Table 1 Saturation effect between urinary nitrate and systolic blood pressure
|
β (95%CI) P-value
|
Model 1: Fitting model by standard linear regression
|
-0.15 (-0.27, -0.03)
|
0.013
|
Model 2: Fitting model by two-piecewise linear regression
|
|
Inflection point (mmHg)
|
4.34 (3.67, 5.18)
|
< 4.34
|
-0.84 (-1.28, -0.41)
|
<0.001
|
> 4.34
|
0.00 (-0.15, 0.15)a
|
0.989
|
P for log-likelihood ratio test
|
0.001
|
aThe effect size was -0.001;
The model adjusted for sex, age, race, body mass index, smoking status, drinking status, medical history (coronary heart disease, stroke, diabetes, heart failure), biochemical indexes (glycohemoglobin, total cholesterol, serum creatinine), medication use (antihypertensive medication, cholesterol-lowering medication, glucose-lowering medication).
Table 2 Association between urinary nitrate and diastolic blood pressure
|
β (95%CI)
|
P-value
|
Model 1
|
0.18 (0.07, 0.29)
|
0.003
|
Model 2
|
-0.02 (-0.13, 0.10)
|
0.785
|
Model 3
|
-0.03 (-0.14, 0.08)
|
0.615
|
Model 1 adjusted for no variable; Model 2 adjusted for age, sex and race; Model 3 adjusted for age, sex, race, body mass index, smoking status, drinking status, medical history (coronary heart disease, stroke, diabetes, heart failure), biochemical indexes (glycohemoglobin, total cholesterol, serum creatinine), medication use (antihypertensive medication, cholesterol-lowering medication, glucose-lowering medication).
3.3 Urinary nitrate and the prevalence of cardiovascular diseases
As shown in Additional file 1, angina pectoris occurred in 271 of the 5989 participants at a rate of 4.22% (95% CI: 3.38%, 5.25%); myocardial infarction occurred in 428 participants at a rate of 6.18% (95% CI: 5.31%, 7.18%); CHD occurred in 416 participants at a rate of 6.12% (95% CI: 5.12%, 7.31%); heart failure occurred in 346 participants at a rate of 4.40% (95% CI: 3.70%, 5.24%). A logistic regression model was developed to analyze the relationship between urinary nitrate and the prevalence of CVD (as shown in Table 3). The prevalence of heart failure decreased with increasing urinary nitrate levels. The unadjusted, partially
Table 3 Association of urinary nitrate with the prevalence of cardiovascular diseases
|
Cases
|
OR (95% CI)
|
P-value
|
Angina pectoris
|
271
|
|
|
Model 1
|
|
0.958 (0.908, 1.010)
|
0.117
|
Model 2
|
|
0.980 (0.930, 1.032)
|
0.443
|
Model 3
|
|
0.982 (0.928, 1.040)
|
0.531
|
Myocardial infarction
|
428
|
|
|
Model 1
|
|
0.978 (0.946, 1.012)
|
0.215
|
Model 2
|
|
1.001 (0.968, 1.034)
|
0.969
|
Model 3
|
|
1.008 (0.975, 1.042)
|
0.648
|
Coronary heart disease
|
416
|
|
|
Model 1
|
|
0.976 (0.935, 1.019)
|
0.281
|
Model 2
|
|
1.005 (0.967, 1.044)
|
0.811
|
Model 3
|
|
1.015 (0.981, 1.050)
|
0.395
|
Heart failure
|
346
|
|
|
Model 1
|
|
0.859 (0.810 0.911)
|
<0.001
|
Model 2
|
|
0.885 (0.834, 0.938)
|
<0.001
|
Model 3
|
|
0.914 (0.859, 0.972)
|
0.006
|
Model 1 adjusted for no variable; Model 2 adjusted for age, sex and race; Model 3 adjusted for age, sex, race, body mass index, smoking status, drinking status, systolic blood pressure, diastolic blood pressure, medical history (stroke, diabetes), biochemical indexes (glycohemoglobin, total cholesterol, serum creatinine), medication use (antihypertensive medication, cholesterol-lowering medication, glucose-lowering medication).
adjusted and fully adjusted OR (95% CI) were 0.859 (0.810, 0.911; P<0.001), 0.885 (0.834, 0.938; P<0.001) and 0.914 (0.859, 0.972; P=0.006), respectively. Whether or not adjusting for confounders, a decline in the prevalence of angina pectoris was observed with increasing urinary nitrate levels, but the effect size did not reach statistical significance. The unadjusted, partially adjusted and fully adjusted OR (95% CI) were 0.958 (0.908, 1.010; P=0.117), 0.980 (0.930, 1.032; P=0.443) and 0.982 (0.928, 1.040; P=0.531), respectively. The current results did not support that urinary nitrate was associated with the prevalence of myocardial infarction and CHD. Multivariable adjusted OR (95% CI) were 1.008 (0.975, 1.042; P=0.648) and 1.015 (0.981, 1.050; P=0.395), respectively.
