MASLD, formerly known under the umbrella of non-alcoholic fatty liver disease, has quietly become the most common chronic liver condition on the planet. It arises when fat accumulates inside liver cells in the context of metabolic problems such as obesity, insulin resistance, diabetes, and dyslipidaemia. In many people the fat deposition remains benign, but in a substantial subset it triggers inflammation, cell death, and the gradual activation of hepatic stellate cells, the resident cells responsible for producing scar tissue. That scarring, or fibrosis, is the single most important determinant of long-term outcomes: the more advanced the fibrosis, the higher the risk of cirrhosis, liver failure, and hepatocellular carcinoma. Identifying modifiable factors that accelerate or mark fibrosis progression is therefore one of the field’s most urgent priorities, and vitamin D has long been a candidate.

The biological rationale for suspecting vitamin D is genuinely intriguing. The vitamin D receptor is expressed on hepatic stellate cells, and experimental work has suggested that when activated vitamin D binds to this receptor, it can suppress the fibrogenic programs that turn quiescent stellate cells into collagen-producing myofibroblasts. Vitamin D also modulates immune responses and insulin signaling, both of which are central to the inflammatory milieu that drives MASLD progression. Deficiency is exceptionally common in people with metabolic disorders, partly because vitamin D is sequestered in adipose tissue, partly because sedentary indoor lifestyles reduce cutaneous synthesis, and partly because dietary intake is often inadequate. In Middle Eastern populations in particular, despite abundant sunshine, cultural and climatic factors conspire to produce some of the highest deficiency rates recorded anywhere in the world.

To interrogate the relationship rigorously, the research team turned to transient elastography, a non-invasive imaging technique marketed widely under the brand name FibroScan. The technology works by inducing gentle vibrations on the skin over the liver and measuring how fast shear waves propagate through the organ. Because stiff, scarred tissue transmits vibrations faster than healthy, fatty tissue, the resulting liver stiffness measurement provides a reliable surrogate for fibrosis stage, graded on a scale from F0, meaning no fibrosis, to F4, meaning cirrhosis. The same device estimates hepatic fat content through the controlled attenuation parameter, or CAP, which quantifies how strongly ultrasound waves are attenuated as they pass through the liver. Combining these two readouts allowed the investigators to assess both scarring and steatosis without subjecting patients to liver biopsy, the invasive gold standard that most large studies now seek to avoid.

The study population consisted of adults with MASLD who underwent vibration-controlled transient elastography and had a serum 25-hydroxyvitamin D measurement taken within ninety days, ensuring that the vitamin D status and the liver assessment reflected the same physiological moment. The cohort was relatively young, with a mean age of just under forty-two years, and nearly evenly split between sexes. The distribution of fibrosis was encouragingly skewed toward the mild end: roughly eighty-one percent of patients had no or mild fibrosis, about fifteen percent had significant fibrosis at stage F2, under four percent had advanced fibrosis at stage F3, and fewer than one percent had cirrhosis. Nevertheless, vitamin D deficiency was strikingly prevalent, affecting almost seventy percent of the entire cohort, a figure consistent with regional epidemiological data but still remarkable when viewed against the backdrop of a sun-drenched Gulf city.

When the researchers cross-tabulated vitamin D status against disease severity, the statistical signals were clear. Deficiency was significantly associated with fibrosis stage, with a p-value of 0.01, and independently associated with the severity of hepatic steatosis as quantified by CAP, also with a p-value of 0.01. Intriguingly, the associations with individual metabolic comorbidities were absent: body mass index, diabetes or prediabetes, hypertension, and dyslipidaemia showed no significant relationship with vitamin D deficiency in this cohort. What did stand out was metabolic syndrome as a composite diagnosis, which was strongly associated with deficiency at a p-value below 0.01. This pattern hints that vitamin D deficiency may track with overall metabolic burden rather than with any single component, a nuance that could matter for interpreting previous studies that focused on isolated risk factors.

The critical test came in the form of multivariable logistic regression, where vitamin D deficiency was evaluated as a predictor of clinically significant fibrosis while accounting for age, sex, and other covariates. The result was a resounding null: the adjusted odds ratio for vitamin D deficiency was 0.89, with a ninety-five percent confidence interval spanning 0.44 to 1.80 and a p-value of 0.754. In plain terms, once other factors were controlled for, knowing a patient’s vitamin D status added essentially no predictive value for identifying meaningful liver scarring. Instead, two far more conventional variables emerged as independent drivers. Each additional year of age increased the odds of significant fibrosis by about five percent, with an adjusted odds ratio of 1.05, and male sex carried a dramatically elevated risk, with an adjusted odds ratio of 6.23, meaning men in this cohort were more than six times as likely as women to have clinically significant fibrosis.

Why does this matter, and why should the finding still be considered news? The answer lies in the recalibration of expectations. Observational studies across many chronic diseases have linked low vitamin D to worse outcomes, fueling a wave of supplementation trials that have, almost uniformly, failed to deliver the benefits that the observational data promised. The present study adds liver fibrosis to the growing list of conditions where the crude association is real but likely non-causal. Vitamin D deficiency may function as a marker, a biological footprint of greater adiposity, reduced physical activity, more severe metabolic derangement, or simply older and sicker populations, rather than as an agent of harm in its own right. Treating it as such a marker could still be useful: a deficient patient with MASLD might warrant closer surveillance for fibrosis progression, even if correcting the deficiency itself is not expected to halt scarring.

The authors are careful and appropriately restrained in their conclusions. They explicitly state that causality cannot be established from this retrospective cross-sectional design, and they call for prospective studies to determine whether vitamin D plays any genuine role in disease progression. Such studies would need to follow patients over time, measure serial liver stiffness, and ideally randomize supplementation in deficient individuals to test whether repletion alters the fibrosis trajectory. Until then, the practical message for clinicians is nuanced: vitamin D deficiency is ubiquitous in MASLD patients, it correlates with disease severity on simple analysis, and it deserves correction for bone health and general wellbeing, but it should not be mistaken for a therapeutic lever against liver scarring. Meanwhile, the strongest signals in this data, age and male sex, serve as a reminder that the demographic profile of the patient in front of you may tell you more about their fibrosis risk than any single laboratory value.

For the millions of people living with MASLD worldwide, the study offers both reassurance and direction. The vast majority of patients in this cohort had no or only mild fibrosis, suggesting that early disease, when detected, remains the norm and the window for intervention is wide. The research also validates transient elastography as a practical, biopsy-free tool for stratifying risk in routine clinical settings, particularly in regions where deficiency and fatty liver overlap so extensively that disentangling their contributions demands precisely this kind of quantitative, mechanistically informed approach. The sunshine vitamin, it turns out, may simply be reflecting the light of underlying metabolic disease rather than driving the darkness of liver scarring, and knowing the difference is the first step toward treating the right target.

Subject of Research: Association between vitamin D deficiency and liver fibrosis in metabolic dysfunction-associated steatotic liver disease assessed by transient elastography

Article Title: Association between vitamin D deficiency and liver fibrosis assessed by transient elastography in patients with metabolic dysfunction-associated steatotic liver disease

Article References: Jose, J., Muhammad Soomar, S., Cristobal, X. F., & Khan, Z. (2026). Association between vitamin D deficiency and liver fibrosis assessed by transient elastography in patients with metabolic dysfunction-associated steatotic liver disease. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02527-0

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02527-0

Keywords: MASLD, vitamin D deficiency, liver fibrosis, transient elastography, FibroScan, hepatic steatosis, liver stiffness measurement, metabolic syndrome, hepatic stellate cells, cross-sectional study, BMC Endocrine Disorders, Dubai

The triglyceride-glucose index, usually abbreviated TyG, is computed as the natural logarithm of fasting triglycerides in milligrams per deciliter multiplied by fasting glucose, divided by two. Because both ingredients come from standard lipid and glucose panels, the index costs essentially nothing to calculate and has become a popular surrogate for insulin resistance in large epidemiological studies, where direct clamp-based measurements of insulin sensitivity are impractical. Prior cross-sectional work had already linked higher TyG values to higher urate levels, but those snapshots could not disentangle whether a person’s trajectory, rather than their standing, carried the risk signal.

That distinction matters because longitudinal studies of repeated measurements face a subtle statistical trap. If you measure an exposure, such as the TyG index, at the same visit at which you also confirm the outcome, such as hyperuricemia, the two overlap in time and the analysis can become circular. The research team, led by LingYi Feng and Ke Li of the Department of Clinical Laboratory at The Third People’s Hospital of Chengdu, addressed this with a landmark cohort design, a technique borrowed from survival analysis in which participants must remain free of the outcome up to a fixed landmark time before their subsequent risk is assessed.

