Abstract
The degree of albuminuria predicts cardiovascular and renal outcomes, but it is not known whether changes in albuminuria also predict similar outcomes. In two multicenter, multinational, prospective observational studies, a central laboratory measured albuminuria in 23,480 patients with vascular disease or high-risk diabetes. We quantified the association between a greater than or equal to twofold change in albuminuria in spot urine from baseline to 2 years and the incidence of cardiovascular and renal outcomes and all-cause mortality during the subsequent 32 months. A greater than or equal to twofold increase in albuminuria from baseline to 2 years, observed in 28%, associated with nearly 50% higher mortality (HR 1.48; 95% CI 1.32 to 1.66), and a greater than or equal to twofold decrease in albuminuria, observed in 21%, associated with 15% lower mortality (HR 0.85; 95% CI 0.74 to 0.98) compared with those with lesser changes in albuminuria, after adjustment for baseline albuminuria, BP, and other potential confounders. Increases in albuminuria also significantly associated with cardiovascular death, composite cardiovascular outcomes (cardiovascular death, myocardial infarction, stroke, and hospitalization for heart failure), and renal outcomes including dialysis or doubling of serum creatinine (adjusted HR 1.40; 95% CI 1.11 to 1.78). In conclusion, in patients with vascular disease, changes in albuminuria predict mortality and cardiovascular and renal outcomes, independent of baseline albuminuria. This suggests that monitoring albuminuria is a useful strategy to help predict cardiovascular risk.
Albuminuria is a powerful independent predictor of progression of renal disease, cardiovascular disease and death in people with renal disease, hypertension, diabetes, vascular disease, and in the general population.1–9 The association of albuminuria with cardiovascular outcomes and death is also observed in the range considered to be “normal” and in apparently healthy individuals.8,10,11
By contrast, it is unknown whether a patient's risk is modified by changes in albuminuria. We hypothesized that changes in albuminuria would have an effect on the occurrence of cardiovascular and renal outcomes in either direction, independent of baseline albuminuria. To our knowledge, the above hypothesis was tested only in two specific populations, namely, in people with hypertension and electrocardiographic evidence of left ventricular hypertrophy and in hypertensive people with overt diabetic nephropathy, and neither study reported the predictive value of changes in albuminuria for mortality.11,12
In two prospective trials, The ONgoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial (ONTARGET) and The Telmisartan Randomized AssessmeNt Study in ACE iNtoleranT subjects with cardiovascular Disease (TRANSCEND), high-risk patients with coronary, cerebrovascular, or peripheral artery disease or diabetes with end-organ damage were enrolled, treated according to current standard therapy, and randomized to 80 mg/d telmisartan, 10 mg/d ramipril, or both (ONTARGET), and 80 mg/d telmisartan or placebo (TRANSCEND).13–15 We analyzed the predictive value of changes in albuminuria for mortality as well as for cardiovascular and renal outcomes in this population with vascular disease.
RESULTS
Baseline Data
The clinical and demographic characteristics of our patients are given in Table 1. Albuminuria at baseline was associated to the incidence of cardiovascular and renal events and mortality. In the groups with UACR <10 mg/g, ≥10 and <30 mg/g, ≥30 and <100 mg/g, ≥100 mg and <300 mg/g, and ≥300 mg/g, the yearly mortality rates were 1.85%, 2.88%, 3.90%, 4.86%, and 7.92% (P < 0.0001 for trend). Cardiovascular deaths, and cardiovascular and renal outcomes, were also independently related to baseline albuminuria (data not given).
Table 1.
Demographic and clinical characteristics of the study population with measurements of albuminuria at baseline and 2-year examination
| Label | Patients with Measurement of Albuminuria at Baseline and at 2-Year Visit (N = 23,480) | Patients with Decrease in Albuminuria (UACR >50%) N = 4994 (21.3%) | Patients with Minor Change in Albuminuria N = 11,968 (51.0%) (Less Than Doubling or Halving) | Patients with an Increase in Albuminuria (UACR >100%) N = 6518 (27.8%) | P (Test for Difference between Albuminuria Change Groups) |
|---|---|---|---|---|---|
| Clinical data | |||||
| age (years) | 66.3 (7.1) | 66.3 (7.2) | 66.1 (7) | 66.6 (7.1) | <0.0001 |
| body mass index (kg/m2) | 28.2 (4.7) | 28.4 (5) | 28 (4.6) | 28.3 (4.8) | <0.0001 |
| baseline UACRa percentage (mg/g)b | 6.5 (2.2 to 12.8) | 19.1 (6.2 to 45.9) | 5.3 (2.2 to 8.0) | 4.2 (1.6 to 7.3) | <0.0001 |
| SBP sitting (mmHg) | 142 (17) | 143 (17) | 141 (17) | 142 (17) | <0.0001 |
| DBP sitting (mmHg) | 82 (10) | 83 (11) | 82 (10) | 82 (10) | <0.0001 |
| pulse rate sitting (beats per min) | 68 (12) | 69 (12) | 67 (12) | 68 (12) | <0.0001 |
| SBP change at 2 years (mmHg) | −3.9 (20.2) | −6.5 (20.1) | −4.3 (19.7) | −1.3 (21) | <0.0001 |
| DBP change at 2 years (mmHg) | −3.0 (12.0) | −4.2 (12) | −3.1 (11.8) | −1.9 (12.3) | <0.0001 |
| change of pulse rate at 2 years (beats per min) | 0.5 (13.1) | −0.2 (12.5) | 0.5 (13) | 1.2 (13.6) | <0.0001 |
| cholesterol (mg/dl) | |||||
| total (mg/dl) | 190.2 (43.0) | 192.5 (43.6) | 188.8 (42.6) | 191.1 (43.1) | <0.0001 |
| LDL (mg/dl) | 112.4 (37.4) | 113.8 (38.3) | 111.5 (37.2) | 113.2 (37.2) | 0.0004 |
| HDL (mg/dl) | 48.4 (15.4) | 49 (16.1) | 48.4 (15) | 47.9 (15.6) | 0.0006 |
| triglycerides (mg/dl) | 158.0 (104.9) | 161.8 (121.3) | 154.5 (96.9) | 161.6 (105.2) | <0.0001 |
| glucose (mg/dl) | 117.0 (43.8) | 120.6 (48) | 114.5 (41.3) | 118.9 (44.7) | <0.0001 |
| serum creatinine (mg/dl) | 1.1 (0.3) | 1.1 (0.3) | 1 (0.3) | 1.1 (0.3) | <0.0001 |
| change of serum creatinine (mg/dl) at 2 years | 0.05 (0.25) | 0.08 (0.28) | 0.05 (0.23) | 0.04 (0.25) | <0.0001 |
