Abstract
Introduction
This study aims to investigate the association between change in urine albumin-to-creatinine ratio (UACR) and clinical outcomes in patients with chronic kidney disease (CKD) and type 2 diabetes.
Research design and methods
Adult patients with elevated UACR (≥30 mg/g in initial testing) after the diagnosis of type 2 diabetes and CKD were identified from the Optum electronic health records database (01/2007–09/2021). UACR change from initial to last test (6–24 months) was categorized as >30% decrease, stable (−30% to 30%), or >30% increase. Risk of all-cause mortality, composite cardiovascular (CV) outcome (CV death, myocardial infarction, stroke, and hospitalization for heart failure), and CKD progression (≥40% decline in estimated glomerular filtration rate or kidney failure) were estimated with Cox proportional hazard models adjusted for baseline characteristics.
Results
Compared with patients with a stable UACR (n=35 117), those with a >30% UACR decrease (n=89 562) had lower risk of all-cause mortality (adjusted HR (aHR)=0.93, 95% CI 0.90 to 0.96), composite CV outcomes (aHR=0.93, 95% CI 0.90 to 0.95), and CKD progression (aHR=0.84, 95% CI 0.81 to 0.86) (all p<0.001), and patients with a >30% UACR increase (n=35 703) had higher risk of each endpoint (aHR=1.24, 95% CI 1.19 to 1.28; aHR=1.24, 95% CI 1.20 to 1.28; and aHR=1.41, 95% CI 1.36 to 1.46, respectively; all p<0.001).
Conclusions
In patients with CKD and type 2 diabetes, a >30% UACR decrease was associated with lower risk of mortality, CV events, and CKD progression, whereas a >30% UACR increase was associated with higher risk of these clinical outcomes. These findings highlight the importance of albuminuria monitoring and potential clinical benefits of targeted UACR reductions in this population.
Keywords: Kidney Diseases; Diabetes Mellitus, Type 2; Albuminuria
WHAT IS ALREADY KNOWN ON THIS TOPIC.
Guideline-recommended treatments for patients with chronic kidney disease (CKD) and type 2 diabetes have been shown to reduce urine albumin-to-creatinine ratio (UACR); however, the clinical impact of this reduction has not been fully quantified.
WHAT THIS STUDY ADDS
This study aimed to examine the association of change in UACR with kidney and cardiovascular (CV) outcomes and mortality in this patient population.
Achieving a 30% decrease in UACR or maintaining a stable UACR was associated with improved CV outcomes and lower risks of CKD progression and all-cause mortality compared with having a 30% increase in UACR.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These results suggest better adherence to guideline-recommended UACR monitoring and management may improve CV and kidney outcomes in patients with CKD and type 2 diabetes.
Introduction
Chronic kidney disease (CKD) affects one-quarter to one-third of adults with diabetes in the USA.1 2 Patients with CKD and type 2 diabetes have an elevated risk of cardiovascular (CV) disease and can eventually progress to end-stage kidney disease (ESKD).3–5 Furthermore, the co-occurrence of type 2 diabetes and CKD accelerates disease progression, with higher mortality than CKD alone.6
Albuminuria is an early indicator of CKD that is commonly defined as an elevated urine albumin-to-creatinine ratio (UACR), which measures the level of excreted albumin relative to the creatinine.7 UACR is used in conjunction with estimated glomerular filtration rate (eGFR) to diagnose end-stage CKD, as well as to estimate CKD progression risk according to the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines.8
Given that elevated UACR is associated with increased risk of kidney failure, CV outcomes, and mortality,9–11 it is a key measure to include as part of comprehensive kidney–heart risk management.8 12 The American Diabetes Association (ADA) recommends targeting a decrease of ≥30% for patients with CKD with UACR ≥300 mg/g to slow CKD progression.12 In 2019, Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and the Chronic Kidney Disease Prognosis Consortium (CKD-PC) published two meta-analyses that supported the surrogacy of UACR change for CKD progression and ESKD.13 14 Specifically, in the meta-analysis of 41 randomized clinical trials, a 30% decrease in UACR reduced the risk for ESKD (eGFR <15 mL/min/1.73 m2 or doubling of serum creatinine) by 27%, with stronger association for patients with baseline UACR >30 mg/g.14 Similarly, in another meta-analysis of individual-level data from 28 observational study cohorts, a 30% decrease in UACR during a 2-year baseline period was associated with a 22% reduction in the risk of ESKD.13
Although the relationship between UACR change and kidney outcomes has been established by these previous studies, less is known about its impact on CV events or all-cause mortality—both of which are more common than kidney failure in patients with CKD.15 16 To address this gap in the literature, the present study aimed to assess the association between patterns of UACR change and kidney outcomes, CV outcomes, and mortality in patients with CKD and type 2 diabetes with albuminuria.
Methods
Data source
This was a retrospective, longitudinal, observational cohort study using data from the Optum electronic health records (EHRs) database (01/2007–09/2021), which encompasses more than 150 000 medical providers at over 2000 hospitals and 7000 clinics in the USA. The database contains information on demographic characteristics, type of healthcare provider, medical history, diagnoses, detailed area of care during hospitalization, in-hospital procedures, inpatient medications, physician prescriptions, and laboratory data for all types of encounters within the network.
Study population and design
Adult patients with CKD and type 2 diabetes who had albuminuria (ie, UACR ≥30 mg/g) were included in the current study. Type 2 diabetes was identified using a modified EMERGE algorithm17 that is applicable to EHRs and incorporates International Classification of Diseases diagnosis codes, type 2 diabetes medications, and laboratory measurements. CKD on or after the date of type 2 diabetes diagnosis was identified based on the 2020 KDIGO clinical guidelines18 using laboratory measurements and diagnosis codes. Patients were required to have at least one elevated UACR test (≥30 mg/g in the initial testing) on or after CKD diagnosis following type 2 diabetes diagnosis, as well as a follow-up UACR measurement (the last UACR measurement within 0.5–2 years after the initial UACR test). Patients were also required to have continuous eligibility from 1 year before to 2 years after the initial UACR test, with continuous enrollment windows defined as periods of time with records indicating any clinical activity (ie, disease diagnoses, procedures, laboratory visits, medication prescriptions, medication administrations, and other clinical visits and encounters) with a gap time less than 6 months. The 6-month period preceding the date of follow-up UACR test was defined as the baseline period.
