Abstract
Objective
Whether or not combined lifestyle factors are associated with similar decreases in risks of incident hypertension and diabetes among individuals with and without chronic kidney disease (CKD) remains unclear.
Methods
This population-based prospective cohort study included participants 40-74 years old who were free from heart disease, stroke, renal failure, hypertension, diabetes, and hypercholesterolemia at baseline (n=60,234). Healthy lifestyle scores (HLSs) were calculated by adding the total number of 5 healthy lifestyle factors (non-smoking, body mass index <25 kg/m2, regular exercise, healthy eating habits, and moderate or less alcohol consumption). Cox proportional hazards models were used to examine associations between the HLS and incident hypertension or type 2 diabetes and whether or not CKD modified these associations.
Results
During a median of 4 years, there were 2,773 incident hypertension cases (30.1 cases per 1,000 person-years) and 263 incident diabetes cases (2.4 cases per 1,000 person-years). The risk of developing hypertension and diabetes decreased linearly as participants adhered to more HLS components. Compared with adhering to 0, 1, or 2 components, adherence to all 5 HLS components was associated with a nearly one-half reduction in the risk of hypertension [hazard ratio (HR) =0.52; 95% confidence interval (CI), 0.45-0.60] and diabetes (HR=0.51; 95% CI, 0.32-0.81) in fully adjusted models. CKD did not have a modifying effect on associations between the HLS and incident hypertension (Pinteraction=0.6) or diabetes (Pinteraction=0.3).
Conclusion
Adherence to HLS components was associated with reduced risks of incident hypertension and diabetes, regardless of CKD status.
Keywords: CKD, diabetes mellitus, hypertension, obesity, physical activity
Introduction
Adherence to healthy lifestyle factors is associated with a reduced risk of incident type 2 diabetes and hypertension in the general population (1-6). A systematic review and meta-analysis of prospective cohort studies summarized the relationship between combined lifestyle factors (including, but not limited to, smoking, alcohol drinking, physical activity, diet, and being overweight or obese) and incident type 2 diabetes (1). Compared with participants considered to have the least-healthy lifestyle, those with the healthiest lifestyle had a 75% lower risk of incident diabetes (1). A dose-dependent relationship between combined lifestyle factors and incident hypertension has also been reported (2-6). Furthermore, a population-based cohort study using data from the United Kingdom (UK) Biobank found no significant interaction between lifestyle factors and genetic risk, indicating that adherence to multiple ideal health behaviors effectively prevents diabetes and hypertension, even in individuals with a high genetic risk (6).
It remains unclear, however, whether or not combined lifestyle factors are associated with similar decreases in risks of incident hypertension and diabetes among individuals with and without chronic kidney disease (CKD). This information is clinically relevant because individuals with CKD may be at high risk of developing type 2 diabetes and hypertension. The rate of incident type 2 diabetes among individuals with CKD is markedly higher than in the general population (7,8). The prevalence of hypertension is also high among patients with CKD (ranging from 60-90%, depending on the stage of CKD and its cause) (9). Furthermore, higher blood pressure levels are associated with a higher risk for incident CKD (10) as well as CKD progression (11). Thus, establishing the effect of combined lifestyle factors on preventing hypertension would be of clinical importance for both individuals with and without CKD.
Against this backdrop, the present study aimed to use data from a Japanese nation-wide prospective cohort study to calculate a healthy lifestyle score (HLS) from five lifestyle risk factors that are largely modifiable [smoking, body mass index (BMI), physical activity, eating habits, and alcohol consumption] (12-15) and to examine the prospective association between the HLS and incident hypertension and diabetes. Our goals were to (i) evaluate the association between the HLS and incident hypertension and diabetes in the total cohort (including participants with and without CKD), (ii) examine whether the associations varied by the presence of CKD, and (iii) estimate the population attributable fractions (PAFs) of HLS for incident hypertension and diabetes. As a secondary objective, the present study aimed to explore the independent association between each HLS component and incident hypertension and diabetes.
Materials and Methods
Data source and study design
This is a longitudinal study of individuals who participated in the nationwide Specific Health Checkup program in Japan between 2008 and 2014 [the Japan Specific Health Checkups (J-SHC) study] (16-19). Details of this cohort study and the program have been published previously (20), and more recent data have been added since that report (16-19). In brief, the Japanese government initiated the program in 2008 to support the early diagnosis of and intervention for metabolic syndrome for all insured persons and their dependents 40-74 years old throughout Japan. The databases used in this study were based in Fukushima, Ibaraki, Niigata, Osaka, Fukuoka, Miyazaki, and Okinawa Prefectures. Participants in the program answer a self-administered questionnaire that covers their medical history, smoking habits, alcohol intake, exercise habits, and eating habits. Trained staff members measure the height, weight, and blood pressure of each participant. Serum and spot urine samples are collected to measure chemical data.
This study was conducted according to the principles of the Declaration of Helsinki, as well as the Ethical Guidelines for Medical and Health Research Involving Human Subjects published by the Ministry of Education, Science, and Culture and the Ministry of Health, Labour and Welfare in 2015. The study protocol was approved by the Ethics Committee of Fukushima Medical University. The need for informed consent was waived due to the use of de-identified information.
Study population
Participants who underwent the Specific Health Checkup in 2008 with available lifestyle factors; who were free from heart disease, stroke, renal failure, hypertension (blood pressure ≥140/90 mmHg or the use of antihypertensive drugs), diabetes [hemoglobin A1c (HbA1c) ≥6.5%, fasting blood sugar level ≥126 mg/dL, or the use of anti-hyperglycemic drugs], and hypercholesterolemia [low-density lipoprotein (LDL) cholesterol level ≥140 mg/dL or the use of cholesterol-lowering medication]; and who underwent the Specific Health Checkup at least once between 2009 and 2014 were eligible. Exclusion criteria were participants with missing information on baseline serum creatinine and urinalysis. Participants with a prevalent prediabetic state (HbA1c >5.6% and/or fasting blood sugar ≥110 mg/dL) or prehypertension (blood pressure ≥120/80 mmHg) were also excluded per outcome.