| eGFR MDRD (ml/min per 1.73 m2) | 73.6 (19.2) | 73.9 (19.9) | 74.2 (18.7) | 72.4 (19.4) | <0.0001 |
| change of eGFR MDRD (ml/min per 1.73 m2) at 2 years | −3.3 (16.4) | −4.9 (16.7) | −3.2 (16.0) | −2.1 (16.8) | <0.0001 |
| potassium (mmol/L) | 4.4 (0.4) | 4.4 (0.5) | 4.4 (0.4) | 4.4 (0.4) | 0.6596 |
| female gender (no.[%]) | 6757 (28.8) | 1507 (30.2) | 3272 (27.3) | 1978 (30.3) | <0.0001 |
| Ethnic group (no. [%])c | <0.0001 | ||||
| Asian | 3851 (16.4) | 829 (16.6) | 1867 (15.6) | 1155 (17.7) | |
| African (black) | 516 (2.2) | 135 (2.7) | 210 (1.8) | 171 (2.6) | |
| European (white) | 16,768 (71.4) | 3543 (70.9) | 8714 (72.8) | 4511 (69.2) | |
| other ethnic group | 2342 (10) | 486 (9.7) | 1175 (9.8) | 681 (10.4) | |
| missing data of ethnic group | 3 (<0.1) | 1 (0) | 2 (0) | 0 (0) | |
| Diagnosis for study entry (no. [%]) | |||||
| coronary artery disease | 17,644 (75.1) | 3621 (72.5) | 9278 (77.5) | 4745 (72.8) | <0.0001 |
| peripheral artery disease | 2880 (12.3) | 627 (12.6) | 1396 (11.7) | 857 (13.1) | 0.0104 |
| stroke | 3667 (15.6) | 834 (16.7) | 1744 (14.6) | 1089 (16.7) | <0.0001 |
| transient ischemic attack (TIA) | 820 (3.5) | 200 (4) | 385 (3.2) | 235 (3.6) | 0.0328 |
| high-risk diabetes | 6300 (26.8) | 1475 (29.5) | 2904 (24.3) | 1921 (29.5) | <0.0001 |
| left ventricular hypertrophy | 1203 (5.1) | 264 (5.3) | 567 (4.7) | 372 (5.7) | 0.0141 |
| microalbuminuria | 1443 (6.1) | 395 (7.9) | 593 (5) | 455 (7) | <0.0001 |
| Clinical history (no. [%]) | |||||
| myocardial infarction | 11,379 (48.5) | 2330 (46.7) | 6031 (50.4) | 3018 (46.3) | <0.0001 |
| angina pectoris | 10,891 (46.4) | 2171 (43.5) | 5726 (47.8) | 2994 (45.9) | <0.0001 |
| stable | 8466 (36.1) | 1693 (33.9) | 4443 (37.1) | 2330 (35.7) | 0.0003 |
| unstable | 3628 (15.5) | 720 (14.4) | 1908 (15.9) | 1000 (15.3) | 0.0421 |
| stroke/TIA | 4865 (20.7) | 1125 (22.5) | 2328 (19.5) | 1412 (21.7) | <0.0001 |
| hypertension | 16,409 (69.9) | 3606 (72.2) | 8042 (67.2) | 4761 (73) | <0.0001 |
| diabetes | 8543 (36.4) | 1981 (39.7) | 3925 (32.8) | 2637 (40.5) | <0.0001 |
| coronary artery bypass grafting | 5144 (21.9) | 1109 (22.2) | 2697 (22.5) | 1338 (20.5) | 0.0059 |
| percutaneous transluminal coronary angioplasty | 6902 (29.4) | 1408 (28.2) | 3716 (31) | 1778 (27.3) | <0.0001 |
| Smoking status (no. [%]) | 0.0123 | ||||
| current smoker | 2765 (11.8) | 567 (11.4) | 1419 (11.9) | 779 (12) | |
| past smoker | 11,991 (51.1) | 2539 (50.8) | 6222 (52) | 3230 (49.6) | |
| never smoked | 8695 (37.0) | 1882 (37.7) | 4313 (36) | 2500 (38.4) | |
| Alcohol consumption (%) | 9394 (40.0) | 1936 (38.8) | 5007 (41.8) | 2451 (37.6) | <0.0001 |
DBP, diastolic blood pressure; SBP, systolic blood pressure; eGFR MDRD, estimated glomerular filtration rate according to the MDRD formula.17
aValues are given as mean (SD).
bFor albuminuria, geometric mean and interquartile range are given.
cEthnic group was self-reported.
Changes in Albuminuria and Association to BP, Heart Rate, and Kidney Function
Albuminuria increased from 6.5 mg/g (interquartile range: 2.2 to 12.8) at baseline to 7.4 mg/g (interquartile range: 2.4 to 14.9) at the 2-year examination (P < 0.0001); it decreased by >50% in 4994 (21.3%), increased by >100% in 6518 (27.8%), and exhibited minor changes in 11,968 (51.0%). Multiple regression analysis revealed that changes in systolic blood pressure (SBP), heart rate, and serum creatinine were due to the large sample size significantly associated with changes in albuminuria after 2 years (all P < 0.0001). A 10 mmHg increase in SBP predicted a 6% increase in UACR, an increase of 10 beats in heart rate predicted an increase of 4% in UACR, and an increase of 0.1 mg/dl in serum creatinine predicted a decrease of 3% in UACR.
Changes in Albuminuria and All-Cause Mortality
Overall, in a main effect model mortality rates were significantly higher in patients with more than doubling of albuminuria (hazard ratio [HR] 1.56, 95% confidence interval [CI] 1.40 to 1.75, P < 0.0001), and lower in patients with decreases in albuminuria by more than half (HR 0.84, 95% CI 0.73 to 0.97, P = 0.015), as compared with those with minor changes in albuminuria, after adjusting for baseline albuminuria. After adjustment for other predictors including changes in systolic and diastolic BP (see Concise Methods), these associations were maintained (adjusted HR for increase in albuminuria: 1.48, 95% CI 1.32 to 1.66, P < 0.0001; adjusted HR for decrease in albuminuria: 0.85, 95% CI, 0.74 to 0.98, P = 0.026) (Figure 1). In Table 2A, the adjusted HRs for each baseline albuminuria category and each albuminuria change category are displayed (with patients who have a UACR <10 at baseline and no change within 2 years as the reference group, i.e., HR = 1). For example, in patients with UACR ≥10 and <30 mg/g creatinine the adjusted HR in patients with doubling or more in albuminuria was 2.48, compared with 1.20 in patients with a minor change in UACR corresponding to a 107% higher risk.
Figure 1.
Adjusted HR (95 CI %) [x axis] of changes in UACR from baseline to 2-year visit for (A) all-cause mortality, (B) cardiovascular death, (C) composite cardiovascular outcome, and (D) composite renal outcome after the 2-year visit with mean follow-up of 32 months in the whole study group. Increase refers to at least doubling and decrease to at least halving of albuminuria. Minor change of albuminuria was taken as reference group (HR = 1.0) and P values of the adjusted HR are given in the figure next to the circles. (Hazard ratios were adjusted for age, sex, body mass index, smoking, alcohol consumption, eGFR, plasma glucose, systolic and diastolic BP and HR at baseline, BP change and eGFR change within 2-year study (ONTARGET, TRANSCEND), treatment, and diagnosis at study entry such as coronary artery disease, peripheral artery disease, previous stroke, transient ischemic attack, and high-risk diabetes.)
Table 2.