Patients were excluded from the study if they had records indicating pre-existing albuminuria/proteinuria or ESKD (including chronic dialysis and kidney transplant) or pre-existing conditions associated with kidney function decline (including systemic lupus erythematosus, polycystic kidney disease, and kidney cancer) during the 1-year period before the initial UACR test. Additionally, patients with pre-existing myocardial infarction or stroke on or before the date of the follow-up UACR test were excluded from the analysis of CV outcomes, while those with pre-existing ESKD on or before this date or without a valid eGFR measurement during the baseline period were excluded from the analysis of kidney disease progression.
UACR change patterns
UACR measurements were either directly reported values from UACR laboratory tests or calculated from the reported urine albumin and creatinine concentrations measured on the same date. UACR change was assessed as the per cent change in UACR from the initial UACR test to the follow-up UACR test and was categorized as a >30% increase, stable (30% decrease to 30% increase), or a >30% decrease.12–14 UACR change was also evaluated as the transition from microalbuminuria (UACR ≥30 to <300 mg/g) or macroalbuminuria (UACR ≥300 mg/g) at the initial UACR test to normoalbuminuria (UACR <30 mg/g), microalbuminuria, or macroalbuminuria at the follow-up UACR test.
Clinical outcomes
Clinical outcomes including all-cause mortality, a composite CV outcome (CV death, myocardial infarction, stroke, or hospitalization for heart failure (HHF)), individual CV events, a composite kidney disease progression outcome (≥40% eGFR decline or kidney failure), and kidney failure were evaluated from the follow-up UACR test until the end of follow-up. CV death was defined as a death event with a diagnosis record of ischemic heart diseases (including myocardial infarction), heart failure, cardiac arrest, arrhythmia, or stroke within 60 days before or after the death record. Kidney failure was defined as developing ESKD, including being on chronic dialysis, receipt of kidney transplantation, or reaching eGFR <15 mL/min/1.73 m2.
Statistical analysis
Patient demographics and clinical characteristics were summarized for the overall patient population and by initial UACR category (ie, ≥30 to <300 mg/g, ≥300 to <1000 mg/g, and ≥1000 mg/g) and UACR change category. Clinical outcomes were evaluated in the overall population and by UACR change categories using Kaplan-Meier curves. The log-rank test was used to compare survival curves across UACR change categories for each clinical outcome.
The association between UACR change categories and clinical outcomes was further evaluated with a Cox proportional hazards model adjusting for baseline demographics (age, sex, race, region, type of insurance, year of follow-up UACR test) and clinical characteristics (eGFR, body mass index, hemoglobin A1c (HbA1c), time from initial UACR test to follow-up UACR test, and comorbidities, including hypertension, ischemic heart disease, heart failure, stroke, diabetes-related microvascular complications, hyperlipidemia, anemia, and acidosis). Predictors of UACR change patterns were assessed using a multinomial logistic regression model with the stable UACR group as the reference category.
All statistical analyses were conducted using SAS (SAS Studio Release: 3.8 (Enterprise Edition), SAS Institute Inc, Cary, North Carolina, USA) and R V.4.2.2 (R Foundation for Statistical Computing). P <0.05 was considered statistically significant for all analyses, and tests were two-tailed.
Results
Characteristics of the study population
A total of 160 382 patients with CKD and type 2 diabetes who had an elevated UACR and a follow-up UACR measurement within 0.5–2 years after the initial UACR test were included in the study (online supplemental figure S1). The demographic and clinical characteristics of the study population by UACR change category are shown in table 1. Mean age (SD) at the date of follow-up UACR test for overall population was 65.9 (11.9) years. Most patients were Caucasian (82.0%), resided in the Midwest (66.6%), and had commercial insurance (57.5%). Mean (SD) eGFR was 76.2 (23.5) mL/min/1.73 m2 and median (first quartile (Q1), third quartile (Q3)) UACR was 52.7 (37.2, 98.0) mg/g. Over two-thirds of patients (68.9%) had an eGFR stage of G1 (eGFR ≥90 mL/min/1.73 m2) or G2 (eGFR between 60 and <90 mL/min/1.73 m2); 22.4% of patients were in G3 (14.5% in G3a (eGFR between 45 and <60 mL/min/1.73 m2) and 7.9% in G3b (eGFR between 30 and <45 mL/min/1.73 m2)); and 2.4% patients were in G4/5 (eGFR <30 mL/min/1.73 m2).
Table 1.