HLS
We assigned scores for individual HLS components as previously described (Table 1) (12-15). The HLS was calculated as the sum of the score from each component and ranged from 0 (least healthy) to 5 (healthiest), in accordance with previous studies (12-15,21-31). The HLS was the primary exposure, and its individual components were secondary exposures.
Table 1.
Components and Definitions of Healthy Lifestyle Score (HLS).
| Component | Unhealthy (0 point) | Healthy (1 point) |
|---|---|---|
| Smoking | Current smoker | Never or ex-smoker |
| Body mass index | ≥25 kg/m2 | <25 kg/m2 |
| Regular exercise | ||
| Two questions were posed: | ||
| Are you in the habit of exercising to light sweat for more than 30 minutes at a time, 2 times weekly, for over a year? | ’No’ to one or two questions | ’Yes’ to both questions |
| In your daily life, do you walk or do any equivalent amount of physical activity for longer than one hour a day? | ||
| Eating habits | ||
| Two questions were posed: | ||
| Do you skip breakfast more than 3 times per week? | ’Yes’ to one or two questions | ’No’ to both questions |
| Do you eat snacks after supper more than 3 times a week? | ||
| Alcohol intake | ≥20 g/day | <20 g/day |
HLS is the sum of each component, ranging from 0 (least healthy) to 5 (healthiest).
For smoking, a low risk was defined as not currently smoking (12-15). Optimal body weight was defined as a BMI <25 kg/m2, the standard World Health Organization cut-off for healthy weight, in agreement with previous studies (12-15,21-23). For exercise habits, those who answered ‘Yes’ to both questions were considered to be at low risk (12-15) based on current Japanese guidelines (32). Healthy eating habits were defined as eating breakfast and not eating snacks after dinner because our dataset did not include specific nutrition or food data. Those who answered ‘No’ to both skipping breakfast and eating snacks after dinner were considered to be at low risk (12-15). For alcohol consumption, low risk was defined as an average daily alcohol consumption <20 g (12-15).
Baseline covariates
Estimated glomerular filtration rate (eGFR) was calculated using the eGFR formula for Japanese individuals (33). Proteinuria was defined as a dipstick urinalysis score of 1+ or greater (equivalent to ≥30 mg/dL). CKD was defined as the presence of proteinuria, an eGFR <60 mL/min/1.73 m2, or both at baseline (34). Absence of proteinuria with an eGFR ≥60 mL/min/1.73 m2 was defined as “non-CKD.”
Outcome measurement
Study outcomes were incident hypertension or diabetes at the annual medical checkup program during the follow-up period (2009-2014). New-onset hypertension was defined as blood pressure ≥140/90 mmHg or the use of antihypertensive medication. New-onset diabetes was defined as HbA1c ≥6.5%, fasting blood sugar level ≥126 mg/dL, or the use of anti-hyperglycemic drugs (35). The date of participation in the Specific Health Checkup in 2008 was used as the index date. The date of the annual medical checkup program when fulfilling the above criteria was defined as the date of onset.
Statistical analyses
We first examined the association between the HLS and incident hypertension or diabetes, adjusted for CKD status, in the total cohort (including participants with and without CKD). Baseline HLSs of 0, 1, and 2 were combined into a single category due to the limited number of cases. Associations of the baseline HLS (primary exposure) and its individual components (secondary exposures) with incident events of hypertension or diabetes were assessed in univariable and multivariable Cox proportional hazards regression analyses. The assumption of proportional hazards was checked using Schoenfeld residuals and evaluated graphically and found to not be in violation. All models were analyzed unadjusted and adjusted for potential covariates identified a priori. The initial model tested the main effect of HLS as a categorical variable on incident events of hypertension or diabetes, after adjusting for age and sex. The second, fully adjusted model added terms for systolic blood pressure, HbA1c, LDL cholesterol, and CKD status.
Several analyses were performed to test the robustness of the data. First, the analysis was repeated after excluding early events (<2 years) to rule out reverse causality. Second, the analysis was repeated after adjusting for the eGFR and proteinuria, rather than the CKD status. Third, the analysis was repeated after adjusting for fasting blood sugar levels, rather than HbA1c. Fourth, the analysis was repeated using the average HLS during the study period, instead of the baseline HLS. The HLS was calculated for each visit, and the visit-specific HLS was then averaged across all non-missing visits to yield the average HLS. Stratified analyses were also conducted to test the consistency of results by sex and age (≤60 and >60 years).
Associations stratified by the CKD status were then examined to determine whether or not associations between the HLS and outcomes varied by the presence of CKD. Interactions were tested by including a cross-product term along with the main effect terms in the models. To investigate the combined impact of the HLS and CKD status on outcomes, multivariable-adjusted hazard ratios (HRs) and 95% confidence interval (CIs) of participants with and without CKD with a higher HLS were estimated, as well as for participants with CKD with a lower HLS, and compared to those without CKD and with the lowest HLS as the reference category.
Finally, adjusted PAFs and 95% CIs were calculated to estimate the proportion of incident cases that would not have occurred if all participants had been in the healthiest lifestyle group, assuming that the observed associations represent causal effects. All reported p values were two-sided, and values <0.05 were considered statistically significant except for tests of interaction. All statistical analyses were performed using Stata 16 (StataCorp., College Station, USA).
Results
Participant flow and characteristics
Of the 423,930 individuals 40-74 years old with available lifestyle factors who participated in the Specific Health Checkup program in 2008, 74,729 were free from heart disease, stroke, renal failure, hypertension, diabetes, and hypercholesterolemia at baseline (Fig. 1). Among these participants, 60,234 with baseline serum creatinine and urinalysis data were included in the analyses. Characteristics of participants with and without baseline data on the kidney function were similar (Supplementary material 1). Participants with advanced CKD stage were more likely to be excluded due to the high prevalence of comorbidities (Supplementary material 2). Participants with prevalent prehypertension or prediabetic status were excluded per outcome, leaving 26,457 participants for the analysis for hypertension and 31,039 for diabetes.