Number of patients and events and predicted hazard ratio of a decrease or increase in UACR for all-cause mortality (A), cardiovascular death (B), composite cardiovascular outcome (C), and composite renal outcome (D), in the five groups divided according to baseline albuminuria and to change in albuminuria
| Predicted Hazard Ratioa/Baseline UACR | Patients | Events | Decrease in UACR >50% | Patients | Events | Minor Change in UACR | Patients | Events | Increase in UACR >100% | |
|---|---|---|---|---|---|---|---|---|---|---|
| (A) All-cause mortalityb (P = 0.017 in test for interaction) | ||||||||||
| UACR | <10 | 2003 | 87 | 1.05 (0.83 to 1.32) | 9453 | 404 | 1 (1 to 1) | 5175 | 350 | 1.52 (1.31 to 1.76) |
| UACR | ≥10 and <30 | 1376 | 83 | 1.18 (0.92 to 1.52) | 1298 | 86 | 1.20 (0.95 to 1.53) | 704 | 94 | 2.48 (1.97 to 3.14) |
| UACR | ≥30 and <100 | 887 | 71 | 1.52 (1.17 to 1.97) | 583 | 63 | 1.92 (1.47 to 2.53) | 390 | 60 | 2.73 (2.06 to 3.62) |
| UACR | ≥100 and <300 | 436 | 42 | 1.67 (1.20 to 2.32) | 308 | 50 | 2.95 (2.18 to 3.99) | 168 | 25 | 2.80 (1.85 to 4.23) |
| UACR | ≥300 | 292 | 50 | 2.55 (1.86 to 3.52) | 326 | 73 | 4.02 (3.06 to 5.27) | 81 | 18 | 3.39 (2.09 to 5.51) |
| (B) Cardiovascular deathb (P = 0.124 in test for interaction) | ||||||||||
| UACR | <10 | 2003 | 56 | 0.88 (0.73 to 1.05) | 9453 | 248 | 1 (1 to 1) | 5175 | 221 | 1.54 (1.33 to 1.78) |
| UACR | ≥10 and <30 | 1376 | 52 | 1.30 (1.03 to 1.63) | 1298 | 60 | 1.48 (1.24 to 1.78) | 704 | 67 | 2.28 (1.81 to 2.89) |
| UACR | ≥30 and <100 | 887 | 49 | 1.67 (1.31 to 2.13) | 583 | 37 | 1.90 (1.54 to 2.35) | 390 | 42 | 2.94 (2.27 to 3.80) |
| UACR | ≥100 and <300 | 436 | 27 | 1.91 (1.43 to 2.55) | 308 | 26 | 2.18 (1.67 to 2.85) | 168 | 19 | 3.37 (2.48 to 4.58) |
| UACR | ≥300 | 292 | 35 | 3.00 (2.25 to 4.01) | 326 | 49 | 3.42 (2.64 to 4.44) | 81 | 9 | 5.28 (3.87 to 7.21) |
| (C) Composite cardiovascular outcomeb (P = 0.086 in test for interaction) | ||||||||||
| UACR | <10 | 2003 | 154 | 0.88 (0.79 to 0.99) | 9453 | 732 | 1 (1 to 1) | 5175 | 535 | 1.30 (1.19 to 1.43) |
| UACR | ≥10 and <30 | 1376 | 143 | 1.24 (1.07 to 1.42) | 1298 | 170 | 1.40 (1.25 to 1.57) | 704 | 133 | 1.82 (1.57 to 2.11) |
| UACR | ≥30 and <100 | 887 | 113 | 1.30 (1.11 to 1.52) | 583 | 78 | 1.47 (1.27 to 1.69) | 390 | 76 | 1.91 (1.61 to 2.26) |
| UACR | ≥100 and <300 | 436 | 64 | 1.59 (1.32 to 1.92) | 308 | 58 | 1.80 (1.51 to 2.14) | 168 | 42 | 2.34 (1.91 to 2.86) |
| UACR | ≥300 | 292 | 52 | 1.83 (1.49 to 2.24) | 326 | 86 | 2.07 (1.72 to 2.5) | 81 | 20 | 2.69 (2.17 to 3.35) |
| (D) Composite renal outcomeb (P = 0.061 in test for interaction) | ||||||||||
| UACR | <10 | 2003 | 26 | 0.73 (0.56 to 0.95) | 9453 | 84 | 1 (1 to 1) | 5175 | 54 | 1.40 (1.11 to 1.78) |
| UACR | ≥10 and <30 | 1376 | 26 | 1.31 (0.91 to 1.90) | 1298 | 18 | 1.79 (1.32 to 2.44) | 704 | 21 | 2.52 (1.69 to 3.74) |
| UACR | ≥30 and <100 | 887 | 20 | 1.63 (1.10 to 2.40) | 583 | 14 | 2.22 (1.58 to 3.12) | 390 | 20 | 3.12 (2.05 to 4.76) |
| UACR | ≥100 and <300 | 436 | 17 | 2.52 (1.68 to 3.78) | 308 | 9 | 3.44 (2.39 to 4.94) | 168 | 20 | 4.83 (3.09 to 7.53) |
| UACR | ≥300 | 292 | 24 | 3.98 (2.68 to 5.91) | 326 | 56 | 5.43 (3.88 to 7.59) | 81 | 16 | 7.62 (4.89 to 11.88) |
Reference group is UACR <10 mg/g and minor change of albuminuria of the whole study population; urinary albumine to creatinine ratio (mg/g). (For details see Concise Methods.)
aHazard ratios were adjusted for age, gender, BMI, smoking, alcohol consumption, eGFR, plasma glucose, systolic and diastolic BP and HR at baseline, BP change and eGFR change within 2-year study (ONTARGET, TRANSCEND), treatment, and diagnosis at study entry such as coronary artery disease, peripheral artery disease, previous stroke, transient ischemic attack, and high-risk diabetes.
bIn the COX regression model interaction between changes in albuminuria and baseline albuminuria was taken into account.
Changes in Albuminuria and Cardiovascular Death
Cardiovascular death rates were higher in patients with increases in albuminuria >100% (HR 1.65, 95% CI, 1.43 to 1.90, P < 0.0001), also after adjusting for confounders (adjusted HR 1.54, 95% CI 1.34 to 1.78, P < 0.0001), as compared with those with minor changes in albuminuria (Figure 1). Decreases in albuminuria were not related to reduced cardiovascular death rates (HR 0.88, 95% CI 0.74 to 1.04, P = 0.129), also after adjusting for confounders (Figure 1). In Table 2B, the adjusted HRs for each baseline albumine category are given; for example, in patients with UACR >300 mg/g and an increase of UACR the adjusted HR was 5.28, compared with an adjusted HR of 3.42 in patients with minor changes, corresponding to a 54% higher risk.
Changes in Albuminuria and Cardiovascular Outcome
The incidence of the composite cardiovascular outcome was increased in patients with increases in albuminuria (HR 1.38, 95% CI 1.26 to 1.51, P < 0.0001) and reduced in patients with decreases in albuminuria (HR 0.85, 95% CI 0.76 to 0.95, P = 0.004). These hazard ratios were consistent across all levels of baseline albuminuria and were virtually unchanged after adjustment for confounding covariates (Figure 1). In Table 2C, the adjusted HRs for each baseline albuminuria category are listed.
Changes in Albuminuria and Renal Outcome
The incidence of the composite renal outcome was increased in patients with increases in albuminuria (HR 1.52, 95% CI 1.22 to 1.91, P = 0.0003). When all potential confounders were included, increase of albuminuria was associated with an increase of ESRD or doubling of serum creatine (HR 1.40, 95% CI 1.11 to 1.78, P = 0.005), whereas decrease of albuminuria was associated with a decrease of the combined renal outcome (HR 0.73, 95% CI 0.57 to 0.95, P = 0.019). In Table 2D, the adjusted HRs for each baseline albuminuria category are listed.
Descriptive Analysis of Yearly Event Rates
Results of a model-free analysis of yearly event rates in the baseline-by-change groups are displayed in Table 3. In patients with increases in albuminuria, mortality rates were on average 0.37% per year higher (corresponding to a relative increase of 19% [weighted means]), and cardiovascular death and composite cardiovascular events rates by 0.79% and 1.33% per year, respectively (corresponding to a relative increase of 67% and 38% [weighted means]) as compared with those with minor changes in albuminuria. Decreases in albuminuria were related to slightly decreased event rates by 0.29%, 0.12%, and 0.44% per year in all-cause mortality, cardiovascular death, and composite cardiovascular events, respectively. No such weighted mean calculations were performed for the composite renal outcome because of the apparent heterogeneity between the various baseline levels in this unadjusted analysis.
