Today, Nicole begins saying goodbye to her Medtronic insulin pump. It’s been five years of a very questionable relationship. The next few weeks will show whether blood sugar control is better without this $3,600 device working away day and night.
“Of the new patients who visit me using pumps,” says Dr. Bernstein in an interview with Diabetes Health, “there was only one whom I was able to get near normal blood sugars. It was because he was still in his honeymoon period of diabetes. After a year on the pump, his blood sugar started getting unpredictable. Why? I believe it is because of the scar tissue that forms where you have a foreign body inserted for days at a time.”
It was for this very reason that we decided to go back in time to needles and syringes.
The Transition from Insulin Pump Back to Needles
Today, however, the transition period begins. Determining doses, of course, isn’t going to be a matter of some simple calculations, followed by perfect blood sugars. Adjustments will have to be made. So the first decision was whether to risk low blood sugars or high blood sugars.
“I’d rather go low,” Nicole voted. Shocking words, really. Prior to following a ketogenic diet she’d much rather have preferred to be high. Now, living off ketones and 20g of carbohydrates a day, she’d rather have a blood sugar of 3.0 mmol/L (54 mg/dL) than 6.0 mmol/L (108 mg/dL).
Also, to help with the transition, Nicole still has her insulin pump pumping away. We’ll use it to maintain her basal needs, while we master the meal bolus doses. Once the meal boluses are going smoothly, she’s going to switch to long acting insulin (the newly prescribed bottle of Levemir awaits anxiously in the fridge).
Letting the pump manage the basal, however, may not have been the best move. After all, it never seems to do a great job at it. Case in point, she woke up with a blood sugar of 2.9 mmol/L (52 mg/dL)at 6am this morning.
Usually she’d use the pump to inject an half-unit of Humalog (using a 30-mintue square bolus) at 6am to counter the dawn effect. Considering her low blood sugar, she delayed this until 7am.
One-and-half Glucolifts and 45 minutes later, her blood sugar had reached 4.2 mmol/L (75 mg/dL) (within her normal 4.2-5.2 mmol/L target range).
She then took 1 unit of Regular to cover the dawn effect. 1 unit, however, may have been too much. We decided upon 1 unit because back in her Humalog days (i.e. yesterday) she’d take 0.5 units of Humalog. The equivalent with Regular insulin is about 1 unit.
One Problem with the Pump
This is where our suspicions about the insulin pump begin. Even though the pump is supposed to micromanage 1/10 a unit of Humalog, it often seems like micro-doses don’t get properly absorbed.
Case in point, for lunch and dinner, which include 2 cups of low-carb vegetables and 4 ounces of eggs, fish or poultry, Nicole would take 1.3 units of Humalog with her pump. This would work fine. According to Dr. Berstein’s calculations (in the Chapter 19 of Diabetes Solution), for a 116-pound woman, this is the exact amount of insulin she’d need.
Strangely, though, when she eats a carb-free breakfast, with only 1.5 servings of protein (in the form of highly fermented, homemade kefir) she still needs to take 1.2 units of Humalog. Even when you consider the dawn phenomenon, this amount of insulin should send her blood sugar plummeting. But it doesn’t. Which made us suspect that not everything the pump pumps out, get absorbed. It was either that, or her liver was simply dumping lots of glucagon (“The Chinese Restaurant Effect”) in response to the kefir.
So was 1 unit of Regular, when injected using a needle, more than she needed for the dawn effect? One way or the other, we’d find out…
Breakfast Bolus (Trial 1)
Erring on the same side, Nicole only injected 3 units of Regular to cover her one and half cups of kefir this morning. Based on what was working with the Humalog in her insulin pump, she should have taken 3.5 units or Regular (as Humalog is 2.5 times stronger than Regular).
