#28 August 2016 Update on My T1D Management

I am pleased to announce the 2nd Annual Conference on Nutritional Ketosis and Metabolic Therapeutics to be held in Tampa, Florida, February 1-4, 2017. I would like to thank the organizers, Angela Poff, PhD and Dominic D’Agostino, PhD, for inviting me to speak on the “Management of Type 1 Diabetes with a Ketogenic Diet.” I think anyone interested in ketogenic diets would enjoy this conference. You can see some of the talks from the inaugural event at the Epigenix Foundation YouTube channel. You can register for the event here.

In August 2016, I achieved a personal best in glycemic control. Although glucose management in diabetes is subject to many factors that are difficult to quantify, I attribute this month’s results to my most recent exercise schedule. Starting the second week of August, I changed to an alternating schedule of weightlifting with swimming or cycling. This facilitated recovery from each activity and the daily exercise hopefully stabilized my insulin sensitivity and improved glycemic control. I hope this continues and results in some improvement in the weights I am able to lift. Currently, I’m at 135 lb. snatch and 175 lb. clean and jerk which is just short of my personal bests of 140 lb. and 185 lb., respectively. There has been no significant change in my low carb healthy fat ketogenic lifestyle.

The summary table of current and previous mean blood glucose (BG) and insulin doses is shown below. Compared to July 2016, my mean BG decreased from 101 to 95 mg/dl in part as a result of an increase in mean total daily insulin dose from 28.1 to 30.4 IU/day due to an increase in basal insulin (from a mean of 21.2 to 21.8 IU/day Lantus) and an increase in rapid-acting insulin (from a mean of 6.9 to 8.5 IU/day Humalog). I had 3 of 130 (2%) BG readings < 51 mg/dl, but all were without symptoms due to either nutritional ketosis or hypoglycemia unawareness (see blog post #12 for more details). I had 102 of 130 (79%) BG readings between 51 and 120 mg/dl, 24 of 130 (18%) between 121 and 200 mg/dl, and 1 of 130 (1%) > 200 mg/dl.

Post 28 Means Table

Below are my BG readings along with exercise type and time so you can see how the type and duration of exercise affected my glycemic control.

Post 28 Blood Glucose and Exercise

The table below shows the summary of current and previous BG variability results. Most of the results were improved compared to previous months and years with improvement in the following variability measures: standard deviation = 32, coefficient of variation = 34%, mean daily BG range = 66, mean time of hyperglycemia = 3.7 hr/day, mean time of hypoglycemia = 0.1 hr/day, interquartile BG range = 47. The measures of BG variability were defined and explained in blog post #10.

Post 28 Variability Table

The actual daily insulin doses and daily insulin dose totals are shown in the graphs below.

Post 28 Insulin Doses

The Ketonix breath ketone (acetone) monitor results beginning June 2015 are shown below (see blog post #19 for more details). I have been in continuous nutritional ketosis due to my low carbohydrate ketogenic diet. There is a noticeable gradual decline in the readings. I would guess this is related to a gradual increase in carbohydrate intake from nuts and seeds.

Post 28 Ketonix Breath Ketones

The graphs below show the change in BG that results from mealtime rapid-acting insulin. On the x-axis is the Breakfast and Dinner mealtime insulin dose (Humalog) plotted against the change in BG, i.e. Post meal BG minus Pre meal BG on the y-axis. Thus positive values represent an increase in BG and negative values represent a decrease in BG after the meal. Also, note that I removed 3 of 31 Breakfast points and 3 of 31 Dinner points from the graphs where the Post meal BG was not in the range of 51-120 mg/dl. The rationale for this is to eliminate graphing mealtime doses that were “incorrect” so to speak. This way the graph shows both that larger doses of insulin reduce BG more (obvious), but more importantly even when the resulting Post BG was in an acceptable range, that there is a wide variation in the amount of BG reduction for any given dose of insulin. This variation can be due to varying absorption of injected insulin, variation in insulin sensitivity from exercise, or variation in food consumed during the meal from one day to the next. Reducing this variation is an important goal of mine.

