Diabetic Kidney Disease

Meet the patient

A 58-year-old man with 15 years of type 2 diabetes comes for review. His blood pressure is 148/92 mmHg, his HbA1c is 8.4 percent, and a routine dipstick is positive for protein. A spot UACR is 180 mg/g and his eGFR has slipped from 78 to 64 mL/min over the last year. The fundus shows background retinopathy.[13]

Two exam questions are now live: is this classical DKD, or is something else hiding behind the diabetes? and which of the four pillars is he missing? The first decides whether you biopsy; the second decides whether his kidneys — and his heart — survive the next decade.[1][2]

What DKD is — and the three things it is not

DKD is a specific microvascular complication of diabetes. Most patients are diagnosed clinically, without biopsy, on the co-existence of diabetes with albuminuria (UACR at least 30 mg/g), a declining eGFR and — where present — diabetic retinopathy; the characteristic biopsy findings, when tissue is obtained, are basement membrane thickening, mesangial expansion and Kimmelstiel-Wilson nodules (nodular glomerulosclerosis).[15][13]

It is not inevitable. Historically a third of type 1 diabetics reached ESKD; with intensive therapy that is now rare, and the same therapy applied early in type 2 markedly slows progression.[1][12]

It is not "just diabetes with a rising creatinine". Not every proteinuric diabetic has DKD — a missed glomerulonephritis is the classic exam and clinical pitfall. Short duration, absent retinopathy, haematuria or a rapid decline all mandate biopsy.[1]

It is not a one-drug disease anymore. The single most important conceptual shift since the 2010s is that SGLT2 inhibitors are disease-modifying independent of glucose lowering (CREDENCE, DAPA-CKD, EMPA-KIDNEY), and finerenone adds further protection on top of RAAS blockade. DKD is now managed as a chronic, combination-treated disease.[3][4][5][9]

The Mogensen ladder — five stages and the silent middle

Classical DKD climbs a predictable five-stage ladder, and the treatable window is the silent middle. Stating the natural history is a guaranteed viva point.[1]

• Stage I — glomerular hyperfiltration and hypertrophy (at diagnosis). GFR is raised, kidney size increased, albuminuria normal; reversible with good control.
• Stage II — silent/latent phase (over years). Histological lesions present; albuminuria normal at rest but may rise with exercise or poor control.
• Stage III — incipient nephropathy (microalbuminuria). Persistent UACR 30 to 300 mg/g, below the detection threshold of the routine dipstick; GFR may be normal or raised; blood pressure begins to rise. This is the earliest clinically detectable stage — the treatment window.[13]
• Stage IV — overt nephropathy (macroproteinuria). UACR over 300 mg/g — now in the dipstick-positive range; GFR declines steadily; hypertension is universal; oedema may develop.
• Stage V — ESKD (G5). GFR under 15 mL/min/1.73 m²; requires renal replacement therapy.[20]

The KDIGO heat-map — albuminuria times GFR

Two axes, one heat-map, and it drives both risk and referral. The KDIGO albuminuria categories (A1–A3) crossed with the GFR categories (G1–G5) place every patient in a green, yellow, orange or red cell.[1]

[13] [3]

KDIGO GFR categories:[20]

• G1 — 90 or above | G2 — 60 to 89 | G3a — 45 to 59 | G3b — 30 to 44 | G4 — 15 to 29 (plan RRT) | G5 — under 15 (kidney failure; initiate RRT).[20]

The intersection gives the risk: patients in G3b or worse with A3 are at the highest risk of progression and of cardiovascular events, and are referred for nephrology care.[20]

How common, and who

DKD is the leading single cause of ESKD worldwide — roughly 30 to 50 percent of all patients reaching dialysis or transplantation in industrialised countries, and an even larger share of incident ESKD in India, China, the Middle East and the Pacific Islands. It develops in roughly 30 to 40 percent of patients with either type 1 or type 2 diabetes over their lifetime.[1]

Non-modifiable risk factors: long duration of diabetes (the dominant factor — DKD is rare within the first 5 years of T1DM), older age at onset, male sex, family history of DKD or hypertension, and genetic susceptibility with ethnic clustering — Pima Indians, South Asians, Black Americans, Hispanic Americans, Pacific Islanders. Variants in UMOD, APOL1 (African ancestry) and ACACB modulate risk, but no gene is yet used clinically.[1]

Modifiable risk factors (the targets of therapy):[1]