Figure 1.
Flowchart for the selection of the analyzed study sample.
Hypertension
Among the 26,457 participants who were free from prehypertension at baseline, those with a higher HLS at baseline were more likely to be older, have CKD, and less likely to be male (Table 2). Among the included participants, 2,975 (11.2%) had CKD, of which the majority (79.5%) had an eGFR in the range of 45-59 mL/min/1.73 m2 (Supplementary material 3). During a median follow-up of 4 years (25th percentile: 2; 75th percentile: 5), we identified 2,773 incident hypertension cases (30.1 cases per 1,000 person-years). Both the HLS and CKD were significantly associated with incident hypertension (Table 3). In unadjusted and adjusted models, a higher HLS was monotonically associated with a lower risk of incident hypertension. Compared with the lowest category of HLS (0-2), participants in the healthiest group exhibited a 48% relative reduction in the risk of developing hypertension (HR: 0.52; 95% CI: 0.45-0.60) in the fully adjusted model. The PAF was 22.4% (95% CI: 15.7-28.6%), suggesting that nearly one-fifth of hypertension cases in this cohort might have been prevented if all participants had been in the low-risk group.
Table 2.
Baseline Characteristics of 26,457 Participants Included in the Analysis of Hypertension According to Healthy Lifestyle Score.
| Characteristics | Healthy lifestyle score at baseline | p for trend | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 [n=27 (0.1%)] |
1 [n=555 (2.1%)] |
2 [n=2,338 (8.8%)] |
3 [n=6,531 (24.7%)] |
4 [n=12,449 (47.1%)] |
5 [n=4,557 (17.2%)] |
|||||||||
| Age, years | 52±9 | 52±9 | 55±10 | 57±10 | 59±9 | 63±8 | <0.001 | |||||||
| Males (%) | 74.1 | 70.6 | 60.8 | 41.6 | 24.6 | 27.3 | <0.001 | |||||||
| Current smoker (%) | 100.0 | 92.1 | 67.8 | 25.5 | 3.0 | 0.0 | <0.001 | |||||||
| Body mass index, kg/m2 | 26.3±0.9 | 23.4±3.6 | 22.7±3.4 | 22.0±3.1 | 20.9±2.4 | 20.9±2.1 | <0.001 | |||||||
| Alcohol >20 g/day (%) | 100.0 | 75.0 | 43.4 | 18.0 | 3.9 | 0.0 | <0.001 | |||||||
| Exercise habit | ||||||||||||||
| Exercise to light sweat (%) | 25.9 | 13.0 | 17.6 | 22.1 | 29.2 | 100.0 | <0.001 | |||||||
| Walking >1 h per day (%) | 29.6 | 28.6 | 35.8 | 38.3 | 43.6 | 100.0 | <0.001 | |||||||
| Eating habit | ||||||||||||||
| Snacks after supper (%) | 48.1 | 43.8 | 37.8 | 34.7 | 5.5 | 0.0 | <0.001 | |||||||
| Skipping breakfast (%) | 70.4 | 66.3 | 37.5 | 18.8 | 2.4 | 0.0 | <0.001 | |||||||
| Systolic BP, mmHg | 110±7 | 109±7 | 108±8 | 107±8 | 107±8 | 108±8 | <0.001 | |||||||
| Diastolic BP, mmHg | 66±7 | 68±7 | 67±7 | 66±7 | 65±7 | 66±7 | <0.001 | |||||||
| Fasting blood sugar, mg/dL | 94±10 | 92±10 | 91±10 | 90±9 | 89±9 | 90±9 | <0.001 | |||||||
| Hemoglobin A1c, % | 5.4±0.4 | 5.4±0.3 | 5.5±0.3 | 5.5±0.3 | 5.5±0.3 | 5.6±0.3 | <0.001 | |||||||
| LDL cholesterol, mg/dL | 109±22 | 104±22 | 109±21 | 110±19 | 111±19 | 112±18 | <0.001 | |||||||
| CKD (%) | 7.4 | 6.8 | 10.1 | 9.8 | 11.2 | 14.6 | <0.001 | |||||||
| Stage 1-2 (%) | 3.7 | 3.2 | 3.3 | 1.9 | 1.8 | 1.4 | <0.001 | |||||||
| G3a (%) | 3.7 | 3.6 | 6.5 | 7.7 | 9.0 | 12.6 | ||||||||
| G3b-5 (%) | 0.0 | 0.0 | 0.3 | 0.2 | 0.4 | 0.7 | ||||||||
| Proteinuria (%) | 3.7 | 3.4 | 3.8 | 2.2 | 2.1 | 1.6 | <0.001 | |||||||
| Serum Cr, mg/dL | 0.78±0.16 | 0.73±0.14 | 0.73±0.16 | 0.69±0.16 | 0.67±0.16 | 0.70±0.29 | <0.001 | |||||||
| eGFR, mL/min/1.73 m2 | 79±15 | 84±15 | 81±15 | 80±16 | 78±16 | 74±14 | <0.001 | |||||||
Data are presented as mean±standard deviation or percentage.
BP: blood pressure, CKD: chronic kidney disease, Cr: creatinine, eGFR: estimated glomerular filtration rate, LDL: low-density lipoprotein
Table 3.
