Highlights
Adults with type 1 diabetes showed a larger postprandial rise in hepatic deuterium-labeled glucose than matched healthy control individuals, despite similar systemic insulin exposure.
Healthy control individuals demonstrated clear hepatic glycogen accumulation after oral glucose, whereas the type 1 diabetes group showed no significant glycogen rise by 150 minutes.
Stable-isotope minimal-model analysis indicated impaired suppression of endogenous glucose production in type 1 diabetes, while insulin-dependent peripheral glucose disposal was broadly similar between groups.
Within the type 1 diabetes cohort, imaging identified two metabolic subphenotypes with distinct hepatic glucose and glycogen trajectories despite no obvious clinical differences, underscoring the potential for precision phenotyping.
Background
Postprandial glucose homeostasis depends on tightly coordinated interactions among the gut, pancreas, liver, and peripheral tissues. In healthy physiology, insulin is secreted into the portal circulation, exposing the liver to a higher insulin concentration than the systemic circulation. This porto-systemic gradient is central to restraining hepatic glucose production and promoting hepatic glucose uptake and glycogen synthesis after meals.
Type 1 diabetes disrupts this physiology in a fundamental way. Patients rely on exogenous insulin administered subcutaneously, which reaches the systemic circulation before the portal vein. Even when overall insulin exposure appears adequate, the liver may remain relatively underinsulinized compared with normal physiology. This can impair suppression of endogenous glucose production, alter hepatic glucose uptake, and reduce glycogen storage. These abnormalities are clinically relevant because postprandial hyperglycemia contributes materially to glycemic variability and long-term diabetes complications.
Yet liver-specific dynamics in type 1 diabetes are difficult to characterize non-invasively. Standard glucose monitoring, plasma hormone measurements, and even sophisticated tracer methods primarily capture systemic behavior. They do not directly visualize how the liver handles ingested glucose in real time. Magnetic resonance approaches offer that possibility. In particular, deuterium metabolic imaging, combined with carbon-13 magnetic resonance spectroscopy, may permit simultaneous tracking of hepatic labeled glucose and glycogen formation after an oral glucose load.
The present case-control study by Brunasso and colleagues addresses this gap by combining ultra-high-field liver imaging with stable-isotope modeling to assess postprandial hepatic and systemic glucose homeostasis in adults with type 1 diabetes versus healthy control individuals.
Study Design
Design and participants
This was a cross-sectional case-control study including 20 adults: ten with type 1 diabetes and ten healthy control individuals. The groups were matched for age, BMI, and gender distribution. The study focused on adults with well-managed type 1 diabetes, although baseline fasting hyperglycemia remained present in the diabetes group.
Intervention and metabolic protocol
After an overnight fast, participants ingested 60 g of [6,6′-2H2]-glucose, referred to as deuterium-labeled glucose or D-Glc. Participants with type 1 diabetes received subcutaneous insulin according to their individualized carbohydrate-to-insulin ratio, reflecting real-world prandial insulin replacement rather than an experimentally fixed insulin infusion.
Imaging methods
The investigators performed interleaved deuterium metabolic imaging and 13C-magnetic resonance spectroscopy at 7 T from before glucose ingestion through 150 minutes after ingestion. These techniques were used to quantify hepatic D-Glc and hepatic glycogen concentrations over time. This is the major technical innovation of the study: a non-invasive dynamic readout of liver substrate handling after an oral glucose challenge.
Systemic measurements and modeling
Serial blood sampling measured plasma glucose, insulin, and glucagon. Whole-body glucose-insulin dynamics were estimated using the single-tracer oral minimal model, adapted to account for non-steady-state insulin exposure. This allowed the investigators to examine endogenous glucose production and insulin-dependent glucose disposal in parallel with liver imaging findings.
Endpoints
The principal endpoints were the temporal profiles of hepatic D-Glc and hepatic glycogen after oral glucose administration. Secondary endpoints included plasma glucose and hormone responses, suppression of endogenous glucose production, insulin-dependent glucose disposal, and exploratory clustering of metabolic phenotypes within the type 1 diabetes group.
Key Findings
Baseline metabolic state
At baseline, fasting plasma glucose was significantly higher in the type 1 diabetes group than in the control group: 10.7 ± 2.3 mmol/l versus 5.2 ± 0.4 mmol/l, with p<0.001. By contrast, preprandial hepatic glycogen levels did not differ significantly. This is noteworthy because it suggests that fasting glycogen content alone may not capture the dynamic defect in postprandial hepatic handling.
Greater hepatic labeled glucose accumulation in type 1 diabetes
Following ingestion of D-Glc, hepatic deuterium-labeled glucose rose more strongly in participants with type 1 diabetes than in control individuals. Peak hepatic D-Glc values were 4.7 ± 2.0 mmol/l in type 1 diabetes and 3.0 ± 0.8 mmol/l in control individuals, with p=0.02. In practical terms, the liver in type 1 diabetes appeared to experience a larger postprandial labeled glucose burden.
