#58 October 2018 Update on My Type 1 Diabetes 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 have made several changes over the past three months to further improve the glycemic control of my T1DM. These include:

  • I returned to two meals per day on 8/8/2018, but now eating breakfast (8 AM) and late lunch (3 PM) instead of breakfast and dinner. This will allow me to administer Humalog at 8 AM and 3 PM and Lantus at 6 PM separately rather than Humalog and Lantus together at 6 PM as I had been doing for many years. Turns out my basal Lantus dose was too high and my dinner-time Humalog dose was too low (many days I was not even taking Humalog at dinner-time). I check my blood glucose (BG) five times a day at 8:00 AM (fasting i.e. before breakfast) and 11:00 AM, at noon I exercise, another BG check and lunch at 3 PM, another BG check at 6 PM and take my dose of Lantus, then check BG at 10 PM (bedtime).
  • I returned to weighing my food on a kitchen scale last month to more accurately balance the Humalog dose with food. The last time I did this was about 10 years ago when I was carbohydrate counting. I did it for two years with very unsatisfactory results. IMO carbohydrate counting does not work for T1DM (or for T2DM for that matter) primarily because eating a lot a carbs in those with diabetes does not work.
  • Last month, I increased my dose of metformin to the maximum of 2,500 mg/day, 1,500 mg with breakfast and 1,000 mg with lunch. I have been tolerating this maximal dose without any side effects. I am convinced that even though I am relatively insulin sensitive, the metformin helps control post-meal BG by suppressing liver glucose production in response to meals and may be increasing muscle glucose uptake as well. These are the known mechanisms of metformin in helping to control BG in diabetes. I believe metformin is helping me because on the several occasions when I forgot to take the dose my post-meal BG was significantly elevated (by 30 – 40 mg/dl) compared to the previous days when I took the metformin. It makes sense that metformin would help T1DM because as I reviewed in other articles on my website lowcarbdiabetesdoctor.com, exogenous insulin is at a relatively low concentration around the pancreatic alpha-cells compared to normal and thus glucagon secretion is chronically elevated and particularly elevated after meals since amino acids (from the protein in the meal) directly stimulate alpha-cell glucagon secretion. Glucagon in turn stimulates liver glucose production (and ketone production). Less liver glucose production by taking metformin in turn means either lower BG or since my BG is low already, means lower insulin doses. I think lower insulin doses while BG is controlled is beneficial in terms of prevention of insulin resistance (and therefore “double diabetes”), cardiovascular disease, cancer, and Alzheimer’s dementia. These chronic conditions constitute the top causes of death amongst Americans. See the section below about euglycemic diabetic ketoacidosis where exogenous insulin doses can be too low!
  • Last month, all of the above resulted in a reduction in total daily insulin dose from 24.5 IU/day on 8/8/2018 to 16 IU/day on 9/30/2018 as well as a reduction in body weight from 163.4 to 150.6 lb. (over a 10 week period). I am sure the reduction in body weight contributed greatly to the reduction in insulin dose. However, insulin dose is usually expressed as IU/kg body weight and when expressed this way it is a reflection of insulin sensitivity. In general the lower the dose, the more insulin sensitive (assuming BG control is the same) the person is. Expressed this way, my total daily insulin dose decreased from 0.33 IU/kg BW/day on 8/8/2018 to 0.23 IU/kg BW/day on 9/30/2018. Thus, my insulin sensitivity improved due to the weight reduction and the increase in metformin dose from 2,000 mg/day to 2,500 mg/day. This is the lowest dose of insulin I have taken since my diagnosis in 1998. Fortunately, my insulin doses remained just as low in October, on average 15.7 IU/day, and my weight declined slightly more as well to 147.3 lb. Thus, the total daily insulin dose based on body weight was 0.24 IU/kg/day which is almost one-fourth the amount I was taking prior to starting my low carb ketogenic diet.
  • The last change I made last month was to decrease the volume of vegetables I was eating. This was due to some GI disturbances after meals. I was really eating more than I needed in terms of getting my micronutrients and fiber. I am glad to report that the GI disturbances resolved completely in October.

