Abstract
Abstract
Objectives
To investigate the correlation between the urinary albumin-to-creatinine ratio (UACR) and adverse cardiovascular outcomes in the Beijing community population.
Design
Prospective cohort study.
Setting
Beijing, China, from May 2014 to December 2021.
Participants
Recruited from a survey conducted as part of an ongoing atherosclerosis cohort study in the communities of Gucheng and Pingguoyuan, Shijingshan District in Beijing, China. Excluded participants who already had a history of stroke or myocardial infarction at baseline. Finally, 3627 eligible participants were included in this analysis.
Exposure
The participants were divided into three groups on the basis of baseline UACR: the normal group (UACR<30 mg/g), the microalbuminuria group (30 mg/g≤UACR<300 mg/g) and the dominant proteinuria group (UACR≥300 mg/g).
Primary and secondary outcome measures
The primary endpoint was a composite endpoint (major adverse cardiovascular event, MACE) of cardiovascular death, first acute myocardial infarction or first stroke, whereas secondary endpoints included cardiovascular death, first acute myocardial infarction, first stroke or all-cause death.
Results
The study included 3627 participants. According to the multivariable Cox model, compared with those in the normal group, the risks of MACE (HR=1.47; 95% CI 1.06 to 2.06; p=0.023), cardiovascular death (HR=3.03; 95% CI 1.56 to 5.88; p=0.001) and all-cause mortality (HR=1.91; 95% CI 1.23 to 2.97; p=0.004) were significantly greater in the microalbuminuria group. The risk of MACE (HR=3.65; 95% CI 2.14 to 6.23; p<0.001), cardiovascular death (HR=7.91; 95% CI 2.92 to 21.43; p<0.001), stroke (HR=2.57; 95% CI 1.30 to 5.08; p=0.007) and all-cause death (HR=3.59; 95% CI 1.63 to 7.89; p=0.001) in the group with dominant proteinuria was significantly greater than that in the normal group. The absolute risk differences (per 1000 person-years) for MACE were 14.86 (95% CI 7.20 to 22.51) in the microalbuminuria group and 64.85 (95% CI 26.76 to 102.94) in the dominant proteinuria group, compared with the normal group (incidence rates: 25.24 and 75.23 vs 10.38, respectively). In populations with a UACR less than 30 mg/g, there was a significant increase in the risk of MACE as the UACR increased (HR=1.02; 95% CI 1.00 to 1.04; p=0.036).
Conclusions
This study indicates that an elevated UACR is a significant risk factor for adverse cardiovascular outcomes within the community population. This association remains consistent in individuals with low-grade albuminuria.
Keywords: Cardiovascular Disease, China, Risk Factors, Prognosis, Mortality
STRENGTHS AND LIMITATIONS OF THIS STUDY.
Cox regression was applied to evaluate the impact of urinary albumin-to-creatinine ratio (UACR) on various adverse cardiovascular outcomes.
Outcomes were ascertained through linkage to national mortality and hospital-discharge registers.
The study was conducted in a single centre, and findings require validation in other populations.
Only baseline UACR was analysed, and changes during follow-up were not evaluated.
Introduction
At present, common cardiovascular risk factors such as hypertension, smoking, diabetes, obesity and dyslipidaemia can explain 90% of the risk of myocardial infarction and stroke,1 but there are still some residual cardiovascular disease risks that these risk factors cannot explain. Chronic kidney disease is known to be an independent risk factor for long-term cardiovascular events and all-cause mortality. Urinary albumin quantification is an important marker of renal injury and can predict long-term prognosis in patients with chronic kidney disease.2 3 The urinary albumin-to-creatinine ratio (UACR) is a reliable method for monitoring urinary protein excretion and can reliably reflect the 24-hour urine protein content. It is fast, simple and accurate and has become a clinical qualitative and quantitative diagnostic indicator for proteinuria, replacing traditional 24-hour urine protein quantification.4
Previous studies have shown that an elevated UACR is associated with increased cardiovascular events and mortality, especially in patients with diabetes, hypertension and cardiovascular history.5,7 Although numerous large-scale studies in Western populations have established the association between UACR and cardiovascular risk (eg, EPIC-Norfolk, MESA studies),8 9 evidence remains limited and inconsistent in East Asian community-based populations, particularly among individuals without known cardiovascular disease.10,12 In addition, in the past decade, low-grade albuminuria (LGA, 24-hour urine protein quantification below 30 mg), which does not meet the diagnostic criteria for microalbuminuria, has received increasing attention, and the predictive value of LGA for the risk of hard endpoint events has been reported.13 14 However, there is controversy over the use of LGA as a risk indicator to predict all-cause mortality in healthy individuals.12 15 There is also little research exploring the predictive value of LGA for myocardial infarction and stroke in East Asian community populations. Therefore, this study aimed to systematically evaluate the relationship between UACR (including within the LGA range) and multiple adverse cardiovascular outcomes and all-cause mortality in a cardiovascular disease-free community-based cohort from Beijing, China, with long-term follow-up (median 7.4 years), to provide more population-specific evidence for risk assessment in this group.
Methods
Study population
Participants were recruited from a survey conducted as part of an ongoing atherosclerosis cohort study in the communities of Gucheng and Pingguoyuan, Shijingshan District in Beijing, China, between May 2014 and July 2014. The detailed procedures of this cohort study have been described previously.16 17 We further excluded participants who already had a history of stroke or coronary heart disease at baseline. Finally, 3627 eligible participants were included in this analysis. We adhered to the principles of the Declaration of Helsinki. The procedures followed were in accordance with institutional guidelines. We followed the Strengthening the Reporting of Observational Studies in Epidemiology Statement for cohort studies to ensure comprehensive reporting.18
Data collection
Baseline data were collected by trained research staff according to a standard operating procedure. All participants were interviewed via a standardised questionnaire, including sociodemographic status, education, occupation, diet, lifestyle, health behaviour and medical history. Anthropometric measurements were taken according to a standard operating procedure. Current smoking was defined as smoking one cigarette per day for at least half a year. Current drinking was defined as drinking once per week for at least half a year. Body mass index (BMI) was calculated as weight (kg) divided by height (m) squared.16 After the participants had rested for 5 min, brachial blood pressure was measured via an Omron HEM-7117 electronic sphygmo manometer, and the mean value of three consecutive measurements was recorded for data analysis.
