Diabetic ketoacidosis, or DKA, is one of the most dramatic emergencies in medicine. When the body cannot use glucose for fuel because insulin is absent or ineffective, it begins breaking down fat at an uncontrolled rate, flooding the bloodstream with acidic ketone bodies, driving blood sugar to dangerous heights, and shifting the internal chemistry of the body toward a state that can quickly become life-threatening. But DKA carries a second, subtler danger that emergency physicians and intensivists wrestle with every day: it masquerades as infection. The metabolic storm of DKA produces many of the same signs that clinicians use to identify bacterial infection, including fever-like physiology, elevated white blood cell counts, tachycardia, and widespread inflammatory activation. Distinguishing a patient whose DKA was triggered by a urinary tract infection or pneumonia from a patient whose crisis is purely metabolic is therefore a genuine diagnostic challenge, and one with real consequences, because unnecessary antibiotics carry their own risks while missed infections can be fatal.
A new study from researchers at Eskisehir Osmangazi University in Turkey offers a potential way to cut through this diagnostic fog. Writing in BMC Endocrine Disorders, Onur Gökçe and colleagues report that a simple ratio of two routinely measured blood values, procalcitonin and lactate, may help clinicians identify bacterial infection in patients admitted with DKA. The procalcitonin-to-lactate ratio, abbreviated PLR, showed meaningful diagnostic power in their retrospective cohort of 158 adults, outperforming the more widely used neutrophil-to-lymphocyte ratio, or NLR, in adjusted analyses. The findings, while preliminary in nature, add to a growing effort to refine biomarker-based decision-making in metabolic emergencies where traditional markers of infection lose their usual reliability.
The scientific logic behind the study rests on the distinct biology of its two components. Procalcitonin is the prohormone of calcitonin, a peptide normally produced by thyroid C cells, but during systemic bacterial infections it is produced in large quantities by extra-thyroid tissues under the influence of pro-inflammatory cytokines, particularly interleukin-1 beta, tumor necrosis factor-alpha, and interleukin-6. This makes procalcitonin one of the most infection-specific markers available in routine practice, rising sharply within hours of a bacterial insult and falling again when the infection is controlled. Lactate, by contrast, is a product of anaerobic metabolism and a marker of tissue hypoperfusion and physiological stress. It rises in DKA itself, driven by dehydration, reduced tissue perfusion, and shifts in the cellular redox state, which means lactate alone says little about infection. The insight behind the PLR is that dividing procalcitonin by lactate effectively adjusts the infection signal for the severity of the metabolic derangement. In theory, a patient with mild infection but severe DKA, or severe infection with only modest metabolic disturbance, could be characterized more accurately by the ratio than by either number alone.
The Turkish team tested this idea in 158 adult patients admitted with DKA, performing a retrospective analysis of laboratory data collected at presentation. The study was approved by the Institutional Ethics Committee of Eskisehir Osmangazi University and conducted in accordance with the Declaration of Helsinki. Patients were categorized according to whether a bacterial infection was ultimately identified, and the diagnostic performance of PLR and NLR was assessed using receiver operating characteristic, or ROC, analysis, a standard statistical technique that quantifies how well a test discriminates between two groups across all possible cutoff values. The area under the ROC curve, or AUC, provides a single number summarizing this discrimination: an AUC of 0.5 means the test performs no better than chance, while an AUC of 1.0 means perfect classification.
The results revealed several important patterns. Patients with bacterial infection were significantly older than those without, averaging about 54 years compared with roughly 44 years, a difference that was highly statistically significant. This age gap matters, because older patients tend to have more comorbidities, blunted physiological reserves, and a higher baseline risk of both infection and diagnostic ambiguity. Procalcitonin, lactate, PLR, and NLR were all significantly elevated in the infected group, providing the first signal that these markers carried real information about infection status even amid the inflammatory noise of DKA.
When the researchers formally quantified diagnostic performance, PLR achieved an AUC of 0.791 with a 95 percent confidence interval of 0.722 to 0.861, and NLR achieved an AUC of 0.730 with a 95 percent confidence interval of 0.651 to 0.808. Both values fall into a range generally considered to indicate moderate diagnostic utility. Interestingly, procalcitonin alone produced the highest observed AUC in the study, at 0.812 with a confidence interval of 0.747 to 0.877, a finding the authors themselves highlight with appropriate caution. The cruder but free NLR, calculated simply by dividing the absolute neutrophil count by the absolute lymphocyte count from a standard complete blood count, captured the stereotypical stress response of acute illness: neutrophilia driven by cortisol and catecholamines combined with lymphocyte redistribution and depletion.
The most clinically significant result, however, emerged from multivariable analysis. When the investigators adjusted for age, renal function, and C-reactive protein, PLR remained independently associated with bacterial infection, with an odds ratio of 4.337 and a 95 percent confidence interval of 1.433 to 13.127, reaching statistical significance at p equals 0.009. In other words, higher PLR values multiplied the odds of infection by more than fourfold, even after accounting for the classic confounders. NLR, by contrast, lost its independent association once these covariates were included, with an odds ratio of 1.020 that was not statistically significant. This contrast suggests that the mechanistic specificity of procalcitonin, normalized by the stress-related component of lactate, carries diagnostic information that the generic leukocyte stress response cannot provide in a DKA population.
Renal function deserves particular attention in this context, and the authors were right to adjust for it. Procalcitonin is cleared in part by the kidneys, and patients with diabetic kidney disease, which is common in the DKA population, may show elevated procalcitonin levels even in the absence of infection. Lactate clearance is also impaired in renal failure, and advanced diabetes often brings chronic low-grade inflammation that elevates baseline NLR. By including renal function in the adjusted model, the study strengthened the case that PLR is genuinely tracking infection rather than simply reflecting organ dysfunction. C-reactive protein adjustment was equally important, since CRP is itself an acute-phase marker that rises nonspecifically in severe metabolic stress.