• Chronic hyperglycaemia — the central driver; the DCCT (T1DM) and UKPDS (T2DM) showed intensive control reduces microalbuminuria and overt nephropathy, with a durable metabolic memory effect.
• Hypertension — systolic over 130 mmHg doubles the rate of GFR decline.
• Smoking — independent risk for progression and cardiovascular death.
• Obesity and dyslipidaemia — promote inflammation and progression.
• Dietary protein and sodium intake — high intake accelerates hyperfiltration.
• Insulin resistance (T2DM) — drives hyperfiltration independent of glucose.[1]

The retinopathy correlation — and its trap. In type 1 diabetes the correlation between DKD and proliferative retinopathy is strong: a proteinuric patient almost always has retinopathy, and its absence is a red flag for an alternative diagnosis. In type 2 diabetes the correlation is weaker — 20 to 30 percent of classical DKD patients have no detectable retinopathy — so absent retinopathy is a softer pointer in T2DM.[1]

Why the glomerulus scars — three converging axes

The pathogenesis integrates a haemodynamic, a metabolic, and an inflammatory/fibrotic axis — and every modern therapy interrupts one of them. Understanding each is exactly what the examiner probes at viva depth.[1]

Axis 1 — Haemodynamic (glomerular hypertension). Chronic hyperglycaemia activates the intrarenal RAAS, raising local angiotensin II, which preferentially constricts the efferent arteriole, raising intraglomerular capillary pressure and producing hyperfiltration. The pressure and flow stress podocytes and the GBM and stimulate mesangial matrix. This is the mechanistic rationale for RAAS blockade: ACE inhibitors and ARBs lower intraglomerular pressure and reduce proteinuria beyond blood-pressure lowering alone.[10]

Axis 2 — Metabolic (intracellular glucose toxicity). Persistently raised intracellular glucose over-activates four biochemical pathways:[1]

• Polyol — aldose reductase converts glucose to sorbitol, consuming NADPH and generating osmotic stress.
• Hexosamine — generates UDP-GlcNAC that O-glycosylates transcription factors (Sp1), up-regulating TGF-beta and PAI-1.
• Protein kinase C (PKC) — diacylglycerol activates PKC isoforms, up-regulating VEGF, TGF-beta, NF-kappaB and extracellular matrix.
• Advanced glycation end-products (AGEs) — non-enzymatic glycation forms irreversible cross-links binding the RAGE receptor, driving inflammation, oxidative stress and matrix accumulation.[1]

The net result is reactive oxygen species, mitochondrial dysfunction, and up-regulation of profibrotic cytokines — TGF-beta and CTGF — driving mesangial matrix expansion and tubulointerstitial fibrosis.[1]

Axis 3 — The structural lesions (the histology the examiner wants you to draw).[1]

• Diffuse glomerulosclerosis — mesangial matrix expansion and GBM thickening; the commonest lesion, not specific to diabetes.
• Nodular glomerulosclerosis (Kimmelstiel-Wilson nodules) — rounded, acellular, eosinophilic nodules in the mesangial centre of glomerular lobules; pathognomonic of diabetes (though similar nodules occur in light-chain deposition disease and amyloidosis, distinguished by immunofluorescence and Congo red).
• Arteriolar hyalinosis — affecting both afferent and efferent arterioles (the latter highly specific to diabetes, unlike hypertension, which affects only the afferent).[1]

Additional lesions include exudative lesions (fibrin caps, capsular drop), tubular basement membrane thickening, and interstitial fibrosis with tubular atrophy — and it is the degree of interstitial fibrosis that best correlates with the rate of GFR decline, more than the glomerular changes themselves.[1]

Why some patients deviate — non-albuminuric DKD. A substantial minority (especially T2DM) show a progressive fall in GFR without significant albuminuria, attributed to tubulointerstitial and vascular injury and to prior RAAS blockade lowering albuminuria. It is not benign — these patients still progress and still carry high cardiovascular risk.[1][2]

The bedside round — kidney plus the rest of the microvasculature

Early DKD is asymptomatic and detected only by annual UACR and eGFR. Examination integrates kidney-specific findings with the systemic micro- and macrovascular disease that travels with it — because DKD is the renal face of a multi-system process.[1]

• Kidney and volume — blood pressure in both arms (target under 130/80 mmHg in diabetes or kidney disease), volume status (JVP, sacral and ankle oedema, basal crackles), and consequences of advanced CKD (pallor of anaemia, uraemic scratch marks, asterixis). Bedside urinalysis: dipstick-positive protein means macroproteinuria, but a negative dipstick does not exclude microalbuminuria — a spot UACR is mandatory.[13][2]
• Integrated cardiovascular and microvascular exam — fundus for retinopathy (its presence supports classical DKD; its absence in T1DM is a red flag), heart for ischaemic disease, heart failure and an S4 gallop, peripheral pulses and carotid/abdominal/femoral bruits, feet for neuropathy (10 g monofilament, 128 Hz tuning fork) and ulcers, and postural blood pressure for autonomic neuropathy. The cardiovascular burden is what these patients die of.[1]

