#66 May 2019 Update on My T1D Management

This is a monthly update on my glycemic management of type 1 diabetes (T1DM) using Humalog and Lantus insulin injections with resistance exercise and a ketogenic whole-food diet as described in my book, The Ketogenic Diet for Type 1 Diabetes also available on Amazon in print. My other book, Conquer Type 2 Diabetes with a Ketogenic Diet, is also available on Amazon in print. I would appreciate anyone who has read and benefited from either of these books to leave a review on Amazon. The number and ratings of the reviews are used by Amazon to order the search results when people are looking for books on diabetes.

I now also offer online coaching for those who need some individual attention in optimizing their glycemic control for either type 1 or type 2 diabetes and/or weight management. See the Coaching page for more info.

May 2019 was an interesting month with more changes. I now believe I have tried every possible combination of Lantus dosing. I have given it at 4 different times of the day 7 AM, 12 PM, 6 PM, and 11 PM once daily. I have given it twice daily and three times daily. I have changed the doses up and down the entire spectrum from A to Z. This month I changed my BG check times to 7 AM, 12 PM, 6 PM, and 10 PM. I also decided to change my insulin-dosing strategy and add a fourth meal daily. The reason for the fourth meal was two-fold. First, based on research in the sports science field that I mentioned last month which shows that muscle protein synthesis (MPS) is maximized by consuming 0.4 – 0.55 grams of protein/kg/meal in each meal. This stimulation of MPS is not as powerful as that produced by resistance exercise, but the benefits of exercise on MPS is also dependent on eating enough protein within the next 3-4 hours. Eating more protein than 0.55 g/kg/meal does not further increase MPS. Each meal results in net MPS over and above the muscle protein breakdown that occurs during fasting between meals and especially fasting overnight. Adding the fourth meal at bedtime potentially results in more net MPS per day. This is a hypothesis rather than a proven fact. Understand that the number of studies in humans is small and the ones that do exist are all short-term studies. Most of the studies on MPS were done in animals, primarily rats, and they too are short-term studies. However my decision to add the fourth meal was not just related to this potential, but uncertain, benefit on MPS. My second purpose was to test a different insulin-dosing strategy. The idea was to take my basal insulin (Lantus) dose with the bedtime meal, now 10 PM, such that I inject a dose of Humalog with each of the four meals daily. This means taking a total of five injections daily and no more. The Humalog dose would now have three purposes. 1) cover the meal, 2) correct for either high or low BG readings, 3) supplement my basal insulin needs for 20 of the 24 hours per day. The 20 hours represents the duration of action of Humalog of 5 hours x 4 doses daily. This last purpose is the new part. I am specifically choosing the basal insulin (Lantus) dose to cover my insulin needs between 3 AM and 7 AM, the time period where all Humalog is done. The hope is that the Lantus dose required to accomplish this will be low enough to never cause hypoglycemia at any other time of the 24-hour day. Another way of stating this is that I hope I can find a Lantus dose that is low enough that I always have to take greater than or equal to 0.4 IU of Humalog with each meal. If the Humalog dose at a meal is less than 0.4 IU, then the accuracy of drawing up my 5-fold diluted Humalog is questionable. I thing this approach makes sense because rapid-acting insulin doses can be precisely adjusted daily whereas any change in a Lantus dose takes 3-5 days to take full-effect. This method does not require eating 4 meals per day. If one chooses 3 meals per day, then Humalog could be given at 10 PM with the Lantus without eating a meal. I figured why not use the Humalog for both of the purposes stated above. I started this new regime on May 14 with 7 IU Lantus at 10 PM. I had to subsequently increase to 8, 9, 10, 11 IU, and back down to 10 IU. I hope that 10 IU at 10 PM will do the trick from this point forward. My BG values started looking pretty good on May 24th and the two lows I had subsequently is what prompted the reduction in Lantus from 11 IU to 10 IU.

This month I reached my goal weight of 67 kg at 2000 kcal/day, but today, June 1, I got hungry and fatigued again and increased my caloric intake back to 2100 kcal/day. Hopefully my weight will stay close to 67 kg so that when it comes time to compete in olympic weightlifting, I won’t need to cut much weight. An interesting observation I have had with this body weight adjustment process is that the symptoms of insufficient caloric intake is delayed. In other words, I can feel fine on the caloric intake that leads to the weight loss for several weeks before any symptoms develop. This sounds like the observation noted by overweight persons who lose weight. They find it is easier to lose the weight than to maintain the weight loss. This might be why. I think that covers all the new developments in May 2019.