This study was conducted by trained professionals who collected fasting blood from the vein in the anterior elbow area of the arms of the study subjects to measure biochemical indicators. All study subjects fasted for ≥12 hours before fasting blood was collected. All blood samples were centrifuged within 30 min and stored in a −80°C freezer. Fasting blood glucose, total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol and triglyceride (TG) levels were measured via a Hitachi 7180 Automatic Analyser. Serum creatinine (Scr) was measured via the same instrument with Jaffe’s kinetic method. The estimated glomerular filtration rate (eGFR) was determined according to the Modification of Diet in Renal Disease formula corrected for the Chinese population: eGFR (mL∕ min ∕1.73 m2)=175 × Scr (mg/dL)−1.234 × age−0.179× (×0.79 if female sex).19 Baseline random urine samples were collected to detect the UACR via the Beckman IMMAGE 800 Immunochemistry System.
Definitions of diseases and indicators
Hypertension was defined as a self-reported medical history of hypertension and/or the use of antihypertensive drugs and/or an increase in blood pressure on the day of investigation, with an average systolic blood pressure value of ≥140 mm Hg and/or an average diastolic blood pressure value of ≥90 mm Hg measured three times. Diabetes was defined as diabetes with a self-reported medical history and/or the use of hypoglycaemic drugs and/or fasting blood glucose ≥7.0 mmol/L on the day of examination. Dyslipidaemia was defined as a self-reported medical history of dyslipidaemia and/or receiving lipid-lowering treatment and/or having at least one abnormal blood lipid index on the day of investigation, including TC≥5.18 mmol/L, TG≥1.70 mmol/L, LDL-C≥3.37 mmol/L and HDL-C<1.04 mmol/L.
Definitions of outcomes and data sources
The primary endpoint defined in this study was a composite endpoint (major adverse cardiovascular events, MACE) of cardiovascular death, first acute myocardial infarction (AMI, fatal and non-fatal), or first stroke (fatal and non-fatal), whereas secondary endpoint events include cardiovascular death, first AMI, first stroke or all-cause death. Data on participants’ endpoints were collected from the Chinese Center for Disease Control and Prevention (National Mortality Surveillance System) and the Beijing Municipal Health Commission (Inpatient Medical Record Home Page System)20 21 until 31 December 2021. The disease code was based on the International Classification of Diseases in the 10th Revision, as detailed in online supplemental table 1.
Statistical analysis
Data are presented as the mean±SD for continuous variables that were normally distributed and as the median and IQR for continuous variables that were non-normally distributed. The mean values were compared via analysis of variance when the data were normally distributed and via the Kruskal-Wallis test otherwise. Categorical variables are presented as frequencies (proportions) and were compared with the χ² test or Fisher’s exact test, as appropriate. The Kaplan-Meier method was used to estimate the cumulative hazards of MACE stratified by the UACR, and log-rank tests were used to calculate group differences. Multivariate Cox regression adjusted for age, sex, BMI, eGFR, smoking and drinking status, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia was used to determine the associations between the UACR and the risk of adverse cardiovascular outcomes. Furthermore, the dose‒response associations between UACR and MACE were examined with restricted cubic spline in the fully adjusted models, and we analysed the UACR after subjecting it to natural logarithmic transformation (lnUACR). We further examined whether the associations between UACR and the outcomes were modified by the covariates listed above. The proportional hazards assumption for all Cox models was tested using Schoenfeld residuals, and no significant violations were observed (global test p>0.05 for all models). For cause-specific endpoints (cardiovascular death, AMI and stroke), we additionally performed competing risk analyses using Fine-Gray subdistribution hazard models, with non-cardiovascular death treated as a competing event. A two-sided p<0.05 was considered statistically significant. All the statistical analyses were performed via Empower (X&Y Solutions, Boston, Massachusetts) or R V.4.1.1 (http://www.R-project.org).
Patient and public involvement
Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Results
The study included 3627 participants with an average age of 58.7±8.5 years, of whom 1268 (34.96%) were male. The mean follow-up duration was 7.4±0.7 years. Throughout the study, there were 308 (8.49%) instances of MACE, 147 (4.05%) total deaths, 51 (1.41%) cardiovascular deaths, 60 (1.65%) AMI and 239 (6.59%) strokes.
The participants were categorised into three groups on the basis of their UACR: the normal group (UACR <30 mg/g), the microalbuminuria group (30 mg/g≤UACR<300 mg/g) and the dominant proteinuria group (UACR ≥300 mg/g). Intergroup comparisons revealed that with increasing UACR, the participants were older, had higher BMIs, lower eGFRs, had higher proportions of hypertension and diabetes, and used more antihypertensive and hypoglycaemic drugs (all p values <0.001) (table 1).
