Long-Term Experiment: Switching to Novolog from Humalog in Omnipod, Try #2

I’m switching to Novolog from Humalog in my Omnipod. I tried this last year due to multiple skin issues, site issues, and stuck highs. Novolog was more effective for me for the month that I tried it…but its effectiveness spooked me a few times and I switched back to Humalog.

Many of my skin issues had resolved when I switched back to Humalog, to the extent that I thought it would meet my needs again.

Here I am again, switching back to Novolog. My goals are:

  1. Fewer site failures
  2. Fewer problems with tunneling
  3. Tighter control
  4. More consistent insulin effectiveness in repetitive situations

In the last few months, Humalog has been behaving like a sloppy drunk. Correction doses tend not to do what they used to do. Interestingly, when I inject either Humalog or Novolog, both result in what I believe are red erythemas. That has been going on for the last year now that I look at my notes. No one knows if this means I am allergic to something in the insulin, or if it’s something else. One CDE believes I may need to switch monthly between Humalog and Novolog if I am developing an allergy. I hope it doesn’t come to that. And I hope if it does come to that, that it will actually work.

I use Omnipod and Dexcom G5 for anyone new to FUD reading this. I have used Humalog for 13 years.

4 Likes

Initial Observations:

Correction doses on Novolog are working about 4x more effectively for me than on Humalog. I’ve also changed my pump calculator’s duration setting from 3.5 hours for Humalog to 4 hours for Novolog…although looking at Dexcom data it might even be 4.5 hours for me. Too soon to tell.

6 Likes

The 4x effectiveness is shocking. Clearly there is something special going on.

2 Likes

Agreed.

Thinking back, the Novolog correction effectiveness (via pod) seems stronger than Humalog corrections via pen back when I was on MDI five years ago. I’d use 1:50 for corrections then. Last night, 0.8 units of Novolog (via pod) brought be down 100 points. If I’d had to pump Humalog for the same correction (I was correcting a 197), it would have taken several tries of my correction dose.

The interesting thing…so far, is that basal seems to be spot on for both Novolog and Humalog. In a fasting state with Humalog via pod, my basals would work. It seems like for some reason that correction doses and meal boluses on Humalog just make it forget how to do its job. And it would just wander off and then quit working way before its duration should be up. I’d been checking for tunneling, bleeding, etc etc etc for months to attain optimum technique for podding with Humalog. Hopefully Novolog will behave for me.

3 Likes

I did an extended hot shower experiment tonight. I waited until my dinner bolus had run its course. With only Novolog basal in my system, I did not see a scary bg drop.

I suspect that if I’d had any significant IOB, I’d have seen the evil vacuum effect on my bg that I saw the last time I tried Novolog. Will keep an eye on that going forward. I still don’t know if the hot shower effect just increases how quickly IOB works, or if it actually makes it more effective.

image

2 Likes

FWIW…and I could be wrong or this could be just specific to my body’s interaction with Humalog and Novolog…but at this point, it seems that Novolog is better for meals and corrections, and that Humalog is better for basal.

Novolog seems a bit flimsy when it comes to my basals (based on the last two nights…so too soon to really tell…but my bg has jumped and tanked and jumped on its own overnight on Novolog whereas it’s been rock steady on Humalog for months). Novolog is nimble on correction doses and meal doses. Impressively so compared to Humalog. I’ve also noticed that so far I can reverse downward trends much more quickly on Novolog than on Humalog.

Novolog definitely seems like a much better choice (in my body) for Sugar Surfing due to its nimbleness. Humalog just doesn’t function that nimbly in my body so true Surfing is not really an option.

My ideal (with very limited data at this point) would be to have Humalog for basal and Novolog for meals/corrections/surfing.

5 Likes

Do you intend to try this? It seems perfectly feasible to me (insurance obstacles notwithstanding). For a little while I used the pod just for Humalog basal and did boluses with injected Fiasp.

1 Like

I will probably try to do a month of podding with Novolog to try to get a good big picture baseline of how it performs on all fronts compared to Humalog. But I’m not opposed to going with that regimen if my experience supports it!

Looking back at my A1C’s since after my second pregnancy (2014 - which is when I became super low adverse (and was tanking all the time) due to the responsibilities of motherhood and started podding), they’ve all been mid-6’s up to a peak of 7.56. I would like to get back to mid-6’s, but still while avoiding deep or chronic lows. I maintain optimism that it is possible, but allowing lows certainly leads to lower A1C’s for a variety of reasons. My control is so much more consistent while podding compared to MDI (Lantus/Humalog), so I am definitely safer and seeing a smaller SD in my bg. But I’d really like to bring the average BG that I revolve around lower.