Investigations — screen, stage, exclude

Investigations serve three purposes: screen and detect early, stage established DKD, and exclude the mimics.[1]

First-line screening bundle for every diabetic:[1][2]

• Spot UACR on a random sample — the earliest marker, below dipstick detection. A1 under 30, A2 30 to 300, A3 over 300 mg/g.
• Serum creatinine with eGFR using a current creatinine-based equation (CKD-EPI).
• HbA1c, lipid profile, blood pressure and weight/BMI.[13][14]

Screening schedule: annual UACR screening in every person with diabetes — from diagnosis in T2DM, from 5 years post-diagnosis in T1DM — confirmed on repeat sampling before labelling persistent albuminuria, excluding the false-positive triggers: UTI, heavy exercise in the prior 24 hours, fever, heart failure decompensation, uncontrolled hypertension, menstruation, and marked hyperglycaemia.[14][13]

Confirmatory and staging tests: eGFR (CKD-EPI 2021; use cystatin C-based eGFR if creatinine is unreliable — low/high muscle mass, amputation, pregnancy); renal ultrasound (exclude obstruction; small symmetric kidneys suggest advanced CKD, asymmetric kidneys suggest renovascular disease; DKD kidneys are often normal or enlarged early, small late); urinalysis and microscopy (exclude haematuria and active sediment); ECG and cardiac work-up.[1]

Excluding alternative diagnoses (when features are atypical): serum immunoglobulins, serum free light chains and SPEP (myeloma), ANA, anti-dsDNA, ANCA, complement (lupus, vasculitis), hepatitis B, C and HIV serology (membranous, cryoglobulinaemia), anti-PLA2R (primary membranous), cryoglobulins and rheumatoid factor, and renal biopsy.[2]

When to biopsy — the atypical list, verbatim

Renal biopsy is reserved for features that do not fit classical DKD. Know these verbatim:[15]

1. Rapidly declining kidney function out of keeping with the prior course.
2. Acute-onset nephrotic-range proteinuria.
3. Active urinary sediment — haematuria, dysmorphic red cells, red-cell casts.
4. Short duration of diabetes.
5. Absence of diabetic retinopathy.
6. Suspicion of immune-mediated or other non-diabetic glomerular disease (rash, arthralgia, haemoptysis, systemic features).[15][19]

The differential — not every proteinuric diabetic has DKD

Missing an alternative glomerulonephritis is the classic pitfall. The discriminator line examiners reward: albuminuria plus long-standing diabetes plus retinopathy equals classical DKD; anything else mandates a renal ultrasound and consideration of biopsy.[1][2]

[19] [15]

The can't-miss diagnoses in a diabetic with a rising creatinine or new proteinuria: AKI (volume depletion, sepsis, contrast, NSAIDs, or the triple whammy), renal artery stenosis (a marked eGFR fall on RAAS blockade), obstructive uropathy (older men with autonomic neuropathy — check the bladder scan), contrast-induced nephropathy, and superimposed glomerulonephritis — biopsy series find non-diabetic renal disease in roughly half of type 2 diabetics biopsied for atypical features.[19]

The four pillars — combination therapy, deployed early

The modern paradigm is a four-pillar combination deployed early, with escalation to renal replacement therapy planning as eGFR falls. Remember it as R-S-F-M: RAAS blockade, SGLT2 inhibitor, Finerenone, Multifactorial.[1]

Pillar 1 — RAAS blockade (ACE inhibitor or ARB)

RAAS blockade is first-line in albuminuric DKD. The mechanism — efferent arteriolar dilatation lowering intraglomerular pressure — reduces proteinuria beyond blood-pressure lowering. In T1DM, the captopril trial (Lewis 1993) halved the combined endpoint of death, dialysis and transplantation.[6] In T2DM, IDNT (irbesartan) and RENAAL (losartan) each reduced the risk of doubling creatinine or ESKD by about 16 to 20 percent versus placebo.[7][8]

Key points:[1]