The clinical implications, if confirmed, could be practical and immediate. Both procalcitonin and lactate are already measured routinely in patients presenting with severe DKA, particularly those requiring intensive care, where lactate is a standard component of the metabolic workup and procalcitonin is frequently ordered when infection is suspected. Calculating the ratio between them requires no additional blood draws, no laboratory infrastructure, and no cost, which makes it an attractive candidate for adoption in emergency departments in resource-limited settings where advanced diagnostics such as procalcitonin-guided algorithms, cultures with rapid turnaround, or molecular panels are unavailable. A single number that helps triage which DKA patients truly need empiric antibiotics could reduce both the under-treatment of occult sepsis and the overuse of antimicrobials, a growing concern in the era of rising antibiotic resistance.
Still, the authors are measured in their conclusions, and their caution is well founded. The study is retrospective, which means the analysis depended on the quality of documentation and the clinical judgment used to classify infections in real time. The cohort of 158 patients is modest, and subgroup analyses across different severities of DKA, mild, moderate, and severe, are constrained by sample size in each stratum. The acknowledged statistical reality that procalcitonin alone achieved the highest raw AUC raises the critical question of whether the ratio adds incremental diagnostic value beyond the unadjusted hormone level, particularly in specific patient subgroups such as those with impaired renal function. The authors explicitly call for prospective studies to answer this question, and such studies would ideally incorporate standardized infection adjudication, serial biomarker measurements, and outcomes analysis to determine whether PLR-guided decisions actually change antibiotic use, length of stay, or mortality.
What makes this study noteworthy in the broader landscape of diagnostic research is its reflection of a wider trend: the rediscovery of ratio-based and composite biomarkers derived from routine laboratory data. The NLR has been studied across hundreds of conditions precisely because it is free and universally available, and the PLR extends this logic by combining two mechanistically distinct markers, one tracking the inflammatory specificity of bacterial infection and the other tracking the metabolic stress that confounds it. In DKA, where the baseline inflammatory state is so profoundly perturbed that conventional cutoffs for white blood cell count and CRP lose their usual meaning, this kind of normalization strategy may be exactly what clinical decision-making needs. The work by Gökçe and colleagues does not settle the question of infection detection in DKA, but it adds a plausible, cheap, and readily testable tool to the clinician’s diagnostic toolkit, and it charts a clear path for the prospective trials that will determine whether the procalcitonin-to-lactate ratio earns a permanent place at the bedside.
Subject of Research: Diagnostic performance of the procalcitonin-to-lactate ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) in predicting bacterial infection in patients with diabetic ketoacidosis
Subject of Research: Medicine
Article Title: Diagnostic performance of procalcitonin/lactate ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) in predicting bacterial infection across different severities of diabetic ketoacidosis
Article References: Gökçe, O., Öztürk Gökçe, B., Yorulmaz, G., Akalın, A., Kebapçı, M. N., & Kalkan, A. T. (2026). Diagnostic performance of procalcitonin/lactate ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) in predicting bacterial infection across different severities of diabetic ketoacidosis. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02480-y
Image Credits: AI Generated
DOI: 10.1186/s12902-026-02480-y
Keywords: diabetic ketoacidosis, procalcitonin-to-lactate ratio, PLR, neutrophil-to-lymphocyte ratio, NLR, bacterial infection, biomarkers, diagnostic markers, ROC analysis, procalcitonin, lactate, endocrine emergencies
Cite Scienmag News
APA MLA Chicago
Kristina Jarvis. (September 7, 2026). Procalcitonin/lactate and neutrophil/lymphocyte ratios predict infection in diabetic ketoacidosis. Scienmag. https://scienmag.com/procalcitonin-lactate-and-neutrophil-lymphocyte-ratios-predict-infection-in-diabetic-ketoacidosis/
Kristina Jarvis. “Procalcitonin/lactate and neutrophil/lymphocyte ratios predict infection in diabetic ketoacidosis.” Scienmag, 7 September 2026, https://scienmag.com/procalcitonin-lactate-and-neutrophil-lymphocyte-ratios-predict-infection-in-diabetic-ketoacidosis/. Accessed 7 September 2026.
Kristina Jarvis. “Procalcitonin/lactate and neutrophil/lymphocyte ratios predict infection in diabetic ketoacidosis.” Scienmag. September 7, 2026. https://scienmag.com/procalcitonin-lactate-and-neutrophil-lymphocyte-ratios-predict-infection-in-diabetic-ketoacidosis/
Copy citation Download RIS
Tags: biomarkers for differentiating infection in DKA patientsbiomarkers for infection prediction in diabetic emergencieschallenges in diagnosing infection during diabetic ketoacidosischallenges in diagnosing infection in diabetic ketoacidosisclinical significance of procalcitonin and lactate levels inclinical utility of infection biomarkers in diabeticdistinguishing infection from metabolic crisis in diabetic ketoacidosisdistinguishing metabolic versus infectious causes in DKAimportance of accurate infection detection in DKAimportance of infection markers in diabetic emergenciesneutrophil to lymphocyte ratio in diabetic ketoacidosisneutrophil to lymphocyte ratio in DKA diagnosispredictive value of lactate levels in diabetic emergenciesProcalcitonin and lactate as infection markers in diabetic ketoacidosisProcalcitonin and lactate in infection diagnosisrole of inflammatory markers in DKA managementrole of inflammatory ratios in DKA managementuse of blood test ratios to differentiate infection from DKA symptomsuse of procalcitonin and NLR in emergency medicine
