation of clinical significance based on such data requires more caution and further exploration by prospective cohort studies.

In conclusion, the results from this nationwide, cross-sectional, population-based study have revealed a U-shaped relationship between sleep duration and CKD prevalence, in which inadequate or excessive sleep duration was associated with a greater prevalence of CKD. For long sleepers in particular, the ISM analysis indicated that replacing excessive sleeping time with equivalent time of either sitting, walking, or MVPA was associated with a lower prevalence risk of overall CKD, and an active leisure-time activity was always favorable. Moreover, a more pronounced association in participants with diabetes was found compared with those with prediabetes. Further prospective studies on sleeping, sitting, and physical activities are warranted to confirm causality.

References

[1] Liyanage T, Toyama T, Hockham C, Ninomiya T, Perkovic V, Woodward M, Fukagawa M, Matsushita K, Praditpornsilpa K, Hooi LS, Iseki K, Lin MY, Stirnadel-Farrant HA, Jha V, Jun M. Prevalence of chronic kidney disease in Asia: a systematic review and analysis. BMJ Glob Health 2022; 7(1): e007525

[2] Kalantar-Zadeh K, Jafar TH, Nitsch D, Neuen BL, Perkovic V. Chronic kidney disease. Lancet 2021; 398(10302): 786–802

[3] Bo Y, Yeoh EK, Guo C, Zhang Z, Tam T, Chan TC, Chang LY, Lao XQ. Sleep and the risk of chronic kidney disease: a cohort study. J Clin Sleep Med 2019; 15(3): 393–400

[4] Hao Q, Xie M, Zhu L, Dou Y, Dai M, Wu Y, Tang X, Wang Q. Association of sleep duration with chronic kidney disease and proteinuria in adults: a systematic review and dose-response meta-analysis. Int Urol Nephrol 2020; 52(7): 1305–1320

[5] Sun H, Qin K, Zou C, Wang HH, Lu C, Chen W, Guo VY. The association of nighttime sleep duration and quality with chronic kidney disease in middle-aged and older Chinese: a cohort study. Sleep Med 2021; 86: 25–31

[6] Hirano K, Komatsu Y, Shimbo T, Nakata H, Kobayashi D. Longitudinal relationship between long sleep duration and future kidney function decline. Clin Kidney J 2022; 15(9): 1763–1769

[7] Li Y, Yang L, Wang H, Jiang H, Qiu G, Liu Y, Xiao Y, Yang H, Wu T, Zhang X. Longer time spent in bed attempting to sleep is associated with rapid renal function decline: the Dongfeng-Tongji cohort study. Ann Med 2018; 50(2): 172–179

[8] Choi H, Kim HC, Lee JY, Lee JM, Choi DP, Suh I. Sleep duration and chronic kidney disease: the Korean Genome and Epidemiology Study (KoGES)-Kangwha study. Korean J Intern Med (Korean Assoc Intern Med) 2017; 32(2): 323–334

[9] Kim CW, Chang Y, Sung E, Yun KE, Jung HS, Ko BJ, Kwon MJ, Hyun YY, Lee KB, Kim H, Shin H, Ryu S. Sleep duration and quality in relation to chronic kidney disease and glomerular hyperfiltration in healthy men and women. PLoS One 2017; 12(4): e0175298

[10] Lin M, Su Q, Wen J, Wei S, Yao J, Huang H, Liang J, Li L, Lin W, Lin L, Lu J, Bi Y, Wang W, Ning G, Chen G. Self-reported sleep duration and daytime napping are associated with renal hyperfiltration in general population. Sleep Breath 2018; 22(1): 223–232

[11] Yu JH, Han K, Kim NH, Yoo HJ, Seo JA, Kim SG, Choi KM, Baik SH, Kim NH. U-shaped association between sleep duration and urinary albumin excretion in Korean adults: 2011–2014 Korea National Health and Nutrition Examination Survey. PLoS One 2018; 13(2): e0192980

[12] Mazidi M, Shekoohi N, Katsiki N, Banach M. Longer sleep duration may negatively affect renal function. Int Urol Nephrol 2021; 53(2): 325–332

[13] Cao Z, Xu C, Zhang P, Wang Y. Associations of sedentary time and physical activity with adverse health conditions: outcome-wide analyses using isotemporal substitution model. EClinicalMedicine 2022; 48: 101424

[14] Lee J, Walker ME, Gabriel KP, Vasan RS, Xanthakis V. Associations of accelerometer-measured physical activity and sedentary time with chronic kidney disease: the Framingham Heart Study. PLoS One 2020; 15(6): e0234825