Post 28 Breakfast &amp; Dinner Post BG - Pre BG

I required quite a few doses of rapid-acting insulin (Humalog) to correct hyperglycemia this month. The resulting change in BG is shown in the graph below. Negative values represent a decrease in BG. Note that I removed 2 of 15 correction doses from the graphs where the Post meal BG was not in the range of 51-120 mg/dl. The rationale for this is to eliminate graphing correction doses that were “incorrect” so to speak. Again, the main feature is variability and unpredictability of the BG response to insulin injections. In contrast to the mealtime insulin doses, this may be an over interpretation. To make this statement, I would have to recheck the BG about 3 hr. after the correction dose. This occurred the minority of the time. Instead in most instances, I wait until my next “usual” time to check the BG. Thus, other factors can affect the subsequent BG result other than the correction dose of Humalog.

Post 28 Humalog Correction Doses

The graph below shows the change in BG due to weightlifting. On the y-axis, positive values represent an increase in BG due to weightlifting. There were thirteen positive values representing an increase in BG due to weightlifting, but the degree of increase declined through the month (red linear regression line). The main feature is that intense resistance exercise stimulates catabolic hormones (epinephrine, cortisol, glucagon) to release nutrients into the blood stream and to stimulate muscle contractions. In a person with T1DM who cannot make insulin to correct the rise in BG, this results in hyperglycemia as shown in the graph.

Post 28 Post BG - Pre BG due to weightlifting

Since I did more swimming this month I decided to graph the change in BG due to swimming. On the y-axis, positive values represent an increase in BG whereas negative values represent a decrease in BG due to swimming. There were four negative values representing a decrease and five positive values representing an increase in BG due to swimming. As I have mentioned before, changes in BG with exercise are not very predictable, i.e. sometimes up, sometimes down. It really keeps you on your toes so to speak. The red linear regression line indicates a general increase in BG with swimming during the month, but the changes in BG were rather modest.

Post 28 Post BG - Pre BG due to swimming

In September, I will continue weightlifting and aerobic exercise on alternate days. Thus I will continue weightlifting to 3.5 days/week and aerobic exercise to 3.5 days/week. The olympic weightlifting consists of only 4 exercises (snatch, clean and jerk, snatch overhead squat, front squat) which takes about 2 hours to complete including warmup, cool down, and stretching to maintain mobility. The aerobic exercise consists of swimming or cycling at low intensity for 0.5 – 2 hours. The goal is to exercise consistently to maintain insulin sensitivity while avoiding injury and overtraining.

The goal of glycemic management in type 1 diabetes is to keep BG as close to normal (83 mg/dl) as is safely possible (i.e. avoiding hypoglycemia) to avoid both diabetic complications and a reduction in lifespan. For me, a nutrient-dense whole food low carbohydrate ketogenic diet (see blog post #9 for more details) combined with insulin analogs (Humalog/Lantus) have been the best tools so far in accomplishing this goal. I also feel, but cannot prove, that this eating plan and the resulting nutritional ketosis reduces the symptoms of hypoglycemia and protects the brain from the consequences of mild degrees of hypoglycemia (see blog post #12 for more details). Exercise with its resulting varying insulin sensitivity and hormonal changes actually makes glycemic management more difficult, but I both enjoy it and feel exercise has other health and lifespan extending benefits. Hopefully, my BG values and insulin doses are close enough to normal to avoid both a reduction in lifespan and diabetic complications. Only time will tell.

Till next time ….


      • Daisy

        ok thanks I contacted them 🙂 I apologize I have not read every post (as it seems like there are a ton of graphics and lost of information per post, which is not a bad thing just dont want to rush through everything and not take everything in) but a few questions: what is low intensity to you for when you do aerobic training? Do you follow something like the MAF method as you are ketogenic and trying to utilize fat not glucose as your fuel source (https://philmaffetone.com/maf-test/ ) ? Do you use a dexcom CGM to track the BGs?