• Use either an ACE inhibitor or an ARB — telmisartan was equivalent to ramipril in ONTARGET, with less cough and less angioedema.[16]
• Indicated in albuminuric DKD — the landmark renoprotection trials all enrolled albuminuric patients.[6][7]
• Titrate to the maximum tolerated licensed dose — losartan 50 to 100 mg once daily (RENAAL) or irbesartan 300 mg daily (IDNT); ramipril 10 mg daily was the ONTARGET comparator.[7][8][16]
• Monitor creatinine and potassium after initiation or dose change — a modest creatinine rise is expected; a marked fall in eGFR should prompt review and imaging for renal artery stenosis.[2]
• Do not combine an ACE inhibitor with an ARB — ONTARGET showed dual blockade added hypotensive symptoms, syncope and renal dysfunction (13.5 vs 10.2 percent) with no increase in benefit.[16]
• Continue RAAS blockade in advanced CKD unless hyperkalaemia or hypotension prevent it.[14]

Pillar 2 — SGLT2 inhibitor

SGLT2 inhibitors are the most important recent change in DKD management — disease-modifying, independent of glucose lowering. Trial populations extended to eGFR 20 mL/min/1.73 m² — EMPA-KIDNEY enrolled patients with an eGFR of at least 20 — and the benefit was consistent in patients with and without diabetes, evidence that protection is not a glucose-lowering effect.[3][4][5]

The three pivotal trials:[1]

• CREDENCE (2019) — canagliflozin in T2DM with albuminuric CKD (eGFR 30 to 90, UACR over 300). Reduced the composite of ESKD, doubling of creatinine, or renal/CV death by 30 percent; stopped early for efficacy.[3]
• DAPA-CKD (2020) — dapagliflozin in CKD with and without diabetes (eGFR 25 to 75, UACR over 200). Reduced the composite by 39 percent; benefit identical in diabetics and non-diabetics.[4]
• EMPA-KIDNEY (2023) — empagliflozin in a broad CKD population including eGFR down to 20. Reduced progressive kidney disease or CV death by 28 percent, extending the benefit to lower eGFR and less albuminuric phenotypes.[5]

[3] [4] [5] [17]

Practical use:[14]

• Trial evidence supports use down to eGFR 20; the kidney protection is independent of glycaemic control.[4][5]
• Monitor volume status, and check ketones when unwell.
• Euglycaemic DKA: suspect it in any SGLT2i-treated patient with nausea, vomiting or malaise, or a metabolic acidosis — evaluate urine and/or serum ketones promptly.[17]
• Type 1 diabetes: SGLT2 inhibitors should be used only with great caution, extensive counselling and close monitoring.[17]

Pillar 3 — Finerenone (non-steroidal MRA)

Finerenone adds kidney and cardiovascular protection on top of ACE inhibitor or ARB therapy. FIDELIO-DKD (2020) reduced the primary composite kidney outcome by 18 percent (hazard ratio 0.82) in CKD with type 2 diabetes on maximally tolerated RAS blockade; hyperkalaemia-related discontinuation was 2.3 percent versus 0.9 percent with placebo.[9] Earlier dose-ranging trials showed albuminuria reduction across once-daily doses from 1.25 to 25 mg.[11]

[9] [11]

Pillar 4 — Multifactorial risk reduction

The Steno-2 study proved that intensive, multifactorial intervention (RAAS blockade, BP and glycaemic control, statins, aspirin, lifestyle) in T2DM with microalbuminuria reduced cardiovascular events, nephropathy, retinopathy and autonomic neuropathy — and at 21 years the cohort gained an average of 8 years of life.[12] Multifactorial care is the foundation of all four-pillar therapy.

Glycaemic control. Target an individualised HbA1c — under 7 percent for most, relaxed in advanced CKD, frailty or recurrent hypoglycaemia (ADA/KDIGO). As eGFR falls, the diabetes drug choice changes:[14][13]

• Metformin — first-line; dose-reduce once eGFR falls below 45 and stop below 30 (lactic acidosis risk, though rare).[14]
• Sulfonylureas — avoid in advanced CKD (hypoglycaemia risk).[14]
• SGLT2 inhibitor — recommended for kidney protection.[14]
• GLP-1 receptor agonist — recommended where an SGLT2 inhibitor is not tolerated or targets are unmet.[14]
• Insulin — doses often need reduction as eGFR falls (insulin clearance drops).[14]

Blood pressure control. Target under 130/80 mmHg in diabetes with kidney disease (National Kidney Foundation). Use ACE inhibitor or ARB first-line; add a dihydropyridine calcium channel blocker or a thiazide-type diuretic as second and third agents; loop diuretics replace thiazides in advanced CKD.[13][14]

Lipid management. A statin is indicated for adults with diabetes and CKD — KDIGO recommends statin treatment for all adults with CKD aged 50 or older — because cardiovascular disease is the leading cause of death in this population.[14][20]