Concretely, the investigators drew on de-identified health-examination records from 2023 through 2025 at their tertiary hospital. Only adults who remained normouricemic through the 2024 landmark were eligible, ensuring that the exposure window, the change in TyG between the 2023 and 2024 examinations, strictly preceded the outcome window, the development of hyperuricemia in 2025. The prespecified primary analysis then modeled that one-year TyG change jointly with baseline TyG, age, and sex, while also adjusting for the most recent pre-outcome values of serum urate and estimated glomerular filtration rate, the standard measure of kidney function, both taken from 2024. Because the kidneys excrete the bulk of uric acid, renal function is a critical confounder in any urate study, and using the 2024 values meant the model controlled for kidney status as close to the outcome as the data allowed.

The statistical machinery was modified Poisson regression with robust variance, an approach that estimates risk ratios directly rather than odds ratios, which can overstate effects in common outcomes like hyperuricemia. Among the 21,188 participants who met the eligibility criteria, 1,171, or 5.5 percent, developed hyperuricemia during 2025. In the primary model, each one-standard-deviation greater increase in the TyG index over the 2023-to-2024 year was associated with a 12 percent higher risk of subsequent hyperuricemia, with a risk ratio of 1.120 and a 95 percent confidence interval of 1.061 to 1.181. The confidence interval sits comfortably above 1.0, indicating a statistically robust association.

To test whether the result depended on the choice of adjustment variables, the team ran a complementary model that instead used 2023 serum urate and eGFR, the baseline-covariate version. The estimate barely moved, coming in at a risk ratio of 1.139 with a confidence interval of 1.080 to 1.201. This kind of sensitivity check is important in observational epidemiology, where results can hinge on modeling decisions. The consistency across both specifications suggests the association between rising insulin resistance and incident hyperuricemia is not an artifact of how kidney function or urate levels were anchored in the models.

Yet the study is equally notable for what it did not find. When the researchers asked whether adding one-year TyG change improved the model’s ability to discriminate who would and would not develop hyperuricemia, the gain was almost negligible: the C-statistic, a measure of predictive accuracy where 0.5 represents coin-flip performance and 1.0 represents perfection, rose only from 0.801 to 0.803. In plain terms, the most recent serum urate measurement and kidney function already carried nearly all the predictive information, and knowing how much a person’s TyG index had climbed added almost nothing to individual-level forecasting. The authors are explicit on this point, cautioning that the findings should be interpreted as observational associations rather than evidence of clinical predictive utility or causality.

That honesty is refreshing in a field where novel biomarker studies often overstate their practical implications. The biology linking insulin resistance to urate elevation is plausible and increasingly well mapped. Insulin resistance drives compensatory hyperinsulinemia, which stimulates renal sodium reabsorption and, through shared transporters in the proximal tubule, reduces urinary uric acid excretion. Elevated triglycerides and free fatty acids can also accelerate purine synthesis in the liver, increasing urate production from both ends of the balance. Transporter proteins such as URAT1 and ABCG2, which the study’s abbreviation list flags, sit at the molecular crossroads of this pathway, shuttling urate across renal and intestinal epithelia in ways that are sensitive to the metabolic milieu.

The study’s strengths lie in its scale, its clean temporal separation of exposure and outcome, and its reproducibility posture. The authors released Python analysis scripts, a data dictionary, locked machine-readable expected results, and figure-reproduction files as supplementary material, an unusually transparent package for a clinical epidemiology paper. The analysis followed STROBE reporting guidelines for observational studies, and the protocol, including the prespecified primary model, was defined before the analysis. Ethics approval came from the hospital’s medical ethics committee, with the consent requirement waived because only de-identified routine examination data were used.

Limitations remain, as they always do in single-center retrospective work. All participants came from one tertiary hospital in Chengdu, so the findings may not generalize to other populations or health systems. Health-examination cohorts also tend to skew toward people engaged enough with health care to show up annually, which can underrepresent the most metabolically vulnerable. Hyperuricemia was defined by serum urate thresholds rather than by clinical gout diagnoses, and the one-year measurement interval cannot capture shorter-term fluctuations in diet, weight, or medication use that might influence both TyG and urate. Residual confounding by unmeasured factors, from alcohol intake to diuretic prescriptions, cannot be excluded.

For readers wondering what to take away, the practical message is modest but meaningful. A rising TyG index over a year, visible in ordinary bloodwork, tracks with a modestly elevated chance of crossing the hyperuricemia threshold the following year, even after accounting for where you started and how your kidneys are doing. It is a signal worth watching in population health and in clinical conversations about metabolic trajectory, but it is not yet a tool that will refine anyone’s personal risk prediction. The authors’ own framing sets the right expectation: this is an observational association, a well-executed one, that adds a longitudinal brick to the growing wall of evidence connecting insulin resistance and uric acid metabolism, while leaving the questions of causation and clinical utility for future interventional and multi-center studies to answer.

Subject of Research: Association between one-year change in the triglyceride-glucose index and incident hyperuricemia in adults

Article Title: One-year change in the triglyceride-glucose index and subsequent incident hyperuricemia: a retrospective landmark cohort study

Article References: Feng, L., Xia, Y., Xie, Q., Wang, Y., & Li, K. (2026). One-year change in the triglyceride-glucose index and subsequent incident hyperuricemia: a retrospective landmark cohort study. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02663-7

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02663-7

Keywords: triglyceride-glucose index, hyperuricemia, insulin resistance, serum urate, cohort study, landmark analysis, gout, kidney function, epidemiology, metabolic health, modified Poisson regression, BMC Endocrine Disorders

The research, led by Tingting Zheng, Xingwei Cai, Lina Zhou and colleagues under the guidance of corresponding authors Zhongmin Wen and Yuan Zhang, enrolled 1,267 adults aged 45 and older who underwent health examinations at a single center. Of these, 488 had MASLD and 779 did not, giving the team a substantial dataset on which to train and test their algorithms. The participants provided demographic information, biochemical blood panels, anthropometric measurements such as waist circumference and body mass index, and body composition data captured through bioelectrical impedance analysis, a technique that estimates fat mass, muscle mass, water compartments and related metrics by measuring how the body resists a weak electrical current.

What makes the study methodologically interesting is its incremental design. Rather than throwing every available variable into a single black-box model, the researchers constructed four models of increasing complexity. The first relied on demographic data alone and achieved an area under the receiver operating characteristic curve, or AUC, of just 0.648, barely better than a coin flip weighted by age. Adding biochemical parameters pushed the AUC to 0.839, and incorporating anthropometric measurements lifted it further to 0.873. Only then did the team add the bioelectrical impedance metrics, which nudged the AUC to 0.877, a difference that was not statistically significant, with a p-value of 0.604. In plain terms, the expensive body composition scan contributed almost nothing that the cheap blood tests and tape measure had not already captured.

The variable selection process was deliberately rigorous. The team screened candidate predictors using correlation analysis, checks for multicollinearity quantified by variance inflation factors, Elastic Net regression, a technique that blends two penalty approaches to shrink irrelevant variables toward zero, and Bootstrap stability screening, which tests whether a variable survives repeated resampling of the data. Twelve variables emerged as stable and informative: the triglyceride-glucose index adjusted for body mass index, known as TyG-BMI, alanine aminotransferase, aspartate aminotransferase, the derived appendicular skeletal muscle index, albumin, low-density lipoprotein cholesterol, uric acid, a history of metabolic diseases, the derived phase angle, high-density lipoprotein cholesterol, age, and waist-to-height ratio.

Eight machine learning algorithms were then compared on these twelve predictors, and the winner was not a fashionable deep learning architecture but humble logistic regression. That model achieved a validation AUC of 0.869, with a sensitivity of 0.836, meaning it correctly caught about 84 percent of true MASLD cases, and a specificity of 0.755, meaning it correctly cleared about three-quarters of disease-free individuals. Performance was evaluated not only by discrimination but also by the Brier score, which penalizes overconfident wrong predictions, by calibration slope, which assesses whether predicted probabilities match observed reality, and by decision curve analysis, which estimates the net clinical benefit of acting on the model’s output across a range of risk thresholds.

To open the black box, the researchers applied SHAP analysis, or Shapley Additive exPlanations, a framework borrowed from cooperative game theory that assigns each variable a fair share of credit for every individual prediction. The SHAP results confirmed that TyG-BMI, a composite marker combining blood triglycerides, fasting glucose and body mass index, was the dominant contributor, followed by the liver enzymes ALT and AST, the muscle index, albumin and the lipid and uric acid measures. This transparency matters clinically: a physician using the model can see exactly why a particular patient was flagged, rather than trusting an opaque score, which is essential for building trust and for identifying when the model might be reasoning from a spurious correlation.

The team also confronted an uncomfortable methodological issue head-on. Several of their strongest predictors overlap with, or closely approximate, the components of the MASLD definition itself, which is based on hepatic steatosis in the presence of metabolic dysfunction criteria such as elevated waist circumference, triglycerides, blood pressure or glucose measures. This creates a risk of incorporation bias, where the model appears accurate partly because it is essentially reconstructing the diagnostic criteria rather than discovering independent biological signals. To probe this, the researchers ran sensitivity analyses showing that a model built from non-diagnostic overlapping variables still retained substantial diagnostic capacity, and in a separate criteria-based analysis, adding eight non-overlapping predictors increased the AUC from 0.749 to 0.809, a statistically significant gain of 0.060 with a p-value of 0.0003. This indicates the model does carry information beyond the metabolic criteria, although the authors caution that potential incorporation bias could not be fully eliminated.

