Statistics — Chronic kidney disease¶
Last curated: 2026-09-02 (deepening pass). Every figure carries its source, year, population and method. Conflicting estimates are shown side by side and never averaged.
Global burden¶
| Measure | Estimate | Source year / population | Method | Source |
|---|---|---|---|---|
| Prevalent CKD cases | 697.5 million (95% UI 649.2–752.0m) | 2017, global | GBD modelling | GBD CKD Collaboration 2020, PMID 32061315 |
| Global CKD prevalence | 9.1% (95% UI 8.5–9.8) | 2017, global | GBD modelling | PMID 32061315 |
| Deaths attributed directly to CKD | 1.2 million (95% UI 1.2–1.3m) | 2017, global | GBD cause-of-death model | PMID 32061315 |
| Change in all-age CKD mortality rate | +41.5% (95% UI 35.2–46.5); age-standardised rate change not significant (2.8%, −1.5 to 6.3) | 1990–2017, global | GBD time trend | PMID 32061315 |
| Cardiovascular deaths attributable to impaired kidney function | 1.4 million (95% UI 1.2–1.6m) | 2017, global | Comparative risk assessment | PMID 32061315 |
Definition and measurement¶
| Measure | Estimate | Population | Method | Source |
|---|---|---|---|---|
| Duration required for CKD | >3 months | Global guideline population | Consensus definition + evidence review | KDIGO 2024, PMID 38490803 |
| G3a | eGFR 45–59 mL/min/1.73 m² | Adults | GFR category | PMID 38490803 |
| G3b | eGFR 30–44 mL/min/1.73 m² | Adults | GFR category | PMID 38490803 |
| G4 | eGFR 15–29 mL/min/1.73 m² | Adults | GFR category | PMID 38490803 |
| G5 | eGFR <15 mL/min/1.73 m² | Adults | GFR category | PMID 38490803 |
| A2 albuminuria | uACR 30–300 mg/g | Adults | Albuminuria category | PMID 38490803 |
| A3 albuminuria | uACR >300 mg/g | Adults | Albuminuria category | PMID 38490803 |
| Race-free equations within 30% of measured GFR | ≥85% | Validation set, 4,050 participants | Equation validation against measured GFR | Inker 2021, PMID 34554658 |
| New creatinine-equation median bias, Black participants | −3.6 mL/min/1.73 m² | Validation set | Median eGFR minus measured GFR | PMID 34554658 |
| New creatinine-equation median bias, non-Black participants | +3.9 mL/min/1.73 m² | Validation set | Median eGFR minus measured GFR | PMID 34554658 |
Disease-modifying trials¶
| Trial / measure | Treatment | Control | Effect (95% CI) | Population / follow-up | Source |
|---|---|---|---|---|---|
| DAPA-CKD primary composite | 9.2% | 14.5% | HR 0.61 (0.51–0.72); NNT 19 (15–27) | 4,304; median 2.4 y | PMID 32970396 |
| DAPA-CKD kidney-specific composite | — | — | HR 0.56 (0.45–0.68) | Same trial | PMID 32970396 |
| DAPA-CKD all-cause death | 4.7% | 6.8% | HR 0.69 (0.53–0.88) | Same trial | PMID 32970396 |
| DAPA-CKD total eGFR-slope difference | — | — | +0.95 (0.63–1.27) mL/min/1.73 m²/y | Prespecified analysis | PMID 34619108 |
| EMPA-KIDNEY primary composite | 13.1% | 16.9% | HR 0.72 (0.64–0.82) | 6,609; median 2.0 y | PMID 36331190 |
| EMPA-KIDNEY all-cause hospitalization | — | — | HR 0.86 (0.78–0.95) | Same trial | PMID 36331190 |
| EMPA-KIDNEY acute eGFR dip | — | — | −2.12 (−2.41 to −1.83) mL/min/1.73 m² | First 2 months | PMID 38061371 |
| EMPA-KIDNEY chronic slope | −1.37/y | −2.75/y | 50% relative reduction (42–58) | Prespecified slope analysis | PMID 38061371 |
| CREDENCE primary event rate | 43.2/1,000 py | 61.2/1,000 py | HR 0.70 (0.59–0.82) | 4,401; median 2.62 y | PMID 30990260 |
| CREDENCE kidney-specific composite | — | — | HR 0.66 (0.53–0.81) | Same trial | PMID 30990260 |
| FIDELIO kidney composite | 17.8% | 21.1% | HR 0.82 (0.73–0.93) | 5,734; median 2.6 y | PMID 33264825 |
| FIDELIO hyperkalaemia discontinuation | 2.3% | 0.9% | +1.4 percentage points | Same trial | PMID 33264825 |
| FIDELITY kidney composite | 5.5% | 7.1% | HR 0.77 (0.67–0.88) | 13,026; median 3.0 y | PMID 35023547 |
| FIDELITY cardiovascular composite | 12.7% | 14.4% | HR 0.86 (0.78–0.95) | Same pooled analysis | PMID 35023547 |
| FLOW primary composite | 331 events | 410 events | HR 0.76 (0.66–0.88) | 3,533; median 3.4 y | PMID 38785209 |
| FLOW eGFR-slope difference | — | — | +1.16 mL/min/1.73 m²/y | Same trial | PMID 38785209 |
| FLOW all-cause death | — | — | HR 0.80 (0.67–0.95) | Same trial | PMID 38785209 |
