Fasting on Insulin: Landmark Meta-Analysis Finds Ramadan Is Safer for Type 2 Diabetes Than Feared

Every year, hundreds of millions of Muslims around the world abstain from food and drink from dawn to sunset during the holy month of Ramadan. For the roughly 150 million of them who live with diabetes, this annual spiritual obligation has long been framed as a medical dilemma, particularly for those who depend on insulin, the hormone whose dosing is built around meals. Skipping meals while injecting a glucose-lowering drug is, on paper, a recipe for dangerously low blood sugar. Yet a new systematic review and meta-analysis published in BMC Endocrine Disorders suggests that the reality may be far less alarming than the textbook logic implies, offering the most comprehensive quantitative picture to date of what actually happens to insulin-treated patients with type 2 diabetes when they fast.

The research team, led by Hoda Gad of the Lunenfeld-Tanenbaum Research Institute at Mount Sinai Hospital in Toronto together with clinicians from Hamad Medical Corporation and Weill Cornell Medicine-Qatar in Doha, set out to answer two deceptively simple questions. First, does glycemic control, measured by HbA1c, the blood marker that reflects average glucose levels over the preceding two to three months, improve or deteriorate during Ramadan? Second, does fasting on insulin increase the risk of hypoglycemia, the acute drop in blood glucose that remains the most feared complication of diabetes treatment? To find out, the investigators followed the PRISMA reporting guidelines and systematically searched PubMed, Embase and the Cochrane Library from their inception through January 2025, casting a deliberately wide net that captured both randomized controlled trials and observational cohort studies.

That search yielded twenty studies encompassing a combined 2,024 patients assessed before Ramadan and 1,979 assessed during or after the fasting month. The regimens represented in this pooled population were strikingly heterogeneous, mirroring real-world clinical practice: some patients took long-acting basal insulin alone, others used premixed formulations that combine basal and rapid-acting components in a single injection, others followed intensive basal-bolus schedules, and many combined insulin with oral or injectable glucose-lowering agents such as sulfonylureas, sodium-glucose co-transporter-2 inhibitors or glucagon-like peptide-1 receptor agonists. This diversity is both a strength and a limitation, because it reflects the choices clinicians actually make, but it also complicates any attempt to isolate the effect of a single dosing strategy.

The headline finding concerns HbA1c. When the researchers pooled the randomized controlled trials separately using a random-effects model, they found a statistically significant reduction in HbA1c during Ramadan, with a mean difference of minus 0.38 percentage points and a 95 percent confidence interval spanning minus 0.49 to minus 0.27, a result that was highly significant at p less than 0.00001 with modest heterogeneity, indicated by an I-squared statistic of 35 percent. In plain terms, patients in the trials tended to finish the fasting month with better long-term glucose control than they had going in. Observational studies told a similar story, showing a slightly smaller but still significant reduction of minus 0.33 percentage points, though with substantially higher heterogeneity at I-squared 82 percent, a signal that real-world populations, dosing adjustments and measurement practices varied considerably across studies.

The authors did not stop at the primary pooled estimates. Sensitivity analyses using fixed-effect models were run to test whether the conclusions depended on the statistical assumptions of the random-effects approach, and they confirmed the robustness of the findings, with statistical significance persisting and heterogeneity falling. This kind of internal consistency matters in meta-analysis, where pooled results can sometimes be artifacts of how small studies are weighted. Here, the direction and magnitude of the HbA1c effect survived alternative modeling, lending credibility to the central claim that insulin-containing regimens during Ramadan are associated with modest improvements, not deterioration, in glycemic control.

The hypoglycemia story is more nuanced and more cautiously told. Across most of the included studies, the incidence of hypoglycemia was either stable or reduced during Ramadan compared with the pre-fasting period, a finding that runs counter to the intuitive expectation that skipped meals plus insulin should produce more low-glucose events. In the formal pooled analysis of the five studies that reported the actual number of hypoglycemic events, mixing randomized and observational designs, the standardized mean difference was minus 0.06 with a 95 percent confidence interval of minus 0.25 to 0.13 and a p-value of 0.52, meaning there was no statistically significant difference between the fasting and non-fasting periods. Heterogeneity was essentially zero, with an I-squared of 0 percent, suggesting remarkable agreement among the studies that could be pooled.