Table 1. Characteristics of participants presented by different UACR.
| Total (N=3627) | UACR<30 mg/g (N=3344) | 30≤UACR <300 mg/g (N=250) | UACR ≥300 mg/g (N=33) | P value | |
|---|---|---|---|---|---|
| Baseline data | |||||
| Age, year | 58.7±8.5 | 58.5±8.2 | 61.8±10.7 | 61.9±10.4 | <0.001 |
| Male, N (%) | 1268 (34.96%) | 1155 (34.54%) | 96 (38.40%) | 17 (51.52%) | 0.063 |
| UACR, mg/g | 6.09 (4.14–10.68) | 5.70 (4.01–9.12) | 51.22 (37.29–106.86) | 515.30 (376.69–713.93) | <0.001 |
| BMI, kg/m2 | 25.94±3.53 | 25.89±3.51 | 26.45±3.68 | 27.24±3.99 | 0.006 |
| eGFR, mL/min/1.73 m2 | 73.36 (66.70–80.67) | 73.64 (67.13–80.79) | 70.53 (62.32–78.18) | 66.71 (47.65–79.97) | <0.001 |
| Current smoking, N (%) | 582 (16.05%) | 528 (15.79%) | 48 (19.20%) | 6 (18.18%) | 0.346 |
| Current drinking, N (%) | 504 (13.90%) | 447 (13.37%) | 53 (21.20%) | 4 (12.12%) | 0.002 |
| Hypertension, N (%) | 1621 (44.69%) | 1426 (42.64%) | 166 (66.40%) | 29 (87.88%) | <0.001 |
| Diabetes, N (%) | 693 (19.11%) | 591 (17.67%) | 89 (35.60%) | 13 (39.39%) | <0.001 |
| Dyslipidaemia, N (%) | 2745 (75.68%) | 2511 (75.09%) | 207 (82.80%) | 27 (81.82%) | 0.017 |
| Hypertension drug, N (%) | 887 (24.46%) | 779 (23.30%) | 93 (37.20%) | 15 (45.45%) | <0.001 |
| Diabetes drug, N (%) | 307 (8.46%) | 268 (8.01%) | 32 (12.80%) | 7 (21.21%) | <0.001 |
| Dyslipidaemia drug, N (%) | 312 (8.60%) | 283 (8.46%) | 25 (10.00%) | 4 (12.12%) | 0.542 |
| Outcomes | |||||
| MACE, N (%) | 308 (8.49%) | 250 (7.48%) | 43 (17.20%) | 15 (45.45%) | <0.001 |
| Cardiovascular death, N (%) | 51 (1.41%) | 30 (0.90%) | 16 (6.40%) | 5 (15.15%) | <0.001 |
| AMI, N (%) | 60 (1.65%) | 49 (1.47%) | 8 (3.20%) | 3 (9.09%) | <0.001 |
| Stroke, N (%) | 239 (6.59%) | 199 (5.95%) | 31 (12.40%) | 9 (27.27%) | <0.001 |
| All-cause death, N (%) | 147 (4.05%) | 112 (3.35%) | 28 (11.20%) | 7 (21.21%) | <0.001 |
AMI, acute myocardial infarction; BMI, body mass index; eGFR, estimated glomerular filtration rate; MACE, major adverse cardiovascular events; UACR, urinary albumin-to-creatinine ratio.
Absolute risk differences (RD, per 1000 person-years) for adverse outcomes increased progressively with UACR categories (online supplemental table 2). Compared with the normal group, the microalbuminuria group had significant excess risks of MACE (RD=14.86, 95% CI 7.20 to 22.51), cardiovascular death (RD=7.65, 95% CI 3.29 to 12.01) and all-cause mortality (RD=11.00, 95% CI 5.20 to 16.80). The dominant proteinuria group showed even greater absolute excess risks: MACE (RD=64.85, 95% CI 26.76 to 102.94), cardiovascular death (RD=20.02, 95% CI 1.41 to 38.63) and all-cause mortality (RD=25.21, 95% CI 3.18 to 47.24). Absolute RD for AMI was not statistically significant.
Kaplan-Meier curves revealed that the incidence of MACE increased among participants with higher UACR. Compared with the remaining participants, participants with a UACR ≥300 mg/g had the highest HR for MACE (p<0.0001, figure 1).
Figure 1. Kaplan-Meier curves of the hazards of MACE in different UACR groups. MACE, major adverse cardiovascular events; UACR, urinary albumin-to-creatinine ratio.
In the multivariable Cox model adjusted for age, sex, BMI, eGFR, smoking and drinking status, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia (model II of table 2), compared with those in the normal group, the risks of MACE (HR=1.47; 95% CI 1.06 to 2.06; p=0.023), cardiovascular death (HR=3.03; 95% CI 1.56 to 5.88; p=0.001) and all-cause mortality (HR=1.91; 95% CI 1.23 to 2.97; p=0.004) were significantly greater in the microalbuminuria group. The risk of MACE (HR=3.65; 95% CI 2.14 to 6.23; p<0.001), cardiovascular death (HR=7.91; 95% CI 2.92 to 21.43; p<0.001), stroke (HR=2.57; 95% CI 1.30 to 5.08; p=0.007) and all-cause death (HR=3.59; 95% CI 1.63 to 7.89; p=0.001) in the group with dominant proteinuria was significantly greater than that in the normal group; the risk of AMI tended to increase (HR=3.19; 95% CI 0.95 to 10.68; p=0.060). To enhance model robustness, we conducted a sensitivity analysis using a simplified model with a reduced number of covariates (adjusted only for age, sex, BMI and eGFR, model I of table 2). The results were largely consistent with those from the main analysis, indicating that the key conclusions remained robust. However, the number of participants in the dominant proteinuria group was small (n=33), yielding only 15 MACE, 5 cardiovascular deaths and 3 AMI. Consequently, HRs in this group are accompanied by wide CI and should be interpreted cautiously (table 2).
Table 2. The impact of UACR on adverse cardiovascular outcomes by Cox regression analysis.