The implications for aging populations are significant. As people age, fat redistributes toward visceral depots and skeletal muscle mass declines, a combination that reshapes the metabolic phenotype of fatty liver disease in older adults. The derived appendicular skeletal muscle index and phase angle, both bioelectrical impedance-derived markers of muscle quantity and cellular health, ranked among the top contributing variables, suggesting that the muscle-liver axis is genuinely informative in this age group even if it does not dramatically outperform conventional measures. This aligns with a growing body of research linking sarcopenia, the age-related loss of muscle, to metabolic deterioration and hepatic fat accumulation.

Why does the finding that body composition adds little matter beyond academic curiosity? Bioelectrical impedance analyzers are increasingly marketed directly to consumers and embedded in smart scales, and clinics have begun investing in them as screening tools. This study suggests that for MASLD identification specifically, a fasting blood panel measuring glucose, triglycerides, cholesterol fractions, liver enzymes and uric acid, combined with height, weight and waist measurements, captures nearly all of the available discriminative information. That is good news for resource-limited settings, where a simple, interpretable, twelve-variable logistic model requiring no special equipment could serve as an auxiliary screening tool to decide which patients warrant further evaluation with ultrasound or other imaging.

The authors are appropriately measured about what their model can and cannot do. Because this was a single-center, retrospective, cross-sectional study, the model has been internally validated only, meaning its performance was assessed on data from the same population in which it was developed. External validation in independent cohorts, ideally with prospective enrollment and diverse demographic profiles, is required before any clinical implementation. The final logistic regression model should therefore be regarded as an internally validated auxiliary classification tool, not a diagnostic device. Still, in an era when artificial intelligence in medicine often means opaque algorithms trained on opaque data, this study is a refreshing counterexample: a transparent, statistically disciplined model that tells clinicians exactly which routine measurements matter most for catching a silent epidemic of the aging liver.

Subject of Research: Development of an interpretable machine learning model for identifying metabolic dysfunction-associated steatotic liver disease using metabolic and body composition indicators in middle-aged and older adults

Article Title: Development and internal validation of an interpretable model for MASLD identification using metabolic and body composition indicators in middle-aged and older adults

Article References: Zheng, T., Cai, X., Zhou, L., Wang, S., Xu, J., Wen, Z., & Zhang, Y. (2026). Development and internal validation of an interpretable model for MASLD identification using metabolic and body composition indicators in middle-aged and older adults. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02635-x

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02635-x

Keywords: MASLD, fatty liver disease, machine learning, logistic regression, body composition, bioelectrical impedance analysis, TyG index, SHAP, metabolic dysfunction, aging, screening model, BMC Endocrine Disorders

The research team, led by Tongkai Wu, Jin Yu, Yinfeng Chen, Dongrong Yu, and Jun Ni, assembled a cohort of 905 patients with diabetic kidney disease whose baseline estimated glomerular filtration rate, or eGFR, sat between 15 and just under 60 mL/min/1.73 m², the range defining moderate to severe chronic kidney disease. Enrolment stretched from June 2008 to April 2024, giving the study more than fifteen years of real-world clinical data drawn from a single center. Using the urine albumin-to-creatinine ratio, the standard metric for classifying albuminuria, the researchers divided patients into 176 with NADKD and 729 with the conventional albuminuric form of diabetic kidney disease. Because the two groups differed in many baseline characteristics, the investigators applied propensity score matching in a one-to-one fashion, a statistical technique that pairs patients across groups with similar overall risk profiles, ultimately yielding 151 well-balanced patients in each arm with a median age of 73 years.

The comparison revealed a metabolic fingerprint that sets NADKD apart in striking ways. Patients in the non-albuminuric group had higher serum albumin levels but lower fasting blood glucose, lower total cholesterol, lower glycated hemoglobin, and a lower triglyceride-glucose index, a composite marker of insulin resistance. In plain terms, these patients carried a milder metabolic derangement than their albuminuric counterparts, yet their kidneys had still declined to the same functional territory. That paradox is precisely what makes NADKD clinically perplexing: if chronic hyperglycemia and dyslipidemia drive kidney injury in diabetes, why do some patients with comparatively better metabolic control lose kidney function while their urine remains essentially albumin-free?

One of the study’s most eye-catching findings points toward an answer that lies outside the kidney itself. Nearly sixty percent of the NADKD patients, 59.6 percent, had macrovascular disease, meaning clinically manifest atherosclerotic conditions such as coronary artery disease, cerebrovascular disease, or peripheral arterial disease, compared with only 39.7 percent of the albuminuric group, a difference that reached statistical significance at P equals 0.001. This inverted burden, heavier vascular disease in the group with lighter microvascular footprints, suggests that non-albuminuric diabetic kidney disease may not simply be an early or attenuated version of classic diabetic nephropathy but a biologically distinct entity in which large-vessel pathology and ischemic injury to kidney tissue play a proportionally larger role than the microvascular mechanisms classically blamed for diabetic kidney failure.

The prognostic implications of that distinction are substantial. Kaplan-Meier survival analysis, the standard method for tracking event-free survival over time, showed significantly better renal survival in the NADKD group, with a P value below 0.001. The annual rate of eGFR decline told the same quantitative story: patients with non-albuminuric disease lost kidney function at 0.747 mL/min/1.73 m² per year on average, whereas albuminuric patients declined at 3.224 mL/min/1.73 m² per year, more than four times faster. Slippage of that magnitude compounds mercilessly; a patient losing three-plus units annually from a baseline of 40 will typically cross the threshold toward dialysis within a decade, while a slower decliner may never reach end-stage disease within their lifetime. All-cause mortality, intriguingly, did not differ significantly between the groups, with a P value of 0.054 that narrowly missed conventional significance, hinting that while the kidneys fare differently, overall survival risk may be comparable once the vascular burden is accounted for.

To move from description to prediction, the researchers built a multivariable Cox proportional hazards model, the workhorse tool for identifying independent risk factors for time-to-event outcomes. In that analysis, macrovascular disease emerged as an independent risk factor for composite renal outcomes, carrying a hazard ratio of 4.19 with a 95 percent confidence interval spanning 1.45 to 12.09 and a P value of 0.008. A hazard ratio above four is clinically meaningful: after adjusting for other variables, patients with macrovascular disease faced more than quadruple the instantaneous risk of a adverse renal event compared to those without. The final parsimonious model contained just four variables, age, sex, baseline eGFR, and macrovascular disease, an appealingly simple set that any nephrologist or endocrinologist could assemble from a routine clinic visit.

The performance statistics of that model were reported with a methodological rigor that deserves attention. The apparent concordance index, or C-index, a measure of how well the model ranks risk across patients, reached 0.762. Because apparent performance in the development dataset tends to be optimistically biased, the team applied bootstrap internal validation, resampling the data 1,000 times to estimate and subtract that optimism. The optimism-corrected C-index settled at 0.744, still comfortably in the range considered acceptable to good for clinical prediction models. Time-dependent analyses added further support: the area under the receiver operating characteristic curve for predicting renal outcomes at three years was 0.828, and at five years it was 0.745. In practical terms, the model correctly discriminates high-risk from low-risk patients well enough to justify early intensified monitoring in those flagged as vulnerable, particularly those with documented macrovascular disease.

The conceptual takeaway is potentially practice-changing. Screening algorithms built around albuminuria may systematically miss patients whose kidneys are quietly deteriorating without a protein leak, delaying both diagnosis and the initiation of kidney-protective therapies. The study’s authors argue that NADKD represents a distinct subtype of diabetic kidney disease, one characterized by a heavier macrovascular burden and a slower but real trajectory of renal decline. That framing carries a direct clinical message: in a patient with diabetes and falling eGFR, a normal urine albumin-to-creatinine ratio should reassure nobody. Instead, the presence of coronary, cerebral, or peripheral atherosclerosis in such patients should itself be read as a renal risk signal, and serial eGFR monitoring should be treated as the primary surveillance tool for this population in the way albumin measurements have long been for others.

As with all single-center retrospective studies, caveats apply. The cohort came from one hospital in Hangzhou, China, and the retrospective design means classification depended on the quality of routine clinical documentation rather than protocolized assessments; residual confounding by unmeasured variables can never be fully excluded even with propensity matching. Kidney biopsies, the diagnostic gold standard that would confirm whether the pathology of NADKD truly differs at the tissue level, were not part of the analysis, so the mechanistic interpretation, that large-vessel ischemia substitutes for microvascular injury, remains a well-supported hypothesis rather than a histologically proven fact. The prediction model, validated internally by bootstrap but not yet in an external independent cohort, will need confirmation in multi-center populations before it can be embedded in guidelines.