Number, Crude Incidence Rate, and Hazard Ratio (95% CI) of Incident Hypertension According to Healthy Lifestyle Score.
| No. of incident hypertension | Rate (95% CI)a | HR (95% CI) | |||
|---|---|---|---|---|---|
| Unadjusted | Age/sex adjusted | Multivariable adjustedb | |||
| HLS | |||||
| 0-2 | 386 | 38.6 (34.9-42.7) | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| 3 | 743 | 33.1 (30.8-35.6) | 0.86 (0.76-0.97) | 0.78 (0.69-0.88) | 0.80 (0.70-0.90) |
| 4 | 1,204 | 27.5 (26.0-29.1) | 0.71 (0.63-0.80) | 0.59 (0.53-0.67) | 0.62 (0.54-0.69) |
| 5 | 440 | 27.7 (25.2-30.4) | 0.72 (0.62-0.82) | 0.50 (0.44-0.58) | 0.52 (0.45-0.60) |
| HR for trendc | 0.88 (0.84-0.92) | 0.79 (0.75-0.83) | 0.80 (0.76-0.84) | ||
| CKD status | |||||
| Non-CKD | 2,345 | 28.5 (27.4-29.7) | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| CKD | 428 | 43.5 (39.6-47.8) | 1.56 (1.41-1.73) | 1.37 (1.27-1.48) | 1.29 (1.17-1.44) |
aper 1,000 person-years.
bAdjusted for sex, age, systolic blood pressure, hemoglobin A1c, LDL cholesterol, and CKD status.
cHR for trend was calculated by entering the exposure categories as a continuous term in the Cox model.
CI: confidence interval, HLS: healthy lifestyle score, HR: hazard ratio
Results from the sensitivity analyses were consistent with the primary analyses (Supplementary material 4). Excluding early events (<2 years), adjusting for the eGFR and proteinuria instead of CKD status, and adjusting for fasting blood sugar levels instead of HbA1c produced similar results. The results were also similar when using the average HLS instead of the baseline HLS. The average HLS was highly correlated with the baseline HLS (Supplementary material 5A). The results were also similar when stratified by sex (Pinteraction=0.5; Supplementary material 6) or age (Pinteraction=0.1; Supplementary material 7).
No effect modification was observed by the presence of CKD (Pinteraction=0.6; Supplementary material 8). When all groups were compared to a single combined non-CKD group with unhealthy lifestyle category as the reference, the CKD group with HLSs of 4 and 5 had a significantly lower risk of hypertension than the reference (Fig. 2A).
Figure 2.
Associations of the baseline HLS with incident hypertension (A) and diabetes (B) by CKD status. The least healthy, non-CKD group is the reference group. Adjusted for age, sex, systolic blood pressure, HbA1c, and LDL cholesterol. The HR for trend indicates the change in the HR by changing one lifestyle category in a healthy direction. Error bars indicate 95% CIs. CI: confidence interval, CKD: chronic kidney disease, HLS: healthy lifestyle score, HR: hazard ratio
Each HLS component showed significant inverse associations with hypertension for all five components after taking the remaining lifestyle factors into account (Supplementary material 9). When stratified by CKD, only smoking status was modified by CKD status (Pinteraction=0.02; Fig. 3A). Non-smoking was associated with a significantly reduced risk of incident hypertension (HR: 0.58; 95% CI: 0.45-0.75), with a PAF of 8.0% (95% CI: 5.0-10.8%) in participants with CKD (Supplementary material 10).
Figure 3.
Associations between each HLS component and incident hypertension (A) and diabetes (B) by CKD status. Adjusted for age, sex, systolic blood pressure, HbA1c, LDL cholesterol, and other lifestyle factors. Pinteraction shows the results of fitting an interaction term between each HLS component and CKD (i.e. each HLS component ×CKD). BMI: body mass index, CI: confidence interval, HLS: healthy lifestyle score, HR: hazard ratio. Error bars indicate 95% CIs.
Diabetes
Among the 31,039 included participants who were free from prediabetic status at baseline, those with a higher HLS at baseline were more likely to be older, have CKD, and less likely to be male (Table 4). Among the included participants, 3,552 (11.4%) had CKD, of which the majority (76.9%) had an eGFR in the range of 45-59 mL/min/1.73 m2 (Supplementary material 11). During a median follow-up of 4 years (25th percentile: 2; 75th percentile: 5), we identified 263 incident diabetes cases (2.4 cases per 1,000 person-years). HLS, but not CKD, was significantly associated with incident diabetes (Table 5). In unadjusted and adjusted models, a higher HLS was monotonically associated with a lower risk of incident diabetes. Compared with the lowest category of HLS (0-2), participants in the healthiest group exhibited a 49% relative reduction in the risk of developing diabetes (HR: 0.51; 95% CI: 0.32-0.81) in the fully adjusted model. The PAF was 27.4% (95% CI: 1.3-46.6%).
Table 4.
Baseline Characteristics of 31,039 Participants Included in the Analysis of Diabetes According to Healthy Lifestyle Score.