Glycemic Management Results for October 2018

My October glycemic results were about the same in terms of mean BG (96 mg/dl) and standard deviation (30 mg/dl) compared to September. Of course, I was hoping for more improvement (a reduction in standard deviation), but I did not achieve that. I did reach my desired BG goal of about 70% of time spent with a BG value between 61 and 110 mg/dl. I had a reduced frequency of asymptomatic hypoglycemia this month compared to last month. In October, 8.4% of my BG meter readings were less than 61 mg/dl, 61.3% were between 61 and 110 mg/dl, 30.3% were between 111 and 200 mg/dl, and none were in excess of 200 mg/dl. Preventing hypoglycemia is my top priority now so this reduction is asymptomatic hypoglycemia was encouraging.

The graph below shows my actual BG readings and the total daily insulin doses of Humalog and Lantus insulin and the total of both insulin doses.

Post 58 BG Insulin Totals Graph

The graph below shows the BG readings and the Humalog doses given.

Post 58 BG Humalog Doses Graph

I had more fluctuations in insulin doses this month compared to prior months. Hopefully this will smooth out in November.

The graph below shows the % Time spent in different ranges of BG on each day of October.

Post 58 %Time in Range

The graph below shows the % Time my BG was low, in target, or high and the mean BG value in those intervals for the month of October.

Post 58 %Time BG Low, BG in Target, and BG High

The graph below shows the % of meter BG readings in three different ranges at each of the five times of day that I measure it for October.

Post 58 %BG in Range Different Times of Day

New Diet Regimen beginning October 2018

The table below shows my new diet menu.

Daily Menu

Note: the MCT oil is only to be used if my body weight falls below 66 kg or 145 lb. I haven’t needed to use it yet.

From the table above you can see I eat different breakfast menus on Sunday Tuesday Thursday Saturday (beef and egg) compared to Monday Wednesday Friday (salmon). The weight of food and macronutrients are almost identical. Pictures of the meals are shown below. I realize I am not a professional food photographer. LOL.

Breakfast Beef and Egg on STTSBreakfast Salmon on MWF

Below is lunch at 3 PM everyday.Lunch Beef and Egg Photo

The macronutrient counts are as follows:

Macronutrient Table

Note that I am consuming 2.1 grams of protein/kg body weight/day. This is more protein than many who follow a ketogenic diet might consume. I am doing this due my age and and athletic goals as I desire to hold on to as much lean muscle mass as I can as I age. Aging has been shown to create some resistance to dietary protein in promoting skeletal muscle synthesis as well as resistance to the muscle building effects of resistance exercise. Stuart M Phillips, PhD in the Department of Kinesiology, McMaster University, Hamilton, Canada, based on his meta-analysis of multiple studies recommends a dietary protein intake of 1.62 grams protein/kg/day. However the 95% confidence interval is 1.03 to 2.20 grams protein/kg/day,  so I decided to consume protein at the higher end of his recommendation since I have no contraindications to doing so. The paper is here for your reference. At this level of protein intake, my blood ketones last month on two occasions were 0.5 and 0.6 mM.

You may wonder if eating the same menu over and over each day is difficult. For me it is not. I realize not everyone would be satisfied with the monotony. However, I did pick foods that I really enjoy and secondly, for me anyway, the lack of hunger produced by a ketogenic diet eliminates all cravings for other foods. If the above diet will result in the type of BG control that I am seeking, then the lack of variety will be worth it. Additionally, I can change the menu in small ways from time to time as long as I do so infrequently.

The graph below is a pictorial version of the above macronutrient table.

Macronutrient Pie Chart without MCT Oil

The graph below shows the % of the RDA (Recommended Daily Allowances) of vitamins provided by my diet. This was to confirm that I am getting all the micronutrients in adequate amounts. So as not to make this post too long, I will skip for now the meaning of RDA. Turns out that the confidence one should place on the RDA amounts is not very high.

Percent of RDA for Vitamins

The graph below shows the % of the RDA (Recommended Daily Allowances) of minerals provided by my diet.