[15] Kosaki K, Tanahashi K, Matsui M, Akazawa N, Osuka Y, Tanaka K, Dunstan DW, Owen N, Shibata A, Oka K, Maeda S. Sedentary behaviour, physical activity, and renal function in older adults: isotemporal substitution modelling. BMC Nephrol 2020; 21(1): 211

[16] American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010; 33(Suppl 1): S62–S69

[17] Lu J, Bi Y, Wang T, Wang W, Mu Y, Zhao J, Liu C, Chen L, Shi L, Li Q, Wan Q, Wu S, Qin G, Yang T, Yan L, Liu Y, Wang G, Luo Z, Tang X, Chen G, Huo Y, Gao Z, Su Q, Ye Z, Wang Y, Deng H, Yu X, Shen F, Chen L, Zhao L, Dai M, Xu M, Xu Y, Chen Y, Lai S, Ning G. The relationship between insulin-sensitive obesity and cardiovascular diseases in a Chinese population: results of the REACTION study. Int J Cardiol 2014; 172(2): 388–394

[18] Stevens PE, Levin A. Evaluation and management of chronic kidney disease: synopsis of the kidney disease: Improving Global Outcomes 2012 Clinical Practice Guideline. Ann Intern Med 2013; 158(11): 825–830

[19] Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med 1989; 8(5): 551–561

[20] Galmes-Panades AM, Varela-Mato V, Konieczna J, Warnberg J, Martinez-Gonzalez MA, Salas-Salvado J, Corella D, Schroder H, Vioque J, Alonso-Gomez AM, Martinez JA, Serra-Majem L, Estruch R, Tinahones FJ, Lapetra J, Pinto X, Tur JA, Garcia-Rios A, Riquelme-Gallego B, Gaforio JJ, Matia-Martin P, Daimiel L, Mico Perez RM, Vidal J, Vazquez C, Ros E, Garcia-Arellano A, Diaz-Lopez A, Asensio EM, Castaner O, Fiol F, Mira-Castejon LA, Moreno Rodriguez A, Benavente-Marin JC, Abete I, Tomaino L, Casas R, Baron Lopez FJ, Fernandez-Garcia JC, Santos-Lozano JM, Galera A, Mascaro CM, Razquin C, Papandreou C, Portoles O, Perez-Vega KA, Fiol M, Compan-Gabucio L, Vaquero-Luna J, Ruiz-Canela M, Becerra-Tomas N, Fito M, Romaguera D. Isotemporal substitution of inactive time with physical activity and time in bed: cross-sectional associations with cardiometabolic health in the PREDIMED-Plus study. Int J Behav Nutr Phys Act 2019; 16(1): 137

[21] Monk TH, Pfoff MK, Zarotney JR. Depression in the spousally bereaved elderly: correlations with subjective sleep measures. Depress Res Treat 2013; 2013: 409538

[22] Zhai L, Zhang H, Zhang D. Sleep duration and depression among adults: a meta-analysis of prospective studies. Depress Anxiety 2015; 32(9): 664–670

[23] Lane JM, Vlasac I, Anderson SG, Kyle SD, Dixon WG, Bechtold DA, Gill S, Little MA, Luik A, Loudon A, Emsley R, Scheer FA, Lawlor DA, Redline S, Ray DW, Rutter MK, Saxena R. Genome-wide association analysis identifies novel loci for chronotype in 100 420 individuals from the UK Biobank. Nat Commun 2016; 7(1): 10889

[24] Agarwal R, Light RP. Sleep and activity in chronic kidney disease: a longitudinal study. Clin J Am Soc Nephrol 2011; 6(6): 1258–1265

[25] Jhee JH, Joo YS, Han SH, Yoo TH, Kang SW, Park JT. High muscle-to-fat ratio is associated with lower risk of chronic kidney disease development. J Cachexia Sarcopenia Muscle 2020; 11(3): 726–734

[26] Abdalla M, Sakhuja S, Akinyelure OP, Thomas SJ, Schwartz JE, Lewis CE, Shikany JM, Lloyd-Jones D, Booth JN 3rd, Shimbo D, Hall MH, Muntner P. The association of actigraphy-assessed sleep duration with sleep blood pressure, nocturnal hypertension, and nondipping blood pressure: the coronary artery risk development in young adults (CARDIA) study. J Hypertens 2021; 39(12): 2478–2487

[27] Nagai M, Hoshide S, Ishikawa J, Shimada K, Kario K. Visit-to-visit blood pressure variations: new independent determinants for carotid artery measures in the elderly at high risk of cardiovascular disease. J Am Soc Hypertens 2011; 5(3): 184–192

[28] Johansson JK, Kronholm E, Jula AM. Variability in home-measured blood pressure and heart rate: associations with self-reported insomnia and sleep duration. J Hypertens 2011; 29(10): 1897–1905