| 30≤UACR<300 mg/g vs UACR<30 mg/g | UACR ≥300 mg/g vs UACR <30 mg/g | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Primary endpoint | ||||
| MACE | ||||
| Univariable | 2.44 (1.77 to 3.37) | <0.001 | 7.34 (4.36 to 12.37) | <0.001 |
| Multivariable model I* | 1.81 (1.30 to 2.52) | <0.001 | 4.68 (2.76 to 7.95) | <0.001 |
| Multivariable model II† | 1.47 (1.06 to 2.06) | 0.023 | 3.65 (2.14 to 6.23) | <0.001 |
| Secondary endpoint | ||||
| Cardiovascular death | ||||
| Univariable | 7.44 (4.06 to 13.65) | <0.001 | 18.05 (7.00 to 46.54) | <0.001 |
| Multivariable model I* | 4.45 (2.35 to 8.43) | <0.001 | 10.57 (4.00 to 27.90) | <0.001 |
| Multivariable model II† | 3.03 (1.56 to 5.88) | 0.001 | 7.91 (2.92 to 21.43) | <0.001 |
| AMI | ||||
| Univariable | 2.28 (1.08 to 4.82) | 0.030 | 6.99 (2.18 to 22.43) | 0.001 |
| Multivariable model I* | 1.64 (0.77 to 3.50) | 0.204 | 3.74 (1.13 to 12.42) | 0.031 |
| Multivariable model II† | 1.32 (0.61 to 2.87) | 0.476 | 3.19 (0.95 to 10.68) | 0.060 |
| Stroke | ||||
| Univariable | 2.19 (1.50 to 3.20) | <0.001 | 5.18 (2.66 to 10.11) | <0.001 |
| Multivariable model I* | 1.62 (1.10 to 2.39) | 0.014 | 3.30 (1.68 to 6.48) | <0.001 |
| Multivariable model II† | 1.32 (0.89 to 1.95) | 0.163 | 2.57 (1.30 to 5.08) | 0.007 |
| All-cause death | ||||
| Univariable | 3.48 (2.30 to 5.26) | <0.001 | 6.75 (3.15 to 14.49) | <0.001 |
| Multivariable model I* | 2.22 (1.45 to 3.41) | <0.001 | 4.12 (1.91 to 8.91) | <0.001 |
| Multivariable model II† | 1.91 (1.23 to 2.97) | 0.004 | 3.59 (1.63 to 7.89) | 0.001 |
Model I adjusting for age, sex, BMI and eGFR.
Model II adjusting for age, sex, BMI, eGFR, whether smoking and drinking, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia.
AMI, acute myocardial infarction.BMI, body mass index; eGFR, estimated glomerular filtration rate; MACE, major adverse cardiovascular events; UACR, urinary albumin-to-creatinine ratio.
To account for the competing risk of non-cardiovascular death, we performed Fine-Gray subdistribution hazard models for cause-specific endpoints. The results remained consistent with the primary Cox regression analyses (online supplemental table 3). For MACE, the subdistribution HRs were 1.46 (95% CI 1.05 to 2.03) and 3.75 (95% CI 2.36 to 5.96) for the microalbuminuria and dominant proteinuria groups, respectively. As shown in the data, the main conclusions remained unchanged after accounting for competing risks, with effect estimates closely aligned with the original HRs, supporting the robustness of the primary findings.
We further conducted a restricted cubic spline using fully adjusted models to explicitly determine the dose‒response associations between lnUACR and MACE, which revealed that the risk of MACE gradually increased with increasing lnUACR (figure 2).
Figure 2. The restricted cubic spline of the dose-response relationship between lnUACR and MACE. The restricted cubic spline was adjusted for age; sex; BMI; eGFR; smoking and drinking status; hypertension, diabetes and dyslipidaemia history; and medication history of hypertension, diabetes and dyslipidaemia. The blue area indicates the 95% CI for the restricted cubic spline. BMI, body mass index; eGFR, estimated glomerular filtration rate; lnUACR, natural logarithmic transformation UACR; MACE, major adverse cardiovascular events; UACR, urinary albumin-to-creatinine ratio.
We conducted additional subgroup analyses and discovered no significant interactions between UACR and other risk factors, including age, sex, BMI, eGFR, smoking and drinking status, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia (the p values for interactions were all greater than 0.05; refer to table 3 for further details).
Table 3. Subgroup analyses of the impact of UACR on MACE by Cox regression analyses.
| 30≤UACR<300 mg/g vs UACR<30 mg/g | UACR ≥300 mg/g vs UACR <30 mg/g | P for interaction | |||
|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | ||
| Age, year | 0.208 | ||||
| <65 | 1.07 (0.63 to 1.81) | 0.803 | 4.60 (2.13 to 9.93) | <0.001 | |
| ≥65 | 1.82 (1.18 to 2.82) | 0.007 | 3.18 (1.53 to 6.60) | 0.002 | |
| Sex | 0.816 | ||||
| Male | 1.52 (0.98 to 2.37) | 0.063 | 4.10 (2.18 to 7.71) | <0.001 | |
| Female | 1.42 (0.87 to 2.31) | 0.164 | 2.86 (1.05 to 7.78) | 0.040 | |
| BMI, kg/m2 | 0.276 | ||||
| <24 | 2.58 (1.47 to 4.54) | 0.001 | 3.75 (1.15 to 12.24) | 0.029 | |
| ≥24 to <28 | 1.15 (0.68 to 1.96) | 0.601 | 4.38 (1.91 to 10.03) | <0.001 | |
| ≥28 | 1.21 (0.64 to 2.30) | 0.552 | 3.01 (1.28 to 7.09) | 0.012 | |
| eGFR, mL/min/1.73 m2 | 0.081 | ||||
| <90 | 1.30 (0.91 to 1.86) | 0.149 | 3.48 (2.00 to 6.05) | <0.001 | |
| ≥90 | 4.50 (1.77 to 11.44) | 0.002 | 3.50 (0.46 to 26.94) | 0.228 | |
| Current smoking | 0.282 | ||||
| No | 1.25 (0.83 to 1.87) | 0.282 | 3.49 (1.88 to 6.46) | <0.001 | |
| Yes | 2.22 (1.25 to 3.97) | 0.007 | 4.18 (1.48 to 11.84) | 0.007 | |
| Current drinking | 0.780 | ||||
| No | 1.40 (0.94 to 2.07) | 0.094 | 3.46 (1.96 to 6.14) | <0.001 | |
| Yes | 1.69 (0.92 to 3.13) | 0.093 | 5.33 (1.26 to 22.53) | 0.023 | |
| Hypertension | 0.516 | ||||
| No | 1.33 (0.61 to 2.89) | 0.475 | 0 (0 to Inf) | 0.99 | |
| Yes | 1.51 (1.05 to 2.18) | 0.027 | 3.84 (2.25 to 6.58) | <0.001 | |
| Diabetes | 0.319 | ||||
| No | 1.77 (1.14 to 2.73) | 0.010 | 2.82 (1.22 to 6.50) | 0.015 | |
| Yes | 1.20 (0.73 to 1.98) | 0.482 | 4.49 (2.24 to 9.02) | <0.001 | |
| Dyslipidaemia | 0.804 | ||||
| No | 1.87 (0.88 to 3.97) | 0.105 | 3.86 (1.19 to 12.53) | 0.025 | |
| Yes | 1.40 (0.97 to 2.03) | 0.071 | 3.61 (1.99 to 6.53) | <0.001 | |
| Hypertension drug | 0.724 | ||||
| No | 1.31 (0.83 to 2.07) | 0.250 | 3.37 (1.62 to 7.08) | 0.001 | |
| Yes | 1.70 (1.05 to 2.74) | 0.031 | 4.00 (1.84 to 8.68) | <0.001 | |
| Diabetes drug | 0.430 | ||||
| No | 1.58 (1.09 to 2.27) | 0.015 | 3.07 (1.56 to 6.04) | 0.001 | |
| Yes | 1.10 (0.49 to 2.47) | 0.809 | 5.17 (2.15 to 12.41) | <0.001 | |
| Dyslipidaemia drug | 0.093 | ||||
| No | 1.59 (1.13 to 2.23) | 0.008 | 3.34 (1.89 to 5.91) | <0.001 | |
| Yes | 0.38 (0.05 to 2.82) | 0.341 | 10.15 (2.28 to 45.08) | 0.002 | |
Adjusting for age, sex, BMI, eGFR, whether smoking and drinking, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia.