Percent of RDA for Minerals

I would like to give credit to cronometer.com as an easy to use application to enter foods and determine both the macronutrient and micronutrient components of a menu.

Side Bar Regarding Euglycemic Diabetic Ketoacidosis

I have written on this blog in a prior post about nutritional ketosis as apposed to diabetic ketoacidosis (DKA). I think nutritional ketosis is a desirable state in general, but that the ketone levels do not need to be particularly elevated and in fact probably should not be particularly elevated, e.g. > 4 mM. I recently heard a talk by Jake Kushner, MD (a pediatric endocrinologist) at the Low Carb Houston conference where he said that persons with T1DM using a ketogenic diet to help manage their BG should be aware that nutritional ketosis may in rare cases increase the likelihood of euglycemic diabetic ketoacidosis (EDKA). DKA and EKDA are both pathological and potentially life-threatening states where the liver is overproducing ketones due to a high glucagon to insulin ratio in the blood, but in contrast to DKA, in EDKA the BG is not particularly elevated (generally less than 200 mg/dl). The fact that the BG is not particularly elevated makes the person with T1DM not be on the lookout for DKA. This phenomenon of EDKA has become more prominent recently due to some persons with T1DM being prescribed an SGLT2 inhibitor drug (which include the following: canagliflozin (Invokana), dapagliflozin (Farxiga), empagliflozin (Jardiance), empagliflozin/linagliptin (Glyxambi), empagliflozin/metformin (Synjardy), dapagliflozin/metformin (Xigduo XR)) off label to improve BG control and subsequently developed EDKA. SGLT2 inhibitors are FDA approved for T2DM and more commonly used for T2DM. The number of cases of EDKA has also increased in those with T2DM. Thus, this problem is not limited to just T1DM.

Dr. Kushner explained that nutritional ketosis has some physiological similarities to the effect of SGLT2 inhibitors. One is that SGLT2 inhibitors lead to chronically elevated glucagon levels and an elevated glucagon to insulin ratio. This elevated glucagon to insulin ratio occurs in those with T1DM (even not taking an SGLT2 inhibitor) because exogenous insulin does not adequately suppress alpha-cell glucagon production and I believe that nutritional ketosis and the resulting lower exogenous insulin doses probably magnifies this effect (although I don’t think it has been actually measured). So first, in my opinion, SGLT2 inhibitors should not be used in T1DM especially if a low carb ketogenic diet is being consumed. The second condition that can lead to an elevated glucagon to insulin ratio and the potential risk for EDKA in persons with T1DM is restriction of food intake either voluntarily (intentional fasting) or involuntarily due to an illness like gastroenteritis. The reduction in food intake requires a further (beyond what the low carb diet requires) reduction in exogenous insulin doses to avoid hypoglycemia. However, because as mentioned earlier exogenous insulin at very low doses does not adequately suppress alpha-cell glucagon production, the liver receives a signal (a high glucagon to insulin ratio) to overproduce ketones. This would not occur normally in a non-diabetic person who intentionally fasts or does not eat due to illness. So the take home message is that very low (or large reductions in) exogenous insulin doses in persons with T1DM should be a waving red flag. The blood ketone levels should be monitored along with BG and exogenous insulin doses need to be increased to keep blood ketones below, say 4 mM, even if that requires eating a significant amount of dietary carbohydrates. Obviously, illnesses can and do occur and cannot necessarily be avoided. However, voluntary fasting or significant reductions in caloric intake for the purpose of body fat loss, or other potential benefits (e.g. autophagy) can be moderated or avoided in those with T1DM. The benefits of fasting for autophagy and improvements in lifespan or health-span has only been studied in microorganisms and small animals, not in humans. Thus, implementing fasting for these purposes given the potential risks in those with T1DM is unwarranted in my opinion. As far as body fat loss is concerned, achieving a normal amount of body fat is important, but it must be achieved gradually with a modest caloric deficit (e.g. 100 – 200 kcal/day) to be safe in those with T1DM.

