Selection of participants and baseline characteristics
Out of 133,925 individuals who underwent a medical examination in Kanazawa City between 1999 and 2013, 51,938 met the eligibility criteria and were included in the analysis (Supplementary Figure 1). The median follow-up period was 4.0 years.
Table 1 shows the participants’ baseline characteristics. The group with the highest number of participants was the 60–69 years age group (n=19,015). Lower eGFR and higher blood pressure were observed in the older group, and more than half of the participants aged ≥70 years were treated for hypertension. Nearly half of the participants (49%) in the age group of ≥80 years had an eGFR of <60 mL/min/1.73 m2. All groups showed almost the same levels of hemoglobin, BMI, and total cholesterol. Diabetes mellitus, history of coronary artery disease, and history of stroke were more frequent in older people. The proportion of current smokers was the highest in the youngest group (26%).
Table 1. Baseline characteristics of study population
|
|
|
Age (years)
|
|
|
|
|
Variable
|
40 – 49
(n=1,775)
|
50 – 59
(n=4,386)
|
60 – 69
(n=19,015)
|
70 – 79
(n=18,635)
|
80 –
(n=8,127)
|
Overall
(n=51,938)
|
p trend
|
Age (years)
|
45 ± 3
|
57 ± 3
|
65 ± 3
|
75 ± 3
|
85 ± 4
|
70 ± 10
|
<0.01
|
Sex (men)
|
34%
|
32%
|
39%
|
39%
|
33%
|
37%
|
0.71
|
eGFR (mL/min/1.73 m2)
|
86 ± 15
|
79 ± 14
|
75 ± 14
|
69 ± 15
|
61 ± 17
|
72 ± 16
|
<0.01
|
eGFR <60 mL/min/1.73 m2
|
2%
|
6%
|
12%
|
25%
|
49%
|
22%
|
<0.01
|
eGFR <60 mL/min/1.73 m2 and/or proteinuria (≥1+)
|
7%
|
11%
|
17%
|
29%
|
53%
|
26%
|
<0.01
|
Systolic blood pressure (mmHg)
|
118 ± 17
|
127 ± 18
|
132 ± 18
|
136 ± 17
|
139 ± 18
|
134 ± 18
|
<0.01
|
Diastolic blood pressure (mmHg)
|
73 ± 12
|
78 ± 12
|
79 ± 11
|
77 ± 10
|
75 ± 11
|
77 ± 11
|
<0.01
|
Hemoglobin (g/dL)
|
13.5 ± 1.7
|
13.7 ± 1.4
|
13.7 ± 1.3
|
13.2 ± 1.4
|
12.4 ± 1.4
|
13.3 ± 1.4
|
<0.01
|
Body mass index (kg/m2)
|
22.3 ± 3.7
|
22.7 ± 3.4
|
22.9 ± 3.1
|
23.2 ± 3.2
|
22.5 ± 3.4
|
22.9 ± 3.3
|
0.01
|
Total cholesterol (mg/dL)
|
201 ± 34
|
214 ± 36
|
209 ± 34
|
201 ± 33
|
195 ± 33
|
204 ± 34
|
<0.01
|
Proteinuria (≥1+)
|
5%
|
5%
|
6%
|
7%
|
11%
|
7%
|
<0.01
|
Diabetes mellitus
|
2%
|
7%
|
10%
|
13%
|
12%
|
11%
|
<0.01
|
Current smoker
|
26%
|
19%
|
14%
|
10%
|
5%
|
12%
|
<0.01
|
Treatment of hypertension (%)
|
7%
|
22%
|
36%
|
51%
|
60%
|
43%
|
<0.01
|
History of coronary artery disease (%)
|
2%
|
4%
|
9%
|
16%
|
26%
|
13%
|
<0.01
|
History of stroke (%)
|
1%
|
3%
|
5%
|
9%
|
13%
|
7%
|
<0.01
|
Follow-up period (years)
|
2.8 (1.1, 4.0)
|
3.2 (1.8, 4.9)
|
4.0 (2.1, 4.9)
|
4.2 (2.8, 5.0)
|
3.3 (2.0, 4.9)
|
4.0 (2.1, 4.9)
|
<0.01
|
Number of creatinine measurements (n)
|
3 (2, 4)
|
3 (2, 5)
|
4 (3, 6)
|
5 (3, 6)
|
4 (2, 5)
|
4 (3, 6)
|
<0.01
|
Continuous variables are expressed as mean ± standard deviation, or median (25th and 75th percentiles). Categorical variables are expressed as numbers (percentage).
Abbreviation: eGFR, estimated glomerular filtration rate.