Patient baseline demographic and clinical characteristics by UACR change category
| UACR change category* | |||||
|
Total
N=160 382 |
>30% decrease
N=89 562 |
30% decrease to 30% increase
N=35 117 |
>30% increase
N=35 703 |
P value | |
| Demographics† | |||||
| Age, years | <0.001* | ||||
| Mean (SD) | 65.9 (11.9) | 65.0 (12.0) | 66.7 (11.8) | 67.3 (11.7) | |
| Sex | <0.001* | ||||
| Female | 82 753 (51.6%) | 49 586 (55.4%) | 16 777 (47.8%) | 16 390 (45.9%) | |
| Male | 77 586 (48.4%) | 39 946 (44.6%) | 18 334 (52.2%) | 19 306 (54.1%) | |
| Unknown | 43 (0.0%) | 30 (0.0%) | 6 (0.0%) | 7 (0.0%) | |
| Race‡ | <0.001* | ||||
| African American | 18 306 (11.4%) | 9953 (11.1%) | 3968 (11.3%) | 4385 (12.3%) | |
| Asian | 3337 (2.1%) | 1865 (2.1%) | 732 (2.1%) | 740 (2.1%) | |
| Caucasian | 131 479 (82.0%) | 73 679 (82.3%) | 28 830 (82.1%) | 28 970 (81.1%) | |
| Other/unknown | 7260 (4.5%) | 4065 (4.5%) | 1587 (4.5%) | 1608 (4.5%) | |
| Region | <0.001* | ||||
| Midwest | 106 834 (66.6%) | 59 879 (66.9%) | 22 905 (65.2%) | 24 050 (67.4%) | |
| South | 22 654 (14.1%) | 12 211 (13.6%) | 5660 (16.1%) | 4783 (13.4%) | |
| Northeast | 18 309 (11.4%) | 10 478 (11.7%) | 3870 (11.0%) | 3961 (11.1%) | |
| West | 9050 (5.6%) | 5015 (5.6%) | 1929 (5.5%) | 2106 (5.9%) | |
| Other/unknown | 3535 (2.2%) | 1979 (2.2%) | 753 (2.1%) | 803 (2.2%) | |
| Insurance type | <0.001* | ||||
| Commercial | 83 273 (57.5%) | 48 018 (59.1%) | 17 905 (56.7%) | 17 350 (54.3%) | |
| Medicaid | 8034 (5.6%) | 4635 (5.7%) | 1623 (5.1%) | 1776 (5.6%) | |
| Medicare | 48 892 (33.8%) | 25 992 (32.0%) | 11 100 (35.1%) | 11 800 (36.9%) | |
| Other payer type | 939 (0.6%) | 555 (0.7%) | 202 (0.6%) | 182 (0.6%) | |
| Uninsured | 2127 (1.5%) | 1186 (1.5%) | 460 (1.5%) | 481 (1.5%) | |
| Unknown | 1462 (1.0%) | 800 (1.0%) | 302 (1.0%) | 360 (1.1%) | |
| Year of follow-up UACR test | <0.001* | ||||
| 2008–2013 | 53 892 (33.6%) | 29 152 (32.5%) | 12 136 (34.6%) | 12 604 (35.3%) | |
| 2014–2019 | 106 490 (66.4%) | 60 410 (67.5%) | 22 981 (65.4%) | 23 099 (64.7%) | |
| Durations | |||||
| Time from initial to follow-up UACR test (months) | <0.001* | ||||
| Mean (SD) | 16.9 (4.6) | 16.8 (4.5) | 16.6 (4.6) | 17.2 (4.6) | |
| Follow-up time after follow-up UACR test (years) | <0.001* | ||||
| Mean (SD) | 3.4 (2.2) | 3.4 (2.2) | 3.4 (2.1) | 3.3 (2.1) | |
| Clinical characteristics † | |||||
| eGFR stage§ | <0.001* | ||||
| G1 | 50 327 (31.4%) | 29 682 (33.1%) | 10 917 (31.1%) | 9728 (27.2%) | |
| G2 | 60 209 (37.5%) | 34 248 (38.2%) | 13 061 (37.2%) | 12 900 (36.1%) | |
| G3a | 23 186 (14.5%) | 12 035 (13.4%) | 5076 (14.5%) | 6075 (17.0%) | |
| G3b | 12 639 (7.9%) | 6157 (6.9%) | 2874 (8.2%) | 3608 (10.1%) | |
| G4 | 3461 (2.2%) | 1518 (1.7%) | 783 (2.2%) | 1160 (3.2%) | |
| G5 | 400 (0.2%) | 145 (0.2%) | 98 (0.3%) | 157 (0.4%) | |
| Unknown | 10 160 (6.3%) | 5777 (6.5%) | 2308 (6.6%) | 2075 (5.8%) | |
| Body mass index (kg/m2) | <0.001* | ||||
| Mean (SD) | 34.0 (7.8) | 34.1 (7.8) | 33.7 (7.7) | 34.0 (7.9) | |
| Missing (%) | 16 204 (10.1%) | 8800 (9.8%) | 3661 (10.4%) | 3743 (10.5%) | |
| Laboratory measures ¶ | |||||
| eGFR (mL/min/1.73 m2) | <0.001* | ||||
| Mean (SD) | 76.2 (23.5) | 77.8 (22.9) | 75.8 (23.6) | 72.7 (24.4) | |
| Median (Q1, Q3) | 77.9 (58.8, 95.3) | 79.7 (61.1, 96.2) | 77.4 (58.3, 95.2) | 73.4 (54.1, 92.7) | |
| Missing (%) | 10 160 (6.3%) | 5777 (6.5%) | 2308 (6.6%) | 2075 (5.8%) | |
| HbA1c (%) | <0.001* | ||||
| Mean (SD) | 7.3 (1.5) | 7.2 (1.4) | 7.3 (1.5) | 7.5 (1.6) | |
| Missing (%) | 8494 (5.3%) | 4618 (5.2%) | 1853 (5.3%) | 2023 (5.7%) | |
| Initial UACR (mg/g) | <0.001* | ||||
| Mean (SD) | 146.1 (445.5) | 148.3 (481.5) | 153.1 (461.5) | 133.8 (316.8) | |
| Median (Q1, Q3) | 52.7 (37.2, 98.0) | 55.0 (38.3, 101.0) | 47.9 (35.6, 86.3) | 52.0 (37.0, 98.2) | |
| Follow-up UACR (mg/g) | <0.001* | ||||
| Mean (SD) | 162.3 (715.4) | 54.5 (424.8) | 166.4 (632.2) | 428.7 (1163.3) | |
| Median (Q1, Q3) | 36.0 (16.5, 94.1) | 18.4 (10.7, 32.0) | 48.0 (33.9, 91.5) | 147.1 (81.0, 344.7) | |
| Comorbidities† | |||||
| Hypertension | 119 844 (74.7%) | 65 967 (73.7%) | 26 317 (74.9%) | 27 560 (77.2%) | <0.001* |
| Ischemic heart diseases | 27 744 (17.3%) | 14 538 (16.2%) | 6201 (17.7%) | 7005 (19.6%) | <0.001* |