My Goals For November 2018 and Beyond

I am going to raise the bar further in terms of my glycemic management. I am not, by any means, sure that I will be able to achieve these lofty goals, but without goals, there is no direction to move toward. These are my goals:

  • Achieve near-normal BG values including mean BG of 96 mg/dl with a standard deviation as close to 12 mg/dl as possible.
  • Adjust my target BG range up from 61 – 110 mg/dl to 71 – 120 mg/dl and spend 80% of the day in this new higher range, 71 – 120 mg/dl. Honestly, with the elevation in BG that results from olympic weightlifting, this will be difficult to achieve. I do not feel the temporary elevation in BG that results from exercise is harmful. The elevation in BG is a normal response to exercise. The part that is not normal is that without exogenous insulin, the BG would remain elevated. This is why I eat lunch and take insulin right after I finish exercising. The main purpose of increasing the target BG range is to encourage me to avoid hypoglycemia. However, this elevation in BG after olympic weightlifting increases both the mean BG and the standard deviation. Thus, normalizing both of these parameters may not be compatible with olympic weightlifting.
  • Eliminate hypoglycemia or make it a rare event. I my case, my odds of developing long-term diabetic complications due to hyperglycemia are quite small, but I think my risk of having adverse effects of hypoglycemia are significantly higher. I want to eliminate that risk. This is my highest priority.

How Will I Achieve These Goals

In addition to the changes made in the past two months enumerated above, I am going make smaller adjustments in insulin doses by using 0.25 IU increments or decrements on my insulin syringes. This is because my total insulin doses are smaller than in the past and thus I need to have finer adjustments to the doses so the % change remains small. I have become aware that this applies to my basal insulin, Lantus, dose as well. I have noticed that changes in the Lantus dose have significant BG implications the following morning. I had been operating under the assumption that because it was “long-acting” it would take larger adjustments and require several days to have an effect. This is not the case. So any changes in my Lantus dose will be smaller than I have used in the past.

I hope next month I will have some improvements to report.

 

 

Advertisements

13 comments

  1. Rhonda Reed

    Why do you not wear an insulin pump? You can make very minute adjustments in your insulin as opposed to using an insulin pen or syringe with one unit increments. Are you using a continuous glucose monitor or finger sticks?

    Like

    • Keith Runyan, MD

      I use the Freestyle Freedom Lite meter. I have considered using the FreeStyle Libre system, but would prefer not to if I can manage without it. Similarly, for an insulin pump.

      Like

      • Rhonda Reed

        I resisted the use of a pump for many years because I believed it would be restrictive. Actually,the restrictions came from my diabetes management with a short term insulin and a long term insulin that forced me to eat at designated times everyday or risk frequent hypoglycemia. I took 4 to 6 injections daily and tested 8 times per day and still had A1c of 9 to 11. Getting a pump was like getting out of jail. I change my site every 3 days and test 4 to 6 times per day. I just took .1 of a unit that was precisely measured. Having the basal rate going in the background and bolus dosing for meals gave me far improved control and flexibility I did not have. It is still working with my transition to fat burning for fuel. For my lifestyle it is ideal. I plan on using the Cgm soon as I feel this will give me even more freedom. I thank you for the resources you have provided me as I have been Type 1 for almost 50 years. I’m praying ketones will work for me. Keep up the good work!!

        Like

  2. Brian Lucido

    Hi Keith – your recent findings and changes seem somewhat familiar to experiments I have been making. I appreciate the mention of the Euglycemic Ketoacidosis. That is scary because you would never detect it unless you’re monitoring blood ketones. On longer bike rides, I sometimes feel “a little weird”. Blood sugars will be in 70s or 80s, And I don’t eat. I used to think this meant I didn’t need to eat, but now I am starting to think that this is euglycemic ketoacidosis. I assume that you’re not worried about EKA with your lower doses, which are about 1/3 of the average T1 diabetic. Your most recent TDDs are very similar to mine.

    Also, You are using metformin, which I am not – but I did recently start on Symlin, and have had excellent improvements in the way I feel. Also, my standard deviation is better, and I am branching out of the Bernstein diet a little with home cooked beans. Mainly, the feeling of satisfaction after a meal is what I like the most. Previously, I was in a constant state of hunger which had been going on for years. I don’t remember seeing a post where you had tried Symlin; maybe you have already. Thanks for the update on your management. I always look forward to reading these and learning a bit in the process.