Nevertheless, the study lands at a moment when the clinical stakes of diabetic kidney disease are rising sharply. Global diabetes prevalence continues to climb, kidney replacement therapy remains enormously expensive, and newly available kidney-protective drugs, from SGLT2 inhibitors to non-steroidal mineralocorticoid receptor antagonists, work best when started early. Identifying which patients need those therapies most urgently, including the non-albuminuric minority that traditional screening overlooks, could translate directly into preserved kidney function and avoided dialysis. If follow-up studies confirm that macrovascular disease reliably stratifies renal risk in NADKD, the humble urine dipstick may finally have to share its throne with a stethoscope, an electrocardiogram, and a routinely calculated eGFR slope, a small shift in workflow with the potential to change outcomes for millions of patients whose kidneys fail without ever leaking a trace of albumin.

Subject of Research: Clinical characteristics and prognosis of non-albuminuric diabetic kidney disease in patients with reduced kidney function

Article Title: Clinical characteristics and prognosis of non-albuminuric diabetic kidney disease: a single-center retrospective cohort study

Article References: Wu, T., Yu, J., Chen, Y., Yu, D., & Ni, J. (2026). Clinical characteristics and prognosis of non-albuminuric diabetic kidney disease: a single-center retrospective cohort study. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02646-8

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02646-8

Keywords: diabetic kidney disease, non-albuminuric diabetic kidney disease, chronic kidney disease, macrovascular disease, eGFR decline, albuminuria, renal prognosis, risk prediction model, propensity score matching, BMC Endocrine Disorders, diabetic nephropathy, retrospective cohort study

The research team, led by Lei Chen and WeiXia Du of the Department of Gastroenterology at Huanggang Central Hospital in Hubei Province, China, systematically searched the literature for studies on ampullary neuroendocrine tumors published between January 2010 and June 2024. From that search, 16 studies involving 32 patients met inclusion criteria. The investigators collected clinical and pathological data, imaging findings, treatment regimens, and follow-up outcomes, then retrospectively analyzed the pooled dataset. Because these tumors are so uncommon that no single center can accumulate a meaningful series, such pooled case-level analysis is one of the few ways to extract statistically meaningful patterns from the literature.

The demographic picture that emerges is consistent with what fragmentary prior reports suggested. Of the 32 patients, 19 were male and 13 female, a male-to-female ratio of 1.46 to 1, and the median age at diagnosis was 56 years, with patients ranging from 18 to 77. What makes the clinical picture treacherous, however, is how nonspecific it is. The most common presenting complaint was abdominal pain, seen in 35.5 percent of cases, followed by jaundice in 22.5 percent. Perhaps most strikingly, more than a quarter of cases, 25.8 percent, were discovered incidentally during routine health examinations, meaning a substantial fraction of patients had no symptoms at all pointing toward the ampulla.

That nonspecificity has real consequences. Ampullary neuroendocrine tumors are prone to being overlooked or misdiagnosed, the authors note, because their symptoms overlap with far more common conditions such as biliary colic, peptic ulcer disease, or ordinary ampullary adenomas. A patient with intermittent abdominal pain or mild jaundice may cycle through evaluations for gallstones before anyone considers a hormone-producing neoplasm hiding at the papilla. The review underscores that clinicians need a low threshold for examining the ampulla directly, particularly during endoscopic procedures performed for unrelated reasons, since incidental detection is already a common route to diagnosis.

On the imaging side, computed tomography dominated the diagnostic workup, having been used in 90.6 percent of cases, while magnetic resonance imaging was employed in 37.5 percent. But the headline finding of the review concerns tissue diagnosis rather than imaging. Endoscopic ultrasound-guided biopsy and endoscopic retrograde cholangiopancreatography-guided biopsy each achieved a diagnostic accuracy of 100 percent in the pooled cases. In an era when preoperative misdiagnosis remains a persistent problem for ampullary lesions generally, the message is unambiguous: when these endoscopic sampling techniques are deployed, they essentially never fail to establish the correct diagnosis before surgery.

The technical explanation for that accuracy lies in the anatomy and the tools. Endoscopic ultrasound places a high-frequency transducer directly adjacent to the ampulla through the duodenal lumen, allowing not only needle biopsy of the lesion but also precise measurement of tumor size and depth of invasion into the duodenal wall. ERCP, meanwhile, provides direct visual and contrast-based access to the ampullary orifice, permitting forceps biopsy from the exposed tumor surface. The review found the median tumor diameter to be just 15.5 millimeters, underscoring how small these lesions typically are at detection, and how critical high-resolution endoscopic assessment is for lesions that CT alone may characterize poorly.

Two pathological variables emerged as the key predictors of spread. Both larger tumor size and greater depth of invasion were associated with a higher rate of lymph node metastasis, and the authors additionally identified tumor grade as a factor linked to nodal risk. This trio of characteristics, size, depth, and grade, forms the practical decision tree for treatment planning. Small, superficial, low-grade tumors carry a low risk of occult nodal disease and can be candidates for organ-preserving local therapy, whereas larger, deeply invading, or higher-grade lesions demand more aggressive management because the probability that cancer has already reached regional lymph nodes rises substantially.

Treatment in the pooled cohort reflected that stratification. Endoscopic resection was performed in 31.3 percent of patients, while surgical resection accounted for 59.4 percent. Endoscopic resection, typically performed with techniques such as endoscopic papillectomy, spares the patient a major operation but is appropriate only for carefully selected early-stage cases confined to the superficial layers of the ampulla. Radical surgery, most commonly pancreaticoduodenectomy, remains the core therapeutic modality for high-risk or advanced-stage disease, removing the tumor together with regional lymph nodes. The authors emphasize that treatment decisions should be individualized according to tumor size, depth, and grade, and made through multidisciplinary collaboration among gastroenterologists, surgeons, pathologists, and oncologists.

The prognostic news is genuinely encouraging. With a median follow-up duration of 38.9 months, the pooled cohort showed a 5-year overall survival rate of 96.4 percent, a figure that stands in sharp contrast to the far grimmer outcomes of ampullary adenocarcinomas and neuroendocrine carcinomas. Well-differentiated neuroendocrine tumors of the ampulla, detected while still small, are in most cases indolent and curable. Yet the authors caution that the risk of recurrence persists even after apparently curative treatment, and they argue that establishing a long-term, regular follow-up system is crucial, especially for patients with high-risk pathological features such as larger size, deeper invasion, or higher grade.

The review’s authors are candid about the limits of their evidence. With only 32 patients drawn from retrospective published cases, the correlations they report between tumor characteristics and lymph node metastasis need to be verified in larger-scale prospective studies before they can harden into firm guidelines. Still, the analysis delivers practical takeaways that clinicians can apply immediately: consider ampullary neuroendocrine tumors in the differential of unexplained abdominal pain or jaundice, exploit EUS and ERCP-guided biopsy for near-certain preoperative diagnosis, let size, depth, and grade drive the choice between endoscopic resection and radical surgery, and commit every patient to structured long-term surveillance. For a tumor so rare that most physicians will see only a handful of cases in a career, 14 years of pooled global experience is a meaningful step from anecdote toward evidence.

Subject of Research: Clinical characteristics, diagnosis, treatment, and prognosis of ampullary neuroendocrine tumors

Article Title: Systematic review of ampullary neuroendocrine tumors: a pooled analysis of 32 cases reported from 2010 to 2024

Article References: Chen, L., Du, W., Fang, L., Wu, J., & Wang, C. (2026). Systematic review of ampullary neuroendocrine tumors: a pooled analysis of 32 cases reported from 2010 to 2024. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02460-2

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02460-2

Keywords: ampullary neuroendocrine tumors, ampulla of Vater, systematic review, endoscopic ultrasound, ERCP, endoscopic resection, lymph node metastasis, neuroendocrine neoplasms, prognosis, gastroenterology, BMC Endocrine Disorders, pooled analysis

The study, led by Ming-xin Chen and colleagues at the Affiliated Kunshan Hospital of Jiangsu University and Wuxi Ninth People’s Hospital Affiliated to Soochow University, analyzed 2,699 adults aged 50 and older who were admitted with fractures. Each patient contributed a single index hospitalization, defined as the first eligible admission during the study period, and all biochemical measurements were drawn from fasting blood samples obtained within the first 24 hours of admission. That early sampling window was a deliberate design choice: later inpatient values were excluded because surgery, fasting, stress, and fluid shifts can distort lipid and uric acid readings as hospitalization proceeds. Hyperuricemia was defined using sex-specific thresholds, with serum uric acid at or above 416 micromoles per liter in men and at or above 357 micromoles per liter in women, cutoffs consistent with the solubility limit of monosodium urate in physiological conditions.