| FIND-CKD total eGFR slope (no diabetes) | −3.3/y | −4.0/y | Difference 0.7 (0.3–1.1) mL/min/1.73 m²/y | 1,584; 32 months | PMID 42246672 |
| FIND-CKD kidney-or-CV composite | — | — | HR 0.77 (0.60–0.99) | Same trial; hierarchical testing | PMID 42246672 |
| FIND-CKD hyperkalaemia | 17.0% | 13.3% | Discontinuation 1.5% vs 0.1% | Same trial | PMID 42246672 |
| FIND-CKD glomerular-disease subgroup, eGFR slope | −3.50/y | −4.23/y | Difference 0.73 (0.22–1.24) mL/min/1.73 m²/y | 903 of 1,584; exploratory | PMID 42246414 |
| INFINITY pooled kidney composite | 22.3/1,000 py | 28.8/1,000 py | HR 0.76 (0.68–0.86) | 14,574 across FIDELIO, FIGARO, FIND-CKD | PMID 42248158 |
| INFINITY pooled kidney failure alone | — | — | HR 0.85 (0.74–0.99) | Same pooled analysis | PMID 42248158 |
| INFINITY pooled HF hospitalisation or CV death | 19.1/1,000 py | 23.9/1,000 py | HR 0.80 (0.70–0.91) | Same pooled analysis | PMID 42248158 |
| INFINITY pooled all-cause death | — | — | HR 0.88 (0.79–0.99) | Same pooled analysis | PMID 42248158 |
| CONFIDENCE uACR at day 180, combination vs finerenone | — | — | 29% greater reduction; ratio 0.71 (0.61–0.82) | 779 randomized; phase 2, surrogate endpoint | PMID 40470996 |
| CONFIDENCE uACR at day 180, combination vs empagliflozin | — | — | 32% greater reduction; ratio 0.68 (0.59–0.79) | Same trial | PMID 40470996 |
| CONFIDENCE hyperkalaemia by arm | 15.1% combination | 18.8% finerenone / 9.7% empagliflozin | No significant difference combination vs finerenone | Prespecified secondary analysis | PMID 41493296 |
| FINE-ONE uACR at 6 months (type 1 diabetes) | −34% | −12% | Ratio 0.75 (0.65–0.87) | 242 randomized; surrogate endpoint | PMID 41780000 |
| Semaglutide pooled kidney composite | 973 events | 1,134 events | HR 0.84 (0.77–0.91) | 30,787 across SELECT, FLOW, SOUL | PMID 42567173 |
Complications, replacement and symptoms¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| TREAT stroke | — | Darbepoetin vs placebo/rescue; HR 1.92 (1.38–2.68) | PMID 19880844 |
| EVOLVE primary composite | HR 0.93 (0.85–1.02) | 3,883 haemodialysis patients; unadjusted ITT | PMID 23121374 |
| SHARP major atherosclerotic events | RR 0.83 (0.74–0.94) | 9,270 CKD patients; median 4.9 y | PMID 21663949 |
| IDEAL all-cause death | HR 1.04 (0.83–1.30) | Early vs late planned dialysis; 828 adults | PMID 20581422 |
| IDEAL late-arm start above target | 75.9% | Started for symptoms above eGFR 7 | PMID 20581422 |
| Conservative-care review mortality | HR 0.47 (0.34–0.65), favouring dialysis | 25 observational studies | PMID 34507554 |
| Forgoing-dialysis median survival range | 1–41 months | 41 cohorts, 5,102 patients | PMID 35285915 |
| Forgoing-dialysis hospital admissions | 1–2/person-year | Cohort synthesis | PMID 35285915 |
| Forgoing-dialysis in-hospital days | 6–16/person-year | Cohort synthesis | PMID 35285915 |
| Difelikefalin ≥3-point itch response | 49.1% | 27.9% placebo; 378 haemodialysis patients | PMID 31702883 |
| Difelikefalin ≥4-point itch response | 37.1% | 17.9% placebo | PMID 31702883 |
| AMBER retained on spironolactone | 86% | 66% placebo; difference 19.5 points (10.0–29.0) | PMID 31533906 |
| VALOR-CKD kidney composite | HR 0.99 (0.80–1.20) | Veverimer vs placebo | PMID 38261535 |
| FSGS partial proteinuria remission | 42% | 26% irbesartan; week 36 | PMID 37921461 |
| FSGS total eGFR-slope difference | +0.3 (−1.7 to 2.4) mL/min/1.73 m²/y | Sparsentan vs irbesartan | PMID 37921461 |
Burden — updated global estimates (GBD 2023 and pooled cohorts)¶
| Measure | Estimate | Source year / population | Method | Source |
|---|---|---|---|---|
| Prevalent CKD, adults ≥20 y | 788 million (95% UI 743–843m) | 2023, global | GBD modelling | GBD 2023 CKD Collaborators 2025, PMID 41213283 |
| Prevalent CKD, adults ≥20 y | 378 million (354–407m) | 1990, global | Same model | PMID 41213283 |
| Age-standardised adult prevalence | 14.2% (13.4–15.2); relative rise 3.5% (2.7–4.1) since 1990 | 2023, global | GBD modelling | PMID 41213283 |
| CKD deaths | 1.48 million (1.30–1.65m); 9th leading cause | 2023, global | GBD cause-of-death model | PMID 41213283 |
| CKD DALY rate | 769.2 (691.8–857.4) per 100,000, age-standardised; 12th leading cause | 2023, global | GBD | PMID 41213283 |