This finding is physiologically plausible. If portal insulinization is inadequate, the liver may fail to transition efficiently from glucose output toward glucose storage. Labeled glucose may therefore accumulate within hepatic tissue or transit differently through intrahepatic pathways rather than being promptly incorporated into glycogen.
Blunted hepatic glycogen synthesis after oral glucose
The most clinically intuitive result was the contrast in glycogen storage. Healthy control individuals showed a clear postprandial rise in hepatic glycogen, with an incremental area under the curve from 0 to 180 minutes of 2.4 mol/l × min. In the type 1 diabetes group, glycogen levels did not significantly rise by 150 minutes. The implication is that even when exogenous insulin is given and systemic insulin exposure appears comparable, hepatic conversion of incoming glucose into glycogen remains impaired.
This aligns with long-standing physiological observations that the liver is not simply another insulin-responsive tissue. It depends critically on portal insulin delivery and on integrated signaling with glucose and glucagon to regulate glycogenesis and glycogenolysis. Subcutaneous insulin replacement can normalize many systemic features while still leaving the liver metabolically out of phase.
Reduced suppression of endogenous glucose production
Minimal-model analysis showed that individuals with type 1 diabetes had significantly reduced suppression of endogenous glucose production compared with control individuals, with p=0.001. This is one of the most important translational findings in the study. It indicates that the liver in type 1 diabetes continues to contribute glucose to the circulation after oral glucose intake more than it should under physiological conditions.
From a pathophysiological standpoint, this is exactly the abnormality expected from relative portal hypoinsulinization. The liver does not fully switch off its own glucose output after the meal, worsening postprandial hyperglycemia.
Peripheral insulin-dependent glucose disposal was similar
Despite the hepatic abnormalities, insulin-dependent glucose disposal did not significantly differ between groups. This is also highly informative. It suggests that postprandial dysregulation in these adults with type 1 diabetes was driven more by defective hepatic regulation than by major impairment in peripheral insulin action, at least under the study conditions.
Clinically, that distinction matters. It supports the view that postprandial control in type 1 diabetes is not solely a matter of matching insulin dose to carbohydrate amount. Route and timing of insulin delivery, especially with respect to the liver, may be equally important.
Glucagon did not explain the group difference
The investigators found no significant differences in postprandial glucagon concentrations between groups. Therefore, the observed defect in endogenous glucose production suppression was not readily attributable to a glucagon excess signal in the type 1 diabetes group. This strengthens the argument that abnormal insulin distribution, rather than overt glucagon divergence, is a dominant factor in this setting.
Hidden heterogeneity within type 1 diabetes
A particularly interesting part of the study was the exploratory hierarchical clustering analysis. Within the type 1 diabetes cohort, two subgroups emerged.
Subgroup 1 had a steeper increase in both hepatic and systemic D-Glc profiles and showed net glycogen accumulation. Subgroup 2 displayed a divergent D-Glc trajectory and net glycogen depletion rather than storage. The incremental area under the curve for glycogen from 0 to 180 minutes was 2.5 mol/l × min in subgroup 1 and -3.0 mol/l × min in subgroup 2, with p=0.04.
Importantly, these subgroups had no overt clinical differences apparent from routine descriptors. That observation raises the possibility that standard clinical phenotyping in type 1 diabetes misses biologically meaningful differences in hepatic metabolism. Such differences could help explain why patients with apparently similar insulin regimens, HbA1c values, or demographics experience very different postprandial glycemic responses.
Clinical Interpretation
This study adds a valuable liver-centered perspective to type 1 diabetes management. Contemporary care often focuses on systemic metrics such as HbA1c, time in range, continuous glucose monitor profiles, and total daily insulin dose. Those measures are important, but they do not directly reveal whether the liver is appropriately suppressing glucose production and storing meal-derived glucose.
The data suggest that in type 1 diabetes, systemic insulin exposure can look adequate while hepatic regulation remains abnormal. That distinction may partly explain why even advanced insulin pumps and hybrid closed-loop systems do not fully normalize postprandial physiology. Subcutaneous insulin cannot readily reproduce the normal portal insulin peak delivered by endogenous pancreatic secretion.
The findings also have implications for therapeutic development. Approaches that better target the liver, whether through alternative insulin delivery routes, hepatopreferential insulins, adjunctive therapies, or more physiologic meal-time algorithms, may deserve renewed attention. Likewise, imaging-derived metabolic phenotypes could one day help identify which patients are most likely to benefit from specific strategies.
Methodological Strengths
The study has several notable strengths. First, it uses a sophisticated but clinically relevant integrated design: direct liver imaging plus whole-body tracer modeling. This allows local hepatic observations to be interpreted in the context of systemic metabolism.
Second, the oral glucose challenge with individualized subcutaneous insulin mirrors a real-world postprandial scenario more closely than protocols using artificial intravenous clamps alone. Third, the use of 7 T DMI and 13C-MRS provides rare in vivo insight into hepatic glucose and glycogen kinetics without invasive sampling.
Finally, the exploratory within-cohort clustering, although preliminary, is conceptually important. It moves beyond average group comparisons and points toward heterogeneity that may matter clinically.