Lifestyle. Smoking cessation, dietary protein around 0.8 g/kg/day, sodium restriction under 2 g/day, weight management, regular exercise and alcohol moderation are the foundation on which drug therapy is layered (ADA/KDIGO).[14]

Escalation triggers and referral

Refer to nephrology when:[20][15]

• eGFR under 30 mL/min/1.73 m² (plan for RRT).
• A3 albuminuria (over 300 mg/g).
• Rapidly declining eGFR.
• Refractory hypertension or hyperkalaemia.
• Atypical features requiring biopsy.
• Nephrotic-range proteinuria.[15]

Plan vascular access (arteriovenous fistula) and transplant assessment in suitable candidates as eGFR falls towards end-stage kidney disease; initiate dialysis on clinical grounds — symptoms and complications — not on a single number.[20]

The resuscitation scenarios — when DKD decompensates acutely

DKD is chronic, but decompensation is acute and life-threatening. The scenarios the examiner expects you to handle:[1]

1. Hyperkalaemia in DKD. Common — reduced potassium excretion plus RAAS blockade. ECG changes (peaked T waves, widened QRS) warrant emergency treatment: intravenous calcium gluconate, insulin with glucose, nebulised salbutamol (albuterol) and sodium bicarbonate for short-term plasma shifts in acidosis; for chronic control, withdraw exacerbating drugs and add potassium binders (sodium polystyrene sulfonate, patiromer, sodium zirconium cyclosilicate).[21]

2. Euglycaemic DKA on an SGLT2 inhibitor. SGLT2 inhibitors can precipitate euglycaemic DKA — glucose under 250 mg/dL (about 14 mmol/L) with anion-gap metabolic acidosis and ketosis — typically with starvation, infection, surgery, alcohol, pregnancy or very low carbohydrate intake. Suspect it in any SGLT2i-treated patient with nausea, vomiting, malaise or fatigue, or an unexplained metabolic acidosis — check urine and/or serum ketones. Management: resuscitation with IV fluids, insulin and glucose, with treatment of the underlying precipitant.[18][17]

3. AKI superimposed on DKD. Precipitants: volume depletion (vomiting, diuretics, sepsis), contrast, NSAIDs (alone or as the triple whammy), sepsis, hypoglycaemia. Manage ABCDE, stop nephrotoxins, resuscitate with IV fluids, treat the precipitant, withhold ACEi/ARB, SGLT2i, metformin and NSAIDs temporarily, and recheck renal function and potassium daily. Resume disease-modifying therapy once eGFR returns to baseline.[2]

4. Flash pulmonary oedema / accelerated hypertension. Sit the patient up, high-flow oxygen, IV loop diuretic, consider IV nitrate if hypertensive, and investigate for renal artery stenosis if eGFR fell markedly on starting RAAS blockade.[2]

Subtypes and scenarios

Type 1 vs type 2 diabetes. Pathophysiology and histology are essentially identical, but the clinical course and screening differ. In T1DM latency to microalbuminuria is 5 to 15 years, screening starts at 5 years post-diagnosis (or puberty), the retinopathy correlation is strong, and intensive insulin therapy (DCCT/EDIC — metabolic memory) markedly reduces ESKD. In T2DM DKD can be present at diagnosis, screening is annual from diagnosis, the retinopathy correlation is weaker (20 to 30 percent have none), and patients carry a much heavier macrovascular burden (the usual cause of death).[1]

Non-albuminuric DKD (GFR-decline phenotype). Recognised in T2DM: a progressive fall in eGFR without significant albuminuria. These patients still progress — EMPA-KIDNEY enrolled patients with eGFR 20 to below 45 irrespective of albuminuria, and empagliflozin reduced kidney disease progression in that broad population.[5]

Refractory oedema and heavy albuminuria. Treat with loop diuretics and salt restriction; adding finerenone further reduces albuminuria on top of RAAS blockade.[11] Heavy albuminuria itself marks rapid progression and high cardiovascular risk.[9]

Cardiorenal / Cardiovascular-Kidney-Metabolic (CKM) syndrome. DKD is the renal manifestation of a systemic cardiometabolic disorder. The SGLT2 inhibitor and finerenone benefit heart failure as well as kidney, supporting a unified cardiorenal therapy.[3][9]

Complications and pitfalls

Kidney-related: progression to ESKD (now substantially deferred by combination therapy), nephrotic syndrome, AKI superimposed on CKD (NSAIDs, contrast, sepsis, triple whammy), hyperkalaemia, anaemia of CKD (which develops earlier in DKD than in non-diabetic CKD), CKD-MBD (high phosphate, low calcitriol, secondary hyperparathyroidism, vascular calcification), metabolic acidosis, and increased infection risk.[1]