| Characteristics | Healthy lifestyle score at baseline | p for trend | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 [n=67 (0.2%)] |
1 [n=822 (2.6%)] |
2 [n=3,186 (10.3%)] |
3 [n=7,921 (25.5%)] |
4 [n=13,925 (44.9%)] |
5 [n=5,118 (16.5%)] |
|||||||||
| Age, years | 51±9 | 53±9 | 56±10 | 58±10 | 60±10 | 64±8 | <0.001 | |||||||
| Males (%) | 80.6 | 74.6 | 65.9 | 47.6 | 30.7 | 33.7 | <0.001 | |||||||
| Current smoker (%) | 100.0 | 89.8 | 61.1 | 21.7 | 2.6 | 0.0 | <0.001 | |||||||
| Body mass index, kg/m2 | 26.5±1.4 | 23.3±3.5 | 23.0±3.4 | 22.4±3.1 | 21.2±2.4 | 21.2±2.1 | <0.001 | |||||||
| Alcohol >20 g/day (%) | 100.0 | 82.4 | 54.2 | 26.0 | 6.5 | 0.0 | <0.001 | |||||||
| Regular exercise | ||||||||||||||
| Exercise to light sweat (%) | 14.9 | 12.0 | 17.7 | 24.1 | 32.4 | 100.0 | <0.001 | |||||||
| Walking >1 h per day (%) | 23.9 | 29.3 | 35.6 | 39.4 | 45.5 | 100.0 | <0.001 | |||||||
| Eating habit | ||||||||||||||
| Snacks after supper (%) | 44.8 | 39.9 | 33.1 | 30.2 | 4.9 | 0.0 | <0.001 | |||||||
| Skipping breakfast (%) | 68.7 | 66.3 | 34.4 | 18.0 | 2.6 | 0.0 | <0.001 | |||||||
| Systolic BP, mmHg | 122±10 | 119±12 | 119±12 | 117±12 | 117±13 | 118±12 | 0.04 | |||||||
| Diastolic BP, mmHg | 75±9 | 73±8 | 73±9 | 72±9 | 71±9 | 71±8 | <0.001 | |||||||
| Fasting blood sugar, mg/dL | 94±8 | 91±8 | 90±8 | 89±8 | 88±8 | 88±7 | <0.001 | |||||||
| Hemoglobin A1c, % | 5.3±0.2 | 5.2±0.2 | 5.3±0.2 | 5.3±0.2 | 5.3±0.2 | 5.3±0.2 | <0.001 | |||||||
| LDL cholesterol, mg/dL | 105±22 | 100±23 | 106±21 | 109±20 | 110±19 | 111±19 | <0.001 | |||||||
| CKD (%) | 7.5 | 8.0 | 10.4 | 10.2 | 11.3 | 15.0 | <0.001 | |||||||
| Stage 1-2 (%) | 3.0 | 4.1 | 3.4 | 2.4 | 1.9 | 1.7 | <0.001 | |||||||
| G3a (%) | 4.5 | 3.9 | 6.7 | 7.5 | 8.9 | 12.7 | ||||||||
| G3b-5 (%) | 0.0 | 0.0 | 0.3 | 0.3 | 0.5 | 0.6 | ||||||||
| Proteinuria (%) | 3.0 | 4.7 | 3.9 | 2.7 | 2.2 | 2.1 | <0.001 | |||||||
| Serum Cr, mg/dL | 0.76±0.14 | 0.74±0.14 | 0.74±0.16 | 0.71±0.18 | 0.68±0.19 | 0.70±0.16 | <0.001 | |||||||
| eGFR, mL/min/1.73 m2 | 84±16 | 84±16 | 81±16 | 80±16 | 78±16 | 75±15 | <0.001 | |||||||
Data are presented as mean±standard deviation or percentage.
BP: blood pressure, CKD: chronic kidney disease, Cr: creatinine, eGFR: estimated glomerular filtration rate, LDL: low-density lipoprotein
Table 5.
Number, Crude Incidence Rate, and Hazard Ratio (95% CI) of Incident Diabetes According to Healthy Lifestyle Score.
| No. of incident diabetes | Rate (95% CI)a | HR (95% CI) | |||
|---|---|---|---|---|---|
| Unadjusted | Age/sex adjusted | Multivariable adjustedb | |||
| HLS | |||||
| 0-2 | 52 | 3.6 (2.7-4.7) | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| 3 | 73 | 2.6 (2.1-3.3) | 0.72 (0.51-1.03) | 0.75 (0.52-1.08) | 0.76 (0.53-1.09) |
| 4 | 105 | 2.1 (1.7-2.6) | 0.59 (0.42-0.82) | 0.64 (0.45-0.91) | 0.65 (0.45-0.92) |
| 5 | 33 | 1.8 (1.3-2.6) | 0.52 (0.33-0.80) | 0.50 (0.32-0.79) | 0.51 (0.32-0.81) |
| HR for trendc | 0.80 (0.70-0.91) | 0.80 (0.70-0.92) | 0.81 (0.70-0.93) | ||
| CKD status | |||||
| Non-CKD | 226 | 2.3 (2.0-2.6) | 1 (ref.) | 1 (ref.) | 1 (ref.) |
| CKD | 37 | 3.0 (2.2-4.1) | 1.31 (0.93-1.86) | 1.11 (0.78-1.58) | 1.13 (0.80-1.61) |
aper 1,000 person-years.
bAdjusted for sex, age, systolic blood pressure, hemoglobin A1c, LDL cholesterol, and CKD status.
cHR for trend was calculated by entering the exposure categories as a continuous term in the Cox model.
CI: confidence interval, HLS: healthy lifestyle score, HR: hazard ratio
Results from sensitivity analyses were consistent with the primary analyses (Supplementary material 12). Excluding early events (<2 years), adjusting for the eGFR and proteinuria instead of CKD status, and adjusting for fasting blood sugar levels instead of HbA1c produced similar results. The results were also similar when using the average HLS instead of the baseline HLS. The average HLS was highly correlated with the baseline HLS (Supplementary material 5B). The results were also similar when stratified by sex (Pinteraction=0.9; Supplementary material 13). Age (Pinteraction=0.01; Supplementary material 14) modified the association between the HLS and incident diabetes, with a non-significant association with the HLS in the older age group.
No effect modification was observed by the presence of CKD (Pinteraction=0.3; Supplementary material 15). Notably, the PAF in patients with CKD was 72.1% (95% CI: 11.2-91.2%). When all groups were compared to a single combined non-CKD group with the unhealthy lifestyle category as the reference, the CKD group with the healthiest lifestyle category had a significantly lower risk of diabetes than the reference (Fig. 2B).
When each HLS component was analyzed individually, we found significant inverse associations between the risk of incident diabetes and non-smokers (HR: 0.73; 95% CI: 0.53-0.99), BMI <25 kg/m2 (HR: 0.62; 95% CI: 0.45-0.84), and healthy eating habits (HR: 0.73; 95% CI: 0.55-0.97) (Supplementary material 16). When stratified by CKD, only regular exercise was modified by CKD (Pinteraction<0.001; Fig. 3B). Being physically active was associated with a significantly lower risk of incident diabetes (HR: 0.27; 95% CI: 0.10-0.70), with a PAF of 63.2% (95 % CI: 32.6-79.9%) in patients with CKD (Supplementary material 17).