Limitations
Several limitations temper interpretation. The sample size was very small, with only ten participants per group. This is understandable for a demanding ultra-high-field imaging study, but it limits precision, increases vulnerability to chance findings, and makes subgroup analysis exploratory rather than definitive.
The study was cross-sectional, so it cannot establish causal relationships or determine whether the identified metabolic phenotypes are stable over time. Reproducibility of the imaging and clustering findings in independent cohorts will be essential.
Baseline fasting hyperglycemia differed substantially between groups. Although this reflects the lived biology of type 1 diabetes, it may itself influence hepatic glucose handling and glycogen turnover. Distinguishing the effects of chronic disease physiology from the immediate effect of higher starting glucose is challenging.
The imaging window extended to 150 minutes, while some results are summarized as incremental area under the curve through 180 minutes. The main conclusion remains clear, but readers should note the timing framework carefully as reported by the investigators.
Generalizability is also limited. These were adults with established, well-managed type 1 diabetes studied under specialized research conditions at 7 T. The findings may not translate directly to children, older adults, people with brittle diabetes, or those with substantial insulin resistance or hepatic steatosis.
Finally, the study was not designed to test therapeutic interventions. It shows abnormal physiology, but it does not prove which treatment changes would correct it.
Expert Commentary
The biological message of this work is consistent with a large body of prior physiology: the liver is central to postprandial glucose control, and portal insulinization matters. What this study contributes is a direct, non-invasive visualization of that principle in people with type 1 diabetes under meal-like conditions.
Current diabetes guidelines emphasize individualized glycemic targets, time in range, and use of diabetes technologies. This study complements that framework by suggesting that some residual postprandial dysregulation may be specifically hepatic and not obvious from standard clinical metrics alone. If validated, liver-focused phenotyping could eventually refine patient stratification in clinical trials and, perhaps later, routine specialist care.
It is also worth noting that the apparent preservation of insulin-dependent glucose disposal shifts the mechanistic focus away from classic peripheral insulin resistance and toward hepatic control failure. In selected patients, this may support therapeutic thinking that prioritizes meal timing, insulin delivery kinetics, and hepatic-targeted adjuncts over simple dose escalation.
Funding and Trial Registration
The abstract and citation provided do not report funding details. No ClinicalTrials.gov registration number is supplied in the source material. Readers should consult the full published article in Diabetologia for complete disclosures, funding statements, and any registry information.
Conclusion
Brunasso and colleagues provide compelling early evidence that advanced metabolic imaging can expose a clinically important disconnect in type 1 diabetes: systemic insulin replacement may appear adequate while hepatic postprandial physiology remains distinctly abnormal. Compared with healthy control individuals, adults with type 1 diabetes had greater hepatic labeled glucose accumulation, absent or blunted glycogen repletion, and impaired suppression of endogenous glucose production. Just as importantly, the study uncovered metabolic heterogeneity within type 1 diabetes that was not evident from routine clinical characteristics.
For clinicians, the study reinforces that postprandial control in type 1 diabetes is fundamentally a liver problem as well as a dosing problem. For researchers, it introduces a promising non-invasive platform for metabolic phenotyping. The next steps are clear: larger validation studies, longitudinal assessment of phenotype stability, and interventional trials testing whether liver-directed strategies can normalize the abnormalities identified here.
If those steps succeed, deuterium metabolic imaging and related methods may help move type 1 diabetes care closer to true precision medicine.
Citation
Brunasso A, Lange NF, Poli S, Schiavon M, Herzig D, Dalla Man C, Kreis R, Bally L. Novel deuterium metabolic imaging technique reveals distinct patterns of postprandial hepatic glucose homeostasis in individuals with type 1 diabetes and healthy control individuals: a case-control study. Diabetologia. 2026-02-13;69(5):1301-1316. PMID: 41686194. Available at: https://pubmed.ncbi.nlm.nih.gov/41686194/
Selected References
American Diabetes Association Professional Practice Committee. 2. Diagnosis and classification of diabetes: Standards of Care in Diabetes. Diabetes Care. Updated annually. These standards provide current clinical context for type 1 diabetes classification and management.
Petersen KF, Shulman GI. Etiology of insulin resistance. Am J Med. 2006;119(5 Suppl 1):S10-S16. This review helps frame hepatic versus peripheral insulin action, although it is not specific to type 1 diabetes.
Shulman GI, Rothman DL, Jue T, Stein P, DeFronzo RA, Shulman RG. Quantitation of muscle glycogen synthesis in normal subjects and subjects with non-insulin-dependent diabetes by 13C nuclear magnetic resonance spectroscopy. N Engl J Med. 1990;322(4):223-228. A foundational paper demonstrating the translational value of magnetic resonance spectroscopy in human glucose metabolism.
Edgerton DS, Cherrington AD. Glucose homeostasis and the critical role of the liver in normal physiology and in diabetes. Diabetes. 2011;60(3):679-681. A concise perspective on hepatic control of glucose metabolism.
