Hidden Glycation Fingerprint in Blood May Signal Fragile Bones in Diabetes

People living with type 2 diabetes have long presented clinicians with a paradox: their bones often look strong on a standard scan, yet they break more easily than the bones of people without the disease. A new study from researchers at the Second Affiliated Hospital of Army Medical University in Chongqing, China, adds a striking piece to this puzzle. The team found that a simple blood-derived measure, the hemoglobin glycation index, was linked to a substantially higher prevalence of osteoporosis in the lumbar spine among hospitalized patients with type 2 diabetes. The work, published in BMC Endocrine Disorders, suggests that the way glucose attaches to hemoglobin inside red blood cells may carry clues about skeletal health that conventional measures of blood sugar alone do not reveal.

The hemoglobin glycation index, usually abbreviated HGI, is calculated by comparing the HbA1c value actually measured in a patient’s blood with the HbA1c value that would be predicted from that patient’s fasting plasma glucose. HbA1c reflects the percentage of hemoglobin molecules that have been chemically modified by glucose over the roughly two-to-three-month lifespan of red blood cells. Because individuals differ in how readily glucose binds to their hemoglobin, two patients with identical average glucose can show different HbA1c readings. The index captures this individual discrepancy: a positive value means the measured HbA1c runs higher than expected for the observed glucose, a phenomenon sometimes described as high glycators, while a negative value indicates the opposite tendency.

To investigate whether this glycation fingerprint matters for bone, the researchers retrospectively analyzed 757 hospitalized patients with type 2 diabetes who had undergone dual-energy X-ray absorptiometry, the standard technique for measuring bone mineral density, between January and December 2023. Dual-energy X-ray absorptiometry works by scanning the skeleton with X-ray beams at two different energy levels and calculating how much of each beam is absorbed by bone tissue, yielding a precise estimate of mineral content. Lumbar osteoporosis was defined using the T-score, which compares a patient’s bone density at the first through fourth lumbar vertebrae with the average density of a healthy young adult. A T-score of negative 2.5 or lower meets the conventional threshold for osteoporosis.

The results were unambiguous in direction. Overall, 208 of the 757 participants, or 27.5 percent, had lumbar osteoporosis. When the researchers divided the cohort into tertiles, three groups of equal size ranked by HGI, the prevalence of lumbar osteoporosis climbed steadily across the groups, from 15.8 percent in the lowest tertile to 28.6 percent in the middle tertile and 38.1 percent in the highest. In other words, patients whose hemoglobin showed the greatest degree of glycation relative to their measured glucose levels were more than twice as likely to have osteoporotic lumbar spines as those in the lowest group.

Because simple prevalence comparisons can be distorted by age, sex, body weight, kidney function and other confounders, the team turned to logistic regression, a statistical technique that models the probability of a binary outcome, here the presence or absence of osteoporosis, as a function of multiple explanatory variables. In the fully adjusted primary model, which included 687 participants with complete data, each one-unit increase in HGI was associated with 34 percent higher odds of lumbar osteoporosis, with an odds ratio of 1.342 and a 95 percent confidence interval running from 1.168 to 1.546, a result that was highly statistically significant. The confidence interval indicates that, given the data, the true effect very likely lies well above the null value of 1.0, meaning no association.

A critical question for any study involving HbA1c is whether the index is merely re-describing what HbA1c or fasting glucose already reveal. To address this, the researchers performed sensitivity analyses in which they additionally adjusted for HbA1c itself and, separately, for fasting plasma glucose. The association survived both tests. With additional adjustment for HbA1c, the odds ratio was 1.251, with a confidence interval of 1.039 to 1.507 and a P value of 0.018. With additional adjustment for fasting glucose, the odds ratio was 1.337, with a confidence interval of 1.164 to 1.540 and a P value below 0.001. This pattern suggests that HGI carries information about bone health that is not redundant with either component used to compute it.