Across the cohort, 418 patients, or 15.5 percent, met the definition of hyperuricemia. When the researchers modeled the odds of prevalent hyperuricemia against the NHHR expressed per one standard deviation, they found a prevalence odds ratio of 1.188, with a 95 percent confidence interval of 1.061 to 1.330 and a p-value of 0.003. Translated into plain terms, each standard deviation increase in the cholesterol ratio was associated with roughly a 19 percent higher odds of having elevated uric acid. Because odds ratios can overstate effect sizes when an outcome is not rare, the team also computed modified-Poisson prevalence ratios, a technique that estimates the ratio of actual prevalences directly. That complementary estimate came in at 1.129, a 13 percent higher prevalence per standard deviation, still statistically robust with a confidence interval of 1.045 to 1.220.

The statistical machinery behind these numbers deserves attention, because it reflects how modern observational epidemiology tries to protect itself against its own weaknesses. Missing data were handled through multiple imputation across 20 datasets, with the pooled estimates drawn from that ensemble rather than from any single filled-in version of the data. Covariate selection was clinically informed rather than purely algorithmic, and kidney function was accounted for using flexible adjustment for estimated glomerular filtration rate, an important step since the kidneys are the primary route of uric acid excretion and reduced filtration capacity can inflate serum urate independently of any lipid effect. Sensitivity analyses using complete cases, selection weighting, and alternative uric acid thresholds were run to test whether the headline association was an artifact of any single modeling decision.

Those sensitivity checks told a more cautious story. The selection-weighted estimates, at an odds ratio of 1.191 and a prevalence ratio of 1.131, were nearly identical to the primary results, which is reassuring. But the complete-case analysis, restricted to patients with no missing values on the relevant variables, produced an odds ratio of 1.130 with a confidence interval of 0.999 to 1.277, brushing against the boundary of statistical significance. The authors attribute this fragility to lipid missingness and limited positivity, the technical term for situations where certain combinations of covariates are rare or absent in the data, making it hard to estimate effects across the full covariate distribution. In other words, the association is real enough to appear consistently, but its precise magnitude depends on assumptions about why some laboratory values were never measured.

Perhaps the most sobering result concerns what the NHHR adds beyond its component parts. When triglycerides and low-density lipoprotein cholesterol were included in the same joint model, the NHHR odds ratio shrank to 1.046 with a confidence interval of 0.876 to 1.248, comfortably spanning the null. This is a classic demonstration of collinearity in lipid epidemiology: the ratio is mathematically constructed from cholesterol fractions that travel together with triglycerides and LDL in the metabolic syndrome cluster, so much of the apparent signal is shared information rather than independent predictive power. Variance inflation factor and correlation analyses in the supplementary material quantify this overlap, and the joint model makes clear that the ratio does not carve out a large independent niche once its ingredients are accounted for.

The team also probed whether the ratio could improve actual risk prediction, using cross-validated area under the receiver operating characteristic curve as the yardstick. Adding NHHR to an existing model nudged the AUC from 0.777 to 0.780, an incremental gain of 0.0026 with a conditional 95 percent interval running from minus 0.0016 to 0.0070, an interval that includes zero. For readers unfamiliar with discrimination metrics, an AUC of 0.5 indicates performance no better than a coin flip, while 1.0 indicates perfect separation; the baseline model already performed respectably, and the cholesterol ratio contributed essentially nothing measurable on top of it. The authors explicitly state that no validated prediction role or clinical NHHR cutoff is established by this work, a disclaimer that guards against the common temptation to convert a statistical association into a screening recommendation overnight.

Nonlinearity, another fashionable question in modern epidemiology, fared no better. The researchers searched for threshold effects or U-shaped relationships between the ratio and uric acid, but the evidence proved unstable after trimming extreme NHHR values, meaning that any apparent curvature in the raw data was driven by a handful of outlying patients rather than a reproducible biological pattern. This matters because nonlinear claims, once published, tend to metastasize into clinical folklore about optimal target values. By reporting the instability transparently, the study models a kind of negative-result honesty that is still too rare in the literature on metabolic biomarkers.

There is also a deeper physiological caveat that the authors flag: fasting blood drawn within 24 hours of a fracture admission cannot fully separate chronic metabolic status from the acute physiology of hospitalization. Trauma, pain, immobility, and the inflammatory response to injury all shift lipid and uric acid levels, and while restricting sampling to the first day minimizes the most dramatic perturbations, it cannot eliminate them. The fracture-hospitalized population is itself a selected one, enriched for older adults, and often for osteoporosis and frailty, so extrapolating the findings to healthy community populations requires caution. Selection-weighted analyses attempted to correct for the gap between the 2,699 analyzed patients and a larger 4,719-person target population, with balance diagnostics confirming that weighting improved covariate balance, but no weighting scheme can conjure data that were never collected.

What, then, is the takeaway from this study? It is a carefully executed, honestly reported observation that a lipid ratio long used as a composite marker of atherogenic burden, capturing both the cholesterol carried in potentially harmful particles and the protective capacity of HDL, travels alongside elevated uric acid in a real-world clinical population. The association is modest, statistically fragile at the margins, largely redundant with triglycerides and LDL, and of no demonstrated predictive value. Yet the biological plausibility is genuine: insulin resistance, visceral adiposity, and renal handling of both urate and lipids form a tightly connected metabolic web, and elevated uric acid has been implicated in oxidative stress, endothelial dysfunction, and inflammation. Studies like this one, which quantify an association and then meticulously document how far it can and cannot be pushed, are the unglamorous scaffolding on which eventual mechanistic and interventional work is built. For now, the NHHR remains what it has always been, a useful lipid summary measure, with a newly documented, cautiously framed companion relationship to the uric acid that crystallizes in gouty joints.

Subject of Research: Association between the non-HDL-C/HDL-C cholesterol ratio and prevalent hyperuricemia in adults hospitalized with fractures

Article Title: Non-HDL-C/HDL-C ratio and prevalent hyperuricemia in adults hospitalized with fractures: a retrospective cross-sectional study

Article References: Chen, M.-X., Feng, L.-L., Huang, X.-Y., Wang, C., Lu, K., & Shan, H.-Q. (2026). Non-HDL-C/HDL-C ratio and prevalent hyperuricemia in adults hospitalized with fractures: a retrospective cross-sectional study. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02619-x

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02619-x

Keywords: NHHR, hyperuricemia, serum uric acid, lipid ratio, HDL cholesterol, fracture, retrospective cross-sectional study, gout, metabolic syndrome, prevalence odds ratio, multiple imputation, BMC Endocrine Disorders

The research, published in BMC Endocrine Disorders by a team led by radiologist Sook Min Hwang of Hallym University Kangnam Sacred Heart Hospital in Seoul, examined 88 consecutive children and adolescents who presented with goiter and normal thyroid function between January 2000 and January 2024. Every patient underwent thyroid ultrasonography and biochemical testing for thyroid autoimmunity within three months of each other, allowing the investigators to compare what the ultrasound saw against what the immune system was actually doing. Based on thyroid antibody status, the cohort was divided into two groups: 34 children who tested positive for autoimmune thyroiditis and 54 who did not. The design is elegant in its simplicity, because it isolates a question that clinicians face every day: when a child has a goiter but a euthyroid blood panel, does the ultrasound add anything that blood work cannot yet reveal?

The answer, according to the data, is a resounding yes. Three ultrasound features separated the two groups with striking clarity. Hypoechogenicity, meaning the gland appears darker than normal on the scan, was present in 41.2 percent of the autoimmune group but only 3.7 percent of the non-autoimmune group. Heterogeneous echotexture, a patchy, uneven appearance of the thyroid tissue, was seen in 85.3 percent versus 33.3 percent. And increased vascularity, detectable with Doppler imaging as heightened blood flow through the gland, appeared in 76.5 percent versus 25.9 percent. All three differences were statistically significant at p less than 0.001, a level of consistency across independent imaging features that is rare in studies of this size and speaks to a genuine biological signal rather than statistical noise.

Each of these findings has a plausible pathophysiological explanation rooted in what autoimmune thyroiditis actually does to thyroid tissue. In Hashimoto thyroiditis, the most common form of the disease in children, lymphocytes infiltrate the gland and progressively destroy the follicular cells that produce thyroid hormone. This inflammatory infiltration disrupts the uniform architecture of healthy thyroid tissue, which is why the gland takes on a heterogeneous, pseudo-lobulated appearance. The destruction of follicular cells and their colloid-rich interior reduces the number of acoustic interfaces that normally reflect ultrasound waves, making the tissue appear darker, or hypoechoic. Meanwhile, the inflammatory process triggers angiogenesis, the growth of new blood vessels to supply the immune cell infiltrate, which manifests as increased vascularity on Doppler examination. In Graves disease, the other major autoimmune thyroid condition, stimulating antibodies drive hyperactivity of the gland, which also produces hypervascularity and diffuse textural change.