Discussion
In this large-scale Japanese population-based cohort study, we confirmed the combined impact of healthy lifestyle factors on preventing incident hypertension and diabetes in a cohort that included participants with and without CKD. Our findings suggest the lack of a modifying effect of CKD on associations between the HLS and incident hypertension and diabetes. When each HLS component was analyzed individually, CKD modified the associations between smoking status and incident hypertension, and between regular exercise and incident diabetes. These findings suggest that adhering to an increasing number of HLS components effectively reduces the risk of both incident hypertension and diabetes, regardless of CKD status.
CKD was found to be significantly associated with incident hypertension but not diabetes, suggesting that CKD itself is not a risk factor for developing diabetes. This is in line with the results from a previous cohort study showing that neither of the two indicators of the kidney function and damage (eGFR and urinary albumin-creatinine ratio) was significantly associated with incident type 2 diabetes (8). Since type 2 diabetes and CKD share common lifestyle risk factors (8), the reported high incidence of type 2 diabetes among individuals with CKD (7,8). might be explained by a lower adherence to a healthy lifestyle. This is supported by a previous study showing that the association between moderate-severe CKD and reduced insulin sensitivity was attenuated after adjusting for lifestyle factors and body composition (36).
We also found that CKD modified the association between smoking status and incident hypertension, and the effect was more pronounced among participants with CKD than among those without CKD. Considering the known correlations between the GFR and clearance of nicotine (37), one of the toxic components in tobacco that can cause an acute increase in blood pressure, the toxic effects of smoking may persist longer in patients with CKD than in individuals with a normal kidney function.
CKD also modified the association between regular exercise and incident diabetes. Being physically active was associated with a significantly lower risk of incident diabetes in participants with CKD, but not those without. Our latter finding is inconsistent with a previous systematic review and meta-analysis of cohort studies showing the protective effects of exercise on the development of diabetes (38,39). This may be due to high-risk participants being more likely to engage in regular exercise habits than others. The present study included participants who fulfilled the following criteria: 1) HbA1c ≤5.6%, 2) fasting blood sugar <110 mg/dl, and 3) self-reported nonuse of anti-hyperglycemic drugs. These criteria, however, would encompass those with well-controlled diabetes who are not on medication. Such individuals are more likely to have regular exercise habits but still be at high risk of developing diabetes. For the same reason, a graded association between the HLS and incident diabetes may not be observed in the older group.
Our findings contribute to efforts to tackle the burden of CKD. The major worldwide risk factors for CKD and end-stage kidney disease (ESKD) are diabetes mellitus and hypertension (40). Preventing both diseases would lessen the burden of ESKD. Furthermore, we previously reported that incident proteinuria decreased as the number of HLS components adhered to increased (12-14). A recent cohort study of the Japanese general population also showed that subjects with a greater number of healthy lifestyle factors (noncurrent smoking, BMI <25, and healthy eating habits) showed a lower incidence of trace/positive proteinuria by dipstick test and rapid eGFR decline (eGFR decline ≥20%) than those with fewer factors (41). These findings strongly suggest that adherence to healthy lifestyle factors can reduce the risk of hypertension, diabetes, proteinuria, and rapid eGFR decline. Tackling multiple risk factors, rather than concentrating on one lifestyle factor, should be the cornerstone of efforts to reduce the global burden of CKD.
Strengths of the present study include the large sample size, large number of incident cases, use of a representative study population from throughout Japan, and detailed information on many lifestyle factors. However, this study also has several limitations that should be noted. First, the scores for HLS components, with the exception of BMI, were determined from self-reported questionnaires, raising the possibility of misclassification. Yet, in prospective studies, misclassification is typically considered nondifferential and therefore is expected to result in an underestimation of risk. Second, CKD was defined based on a single measurement of eGFR and proteinuria, which can also lead to misclassification. Furthermore, many participants, especially those with advanced CKD, were excluded due to comorbidities. This may have introduced selection bias. Third, despite adjusting for potential confounding factors, residual confounding is possible. Finally, this study was conducted in Japan, and thus the results might not be generalizable to other populations. However, the dose-dependent effect of adhering to multiple healthy lifestyle factors on incident hypertension and diabetes in the general population has been observed across different populations (1-6).
In conclusion, this large-scale Japanese population-based cohort study revealed that risks of incident hypertension and diabetes decreased as the number of HLS components adhered to increased, irrespective of CKD status. Our results strongly suggest that adherence to healthy lifestyle factors can reduce the development of incident hypertension and diabetes in individuals with and without CKD.
The authors state that they have no Conflict of Interest (COI).
Financial Support
This work was supported by a Health and Labour Sciences Research Grant for “Study on the design of the comprehensive health care system for chronic kidney disease (CKD) based on the individual risk assessment by Specific Health Checkup” from the Ministry of Health, Labour and Welfare of Japan, and a Grant-in-Aid for “Research on Advanced Chronic Kidney Disease (REACH-J), Practical Research Project for Renal Diseases” from the Japan Agency for Medical Research and Development (AMED).
Supplementary Materials
Comparison of baseline characteristics between participants with and without serum creatinine and/or urinalysis data
Prevalence of comorbidities among 423,930 participants according to kidney function
Baseline characteristics of 26,457 participants included in the analysis of hypertension according to CKD status
Sensitivity analysis of adjusted hazard ratio (95% CI) of healthy lifestyle score for incident hypertension
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by sex
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by age
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by CKD status
Hazard ratio and 95% CI for incident hypertension according to each component of healthy lifestyle score
Hazard ratio and 95% CI for incident hypertension according to each component of healthy lifestyle score, stratified by CKD status
Baseline characteristics of 31,039 participants included in the analysis of diabetes according to CKD status
Sensitivity analysis of adjusted hazard ratio (95% CI) of healthy lifestyle score for incident diabetes
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by sex
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by age.