Table 3.
Yearly event rate (absolute difference and ratio of absolute to reference group) of decreases or increases in UACR for all-cause mortality (A), cardiovascular death (B), composite cardiovascular outcome (C), and composite renal outcome (D), in the five groups divided according to baseline albuminuria and to change in albuminuria
| >50% Decrease in UACR |
Minor Change in UACR (Reference Group) Yearly Event Rate (%) | >100% Increase in UACR |
||||||
|---|---|---|---|---|---|---|---|---|
| Yearly Event Rate (%) | Absolute Difference (%) | Ratio to Reference | Yearly Event Rate (%) | Absolute Difference (%) | Ratio to Reference | |||
| (A) All-cause mortality | ||||||||
| UACR | <10 | 1.59 | 0.04 | 1.03 | 1.55 | 1.85 | 0.30 | 1.19 |
| UACR | ≥10 and <30 | 2.21 | −0.22 | 0.91 | 2.43 | 2.88 | 0.45 | 1.19 |
| UACR | ≥30 and <100 | 3.00 | −1.00 | 0.75 | 4.00 | 5.84 | 1.84 | 1.46 |
| UACR | ≥100 and <300 | 3.62 | −2.56 | 0.59 | 6.18 | 5.75 | −0.43 | 0.93 |
| UACR | ≥300 | 6.59 | −2.37 | 0.74 | 8.96 | 8.70 | −0.26 | 0.97 |
| Weighted mean | −0.29 | 0.94 | 0.37 | 1.19 | ||||
| (B) Cardiovascular death | ||||||||
| UACR | <10 | 1.02 | 0.07 | 1.07 | 0.95 | 1.58 | 0.63 | 1.66 |
| UACR | ≥10 and <30 | 1.38 | −0.32 | 0.81 | 1.70 | 3.64 | 1.94 | 2.14 |
| UACR | ≥30 and <100 | 2.07 | −0.28 | 0.88 | 2.35 | 4.09 | 1.74 | 1.74 |
| UACR | ≥100 and <300 | 2.32 | −0.89 | 0.72 | 3.21 | 4.37 | 1.16 | 1.36 |
| UACR | ≥300 | 4.61 | −1.41 | 0.77 | 6.02 | 4.35 | −1.67 | 0.72 |
| Weighted mean | −0.12 | 0.98 | 0.79 | 1.67 | ||||
| (C) Composite cardiovascular outcome | ||||||||
| UACR | <10 | 2.87 | −0.02 | 0.99 | 2.89 | 3.96 | 1.07 | 1.37 |
| UACR | ≥10 and <30 | 3.94 | −1.11 | 0.78 | 5.05 | 7.64 | 2.59 | 1.51 |
| UACR | ≥30 and <100 | 4.95 | −0.22 | 0.96 | 5.17 | 7.94 | 2.77 | 1.54 |
| UACR | ≥100 and <300 | 5.83 | −1.79 | 0.77 | 7.62 | 10.58 | 2.96 | 1.39 |
| UACR | ≥300 | 7.09 | −4.30 | 0.62 | 11.39 | 10.67 | −0.72 | 0.94 |
| Weighted mean | −0.44 | 0.93 | 1.33 | 1.38 | ||||
| (D) Composite renal outcome | ||||||||
| UACR | <10 | 0.54 | 0.21 | 1.64 | 0.33 | 0.37 | 0.04 | 1.12 |
| UACR | ≥10 and <30 | 0.72 | 0.26 | 1.57 | 0.46 | 1.06 | 0.60 | 2.30 |
| UACR | ≥30 and <100 | 0.90 | −0.15 | 0.86 | 1.05 | 1.68 | 0.63 | 1.60 |
| UACR | ≥100 and <300 | 1.42 | 0.72 | 2.03 | 0.70 | 4.84 | 4.14 | 6.91 |
| UACR | ≥300 | 3.16 | −4.03 | 0.44 | 7.19 | 7.74 | 1.55 | 1.22 |
UACR, mg/g. Please note that the yearly event rates are not adjusted for confounders.
Subgroups and Shift Table Analysis
The prognostic effect of changes in albuminuria and all-cause mortality were consistently observed in each of the two studies (ONTARGET versus TRANSCEND) or between the allocated treatments (telmisartan, ramipril, combination of both, or placebo) in the analyses adjusted for covariates (Figure 2). The percentage of patients with increases and decreases in albuminuria differed between placebo and active treatments (P < 0.0001) and between ramipril and each telmisartan group (P < 0.0001; Figure 2).
Figure 2.
Adjusted HR (95 CI%) for all-cause mortality in patients with a mean follow-up of 32 months: decrease of ≥50% or more (upper part); increase of ≥100% in all patients (lower part) in whole study group (N = 23,480) and in the five treatment groups. No heterogeneity between treatments has been observed (P = 0.12). Likewise, no heterogeneity between treatments were observed for cardiovascular death (P = 0.32) and composite cardiovascular outcome (P = 0.70). Minor change of albuminuria was taken as reference group (HR = 1.0) increase refers to at least doubling and decrease to at least halving of albuminuria. (P values of the adjusted HR are given in the figure, % numbers add to 100% per treatment group.) (Hazard ratios were adjusted for age, sex, body mass index, smoking, alcohol consumption, eGFR, plasma glucose, systolic and diastolic BP and HR at baseline, BP change and eGFR change within 2-year study (ONTARGET, TRANSCEND), treatment, and diagnosis at study entry such as coronary artery disease, peripheral artery disease, previous stroke, transient ischemic attack, and high-risk diabetes.)
Because of the well-known high variability of albuminuria measurements, the use of relative scales (halving, doubling) may exaggerate small changes within the normal range. Therefore, we additionally analyzed post hoc our data in a categorial manner (Table 4). This shift analysis confirmed the results of the analysis based on changes in albuminuria. Deteriorations in albuminuria by more than one category were associated with increased all-cause mortality, cardiovascular mortality, composite cardiovascular outcome, and renal outcome (all P < 0.0001), whereas shifts toward lower albuminuria categories were associated with improved patients' prognosis.
Table 4.