It seems to me that the difference between where I want to be bg-wise and the where I am bg-wise is the margin of error that comes with diabetes. If it’s not revolving hormones, it’s a skin issue, or a knocked cannula that sort-of-might-be-but-not-really leaking, or a miscounted meal, or unplanned activity or sedentariness. I don’t say that to be negative. But I do want to keep a clear eye for myself on the the trade-off between aggressively chasing a lower bg average vs keeping an overall quality of life that comes with a smaller bg SD (with fewer scares, especially since I do not feel lows and have not for some time) while keeping in mind the realities of manually controlling blood sugar while living life. Today’s numbers are sluggish on Novolog while yesterday’s were pretty slick. Will keep investigating and sharing but that’s where I’m at currently.

4 Likes

Here’s what bothers me mathematically right now:

On Humalog, I used:
1 unit to bring me down 50 points
4g of carb would raise me 20 points
I:C (other than breakfast) of 1:10
The math on that supports each other. (To raise myself 50 points, I’d need 10 grams of carbs. 1 unit of insulin could handle either dropping me 50 points or eating 10g of carb.)

On Novolog, I’m finding that:
1 unit brings me down 100 points
The other night I noted that 12g of carbs raised me 100 points. So that would be 4g can raise me 33 points.
Reversing the math, that would bear out an I:C ratio of 1:12 if I time it correctly.

Thoughts? Do you all think it actually works like this mathematically…if Diabetes occurred in a vacuum?

1 Like

I have no idea if it is possible or not, but I can say I’d test it out a few more times before drawing a conclusion that really surprises you. So many factors at play— If it’s legitimate, another test or two will produce similar results.

Because diabetes moments are like snowflakes…

2 Likes

I have tried many times to close the circle of calculations: ICR → carb raise → CR and it never works for us, by a very large factor. I have always wondered why, but possibly it should be the opposite: why should it work? I have always thought we should make a thread on that…

2 Likes

You can’t make a direct correlation between what you eat raising you a certain amount, and expect a certain amount of insulin to drop you the same corresponding amount every time. It does not work that way.

When you eat, your body is not putting all of that into the blood. The liver is taking some of it. The numbers just don’t line up like people want them to.

When you eat, the liver takes some of the glucose. The liver does not need insulin for glucose uptake. It uses the GLUT2 transport, which is a low affinity glucose transporter. So glucose uptake by the liver is generally slow, but is in proportion to the amount of glucose in the blood.

There is cross-talk between the liver and other organs and cells, particularly the digestive system. The digestive system signals the liver in response to food ingestion in a very complex way.

When carbs are abundant after a meal, the liver converts glucose into glycogen and fatty acids for later use. When carb intake is less abundant, the amount stored as fatty acids is much less.

The rate of glucose uptake by the liver is also dependent on the amount of glycogen currently stored in the liver. And it is also affected by the amount of energy being used by skeletal muscle - active versus resting state.

There are different amounts of glucose uptake and release by the liver in low energy/fasting states, low energy/fed states, high energy/fasting states, and high energy/fed states (waking up in the morning, lunch, dinner, bedtime, exercise, etc.).

Additionally, high blood glucose changes insulin sensitivity, so the numbers are not linear for corrections.

All of the above is simply to say that you really do not have any way of knowing for sure exactly what your liver is doing in response to a meal. 20 grams of carbs that you eat does not mean 20 grams of carbs will go into your blood.

On the other hand, if your blood glucose is 150, you know exactly how many grams of glucose is in your blood at that moment. That’s what that 150 number tells you. The glucose is already there.

So IC and correction factor to do not always line up the same. They might sometimes, but there is no guarantee that they will always line up.

4 Likes

I appreciate the responses.

I’m not trying to discern ironclad numbers to use blindly in pump calculations. I’m trying to establish a safe ballpark for getting through the acclimation to a new insulin.

Pumping, to correct a 200 on Humalog, would take about 4 units in practice (start with two units, then another, then another). Via Novolog, I need 1 unit or less, so far, and I don’t have to keep throwing correction doses at it. I have to pay attention to that. I don’t think it is too far out of the realm of possibility to wonder if that big of a change in correction factor should lead me to re-examine my carb ratios for Novolog as I keep wading into this.