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by CKD status
Hazard ratio and 95% CI for incident diabetes according to each component of healthy lifestyle score
Hazard ratio and 95% CI for incident diabetes according to each component of healthy lifestyle score, stratified by CKD status
Violin pots showing relationship between baseline HLS and average HLS during the study period among study participants included in the analysis of hypertension (A) and diabetes (B)
Acknowledgement
The authors are thankful for contributions from staff members who collected data and instructed participants with metabolic syndrome at screening centers in each Japanese region.
References
- 1. Zhang Y, Pan XF, Chen J, et al. Combined lifestyle factors and risk of incident type 2 diabetes and prognosis among individuals with type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies. Diabetologia 63: 21-33, 2020. [DOI] [PubMed] [Google Scholar]
- 2. Forman JP, Stampfer MJ, Curhan GC. Diet and lifestyle risk factors associated with incident hypertension in women. JAMA 302: 401-411, 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bai G, Zhang J, Zhao C, Wang Y, Qi Y, Zhang B. Adherence to a healthy lifestyle and a DASH-style diet and risk of hypertension in Chinese individuals. Hypertens Res 40: 196-202, 2017. [DOI] [PubMed] [Google Scholar]
- 4. Lelong H, Blacher J, Baudry J, et al. Combination of healthy lifestyle factors on the risk of hypertension in a large cohort of French adults. Nutrients 11: 1687, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Díaz-Gutiérrez J, Ruiz-Estigarribia L, Bes-Rastrollo M, RuizCanela M, Martin-Moreno JM, Martínez-González MA. The role of lifestyle behaviour on the risk of hypertension in the SUN cohort: the hypertension preventive score. Prev Med 123: 171-178, 2019. [DOI] [PubMed] [Google Scholar]
- 6. Said MA, Verweij N, van der Harst P. Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK Biobank study. JAMA Cardiol 3: 693-702, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thornley-Brown D, Wang X, Wright JT Jr, et al. Differing effects of antihypertensive drugs on the incidence of diabetes mellitus among patients with hypertensive kidney disease. Arch Intern Med 166: 797-805, 2006. [DOI] [PubMed] [Google Scholar]
- 8. Jepson C, Hsu JY, Fischer MJ, et al. Incident type 2 diabetes among individuals with CKD: findings from the chronic renal insufficiency cohort (CRIC) study. Am J Kidney Dis 73: 72-81, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Ku E, Lee BJ, Wei J, Weir MR. Hypertension in CKD: core curriculum 2019. Am J Kidney Dis 74: 120-131, 2019. [DOI] [PubMed] [Google Scholar]
- 10. Chang TI, Lim H, Park CH, et al. Associations of systolic blood pressure with incident CKD G3-G5: a cohort study of South Korean adults. Am J Kidney Dis 76: 224-232, 2020. [DOI] [PubMed] [Google Scholar]
- 11. Lee JY, Park JT, Joo YS, et al. Association of blood pressure with the progression of CKD: findings from KNOW-CKD Study. Am J Kidney Dis 78: 236-245, 2021. [DOI] [PubMed] [Google Scholar]
- 12. Wakasugi M, Kazama JJ, Yamamoto S, Kawamura K, Narita I. A combination of healthy lifestyle factors is associated with a decreased incidence of chronic kidney disease: a population-based cohort study. Hypertens Res 36: 328-333, 2013. [DOI] [PubMed] [Google Scholar]
- 13. Wakasugi M, Kazama JJ, Narita I, et al. Association between combined lifestyle factors and non-restorative sleep in Japan: a cross-sectional study based on a Japanese health database. PLoS One 9: e108718, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Wakasugi M, Kazama J, Narita I, et al. Association between overall lifestyle changes and the incidence of proteinuria: a population-based, cohort study. Intern Med 56: 1475-1484, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Wakasugi M, Narita I, Iseki K, et al. The effect of CKD on associations between lifestyle factors and all-cause, cancer, and cardiovascular mortality: a population-based cohort study. Intern Med 60: 2189-2200, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Nishimoto M, Murashima M, Yoshida H, et al. Impact of self-reported walking habit on slower decline in renal function among the general population in a longitudinal study: the Japan specific health checkups (J-SHC) study. J Nephrol 34: 1845-1853, 2021. [DOI] [PubMed] [Google Scholar]
- 17. Kosugi T, Eriguchi M, Yoshida H, et al. Association between chronic kidney disease and new-onset dyslipidemia: the Japan specific health checkups (J-SHC) study. Atherosclerosis 332: 24-32, 2021. [DOI] [PubMed] [Google Scholar]
- 18. Araumi A, Ichikawa K, Konta T, et al. The distribution of eGFR by age in a community-based healthy population: the Japan specific health checkups study (J-SHC study). Clin Exp Nephrol 25: 1303-1310, 2021. [DOI] [PubMed] [Google Scholar]
- 19. Nagai K, Yamagata K, Iseki K, et al. Weight loss reduces the incidence of dipstick proteinuria: a cohort study from the Japanese general population. Clin Exp Nephrol 25: 1329-1335, 2021. [DOI] [PubMed] [Google Scholar]
- 20. Iseki K, Asahi K, Yamagata K, et al. Mortality risk among screened subjects of the specific health check and guidance program in Japan 2008-2012. Clin Exp Nephrol 21: 978-985, 2017. [DOI] [PubMed] [Google Scholar]
- 21. Stampfer MJ, Hu FB, Manson JE, Rimm EB, Willett WC. Primary prevention of coronary heart disease in women through diet and lifestyle. N Engl J Med 343: 16-22, 2000. [DOI] [PubMed] [Google Scholar]
- 22. Chiuve SE, McCullough ML, Sacks FM, Rimm EB. Healthy lifestyle factors in the primary prevention of coronary heart disease among men: benefits among users and nonusers of lipid-lowering and antihypertensive medications. Circulation 114: 160-167, 2006. [DOI] [PubMed] [Google Scholar]
- 23. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med 345: 790-797, 2001. [DOI] [PubMed] [Google Scholar]
