Ferritin rose with dengue severity, but the ratios did not beat routine labs
Dengue remains a high-volume clinical problem in India, where patients can deteriorate quickly from an apparently routine febrile illness to shock or severe hepatitis. In that setting, clinicians want a marker that is cheap, fast, and available early enough to change bedside decisions. This prospective biomarker study examined whether serial serum ferritin and two ferritin-derived ratios could help separate more dangerous dengue from less severe disease in hospitalized adults.
The answer was mixed in a way that matters clinically. Ferritin moved in a clear rise-peak-fall pattern, and patients with severe dengue carried higher values throughout the admission. Yet the composite indices built around ferritin did not outperform the familiar tests already embedded in them: platelet count for the Ferritin-to-Platelet Ratio and ALT for the Ferritin/ALT Ratio. In practical terms, the ratios reflected illness burden, but they did not add enough information to justify replacing simpler laboratory readouts.
Serial ferritin followed a predictable inflammatory course
The study enrolled 100 adults with WHO 2009-confirmed dengue at a tertiary hospital in Surat and measured ferritin on hospital Days 1, 3, and 5. Ferritin peaked on Day 3, when the mean level reached 955 ± 177 ng/mL, with a highly significant within-patient change across time. When the same values were sorted by day of illness, ferritin climbed from a median of 700 ng/mL in illness days 1 to 3 to 950 ng/mL in days 6 to 7 before flattening out.
That trajectory fits dengue biology. Ferritin is not just an iron-storage protein; it also behaves as a marker of macrophage activation and systemic inflammation. In dengue, immune activation and endothelial stress intensify as patients approach the critical phase, which helps explain why ferritin did not simply stay elevated from the first blood draw but rose further after admission. A one-sentence bedside takeaway follows directly from the kinetics: ferritin was dynamic, not static.
Severity separation was real, but modest for the admission ferritin signal
Severity classification divided the cohort into non-severe dengue and severe dengue, with 32 patients in the severe category. Admission ferritin helped distinguish these groups only to a moderate degree. The Ferritin-to-Platelet Ratio on Day 1 produced an area under the curve of 0.704 for severity discrimination, and severe cases had a five-fold higher median ratio than non-severe cases, 39.0 versus 7.5.
Against specific complications, the picture sharpened. For shock or dengue shock syndrome, Day 1 ferritin reached an AUC of 0.789, while the Ferritin-to-Platelet Ratio reached 0.801. Severe thrombocytopenia and shock clustered where clinicians would expect them to cluster: around platelet consumption, capillary leak, and systemic inflammatory burden. Ferritin also showed a negative association with platelet count, with a partial Spearman correlation of −0.392 after demographic adjustment. The biology is plausible enough; intense inflammation and marrow or peripheral platelet suppression often travel together in severe arboviral illness.
The composite ratios did not outperform the variables already inside them
The strongest signal in the paper came not from the ratios themselves, but from what happened when they were compared with their component tests. For shock or dengue shock syndrome, the Ferritin-to-Platelet Ratio did not beat platelet count alone. Platelet count had an AUC of 0.784, and the paired DeLong comparison gave p = 0.41. In other words, the ratio looked useful, but not more useful than the routine laboratory value already available on any standard complete blood count.
The same pattern appeared with the liver injury signal. The Ferritin/ALT Ratio delivered an almost perfect AUC of 0.995 for severe hepatitis, yet ALT alone reached 1.000, and the difference was not significant. Because ALT defines the outcome threshold of severe hepatitis here, a composite that includes ALT cannot logically be expected to surpass it by much. The ratio may be numerically elegant, but the diagnostic advantage disappears once the denominator already encodes the endpoint.
How the study design limited predictive claims
A critical methodological detail shaped the interpretation. Platelet count and transaminases were taken as the lowest and highest values recorded during the hospital stay, while outcomes were counted if they occurred at any time between admission and discharge. That means the ratios were constructed from values that were not strictly anchored to the admission moment. The result is a marker of overall disease severity during hospitalization, not a clean predictor of later deterioration from the first day in the ward.
Ferritin itself still had value as a severity-associated biomarker. Against shock or dengue shock syndrome, ferritin on Day 1 reached an AUC of 0.789, and it also discriminated severe thrombocytopenia with an AUC of 0.727. But the study’s own design explains why those numbers should be read cautiously. If the laboratory values are drawn from the entire admission rather than the moment of presentation, then any ratio built from them is partly retrospective by construction.
What this means for dengue care in resource-limited hospitals
For frontline dengue management, the message is straightforward. Ferritin rises with worsening disease and may help characterize the inflammatory burden in hospitalized adults, but the ferritin-derived ratios do not yet justify replacing platelet count or ALT in routine decision-making. A hospital that already has platelet counts and liver enzymes does not gain much by adding a more complicated arithmetic version of the same information.
This does not make ferritin irrelevant. It may still have a role as part of a broader severity picture, especially where clinicians want to understand inflammatory activation rather than only count platelets or transaminases. It may also help in future prognostic work if measured at presentation and paired with outcomes recorded from that same time point forward. What this study rules out is a stronger claim: these ratios are not ready to function as superior predictors of shock or severe hepatitis.
The next question is timing, not biology
The most important unanswered issue is not whether ferritin tracks dengue inflammation. It does. The open question is whether an admission-time ferritin strategy, measured before complications emerge, can predict who will worsen before the ward course itself reveals the answer. Prospective multicentre studies that align laboratory sampling with outcome timing will be needed to answer that question cleanly.
Until then, ferritin remains a biologically credible severity marker with limited incremental diagnostic value over standard tests. The ratios studied here are inexpensive and easy to calculate, but in this cohort they were more descriptive than predictive.