Why the authors urge caution, however, is that hypoglycemia is notoriously difficult to compare across studies. There is no single agreed definition: some trials counted any glucose reading below 3.9 millimoles per liter, others relied on patient-reported symptoms, and still others recorded only severe events requiring third-party assistance. Reporting methods varied just as widely, from continuous glucose monitoring to retrospective questionnaires. Because of this inconsistency, a subgroup analysis by study design was not feasible, and the reassuring pooled null result rests on only five studies. Absence of evidence for harm, the authors emphasize, is not the same as evidence of absence of harm, particularly when the underlying data collection was so uneven.

The clinical significance of the HbA1c improvement also deserves scrutiny. A drop of roughly 0.3 to 0.4 percentage points is modest by the standards of diabetes pharmacotherapy, where new drug classes often deliver reductions of 0.5 to 1.5 points. It is plausible that some of the improvement reflects a Hawthorne-like effect: patients who choose to fast may simultaneously pay closer attention to their diet, adhere more faithfully to their medication schedule and engage more actively with their care teams in the pre-Ramadan counseling that many diabetes clinics in Muslim-majority countries provide. Structured pre-Ramadan education, medication dose adjustment and glucose self-monitoring are cornerstones of the international guidance on diabetes and fasting, and the studies in this review were conducted in settings where such support is increasingly routine.

What the analysis does not resolve is the question of who should fast and who should not. Professional guidelines, including those from international diabetes federations, have long stratified patients by risk, generally advising against fasting for those with recurrent severe hypoglycemia, poor overall control, pregnancy or advanced complications. The twenty studies included here almost certainly enrolled a selected population, patients well enough and motivated enough to attempt the fast under medical supervision, which means the favorable safety signal cannot simply be extrapolated to frail or poorly controlled individuals. The authors are explicit on this point: individualized insulin management and structured pre-Ramadan education remain essential to optimize safety during fasting, and the pooled data should inform, not replace, that individualized conversation.

Nevertheless, the study arrives at a moment when the demographics of diabetes and the demographics of Islam are colliding at scale. The International Diabetes Federation estimates that the number of Muslims with diabetes who fast during Ramadan runs into the tens of millions each year, and that figure is projected to grow. For clinicians in the Gulf, South Asia, North Africa and increasingly in Western diaspora communities, the practical question is not whether patients will fast, because many will regardless of advice, but how to make fasting as safe as possible. By aggregating two decades of evidence into a single quantitative framework, this meta-analysis shifts the default narrative from one of prohibition and fear toward one of informed, monitored participation. The modest HbA1c improvement and the absence of a detectable rise in hypoglycemia together suggest that, with the right regimen adjustments and education, insulin-treated type 2 diabetes and Ramadan fasting can coexist far more safely than has often been assumed, while the gaps in hypoglycemia definitions and the small number of pooled studies mark out precisely where the next generation of trials should focus.

Subject of Research: Glycemic control and hypoglycemia risk in insulin-treated type 2 diabetes during Ramadan fasting

Article Title: Safety and glycaemic outcomes of insulin-containing regimens in patients with type 2 diabetes mellitus during Ramadan: a systematic review and meta-analysis

Article References: Gad, H., Hayat, T., Dabbous, Z., Toufaili, Z., Elgassim, E., Khalil, S. K., Zara, K., Bennabhaktula, S., Sattar, A., MacDonald, R., Malik, R. A., & Elhadd, T. (2026). Safety and glycaemic outcomes of insulin-containing regimens in patients with type 2 diabetes mellitus during Ramadan: a systematic review and meta-analysis. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02574-7

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02574-7

Keywords: Ramadan fasting, type 2 diabetes, insulin therapy, HbA1c, hypoglycemia, meta-analysis, systematic review, glycemic control, diabetes management, PRISMA, randomized controlled trials, pre-Ramadan education