[29] Pearce M, Strain T, Wijndaele K, Sharp SJ, Mok A, Brage S. Is occupational physical activity associated with mortality in UK Biobank? Int J Behav Nutr Phys Act 2021; 18(1): 102 doi:10.1186/s12966-021-01154-3

[30] Holtermann A, Schnohr P, Nordestgaard BG, Marott JL. The physical activity paradox in cardiovascular disease and all-cause mortality: the contemporary Copenhagen General Population Study with 104 046 adults. Eur Heart J 2021; 42(15): 1499–1511

[31] Cillekens B, Huysmans MA, Holtermann A, van Mechelen W, Straker L, Krause N, van der Beek AJ, Coenen P. Physical activity at work may not be health enhancing. A systematic review with meta-analysis on the association between occupational physical activity and cardiovascular disease mortality covering 23 studies with 655 892 participants. Scand J Work Environ Health 2022; 48(2): 86–98

[32] Byambasukh O, Snieder H, Corpeleijn E. Relation between leisure time, commuting, and occupational physical activity with blood pressure in 125 402 adults: the lifelines cohort. J Am Heart Assoc 2020; 9(4): e014313

[33] Clays E, De Bacquer D, Janssens H, De Clercq B, Casini A, Braeckman L, Kittel F, De Backer G, Holtermann A. The association between leisure time physical activity and coronary heart disease among men with different physical work demands: a prospective cohort study. Eur J Epidemiol 2013; 28(3): 241–247

[34] Wennberg P, Lindahl B, Hallmans G, Messner T, Weinehall L, Johansson L, Boman K, Jansson JH. The effects of commuting activity and occupational and leisure time physical activity on risk of myocardial infarction. Eur J Cardiovasc Prev Rehabil 2006; 13(6): 924–930

[35] Honda T, Kuwahara K, Nakagawa T, Yamamoto S, Hayashi T, Mizoue T. Leisure-time, occupational, and commuting physical activity and risk of type 2 diabetes in Japanese workers: a cohort study. BMC Public Health 2015; 15(1): 1004

[36] Ketels M, Rasmussen CL, Korshoj M, Gupta N, De Bacquer D, Holtermann A, Clays E. The relation between domain-specific physical behaviour and cardiorespiratory fitness: a cross-sectional compositional data analysis on the physical activity health paradox using accelerometer-assessed data. Int J Environ Res Public Health 2020; 17(21): 7929

[37] Lee J, Kim HR, Jang TW, Lee DW, Lee YM, Kang MY. Occupational physical activity, not leisure-time physical activity, is associated with increased high-sensitivity C reactive protein levels. Occup Environ Med 2021; 78(2): 86–91

[38] Fang Y, Son S, Yang J, Oh S, Jo SK, Cho W, Kim MG. Perturbation of circadian rhythm is associated with increased prevalence of chronic kidney disease: results of the Korean Nationwide Population-Based Survey. Int J Environ Res Public Health 2022; 19(9): 5732

[39] Yang L, Yang H, He M, Pan A, Li X, Min X, Zhang C, Xu C, Zhu X, Yuan J, Wei S, Miao X, Hu FB, Wu T, Zhang X. Longer sleep duration and midday napping are associated with a higher risk of CHD incidence in middle-aged and older Chinese: the Dongfeng-Tongji Cohort Study. Sleep 2016; 39(3): 645–652

[40] Zelle DM, Klaassen G, van Adrichem E, Bakker SJ, Corpeleijn E, Navis G. Physical inactivity: a risk factor and target for intervention in renal care. Nat Rev Nephrol 2017; 13(5): 318

[41] Stamatakis E, Rogers K, Ding D, Berrigan D, Chau J, Hamer M, Bauman A. All-cause mortality effects of replacing sedentary time with physical activity and sleeping using an isotemporal substitution model: a prospective study of 201 129 mid-aged and older adults. Int J Behav Nutr Phys Act 2015; 12(1): 121

[42] Michishita R, Matsuda T, Kawakami S, Tanaka S, Kiyonaga A, Tanaka H, Morito N, Higaki Y. The joint impact of habitual exercise and glycemic control on the incidence of chronic kidney disease (CKD) in middle-aged and older males. Environ Health Prev Med 2017; 22(1): 76

[43] Wang IK, Tsai MK, Liang CC, Yen TH, Huang CC, Wen SF, Wen CP. The role of physical activity in chronic kidney disease in the presence of diabetes mellitus: a prospective cohort study. Am J Nephrol 2013; 38(6): 509–516

RIGHTS & PERMISSIONS

Higher Education Press