Glycemic Results For May 2019

The table below shows my mean blood glucose (BG), standard deviation (SD), coefficient of variation (CV), body weight, and mean insulin dose totals for May 2019. I experienced a slight reduction in insulin doses, mean BG, SD, and CV compared to April.

The table below shows the percentage of BG values in the indicated ranges of low, goal, and high values for May 2019. Two of the three values were improved compared to April.

The graphs below show all of the daily insulin dose totals and all of the BG readings for May 2019. HUM = Humalog, LAN = Lantus, INS = total daily insulin dose. Note: I accidentally forgot to take a Lantus Dose on May 8th. I attributed that to a “senior moment” and to my frequent changes in dosing schedules. I don’t expect that to continue.

In June, I will continue with the above plan. I do not have plans for any more experiments. I think I need to give the current plan some time to settle out and see if I can get normal blood sugars in June.

I found a new study I plan to review next month regarding measuring interstitial glucose in nondiabetic subjects. It was published in 2009, so I don’t understand why it took me so long to come across it, but I guess better late than never. It is a better study than any of the others I have come across on this topic and involved 434 metabolically healthy nondiabetic subjects.

Finally, I would appreciate your comments on the idea of using inhaled insulin for the sole purpose of eating candy “as a treat.” I assume that means not very often, but I’m not sure how often that is. For me, I’m afraid that would send me down a path of having “treats” more and more often. I believe I was once addicted to chocolate and sweets in general. My low-carbohydrate diet fixed that, thank goodness. Also, not sure if this matters, but the person doing this is an endocrinologist with T1DM who states he/she follows a “not very strict low-carbohydrate diet.” To me, this means he/she understands the purpose of the low-carbohydrate diet. To me, the purpose of eating carbohydrates is to get nutrients from plant foods that are difficult to get, or are not available, in animal foods. The nutrients in my diet that come almost exclusively from plants include vitamin C, vitamin K1, manganese, lutein, and zeaxanthin. The nutrients from plants in my diet that make up at least 40% of my daily requirements include folate, vitamin E, copper, magnesium, and potassium. I would be interested in hearing your thoughts on the idea of eating candy as a “treat” with T1DM.

Well that is all I have for this month. Wishing you smooth blood sugars in June 2019.

Published by Keith Runyan, MD

I'm a physician with type 1 diabetes who uses a ketogenic diet and exercise to manage my diabetes. The purpose of the blog is to describe the lifestyle changes I made that have resulted in near-normal blood glucose, reduction in hypoglycemic events and symptoms, and reduction in insulin doses.

10 replies on “#66 May 2019 Update on My T1D Management”

  1. Keith – I’m intrigued by your new dosing strategy, and very curious to hear how it works out. Mainly, the idea of reducing Lantus and just using the Humalog to modulate throughout the day – it seems like it gives you the flexibility to adjust for activity throughout the day, and the number of injections isn’t really that different. It is kind of like going “untethered” with the pump. (I listen to a podcast by Stacey Simms where her son uses a pump in conjunction with Lantus and has had great results. They call this “untethered”, which I suppose is a bit of a misnomer). Once concern I would have with your strategy is that the action profile of the humalog might be a little too spike-y for a low carb diet. If you have good results, I may even want to give it a try.

    For your question: I have used inhaled insulin and have it in the fridge. I am afraid to use it often because I fear that it could promote un-regulated cell growth in my lungs. It seems to work slightly faster than an injection, and in spite of the (apparently) huge minimum dose of 4U, I haven’t had a hypo from the inhaled insulin. My version of a “treat” is in June when our cherry trees produce all of their fruit. I worked so hard to cultivate these trees, and it saddens me to just give all that fruit away – so sometimes I pull out the Afrezza to enjoy 30 cherries at once. I guess it is lucky for me that I don’t care for candy and anything other than a small quantity of chocolate makes me sick to my stomach.