| Heart failure | 11 382 (7.1%) | 5945 (6.6%) | 2358 (6.7%) | 3079 (8.6%) | <0.001* |
| Stroke | 2872 (1.8%) | 1501 (1.7%) | 602 (1.7%) | 769 (2.2%) | <0.001* |
| Diabetic ketoacidosis | 116 (0.1%) | 50 (0.1%) | 26 (0.1%) | 40 (0.1%) | <0.01* |
| Diabetes-related microvascular complications | 26 058 (16.2%) | 14 326 (16.0%) | 5340 (15.2%) | 6392 (17.9%) | <0.001* |
| Hyperlipidemia | 114 473 (71.4%) | 64 310 (71.8%) | 25 022 (71.3%) | 25 141 (70.4%) | <0.001* |
| Hyperkalemia | 2503 (1.6%) | 1224 (1.4%) | 548 (1.6%) | 731 (2.0%) | <0.001* |
| Hypoglycemia | 762 (0.5%) | 423 (0.5%) | 153 (0.4%) | 186 (0.5%) | 0.252 |
| Hyponatremia | 2240 (1.4%) | 1116 (1.2%) | 481 (1.4%) | 643 (1.8%) | <0.001* |
| Anemia (non-hereditary) | 18 327 (11.4%) | 9781 (10.9%) | 3915 (11.1%) | 4631 (13.0%) | <0.001* |
| Acidosis | 857 (0.5%) | 429 (0.5%) | 173 (0.5%) | 255 (0.7%) | <0.001* |
| Acute kidney injury | 3591 (2.2%) | 1742 (1.9%) | 723 (2.1%) | 1126 (3.2%) | <0.001* |
| Volume depletion | 2111 (1.3%) | 1135 (1.3%) | 422 (1.2%) | 554 (1.6%) | <0.001* |
| Edema | 9275 (5.8%) | 4942 (5.5%) | 1909 (5.4%) | 2424 (6.8%) | <0.001* |
| Urinary tract infections | 10 008 (6.2%) | 5457 (6.1%) | 2059 (5.9%) | 2492 (7.0%) | <0.001* |
*UACR change was calculated as percentage change in UACR from initial UACR to the follow-up UACR test, defined as the last UACR test within 0.5–2 years after the initial UACR test.
† Patient demographic characteristics were assessed on the date of follow-up UACR test and clinical characteristics were assessed during the baseline period.
‡Race data were patient-reported or provider documented.
§ eGFR categories were as follows: G1—normal or high: eGFR ≥90 mL/min/1.73 m2; G2—mildly decreased: eGFR between 60 and <90 mL/min/1.73 m2; G3a—mildly to moderately decreased: eGFR between 45 and <60 mL/min/1.73 m2; G3b—moderately to severely decreased: eGFR between 30 and <45 mL/min/1.73 m2; G4—severely decreased: eGFR between 15 and <30 mL/min/1.73 m2; G5—kidney failure: eGFR <15 mL/min/1.73 m2.
¶For laboratory measures, the most recent records before the date of follow-up UACR test were used.
UACR, urine albumin-to-creatinine ratio.eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; Q1, first quartile; Q3, third quartile.
bmjdrc-13-5-s001.pdf (450KB, pdf)
The most common comorbid condition was hypertension, which was present in nearly three-quarters (74.7%) of patients. Accordingly, the majority (73.6%) of patients were using antihypertensives at the initial UACR test, among which most (73.3%) of the patients used at least one type of ACE inhibitors/angiotensin receptor blockers (online supplemental table S1). By contrast, <5% of patients used antidiabetes therapies with a UACR-lowering effect (glucagon-like peptide 1 receptor agonists (GLP-1 RAs), 4.8%; sodium-glucose cotransporter-2 inhibitors (SGLT2is), 3.3%), with similar proportions across initial UACR categories.
UACR change patterns
The median time from initial to follow-up UACR tests was approximately 17 months for all initial UACR categories (table 1). There were 89 562 patients (55.8%) who experienced a >30% decrease in UACR level and 35 703 (22.3%) who had a >30% increase (table 2).
Table 2.
Summary of UACR change patterns by initial UACR level
| Initial UACR category | ||||
|
Total
N=160 382 |
30–<300 mg/g
N=146 890 |
300–<1000 mg/g
N=9928 |
≥1000 mg/g
N=3564 |
|
| UACR change category | ||||
| >30% decrease | 89 562 (55.8%) | 82 016 (55.8%) | 5678 (57.2%) | 1868 (52.4%) |
| 30% decrease to 30% increase | 35 117 (21.9%) | 32 183 (21.9%) | 1937 (19.5%) | 997 (28.0%) |
| >30% increase | 35 703 (22.3%) | 32 691 (22.3%) | 2313 (23.3%) | 699 (19.6%) |
UACR, urine albumin-to-creatinine ratio.
Among 146 890 patients with microalbuminuria at the initial UACR test, 72 807 (49.6%) had a reversal to normoalbuminuria, 67 026 (45.6%) remained in the microalbuminuria category, and 7057 (4.8%) progressed to macroalbuminuria (table 3). In contrast, among 13 492 patients with macroalbuminuria at the initial UACR test, 6076 reverted to normoalbuminuria (1788; 13.3%) or microalbuminuria (4288; 31.8%) and 7416 (55%) remained in the macroalbuminuria category.