| CV deaths attributable to impaired kidney function | 11.5% (8.4–14.5) of all CV deaths | 2023, global | Comparative risk assessment | PMID 41213283 |
| Highest regional age-standardised prevalence | 18.0% (16.9–19.4), north Africa and Middle East | 2023 | GBD | PMID 41213283 |
| Stage 1–3 combined prevalence | 13.9% (13.1–15.0) of 14.2% | 2023, global | GBD | PMID 41213283 |
| Pooled prevalence, stages 1–5 | 13.4% (11.7–15.1) | ≤2016; 100 studies, 6,908,440 participants | Random-effects meta-analysis | Hill 2016, PMID 27383068 |
| Pooled prevalence, stages 3–5 | 10.6% (9.2–12.2) | Same | Same | PMID 27383068 |
| Stage-specific pooled prevalence | S1 3.5% (2.8–4.2); S2 3.9% (2.7–5.3); S3 7.6% (6.4–8.9); S4 0.4% (0.3–0.5); S5 0.1% (0.1–0.1) | Same | Same | PMID 27383068 |
| Lifetime risk of dialysis-requiring kidney failure | Men 3.14% (3.10–3.18); women 1.42% (1.39–1.44) — ~1 in 32 and 1 in 71 | 2017, Japan; competing-mortality adjusted | National registry life-table | Wakasugi 2020, PMID 32040656 |
| Projected CKD population growth at current diagnosis rates | +5.8% population, +9.3% costs by 2027 | 31 countries | Microsimulation | Wish 2025, PMID 41141496 |
Measurement — variability, discordance and equation performance¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| Within-person CV, serum creatinine | 5.4% (RCV +16%/−14%) | 50 stable CKD outpatients, 3 visits <4 weeks | Waikar 2018, PMID 30031564 |
| Within-person CV, cystatin C | 4.1% (+12%/−11%) | Same | PMID 30031564 |
| Within-person CV, first-morning uACR | 32.5% (+141%/−58%) | Same | PMID 30031564 |
| Within-person CV, random-spot uACR | 29.7% (+124%/−55%) | Same | PMID 30031564 |
| Within-person CV, random-spot urine albumin concentration | 50.6% (+276%/−73%) | Same | PMID 30031564 |
| Prevalence of eGFRcys ≥30% below eGFRcr, outpatients | 11% (cohort range 3–50%) | 821,327 outpatients, 23 cohorts | Estrella 2025, PMID 41202182 |
| Same, inpatients | 35% | 39,639 inpatients, 2 cohorts | PMID 41202182 |
| Large negative eGFR difference — all-cause mortality | 28.4 vs 16.8 per 1,000 py; HR 1.69 (1.57–1.82) | Mean 11 y follow-up | PMID 41202182 |
| Same — CV mortality / HF / ASCVD / KFRT | HR 1.61 (1.48–1.76) / 1.54 (1.40–1.68) / 1.35 (1.27–1.44) / 1.29 (1.13–1.47) | Same | PMID 41202182 |
| Variance in eGFR difference explained by putative determinants | 36% | 1,290 CRIC participants with iothalamate GFR | McCoy 2025, PMID 40373998 |
| P30 when eGFRcys ≥20% below eGFRcr | eGFRcr 50%, eGFRcys 73%, eGFRcr-cys 84% | 9,404 paired samples vs iohexol clearance | Fu 2023, PMID 36995139 |
| Median bias in the same stratum | +15.0 / −8.5 / +0.8 mL/min/1.73 m² | Same | PMID 36995139 |
| P30 in US mGFR cohorts | CKD-EPI 2021 79.2% (78.5–79.9); EKFC race-free 80.1% (79.4–80.7); EKFC population-specific 81.1% (80.5–81.8) | 12,854 participants, 9 studies | Delanaye 2024, PMID 38101514 |
| Median bias, same cohorts | CKD-EPI 2021 +1.22 (0.99–1.47); EKFCPS +0.14 (−0.07 to 0.35) mL/min/1.73 m² | Same | PMID 38101514 |
| EKFC creatinine equation, adult bias | −0.9 mL/min/1.73 m²; 3.1% of estimates in error >30% | 11,251 development + 8,378 external validation | Pottel 2021, PMID 33166224 |
| Dipstick ≥1+ for uACR ≥30 mg/g | Sensitivity 57.8% (54.1–61.4); specificity 95.4% | 10,944 Australian adults | White 2011, PMID 21411199 |
| Dipstick ≥1+ for uACR ≥300 mg/g | Sensitivity 98.9%; specificity 92.6% | Same | PMID 21411199 |
| Dipstick trace-or-greater for uACR ≥30 / stage A2 | Sensitivity 62% / 36%; specificity 88% / 88% | 919,383 adults, 33 cohorts | Sumida 2020, PMID 32658569 |
| PCR conversion for screening / A2 / A3 | Sensitivity 91% / 75% / 87%; specificity 87% / 89% / 98% | Same | PMID 32658569 |
Age-adaptation, screening economics and capacity¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| Incidence under fixed vs age-adapted CKD criteria | 537 vs 343 new cases per 100,000 person-years | Alberta population, 2009–2017 | Liu 2021, PMID 34459844 |
| 5-year kidney failure / death, fixed vs age-adapted cohorts | 1.7% vs 3.0% / 21.9% vs 25.4% | Same | PMID 34459844 |
| Fixed-criteria-only cases aged ≥65 with eGFR 45–59 and A1 | 54,342 of 72,703 (75%); 5-year kidney-failure risk ≤0.12% | Same | PMID 34459844 |