    Like

    • Keith Runyan, MD

      Thanks for the comment and kind words, Brian.
      Dr. Kushner was trying to point out to persons with T1DM who become ill with the typical symptoms of diabetic ketoacidosis like nausea, vomiting, non-specific abdominal pain and are taking an SGLT2 inhibitor drug or following a very low carb diet and have a normal blood glucose, to consider the possibility that they may have euglycemic diabetic ketoacidosis (EDKA), rather than assuming that because they do not have hyperglycemia, that they do not have diabetic ketoacidosis. He says this is when blood ketones should be checked to rule in or rule out EDKA.
      Personally, I do not worry about developing any medical problems, rather I remain vigilant and know that any person with T1DM can develop diabetic ketoacidosis or any of a number of different medical problems. To me, worrying about things that may or may not ever occur is a waste of emotion.
      I am glad you were able to try Symlin and have had good results with it.
      Just to add another option to that down the road are the GLP-1 agonists. They can be used off-label for T1DM, but currently are only FDA approved for T2DM. They are useful for controlling hunger and they also inhibit glucagon secretion similar to what Symlin does. Here is an article about these medications that can be used for T1DM.
      https://bjd-abcd.com/index.php/bjd/article/view/12/37
      I have not tried Symlin before. Fortunately, I have never had a problem with hunger. My weight gain since starting olympic weightlifting 3 years ago was not related to being hungry, rather it was voluntary. I thought I was gaining mainly muscle because I could see larger muscles in the mirror. However, in retrospect, I think I was overestimating the weight of that additional muscle and I was really adding more fat than muscle. I feel good at my current weight and am glad that I got rid of the unneeded fat.

      Liked by 1 person

  3. ÉRIC

    Hello, Keith.

    I remember having written some comments some months ago about the EDKA in which I showed my fears. It was more or less when I started the LCHF diet.

    I remember having found a year ago two studies in PUBMED which discussed the SGLT2-EDKA linkage, but I was assured by my former endocrinologist that even here in my town there had been reported other rare cases with no relation to SGLT2.

    Since then, and specially during the last months, I have been monitoring my blood ketone levels (BKL). According to my own data and my meetings with my current endocrinologist, I have found some interesting points which I will list below:

    1) First of all I should define the context. My CH daily intake is between 47 and 48,5g, and between 3200 and 3700kcal (79-80% of kcal intake comes from fat; 14-15% from protein; 5-6% from CH). My dietary protein intake is of 2,2g/kg/day. I have 3 meals per day: 6:15am, 14:45am and 21:15am.

    2) I practice 3 different sports: mountainbiking, swimming and bike training. I spend about 20 hours/week, specially in mountainbiking. Swimming and bike training sessions do not exceed 1,5 hours, but are much more vigorous during the whole workout. Mountainbiking, however, obviously combines moments of varied intensity.

    3) My HBA1C is 5,3%, and 100% of my BG levels are in desirable range (60-130mg). My insulin doses are 2u Toujeo basal insulin (3u in stressed days), 2+1u Novorapid at breakfast (1u less if I start training less than 1,5 hours after the injection) , 2u Novorapid at lunch (1u after long montainbiking workout, or after swimming or bike training if these workouts are close to lunch time), 2+1u Novorapid at dinner (2u if long mountainbiking workouts for two or more days in a row).

    4) In these conditions I find my fasting BKL very difficult to exceed 1,5mmol. I usually register between 1mmol and 1,4mmol, regardless of the previous day exercise. My endocrinologist told me that these BKL levels are already dangerous when combined with high BG, so patients are already admitted at hospital in these cases.

    5) To my surprise (and my endocrinologist’s), practising sport hardly increases BKL, which helps to mitigate one of my most persistent fears: suffering from EDKA while exercising. According to my experience, my BKL always goes down when practising intense workouts (swimming, bike riding): it may fall, let’s say, from 1,2 to 0,6mmol, or from 1,4 to 0,6mmol… I have never registered an increase. With mountainbiking, however, things are a bit differet: my BKL may remain almost stable (sometimes I observe a very low rise, sometimes it goes down very slightly; let’s say 0,2mmol, even if I ride for 5 or 6 hours). I assume -correct me,Keith- this is due to the fact that vigorous workout implies a huge consumtion of ketones (this consumption is much more moderate with a not so intense workout, like mountainbiking).