Table 3.
Summary of change in UACR clinical category from initial to follow-up UACR test
| Initial UACR categories | |||
|
Total
N=160 382 |
Microalbuminuria
N=146 890 |
Macroalbuminuria
N=13 492 |
|
| Follow-up UACR clinical category* | |||
| Normoalbuminuria | 74 595 (46.5%) | 72 807 (49.6%) | 1788 (13.3%) |
| Microalbuminuria | 71 314 (44.5%) | 67 026 (45.6%) | 4288 (31.8%) |
| Macroalbuminuria | 14 473 (9.0%) | 7057 (4.8%) | 7416 (55.0%) |
*UACR categories were defined as follows: normoalbuminuria, UACR <30 mg/g; microalbuminuria, UACR ≥30 and <300 mg/g; macroalbuminuria, UACR ≥300 mg/g.
UACR, urine albumin-to-creatinine ratio.
Association between UACR change and all-cause mortality
The median follow-up time after the follow-up UACR test was 3.4 years for patients with >30% decrease in UACR, 3.4 years for patients with stable UACR, and 3.3 years for patients with >30% increase in UACR (table 1). Patients with a >30% decrease had the longest overall survival (median time to death, 10.3 years), followed by patients with a stable UACR (9.1 years) and patients with a >30% increase (7.8 years), with the trend persisting over the follow-up period (figure 1A). The associations were consistent after adjusting for demographics, clinical characteristics, and comorbidities at baseline. Compared with patients with a stable UACR, those with a >30% decrease had a significantly lower risk of all-cause mortality (adjusted HR (aHR)=0.93, 95% CI 0.90 to 0.96; p<0.001, while patients with a >30% increase had a significantly higher risk of all-cause mortality (aHR=1.24, 95% CI 1.19 to 1.28; p<0.001).
Figure 1.
Overall survival (a), CV event-free survival (b), kidney disease progression-free survival (c), and kidney failure-free survival (d) by UACR change category.CV, cardiovascular; OS, overall survival; UACR, urine albumin-to-creatinine ratio.
Association between UACR change and CV outcomes
The median time to the composite CV outcome (ie, CV mortality, myocardial infarction, stroke, HHF) was 11.1 years for patients with a >30% decrease in UACR, 9.9 years for patients with a stable UACR, and 8.1 years for patients with a >30% increase in UACR (figure 1B). Compared with patients in the stable UACR group, those with a >30% decrease had a significantly lower adjusted risk of the composite CV outcome (aHR=0.93, 95% CI 0.90 to 0.95; p<0.001), whereas those with a >30% increase had a significantly higher risk (aHR=1.24, 95% CI 1.20 to 1.28; p<0.001). A similar association was observed between UACR change and individual CV outcomes, including CV mortality, myocardial infarction, stroke, and HHF (online supplemental figure S2).
Association between UACR change and kidney disease progression
The median time to kidney disease progression was 11.4 years for patients with a >30% decrease in UACR, 9.2 years for patients with a stable UACR, and 6.8 years for patients with a >30% increase in UACR (figure 1C). Compared with patients in the stable UACR group, those with a >30% decrease had a significantly lower adjusted risk of kidney disease progression (aHR=0.84, 95% CI 0.81 to 0.86; p<0.001), whereas patients with >30% increase in UACR had a significantly higher risk (aHR=1.41, 95% CI 1.36 to 1.46; p<0.001). A similar association was observed between UACR change and kidney failure (figure 1D).
Factors associated with UACR change patterns
Male sex, non-Caucasian race, more advanced CKD stage, and hypertension were associated with lower odds of having a >30% UACR decrease, while younger age, year of follow-up UACR test after 2014, and well-controlled HbA1c (<6.5%) were associated with higher odds of having a >30% UACR decrease (online supplemental table S2). Higher levels of HbA1c showed a dose-dependent association with higher odds of >30% UACR increase. Other patient characteristics associated with 30% UACR increase included male sex, non-Caucasian race, more advanced CKD stage, and hypertension.
Discussion
In this large real-world cohort study of patients with CKD and type 2 diabetes with elevated UACR, we found that an increasing UACR was associated with a higher risk of kidney disease progression (including kidney failure), CV events, and all-cause mortality. Conversely, achieving a >30% decrease in UACR was associated with reduced risks of these outcomes. These results further demonstrated the potential benefits of UACR monitoring and management in patients with CKD and type 2 diabetes. Furthermore, the findings provide real-world evidence that supports the recommendation of American Diabetes Association (ADA) clinical guidelines12 to target a ≥30% decrease in urinary albumin for those with UACR levels ≥300 mg/g.
The association between elevated UACR and CV events has been extensively described,9 19–21 and an improvement in CV outcomes is a known effect of UACR-lowering pharmacotherapies.22 Less is known, however, about the change in the UACR over time, and its effect on CV outcomes. Our study results address this gap and further improve the chain of evidence, showing a strong and consistent association between change in UACR and a composite CV outcome along with its individual components, as well as all-cause mortality. Our results are consistent with those reported in a post-hoc analysis of the LEADER trial, in which a >30% UACR decrease was shown to be associated with a lower risk of major CV events.23
Change in UACR has been proposed as a surrogate endpoint for predicting kidney outcomes in patients with diabetes and kidney diseases.24 25 The value of such surrogacy is that it can shorten study duration and allow assessment of clinically important events (eg, CKD progression, for which ESKD is a frequently used endpoint) at earlier stages of disease.25 Recent large-scale meta-analyses evaluating the association between UACR change and kidney outcomes found that a 30% decrease in UACR was associated with a 27% lower risk of CKD progression and 22% lower risk of ESKD, with stronger associations observed in patients with higher baseline UACR (>30 or ≥300 mg/g).13 14 Retrospective studies of patients with type 2 diabetes in Japan also found that a ≥30% increase in UACR was associated with an increased risk of kidney disease progression.26 27 Although we used slightly different definitions, our results are broadly consistent with these earlier reports: a >30% UACR decrease was associated with a 16% lower risk of kidney disease progression and 19% lower risk of kidney failure.