Shift table analysis: Number of patients (A), predicted hazard ratios (HR) of all-cause mortality (B), cardiovascular death (C), cardiovascular outcome (D), and renal outcome (E) categorized by change of albuminuria (2 years) and baseline UACR
| Changes in Albuminuria Categories |
|||||
|---|---|---|---|---|---|
| >1 Category Better | 1 Category Better | No Change | 1 Category Worse | >1 Category Worse | |
| (A) No. patientsa | |||||
| Baseline UACRb | |||||
| <10 | 13,685 | 2026 | 920 | ||
| 10 to <30 | 1647 | 929 | 571 | 231 | |
| 30 to <100 | 522 | 423 | 481 | 316 | 118 |
| 100 to <300 | 252 | 221 | 248 | 191 | |
| ≥300 | 164 | 123 | 412 | ||
| (B) All-cause mortalitya | |||||
| Baseline UACR | |||||
| <10 | |||||
| events (%) | 578 (4.2%) | 146 (7.2%) | 117 (12.7%) | ||
| HR (95% CI)b | 1.00 | 1.40 (1.22 to 1.62) | 2.56 (2.16 to 3.04) | ||
| 10 to <30 | |||||
| events (%) | 100 (6.1%) | 63 (6.8%) | 58 (10.2%) | 42 (18.2%) | |
| HR (95% CI)b | 1.11 (0.92 to 1.34) | 1.33 (1.13 to 1.57) | 1.87 (1.53 to 2.28) | 3.41 (2.72 to 4.26) | |
| 30 to <100 | |||||
| events (%) | 41 (7.9%) | 35 (8.3% | 54 (11.2%) | 42 (13.3%) | 22 (18.6%) |
| HR (95% CI)b | 1.23 (0.95 to 1.60) | 1.54 (1.23 to 1.92) | 1.84 (1.54 to 2.20) | 2.59 (2.09 to 3.21) | 4.72 (3.71 to 6.01) |
| 100 to <300 | |||||
| events (%) | 20 (7.9%) | 29 (13.1%) | 37 (14.9%) | 31 (16.2%) | |
| HR (95% CI)b | 1.63 (1.23 to 2.17) | 2.03 (1.59 to 2.61) | 2.44 (1.96 to 3.04) | 3.42 (2.67 to 4.38) | |
| ≥300 | |||||
| events (%) | 25 (15.2%) | 16 (13.0%) | 100 (24.3%) | ||
| HR (95% CI)b | 2.55 (1.92 to 3.40) | 3.18 (2.44 to 4.14) | 3.81 (3.07 to 4.74) | ||
| (C) Cardiovascular deathc | |||||
| Baseline UACR | |||||
| <10 | |||||
| events (%) | 357 (2.6%) | 91 (4.5%) | 77 (8.4%) | ||
| HR (95% CI)b | 1.00 | 1.39 (1.16 to 1.67) | 2.61 (2.12 to 3.22) | ||
| 10 to <30 | |||||
| events (%) | 64 (3.9%) | 45 (4.8%) | 40 (7.0%) | 30 (13.0%) | |
| HR (95% CI)b | 1.18 (0.93 to 1.49) | 1.46 (1.20 to 1.78) | 2.04 (1.60 to 2.60) | 3.82 (2.91 to 5.02) | |
| 30 to <100 | |||||
| events (%) | 32 (6.1%) | 20 (4.7%) | 34 (7.1%) | 26 (8.2%) | 16 (13.6%) |
| HR (95% CI)b | 1.40 (1.02 to 1.91) | 1.57 (1.19 to 2.07) | 1.95 (1.56 to 2.43) | 2.72 (2.08 to 3.55) | 5.09 (3.79 to 6.83) |
| 100 to <300 | |||||
| events (%) | 10 (4.0%) | 22 (10.0%) | 21 (8.5%) | 19 (9.9%) | |
| HR (95% CI)b | 1.74 (1.23 to 2.48) | 1.96 (1.43 to 2.69) | 2.44 (1.84 to 3.22) | 3.39 (2.48 to 4.64) | |
| ≥300 | |||||
| events (%) | 18 (11.0%) | 7 (5.7%) | 68 (16.5%) | ||
| HR (95% CI)b | 2.90 (2.04 to 4.12) | 3.25 (2.34 to 4.53) | 4.04 (3.08 to 4.31) | ||
| (D) Composite cardiovascular outcomed | |||||
| Baseline UACR | |||||
| <10 | |||||
| events (%) | 1042 (7.6%) | 242 (11.9%) | 137 (14.9%) | ||
| HR (95% CI)b | 1.00 | 1.28 (1.14 to 1.43) | 1.81 (1.56 to 2.09) | ||
| 10 to <30 | |||||
| events (%) | 177 (10.7%) | 119 (12.8%) | 93 (16.3%) | 57 (24.7%) | |
| HR (95% CI)b | 1.22 (1.06 to 1.40) | 1.37 (1.21 to 1.56) | 1.75 (1.50 to 2.05) | 2.48 (2.06 to 2.98) | |
| 30 to <100 | |||||
| events (%) | 61 (11.7%) | 55 (13.0%) | 77 (16.0%) | 47 (14.9%) | 27 (22.9%) |
| HR (95% CI)b | 1.08 (0.87 to 1.34) | 1.34 (1.12 to 1.61) | 1.51 (1.30 to 1.76) | 1.93 (1.62 to 2.30) | 2.73 (2.22 to 3.35) |
| 100 to <300 | |||||
| events (%) | 28 (11.1%) | 46 (20.8%) | 46 (18.5%) | 44 (23.0%) | |
| HR (95% CI)b | 1.37 (1.08 to 1.74) | 1.71 (1.39 to 2.10) | 1.93 (1.60 to 2.32) | 2.46 (2.00 to 3.01) | |
| ≥300 | |||||
| events (%) | 27 (16.5%) | 17 (13.8%) | 114 (27.7%) | ||
| HR (95% CI)b | 1.65 (1.28 to 2.12) | 2.06 (1.64 to 2.58) | 2.32 (1.91 to 2.81) | ||
| (E) Composite renal outcomee | |||||
| Baseline UACR | |||||
| <10 | |||||
| events (%) | 122 (0.9%) | 18 (0.9%) | 24 (2.6%) | ||
| HR (95% CI)b | 1.00 | 1.39 (1.03 to 1.88) | 2.43 (1.70 to 3.46) | ||
| 10 to <30 | |||||
| events (%) | 28 (1.7%) | 12 (1.3%) | 15 (2.6%) | 10 (4.3%) | |
| HR (95% CI)b | 1.40 (0.97 to 2.01) | 1.81 (1.30 to 2.52) | 2.51 (1.66 to 3.79) | 4.39 (2.73 to 7.05) | |
| 30 to <100 | |||||
| events (%) | 10 (1.9%) | 11 (2.6%) | 10 (2.1%) | 14 (4.4%) | 9 (7.6%) |
| HR (95% CI)b | 1.27 (0.78 to 2.05) | 1.77 (1.16 to 2.70) | 2.29 (1.61 to 3.26) | 3.18 (2.06 to 4.93) | 5.56 (3.44 to 9.00) |
| 100 to <300 | |||||
| events (%) | 11 (4.4%) | 7 (3.2%) | 9 (3.6%) | 19 (9.9%) | |
| HR (95% CI)b | 2.19 (1.33 to 3.59) | 3.06 (1.97 to 4.76) | 3.96 (2.70 to 5.81) | 5.50 (3.50 to 8.66) | |
| ≥300 | |||||
| events (%) | 5 (3.0%) | 11 (8.9%) | 80 (19.4%) | ||
| HR (95% CI)b | 3.64 (2.20 to 6.03) | 5.10 (3.25 to 8.00) | 6.60 (4.65 to 9.36) | ||
aNo baseline-by-change interaction, P = 0.29. Main effect model: >1 category better versus no change: HR 0.67, 95% CI, 0.52 to 0.86, P = 0.002; 1 category better versus no change: HR 0.83, 95% CI 0.69 to 1.01, P = 0.061; 1 category worse versus no change: HR 1.40, 95% CI, 1.22 to 1.62, P < 0.0001; >1 category worse versus no change: HR 2.56, 95% CI 2.16 to 3.04, P < 0.0001.
bHazard ratios were adjusted for age, gender, body mass index, smoking, alcohol consumption, eGFR, plasma glucose, systolic and diastolic BP and HR at baseline, BP change and eGFR change within 2-year study (ONTARGET, TRANSCEND), treatment, and diagnosis at study entry such as coronary artery disease, peripheral artery disease, previous stroke, transient ischemic attack, and high-risk diabetes.
cNo baseline-by-change interaction, P = 0.11. Main effect model: >1 category better versus no change: HR 0.72, 95% CI, 0.53 to 0.98, P = 0.034; 1 category better versus no change: HR 0.80, 95% CI 0.63 to 1.02, P = 0.072; 1 category worse versus no change: HR 1.39, 95% CI, 1.16 to 1.67, P = 0.0003; >1 category worse versus no change: HR 2.61, 95% CI 2.12 to 3.22, P < 0.0001.
dPossible baseline-by-change interaction, P = 0.027. Model allowing for baseline-by-change interaction.
eNo baseline-by-change interaction, P = 0.16. Main effect model: >1 category better versus no change: HR 0.551, 95% CI 0.36 to 0.86, P = 0.008; 1 category better versus no change: HR 0.77, 95% CI 0.55 to 1.08, P = 0.13; 1 category worse versus no change: HR 1.39, 95% CI 1.03 to 1.88, P = 0.033; >1 category worse versus no change: HR 2.43, 95% CI 1.70 to 3.46, P < 0.0001.