The most clinically actionable result emerged when the researchers combined these features. When a child showed two or more of the three sonographic abnormalities, the composite criterion achieved a sensitivity of 85.3 percent for autoimmune thyroiditis. In multivariate logistic regression analysis, the composite criterion remained an independent predictor, with an odds ratio of 4.63 and a 95 percent confidence interval of 1.39 to 15.44, corresponding to a p value of 0.013. In practical terms, a child with a goiter and normal thyroid function who displays at least two of these ultrasound findings is more than four and a half times as likely to have autoimmune thyroiditis as a child whose gland looks uniform, normally echogenic, and normally vascular. That kind of risk stratification, derived from an imaging study that is already being performed, costs nothing extra.

The clinical context matters enormously here. Autoimmune thyroiditis is the leading cause of diffuse thyroid disease in the pediatric population, and if it goes undetected it can seriously impair growth, pubertal development, bone mineralization, and cognitive function. The paradox that complicates early diagnosis is that many children present with goiter while their thyroid function remains normal, a state sometimes called euthyroid goiter. Standard practice in many settings is to reassure these patients and monitor them, reserving antibody testing for those who develop abnormal hormone levels. But antibodies can lag behind tissue changes, and the study’s data suggest that the gland often announces the autoimmune process sonographically before the blood test confirms it. Waiting for the laboratory to catch up may mean missing a window in which intervention or closer surveillance could prevent complications.

Beyond the imaging findings themselves, the study identified clinical variables that independently associated with autoimmune thyroiditis. Children in the autoimmune group were significantly older, with a mean age of 14.29 years compared with 11.56 years in the non-autoimmune group, a difference that was statistically significant at p equal to 0.01. They were also significantly more likely to have a family history of autoimmune thyroid disease, at 26.5 percent versus 9.3 percent, with p equal to 0.03. Both findings align with the known epidemiology of the condition: autoimmune thyroiditis has a strong genetic component, with familial clustering well documented, and its incidence rises through childhood and peaks in adolescence, particularly in girls. Age and family history therefore serve as useful contextual clues, but they are far less discriminating on their own than the ultrasound features, which showed much larger absolute differences between the groups.

The authors are careful to frame the composite criterion as a risk stratification tool rather than a definitive diagnostic test, and that framing is scientifically appropriate. Sensitivity of 85.3 percent means that roughly one in seven children with autoimmune thyroiditis would still be missed by the two-abnormality threshold, which is why the researchers position it as a trigger for expedited serological testing rather than a replacement for it. A negative ultrasound pattern, with fewer than two abnormalities, should not end the workup in a child with a concerning presentation. Moreover, the study was retrospective and conducted at a single institution, and the authors explicitly state that prospective multicenter validation is needed before the approach is widely implemented. Ultrasound interpretation also carries an inherent operator dependency, and the prevalence of autoimmune thyroiditis in the study cohort, nearly 39 percent, may be higher than in broader primary care populations, which would affect the predictive values of the criterion in routine practice.

Even with those caveats, the implications for clinical workflow are substantial, particularly in resource-limited settings where every antibody panel represents a real cost. Thyroid ultrasonography is painless, involves no ionizing radiation, and is already the standard first-line imaging modality for pediatric goiter in most guidelines. Embedding a structured assessment of echogenicity, echotexture, and vascularity into the routine interpretation of these scans, and using the two-or-more-abnormality rule to prioritize which children receive immediate antibody testing, could shorten the diagnostic odyssey for many families. It could also reduce unnecessary testing in children whose glands look entirely normal, freeing resources while sparing them the anxiety of prolonged uncertainty. The approach fits into a broader trend in pediatric imaging toward using routinely acquired data to extract earlier, more actionable signals.

What makes the study resonate beyond its immediate clinical utility is the picture it paints of disease unfolding in real time. The thyroid gland of a fourteen-year-old with early Hashimoto thyroiditis is not simply a normal organ that happens to be large; it is an organ under immunological siege, darker and patchier on the scan, flushed with the extra blood flow of active inflammation, all while the endocrine system compensates well enough to keep hormone levels within reference range. The ultrasound, in this light, functions as a window into the earliest phase of the autoimmune process, capturing the tissue-level battle that blood tests will only register once the gland’s functional reserve begins to fail. For the children in this study, that window translated into earlier identification and, potentially, earlier protection of growth and development. For clinicians everywhere who face the daily question of what to do with a child’s painless neck swelling and a normal lab panel, the message is that the image on the screen deserves to be read not just for size, but for texture, brightness, and flow, because those details may hold the diagnosis the blood has not yet revealed.

Subject of Research: Thyroid ultrasonography for early detection of autoimmune thyroiditis in euthyroid children and adolescents with simple goiter

Article Title: Thyroid ultrasonography as an early diagnostic tool for autoimmune thyroiditis in children and adolescents with simple goiter

Article References: Hwang, S. M., Park, C. Y., Moon, J. H., Woo, J. Y., & Lee, H. J. (2026). Thyroid ultrasonography as an early diagnostic tool for autoimmune thyroiditis in children and adolescents with simple goiter. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02432-6

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02432-6

Keywords: autoimmune thyroiditis, thyroid ultrasonography, pediatric goiter, Hashimoto thyroiditis, Graves disease, hypoechogenicity, thyroid vascularity, risk stratification, thyroid antibodies, pediatric endocrinology, early diagnosis, BMC Endocrine Disorders

Oral frailty is a concept that has gained momentum in geriatric medicine over the past decade. It describes an age-related, potentially reversible decline in oral function, encompassing diminished chewing ability, difficulty swallowing, reduced tongue and lip strength, a dry mouth, and a tendency to avoid harder foods. Unlike tooth loss alone, oral frailty is a functional syndrome, capturing how well the entire oral apparatus performs rather than simply counting missing teeth. Researchers have already linked it to malnutrition, sarcopenia, cognitive decline, and cardiovascular risk. What has remained far less clear is whether oral frailty tracks with kidney abnormalities, particularly in older adults with type 2 diabetes, a population in which both oral deterioration and diabetic kidney disease are common and often progress silently for years.

To probe that question, the research team led by Xiaoling Deng and Wenbo Xia, with Lili Shi as corresponding author, recruited 201 hospitalized patients aged 60 and older who had been diagnosed with type 2 diabetes. The study was approved by the hospital’s ethics committee and conducted in accordance with the Helsinki declaration, with written informed consent from all participants. Each patient underwent a structured oral assessment using two validated instruments: the Oral Frailty Five-Item Checklist, known as OF-5, and the eight-item Oral Frailty Index, or OFI-8. The OF-5 is a rapid screening tool that asks about practical difficulties such as chewing tough foods, swallowing, and social dining, while the OFI-8 provides a more granular functional profile. Alongside these oral measures, the researchers collected standard clinical data, including age, sex, diabetes duration, glycated hemoglobin, and hypertension status, as well as laboratory values covering renal function, blood counts, and inflammatory indices.

The results were striking. According to the OF-5 checklist, 46.3 percent of the participants screened positive for oral frailty, meaning nearly one in two older diabetic inpatients showed measurable decline in oral function. The OFI-8 classified 35.3 percent as high risk. When the researchers cross-referenced these oral findings with kidney measures, a consistent pattern emerged. Patients positive on the OF-5 had roughly 4.7 times higher odds of having an estimated glomerular filtration rate below 90 milliliters per minute per 1.73 square meters, the conventional threshold indicating reduced kidney filtration capacity. Even more pronounced was the association with albuminuria: OF-5 positivity was linked to about 7.1 times higher odds of a first-morning urinary albumin concentration at or above 30 milligrams per liter, a level that signals glomerular damage and is a cornerstone marker for early diabetic kidney disease.

These associations held up after statistical adjustment. The team used multivariable logistic regression models that controlled for age, sex, diabetes duration, HbA1c, and hypertension, the major clinical covariates that typically confound relationships between diabetes complications. The fact that oral frailty remained independently associated with both renal outcomes after these adjustments suggests the link is not merely a byproduct of older age, longer disease duration, poorer glycemic control, or coexisting high blood pressure, each of which independently damages kidneys. The magnitude of the adjusted odds ratios, particularly for urinary albumin, is notable for a screening tool that costs nothing to administer and takes only moments to complete.

Perhaps the most technically interesting part of the study is how the authors quantified whether oral frailty adds genuine predictive value beyond routine clinical information. They built a clinical base model using the standard covariates and then compared it with extended models that additionally incorporated the OF-5 or OFI-8 scores. Model performance was measured with the area under the receiver operating characteristic curve, or AUC, and was internally validated using 1,000 bootstrap resamples with optimism correction, a rigorous approach that guards against the inflated performance estimates that plague many small clinical prediction studies. For the eGFR outcome, adding the OF-5 raised the optimism-corrected AUC from 0.713 to 0.771. For elevated urinary albumin, the base model’s AUC of 0.675 improved to 0.783 with the OF-5 and to 0.779 with the OFI-8. Optimism-corrected calibration slopes for the extended models ranged from 0.852 to 0.887, indicating reasonably faithful agreement between predicted and observed risks.