      • Keith Runyan, MD

        I do use the MAF method for cycling and there seems to be a good correlation between heart rate and perceived exertion. Therefore, I use perceived exertion while swimming. That said, I seem to have different blood glucose responses to different types of exercise (cycling vs swimming vs walking vs running, etc.) despite similar degrees of perceived exertion. Exercising at a MAF heart rate doesn’t necessarily mean you will be burning just fat, nor does it mean it will necessarily prevent hypoglycemia during exercise, but it may make it less likely.
        I don’t use a CGM, but if others like them and derive benefit, they should use them.


  1. Daisy

    I agree with the different BG responses on different activities, same here. I apologize maybe I dont understand MAF appropriately then, cause I thought it was to ensure or strongly utilize fat and not sugar for physical activity. May I ask why you use the MAF method for cycling, is it for BG control or endurance if it is not used for this?
    I only thought the CGM would be a great tool for for information and statistics as it shows trends rather then single points in time. I hope this is ok to be asking these questions on here, if it would be more appropriate to have a PM conversation please let me know. I am not looking for MD advise, I am always just looking at what other keto type 1 do and are successful with 🙂 I am already a member of type1grit, optimisingnutrition etc. but always interesting to get an MD perspective and an MD who has chosen to do what works for him despite the advise of the ADA etc. is amazing!


    • Keith Runyan, MD

      MAF is designed to help carb-fueled athletes train at a lower intensity so as to start burning more fat. Eating a low-carb diet will aid fat-burning further. In type 1 diabetes (T1D), blood glucose (BG) regulation is altered by factors not present in non-diabetics. The normal response to low to moderate intensity exercise is a reduction in insulin secretion. This facilitates increased glucose production by the liver and release of fatty acids by fat cells to supply exercising muscles with fuel. In T1D, injected insulin continues to be released and can lead to hypoglycemia. An insulin pump if turned down or off prior to exercise could serve this role. In a person with T1DM using injected insulin, hypoglycemia is more likely when the intensity exceeds the MAF heart rate for extended periods because the muscles rely more on glucose via glycolysis to produce ATP and less so on oxidative phosphorylation from fatty acids and ketones. An all-out sprint is different still. During extremely intense exercise, catabolic hormones are released by the alpha-cells (glucagon), adrenal glands (cortisol), and sympathetic nervous system (epinephrine/norepinephrine) which stimulate the liver to make glucose and this results in mild hyperglycemia in non-diabetics and moderate or more severe hyperglycemia in T1DM.
      When I do MAF-type cycling rather than alter my injected insulin doses, I choose to eat small amounts of glucose (4 grams every 5 miles) to prevent hypoglycemia. I take my glucose meter and check BG every 2 hours and adjust the glucose eating if needed. A CGM would be handy in this scenario, but I have gotten better and better results over the years without one. So your understanding of MAF is essentially correct, but I doesn’t exactly translate to T1D due to reasons above. Thus, I use MAF to keep my glucose intake to a minimum. I also use cycling and swimming to recover from weightlifting (my intense exercise), but still remain insulin sensitive. My insulin sensitivity starts declining within 24 hrs after my last exercise session. I know this because if I don’t increase insulin doses by dinner-time of a rest day, I will develop hyperglycemia. I hope that clarifies things for you. If you use a CGM during exercise, I would interested in hearing about your experience.


      • Daisy

        Thanks so much for explaining that further for me, much appreciated. I always find it hard as type1 (no insulin) and type 2 diabetes (too much insulin) are complete opposite diseases however everyone uses them interchangeably and makes it hard to find something that would work for me. I was thinking of using the MAF approach for my cardio ( I also do HIIT training with easier spinning or long run sessions for recovery days) to reduce the hypos but I have an insulin pump and already utilize it with reduced basal and bolus depending on the day and what I am eating and doing. I have recently tried to do both HIIT and cardio sessions with no IOB and just have my normal coffee with some added MCT oil to last me until lunch, but that has had some unexpected results that I am still trying to figure out what the trends are and how to dose for it. I do have a CGM its a dexcom, if you would like to chat privately let me know (i think my private email is attached to the messages) 🙂 thanks again for explaining some things, you have a nice way of making it understandable 🙂