The present study showed that both reducing albuminuria and maintaining a stable UACR level are associated with lower risk of adverse clinical outcomes compared with having a UACR increase. This is in line with the clinical guidelines targeting UACR for CKD management,8 12 highlighting the relevance of UACR control as a meaningful objective measure for clinicians. A recent post-hoc mediation analysis of data from two phase III clinical trials on finerenone, a drug recommended for UACR control and CV/CKD risk reduction, concluded that 84% of finerenone-associated improvement in renal outcomes and 37% of its improvement in CV outcomes were mediated by finerenone’s effect on decreasing UACR.28 While achieving reduction in UACR would lead to greater clinical benefits, maintaining a stable UACR level may be a suboptimal treatment target; it can nonetheless confer clinical benefits when a sustained reduction is not feasible.
It is important to note that the current study characterized UACR change patterns in real-world practice before the implementation of the most recent (2022) KDIGO guidelines,8 which emphasized UACR control and recommended CKD treatments with UACR reduction effects (ie, SGLT2i and finerenone).29–31 Nonetheless, over half of patients had a >30% UACR decrease 2 years after the initial test, which could be at least partially attributable to the widespread use of renin–angiotensin–aldosterone system (RAAS) inhibitors, which may have been initiated or increased in response to the elevated UACR observed during the initial testing.32
Despite a large proportion of patients with decreased or stable UACR, there were still more than 20% of patients who experienced an over 30% increase in UACR, indicating inadequate disease control. In addition, and more importantly, we also found that 47% of patients with elevated UACR did not have a valid subsequent UACR test within 0.5–2 years after the initial test. These results suggest an unmet need of UACR monitoring and management in this patient population, which may in turn lead to worsened clinical outcomes.
A strength of this study was the use of a large EHR database as the data source, which allowed the selection of a representative sample of patients with incident CKD and prior type 2 diabetes across all age groups and geographic regions in the USA. In addition, the use of real-world data helps capture the variability in patient characteristics and disease management strategies that are typically not reflected in the randomized clinical trials. Another strength was that unlike many other real-world studies which relied on diagnosis codes alone (which can result in underdetection of disease and delayed identification of incident disease onset), the current study used eGFR and UACR laboratory measures in addition to CKD diagnosis codes to identify patients with CKD and characterize disease stage. In addition, our use of the EMERGE algorithm,17 which required at least two separate modalities to identify a patient with type 2 diabetes, resulted in greater specificity in identifying patients with type 2 diabetes than studies that rely on diagnosis codes alone. By identifying patients with both CKD and type 2 diabetes, rather than relying exclusively on diagnosis codes for diabetic kidney disease, the study captured a broader and more representative patient population, which is also consistent with the target patient population for whom the ADA guidelines recommend treatments for the management of kidney disease and CV risk, such as SGLT2is, GLP-1 RAs, and non-steroidal mineralocorticoid receptor antagonists.12 Moreover, the study population includes a large proportion of patients with early-stage CKD who had elevated UACR, which is a critical subpopulation for patients with CKD and type 2 diabetes. Additionally, UACR change was carefully defined to minimize potential selection bias and reflect clinically meaningful categorization to inform clinical practice. Lastly, this study had a long duration of follow-up (median of 3.4 years) allowing the evaluation of long-term clinical outcomes.
There were also some limitations to the study that warrant mention. First, UACR measurements were less frequently recorded in the dataset, which has led to the use of a single UACR test to define albuminuria and UACR change and a relatively long interval to identify the follow-up UACR test (ie, 0.5–2 years after the initial UACR). Therefore, the UACR change reflects relative change measured at different timepoints compared with the initial UACR measurement. Second, as UACR measurements are highly variable, there may have been misclassification of UACR category and change status. However, this was partially mitigated by our requirement of observing large changes in UACR (ie, >30%) to be classified as having either an increase or decrease in UACR, which would reduce misclassification errors introduced by the fluctuation in UACR levels. Finally, to ensure that the selection of a study population with a follow-up UACR was independent of the duration of continuous eligibility, patients were required to have at least 2 years of continuous eligibility after the initial UACR test. Consequently, patients with rapid disease progression after the initial UACR test that led to early death were excluded.
Conclusions
In summary, our findings demonstrate that achievement of decreased or stable UACR is associated with lower risk of CV and kidney adverse events compared with patients with an increase in UACR. These results add to previous studies examining the impact of UACR change on kidney outcomes and establish the benefit of measuring and targeting UACR for both CV and kidney disease. Together, our findings and the ADA guidelines support measurement of albuminuria at every visit in patients with kidney disease and type 2 diabetes, and its importance as a modifiable risk factor.
Acknowledgments
Medical writing support was provided by Janice Imai, an employee of Analysis Group, Inc, which provided paid consulting services to Bayer U.S., LLC for the development and conduct of this study and development of the manuscript.
Footnotes
Contributors: NT, RS, YC, KAB, YMKF, SB, YD, SXK, TW, QL, AW, and KP conceived the study idea and design. RS, YC, KAB, YD, QL, and AW contributed to the data collection of the study. YC, KAB, AW, MS, and BR performed the statistical analysis of the study. All authors contributed critically to the interpretation of findings. All authors contributed to the drafting of the manuscript, provided crucial input on multiple iterations of the manuscript and approved the submitted version and each author satisfies the authorship criteria of the International Committee of Medical Journal Editors. KAB is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Funding: This study was funded by Bayer U.S., LLC. The sponsor was involved in the study design, analysis, and interpretation of data, writing the manuscript, and the decision to submit the report for publication.