Systemic: the dominant one is accelerated atherosclerosis — coronary artery disease, stroke and peripheral arterial disease are the leading causes of death. Other: progression of diabetic retinopathy, neuropathy (sensory, autonomic, gastroparesis), foot ulcers and amputation, erectile dysfunction, and hypoglycaemia unawareness.[1]

Drug-related pitfalls (high-yield):[1]

• The triple whammy — NSAID + ACEi/ARB + diuretic to AKI. Avoid the combination; stop NSAIDs in any CKD patient.
• Hypoglycaemia with insulin and sulfonylureas as eGFR falls — reduce doses, prefer glipizide if a sulfonylurea is needed.
• Metformin lactic acidosis — stop below eGFR 30 or in acute illness.
• SGLT2i — euglycaemic DKA, volume depletion, genital mycotic infection, rare Fournier gangrene; stop perioperatively and in acute illness.[3]
• RAAS blockade — hyperkalaemia, fall in eGFR (under 30 percent acceptable), contraindicated in bilateral renal artery stenosis.
• Statins — start at diagnosis in all diabetics with CKD aged 40 to 75; stop in advanced CKD if not already on, continue if established.[2]

Classic diagnostic pitfalls:[1]

• Missing an alternative glomerulonephritis because "it is just diabetes" — always check for haematuria, short diabetes duration, absent retinopathy.
• Not screening with UACR — the dipstick misses microalbuminuria.
• Under-dosing RAAS blockade — the benefit is dose-dependent; titrate to max tolerated.
• Forgetting finerenone — a third pillar of kidney protection.
• Failing to plan dialysis access early — an arteriovenous fistula needs months to mature.
• Stopping ACEi/ARB for a modest creatinine rise — under 30 percent is expected and benign.[1]

Prognosis and disposition

The strongest predictors of progression: clinically, the albuminuria level and the rate of eGFR decline; histologically, tubulointerstitial fibrosis and tubular atrophy predict progression better than the glomerular lesions, and arteriolar hyalinosis carries independent renal and cardiovascular prognostic weight.[15][1]

Mortality. Patients with DKD have a two- to four-fold higher mortality than non-diabetic CKD, and cardiovascular disease is the leading cause of death — more die with DKD than ever reach dialysis. This is why statins, blood-pressure control and SGLT2i (which reduce CV events) are central, not adjunctive.[1]

Disposition. Primary care manages stable early DKD (A1/G1-2, A2/G1-2 with stable eGFR). Refer to nephrology for atypical features, eGFR under 30, A3 albuminuria, rapidly declining GFR, refractory hypertension, or biopsy. Coordinate integrated care with ophthalmology, podiatry, cardiology, and diabetes/endocrinology.[2]

Impact of modern therapy. With early detection and the four-pillar combination, DKD progression can be markedly slowed. The Steno-2 21-year follow-up showed an 8-year gain in median lifespan with intensive multifactorial intervention.[12] Historically a third of T1DM patients developed ESKD; with DCCT-era control and modern combination therapy this is now well under 10 percent.[12]

Special populations

Pregnancy. Pre-conception counselling is essential. SGLT2 inhibitors are stopped around conception and in pregnancy — pregnancy is itself a recognised low-glucose state that predisposes to euglycaemic ketoacidosis.[18] A full medication review before conception is standard in diabetes care planning — including switching agents that are unsafe in pregnancy — and CKD in pregnancy is co-managed with obstetric medicine and nephrology.[2]

The elderly. Less tight glycaemic target (HbA1c 7.5 to 8.5 percent) to avoid hypoglycaemia and falls; less tight blood pressure (under 140/90 is reasonable in the frail). The SGLT2i benefit persists, but watch volume status (especially with loop diuretics) and stop in acute illness or poor oral intake. Continue RAAS blockade unless hyperkalaemia or hypotension intervene.[1]

Children and adolescents. In T1DM screen from puberty plus 5 years after diagnosis (whichever first) — puberty accelerates microvascular complications. No licensed SGLT2i indications in children for DKD at present; management focuses on intensive glycaemic control and early RAAS blockade for albuminuria.[1]

The immunocompromised and post-transplant patient. Calcineurin inhibitors (ciclosporin, tacrolimus) and steroids worsen hypertension, glucose intolerance and proteinuria. Diabetic nephropathy can recur in a transplanted kidney, typically 10 to 15 years post-transplant, managed with the same four-pillar approach.[2]