DISCUSSION
This analysis establishes that changes in albuminuria predict mortality and cardiovascular and renal outcomes in treated patients aged 55 years or older with vascular disease or complicated diabetes at all levels of baseline albuminuria. The predictive value of increase >100% in albuminuria for mortality, cardiovascular death, cardiovascular and renal outcome was mostly independent of the degree of underlying albuminuria and of other cardiovascular risk factors. The predictive power of the decrease of >50% in albuminuria was found for all-cause mortality and cardiovascular and renal outcome, but not clearly for cardiovascular death.
The results were obtained in patients most of whom received standard cardioprotective therapies, including blockers of the renin angiotensin system, statins, beta blockers, and antiplatelet agents. Treatment allocation according to the ONTARGET and TRANSCEND study protocol was included in the analyses, adjusting for other covariates, but did not materially influence the results. Although the percentage of patients with increases in albuminuria was highest and with decreases in albuminuria lowest in the placebo group, the association of changes in albuminuria with mortality and cardiovascular outcome in the placebo arm was not different from that found in the active treatment groups (Figure 2).
Taken together, these findings suggest that changes in albuminuria are clinical indicators of changes in cardiovascular risk and renal in patients with atherosclerotic vascular disease. In the diabetic population, approximately one third of the whole study population, similar findings were observed (data not shown). However, our results cannot necessarily be extrapolated to other study populations, for example, patients with lower cardiovascular risk or patients with renal disease and reduced renal function and/or macroalbuminuria.
Previous trials have shown that baseline values of albuminuria are predictive of cardiovascular outcomes. But little is known about whether changes in albuminuria translate into changes in these outcomes.11,12,16,17 In 8206 hypertensive patients with electrocardiographic left ventricular hypertrophy followed for 4.8 years, patients with decreases in albuminuria during the first 6 months of the study had lower cardiovascular events than those with increases in albuminuria.11 No data on cardiovascular death and all-cause death were reported. In another study 1513 patients with type 2 diabetes with overt nephropathy were followed for 3.4 years and a reduction in albuminuria in the first 6 months was related to lower cardiovascular events. In this specific population, a reduction in albuminuria by 50% translated into a 18% reduction of cardiovascular risk,12 but no relation to cardiovascular or total mortality could be analyzed because of the low number of patients and events. In an observational study with 983 Pima Indians with type 2 diabetes changes in albuminuria were found to be of minimal predictive value beyond the latest measurement of albuminuria.16 In another retrospective analysis of 216 patients with diabetes, change in albuminuria was considered an integrated indicator for renal and cardiovascular risk reduction.17 Our results in 23,480 patients with vascular disease suggest that overall a reduction in albuminuria of 50% or more translated into a 15% decrease of mortality and conversely, and more striking, an increase in albuminuria by 100% or more translated into a 30% increase in cardiovascular events, 54% increase in cardiovascular death, and 48% increase in total mortality.
We confirm in this population the predictive power of baseline albuminuria, even below the threshold of microalbuminuria, for total mortality, cardiovascular and renal outcomes independent of classical risk factors and renal function.8,18 Compared with patients with albuminuria <10 mg/g, we observed an increased risk of total mortality and cardiovascular and renal events in patients with albuminuria between 10 and 30 mg/g creatinine, which is below the so-called normal threshold of microalbuminuria. In the current trial the incidence of total mortality was 1.85% per year in patients with UACR below 10 mg/g and 2.88% in those with UACR 10 to 30 mg/g. Likewise, an increase >100% in albuminuria was related to increased all-cause death (Table 3). These data question the currently applied cutoff level of microalbuminuria (and macroalbuminuria) and support the concept that any measurement of albuminuria should be considered as part of a continuous risk indicator of cardiovascular and renal disease.8,18,19,20
Our measurements of albuminuria were restricted to a single urine specimen at baseline and at the 2-year examination. Urine albumin excretion has considerable intraindividual variability and several measurements on consecutive days are preferable. However, such repeated measurements in a large prospective trial with over 30,000 participants face logistic and financial obstacles. Relying on one sample increases the variability of the measurement. On the other hand, the large sample size and the central laboratory assay of urine albumin would reduce the variability. Recently, it was documented that urinary albumin concentration or albumin to creatinine ratio measured once in a first morning sample are good predictors of 24-hour albumin excretion.18,21,22 Furthermore, single random urine sampling has been found to be useful for screening and predict cardiovascular mortality.23,24 Although it is possible that multiple measures are preferable, the clear relationships obtained in this and previous analyses using a single urine measurement emphasizes the robustness and clinical utility of such a simple approach.
Measurements of changes in albuminuria are subject to regression-to-mean phenomenon. To account for that phenomenon, we divided the population into five groups stratified according to albuminuria at baseline and analyzed the change in albuminuria separately in these groups. Furthermore, in a multivariate Cox-regression model the effect of change in albuminuria was analyzed with adjustment for albuminuria at baseline and covariates identified to be potential determinants of albuminuria were also included in the adjusted multivariate analysis. Because of the well-known high variability of albuminuria measurements, the use of relative scales (halving, doubling) may exaggerate small changes within the normal range. We additionally analyzed post hoc our data in a categorial manner, which confirmed the results of the analysis based on changes in albuminuria. Finally, we have calculated propensity scores for the three UACR change groups (data not given) and in all cases the results were similar to our preferred analyses, which are displayed in this article.
Our analysis did not intend to analyze determinants of changes in albuminuria. Our findings that changes in systolic BP, heart rate, and renal function were identified as independent determinants for changes in albuminuria are far from being conclusive, though in accordance with the literature.12,25,26 The time slot between the baseline and second urine sample was 2 years, which is by far too long to allow any meaningful analyses of clinical determinants for the changes in albuminuria.
We conclude that changes in albuminuria >100% during follow-up are valuable predictors of total mortality and cardiovascular and to lesser extent renal outcomes in people with atherosclerotic vascular disease independent of other risk factors including the initial level of albuminuria. In a population that is characterized by an increase of albuminuria over time due to the progressive nature of cardiovascular disease, changes in albuminuria clearly are indicative of a worse prognosis, and serial measurements of albuminuria characterize changes in cardiovascular and renal prognosis and all-cause mortality. Additional studies should examine whether adjusting cardiovascular therapy in response to adverse changes in albuminuria will improve outcomes.
CONCISE METHODS
Study Design and Population
The design of the ONTARGET and TRANSCEND study and detailed eligibility criteria have been described previously.13–15 In brief, 2001 through 2003 we enrolled patients aged 55 years or older with established atherosclerotic vascular disease, that is, coronary, cerebrovascular, or peripheral artery disease or diabetes with end-organ damage. Patients with serum creatinine above 3.0 mg/dl, any known renal disease, renal artery stenosis, or uncontrolled hypertension (>160 mmHg systolic or >100 mmHg diastolic) or macroalbuminuria (>300 mg/g creatinine) in TRANSCEND had to be excluded. When urine was analyzed centrally, macroalbuminuria was detected in 699 patients (ONTARGET 652 patients, TRANSCEND 47 patients).13 Before inclusion in ONTARGET, two thirds of the study participants had received an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker. In TRANSCEND only patients intolerant to ACE inhibitors were included.