Competing interests: NT received grants from CIHR, NIH, Kidney Foundation of Canada, AstraZeneca, Boehringer Ingelheim, Janssen Pharmaceuticals, Research Manitoba, Otsuka Pharmaceutical, Tricida, Eli Lilly, and Bayer U.S., LLC, which funded the development and conduct of this study and development of the manuscript. NT provided consulting services to AstraZeneca, Boehringer Ingelheim, GSK, Janssen Pharmaceuticals, Otsuka Pharmaceutical, Prokidney, Roche, Tricida, and Bayer U.S., LLC. NT received honoraria for lectures, presentations, speakers, bureaus, manuscript writing or educational events from AstraZeneca, Boehringer Ingelheim, Janssen Pharmaceuticals, Eli Lilly, Otsuka Pharmaceutical and Tricida. NT has a pending patent for a microfluidic device for point of care detection of urine albumin. NT served in the advisory board of AstraZeneca, Janssen Pharmaceuticals, BI-Lilly, Otsuka Pharmaceutical and National Kidney Foundation. NT reported ownership of ClinPredict, Klinrisk, Quanta and Marizyme; and holds stock options of Mesentech, Renibus Therapeutics, PulseData and Tricida. KP provided consulting services to AstraZeneca, Corcept, Diasome, Eli Lilly, Merck, Novo Nordisk, Sanofi Aventis, Twinhealth and Bayer U.S., LLC, which funded the development and conduct of this study and development of the manuscript. KP conducted teaching and speaking for AstraZeneca, Corcept, Merck and Novo Nordisk; and research for Eli Lilly, Merck, Novo Nordisk, Twinhealth and Bayer U.S., LLC. RS, YMKF, SB, YD, SXK, TW and QL are employees of Bayer U.S., LLC, which funded the development and conduct of this study and development of the manuscript. YC, KAB, AW, MS and BR are employees of Analysis Group, Inc, a company that has provided paid consulting services to Bayer U.S., LLC, which funded the development and conduct of this study and development of the manuscript.
Provenance and peer review: Not commissioned; externally peer reviewed.
Supplemental material: This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Data availability statement
No data are available. All data generated or analyzed during this study were derived from the privately held Optum electronic health records database and are not publicly available.
Ethics statements
Patient consent for publication
Not applicable.
References
- 1. US Centers for Disease Control and Prevention . Diabetes and chronic kidney disease. 2022. Available: https://www.cdc.gov/diabetes/managing/diabetes-kidney-disease.html#:~:text=Approximately%201%20in%203%20adults,diabetes%20can%20cause%20kidney%20disease.&text=Kidney%20diseases%20are%20the%209th,adults%20with%20diabetes%20has%20CKD
- 2. Zelnick LR, Weiss NS, Kestenbaum BR, et al. Diabetes and CKD in the United States Population, 2009-2014. Clin J Am Soc Nephrol 2017;12:1984–90. 10.2215/CJN.03700417 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Oellgaard J, Gæde P, Rossing P, et al. Intensified multifactorial intervention in type 2 diabetics with microalbuminuria leads to long-term renal benefits. Kidney Int 2017;91:982–8. 10.1016/j.kint.2016.11.023 [DOI] [PubMed] [Google Scholar]
- 4. Wong MG, Perkovic V, Chalmers J, et al. Long-term Benefits of Intensive Glucose Control for Preventing End-Stage Kidney Disease: ADVANCE-ON. Diabetes Care 2016;39:694–700. 10.2337/dc15-2322 [DOI] [PubMed] [Google Scholar]
- 5. Zoungas S, Chalmers J, Neal B, et al. Follow-up of blood-pressure lowering and glucose control in type 2 diabetes. N Engl J Med 2014;371:1392–406. 10.1056/NEJMoa1407963 [DOI] [PubMed] [Google Scholar]
- 6. Afkarian M, Sachs MC, Kestenbaum B, et al. Kidney disease and increased mortality risk in type 2 diabetes. J Am Soc Nephrol 2013;24:302–8. 10.1681/ASN.2012070718 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thomas MC, Brownlee M, Susztak K, et al. Diabetic kidney disease. Nat Rev Dis Primers 2015;1:15018. 10.1038/nrdp.2015.18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Rossing P, Caramori ML, Chan JCN, et al. KDIGO 2022 Clinical Practice Guideline for Diabetes Management in Chronic Kidney Disease. Kidney Int 2022;102:S1–127. 10.1016/j.kint.2022.06.008 [DOI] [PubMed] [Google Scholar]
- 9. Ärnlöv J, Nowak C. Association between albuminuria, incident cardiovascular events, and mortality in persons without hypertension, diabetes, and cardiovascular disease. Eur J Prev Cardiol 2022;29:e4–6. 10.1093/eurjpc/zwaa095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Matsushita K, Coresh J, Sang Y, et al. Estimated glomerular filtration rate and albuminuria for prediction of cardiovascular outcomes: a collaborative meta-analysis of individual participant data. Lancet Diabetes Endocrinol 2015;3:514–25. 10.1016/S2213-8587(15)00040-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Pasternak M, Liu P, Quinn R, et al. Association of Albuminuria and Regression of Chronic Kidney Disease in Adults With Newly Diagnosed Moderate to Severe Chronic Kidney Disease. JAMA Netw Open 2022;5:e2225821. 10.1001/jamanetworkopen.2022.25821 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. ElSayed NA, Aleppo G, Aroda VR, et al. Chronic Kidney Disease and Risk Management: Standards of Care in Diabetes—2023. Diabetes Care 2023;46:S191–202. 10.2337/dc23-S011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Coresh J, Heerspink HJL, Sang Y, et al. Change in albuminuria and subsequent risk of end-stage kidney disease: an individual participant-level consortium meta-analysis of observational studies. Lancet Diabetes Endocrinol 2019;7:115–27. 10.1016/S2213-8587(18)30313-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Heerspink HJL, Greene T, Tighiouart H, et al. Change in albuminuria as a surrogate endpoint for progression of kidney disease: a meta-analysis of treatment effects in randomised clinical trials. The Lancet Diabetes & Endocrinology 2019;7:128–39. 10.1016/S2213-8587(18)30314-0 [DOI] [PubMed] [Google Scholar]
- 15. Gansevoort RT, Correa-Rotter R, Hemmelgarn BR, et al. Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet 2013;382:339–52. 10.1016/S0140-6736(13)60595-4 [DOI] [PubMed] [Google Scholar]
- 16. Dalrymple LS, Katz R, Kestenbaum B, et al. Chronic kidney disease and the risk of end-stage renal disease versus death. J Gen Intern Med 2011;26:379–85. 10.1007/s11606-010-1511-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Pantalone KM, Misra-Hebert AD, Hobbs TM, et al. Antidiabetic treatment patterns and specialty care utilization among patients with type 2 diabetes and cardiovascular disease. Cardiovasc Diabetol 2018;17:54:54. 10.1186/s12933-018-0699-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kidney disease: improving global outcomes (kdigo) diabetes work group. kdigo 2020 clinical practice guideline for diabetes management in chronic kidney disease. 2020.