| Ratio of death risk to kidney-failure risk in that group | 69× (65–69 y), 122× (70–74), 279× (75–79), 935× (≥80) | Same | PMID 34459844 |
| ICER, one-time albuminuria screening at age 55 | $86,300/QALY; kidney failure −0.29 percentage points; life expectancy 17.29→17.45 y | US Markov model, NHANES-calibrated | Cusick 2023, PMID 37216661 |
| ICER, 5-yearly screening ages 55–75 with SGLT2i | $128,400/QALY; KRT 2.4%→1.9%; +0.13 life-years | Same model family | Cusick 2024, PMID 39514193 |
| ICER if SGLT2i efficacy 30% lower | $145,400–$182,600/QALY for decennial screening | Sensitivity analysis | PMID 37216661 |
| Dapagliflozin ICER (UK / Germany / Spain) | $8,280 / $17,623 / $11,687 per QALY | DAPA-CKD lifetime Markov model | McEwan 2022, PMID 36323444 |
| Median nephrologists per million population | 11.75 (IQR 1.78–24.76) globally; Africa 1.12; South Asia 1.81; Oceania/SE Asia 3.18 | ISN-GKHA, 167 countries | Okpechi 2024, PMID 39235198 |
| Countries reporting workforce shortages | Transplant surgeons 65%; nephrologists 64%; access coordinators 59%; dialysis nurses 58%; interventional radiologists 54% | Same | PMID 39235198 |
Surrogate-endpoint performance¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| eGFR slope 0.75 mL/min/1.73 m²/y slower over 2 y → ESKD | HR 0.70 (0.68–0.72) at eGFR ≥60; 0.71 (0.68–0.74) at <60 | 3,758,551 + 122,664 participants, 14 cohorts | Grams 2019, PMID 31292199 |
| Absolute 5-year ESKD reduction from that slope change | 1.6% (high-risk, 8.3% baseline) vs 0.13% (low-risk, 0.58% baseline) | Same | PMID 31292199 |
| 30% uACR reduction → kidney failure or doubled creatinine | 19% lower hazard (95% BCI 5–30%); median R² 0.66 (BCI 0.06–0.98) | 48 RCTs, 85,681 participants | Heerspink 2026, PMID 41198855 |
| Finerenone: proportion of kidney effect mediated by 4-month uACR change | 84% continuous; 64% at the ≥30% threshold | 12,512 patients, FIDELIO + FIGARO | Agarwal 2023, PMID 38048573 |
| Finerenone: proportion of CV effect so mediated | 37% continuous; 26% at threshold | Same | PMID 38048573 |
Mechanism and progression¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| Nephron number per kidney, healthy donors | 860,000 ± 370,000 | 1,388 living donors; CT cortical volume × biopsy glomerular density | Denic 2017, PMID 28614683 |
| Single-nephron GFR | 80 ± 40 nL/min; flat across age <70, sex, height ≤190 cm | Same | PMID 28614683 |
| Human kidney atlas resolution | 51 cell types; 28 injury-associated cell states; >400,000 nuclei/cells; 45 reference + 48 diseased kidneys | Multi-omic single-cell and spatial | Lake 2023, PMID 37468583 |
| AKI → incident CKD | HR 8.8 (3.1–25.5); 25.8 per 100 py | 13 cohort studies | Coca 2012, PMID 22113526 |
| AKI → new or progressive CKD (consensus definitions) | HR 2.67 (1.99–3.58); 17.76 vs 7.59 per 100 py | 82 studies, 2,017,437 participants | See 2019, PMID 30473140 |
| AKI → ESKD / death | HR 4.81 (3.04–7.62) / 1.80 (1.61–2.02) | Same | PMID 30473140 |
| Allopurinol effect on eGFR change at 104 weeks | −0.10 mL/min/1.73 m²/y (−1.18 to 0.97); p=0.85 | CKD-FIX, 363 analysed | Badve 2020, PMID 32579811 |
| Allopurinol effect on iohexol-measured GFR after washout | 0.001 mL/min/1.73 m² (−1.9 to 1.9); p=0.99 | PERL, 530 randomized, type 1 diabetes | Doria 2020, PMID 32579810 |
| Canakinumab MACE in CKD subgroup | HR 0.82 (0.68–1.00); 0.68 (0.53–0.86) if on-treatment hsCRP <2 mg/L | 1,875 of 10,061 CANTOS participants | Ridker 2018, PMID 29793629 |
| Ambient heat and eGFR | −0.6% (−0.9 to −0.3) per 30 days of heat index >30 °C in a 120-day window | 4,017 DAPA-CKD participants, 21 countries | Zhang 2024, PMID 38580424 |
| Adiposity and GFR decline (BMI 30/35/40 vs 25) | HR 1.18 (1.09–1.27) / 1.69 (1.51–1.89) / 2.02 (1.80–2.27); attenuated to 1.03/1.28/1.46 after comorbidity adjustment | 5,459,014 adults, 39 cohorts, 40 countries | Chang 2019, PMID 30630856 |
Aetiology¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| APOL1 and FSGS / hypertension-attributed ESKD | OR 10.5 (6.0–18.4) / 7.3 (5.6–9.5) | African American case-control | Genovese 2010, PMID 20647424 |