Assessment of Albuminuria and Outcome
Urine albumin and creatinine concentration, and thereby the urinary albumin creatinine ratio (UACR), were measured centrally before run-in (baseline) and after 2 years from a spot urine (first morning urine).27 Urine albumin was measured by a turbidimetric method (Unicel DxC600 Synchron Systems, Beckman Coulter, Bea, CA). The coefficient of variation at 32.2 mg/L was 4.4% and at 105.5 mg/L was 2.4%. A human serum pool at a concentration of 10.9 mg/L gave a coefficient of variation of 9.2%, and at 159.8 mg/L gave a coefficient of variation of 2.7%. Creatinine in urine was measured centrally by a modified Jaffe method (Unicel DxC600 Synchron Systems). The coefficient of variation at 7.9 mg/L was 2.9%, and at 23.1 mg/L was 2.8%. Serum creatinine was measured locally at the study sites. From the serum creatinine, eGFR was calculated using the four-variable Modification of Diet in Renal Disease formula.28 Serum creatinine values below 0.3 mg/dl (36 at baseline, 119 at follow-up) or above 12.0 mg/dl (0 at baseline, 1 at follow-up) were deemed implausible and excluded from this analysis.27
The follow-up period for this analysis started with the 2-year measurement of albuminuria, and covered on average 32 months for the collection of outcomes. All main study outcomes (death, cardiovascular death, myocardial infarction, stroke, and hospitalization for heart failure) were adjudicated by a blinded central committee utilizing standard definitions. Adjudicators were unaware of treatment and albuminuria status.13 The primary composite cardiovascular outcome was death from cardiovascular causes, nonfatal myocardial infarction, nonfatal stroke, or hospitalization for heart failure.14,15 The composite renal outcome was defined as any dialysis or doubling of baseline serum creatinine.27 The protocol was approved by the ethics committees at each participating institution and regulatory authorities in each country. Each participant provided written informed consent. Both trials (ONTARGET, TRANSCEND) are registered at ClinicalTrials.gov number NCT00153101.
Statistical Analysis
Before study closeout, we specified the analysis plan for the renal outcomes and for the analysis reported here (September 2007). “Decrease of albuminuria” was defined as a reduction of UACR >50% (i.e., less than half), “minor change of albuminuria” as UACR change between −50% and +100%, and “increase of albuminuria” as an increase in UACR of >100% (i.e., more than double). The cutoff criteria “halving” and “doubling” of UACR take into account the skewed distribution of UACR and reflect that UACR was log-transformed before analysis. To identify determinants of change in albuminuria, multiple regression analysis with forward selection was applied with changes in systolic and diastolic BP, changes in heart rate, and changes in serum creatinine as potential predictors.
To account for the fact that patients with low UACR at baseline are more likely to show increases, or patients with high UACR are more likely to show decreases (regression-to-the-mean), the analysis of changes in albuminuria was always adjusted for baseline albuminuria. Albuminuria at baseline was categorized into five groups: UACR <10, ≥10 and <30, ≥30 and <100, ≥100 and <300, and ≥300 mg of albumine per g of urinary creatinine excretion.19 The associated hazard ratios for changes in albuminuria were calculated separately in each of the five subgroups categorized according to albuminuria at baseline. Hazard ratios were related to the group with UACR <10 mg/g creatinine at baseline and minor change of albuminuria. In the case of statistical interaction, the association of the change in albuminuria with outcome is presented by taking the interaction into account.
SAS 9.2 (SAS Institute) was used for statistical analysis. Cox regression was applied to assess the effects of baseline UACR and of changes in UACR on outcomes. Analyses were performed unadjusted and adjusted for age, sex, BMI, smoking, alcohol consumption, eGFR, plasma glucose, systolic and diastolic BP and HR at baseline, BP change and eGFR change within 2 years, study (ONTARGET, TRANSCEND), treatment (ramipril, telmisartan, combination, and placebo), and diagnosis for study entry such as coronary artery disease, peripheral artery disease, previous stroke, transient ischemic attack, and high-risk diabetes. Comparisons of categorical data were done by the χ2 test, and of continuous data by t tests. All P values are two-sided. Continuous data are presented as mean ± SD (UACR as geometric mean with interquartile range) and categorical data as frequencies and percentages. Data are presented for the two studies (ONTARGET and TRANSCEND) combined but the adjusted analyses took into account factors for study and randomized treatment (80 mg/d telmisartan, 10 mg/d ramipril, or both and 80 mg/d telmisartan or placebo, respectively).14,15
DISCLOSURES
R.E.S., J.F.M., G.M., M.W., T.K., and S.Y. report receiving consulting and lecture fees and research grants from Boehringer Ingelheim and from other companies manufacturing angiotensin receptor blockers, ACE inhibitors, and other BP-lowering drugs; H.S. is an employee of Boehringer Ingelheim.
Acknowledgments
Funding for this research was provided by Boehringer Ingelheim.
All authors were members of the trials steering committee and contributed to the discussions and interpretation of the data and to the writing of the report. The analysis was planned by R.S., J.M., H.S., T.K., and S.Y.; data were analyzed by H.S. and P.G. M.Q. was responsible for the laboratory analysis. All authors had full access to the data. No medical writer or other people were involved in the design, analysis, or writing of this manuscript. A full list of all investigators has been published.13
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
See related editorial, “The Fourth Dimension: Associations of Change in Albuminuria over Time,” on pages 1186–1188.