The team also examined bone density as a continuous outcome rather than a binary diagnosis. Higher HGI was associated with a lower T-score across the first through fourth lumbar vertebrae, with a standardized coefficient of negative 0.134 and a confidence interval of negative 0.236 to negative 0.031, statistically significant at P equal to 0.011. Restricted cubic spline models, which allow the relationship between exposure and outcome to bend flexibly rather than being forced into a straight line, were used to probe for nonlinearity in the association. Together, these secondary analyses reinforced the primary finding that the glycation index tracks with progressively weaker lumbar bone density.

Perhaps the most intriguing results came from the exploratory analysis of bone turnover markers, molecules released into the bloodstream during the constant remodeling of skeletal tissue. Osteocalcin, a protein secreted by osteoblasts, the cells that build new bone, serves as a marker of bone formation. PINP, a fragment of type I collagen propeptide, marks the synthesis of new bone matrix, while beta-CTX, a collagen degradation product, reflects bone resorption by osteoclasts. In this cohort, higher HGI was inversely associated with osteocalcin, with a coefficient of negative 0.057 and a confidence interval of negative 0.103 to negative 0.011, statistically significant at P equal to 0.016. Associations with PINP and beta-CTX did not reach statistical significance. The authors are careful to characterize the osteocalcin finding as modest and exploratory, but it hints at a possible mechanism: chronic non-enzymatic glycation of bone matrix proteins, including osteocalcin itself, may impair bone quality even when the quantity of mineral looks adequate on a scan.

This mechanistic hypothesis fits a broader picture that bone scientists have been assembling for years. In diabetes, skeletal fragility is thought to arise less from low bone mass and more from degraded bone material properties. Glucose can attach to collagen and other matrix proteins without the help of enzymes, forming advanced glycation end-products that stiffen collagen fibers and make bone more brittle under stress, even though the tissue appears denser on absorptiometry. A hemoglobin glycation index that runs high may flag patients whose tissues are more exposed to this process, offering a window into bone quality that bone mineral density measurements alone cannot provide. The inverse link with osteocalcin, a formation marker, could indicate suppressed or altered bone-building activity in high-glycating patients, though the cross-sectional data cannot establish which comes first.

The researchers are equally clear about the limits of their work. Because the study is cross-sectional, capturing a single moment in time, it demonstrates association rather than causation; it cannot show whether a high glycation index precedes bone loss or merely accompanies it. The cohort consisted of hospitalized patients, a group that may differ systematically from the broader population of people with type 2 diabetes, and the authors note the absence of formal discrimination or reclassification analyses, meaning the study did not test whether adding HGI to existing models actually improves the prediction of osteoporosis beyond HbA1c. Still, the findings open a compelling line of inquiry for a disease in which fractures carry enormous costs in mobility, independence and mortality. If future longitudinal studies confirm that the hemoglobin glycation index identifies patients at skeletal risk that standard glucose metrics miss, this inexpensive calculation, derivable from two routine blood tests, could become a practical tool for flagging which of the hundreds of millions of people worldwide living with type 2 diabetes need their bones watched as closely as their blood sugar.

Subject of Research: Association between the hemoglobin glycation index and lumbar osteoporosis in patients with type 2 diabetes

Article Title: Association between hemoglobin glycation index and lumbar osteoporosis in patients with type 2 diabetes mellitus: a cross-sectional study with exploratory analysis of bone turnover markers

Article References: Chen, C., Yang, J., Jiang, X., Wang, Q., Wang, X., & Zhang, X. (2026). Association between hemoglobin glycation index and lumbar osteoporosis in patients with type 2 diabetes mellitus: a cross-sectional study with exploratory analysis of bone turnover markers. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02615-1

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02615-1

Keywords: hemoglobin glycation index, type 2 diabetes, osteoporosis, bone mineral density, HbA1c, lumbar spine T-score, bone turnover markers, osteocalcin, dual-energy X-ray absorptiometry, advanced glycation end-products, endocrinology, metabolic bone disease