| Microalbuminuria without diabetes by APOL1 status | 2.3% European American; 6.0% AA 0–1 allele; 16.5% AA 2 alleles | Dallas Heart Study, 1,825 + 1,042 | Friedman 2011, PMID 21997396 |
| eGFR <60 without diabetes by APOL1 status | 1.5% / 1.7% / 6.7% | Same | PMID 21997396 |
| Sickle cell trait and incident ESRD | HR 2.03 (1.44–2.84); 8.5 vs 4.0 per 1,000 py | REGARDS, 9,909 Black participants | Naik 2017, PMID 28280138 |
| APOL1 high-risk genotype in same cohort | HR 1.77 (1.31–2.38); 6.6 per 1,000 py | Same | PMID 28280138 |
| Exome sequencing diagnostic yield in adult CKD | 9.3% (307/3,315) across 66 disorders; 23.9% congenital/cystic; 17.1% unknown-origin nephropathy | 2 cohorts | Groopman 2019, PMID 30586318 |
| Non-diabetic kidney disease at biopsy in diabetes | 58.8% (35.9% without concurrent DN; 22.9% as second diagnosis); 2.56-fold lower ESKD risk | 49,075 biopsied patients, 2001–2024 | Caza 2026, PMID 42034202 |
| Obesity-related glomerulopathy in a biopsy series | 0.89% (90/10,093), rising 0.62%→1.0% over 5 years | Nanjing, 2002–2006 | Chen 2008, PMID 18423814 |
| PPI use and incident CKD | HR 1.50 (1.14–1.96) adjusted | 10,482 ARIC; replicated in 248,751 | Lazarus 2016, PMID 26752337 |
| Lithium and CKD | 0.012 cases per exposed patient-year; eGFR decline 1.8 mL/min/y in those reaching G3; no ESKD | 1,012 patients, 2000–2015 | Van Alphen 2021, PMID 33392830 |
| Primary tubulointerstitial disease among biopsied CKDu referrals | 43/87 (49%); 30% with moderate-severe active disease | Sri Lanka, 600 new referrals over 1 year | Anand 2019, PMID 30659059 |
| Harvest-season eGFR fall in highest-biomarker sugarcane workers | 4.6 (1.0–8.2) mL/min/1.73 m² | 284 Nicaraguan workers | Laws 2016, PMID 26454687 |
Therapy — pooled class effects¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| SGLT2i and kidney disease progression | RR 0.63 (0.58–0.69); similar with and without diabetes | 13 trials, 90,409 participants | Nuffield Renal Studies Group 2022, PMID 36351458 |
| SGLT2i kidney progression, by diabetes status | 33 vs 48 per 1,000 py, HR 0.65 (0.60–0.70) with diabetes; 32 vs 46, HR 0.74 (0.63–0.85) without | 8 trials, 58,816 participants | Staplin 2026, PMID 41202026 |
| SGLT2i and acute kidney injury | 14 vs 18 per 1,000 py, HR 0.77 (0.69–0.87) with diabetes; 13 vs 18, HR 0.72 (0.56–0.92) without | Same | PMID 41202026 |
| SGLT2i vs other glucose-lowering classes for AKI | OR 0.76 (0.66–0.88) vs placebo; 0.79 (0.65–0.97) vs GLP-1 RA; 0.68 (0.54–0.86) vs DPP-4i | 20 trials, 156,690 participants | Zhao 2020, PMID 33376101 |
| Empagliflozin and serum uric acid | −25.6 µmol/L (−30.3 to −21.0); gout events HR 0.87 (0.74–1.02) | 6,609 EMPA-KIDNEY participants | Mayne 2025, PMID 39277784 |
| Semaglutide kidney composite without diabetes | 1.8% vs 2.2%; HR 0.78 (0.63–0.96) | 17,604 SELECT participants | Colhoun 2024, PMID 38796653 |
| GLP-1 RA MACE by background SGLT2i use | HR 0.77 (0.54–1.09) with vs 0.79 (0.71–0.87) without; p-heterogeneity 0.78 | 3 trials, 17,072 participants (1,743 exposed) | Neuen 2024, PMID 39210781 |
| Intensive glucose control and kidney events | HR 0.80 (0.72–0.88) for −0.90% HbA1c | ACCORD, ADVANCE, UKPDS, VADT; 27,049 | Zoungas 2017, PMID 28365411 |
| BP lowering, 5 mmHg systolic, CV events | HR 0.91 (0.87–0.94) with CKD; 0.90 (0.88–0.93) without; p-interaction >0.99 | 46 trials, 285,124 participants | Zeng 2026, PMID 42035778 |
| Same, CKD with vs without diabetes | HR 0.96 (0.90–1.02) vs 0.88 (0.84–0.93); p-interaction 0.044 | Same | PMID 42035778 |
| Chlorthalidone, 24-h ambulatory systolic BP at 12 weeks | −10.5 mmHg (−14.6 to −6.4) vs placebo | CLICK, 160 patients, mean eGFR 23.2 | Agarwal 2021, PMID 34739197 |
| Benazepril in advanced CKD (creatinine 3.1–5.0 mg/dL) | Composite 41% vs 60%; 43% RRR (p=0.005) | 224 patients, mean 3.4 y | Hou 2006, PMID 16407508 |
| RAS-inhibitor withdrawal at eGFR <30 | eGFR difference −0.7 (−2.5 to 1.0); ESKD/KRT HR 1.28 (0.99–1.65) | STOP-ACEi, 411 patients | Bhandari 2022, PMID 36326117 |
| Finerenone hyperkalaemia, mild / moderate | 21.4% vs 9.2% / 4.5% vs 1.4% over median 2.6 y | FIDELIO-DKD | Agarwal 2022, PMID 34732509 |
Complications and replacement — additional figures¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| PIVOTAL primary composite, high- vs low-dose IV iron | 29.3% vs 32.3%; HR 0.85 (0.73–1.00) | 2,141 haemodialysis patients, median 2.1 y | Macdougall 2019, PMID 30365356 |