Evidence, guidelines and regional differences

Landmark trials and what they changed:[1]

• Captopril trial, Lewis 1993 (NEJM) — captopril halved the risk of death, dialysis or transplantation in T1DM nephropathy. Established ACE inhibitor as standard of care in T1DM DKD.[6]
• IDNT (2001) — irbesartan and RENAAL (2001) — losartan — ARBs reduced the risk of doubling creatinine or ESKD by 16 to 20 percent in T2DM nephropathy. Established ARB as standard of care in T2DM DKD.[7][8]
• CREDENCE (2019) — canagliflozin — first trial to show an SGLT2i reduced ESKD in T2DM nephropathy by 30 percent; stopped early. Transformed the field.[3]
• DAPA-CKD (2020) — dapagliflozin — extended benefit to non-diabetic CKD and lower eGFR; 39 percent risk reduction.[4]
• EMPA-KIDNEY (2023) — empagliflozin — extended benefit to eGFR as low as 20 and less albuminuric phenotypes; 28 percent risk reduction.[5]
• FIDELIO-DKD (2020) — finerenone — added a fourth pillar: 18 percent reduction in kidney outcomes on top of RAAS blockade.[9]
• Steno-2 (1993–2016, 21-year follow-up) — intensive multifactorial intervention gained 8 years of life in T2DM with microalbuminuria.[12]

The metabolic memory / legacy effect. The DCCT/EDIC follow-up in T1DM and UKPDS post-trial follow-up in T2DM both showed that early intensive glycaemic control produces benefits that persist for decades, even if later control worsens — metabolic memory. This is why early, intensive therapy in the first years of diabetes is disproportionately important.[1]

[23] [14] [20] [22]

Controversies:[1]

• ACEi vs ARB equivalence — telmisartan equivalent to ramipril with better tolerability; combination is harmful (ONTARGET: more hypotension, syncope and renal dysfunction without added benefit).[16]
• Dual SGLT2i + GLP-1 RA — both classes carry kidney and cardiovascular benefit in the ADA/KDIGO consensus; long-term safety and cost data still evolving.[14]
• Low-protein diet (about 0.8 g/kg/day) — modest slowing; hard to adhere; reserved for selected patients.[14]
• SGLT2i in T1DM — use only with great caution, extensive counselling and close monitoring (euglycaemic DKA).[17]
• Metformin in advanced CKD — dose-reduce below eGFR 45 and stop below 30, rather than abandoning the drug in moderate CKD.[14]

The mantra

Screen with UACR every year; biopsy only if atypical; treat all four pillars early; remember the heart kills before the kidney. Most diabetics die with their kidneys, not of them.[1][12]

Ward-round test

A diabetic with a UACR of 180 mg/g and a dipstick that is negative — explain.

The routine dipstick does not become positive until protein excretion exceeds roughly 300 mg per day, so it misses microalbuminuria (30 to 300 mg/g) — the earliest clinical sign of diabetic nephropathy. The albumin-to-creatinine ratio on a spot morning sample, checked annually, is the recommended screen. This patient has incipient nephropathy: start combination therapy.[13][14]

A type 1 diabetic with proteinuria and a normal fundus — what do you do?

In T1DM the retinopathy correlation is strong, so absent retinopathy is a red flag. This is atypical for classical DKD — biopsy to exclude a superimposed glomerulonephritis. (In T2DM the correlation is weaker, so the pointer is softer.)[1]

Creatinine rises 35 percent two weeks after starting ramipril — your move?

A rise under 30 percent is expected and benign — keep going. A rise over 30 percent mandates stopping the drug and imaging for renal artery stenosis. Do not abandon RAAS blockade for a modest bump.[1]

A diabetic on dapagliflozin presents with vomiting, abdominal pain and a glucose of 9 mmol/L — what is this?

Euglycaemic DKA — an SGLT2i hazard: glucose under 250 mg/dL with ketosis, typically with fasting, illness, surgery or very low carbohydrate intake. Stop the drug, give IV fluids, insulin with glucose, correct electrolytes, treat the precipitant. The trap is missing DKA because the glucose is not high.[18][17]

Name the four pillars, and the one trial that anchors each.