    By the way, I’ve been doing the low carb thing for a while – BUT I’ve been eating beans and raw potatoes which are both high carb sources. The interesting thing about these two items is that the insulin requirement for both is MUCH lower than would be expected from the carb count. It is difficult to be objective on this one because every day seems to be different, but I get the general feeling that eating a lot of beans and raw potatoes actually reduces my daily insulin requirement (i.e. insulin to carb ratio) and tends to produce more stable blood sugars. Of course, this could be different for everyone. My guess is that this is related to resistant starch. For the raw potato, I take zero rapid insulin – really!


    1. OK, several interesting points there. Seems like if you have cherry trees, you could frequently eat those 30 cherries (BTW that is 39.4 grams total carbs) especially if your blood glucose (BG) is reasonably controlled with insulin (injected or inhaled). I know I would be tempted to overdo it. Maybe that is a function of how my brain works. Along those same lines, I used to eat berries because they are lower in carbs. But many times I would eat more than I had intended and BG would end up too high. This was before I started weighing my food, which helps me a lot now. So I decided I would be better off not eating fruit at all. The comment on the raw potatoes does not surprise me at all. I did read somewhere that raw potato is mainly resistant starch. I think your gut bacteria can use it, but it does not get absorbed by the intestine as glucose the way cooked potato does. I hope you are cooking those beans though. They too are pretty high in fiber (24% of the carbs in black beans are fiber), so you may not get much increase in BG when they are part of a meal.

      Brian, do you know how many grams of carbs you eat per day not including the cherries?


      1. Haha, the cherries only last about 4 weeks… and my diet is both seasonal and exercise based, so there isn’t a carb per day rule unless I’m doing a prolonged experiment. Like you, I’ve experimented a lot in 3 month blocks. With the Bernstein 30g carb per day diet, it seemed like I was approximately doing basal of 12Lantus + 6U rapid. Rather than tweak the insulin, I would increase carbs to roughly 50g per day when exercising moderately (~1 hour per day). At the extreme, when doing exercise all day, it is harder to keep track of the carbs because I’m out in the wilderness and just eating the carbs on an as need basis – but recently I did a race where I saved all of the wrappers. I found that for 30 hours of straight racing without sleep: I carried 10,650 calories, but only consumed 5,820. The macros of the food I consumed during 30 hours were: 512g net carb, 271g fat, and 293g protein. On the first 24 hours, I used 14 units of Lantus, and 7 units of Novolog. During the last 6 hours, another day had dawned, so I injected 12 units Lantus and 4 units Novolog before I finished the race. After the 30 hour race was over, I used Symlin in conjunction with Novolog and tried to eat lightly. I’m all over the place when it comes to carbs & insulin because one day I may be cycling 260 miles… and the next I’m sitting on the couch recovering (during the recovery, I tried to eat as little as possible, but in spite of this, the total daily dose peaked at 41U (22Basal / 19Rapid) while eating an estimated (not weighed) 30g carbs for the day.

        My insulin dosing comes from my meter. If I’m high, I take some more rapid. If I wake up higher than 85mg/dL, I guesstimate how much Lantus will have me wake up at 85mg/dL the next day. I mostly guesstimate these days as I spent a year weighing food with spreadsheets – trying to come up with equations that would be predictive. The prediction algorithms (which were complicated and were designed to “fit” the data to past experiences) still failed half the time to produce repeatable results. In other words, my best attempts at predicting the future were no better than my random guesses. My equations became more complicated to the point that they would even assign weighted values to the previous 3 (and later on 5) days of activity / eating.

        I assume that you too have found that carbohydrate labeling is often inaccurate. For me, some examples are beans and coconut which require less insulin than one would expect based on the carb count. Using the cherries for an example, if I eat the cherries now (they are not totally ripe, but still taste sweet), they have a much smaller impact on my blood glucose than if I eat them late in June when they spike me far beyond what you would predict from the carb count I find on Google. This makes sense to me because I also make wine from our grapes. As part of the winemaking process, I use a refractometer to measure the sugars of the grapes to determine when they are ready. The grapes I picked in November last year were far more concentrated in sugars than the grapes I picked in October. So much so that I had to add water to the November harvest to reach the BRIX necessary to produce a wine with 14% alcohol and no remaining sugar. The refractometer has become a toy of sorts that has proven to me that plant material is going to be really unpredictable when it comes to carb count – and that is before you mix in the variable of preparation (cooking vs raw). P.S. I do cook the beans!