The biological plausibility of a mouth-kidney connection rests on several converging mechanisms. Chronic low-grade inflammation is a central suspect: periodontal disease and oral dysfunction can elevate systemic inflammatory burden, and the inflammatory indices measured in the study, including the systemic immune-inflammation index and the systemic inflammation response index, reflect this axis. Sustained inflammation promotes endothelial injury in glomerular capillaries, accelerating albumin leakage. Another pathway involves nutrition. Oral frailty pushes patients toward soft, often processed foods and away from protein-rich and fibrous options, potentially driving the protein-energy malnutrition and sarcopenia that are themselves established risk factors for faster kidney decline. Shared vascular pathology offers a third explanation: the microvascular damage that impairs glomerular filtration in diabetes also compromises blood flow to the periodontal tissues and salivary glands, so oral decline and renal decline may simply be parallel manifestations of the same underlying vascular injury.

The authors are careful to spell out the limits of what their data can show. This was a single-center, cross-sectional study of hospitalized patients, a design that captures a snapshot in time and cannot establish whether oral frailty precedes kidney dysfunction or results from it. Reverse causation is plausible, since patients with advanced chronic kidney disease often suffer uremic taste changes, xerostomia, and reduced appetite that could worsen oral function. The inpatient setting also means the sample may overrepresent patients with more severe disease, limiting generalizability to community-dwelling older adults. The authors explicitly state that the findings do not establish causality, prospective predictive utility, or external validity, and the modest improvement in discrimination, while statistically meaningful, reflects a limited covariate set rather than a fully specified clinical model.

Even with those caveats, the clinical implications are compelling. Diabetic kidney disease affects an estimated 40 percent of people with diabetes and is the leading cause of end-stage renal disease worldwide, yet it is frequently diagnosed late because early stages produce no symptoms. Current screening relies on annual urine albumin testing and serum creatinine measurements, which many patients, particularly in resource-limited settings, do not receive on schedule. A validated oral frailty checklist could serve as an essentially free triage tool: a positive screen in an older diabetic patient might prompt earlier laboratory testing, closer nephrology follow-up, or intensified metabolic management. It could also motivate dental and medical teams to collaborate more systematically, since treating oral frailty through chewing exercises, denture optimization, and nutritional counseling is low-risk and potentially reversible.

Future research will need to determine whether the association holds prospectively, whether oral frailty scores measured in community settings predict incident kidney disease over years, and whether interventions that restore oral function actually slow renal decline. If those questions are answered affirmatively, the humble act of asking an older diabetic patient whether they can chew hard food or swallow comfortably could become a routine part of kidney risk assessment, joining blood pressure cuffs and glucose meters as a first-line instrument in the fight against one of diabetes’ most feared complications. For now, the Harbin study adds a provocative new entry to the growing list of systemic diseases that leave their fingerprints in the mouth, and it challenges clinicians to stop viewing oral health as separate from the rest of metabolic medicine.

Subject of Research: Association between oral frailty and early renal dysfunction in elderly patients with type 2 diabetes

Article Title: Correlation between oral frailty and renal function and urinary microalbumin in elderly patients with type 2 diabetes mellitus

Article References: Deng, X., Xia, W., Huo, D., Zhang, J., Sun, T., & Shi, L. (2026). Correlation between oral frailty and renal function and urinary microalbumin in elderly patients with type 2 diabetes mellitus. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02543-0

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02543-0

Keywords: oral frailty, type 2 diabetes, chronic kidney disease, estimated glomerular filtration rate, urinary albumin, diabetic nephropathy, geriatric medicine, albuminuria, screening, inflammation, BMC Endocrine Disorders, cross-sectional study

The logic behind the ratio rests on two well-studied biological threads. GGT is an enzyme found mainly in the liver and biliary tract, and elevated blood levels have long been linked not only to liver disease but also to oxidative stress, inflammation, and metabolic dysfunction. Population studies have repeatedly associated higher GGT with diabetes, high blood pressure, fatty liver disease, and atherosclerosis, the gradual narrowing of arteries that underlies heart attacks. HDL cholesterol, meanwhile, is widely regarded as protective because it participates in reverse cholesterol transport, ferrying cholesterol away from artery walls back to the liver. When GGT is high and HDL is low at the same time, the resulting ratio captures two damaging tendencies in a single number: a burden of oxidative and metabolic stress paired with a weakened defense system. Combining the two into one indicator, researchers hoped, might outperform either measurement alone and offer clinicians a cheap, readily available window into cardiovascular risk.

To test that idea, the team led by Sisi Chen and colleagues at Wuhan Fourth Hospital conducted a retrospective longitudinal cohort study of 3,123 female patients. The women were evaluated for coronary heart disease, and the researchers used multivariable logistic regression to ask whether a higher standardized GHR was independently associated with having the disease, after accounting for a broad range of clinical and laboratory covariates. The answer was a clear yes. Each one-standard-deviation increase in the ratio was associated with roughly two and a half times the odds of coronary heart disease, with a fully adjusted odds ratio of 2.466 and a 95 percent confidence interval of 1.943 to 3.130, a result that remained highly statistically significant. In conventional receiver operating characteristic analysis, the ratio showed modest ability to discriminate between women with and without the disease, achieving an area under the curve of 0.741.

The investigators then turned to the forward-looking question: among women who already had coronary heart disease, did the ratio predict future trouble? During a median follow-up of 665 days, 611 major adverse cardiovascular events occurred in the cohort. In a complete-case Cox regression model that included 683 patients with 183 events, and that was additionally adjusted for low-density lipoprotein cholesterol along with other clinical and laboratory variables, each one-standard-deviation rise in GHR was associated with a 49.8 percent higher rate of major adverse cardiovascular events, corresponding to a hazard ratio of 1.498 with a confidence interval of 1.358 to 1.653. The dose-response pattern was striking. When patients were divided into tertiles, or thirds, by their baseline ratio, women in the middle third had nearly double the event rate of those in the lowest third, with a fully adjusted hazard ratio of 1.902, while those in the highest third had more than three times the rate, at 3.291. The trend across tertiles was highly significant.

These findings place the GHR in a growing family of combined lipid-and-enzyme markers that researchers have proposed as refinements to traditional cholesterol panels. Standard risk assessment relies heavily on low-density lipoprotein cholesterol, total cholesterol, and related measures, yet many women who suffer cardiovascular events do not have dramatically abnormal values on those tests. Coronary heart disease in women has historically been under-recognized and under-studied, and sex-specific risk markers remain an active frontier in cardiology. A ratio assembled from two tests that are already ordered routinely—liver function panels include GGT, and lipid panels include HDL cholesterol—would cost nothing extra to calculate, which is precisely what makes such markers attractive. The Wuhan study’s large female-only cohort and its adjustment for numerous confounders, including lipid-lowering therapy, lend considerable weight to the observed associations.

However, the study’s most instructive finding may be the one that tempers enthusiasm. When the researchers applied a more rigorous, time-dependent form of receiver operating characteristic analysis focused specifically on the first 365 days of follow-up—a method that properly accounts for censoring, meaning patients whose follow-up ends before the full year—the ratio’s discriminative power collapsed. The area under the curve fell to just 0.594 overall, and across age subgroups it ranged only from 0.593 to 0.623, with no statistically significant differences between age groups. This is a stark contrast with the exploratory conventional ROC analysis, which had suggested the best discrimination among patients aged 75 and older, with an area under the curve of 0.737. The discrepancy illustrates a methodological pitfall in biomarker research: conventional ROC analyses that ignore censoring can overstate a marker’s predictive performance, and the authors’ decision to run both analyses side by side provides an unusually honest appraisal.

The authors are explicit about the implication. Higher GHR, they conclude, is associated with coronary heart disease and with the occurrence of major adverse cardiovascular events in women with the disease, but baseline GHR alone shows limited discrimination for one-year events and should not be considered a standalone short-term risk prediction marker. In practical terms, a high ratio may flag a woman as belonging to a higher-risk stratum, but it cannot tell a clinician with confidence whether a specific patient will experience a heart attack, stroke, or cardiovascular death within the next twelve months. The researchers also emphasize that prospective external validation is required before GHR can be incorporated into clinical risk assessment. Because the study was retrospective and conducted at a single institution in China, its findings must be reproduced in independent, diverse populations before any change to practice could be justified.

The study also has structural limitations worth understanding. Retrospective designs rely on data collected for clinical rather than research purposes, which can introduce measurement variability and missing information; the complete-case Cox model included only 683 of the 3,123 enrolled patients, a substantial reduction that may affect precision and generalizability. Residual confounding remains possible even after extensive adjustment, since factors such as diet, physical activity, alcohol consumption, and detailed medication adherence are difficult to capture fully in routine records. The authors note that they examined variance inflation factors to check for multicollinearity among covariates, and their supplementary analyses included correlations between GHR and the Gensini score, a measure of coronary artery disease severity, as well as Kaplan-Meier survival curves stratified by GHR tertiles, which visually confirmed the gradient in event-free survival across risk groups.