    I would like to underline that when I swim or practise bike training I never eat, either before or after the workout. When mountainbiking I only eat the second and the following days, if I practise it for two days or more in a row. But the first day, and also sometimes in the second day, I can finish a 5 or 6 hour workout without eating. When I eat, I just eat hard cheese (between 40g -5 hours workouts- and 140g at most -in 8 hours workouts if preceded by more days with 5 or 6 hours of mountainbiking-). This cheese is 24% fat and 23% protein. But even with this diet, my BKL stays pretty stable during the workout, as I said before.

    6) After having exercised, my BKL increases quite quickly: it may increase from 0,6 to 1,4mmol, for example, or from 1 to 1,7mmol, in not more than an hour after having finished the workout. But then it decelerates and remains almost stable for the next 3 or 4 hours.

    7) The highest BKL I have ever registered is 2,2mmol. It has happened 3 times in 5 months. One of them was in the morning, at 0:30pm, after a 14 hour fasting period. The other two were at 2:30pm, 5 hours after having intensely exercised (bike training).

    8) To my surprise (and my endocrinologist’s), BKL don’t fall significantly after meals EVEN IF I mesure them when the insulin is doing its strongest effect (1,5 hours afecter being injected). They may fall, let’s say, from 1,2mmol (before starting meal) to 1-0,8mmol, or from 1,4mmol to 1,2mmol.

    9) Again, to my surprise (and my endocrinologist’s), BKL don’t increase significantly after a fasting period (at night) of 8,5 hours. They may increase 0,2 or 0,4mmol; I haven’t registered a higher growth than that.

    In conclusion, after 1 year and 2 months following a LCHF diet, I still don’t understand the BK fluctuations… I read about people that reach 3mmol or more without any practise of sport, but in my case, even if I train hard or if my fat intake is huge, my BKL remain modest (or, at least, remain in levels much lower than I expected). May it depend on each one’s metabolism, Keith?

    And finally I would like to say some (more) words about Freestyle libre. I still find the same margin of error: between 20% and 30% (sometimes even higher) over the BG during the first 4-5 days of use; between 20% and 30% (sometimes even lower) under the BG during the last 3-4 days. So it works accurately for 5-7 days, but when mountainbiking the error rate increases noticeably also in these 5-7 days. The last sensor I used, for example, threw glucose levels under 40mg when my BG was 90mg. Abbott told me that this error rate was acceptable.

    Liked by 1 person

    • Brian Lucido

      Hi Eric – Interesting post about BKL levels… I have no explanation, but I appreciate seeing that you don’t eat during exercise and are doing fine.

      With regard to the Freestyle Libre, I have a theory that subcutaneous glucose monitors that rely on interstitial fluid work very poorly in athletes. I spent a year in frequent contact with Dexcom and finally had to give up using it due to wild inaccuracy. Meanwhile, people all over face book (presumably non-athletes) are enamored with the device. My theory is that athletes have wild hydration swings over the course of a short period of time. For example, you can go on a mountain bike ride on a hot day and easily lose several pounds of water. This likely relocates water from your subcutaneous space towards your core. I started noticing big swings when I’d ride in freezing temperatures as well. Presumably, your body draws fluids from the periphery into your core. When fluids are drawn away from the sensor, it seems to produce a false low. If you calibrated during this time, you could be setting yourself up for a wild ride on the CGM. I was unable to find a solution, but occasionally on phone support was able to talk to some technicians who shared this theory. I did find that the Dexcom worked well if I took a week (mostly) off from training.

      Like

      • Keith Runyan, MD

        Thanks for that feedback Brian. There does seem to be quite differing experiences with CGMs. Dr. Kushner in that Low Carb Conference talk basically stated that a CGM is a “required” tool when using a low carb diet for T1DM. But yet again, your mileage may vary! Dr. Kushner’s talk starts at 2:00:00 on YouTube https://www.youtube.com/watch?v=6RTC_ZQ79To and the comment I referred to is at 2:45:00.