| PIVOTAL monthly ESA dose difference | −7,539 IU (−9,485 to −5,582) | Same | PMID 30365356 |
| PIVOTAL infection rates, high vs low dose | 46.5% vs 45.5% (63.3 vs 69.4 per 100 py); no dose effect | Prespecified secondary analysis | Macdougall 2020, PMID 32253271 |
| FIND-CKD escalation endpoint | 23.5% high-ferritin IV FCM vs 31.8% oral iron; HR 0.65 (0.44–0.95) | 626 non-dialysis CKD patients | Macdougall 2014, PMID 24891437 |
| Daprodustat vs ESA, first HF hospitalisation | HR 1.22 (0.95–1.56) non-dialysis; 1.10 (0.84–1.45) dialysis | ASCEND-ND + ASCEND-D, 6,836 | Cunningham 2024, PMID 38383961 |
| HIF-PHI vs ESA, MACE | RR 0.99 (0.92–1.08) dialysis; 1.08 (0.95–1.22) non-dialysis | 25 trials, 26,478 participants | Ha 2024, PMID 39186635 |
| LANDMARK cardiovascular events, lanthanum vs calcium carbonate | 4.80 vs 4.30 per 100 py; HR 1.11 (0.88–1.41) | 2,309 haemodialysis patients, median 3.16 y | Ogata 2021, PMID 34003226 |
| PRIMO left ventricular mass index change | +0.34 (−0.14 to 0.83) paricalcitol vs −0.07 (−0.55 to 0.42) placebo g/m^2.7 | 227 patients, 48 weeks | Thadhani 2012, PMID 22337679 |
| AURORA primary composite, rosuvastatin on dialysis | 9.2 vs 9.5 per 100 py; HR 0.96 (0.84–1.11) despite 43% LDL fall | 2,776 haemodialysis patients, median 3.8 y | Fellström 2009, PMID 19332456 |
| ISCHEMIA-CKD death or MI | 36.4% vs 36.7% at 3 y; adjusted HR 1.01 (0.79–1.29) | 777 advanced-CKD patients | Bangalore 2020, PMID 32227756 |
| ISCHEMIA-CKD stroke / death-or-dialysis with invasive strategy | HR 3.76 (1.52–9.32) / 1.48 (1.04–2.11) | Same | PMID 32227756 |
| Incident vs progressive coronary calcification | Incident CAC: ASCVD HR 2.42 (1.23–4.79), death 1.82 (1.03–3.22); progression ≥50 units/y: ASCVD 1.42 (0.85–2.35), death 1.73 (1.31–2.28) | 1,310 CRIC participants | Tian 2025, PMID 39154888 |
| Haemodiafiltration vs haemodialysis, all-cause death | 23.3% vs 27.0%; HR 0.84 (0.74–0.95); graded convection-volume response | IPD meta-analysis, 5 RCTs, 4,153 patients | Vernooij 2024, PMID 39489903 |
| Cooler vs standard dialysate, CV composite | 21.4% vs 22.4%; adjusted HR 1.00 (96% CI 0.89–1.11) | MyTEMP, 84 centres, 15,413 patients | MyTEMP writing committee 2022, PMID 36343653 |
| Transplantation vs remaining waitlisted, all-cause death | HR 0.45 (0.39–0.54); 11 of 48 studies identified strata without benefit | 48 observational studies, 1,245,850 patients | Chaudhry 2022, PMID 35232772 |
| PD vs HD from day 0 / from day 90 | HR 0.92 (0.86–1.00) / 1.05 (0.96–1.16) | 6,337 matched pairs, US 2003 incident cohort | Weinhandl 2010, PMID 20133483 |
| RAASi discontinuation after hyperkalaemia — death | HR 1.32 (1.22–1.41) Manitoba; 1.47 (1.41–1.52) Ontario | 7,200 + 71,290 CKD patients | Leon 2022, PMID 35085685 |
| Same — dialysis initiation | HR 1.65 (1.41–1.85) / 1.11 (1.08–1.16) | Same | PMID 35085685 |
| SZC in haemodialysis — arrhythmic composite | 8.8% vs 8.9%; HR 0.98 (0.76–1.26) despite normokalaemia OR 3.36 (2.64–4.26) | DIALIZE-Outcomes, 2,690, stopped early | Fishbane 2025, PMID 40618849 |
| Conservative care vs dialysis pathway — hospital-free days | 352.7 vs 282.7 per year; adjusted IRR 1.15 (1.09–1.21) | 366 patients ≥70 y, Netherlands | Verberne 2018, PMID 30115028 |
| Same — annual treatment cost | Adjusted cost ratio 0.43 (0.28–0.67) | Same | PMID 30115028 |
| Dialysis withdrawal before death | 262 of 536 deaths (49%); median 7 days (IQR 4–11) to death; 34% received palliative care | 1,226 incident haemodialysis patients | Chen 2018, PMID 30026285 |
Symptoms, function and glomerular disease¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| Fatigue prevalence, CKD not on KRT | 70% (60–79) | 14 studies, 4,139 participants | Fletcher 2022, PMID 35385471 |
| Instruments used in the CKD symptom literature | 67 measures across 449 studies covering 68 symptoms, 199,147 participants, 62 countries | Systematic review | PMID 35385471 |
| Intradialytic exercise, 60-s sit-to-stand at 12 months | +3.85 repetitions (2.22–5.48); 6-min walk +37.5 m (14.7–60.4); hospital days 2 vs 5/year | 917 analysed of 1,211 randomized | Anding-Rost 2023, PMID 38320198 |