        As a side note, I recently played around with a zero carb diet for one entire week – the “carnivore diet”. During this time, my body felt really good, but my conscience felt bad. During the week, the TDD needed to hover around 85mg/dL was roughly 16U/day: 13U Lantus / 3U Novolog. This suggests to me that insulin is truly needed to process fat/protein because during my (almost) fasting experiments, I had to reduce Lantus to nearly zero and Novolog was zero. I believe that this shows that insulin needs are more in line with total calorie intake than they are with carbs per-se. Perhaps you’ve seen the “Mastering Diabetes” program where they claim that diabetics need to eat really high carb and extremely low fat. They claim that their insulin sensitivity has increased because their insulin to carb ratio is much lower. I think that this is sleight of hand because (at least in my case), if you look at the insulin to calorie ratio instead, you find that not a lot changes between extremely low carb diets and extremely low fat diets. Having tried both, though, I definitely prefer the easier control that the very low (and slow) carb diet provides.


      2. Very interesting. You are an animal on the bike!

        Also, I have been doing the spreadsheet thing too and tried many different methods. Seems the only think that works is modeling the Humalog dose against the change in BG, but it only really works when the Lantus dose has been unchanged for 5 days in a row. That is what I am aiming to do in June to see if the model can calculate my Humalog dose.

        I also agree about the nutrition fact data, it not just carb counts, but all the macronutrient data varies widely if you spend time looking on the USDA or cronometer.com databases. That is why I keep my meals constant, that way I know the macros really do stay constant even if I don’t know exactly what they are.

        I would also say what you are doing would be difficult for me, but as long as you are happy with your results, more power to you.

        Why did your conscience feel bad with the “carnivore diet” ?

        I did see a “Mastering Diabetes” video on YouTube. Their insulin sensitivity has not increased, they are fooling themselves. The sleight of hand is that their definition of insulin sensitivity is incorrect. The correct measure is total daily insulin per kg body weight. They have to take a lot of insulin to cover all the carbs they are eating. Different strokes for different folks.

        You also get the prize for the longest comment on this blog to date. Thanks.


  2. Hello Dr. Runyan,
    Are you still taking Metformin?
    What are your conclusions after taking it as a D1.
    I feel that I need less insulin.
    Would love your feedback.
    Thank you.


    1. Hi Carlos, Yes I started metformin Jan. 1, 2017. I currently take 2,500 mg per day divided into 4 doses daily. I think its effect is small, but definitely it is effective in making me more insulin sensitive and lowering my total daily insulin dosage by a few IU/day.

      My conclusion is that persons with T1DM should consider giving it a try and seeing if they see any improvements.


      1. Yes, but these are “potential” mechanisms. There are some ongoing clinical trials that may show metformin can prevent disease or extend lifespan. Metformin has already been shown to delay onset of type 2 diabetes, but did not perform as well as diet and exercise. If you have some references to share, please do so.


  3. Keith,

    Hope you are well. It has been a little while since we last traded emails. I have been Keto now for a year and it is going quite well. Goes well with weight training, and I have found some limitations to endurance cycling, mostly on the big hills.

    In response your question about candy, I put this in the silly camp. Equivalent to “you can eat whatever you want, as long as you dose insulin appropriately”. I am actually concerned that we are on the verge of a tipping point in T1D management, which is aimed at enabling the Type 1 community to truly indulge in the SAD diet…. so we will have Type 1s that will achieve Type 2 status. The closed loop, nasal insulin and other devices are great technology, but their use should not be to enable Type 1s to eat what everyone should be avoiding.

    The great work done on general glucose avoidance by the Keto community and others like Peter Attia speaks volumes to this.

    I think the only place for carbs are with endurance exercise, over 4 hours.

    Curious if you have ever considered reporting on your cholesterol values, in addition to the below data. Lots of very interesting information these days that conflict with the last 50 years of cholesterol management.

    Thank you

    Andrew Rosenbaum



    1. I am well thanks, Andrew. I’ve reported cholesterol before, so did not feel to repeat posting it. Since learning about Dave Feldman’s experiences, LDL-C elevations in the setting of low trigs and high HDL-C is not much of a concern. My CAC score is zero as well. Last labs, Total Chol = 243, HDL-C = 91, Trig. = 78, LDL-C = 136, VLDL-C = 16, all in mg/dl. Hope that helps.


Comments are closed.

%d bloggers like this: