Blood Test May Predict Dangerous Blood Sugar Swings in Young Diabetics

Children and young adults living with diabetes in Bangladesh experience blood sugar swings far more extreme than international guidelines consider safe, and a simple, inexpensive blood test may help identify those at greatest risk, according to a new prospective study conducted at a national paediatric diabetes centre in Dhaka. The research, published in Health Science Reports, offers the first systematic portrait of glycaemic variability in a mixed group of young people with Type 1 and Type 2 diabetes in South Asia, a region where continuous glucose monitoring is only beginning to reach clinics and where most published benchmarks come from wealthy countries with very different resources.

The study followed 97 children, adolescents, and young adults aged 1 to 25 years who attended routine follow-up at the BADAS Paediatric Diabetes Care and Research Centre at BIRDEM in Dhaka between March and November 2024. Each participant wore a blinded professional continuous glucose monitor, the Abbott FreeStyle Libre Pro iQ, on the upper arm for 14 days. Unlike the consumer devices increasingly common in high-income countries, the blinded sensor recorded interstitial glucose every 15 minutes without displaying any readings to the wearer. That design choice matters scientifically: when patients can see their glucose in real time, they tend to correct with extra insulin, adjust meals, or change activity, which artificially smooths the very fluctuations researchers want to measure. By keeping the sensor invisible, the team captured a less confounded estimate of true day-to-day glycaemic instability.

The central measure was the coefficient of variation, calculated as the standard deviation of glucose readings divided by the mean glucose and expressed as a percentage. International consensus guidelines, including those from the International Society for Paediatric and Adolescent Diabetes, recommend keeping this value at or below 36 percent, because higher variability is linked to more hypoglycaemic episodes, particularly dangerous overnight drops in blood sugar. In the Dhaka cohort, the mean coefficient of variation was 39.4 percent, with individual values ranging from 19.3 to 59.1 percent. Nearly 61 percent of participants exceeded the recommended threshold, meaning suboptimal variability was the norm rather than the exception. That mean is substantially above the 32 to 37 percent range typically reported in well-resourced paediatric cohorts, and because those comparison studies mostly used real-time devices that tend to lower observed variability, the true gap is likely even wider.

The contrast between diabetes types was stark. Participants with Type 1 diabetes, who made up just over 70 percent of the cohort, had a mean coefficient of variation of 41.9 percent, compared with 33.3 percent among those with Type 2 diabetes, a difference of 8.6 percentage points. More than three-quarters of the Type 1 group exceeded the variability target, against roughly a quarter of the Type 2 group. Time-in-range, the proportion of readings between 70 and 180 milligrams per decilitre, averaged just 52.5 percent in the Type 1 group versus 70.3 percent in the Type 2 group. Notably, time spent below range, which captures hypoglycaemia, was statistically similar between the two groups, at around 10 to 11 percent each, indicating that dangerous low blood sugar is not a concern confined to Type 1 diabetes in this population.

The study’s most intriguing finding concerns C-peptide, a byproduct of insulin production that serves as a marker of residual beta-cell function. When insulin-producing cells in the pancreas are destroyed, as in Type 1 diabetes, C-peptide falls to undetectable levels; when they survive, as in Type 2 diabetes or early Type 1 disease, measurable amounts persist. Fasting C-peptide was available for 46 of the 97 participants, and its relationship with glucose stability was striking. Higher C-peptide correlated strongly with lower variability, and in a multivariable regression adjusting for diabetes type, age, and insulin dose, each additional nanogram per millilitre of C-peptide was associated with a 1.94 percentage point reduction in the coefficient of variation.

Stratifying participants into three C-peptide categories revealed a clear dose-response pattern. The 20 participants with near-complete beta-cell depletion, defined as C-peptide below 0.2 nanograms per millilitre, had a mean coefficient of variation of 44.8 percent, time-in-range of only 42.1 percent, and time below range of 13.6 percent. The 13 participants with the best-preserved beta-cell function, above 1.0 nanograms per millilitre, achieved a mean variability of 32.2 percent and time-in-range of 65.8 percent, approaching international targets despite having no access to advanced insulin delivery technology. Time below range was nearly doubled in the lowest category compared with the highest, showing that hypoglycaemia risk, not just hyperglycaemia, tracks inversely with residual insulin secretion. The statistical trend across categories was significant for variability, time-in-range, time below range, and mean glucose.

The biology behind this relationship is coherent. Preserved endogenous insulin secretion provides meal-responsive, physiologically titrated insulin delivery that no injection regimen fully replicates, buffering the post-meal spikes that drive variability upward. Intact beta cells also support counter-regulatory glucagon responses that restrain hypoglycaemic swings. What makes the finding clinically important is its practicality: a single fasting C-peptide measurement costs roughly 5 to 10 US dollars in Bangladesh and can be run in ordinary hospital biochemistry laboratories, without the stimulation protocols or antibody assays that are often unavailable in low-resource settings. If confirmed in larger samples, the authors suggest, the test could help clinicians identify children who need prioritised access to glucose monitoring, more frequent clinical contact, or structured hypoglycaemia counselling.

The researchers are careful to frame the C-peptide results as hypothesis-generating. Because diabetes type was itself classified partly using fasting C-peptide, the two variables are partially collinear, and the association is best interpreted as a gradient of residual beta-cell function rather than a fully independent effect. Sensitivity analyses restricted to participants with Type 1 diabetes alone, and to those with at least 10 days of valid sensor data, produced consistent results, but C-peptide was missing for more than half the cohort, and although those with and without measurements differed on no measured baseline characteristic, residual selection bias cannot be excluded. The study was also single-centre, conducted at a tertiary referral facility where patients receive specialist care and may not represent all children with diabetes in Bangladesh. The 14-day monitoring window cannot capture seasonal or illness-related patterns, and the cross-sectional design precludes conclusions about cause and effect.

Even so, the findings carry weight for a region where diabetes in young people is rising and monitoring technology remains scarce and unevenly distributed. The authors point out that the elevated variability likely reflects a combination of resource-limited care realities: limited access to rapid-acting and long-acting analogue insulins, inconsistent meal timing, infrequent clinical contact, and practical barriers to rapid dose adjustment. Notably, even the Type 2 subgroup, which retained meaningful insulin secretion, showed variability exceeding values reported in comparably managed high-income country cohorts, suggesting the resource environment imposes a burden that operates across diabetes types. The team calls for multicentre studies across South Asia and sub-Saharan Africa to establish regionally appropriate benchmarks, for intervention trials testing whether variability-focused protocols reduce both high and low glucose excursions, and for cost-effectiveness analyses comparing professional glucose monitoring against standard HbA1c-only care. In the meantime, they argue, reducing variability, not merely lowering average HbA1c, should become an explicit treatment goal for young people with diabetes of both types in resource-limited settings.

Subject of Research: Glycaemic variability and residual beta-cell function in young people with Type 1 and Type 2 diabetes in Bangladesh

Article Title: Glycaemic Variability and C‐Peptide in Bangladeshi Children, Adolescents, and Young Adults With Diabetes: A Prospective Study

Article References: Flabe, A. H., Zabeen, B., Amirat, S., & Luo, J. (2026). Glycaemic Variability and C‐Peptide in Bangladeshi Children, Adolescents, and Young Adults With Diabetes: A Prospective Study. Endocrinology, Diabetes & Metabolism, 9(6), Article e70347. https://doi.org/10.1002/edm2.70347

Image Credits: AI Generated

DOI: 10.1002/edm2.70347

Keywords: continuous glucose monitoring, glycaemic variability, C-peptide, Type 1 diabetes, Type 2 diabetes, paediatric diabetes, hypoglycaemia, time-in-range, Bangladesh, low- and middle-income countries, beta-cell function, HbA1c