Proteinuria, not GFR alone, may explain the renal benefit of SGLT2 inhibitors
The kidney-protective effect of SGLT2 inhibitors appears to track most closely with proteinuria. In a BMJ rapid response to Turchin and colleagues’ target trial emulation, Saad Saleh Alshohaib argues that the class’s hemodynamic effect matters because it lowers proteinuria and thereby reduces tubular injury, rather than because it simply lowers glomerular filtration rate. The practical implication is blunt: a fall in estimated GFR is not inherently useful unless it translates into less protein delivered to the tubule.
Proteinuria is not a passive marker of damage. It is itself an active driver of progression.
Tubular stress sits at the center of the argument
The letter places proximal tubular injury at the center of diabetic kidney disease progression. Excess filtered protein places stress on tubular cells, and the biologic consequence is oxidative injury. Albumin activates NAD(P)H oxidase activity through Rac1 in proximal tubule cells, a mechanism that helps explain why proteinuria does more than accompany renal decline; it participates in it. In that framework, therapies that reduce proteinuria can interrupt a damaging cycle of tubular stress, inflammation, and structural loss.
The argument is especially relevant for SGLT2 inhibitors because their early hemodynamic effect is well known. The initial dip in estimated GFR has been documented across studies of canagliflozin, dapagliflozin, and empagliflozin, and this response has been interpreted as part of the class effect rather than a reason to stop treatment in most patients. Here, the nephrologist’s position is that the dip is clinically meaningful only when it reflects reduced intraglomerular pressure and less albumin leakage into the urine. If the filtered load falls without a corresponding reduction in proteinuria, the benefit is far less clear.
The acute eGFR dip is not the same as kidney harm
A transient reduction in estimated GFR can look alarming in isolation. In the setting of renoprotective therapy, however, the meaning of that change depends on its severity and on the biologic direction of travel over time. The rapid response argues that the haemodynamic effect of SGLT2 inhibition is not harmful unless it is severe, and that the patients most likely to benefit are those with proteinuria. In other words, the same physiological signal can be either acceptable or problematic depending on whether it serves to relieve tubular protein burden.
This interpretation aligns with prior analyses cited in the letter. Data from the CANVAS Program linked canagliflozin’s renal and cardiovascular benefits to albuminuria level. Post hoc work from CREDENCE showed that early changes in albuminuria predicted kidney and cardiovascular outcomes, while prespecified DAPA-CKD analyses showed that dapagliflozin reduced urinary albumin excretion in chronic kidney disease with and without type 2 diabetes. Together, those studies support a mechanistic chain in which albuminuria reduction is not a side effect but a marker of the pathway most likely to matter clinically.
Albuminuria may be the best clinical compass for renoprotection
For bedside practice, the message is simple and potentially consequential. Proteinuria may be the most useful marker for identifying patients who will extract the greatest kidney benefit from SGLT2 inhibitors, and perhaps from other renoprotective agents with similar hemodynamic profiles. If the therapeutic goal is to limit protein-induced tubular stress, then albuminuria becomes more than a risk marker; it becomes a treatment target and a readout of biologic effect.
That reasoning also reframes how clinicians should interpret the first weeks after starting therapy. An initial fall in eGFR does not automatically mean the kidney is worse off. It may reflect a hemodynamic adjustment that reduces glomerular pressure and lowers protein filtration, which is precisely the pathway that could protect the tubulointerstitium over time. The challenge is not to ignore the dip, but to interpret it in the context of proteinuria, severity, and trajectory.
Why the mechanism matters for the target trial emulation
The BMJ target trial emulation by Turchin and colleagues examined renal outcomes in people with type 2 diabetes with and without albuminuria who were treated with SGLT-2 inhibitors and GLP-1 receptor agonists. The rapid response does not dispute the study design; it interprets its findings through a mechanistic lens. By emphasizing proteinuria, the letter argues that apparent renal protection is most biologically plausible when the intervention reduces the protein load reaching tubular cells.
That view is consistent with the broader nephrology literature cited in the response. Albuminuria has repeatedly emerged as a predictor of benefit across SGLT2 inhibitor studies, while acute eGFR changes appear acceptable when they accompany long-term renal protection rather than abrupt hemodynamic collapse. The response therefore presents proteinuria as the common thread connecting hemodynamic shifts, tubular oxidative stress, and downstream renal preservation.
In clinical terms, the message is pointed. SGLT2 inhibitors should not be judged solely by their immediate effect on filtration rate. Their benefit appears to follow the path of albuminuria reduction, and their early hemodynamic effect becomes valuable only when it lowers the burden of protein reaching the tubular epithelium. For patients without meaningful proteinuria, the mechanistic case is weaker, and the expected kidney benefit may be smaller.
A mechanistic reading, not a new efficacy trial
This rapid response is an interpretation rather than a trial report, and it stays within that lane. It does not add new patient data or outcomes. Its value lies in connecting known physiology to an ongoing therapeutic question: why do SGLT2 inhibitors protect the kidney, and in whom is that protection most likely to appear? The answer offered here is proteinuria, with hemodynamic change serving as the route by which filtered protein load is reduced.
The remaining uncertainty is not whether proteinuria matters, but how much it should govern individual treatment decisions across the spectrum of diabetic kidney disease. That question will continue to depend on prospective renal outcomes, subgroup behavior by albuminuria level, and how clinicians interpret the initial eGFR dip when it appears at the bedside.