BMI, body mass index; eGFR, estimated glomerular filtration rate; MACE, major adverse cardiovascular event; UACR, urinary albumin-to-creatinine ratio.
To further explore the impact of LGA on cardiovascular adverse outcomes and all-cause mortality, we separately selected the normal group (UACR<30 mg/g, N=3344) for analysis. The multivariable Cox regression model revealed that in populations with a UACR <30 mg/g, the risk of MACE significantly increased with increasing UACR (HR=1.02; 95% CI 1.00 to 1.04; p=0.036), and the risk of AMI, stroke and all-cause death tended to increase (table 4).
Table 4. The impact of LGA on adverse cardiovascular outcomes in the normal group (UACR<30 mg/g).
| UACR | T2 vs T1 | T3 vs T1 | ||||
|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| Primary endpoint | ||||||
| MACE | ||||||
| Univariable | 1.04 (1.02 to 1.06) | <0.001 | 1.09 (0.79 to 1.51) | 0.588 | 1.52 (1.12 to 2.05) | 0.007 |
| Multivariable* | 1.02 (1.00 to 1.04) | 0.036 | 1.13 (0.81 to 1.57) | 0.482 | 1.31 (0.96 to 1.80) | 0.089 |
| Secondary endpoint | ||||||
| Cardiovascular death | ||||||
| Univariable | 1.06 (1.01 to 1.12) | 0.019 | 1.00 (0.35 to 2.86) | 0.993 | 2.30 (0.95 to 5.59) | 0.066 |
| Multivariable* | 1.04 (0.99 to 1.10) | 0.147 | 1.05 (0.36 to 3.05) | 0.923 | 1.83 (0.72 to 4.64) | 0.206 |
| AMI | ||||||
| Univariable | 1.05 (1.01 to 1.10) | 0.016 | 1.33 (0.58 to 3.04) | 0.495 | 2.69 (1.29 to 5.58) | 0.008 |
| Multivariable* | 1.04 (0.99 to 1.09) | 0.094 | 1.40 (0.61 to 3.23) | 0.432 | 2.56 (1.20 to 5.46) | 0.015 |
| Stroke | ||||||
| Univariable | 1.04 (1.02 to 1.06) | <0.001 | 1.11 (0.78 to 1.59) | 0.557 | 1.40 (1.00 to 1.97) | 0.051 |
| Multivariable* | 1.02 (1.00 to 1.04) | 0.091 | 1.14 (0.79 to 1.64) | 0.483 | 1.17 (0.82 to 1.67) | 0.375 |
| All-cause death | ||||||
| Univariable | 1.04 (1.01 to 1.07) | 0.004 | 1.22 (0.74 to 2.02) | 0.431 | 1.80 (1.13 to 2.86) | 0.013 |
| Multivariable* | 1.03 (1.00 to 1.06) | 0.065 | 1.24 (0.75 to 2.06) | 0.404 | 1.58 (0.97 to 2.56) | 0.065 |
We divided the population with UACR<30 mg/g into three equal groups: T1 (UACR <4.50 mg/g), T2 (4.50 mg/g≤UACR<7.56 mg/g) and T3 (UACR ≥7.56 mg/g).
Adjusting for age, sex, BMI, eGFR, whether smoking and drinking, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia.
AMI, acute myocardial infarction.BMI, body mass index; eGFR, estimated glomerular filtration rate; LGA, low-grade albuminuria; MACE, major adverse cardiovascular event; UACR, urinary albumin-to-creatinine ratio.
We further divided the population with a UACR <30 mg/g into three equal groups: T1 (UACR <4.50 mg/g), T2 (4.50 mg/g≤UACR<7.56 mg/g) and T3 (UACR ≥7.56 mg/g). According to the univariate Cox model, compared with those in the T1 subgroup, the risks of MACE (HR=1.52; 95% CI 1.12 to 2.05; p=0.007), AMI (HR=2.69; 95% CI 1.29 to 5.58; p=0.008) and all-cause death (HR=1.80; 95% CI 1.13 to 2.86; p=0.013) were significantly greater in the T3 subgroup. In a multivariable Cox regression model adjusted for age, sex, BMI, eGFR, smoking and drinking status, hypertension, diabetes and dyslipidaemia history, and medication history of hypertension, diabetes and dyslipidaemia, compared with those in the T1 group, the risk of AMI (HR=2.56; 95% CI 1.20 to 5.46; p=0.015) was significantly greater in the T3 group. The risk of MACE and all-cause death has an increasing trend. However, it is important to interpret the significantly increased risk of AMI observed in the T3 group as a preliminary finding, particularly in the context of multiple comparisons, and its clinical relevance warrants further validation (table 4).