RAAS blockade (captopril/Lewis 1993 in T1DM; IDNT, RENAAL in T2DM), SGLT2 inhibitor (CREDENCE, DAPA-CKD, EMPA-KIDNEY), Finerenone (FIDELIO-DKD), Multifactorial (Steno-2).[6][7][9][12]

References

1. [1]Thomas MC, Brownlee M, Susztak K, et al. Diabetic kidney disease Nat Rev Dis Primers, 2015.PMID 27188921
2. [2]Ambalavanan J, Caramori ML. Management of Diabetes in Patients with Chronic Kidney Disease Endocr Res, 2025.PMID 40119502
3. [3]Perkovic V, Jardine MJ, Neal B, et al. Canagliflozin and Renal Outcomes in Type 2 Diabetes and Nephropathy N Engl J Med, 2019.PMID 30990260
4. [4]Heerspink HJL, Stefánsson BV, Correa-Rotter R, et al. Dapagliflozin in Patients with Chronic Kidney Disease N Engl J Med, 2020.PMID 32970396
5. [5]Herrington WG, Staplin N, Wanner C, et al. Empagliflozin in Patients with Chronic Kidney Disease N Engl J Med, 2023.PMID 36331190
6. [6]Lewis EJ, Hunsicker LG, Bain RP, Rohde RD. The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. The Collaborative Study Group N Engl J Med, 1993.PMID 8413456
7. [7]Lewis EJ, Hunsicker LG, Clarke WR, et al. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes N Engl J Med, 2001.PMID 11565517
8. [8]Brenner BM, Cooper ME, de Zeeuw D, et al. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy N Engl J Med, 2001.PMID 11565518
9. [9]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.PMID 33264825
10. [10]Ruggenenti P, Cravedi P, Remuzzi G. The RAAS in the pathogenesis and treatment of diabetic nephropathy Nat Rev Nephrol, 2010.PMID 20440277
11. [11]Bakris GL, Agarwal R, Chan JC, et al. Effect of Finerenone on Albuminuria in Patients With Diabetic Nephropathy: A Randomized Clinical Trial JAMA, 2015.PMID 26325557
12. [12]Gæde P, Oellgaard J, Carstensen B, et al. Years of life gained by multifactorial intervention in patients with type 2 diabetes mellitus and microalbuminuria: 21 years follow-up on the Steno-2 randomised trial Diabetologia, 2016.PMID 27531506
13. [13]Bakris GL. Microalbuminuria: what is it? Why is it important? What should be done about it? J Clin Hypertens (Greenwich), 2001.PMID 11416691
14. [14]de Boer IH, Khunti K, Sadusky T, et al. Diabetes Management in Chronic Kidney Disease: A Consensus Report by the American Diabetes Association (ADA) and Kidney Disease: Improving Global Outcomes (KDIGO) Diabetes Care, 2022.PMID 36189689
15. [15]Pieczaba M, Dubniański B, Kuźnik Z, et al. Reappraising Kidney Biopsy in Diabetic Kidney Disease: Histopathology, Clinical Course, and the Future of Precision Nephrology J Clin Med, 2026.PMID 42513273
16. [16]The ONTARGET Investigators; Yusuf S, Teo KK, Pogue J, et al. Telmisartan, ramipril, or both in patients at high risk for vascular events. N Engl J Med, 2008.PMID 18378520
17. [17]Peters AL, Buschur EO, Buse JB, et al. Euglycemic Diabetic Ketoacidosis: A Potential Complication of Treatment With Sodium-Glucose Cotransporter 2 Inhibition. Diabetes Care, 2015.PMID 26078479
18. [18]Long B, Lentz S, Koyfman A, Gottlieb M. Euglycemic diabetic ketoacidosis: Etiologies, evaluation, and management. Am J Emerg Med, 2021.PMID 33626481
19. [19]Kritmetapak K, Anutrakulchai S, Pongchaiyakul C, Puapairoj A. Clinical and pathological characteristics of non-diabetic renal disease in type 2 diabetes patients. Clin Kidney J, 2018.PMID 29942497
20. [20]Levin A, Ahmed SB, Carrero JJ, et al. Executive summary of the KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease: known knowns and known unknowns. Kidney Int, 2024.PMID 38519239
21. [21]McCullough PA, Beaver TM, Bennett-Guerrero E, et al. Acute and chronic cardiovascular effects of hyperkalemia: new insights into prevention and clinical management. Rev Cardiovasc Med, 2014.PMID 24762462
22. [22]Mathan Kumar S, Essakky S, Rajasulochana SR, et al. Cost-effectiveness of population-based screening for microalbuminuria in people with type 2 diabetes mellitus in India. Int J Technol Assess Health Care, 2023.PMID 37960938
23. [23]Forbes AK, Hinton W, Feher MD, et al. Implementation of chronic kidney disease guidelines for sodium-glucose co-transporter-2 inhibitor use in primary care in the UK: a cross-sectional study. EClinicalMedicine, 2024.PMID 38304744