REFERENCES
- 1. Ruggenenti P, Perna A, Mosconi L, Pisoni R, Remuzzi G: Urinary protein excretion rate is the best independent predictor of ESRF in non-diabetic proteinuric chronic nephropathies. “Gruppo Italiano di Studi Epidemiologici in Nefrologia” (GISEN). Kidney Int 53: 1209–1216, 1998 [DOI] [PubMed] [Google Scholar]
- 2. de Zeeuw D, Parving HH, Henning RH: Microalbuminuria as an early marker for cardiovascular disease. J Am Soc Nephrol 17: 2100–2105, 2006 [DOI] [PubMed] [Google Scholar]
- 3. Mann JF, Gerstein HC, Yi QL, Franke J, Lonn EM, Hoogwerf BJ, Rashkow A, Yusuf S: Progression of renal insufficiency in type 2 diabetes with and without microalbuminuria: Results of the Heart Outcomes and Prevention Evaluation (HOPE) randomized study. Am J Kidney Dis 42: 936–942, 2003 [DOI] [PubMed] [Google Scholar]
- 4. Remuzzi G, Bertani T: Pathophysiology of progressive nephropathies. N Engl J Med 339: 1448–1456, 1998 [DOI] [PubMed] [Google Scholar]
- 5. Schmieder RE, Schrader J, Zidek W, Tebbe U, Paar WD, Bramlage P, Pittrow D, Bohm M: Low-grade albuminuria and cardiovascular risk: What is the evidence? Clin Res Cardiol 96: 247–257, 2007 [DOI] [PubMed] [Google Scholar]
- 6. Grimm RH, Jr., Svendsen KH, Kasiske B, Keane WF, Wahi MM: Proteinuria is a risk factor for mortality over 10 years of follow-up. MRFIT Research Group. Multiple Risk Factor Intervention Trial. Kidney Int Suppl 63: S10–S14, 1997 [PubMed] [Google Scholar]
- 7. Ninomiya T, Perkovic V, de Galan BE, Zoungas S, Pillai A, Jardine M, Patel A, Cass A, Neal B, Poulter N, Mogensen CE, Cooper M, Marre M, Williams B, Hamet P, Mancia G, Woodward M, Macmahon S, Chalmers J: Albuminuria and kidney function independently predict cardiovascular and renal outcomes in diabetes. J Am Soc Nephrol 20: 1813–1821, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hillege HL, Fidler V, Diercks GF, van Gilst WH, de Zeeuw D, van Veldhuisen DJ, Gans RO, Janssen WM, Grobbee DE, de Jong PE: Urinary albumin excretion predicts cardiovascular and noncardiovascular mortality in general population. Circulation 106: 1777–1782, 2002 [DOI] [PubMed] [Google Scholar]
- 9. Perkovic V, Verdon C, Ninomiya T, Barzi F, Cass A, Patel A, Jardine M, Gallagher M, Turnbull F, Chalmers J, Craig J, Huxley R: The relationship between proteinuria and coronary risk: A systematic review and meta-analysis. PLoS Med 5: e207, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Gerstein HC, Mann JF, Yi Q, Zinman B, Dinneen SF, Hoogwerf B, Halle JP, Young J, Rashkow A, Joyce C, Nawaz S, Yusuf S: Albuminuria and risk of cardiovascular events, death, and heart failure in diabetic and nondiabetic individuals. JAMA 286: 421–426, 2001 [DOI] [PubMed] [Google Scholar]
- 11. Ibsen H, Olsen MH, Wachtell K, Borch-Johnsen K, Lindholm LH, Mogensen CE, Dahlof B, Devereux RB, de Faire U, Fyhrquist F, Julius S, Kjeldsen SE, Lederballe-Pedersen O, Nieminen MS, Omvik P, Oparil S, Wan Y: Reduction in albuminuria translates to reduction in cardiovascular events in hypertensive patients: Losartan intervention for endpoint reduction in hypertension study. Hypertension 45: 198–202, 2005 [DOI] [PubMed] [Google Scholar]
- 12. de Zeeuw D, Remuzzi G, Parving HH, Keane WF, Zhang Z, Shahinfar S, Snapinn S, Cooper ME, Mitch WE, Brenner BM: Albuminuria, a therapeutic target for cardiovascular protection in type 2 diabetic patients with nephropathy. Circulation 110: 921–927, 2004 [DOI] [PubMed] [Google Scholar]
- 13. Teo K, Yusuf S, Sleight P, Anderson C, Mookadam F, Ramos B, Hilbrich L, Pogue J, Schumacher H: Rationale, design, and baseline characteristics of 2 large, simple, randomized trials evaluating telmisartan, ramipril, and their combination in high-risk patients: The Ongoing Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial/Telmisartan Randomized Assessment Study in ACE Intolerant Subjects with Cardiovascular Disease (ONTARGET/TRANSCEND) trials. Am Heart J 148: 52–61, 2004 [DOI] [PubMed] [Google Scholar]
- 14. Yusuf S, Teo K, Anderson C, Pogue J, Dyal L, Copland I, Schumacher H, Dagenais G, Sleight P: Effects of the angiotensin-receptor blocker telmisartan on cardiovascular events in high-risk patients intolerant to angiotensin-converting enzyme inhibitors: A randomised controlled trial. Lancet 372: 1174–1183, 2008 [DOI] [PubMed] [Google Scholar]
- 15. Yusuf S, Teo KK, Pogue J, Dyal L, Copland I, Schumacher H, Dagenais G, Sleight P, Anderson C: Telmisartan, ramipril, or both in patients at high risk for vascular events. N Engl J Med 358: 1547–1559, 2008 [DOI] [PubMed] [Google Scholar]
- 16. Pavkov ME, Knowler WC, Hanson RL, Bennett PH, Nelson RG: Predictive power of sequential measures of albuminuria for progression to ESRD or death in Pima Indians with type 2 diabetes. Am J Kidney Dis 51: 759–766, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Araki S, Haneda M, Koya D, Hidaka H, Sugimoto T, Isono M, Isshiki K, Chin-Kanasaki M, Uzu T, Kashiwagi A: Reduction in microalbuminuria as an integrated indicator for renal and cardiovascular risk reduction in patients with type 2 diabetes. Diabetes 56: 1727–1730, 2007 [DOI] [PubMed] [Google Scholar]
- 18. Gansevoort RT, Verhave JC, Hillege HL, Burgerhof JG, Bakker SJ, de Zeeuw D, de Jong PE: The validity of screening based on spot morning urine samples to detect subjects with microalbuminuria in the general population. Kidney Int Suppl S28–S35, 2005 [DOI] [PubMed] [Google Scholar]
- 19. Levey AS, Cattran D, Friedman A, Miller WG, Sedor J, Tuttle K, Kasiske B, Hostetter T: Proteinuria as a surrogate outcome in CKD: report of a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration. Am J Kidney Dis 54: 205–226, 2009 [DOI] [PubMed] [Google Scholar]
- 20. Ruggenenti P, Remuzzi G: Time to abandon microalbuminuria? Kidney Int 70: 1214–1222, 2006 [DOI] [PubMed] [Google Scholar]
- 21. Eshoj O, Feldt-Rasmussen B, Larsen ML, Mogensen EF: Comparison of overnight, morning and 24-hour urine collections in the assessment of diabetic microalbuminuria. Diabet Med 4: 531–533, 1987 [DOI] [PubMed] [Google Scholar]
- 22. Witte EC, Lambers Heerspink HJ, de Zeeuw D, Bakker SJ, de Jong PE, Gansevoort R: First morning voids are more reliable than spot urine samples to assess microalbuminuria. J Am Soc Nephrol 20: 436–443, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Yuyun MF, Khaw KT, Luben R, Welch A, Bingham S, Day NE, Wareham NJ: 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 33: 189–198, 2004 [DOI] [PubMed] [Google Scholar]
- 24. Garg AX, Kiberd BA, Clark WF, Haynes RB, Clase CM: Albuminuria and renal insufficiency prevalence guides population screening: results from the NHANES III. Kidney Int 61: 2165–2175, 2002 [DOI] [PubMed] [Google Scholar]
- 25. Mann JF, Schmieder RE, Dyal L, McQueen MJ, Schumacher H, Pogue J, Wang X, Probstfield JL, Avezum A, Cardona-Munoz E, Dagenais GR, Diaz R, Fodor G, Maillon JM, Ryden L, Yu CM, Teo KK, Yusuf S: Effect of telmisartan on renal outcomes: A randomized trial. Ann Intern Med 151: 1–10, W1–W2, 2009 [DOI] [PubMed] [Google Scholar]
- 26. Bohm M, Reil JC, Danchin N, Thoenes M, Bramlage P, Volpe M: Association of heart rate with microalbuminuria in cardiovascular risk patients: Data from I-SEARCH. J Hypertens 26: 18–25, 2008 [DOI] [PubMed] [Google Scholar]
- 27. Mann JF, Schmieder RE, McQueen M, Dyal L, Schumacher H, Pogue J, Wang X, Maggioni A, Budaj A, Chaithiraphan S, Dickstein K, Keltai M, Metsarinne K, Oto A, Parkhomenko A, Piegas LS, Svendsen TL, Teo KK, Yusuf S: Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): A multicentre, randomised, double-blind, controlled trial. Lancet 372: 547–553, 2008 [DOI] [PubMed] [Google Scholar]
- 28. Stevens LA, Coresh J, Greene T, Levey AS: Assessing kidney function–measured and estimated glomerular filtration rate. N Engl J Med 354: 2473–2483, 2006 [DOI] [PubMed] [Google Scholar]