Discussion
Our study revealed that, compared with the normal group, the microalbuminuria group had 0.47-fold, 2.03-fold and 0.91-fold increased risks of MACE, cardiovascular death and all-cause death, respectively. The risk of MACE, cardiovascular death, stroke and all-cause death increased by 2.65 times, 6.91 times, 1.57 times and 2.59 times, respectively, in the group with dominant proteinuria. In the subgroup with normal urine protein, we still observed an increase in the risk of MACE with an increase in the UACR; specifically, for every 1 mg/g increase in the UACR, the risk of MACE increased by 2%.
Previous studies have focused on the impact of the UACR on adverse cardiovascular outcomes or all-cause mortality.22 The predictive value of high-level urinary albumin excretion for cardiovascular death and all-cause death has been established in patients with diabetes, hypertension and a history of cardiovascular disease.623,25 Nevertheless, there is limited research on the predictive value of the UACR for cardiovascular adverse outcomes and all-cause mortality among the general population in the community, particularly in East Asian populations. Previous studies have focused mainly on cardiovascular mortality, all-cause mortality or cardiovascular composite outcomes, and there has been no comprehensive evaluation of the impact of the UACR on multiple adverse outcomes. A study in the UK recruited 20 911 subjects aged 40−79 years and divided them into normal albuminuria, microalbuminuria and macroalbuminuria groups, with a prospective follow-up of 6.3 years. With increasing urinary albumin, the risk of all-cause death and cardiovascular death independently increases.8 Another study included 6770 participants aged between 45 and 84 years from six centres in the USA who had no cardiovascular disease, with an average follow-up of 7.1 years. Elevated urinary albumin and the UACR were found to be independent risk factors for cardiovascular events. In low-weight individuals, the correlations between urinary albumin and the UACR and cardiovascular events are stronger.9 A cohort study in China in 2022 focused on the elderly population in the community, with 2148 participants and an average follow-up of 9.87 years. The UACR is an independent risk factor for all-cause mortality in the elderly population in the community.15
Elevated UACR often coexists with diabetes and chronic kidney disease, conditions known to promote diffuse coronary artery calcification and more complex atherosclerotic plaque morphology. This pathophysiology may lead to more complex percutaneous coronary interventions requiring prolonged and intensified antithrombotic therapy, potentially mediating part of the observed association between UACR and cardiovascular outcomes. The relationship between albuminuria, coronary disease complexity and percutaneous coronary intervention outcomes represents an important pathway through which microvascular and macrovascular disease may interact.26 While our study population was community-based without baseline stroke or myocardial infarction, this mechanism may be particularly relevant for the higher UACR groups where subclinical coronary artery disease burden was likely greater.
In the past decade, the impact of LGA (24-hour urine protein quantification below 30 mg) on clinical adverse events has received increasing attention.27,29 The Framingham Heart Study revealed that in 1568 non-diabetic and non-hypertensive individuals, a low UACR can predict cardiovascular events and all-cause mortality risk during a 6-year follow-up period, independent of traditional risk factors. The risk of cardiovascular events in individuals with UACR above the median is nearly three times greater than that in those with UACR below the median, and the risk of all-cause mortality is also increased by 75%. The results remained valid after 98 subjects with microalbuminuria were excluded.30 A Japanese population study confirmed that among 3599 nondiabetic subjects with normal blood pressure (systolic blood pressure lower than 120 mm Hg and diastolic blood pressure lower than 80 mm Hg), the risk of cardiovascular events and all-cause death from LGA in the top tertile (UACR≥9.6 mg/g for men, ≥12.0 mg/g for women) was 2.79 and 1.69 times higher than that in the first tertile, respectively, which further confirmed the value of LGA in predicting hard endpoints in healthy people.14 A study of 37 091 healthy screening populations in South Korea over 11 years confirmed that LGA can also predict the risk of hypertension and cardiovascular death.11
Research on LGA in the Chinese population is limited, predominantly consisting of cross-sectional studies, or the observed indicators do not represent clinical hard endpoints.28 31 A study involving 32 650 participants aged 40 years and above from seven regional centres in China revealed that LGA was significantly associated with a high 10-year cardiovascular risk score among CVD-free and normoalbuminuric Chinese adults.32 Zhang et al15 focused on the elderly population in Chinese communities and reported that the all-cause mortality rate in the normal high-value group (10 mg/g<UACR<30 mg/g) was 28.9% higher than that in the normal low-value group (UACR<10 mg/g). Our study revealed that within the UACR <30 mg/g group, a statistically significant 2% relative increase in MACE risk per 1 mg/g UACR increment (HR=1.02, 95% CI 1.00 to 1.04). We divided the population with a UACR <30 mg/g into three equal groups. Compared with the first tertile group, the top tertile group had a significantly increased risk of AMI. Additionally, there was a rising trend in the risk of MACE and all-cause death. Our study systematically assessed the impact of LGA on various adverse cardiovascular outcomes in Chinese community populations. However, we recognise that while this association is statistically significant, likely driven by our large sample size, the absolute risk increase per mg/g is modest at the individual level. For example, an increase in UACR from 5 to 15 mg/g would translate to an approximate 20% relative increase in MACE risk, but the absolute RD remains small. Therefore, these findings should be interpreted as evidence of a continuous risk gradient at the population level rather than implying strong prognostic value for individual risk prediction.