| Pre-dialysis supervised exercise | VO2peak 17.9 ± 5.5 vs 15.9 ± 7.0 mL/kg/min at 6 months, not sustained at 12; 6MWT +98 feet | 99 adults, eGFR 15–<45 | Weiner 2023, PMID 35944747 |
| Sertraline vs CBT for depression on haemodialysis | QIDS-C difference −1.84 (−3.54 to −0.13) at 12 weeks; more adverse events | 120 patients, phase 2 of 41-facility trial | Mehrotra 2019, PMID 30802897 |
| Treatment acceptance with vs without engagement interview | 66% vs 64% (risk difference 2.1, −12.1 to 16.4) | 184 patients, phase 1 | PMID 30802897 |
| Frailty prevalence in CKD | Frail 41.8% (range 2.8–81.5%); pre-frail 43.9% (19.1–62.7%) | 18 cohorts, 22,788 participants | Mei 2021, PMID 33218914 |
| Frailty and mortality | Pre-frailty HR 1.68 (1.46–1.94); frailty 1.48 (1.21–1.81) | Same | PMID 33218914 |
| Iptacopan, annualized total eGFR slope | −3.10 vs −6.12 mL/min/1.73 m²/y; difference 3.02 (2.02–4.01) | APPLAUSE-IgAN, 477 patients, 24 months | Barratt 2026, PMID 41910396 |
| Iptacopan, composite kidney failure endpoint | 21.4% vs 33.5%; HR 0.57 (0.40–0.81) | Same | PMID 41910396 |
| Iptacopan, 24-h uPCR at 9 months | −38.3% (26.0–48.6) vs placebo | Prespecified interim, first 250 randomized | Perkovic 2025, PMID 39453772 |
| Ravulizumab, urine protein at 26 weeks | −41.9% (−50.2 to −32.0) vs −16.8% (−31.8 to +1.6); 30.1% treatment effect (p=0.005) | Phase 2, 66 patients | Lafayette 2025, PMID 39455063 |
| Atrasentan, eGFR change at week 136 | −7.5 (−9.2 to −5.8) vs −9.9 (−11.7 to −8.1); difference 2.4 (−0.1 to 4.8), p=0.057; total slope difference 1.4/y (0.5–2.3) | ALIGN, 404 randomized | Heerspink 2026, PMID 42242268 |
| Tolvaptan, total kidney volume growth | 2.8%/y (2.5–3.1) vs 5.5%/y (5.1–6.0) | TEMPO 3:4, 1,445 patients | Torres 2012, PMID 23121377 |
| Tolvaptan, eGFR change over 1 year in later-stage ADPKD | −2.34 (−2.81 to −1.87) vs −3.61 (−4.08 to −3.14); difference 1.27 (0.86–1.68) | REPRISE, 1,370 patients | Torres 2017, PMID 29105594 |
| Rituximab vs cyclosporine remission, membranous nephropathy | 60% vs 52% at 12 months; 60% vs 20% at 24 months (difference 40 points, 25–55) | MENTOR, 130 patients | Fervenza 2019, PMID 31269364 |
| Voclosporin complete renal response at 52 weeks | 41% (73/179) vs placebo | AURORA 1, 357 patients, 27 countries | Rovin 2021, PMID 33971155 |
| Avacopan vs prednisone taper, relapse after remission | 9.4% vs 20.9%; HR 0.43 (0.22–0.85) | ADVOCATE kidney subgroup, 268 patients | Geetha 2025, PMID 40814647 |
Safety and negative-result figures¶
| Measure | Estimate | Population / method | Source |
|---|---|---|---|
| Clinically significant contrast-associated AKI | 1.2% (53/4,418); CA-AKI → 90-day death/dialysis/persistent impairment OR 3.93 (2.82–5.49) | PRESERVE cohort | Weisbord 2020, PMID 32192658 |
| Bicarbonate vs saline / acetylcysteine vs placebo prophylaxis | OR 0.93 (0.72–1.22) / 1.02 (0.78–1.33) | PRESERVE, 4,993 analysed | Weisbord 2018, PMID 29130810 |
| Nephrogenic systemic fibrosis, group II gadolinium in G4–G5 | 0 of 4,931 (0%; upper 95% bound 0.07%) | 16 studies | Woolen 2020, PMID 31816007 |
| Chronic NSAID use and CKD | Pooled OR 1.24 (1.11–1.39); HR 1.50 (1.31–1.70); HR 1.67 (1.38–2.02) with pre-existing CKD; I² >90% | 40 studies, 1,757,118 participants | Soliman 2025, PMID 39412516 |
| Metformin at creatinine >530 µmol/L | Mortality 53% vs 41%; adjusted HR 1.35 (1.20–1.51), dose-dependent | 813 matched to 2,439, Taiwan | Hung 2015, PMID 26094107 |
| Increased water intake, eGFR at 12 months | Urine volume +0.6 L/day (0.5–0.7); eGFR difference −0.3 (−1.8 to 1.2) | CKD WIT, 631 patients | Clark 2018, PMID 29801012 |
| Ultra-processed food and incident CKD | Highest vs lowest quartile HR 1.24 (1.15–1.35); 16.5 vs 14.7 per 1,000 py | 14,679 ARIC participants, median 24 y | Du 2022, PMID 35679994 |
| Ultra-processed food and CKD progression | Tertile 3 vs 1 HR 1.22 (1.04–1.42); 2.61 (1.32–5.18) at eGFR ≥60; null below 60 (p-interaction 0.003) | 1,047 events, CRIC | Sullivan 2023, PMID 37028638 |
| Salt restriction (−80 mEq/day) | Clinic SBP −4.9 mmHg (3.1–6.8); proteinuria −0.39 g/day (0.22–0.55) | 11 RCTs, 738 patients | Garofalo 2018, PMID 29882800 |