Slope of eGFR and relationships between risk factors
The adjusted mean eGFR slope in all participants was -0.39 mL/min/1.73 m2 per year. Significant declines in kidney function were observed in the ≥60 years age groups (Table 2). Most of the known risk factors were associated with a significantly faster decline in eGFR, except for higher total cholesterol, which was related to a slower decline in eGFR (Table 3). Proteinuria, current smoking, and diabetes mellitus were associated with a greater decline in eGFR (-0.58 [95% confidence interval (CI) -0.67, -49], -0.25 [95% CI -0.32, -0.18], and -0.33 [95% CI -0.40, -0.26] mL/min/1.73 m2 per year, respectively). Higher systolic blood pressure (+10 mmHg) and lower hemoglobin (-1 g/dL) were also associated with a greater decline in eGFR (-0.16 [95% CI -0.17, -0.14] and -0.16 [95% CI -0.18, -0.15] mL/min/1.73 m2 per year, respectively).
Table 2. Adjusted mean slope according to age groups
Age (years)
|
Adjusted mean slope*
(mL/min/1.73 m2 per year)
|
p value
|
All participants (n=51,938)
|
-0.39
|
(-0.41, -0.37)
|
<0.01
|
40–49 (n=1,775)
|
0.02
|
(-0.12, 0.16)
|
0.82
|
50–59 (n=4,386)
|
-0.01
|
(-0.08, 0.07)
|
0.86
|
60–69 (n=19,015)
|
-0.16
|
(-0.19, -0.13)
|
<0.01
|
70–79 (n=18,635)
|
-0.53
|
(-0.56, -0.49)
|
<0.01
|
≥80 (n=8,127)
|
-0.87
|
(-0.94, -0.80)
|
<0.01
|
Data are presented as mean and 95% confidence interval.
* Adjusted for sex, systolic blood pressure, diastolic blood pressure, body mass index, proteinuria, hemoglobin, total cholesterol, smoking status, history of coronary disease, history of stroke, and diabetes mellitus.
Table 3. Differences from mean slope according to risk factors
Variables
|
Difference from mean slope*
(mL/min/1.73 m2 per year)
|
p value
|
Men (vs. Women)
|
-0.18
|
(-0.24, -0.13)
|
<0.01
|
Systolic blood pressure (+10 mmHg)
|
-0.16
|
(-0.17, -0.14)
|
<0.01
|
Diastolic blood pressure (+5 mmHg)
|
0.04
|
(0.02, 0.05)
|
<0.01
|
Body mass index (+1)
|
-0.02
|
(-0.02, -0.01)
|
<0.01
|
Proteinuria 1+ (vs. normal/trace)
|
-0.58
|
(-0.67, -0.49)
|
<0.01
|
Hemoglobin (-1 g/dL)
|
-0.16
|
(-0.18, -0.15)
|
<0.01
|
Total cholesterol (+10 mg/dL)
|
0.02
|
(0.01, 0.03)
|
<0.01
|
Current smoking (vs. no current smoking)
|
-0.25
|
(-0.32, -0.18)
|
<0.01
|
History of coronary disease (vs. no history)
|
-0.15
|
(-0.21, -0.08)
|
<0.01
|
History of stroke (vs. no history)
|
-0.10
|
(-0.19, -0.01)
|
0.02
|
Diabetes mellitus (vs. no diabetes)
|
-0.33
|
(-0.40, -0.26)
|
<0.01
|
n=51,938. Data are presented as mean and 95% confidence interval.
* Values are differences from the adjusted mean slope of all participants (n=51,938). Each variable was adjusted for all other variables.
Slope of eGFR according to age groups
The differences in eGFR decline according to age group for each risk factor are shown in Figure 1. Higher systolic blood pressure, proteinuria, and current smoking status were related to a faster decline in eGFR in the older age groups (p trend <0.01, 0.03, and <0.01, respectively). Lower hemoglobin and diabetes mellitus status were related to a significantly greater eGFR decline in each age group except for the youngest age group (age 40–49 years), and no significant trends according to age group were observed (p trend 0.47 and 0.17, respectively).
The representative estimated eGFR slopes according to age groups and risk factors (hypertension, proteinuria, and smoking status) are shown in Figure 2. The participants without any risks (systolic blood pressure 120 mmHg, negative or trace urinary protein, and without current smoking) had almost the same mean eGFR slopes among the age groups. In a stepwise analysis of additional risks, more differences in eGFR slope according to age groups were observed.
For further trend analysis, systolic blood pressure was divided into seven categories (<110, 110–119, 120–129, 130–139, 140–149, 150–159, and ≥160 mmHg). Higher systolic blood pressure was associated with a clearly faster eGFR decline in the ≥60 years age groups. U-shaped associations were not observed in any age group (Supplementary Figure 2).
Similar results were found in sensitivity analyses using systolic and diastolic blood pressure without constant addition according to the treatment status at the time of measurement (Supplementary Figure 3) and in analyses combining the 40–49 years with the 50–59 years (Supplementary Figure 4).