The findings of this study further confirm that elevated UACR is a significant risk factor for cardiovascular events and mortality, consistent with previous research.8 9 22 The incremental contributions of this study are threefold: First, we validated this association in a Chinese community-based cohort free of cardiovascular disease at baseline, extending the applicability of this evidence to East Asian populations. Second, we comprehensively evaluated the impact of UACR on multiple hard endpoints, including MACE, cardiovascular death, myocardial infarction, stroke and all-cause mortality. Third, and a key focus of this study, we explored the prognostic significance of LGA. Even within the normoalbuminuric range, increasing UACR was significantly associated with a higher risk of MACE, and the highest tertile showed a significantly increased risk of myocardial infarction. This finding supports the potential value of more granular UACR risk stratification even in apparently healthy community populations, although its clinical utility requires confirmation through prospective intervention studies.
This study has several limitations. First, as a single-centre cohort, its findings—derived from two urban, well-educated and predominantly Han Chinese communities in Beijing—may not be generalisable to rural populations, other ethnic groups or different socioeconomic settings. Second, UACR was measured only once using a random morning urine sample, potentially introducing nondifferential misclassification and regression-dilution bias, which would likely underestimate the true effect. Third, wide CIs for the dominant proteinuria group reflect limited sample size and low event counts, necessitating larger studies for more precise estimates. Fourth, medication use was recorded only as binary (ever/never) at baseline, lacking details on dosage, duration or treatment changes during follow-up; data on diet, physical activity and socioeconomic status were also unavailable. These gaps may contribute to residual confounding, and the reported HRs should therefore be interpreted with caution. Finally, given the exploratory multiple endpoint and subgroup analyses without adjustment for multiple testing, there is an increased risk of type I error, particularly for marginally significant findings. These results should be considered preliminary until confirmed in future studies.
Conclusions
This study suggests that an elevated UACR is a significant risk factor for adverse outcomes, including cardiovascular death, stroke and all-cause mortality, within the Chinese community population. Furthermore, this trend persists even in the population with LGA. These findings support considering UACR as a potentially useful indicator for cardiovascular risk stratification in this population and suggest that even within the normal range, higher UACR values may identify a higher-risk subgroup. Future studies should validate these findings in multicentre, larger populations and explore whether UACR-based (including LGA) risk management strategies can improve clinical outcomes.
Supplementary material
Acknowledgements
We thank all the members of the research team. We are also grateful to the staff of the Pingguoyuan and Gucheng communities’ health centres for their support of this study.
Footnotes
Funding: This work was supported in part by the National Key Research and Development Program of China (2021YFC2500503), National High Level Hospital Clinical Research Funding (High Quality Clinical Research Project of Peking University First Hospital, 2022CR71), Projects of the National Natural Science Foundation of China (82170452) and the Scientific Research Seed Fund of Peking University First Hospital (2018SF003).
Prepublication history and additional supplemental material for this paper are available online. To view these files, please visit the journal online (https://doi.org/10.1136/bmjopen-2025-104198).
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study involves human participants and was approved by the Medical Ethics Committee of Peking University First Hospital (2014(700)). Participants gave informed consent to participate in the study before taking part.
Data availability free text: The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Data availability statement
Data are available on reasonable request.
References
- 1.Schooling CM, Zhao JV. Insights into Causal Cardiovascular Risk Factors from Mendelian Randomization. Curr Cardiol Rep. 2023;25:67–76. doi: 10.1007/s11886-022-01829-8. [DOI] [PubMed] [Google Scholar]
- 2.Haaskjold YL, Lura NG, Bjørneklett R, et al. Long-term follow-up of IgA nephropathy: clinicopathological features and predictors of outcomes. Clin Kidney J. 2023;16:2514–22. doi: 10.1093/ckj/sfad154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ohkuma T, Jun M, Chalmers J, et al. Combination of Changes in Estimated GFR and Albuminuria and the Risk of Major Clinical Outcomes. Clin J Am Soc Nephrol. 2019;14:862–72. doi: 10.2215/CJN.13391118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ying T, Clayton P, Naresh C, et al. Predictive value of spot versus 24-hour measures of proteinuria for death, end-stage kidney disease or chronic kidney disease progression. BMC Nephrol. 2018;19:55. doi: 10.1186/s12882-018-0853-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gerstein HC, Ramasundarahettige C, Avezum A, et al. A novel kidney disease index reflecting both the albumin-to-creatinine ratio and estimated glomerular filtration rate, predicted cardiovascular and kidney outcomes in type 2 diabetes. Cardiovasc Diabetol. 2022;21:158. doi: 10.1186/s12933-022-01594-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gao F, Zhou Y, Yan X, et al. Effect of Urinary Albumin Creatinine Ratio on Cardiovascular Morbidity and Mortality in Diabetes Patients with Atherosclerotic Disease. Diabetes Metab Syndr Obes. 2023;16:819–28. doi: 10.2147/DMSO.S400970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chong J, Fotheringham J, Tomson C, et al. Renal albumin excretion in healthy young adults and its association with mortality risk in the US population. Nephrol Dial Transplant. 2020;35:458–64. doi: 10.1093/ndt/gfy242. [DOI] [PubMed] [Google Scholar]