- 24. Kurth T, Moore SC, Gaziano JM, et al. Healthy lifestyle and the risk of stroke in women. Arch Intern Med 166: 1403-1409, 2006. [DOI] [PubMed] [Google Scholar]
- 25. Elwood P, Galante J, Pickering J, et al. Healthy lifestyles reduce the incidence of chronic diseases and dementia: evidence from the Caerphilly cohort study. PLoS One 8: e81877, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Chiuve SE, Fung TT, Rexrode KM, et al. Adherence to a low-risk, healthy lifestyle and risk of sudden cardiac death among women. JAMA 306: 62-69, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Platz EA, Willett WC, Colditz GA, Rimm EB, Spiegelman D, Giovannucci E. Proportion of colon cancer risk that might be preventable in a cohort of middle-aged US men. Cancer Causes Control 11: 579-588, 2000. [DOI] [PubMed] [Google Scholar]
- 28. Jiao L, Mitrou PN, Reedy J, et al. A combined healthy lifestyle score and risk of pancreatic cancer in a large cohort study. Arch Intern Med 169: 764-770, 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Sasazuki S, Inoue M, Iwasaki M, et al. Combined impact of five lifestyle factors and subsequent risk of cancer: the Japan Public Health Center study. Prev Med 54: 112-116, 2012. [DOI] [PubMed] [Google Scholar]
- 30. van Dam RM, Li T, Spiegelman D, Franco OH, Hu FB. Combined impact of lifestyle factors on mortality: prospective cohort study in US women. BMJ 337: a1440, 2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Liu K, Daviglus ML, Loria CM, et al. Healthy lifestyle through young adulthood and the presence of low cardiovascular disease risk profile in middle age: the coronary artery risk development in (young) adults (CARDIA) study. Circulation 125: 996-1004, 2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. The Office for Lifestyle-Related Diseases Control GAD, Health Service Bureau, Ministry of Health, Labour and Welfare of Japan. Exercise and physical activity guide for health promotion 2006 - to prevent lifestyle-related diseases - <exercise guide 2006> prepared in August [Internet]. 2006 [cited 2019 Jul 19]. Available from: http://www.nibiohn.go.jp/files/exercise_guide.pdf
- 33. Japanese Society of Nephrology. Essential points from evidence-based clinical practice guidelines for chronic kidney disease 2018. Clin Exp Nephrol 23: 1-15, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Levey AS, de Jong PE, Coresh J, et al. The definition, classification, and prognosis of chronic kidney disease: a KDIGO controversies conference report. Kidney Int 80: 17-28, 2011. [DOI] [PubMed] [Google Scholar]
- 35. Seino Y, Nanjo K, Tajima N, et al. Report of the committee on the classification and diagnostic criteria of diabetes mellitus. J Diabetes Investig 1: 212-228, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. de Boer IH, Zelnick L, Afkarian M, et al. Impaired glucose and insulin homeostasis in moderate-severe CKD. J Am Soc Nephrol 27: 2861-2871, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Molander L, Hansson A, Lunell E, Alainentalo L, Hoffmann M, Larsson R. Pharmacokinetics of nicotine in kidney failure. Clin Pharmacol Ther 68: 250-260, 2000. [DOI] [PubMed] [Google Scholar]
- 38. Aune D, Norat T, Leitzmann M, Tonstad S, Vatten LJ. Physical activity and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis. Eur J Epidemiol 30: 529-542, 2015. [DOI] [PubMed] [Google Scholar]
- 39. Smith AD, Crippa A, Woodcock J, Brage S. Physical activity and incident type 2 diabetes mellitus: a systematic review and dose-response meta-analysis of prospective cohort studies. Diabetologia 59: 2527-2545, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Lees JS, Mark PB. Fruit for thought: lifestyle interventions to reduce the risk of future chronic kidney disease. Nephrol Dial Transplant 36: 963-965, 2021. [DOI] [PubMed] [Google Scholar]
- 41. Okada R, Tsushita K, Wakai K, Kato K, Wada T, Shinohara Y. Healthy lifestyle reduces incidence of trace/positive proteinuria and rapid kidney function decline after 2 years: from the Japan Ningen Dock study. Nephrol Dial Transplant 36: 1039-1048, 2021. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Comparison of baseline characteristics between participants with and without serum creatinine and/or urinalysis data
Prevalence of comorbidities among 423,930 participants according to kidney function
Baseline characteristics of 26,457 participants included in the analysis of hypertension according to CKD status
Sensitivity analysis of adjusted hazard ratio (95% CI) of healthy lifestyle score for incident hypertension
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by sex
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by age
Number, crude incidence rate, and hazard ratio of incident hypertension according to healthy lifestyle score, stratified by CKD status
Hazard ratio and 95% CI for incident hypertension according to each component of healthy lifestyle score
Hazard ratio and 95% CI for incident hypertension according to each component of healthy lifestyle score, stratified by CKD status
Baseline characteristics of 31,039 participants included in the analysis of diabetes according to CKD status
Sensitivity analysis of adjusted hazard ratio (95% CI) of healthy lifestyle score for incident diabetes
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by sex
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by age.
Number, crude incidence rate, and hazard ratio of incident diabetes according to healthy lifestyle score, stratified by CKD status
Hazard ratio and 95% CI for incident diabetes according to each component of healthy lifestyle score
Hazard ratio and 95% CI for incident diabetes according to each component of healthy lifestyle score, stratified by CKD status
Violin pots showing relationship between baseline HLS and average HLS during the study period among study participants included in the analysis of hypertension (A) and diabetes (B)