        Like

    • Keith Runyan, MD

      Thanks for the comment Eric. You made a lot of points and I (and I think others will too) think your observations are interesting.
      First, please look at Brian Lucido’s comment and my reply on this post regarding EDKA.
      I would not agree that a blood ketone level (BKL) of 1.4 mM and high BG alone constitutes a reason for hospital admission. In fact, the level of blood ketones is not one of the criteria for making a diagnosis of DKA, although they do need to be present to make the diagnosis. The diagnosis requires both a low blood pH and low serum bicarbonate levels. This article explains DKA quite well. http://spectrum.diabetesjournals.org/content/diaspect/15/1/28.full.pdf
      Second, none of your comments about BKLs are surprising to me. In your case, your insulin requirements are so small that I suspect you are still in the honeymoon phase of T1DM which means you still make insulin and your alpha-cells are still influenced by your remaining beta-cells to keep glucagon more appropriately suppressed compared to a T1DM with essentially no functioning beta-cells. Glucagon, in a T1DM with essentially no functioning beta-cells, is chronically elevated which will signal the liver to make both glucose and ketones and thus make the BKLs higher on a low carb diet.
      As I have stated many times on this blog and in my books, the goal of using a ketogenic diet for diabetes is not to make and use ketones, but to make the management of BG easier and closer to normal levels which occurs as a result of not eating very many carbohydrates, i.e. generating ketones is just a side effect which at moderate levels probably have some fueling and signaling benefits, but do not need to be high to get that benefit.
      Finally, your comments about Freestyle Libre are also appreciated. In fact, personally in addition to not wanting to be “hooked” up to things unless absolutely necessary, my demeanor would be annoyed with having a device that gave inaccurate readings. Since, I stopped doing long duration endurance exercise in 2012 and I don’t have problems with night-time hypoglycemia, I feel a CGM is not absolutely necessary. That said, if I were doing long duration endurance exercise the way you are, I can see that the device has a benefit.
      Best wishes for your continued success!

      Like

  4. Éric

    Thank you both for your answers, and also for the links, Keith. Interesting speech by Kishner.

    Brian, your thoughts about CGM systems can shed some light to the fact that all these systems show great inaccuracy in similar conditions. There must be a physiological explanation; maybe yours is wright. But then we would still need to explain why Freestyle normally shows much higher glucose levels (compared to BG) while training. I mean: in my case, when practising sports, having BG underestimated is very unusual… (although, as I posted, it may occur).

    I thank you also Keith for the clarification about the honeymoon. I agree with you… although my endocrinologist does not! She argues that my fasting serum C-peptide levels are too low to still being in the honeymoon. However, these levels are a bit (only a bit) higher now than 4 years ago, when I was first diagnosed T1D LADA. By the way, I had never considered what you explain about alpha and beta-cells and its relation with BKLs; very interesting. Anyway, I am not interesting in reaching higher BKLs; it is just that I didn’t understand why they were so moderate. Now it makes sense.

    Finally I would like to say that I agree with Brian: it is very useful to read your posts. There are always things to learn, and I appreciate specially the accuracy of your data (we need maths!). Go on, please.

    Like

    • Keith Runyan, MD

      I am glad you appreciate the blog. I enjoy the exchange of information too.
      I could add in addition to fluid shifts (mentioned by Brian) contributing to CGM inaccuracies during exercise, all the CGM companies admit there is a delay between blood glucose (BG) levels and interstitial fluid levels. Exercise can change BG quite rapidly, and undoubtedly there would be a discrepancy for a period of time. I think the technology in general is just an approximation of BG anyway and understanding all the factors involved would be difficult unless we were actually designing and experimenting with the devices.

      Like

      • Éric

        Yes, I knew about the delay.

        Let’s hope in the future we can have a more “democratic” development of technologies the way you expose; for now I think that at least in the short run it may collide with corporation’s profitability.

        Like