What emerges from this research is a nuanced picture rather than a breakthrough headline. The GGT-to-HDL-cholesterol ratio is a genuine and statistically robust signal: it reflects the convergence of oxidative stress and impaired cholesterol handling, it correlates with the presence and severity of coronary disease in women, and it stratifies patients into meaningfully different risk tiers over a follow-up period approaching two years. At the same time, the study demonstrates that association and prediction are not the same thing, and that a biomarker cheap enough to calculate on the back of any lab report still needs to clear a high bar before it can guide individual decisions. For now, the ratio is best viewed as a promising adjunct—a number that may enrich risk conversations between women and their cardiologists—while the search for reliable, sex-specific short-term predictors continues. The full open-access study is available with the DOI 10.1186/s12902-026-02534-1.

Subject of Research: The GGT-to-HDL-cholesterol ratio as a biomarker for coronary heart disease and cardiovascular outcomes in women

Article Title: Association between the gamma-glutamyl transferase to high-density-lipoprotein-cholesterol ratio and clinical outcomes in female patients with coronary heart disease

Article References: Chen, S., Li, Y., Zhu, X., He, F., Li, H., & He, L. (2026). Association between the gamma-glutamyl transferase to high-density-lipoprotein-cholesterol ratio and clinical outcomes in female patients with coronary heart disease. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02534-1

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02534-1

Keywords: GGT-to-HDL ratio, coronary heart disease, women's health, biomarkers, cardiovascular risk, major adverse cardiovascular events, HDL cholesterol, gamma-glutamyl transferase, Cox regression, risk prediction, BMC Endocrine Disorders, retrospective cohort study

The research, led by Muhammet Fatih Şahin of the Department of Internal Medicine at Kestel State Hospital in Bursa and colleagues at collaborating institutions, was published in BMC Endocrine Disorders as an open-access research article. The team set out to answer a deceptively simple question: among people with type 2 diabetes who undergo clinically indicated dual-energy X-ray absorptiometry, do those taking SGLT2 inhibitors show worse bone health than those who do not? To address it, the investigators assembled a primary analytical cohort of 56 patients with type 2 diabetes who were referred for DEXA scanning between June 2025 and February 2026. Of these, 31 patients were not using SGLT2 inhibitors and 25 were. A separate group of 44 participants without diabetes was recruited purely for secondary, descriptive comparisons, providing a reference point for how bone health parameters distribute in people free of the metabolic disease.

The choice of measurement technique matters enormously in this field, and the study leaned on the gold standard. Dual-energy X-ray absorptiometry, or DEXA, works by firing two X-ray beams of different energies through the body and measuring how much each is attenuated by bone tissue. Because calcium in bone absorbs X-rays far more efficiently than soft tissue, the differential attenuation allows clinicians to calculate bone mineral density with remarkable precision at clinically relevant sites. The researchers focused on the lumbar spine and the femoral neck, the narrow segment of the thigh bone just below the hip joint where fractures are both common and devastating. Each participant received a T-score at these sites, a number that expresses bone density as a comparison against the average peak bone mass of a healthy young adult. A T-score of minus 2.5 or lower at either site defines osteoporosis, while values between minus 1.0 and minus 2.5 fall into the intermediate territory of osteopenia.

The headline result was a statistical null. Osteoporosis, defined as a T-score of minus 2.5 or below at the lumbar spine or femoral neck, was present in 32.3 percent of the patients with type 2 diabetes who were not taking SGLT2 inhibitors and in 36.0 percent of those who were, a difference that was statistically indistinguishable with a p-value of 0.769. When the investigators moved beyond raw prevalence and applied logistic regression, adjusting for age, sex, and body mass index, the odds ratio for osteoporosis associated with SGLT2 inhibitor exposure came out at 1.169 with a 95 percent confidence interval stretching from 0.328 to 4.170, and a p-value of 0.810. An odds ratio of 1.0 would indicate no effect at all; the observed value of 1.169 hints at a modestly elevated risk, but the confidence interval is so wide that it comfortably spans everything from a strong protective effect to a more than fourfold increase in risk. In plain terms, the data cannot tell readers which direction, if any, the true effect points.

The null findings extended across every secondary outcome the team examined. There was no significant association between SGLT2 inhibitor use and the worst T-score recorded for each patient, nor with the site-specific T-scores at the lumbar spine or femoral neck. Actual bone mineral density measurements at both anatomical sites likewise showed no meaningful differences between users and non-users of the drug class. Even when patients were sorted into broader bone mineral density risk categories, normal, osteopenic, or osteoporotic, based on the worse of their two T-scores, the distribution did not differ significantly between the groups. A supplementary analysis incorporating the non-diabetic comparison group, evaluated with a Fisher–Freeman–Halton exact test, found no significant difference in bone mineral density category distribution across all three groups, with a p-value of 0.809. For a drug class that has been shadowed by skeletal concerns, the consistency of the null results across so many related endpoints is itself informative.

Why, then, did the skeletal question arise in the first place? The biological plausibility of harm has never been trivial. SGLT2 inhibitors induce a mild state of glucosuria, and the resulting caloric and fluid shifts alter phosphate handling, insulin secretion, and parathyroid hormone dynamics, all of which participate in the delicate regulation of bone remodeling. Some earlier observational studies and post-marketing analyses raised flags about fracture risk, particularly with canagliflozin, prompting regulators to add warnings to drug labels. At the same time, type 2 diabetes itself complicates the picture in a way that standard DEXA measurements can underappreciate. People with diabetes can sustain fragility fractures at bone densities that would be considered safe in the general population, a phenomenon thought to reflect impaired bone quality, altered collagen cross-linking, and the added danger of hypoglycemia-related falls. Any study of diabetes drugs and bone health therefore has to contend with a disease process that already distorts the relationship between measured density and actual fracture risk.

The authors of the new study are careful, and appropriately so, about what their results can and cannot support. In their conclusion they state that no statistically significant association was detected between SGLT2 inhibitor use and DEXA-derived bone health outcomes in patients with type 2 diabetes, but they immediately qualify this by noting that the limited sample size and the wide confidence intervals preclude excluding clinically meaningful associations. They call for larger prospective studies to settle the question. This is the honest language of exploratory research: a study of 56 patients simply lacks the statistical power to detect anything but a very large effect. With confidence intervals as broad as those reported here, a true doubling of osteoporosis risk could hide within the data undetected. Readers and clinicians should therefore interpret the findings as absence of evidence rather than evidence of absence, a distinction that matters when treatment decisions hang in the balance.

The study’s design also carries limitations worth understanding. As an observational, cross-sectional analysis conducted at a single hospital department, it captured patients at one moment in time rather than following them as bone density evolved. Patients were not randomized to receive SGLT2 inhibitors, so the possibility of confounding by indication, where the reasons a clinician chose a particular drug correlate with other health factors, cannot be fully excluded even with adjustment for age, sex, and body mass index. The duration of SGLT2 inhibitor exposure among users was not a central variable in the reported analysis, yet skeletal effects of any drug typically accumulate over years. The inclusion of a non-diabetic comparison group adds descriptive context, but the core inference rests on the 56 patients with diabetes divided into two modest groups. Ethics approval was granted by the Institutional Ethics Committee of Bursa Yüksek İhtisas Training and Research Hospital, and written informed consent was obtained from all participants in accordance with the Declaration of Helsinki, underscoring that the study met standard procedural safeguards even within its constrained scope.

For the millions of people taking SGLT2 inhibitors worldwide, the practical message is one of cautious reassurance rather than definitive clearance. The study found osteoporosis to be strikingly common in this population regardless of medication, affecting roughly a third of patients in both groups, which reinforces the importance of bone health screening for people with type 2 diabetes independent of their glucose-lowering regimen. Clinicians already weigh fracture risk when selecting diabetes therapies, and these findings do not demand a change in that calculus. What the research does provide is a template for the kind of work that must come next: large, prospective cohorts with longitudinal DEXA monitoring, careful documentation of drug duration and dose, and ideally fracture outcomes rather than surrogate density measures alone. Until such studies arrive, the skeleton’s verdict on SGLT2 inhibitors remains open, but the scales have tilted slightly, and measurably, toward the reassuring side.

Subject of Research: The association between SGLT2 inhibitor use and DEXA-derived bone health parameters in patients with type 2 diabetes mellitus

Article Title: SGLT2 inhibitor use and DEXA-derived bone health in type 2 diabetes mellitus

Article References: Şahin, M. F., Şahin, A., Dinar, A. B. K., Avul, S., İzeet, S. N., & Erol, A. (2026). SGLT2 inhibitor use and DEXA-derived bone health in type 2 diabetes mellitus. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02633-z

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02633-z

Keywords: SGLT2 inhibitors, type 2 diabetes, osteoporosis, bone mineral density, DEXA, T-score, BMC Endocrine Disorders, observational study, endocrinology, fragility fractures, glucosuria, bone health