Abstract
Background: The C-peptide-to-insulin ratio (C/I) is associated with hepatic insulin clearance and insulin resistance. We established a novel C-peptide-to-insulin ratio index (CPIRI) and explored its application in the heterogeneity of non-autoimmune diabetes.
Methods: A total of 865 adults diagnosed with new-onset diabetes mellitus (DM) within 1 year and 54 healthy controls (HC) were recruited. Our CPIRI model was established with fasting C/I as the independent variable and homeostasis model assessment of insulin resistance (HOMA-IR) as the dependent variable. Patients with negative-autoantibodies against pancreatic islets were subclassified according to CPIRI, age of onset, hemoglobin A1c (HbA1c), body mass index (BMI) and homeostatic model assessment of β-cell function (HOMA-β) via cluster analysis.
Results: Fasting C/I and HOMA-IR in both diabetic and healthy subjects were hyperbolically correlated, and log(C/I) and log(HOMA-IR) were linearly and negatively correlated; the correlation coefficient was −0.83 (P = 0.000) and −0.76 (P = 0.000), respectively. Finally, we obtained the equations $CPIRI(DM) = 670/(C/I)^{2.24} + 0.25$ and $CPIRI(HC) = 670/(C/I)^{2.24} − 1$ (F = 1904.39, P = 0.000). Pancreatic islet autoantibody-negative patients were further subclassified into three subtypes; subtype 1 (13.8%) had severe insulin resistance, subtype 2 (42.1%) suffered elderly-onset DM with slight symptoms, and subtype 3 (44.1%) suffered severe dysfunction of pancreatic islets. Each subtype exhibited different clinical characteristics and complications.
Conclusions: CPIRI can be used to effectively evaluate insulin resistance. Adult-onset non-autoimmune diabetes can be subclassified into 3 subtypes according to CPIRI along with other symptom-based characteristics of diabetic progression.
Introduction
Diabetes mellitus (DM) is a notoriously heterogeneous condition. While traditional classifications like Type 1 and Type 2 provide a broad framework, they often fail to capture the metabolic nuances of individual disease progression. This lack of granularity frequently leads to a "one-size-fits-all" treatment approach that may not be optimal for all patients.
The current gold standard for measuring insulin resistance in epidemiological studies is the Homeostatic Model Assessment for Insulin Resistance (HOMA-IR). However, HOMA-IR primarily evaluates the relationship between glucose and insulin, often overlooking hepatic insulin clearance—the efficiency with which the liver removes insulin from the blood. As the liver is the primary gatekeeper of systemic insulin levels, ignoring this mechanism leaves a significant gap in our understanding of the metabolic landscape.
Methods
To address this gap, researchers focused on the C-peptide-to-insulin (C/I) ratio. Insulin and C-peptide are secreted by the pancreas in equimolar amounts. However, while the liver clears a substantial portion of insulin, it has minimal effect on C-peptide, which is primarily cleared by the kidneys. Consequently, the C/I ratio serves as a proxy for hepatic insulin clearance.
The study utilized a clinical dataset of 865 adults with new-onset diabetes and 54 healthy controls. The researchers developed the CPIRI model by analyzing the hyperbolic relationship between the C/I ratio and HOMA-IR. Following the establishment of the index, a two-step cluster analysis was performed on 572 non-autoimmune diabetic patients using five key variables: CPIRI, age of onset, HbA1c, BMI, and HOMA-β.
Results
The analysis confirmed a strong negative linear correlation between log(C/I) and log(HOMA-IR) ($r = -0.83$ for diabetic subjects), validating the C/I ratio as a robust proxy for insulin resistance.
By incorporating CPIRI into the clustering model, the researchers identified three distinct subtypes of non-autoimmune diabetes, whereas traditional HOMA-IR-based models only identified two. These subtypes exhibited unique clinical profiles:
- Subtype 1 (13.8% - Severe Insulin-Resistant): Characterized by high BMI and elevated liver enzymes, suggesting significant hepatocyte injury.
- Subtype 2 (42.1% - Elderly-onset): Characterized by an older age of onset and mild symptoms, yet a higher prevalence of peripheral vascular plaques, indicating increased cardiovascular risk.
- Subtype 3 (44.1% - Severe Insulin-Deficient): Characterized by severe failure of pancreatic beta-cells and a higher prevalence of ketosis-onset DM.
Discussion and Conclusion
The development of CPIRI provides a higher-resolution map of diabetes by accounting for hepatic insulin clearance. This index effectively "weights" insulin resistance scores by the liver's performance, offering a more comprehensive metabolic assessment than HOMA-IR alone. The identification of three distinct subtypes suggests that adult-onset non-autoimmune diabetes can be more accurately subclassified, paving the way for precision medicine in blood sugar management.
While promising, the study notes limitations, including potential inaccuracies in patients receiving exogenous insulin and the need for validation in non-Chinese populations. Future research should integrate additional clinical indicators, such as genetic markers or lipidomics, to further refine these clusters. Ultimately, this work moves clinical practice closer to a future where a diabetes diagnosis is accompanied by a specific, personalized roadmap for treatment.
Summary updated