- 8.Yuyun MF, Khaw K-T, Luben R, et al. Microalbuminuria independently predicts all-cause and cardiovascular mortality in a British population: The European Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) population study. Int J Epidemiol. 2004;33:189–98. doi: 10.1093/ije/dyh008. [DOI] [PubMed] [Google Scholar]
- 9.Carter CE, Katz R, Kramer H, et al. Influence of urine creatinine concentrations on the relation of albumin-creatinine ratio with cardiovascular disease events: the Multi-Ethnic Study of Atherosclerosis (MESA) Am J Kidney Dis. 2013;62:722–9. doi: 10.1053/j.ajkd.2013.05.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Liu M, Liang Y, Zhu J, et al. Albumin-to-creatinine ratio as a predictor of all-cause mortality and hospitalization of congestive heart failure in Chinese elder hypertensive patients with high cardiovascular risks. Clin Hypertens. 2018;24:12. doi: 10.1186/s40885-018-0095-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sung K-C, Ryu S, Lee J-Y, et al. Urine Albumin/Creatinine Ratio Below 30 mg/g is a Predictor of Incident Hypertension and Cardiovascular Mortality. J Am Heart Assoc. 2016;5:e003245. doi: 10.1161/JAHA.116.003245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chang DR, Yeh H-C, Ting I-W, et al. The ratio and difference of urine protein-to-creatinine ratio and albumin-to-creatinine ratio facilitate risk prediction of all-cause mortality. Sci Rep. 2021;11:7851. doi: 10.1038/s41598-021-86541-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Danziger J. Importance of low-grade albuminuria. Mayo Clin Proc. 2008;83:806–12. doi: 10.4065/83.7.806. [DOI] [PubMed] [Google Scholar]
- 14.Tanaka F, Komi R, Makita S, et al. Low-grade albuminuria and incidence of cardiovascular disease and all-cause mortality in nondiabetic and normotensive individuals. J Hypertens. 2016;34:506–12. doi: 10.1097/HJH.0000000000000809. [DOI] [PubMed] [Google Scholar]
- 15.Zhang A, Li M, Qiu J, et al. The relationship between urinary albumin to creatinine ratio and all-cause mortality in the elderly population in the Chinese community: a 10-year follow-up study. BMC Nephrol. 2022;23:16. doi: 10.1186/s12882-021-02644-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fan F, Qi L, Jia J, et al. Noninvasive Central Systolic Blood Pressure Is More Strongly Related to Kidney Function Decline Than Peripheral Systolic Blood Pressure in a Chinese Community-Based Population. Hypertension. 2016;67:1166–72. doi: 10.1161/HYPERTENSIONAHA.115.07019. [DOI] [PubMed] [Google Scholar]
- 17.Liu M, Jia J, Fan F, et al. Association of depressive symptoms with the progression of carotid intima-media thickness in a community-based cohort in Beijing, China. BMJ Open. 2021;11:e050254. doi: 10.1136/bmjopen-2021-050254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. The Lancet. 2007;370:1453–7. doi: 10.1016/S0140-6736(07)61602-X. [DOI] [PubMed] [Google Scholar]
- 19.Ma Y-C, Zuo L, Chen J-H, et al. Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease. J Am Soc Nephrol. 2006;17:2937–44. doi: 10.1681/ASN.2006040368. [DOI] [PubMed] [Google Scholar]
- 20.Liu S, Wu X, Lopez AD, et al. An integrated national mortality surveillance system for death registration and mortality surveillance, China. Bull World Health Organ. 2016;94:46–57. doi: 10.2471/BLT.15.153148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Liang Z, Li K, Chen H, et al. The Association of Plasma Homocysteine Concentrations with a 10-Year Risk of All-Cause and Cardiovascular Mortality in a Community-Based Chinese Population. Nutrients. 2024;16:1945. doi: 10.3390/nu16121945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kühn A, van der Giet M, Kuhlmann MK, et al. Kidney Function as Risk Factor and Predictor of Cardiovascular Outcomes and Mortality Among Older Adults. Am J Kidney Dis. 2021;77:386–96. doi: 10.1053/j.ajkd.2020.09.015. [DOI] [PubMed] [Google Scholar]
- 23.Culleton BF, Larson MG, Parfrey PS, et al. Proteinuria as a risk factor for cardiovascular disease and mortality in older people: a prospective study. Am J Med. 2000;109:1–8. doi: 10.1016/s0002-9343(00)00444-7. [DOI] [PubMed] [Google Scholar]
- 24.Oliveras A, Armario P, Sierra C, et al. Urinary albumin excretion at follow-up predicts cardiovascular outcomes in subjects with resistant hypertension. Am J Hypertens. 2013;26:1148–54. doi: 10.1093/ajh/hpt074. [DOI] [PubMed] [Google Scholar]
- 25.Mok Y, Ballew SH, Sang Y, et al. Albuminuria as a Predictor of Cardiovascular Outcomes in Patients With Acute Myocardial Infarction. J Am Heart Assoc. 2019;8:e010546. doi: 10.1161/JAHA.118.010546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Castiello DS, Oliva A, Andò G, et al. Antithrombotic therapy in complex percutaneous coronary intervention. EuroIntervention. 2025;21:e1051–68. doi: 10.4244/EIJ-D-24-00992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Horn JW, Romundstad S, Ellekjær H, et al. Low grade albuminuria as a risk factor for subtypes of stroke - the HUNT Study in Norway. BMC Neurol. 2020;20:170. doi: 10.1186/s12883-020-01746-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang T, Zhong H, Lian G, et al. Low-Grade Albuminuria Is Associated with Left Ventricular Hypertrophy and Diastolic Dysfunction in Patients with Hypertension. Kidney Blood Press Res. 2019;44:590–603. doi: 10.1159/000500782. [DOI] [PubMed] [Google Scholar]
- 29.de Souza RAF, da Silva EF, de Oliveira DM, et al. Low-grade albuminuria and its relationship with cardiovascular disease risk in hypertensive and diabetic patients in primary health care. BMC Nephrol. 2022;23:257. doi: 10.1186/s12882-022-02884-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ärnlöv J, Evans JC, Meigs JB, et al. Low-Grade Albuminuria and Incidence of Cardiovascular Disease Events in Nonhypertensive and Nondiabetic Individuals. Circulation. 2005;112:969–75. doi: 10.1161/CIRCULATIONAHA.105.538132. [DOI] [PubMed] [Google Scholar]
- 31.Ren M, You L, Lin D, et al. Association of metabolic syndrome with the incidence of low-grade albuminuria: a cohort study in middle-aged and elderly Chinese adults. Aging (Milano) 2021;13:7350–60. doi: 10.18632/aging.202592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Li B, Chen K, Wang J, et al. Low-grade albuminuria is associated with high cardiovascular risk in CVD-free and normoalbuminuric Chinese adults: Results from the REACTION study. J Diabetes. 2021;13:648–60. doi: 10.1111/1753-0407.13152. [DOI] [PubMed] [Google Scholar]