- 19. Fangel MV, Nielsen PB, Kristensen JK, et al. Albuminuria and Risk of Cardiovascular Events and Mortality in a General Population of Patients with Type 2 Diabetes Without Cardiovascular Disease: A Danish Cohort Study. Am J Med 2020;133:e269–79. 10.1016/j.amjmed.2019.10.042 [DOI] [PubMed] [Google Scholar]
- 20. Liu S, Niu J, Wu S, et al. Urinary albumin-to-creatinine ratio levels are associated with subclinical atherosclerosis and predict CVD events and all-cause deaths: a prospective analysis. BMJ Open 2021;11:e040890. 10.1136/bmjopen-2020-040890 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Norris KC, Smoyer KE, Rolland C, et al. Albuminuria, serum creatinine, and estimated glomerular filtration rate as predictors of cardio-renal outcomes in patients with type 2 diabetes mellitus and kidney disease: a systematic literature review. BMC Nephrol 2018;19:36. 10.1186/s12882-018-0821-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Li X, Wu H, Peng H, et al. n.d. Comparison the effects of finerenone and SGLT2i on cardiovascular and renal outcomes in patients with type 2 diabetes mellitus: A network meta-analysis. Front Endocrinol 13. 10.3389/fendo.2022.1078686 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Persson F, Bain SC, Mosenzon O, et al. Changes in Albuminuria Predict Cardiovascular and Renal Outcomes in Type 2 Diabetes: A Post Hoc Analysis of the LEADER Trial. Diabetes Care 2021;44:1020–6. 10.2337/dc20-1622 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Palmer BF. Change in albuminuria as a surrogate endpoint for cardiovascular and renal outcomes in patients with diabetes. Diabetes Obes Metab 2023;25:1434–43. 10.1111/dom.15030 [DOI] [PubMed] [Google Scholar]
- 25. Waijer SW, Gansevoort RT, Heerspink HJL. Change in albuminuria as a surrogate endpoint. Curr Opin Nephrol Hypertens 2019;28:519–26. 10.1097/MNH.0000000000000541 [DOI] [PubMed] [Google Scholar]
- 26. Babazono T, Hanai K, Yokoyama Y, et al. Association between 1-year changes in urinary albumin-to-creatinine ratio and kidney disease progression in Japanese individuals with diabetes: a historical cohort study. Clin Exp Nephrol 2023;27:1001–9. 10.1007/s10157-023-02380-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yokoyama Y, Hanai K, Mori T, et al. Changes in urinary albumin as a surrogate for kidney disease progression in people with type 2 diabetes. Clin Exp Nephrol 2023;27:465–72. 10.1007/s10157-023-02328-y [DOI] [PubMed] [Google Scholar]
- 28. Agarwal R, Tu W, Farjat AE, et al. Impact of Finerenone-Induced Albuminuria Reduction on Chronic Kidney Disease Outcomes in Type 2 Diabetes: A Mediation Analysis. Ann Intern Med 2023;176:1606–16. 10.7326/M23-1023 [DOI] [PubMed] [Google Scholar]
- 29. Bakris GL, Agarwal R, Anker SD, et al. Effect of Finerenone on Chronic Kidney Disease Outcomes in Type 2 Diabetes. N Engl J Med 2020;383:2219–29. 10.1056/NEJMoa2025845 [DOI] [PubMed] [Google Scholar]
- 30. Ruilope LM, Pitt B, Anker SD, et al. Kidney outcomes with finerenone: an analysis from the FIGARO-DKD study. Nephrol Dial Transplant 2023;38:372–83. 10.1093/ndt/gfac157 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Górriz JL, Romera I, Cobo A, et al. Glucagon-Like Peptide-1 Receptor Agonist Use in People Living with Type 2 Diabetes Mellitus and Chronic Kidney Disease: A Narrative Review of the Key Evidence with Practical Considerations. Diabetes Ther 2022;13:389–421. 10.1007/s13300-021-01198-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Zhao M, Wang R, Yu Y, et al. Efficacy and Safety of Angiotensin-Converting Enzyme Inhibitor in Combination with Angiotensin-Receptor Blocker in Chronic Kidney Disease Based on Dose: A Systematic Review and Meta-Analysis. Front Pharmacol 2021;12:638611. 10.3389/fphar.2021.638611 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
bmjdrc-13-5-s001.pdf (450KB, pdf)
Data Availability Statement
No data are available. All data generated or analyzed during this study were derived from the privately held Optum electronic health records database and are not publicly available.