Diabetic kidney disease is where this volume comes together. It is the commonest cause of kidney failure in the world, the disease in which the four pillars were largely tested and proven, and the clearest illustration of how mechanism dictates treatment. The glucose-driven hyperfiltration that defines early diabetic kidney injury is the very process SGLT2 inhibition reverses; the proteinuria is the mediator RAAS blockade and finerenone lower; and the result is a regimen, assembled across the previous chapters, that has transformed the outlook for these patients.
How diabetes injures the kidney
Hyperglycaemia damages the kidney on two fronts. Metabolically, excess glucose drives advanced glycation end-products, the polyol and hexosamine pathways, protein kinase C activation, and oxidative stress, all converging on glomerular and tubular injury. Haemodynamically — and this is the part that ties directly to treatment — high filtered glucose is reabsorbed in the proximal tubule together with sodium through SGLT2, so less sodium reaches the macula densa; the macula densa reads the low sodium as under-filtration and, through tubuloglomerular feedback, dilates the afferent arteriole. The result is glomerular hyperfiltration, the earliest functional hallmark of diabetic kidney disease, which raises intraglomerular pressure and feeds the final common pathway of Chapter 2. The elegance, met in Chapter 5, is that an SGLT2 inhibitor blocks exactly this step — restoring sodium delivery, re-engaging tubuloglomerular feedback, and lowering the hyperfiltration that diabetes created. Structurally, the injury shows as mesangial expansion, basement-membrane thickening, podocyte loss, arteriolar hyalinosis, and the classic nodular glomerulosclerosis.
The course — classic and non-albuminuric
The textbook course of diabetic kidney disease proceeds in recognisable stages: early glomerular hyperfiltration, then moderately increased albuminuria, then severe albuminuria, then a declining GFR toward end-stage disease. Albuminuria has long been the central marker, and its appearance and progression track the disease. But a crucial modern refinement is the recognition of non-albuminuric diabetic kidney disease — a substantial proportion of patients, particularly in type 2 diabetes, who lose GFR without ever developing significant albuminuria. The practical lesson is not to require albuminuria to diagnose DKD or to monitor only the urine protein: a falling eGFR in a diabetic patient is diabetic kidney disease until proven otherwise, albuminuric or not, and both axes — GFR and albuminuria — must be tracked, as the staging of Chapter 1 insisted.
Diagnosing it — and when to doubt it
In most cases the diagnosis is clinical, made without a biopsy: long-standing diabetes, albuminuria, and — importantly — diabetic retinopathy, which is strongly concordant with diabetic kidney disease, especially in type 1 diabetes where the two almost always travel together. A typical course and the absence of features pointing elsewhere complete the picture. The skill, then, is knowing when to doubt the easy diagnosis and look for a non-diabetic kidney disease that diabetes is merely accompanying. The warning signs are specific: absent retinopathy, particularly in a type 1 patient with renal disease; a rapid GFR decline; nephrotic-range or rapidly developing proteinuria; an active urinary sediment with haematuria or red-cell casts; a short duration of diabetes insufficient to explain the renal disease; or systemic features suggesting another condition. Any of these should prompt consideration of a kidney biopsy, because a treatable alternative — a glomerulonephritis, say — would otherwise be missed under the assumption of diabetic disease. All diabetics, meanwhile, are screened at least annually with an albumin-to-creatinine ratio and eGFR.
The four pillars converge
The treatment of diabetic kidney disease is the assembled regimen of the previous chapters, and DKD is where each pillar earned its place. RAAS blockade, foundational in proteinuric disease, was proven in the diabetic nephropathy trials. SGLT2 inhibitors, whose mechanism so precisely matches diabetic hyperfiltration, were proven in diabetic CKD and then beyond. Finerenone, the non-steroidal mineralocorticoid antagonist, was tested specifically in albuminuric diabetic CKD on RAAS blockade. GLP-1 receptor agonists, with their weight, glycaemic, cardiovascular, and now kidney benefits, were proven in diabetic CKD. So the modern DKD regimen stacks RAAS blockade and an SGLT2 inhibitor for essentially all, adds finerenone for residual albuminuria, and increasingly adds a GLP-1 agonist — all on a base of blood-pressure control, statin therapy, and the diet and lifestyle measures of Chapter 6, with proteinuria as the shared target. No other CKD population has so complete a set of disease-modifying therapies, and assembling them is the central act of DKD care.
Glycaemic control: necessary, not sufficient
Glycaemic control matters, but its role must be placed correctly. Lowering glucose reduces the microvascular complications of diabetes, including the early development of kidney disease, with a legacy effect that persists — so it is necessary. But once kidney disease is established, the direct kidney protection of the pillar drugs exceeds what glucose-lowering alone provides, so glycaemic control is one component of a regimen, not its centrepiece. The HbA1c target is individualised, around 7% for many, but relaxed in advanced CKD, the elderly, and those at risk of hypoglycaemia — because tight control in these groups buys little and risks much. The framing is important: in established DKD, do not rely on glucose-lowering as the renoprotective strategy; the pillars are that strategy, and glycaemic control supports it.
Choosing glycaemic agents in CKD
The glycaemic agents themselves must be chosen for the kidney. Metformin remains first-line and has cardiovascular benefit, but it is reduced as eGFR falls and held below 30 because of the risk of lactic acidosis. The two agent classes that double as pillars — SGLT2 inhibitors and GLP-1 receptor agonists — are preferred where possible precisely because they lower glucose and protect the kidney and heart at once. DPP-4 inhibitors are safe in CKD with dose adjustment and weight-neutral, but offer no major kidney benefit. Sulfonylureas accumulate in CKD and cause hypoglycaemia, so they are used cautiously or avoided. Insulin requirements fall as GFR declines, because the kidney clears insulin, so doses are reduced to avoid hypoglycaemia. And hypoglycaemia is the overarching hazard: it is more frequent and more dangerous in CKD, so the whole glycaemic regimen is built to avoid it, favouring the agents that do not cause it and dosing the others down.
Where the evidence is firm, and where it refines
Few areas of nephrology rest on firmer ground. Each pillar in diabetic kidney disease is supported by dedicated randomised trials — RAAS blockade, the SGLT2-inhibitor studies, the finerenone trials, and the GLP-1 kidney trial — making the four-pillar regimen grade-A practice. The refinements are at the edges: the optimal sequence and combination of the pillars is still being defined, the management and even the recognition of non-albuminuric DKD is evolving, and the individualisation of glycaemic targets in advanced CKD rests more on avoiding harm than on outcome trials. The honest summary is that diabetic kidney disease is the best-treated CKD there is — a disease where mechanism, evidence, and a stacked regimen align — and the task is to deploy that regimen fully, screen for the non-diabetic mimics, and individualise the glycaemic component to avoid hypoglycaemia.
