Insulin Timing Matters: Results from the T1D Exchange Clinic Registry


I recently attended the annual T1D Exchange clinic investigators meeting and was lucky enough to attend several research presentations. One particularly interesting presentation, summarizing work by Dr. Karishma Datye, MD, MSCI at Vanderbilt University Medical Center, focused on the timing of insulin around meals and its relation to health measures. While GluJill nicely summarized these study findings (among others!) previously on Glu, we decided to take a deeper dive into this important research.

The study looked at the relationship between insulin timing and health outcomes. Currently, there are no official clinical guidelines in pediatrics regarding whether it is best to bolus for a meal before, during, or after eating. It is common practice to administer insulin before a meal, because then it can start metabolizing immediately and be “on-board” for the incoming food. However, some patients and/or their parents are reluctant to give insulin before a meal because if they don’t end up eating the meal, they may have a low blood sugar.

This analysis included data from 12,996 registry participants under the age of 26 who have had type 1 diabetes for at least one year. About 60% of these children and young adults used insulin pumps. Participants reported their typical timing of fast-acting insulin around a meal. While many survey respondents report pre-bolusing, a significant group reported that they usually give insulin during or after eating.

PreBolus Article Graph1

In this group, Dr. Datye found that:

  • Post-meal insulin delivery is correlated with higher HbA1c levels.

– In children under 6, those whose parents gave insulin several minutes before the meal had an average HbA1c of 7.8%, whereas those who received insulin after eating had a mean HbA1c of 8.6%

– The same pattern is seen in other age groups; the average HbA1c in those who pre-bolus is lower than those who receive insulin during or after eating;

  • Patients who gave insulin during or after the meal were more likely to miss at least one insulin bolus during the week
  • Pre-bolusing for meals was not associated with severe hypoglycemic events

In sum, there are measurable benefits related to pre-bolusing before eating, and patients that pre-bolus do not report higher rates of hypoglycemia.

PreBolus Image2


We spoke with Dr. Datye to gain a deeper understanding of these study results.

GluDanielle: How do your experiences as a clinician relate to this research/what made you interested in this topic?

Dr. Datye: I care for many adolescents with type 1 diabetes, and I realized that many of my patients were checking their blood sugars before eating, eating their meals, and then running off to class, sports practice, or other important activities and forgetting to give their meal insulin. I wondered if giving insulin before a meal might make it easier to remember meal insulin. Giving insulin before a meal also makes sense because it takes 15-20 minutes for the insulin to start working, so giving it before a meal helps the insulin start to work as a meal is being eaten. At the time I did not have evidence that there might be an association between timing of meal insulin and missed boluses, so I asked the T1D Exchange if they could help me with this idea.

GluDanielle: It’s always great to have clinicians utilizing the T1D Exchange clinic registry data! So, in your study you found a wide range of percentage of those who administer insulin before meals across multiple clinics- participants at some clinics report pre-bolusing on average on 20% of the time, whereas pre-bolusing in participants at other locations is much more common (around 80% of the time). Why do you think there is such a wide range?

Dr. Datye: I think these data are really interesting! We looked at different clinical centers that are part of the T1D Exchange Clinic Network to see how participants at different sites report giving insulin. I was surprised to see that at some clinics, participants report giving insulin before a meal 20% of the time, while at other clinics, participants report giving insulin before a meal 80% of the time; however, I thought it was really important that all centers have many patients that post-meal bolus and will benefit from these findings. I think some of this variation may be due to age of the participants at particular centers, or perhaps differences in how centers teach patients to give insulin. I am eager to explore these data further-it may be helpful to know how different centers teach their patients to give meal insulin, whether the teaching changes if the patient is a toddler versus an adolescent vs a young adult, and how parents and patients decide when to give meal insulin.

GluDanielle: Thank you for speaking about your work with T1D Exchange data!


I find these results to be particularly compelling, because strategies to improve BG could be helpful to the community overall. This study focused on children, adolescents, and young adults, and these groups all manage their BG differently; for example, parents often manage their children’s BG to a large extent, whereas adolescents are learning independence. However, the same results are found throughout- better timing of insulin is associated with improved HbA1c levels and fewer missed boluses.

Glu, what do you think about this study? Do you give insulin (to yourself or your child with T1D) before eating? Why or why not? Has your regimen changed over time?

-GluDanielle (Danielle Gianferante)

Sign in or Register to view comments.