| MDRD protein restriction, kidney failure | HR 0.89 (0.71–1.12) overall; 0.68 (0.51–0.93) first 6 y; 1.27 (0.90–1.80) after (interaction p=0.008) | 585 Study A participants, registry follow-up | Levey 2006, PMID 17162142 |
| Achieved protein intake and GFR decline | 0.2 g/kg/day lower intake ↔ 1.15 mL/min/y slower decline (p=0.011) | 255 Study B participants | Levey 1996, PMID 8629624 |
| Bicarbonate vs standard care, creatinine doubling / death | 6.6% vs 17.0% (p<0.001) / 3.1% vs 6.8% (p=0.004) | UBI, 740 patients, open-label, 36 months | Di Iorio 2019, PMID 31598912 |
| Veverimer achieved bicarbonate separation | ~1 mEq/L (22.0 ± 3.0 vs 20.9 ± 3.3 at month 3) | VALOR-CKD, 1,480 patients | Tangri 2024, PMID 38261535 |
Known conflicts and caveats¶
- Prevalence based on a single eGFR or uACR does not prove three-month chronicity and will differ from claims-based diagnosed prevalence.
- GBD values are modelled estimates with uncertainty intervals; they should not be combined with registry counts as though methods were identical (PMID 32061315).
- Composite outcomes differ across DAPA-CKD, EMPA-KIDNEY, CREDENCE, FIDELIO and FLOW; relative effects are not a league table.
- Early-stopped trials can overestimate effects; DAPA-CKD, CREDENCE and FLOW stopped early after prespecified review.
- Dialysis-versus-conservative-care comparisons are observational and strongly selected; pooled hazard ratios are not causal treatment effects (PMID 34507554).
- eGFR slope can separate sooner than kidney failure but is sensitive to acute haemodynamic effects (PMIDs: 34619108, 38061371).
- Albuminuria reduction can fail to translate to eGFR benefit, as the FSGS DUPLEX result demonstrates (PMID 37921461). This caveat applies directly to CONFIDENCE and FINE-ONE, whose endpoints are albuminuria (PMIDs: 40470996, 41780000).
- FIND-CKD's primary endpoint was total eGFR slope, not a clinical event count; its kidney-only composite hazard ratio was 0.78 (0.60–1.01) and crossed 1.0 (PMID 42246672).
- The FIND-CKD glomerular-disease analysis is a prespecified exploratory subgroup of 903 participants, not a trial in glomerular disease (PMID 42246414).
- CONFIDENCE hyperkalaemia percentages are denominators of ~259–265 per arm, not the full randomized set (PMID 41493296).
- The 2017 and 2023 GBD CKD estimates use different age denominators (all ages versus adults ≥20 y) and are not a time series; only the within-model 1990-versus-2023 comparison in PMID 41213283 is internally consistent.
- Pooled KFRE discrimination of 0.90 comes from multinational data with wide case-mix; single-country validations give C-statistics near 0.75–0.85 and calibration failures in both directions (PMIDs: 26757465, 38689834, 38184316).
- Frailty prevalence in CKD ranges 2.8–81.5% because instruments differ; the pooled pre-frailty mortality hazard exceeding the frailty hazard is biologically implausible and indicates residual heterogeneity (PMID 33218914).
- Screening ICERs are model outputs conditioned on US prices and on SGLT2 inhibitor efficacy imported from one trial; transportability is untested (PMIDs: 37216661, 39514193).
- Two urate-lowering trials with unambiguous target engagement — one using iohexol-measured GFR — found no effect on kidney function, so the strong urate–progression association is not evidence of a modifiable target (PMIDs: 32579811, 32579810).
- Proteinuria surrogacy is disease- and mechanism-specific: it translated to eGFR slope and kidney failure in IgA nephropathy (PMID 41910396) and did not in FSGS (PMID 37921461).
- Modality and transplantation comparisons rest on 2 randomized trials among 84 studies (PMID 38899545) and on 48 observational studies with unresolved heterogeneity (PMID 35232772); hazard ratios are not treatment effects.
- Finerenone hyperkalaemia rates were obtained under protocol-scheduled potassium measurement with drug withholding and restart, and do not describe unmonitored use (PMID 34732509).
- Contrast and gadolinium risks are now small in absolute terms; a relative risk of 3.93 attaches to a 1.2% absolute clinically significant event rate (PMID 32192658), and no group II gadolinium NSF case was observed in 4,931 patients with G4–G5 (PMID 31816007).