3.4 Urinary nitrate and cardiovascular mortality
After an average follow-up period of 66.97 (95% CI: 64.89, 69.05) months, 184 out of the 5989 participants died from cardiovascular diseases. The weighted cardiovascular mortality rate was 1.69% (95% CI: 1.35%, 2.12%). The Kaplan-Meier curves indicated an association between elevated urinary nitrate and a poorer cardiovascular-related prognosis (Log-rank test, P<0.001, Fig. 3). Cox regression model was applied to further evaluated the relationship between urinary nitrate and cardiovascular mortality (as shown in Table 4). In the hypertensive population, the risk of cardiovascular mortality gradually decreased with increasing urinary nitrate levels. The multivariable-adjusted HR was 0.906 (95% CI: 0.864, 0.951; P<0.001).
Table 4 Association of urinary nitrate with cardiovascular mortality
|
Cases
|
HR (95%CI)
|
P-value
|
Cardiovascular mortality
|
184
|
|
|
Model 1
|
|
0.845 (0.792, 0.902)
|
<0.001
|
Model 2
|
|
0.889 (0.836, 0.945)
|
<0.001
|
Model 3
|
|
0.906 (0.864, 0.951)
|
<0.001
|
Model 1 adjusted for no variable; Model 2 adjusted for age, sex and race; Model 3 adjusted for age, sex, race, body mass index, smoking status, drinking status, systolic blood pressure, diastolic blood pressure, medical history (coronary heart disease, stroke, diabetes, heart failure, cancer), biochemical indexes (glycohemoglobin, total cholesterol, serum creatinine), medication use (antihypertensive medication, cholesterol-lowering medication, glucose-lowering medication).
3.5 Mediating effect analysis
Previous results of this study demonstrated that higher urinary nitrate level was associated with lower baseline SBP, lower prevalence of heart failure, and a decreased risk of cardiovascular mortality. To further clarify whether blood pressure and heart failure acted as mediators of the association between urinary nitrate and cardiovascular mortality, the mediating effect analysis was performed, as shown in Table 5. After multivariable adjustment,
Table 5 Mediating effects between urinary nitrate and cardiovascular mortality
Mediating variable
|
HR (95% CI)
|
P-value
|
Heart failure
|
|
|
Direct effect
|
0.901 (0.844, 0.962)
|
0.002
|
Indirect effect
|
0.995 (0.990, 1.000)
|
0.066
|
Total effect
|
0.897 (0.841, 0.957)
|
0.001
|
Systolic blood pressure
|
|
|
Direct effect
|
0.922 (0.871, 0.975)
|
0.005
|
Indirect effect
|
0.999 (0.997, 1.000)
|
0.102
|
Total effect
|
0.920 (0.870, 0.974)
|
0.004
|
Model adjusted for age, sex, race, body mass index, smoking status, drinking status, systolic blood pressure, diastolic blood pressure, medical history (coronary heart disease, diabetes, stroke, heart failure, cancer), biochemical indexes (glycohemoglobin, total cholesterol, serum creatinine), medication use (antihypertensive medication, cholesterol-lowering medication, glucose-lowering medication). the indirect effect for heart failure was not statistically significant [HR (95% CI), 0.995 (0.990, 1.000); P=0.066]; the indirect effect for SBP was not statistically significant either [HR (95% CI), 0.999 (0.997, 1.000); P=0.102]. Thus, neither SBP nor coexisting heart failure mediated the relationship between urinary nitrate and cardiovascular mortality.
3.6 Subgroup analysis
The results of subgroup analysis between urinary nitrate levels and cardiovascular mortality are shown in Fig. 4. The negative association between urinary nitrate levels and the risk of cardiovascular mortality was influenced by the presence of specific comorbidity at baseline and the history of smoking. In the subgroup with CHD, the HR (95% CI) was 1.019 (0.931, 1.116), while the HR (95% CI) was 0.874 (0.814, 0.939) for non-CHD subgroup (P for interaction=0.015). Similarly, the HR (95% CI) for the subgroup with heart failure and without heart failure were 1.009 (0.906, 1.123) and 0.881 (0.821, 0.945), respectively (P for interaction=0.043). The HR (95% CI) in the smoking subgroup and non-smoking subgroup were 0.948 (0.889, 1.010) and 0.816 (0.722, 0.921), respectively (P for interaction=0.024). It was not proved that the negative association between urinary nitrate and cardiovascular mortality was affected by age, gender, BMI and the presence of diabetes and stroke.