Glad she didn’t. 1-hour post breakfast her blood sugar had gone from 4.2 mmol/L (75 mg/dL) to 4.0 mmol/L (72 mg/dL). No alarms were sounding at this point; but 30 minutes later (around the time Regular nears its peak) she had gone down to 3.2 mmol/L (57 mg/dL). Another Glucolift tab and a glass of water, please…
45 minutes later, her blood sugar rose to 3.6 mmol/L (65 mg/dL). Another Glucolift tab and a glass of water, please…
30 minutes later, she’s having trouble reading and a blood sugar of 2.7 mmol/L (48 mg/dL). Another Glucolift tab and a glass of water, please…
“We should have filmed this,” I said. “We could have made a YouTube video out it. Sort of a reality show about coming off the insulin pump.”
“I don’t think that’s such a great idea,” Nicole said, sitting on the couch in her bathrobe, experiencing her third blood sugar drop.
“Well, it’ll make an exciting blog post,” I said.
“I’m glad you’re happy,” she replied.
Obviously, converting Humalog doses, given by a pump, to Regular doses, given with a needle, is not the way to go. Her ability to absorb insulin with a needle, so far, had proven vastly superior to her ability to absorb insulin from a pump.
45 minutes later she reached 4.4 mmol/L (79.2 mg/dL). Back to normal (for now…).
Her lows were probably caused both by taking 1.25 extra units for breakfast and double what she needed to cover the dawn effect. Tomorrow, we’ll try 0.5 units of Regular for the dawn effect and only 0.75 for the one-and-half cups of zero carb kefir at breakfast.
Lunch Bolus (Trial 1)
As for lunch, we decided to stick with Dr. Bernstein’s calculations (from Chapter 19 of Diabetes Solution), for her lunch bolus:
Carb Dose: 12g carb (2 cups of cabbage) / 8 = 1.5 units of Regular
Protein Dose: 4 ounces of protein (eggs) / 2 = 2.0 units of Regular
Total Dose: Carb Dose + Protein Dose = 3.5 units of Regular
“I think I’ll try three units to start,” said Nicole. A morning of low blood sugar seemed to have spoiled her appreciation for ketones.
In the end, she actually went with 2.5 units; injected at the beginning of the meal. 1.5 hours later, her blood sugar had risen to 6.4 mmol/L (115 mg/dL). 2.5 hours later, her blood sugar climbed to 7.3 mmol/L (131 mg/dL). And at the 5-hour mark, it had returned to 4.8 mmol/L (86 mg/dL).
Since the blood sugar eventually did normalize, it looked like the dose was right, just injected too late. “Let’s try injecting 30-minutes before dinner,” I suggested. “After all, most diabetics need to inject 45-minutes prior to a meal.”
With the insulin pump, she could inject at the beginning of the meal. Of course, the Humalog insulin peaked quickly. Regular offers more of a gradual curve, that should hopefully match her digestion.
Dinner Bolus (Trial 1)
At 5:30pm Nicole injected 3 units of Regular. At 6pm she commenced a dinner of 4 ounces of chicken wings and 2 cups of cooked collard greens. So this time, the insulin had a 30-minute headstart. Here’s how it played out:
6:00pm 4.4 mmol/L (79 mg/dL)
7:00pm 4.1 mmol/L (73 mg/dL)
8:00pm 3.8 mmol/L (68 mg/dL)
8:30pm 4.6 mmol/L (82 mg/dL)
That time the blood sugar did dip a tad too low. But that may of just as easily been caused by her basal insulin, rather than the meal bolus. She also had collard greens for dinner — a very high-fibre leafy green (possibly best suited for goats, not humans). The collards probably digested a little too slowly (despite the fact I cooked them for a half-hour).
The day started off very profitably for Jungell Inc. — Nicole had consumed five and half of their Glucolift tabs by 12pm. The rest of the day went relatively smoothly, with the blood sugar never going beyond 7.3 mmol/L (131 mg/dL).
One of my favourite sayings is: There’s no such thing as mistakes; just learning opportunities.
On the whole, 3 units of Regular taken 30 minutes before lunch and dinner seems like the best strategy at the moment.
As far as breakfast and the dawn effect goes… we’ve decided upon a single shot of 1.25 units of Regular at 6:15am. 1/2 a unit for the dawn effect and 3/4 unit for her 7pm breakfast of zero-carb homemade kefir.