Running with Diabetes: What Systematic Reviews Reveal

For decades, people living with diabetes were routinely discouraged from pushing their bodies to the extreme limits of endurance sport. The fear was understandable: prolonged exercise dramatically alters glucose metabolism, and for athletes who depend on externally administered insulin, the threat of a sudden, dangerous drop in blood sugar mid-race has always loomed large. But a comprehensive new systematic review, published in Sports Medicine – Open, suggests that those fears, while not entirely unfounded, can be managed with remarkable precision. Drawing together more than six decades of scattered case reports and clinical studies, the review concludes that athletes with diabetes — the overwhelming majority of them living with Type 1 diabetes — can safely complete marathons, ultra-marathons, and even multi-day endurance events while maintaining acceptable glycemic control, provided they plan carefully and monitor relentlessly.

The research team, led by Lorin Braschler of the University of Zurich alongside collaborators including Beat Knechtle, Mabliny Thuany, Thomas Züger, Pantelis T. Nikolaidis, Katja Weiss, and Thomas Rosemann, systematically searched seven major databases — PubMed, Medline via Ovid, Scopus, SPORTDiscus, the Cochrane Library, CINAHL, and Web of Science — for studies published through April 2024, with an updated search conducted on May 3rd, 2026. Their initial sweep identified 656 candidate studies, of which only 22 met the strict inclusion criteria. To qualify, studies had to feature runners with a confirmed diagnosis of diabetes mellitus participating in endurance events of at least half-marathon distance, including half-marathons, full marathons, ultra-marathons exceeding the classic 42.195-kilometer distance, and ultra-endurance competitions such as IRONMAN triathlons and multi-day races. Athletes with prediabetes, gestational diabetes, or diabetes following pancreatic islet-cell transplantation were excluded, as were animal and in vitro studies. The final dataset comprised 99 runners with diabetes, a striking 99 percent of whom had Type 1 diabetes and only one of whom had Type 2.

The picture that emerges from these 22 studies is one of deliberate, cautious hyperglycemia as a protective strategy. Rather than chasing the tight glucose targets recommended for sedentary patients, endurance runners with diabetes consistently allowed their blood glucose to run moderately high during competition — essentially building a safety buffer against exercise-induced hypoglycemia, the single greatest danger they face on the course. The weighted mean HbA1c across the pooled population, a measure of average blood glucose over the preceding two to three months, came out at 7.4 percent (95 percent confidence interval 6.9–8.1), slightly above the conventional clinical target of less than 7 percent. Interestingly, when the researchers separated studies by publication date, a clear technological trajectory appeared: studies published before 2010 reported a weighted mean HbA1c of 8.3 percent, while those published after 2010 reported 6.7 percent — a statistically significant difference that the authors attribute largely to the arrival of continuous glucose monitoring, insulin pumps, and automated insulin delivery systems.

Monitoring technology emerged as the defining feature of modern diabetic endurance performance. Half of all runners in the review — 50 percent — used continuous glucose monitoring systems, small sensor devices that read interstitial glucose every few minutes and stream the data to a watch or phone. Another 27.2 percent used insulin pumps, while 63.6 percent relied on multiple daily injections of insulin; some runners combined pump therapy with CGM in hybrid closed-loop configurations. The pattern of glucose readings during races told a consistent story across race distances. Time-in-range — the proportion of race time spent with glucose between 3.9 and 10.0 mmol/l — varied from 40 to 100 percent among half-marathoners, averaged 51.6 percent among marathoners, and ranged from 47 to 73 percent among ultra-marathoners. Critically, no symptomatic hypoglycemic events were recorded during any of the races included in the review, although athletes did spend substantial time in mild hyperglycemia, with some marathon runners exceeding 13.9 mmol/l for more than half the race when pre-race insulin was insufficiently reduced or when CGM signal interference disrupted automated insulin delivery.

Perhaps the most clinically consequential finding concerns what happens after the finish line, not during it. Late-onset hypoglycemia — episodes occurring roughly 6 to 15 hours after strenuous exercise, often overnight — is a well-documented hazard in Type 1 diabetes, driven by the heightened insulin sensitivity and sustained glucose uptake of muscles that are busily replenishing glycogen stores via GLUT-4 transporter translocation. Twenty-seven runners in the review were specifically monitored with CGM during the night and up to 48 hours after racing. Only a small number experienced these delayed dips, with reported time-below-range figures as low as 0.2 percent, translating to a relative incidence of roughly 0.7 to 2 percent. That is markedly lower than figures reported in professional cyclists with Type 1 diabetes, in whom post-exercise hypoglycemia has been measured at 6 to 12 percent of monitoring time. The authors suggest that differences in exercise intensity, insulin adjustment practices, and post-race carbohydrate intake may explain the gap, though they caution that data on runners remain sparse and more research is needed.

Underpinning all of this is insulin dosing strategy, and here the review found remarkable consensus. Among the 18 studies — 81.8 percent of the total — that reported insulin adjustments around race day, the single most common approach was a 50 percent reduction in basal insulin, taken either 24 hours before the race or on race morning, reported by five studies. Four studies documented even more aggressive reductions of 70 to 80 percent before competition. Other athletes kept their usual regimens unchanged, omitted basal insulin entirely on race day, or made modest reductions of 5 to 35 percent. Nutrition worked hand in hand with these adjustments: carbohydrate intake during half-marathons ranged from roughly 12.6 to 39.1 grams per hour, marathoners consumed 25 to 33.9 grams per hour, and ultra-distance athletes took in anywhere from about 10 to 60 grams per hour depending on the event, with an 82-kilometer mountain ultra-marathon runner consuming 33.9 grams per hour while a competitor in a 217-kilometer ultra managed only 10.4 grams per hour.

The physiological toll of these races, meanwhile, appears to hit diabetic and non-diabetic athletes in broadly similar ways. Biomarker data pooled from the included studies showed creatine kinase, an enzyme signaling skeletal muscle damage, spiking between 5.5- and 50.8-fold immediately after racing, peaking at 24 hours and returning to baseline within about a week. Stress hormones followed the expected pattern: serum cortisol rose 1.3- to 2.3-fold post-race before normalizing within 24 hours, and growth hormone increased roughly 1.8-fold. One intriguing finding from the underlying literature concerns glucagon, the counter-regulatory hormone that raises blood glucose. In one study, athletes with Type 1 diabetes showed significantly higher plasma glucagon levels than healthy controls despite similar blood glucose readings — possibly reflecting a compensatory mechanism, since without a functioning glucose-insulin feedback loop, an exaggerated glucagon response may be the primary remaining defense against exercise-induced hypoglycemia.

The authors are careful about what these data can and cannot show. With only 99 runners across 22 studies, most of them case reports and small cross-sectional analyses rather than controlled trials, the evidence base is thin, and the quality of individual studies varied. The review team used the Joanna Briggs Institute critical appraisal tools for cross-sectional studies and case reports, rating each study from very low to high quality based on eight standardized criteria. Nevertheless, the convergent picture across decades, countries, and race formats is compelling: with proper preparation, personalized glycemic targets agreed with endocrinologists and sports physicians, and multidisciplinary support, athletes with diabetes are not merely surviving endurance events — they are excelling in them. The review notes that individual runners with diabetes have completed 1,000 marathon races and 48 24-hour ultra-marathons, achievements that would have seemed fantastical to clinicians a generation ago.

What the review offers in practical terms is a roadmap. Its synthesis of the literature supports aiming for more than 70 percent time-in-range in daily life with less than 4 percent of time below 3.9 mmol/l, accepting moderate hyperglycemia during races as a deliberate trade-off, reducing basal insulin by 50 to 80 percent around competition, monitoring continuously with CGM during and after events, replenishing carbohydrate aggressively in the post-race window, and considering overnight basal insulin reductions of about 20 percent to blunt nocturnal hypoglycemia risk. The question, the authors argue, is no longer whether people with diabetes should run long distances — it is how they can do so safely. As wearable technology continues to improve and automated insulin delivery matures, that answer is likely to keep getting better, and the finishing lines once considered off-limits will keep getting closer.