2 Likes

Not trying to poke fun, but really… there was my explanation:

And Michel’s:

and then Eric’s…

Who you gonna trust? :grin:

4 Likes

I actually think this is a good question that should not be dismissed. I first noticed this relationship when I did the math after reading the Pumping Insulin book many years ago. When I worked out the math I noticed this same relationship and although there will of course be some variance, I believe it should in general be a valid relationship.

Thus:
Carb/Insulin ratio (g/units) = Corr ratio (mgdl/units) / Bg/Carb ratio (mgdl/g)

First, note that Eric’s three arguments are true, but they are largely irrelevant to whether or not the above equation is valid. First, it is true that the liver has an effect, but it will have an effect on both the Carb/Insulin ratio and the BG/Carb ratio so should largely cancel out. Second, we can restrict this to times when exercise is not in play - since for most people that is a small fraction of the day. Finally the Correction ratio being non-linear does not affect this equation either since that is inherent in the calculation of the Correction ratio - and the effect can be minimized by being calculated at a reasonable BG level, say 250 mg/dl or less.

In short, Eric’s arguments say that none of these values can ever be accurate. And of course he is right, these are only approximations and rules of thumb. Yet these rules of thumb are the whole basis of carb counting, and clearly carb counting has value since it is used by most T1 diabetics every day. Of course carb counting has limitations, which is why more people are using postprandial testing (now called sugar surfing) to fine tune the results of carb counting. But carb counting is simply the expectation that if you eat twice as much food, then your BG will climb twice as high - and it will require twice as much insulin to bring down - and there is little doubt that this is a generally valid expectation.

If we accept the value of these carb counting ratios, then I think we can reasonably expect them to be related as in the above equation. What the above equation really tells us, is that it doesn’t matter whether we take our insulin before we eat, or after we eat: that the same amount of insulin will be required to get us back to our starting BG in either case. If these ratios have been calculated correctly, then in my opinion this should absolutely be a reasonably valid expectation.

Now to get back to your ORIGINAL question, I would be very concerned that there could be a problem with the ratios you have computed for Novolog or Humalog. Either that or your body is doing something very strange and unusual with Novolog that it is not doing with Humalog. I compared Humalog and Novolog as a test once, and I found very little difference between them - in fact I could not find any discernable difference at all, so I stuck with Humalog since I already had a prescription for it. I could understand a 125% or 150% difference in potency, but a 400% difference in potency simply doesn’t seem right. So indeed, I would be very careful and would probably do more careful testing with the two insulins (and get another bottle of each to be sure it isn’t the bottle you’re using).

1 Like

Explained in a simpler way.

Why do IC and correction factor not always equate?

Not every carb you eat (IC ratio) requires insulin or ends up as glucose in the blood. Some of it ends up stored in the liver and it does not require insulin to get there. The liver can use the GLUT2 transport which is not dependent on insulin.

But every bit of glucose in the blood (correction factor) is already in the blood.

This does not have anything to do with Humalog vs. NovoLog. It’s just an explanation for why IC and CF do not always equate.

The problem with this formula:
Carb/Insulin ratio (g/units) = Corr ratio (mg dl/units) / Bg/Carb ratio (mg dl/g)

is that Bg/Carb ratio assumes that your blood sugar will rise a certain amount for each gram of carbohydrate ingested. It does not. The liver takes some. And you can not count on what the liver will take out.

That is the point I was trying to make.

4 Likes

First off - interesting topic @T1Allison. For me Humalog is just not as effictive as Novolog or Fiasp. I have tested a couple times. It is sluggish, and I have to use a lot more of it. For example, my I:C ratio is 1:3 on humalog and 1:5 of novolog and my humalog basals are much higher. This is interesting but I have no idea why. I took humalog from it’s introduction in the late 90’s until a few years ago when I was started having serious control issues. The switch to Novolog really helped. Now I use Fiasp. When I first used Humalog it was fine and then it all went downhill.

Not for me… On Fiasp
1 unit to bring me down 56 points (2 mmol/l)
4g of carb would raise me 16 points (0.88 mmol/l) or 4 points (0.22 mmol/l) per gram of carb - this is tested by having a flatline and then eating glucose to see the response.
I:C of 1:5

So based on the pumping insulin formula… that @jag1 mentions.

Carb/Insulin ratio (g/units) = Corr ratio (mgdl/units) / Bg/Carb ratio (mgdl/g)
5 != (56 / 4) = 14

So… I go back to the @Eric thing - when I eat carbs weird things happen in my body.

When I calculate that 1g of carb raises my BG by 0.22 mmol/l, that is just eating glucose. I suspect that when I eat other foods, my body responds differently (as part of the digestion process) that leaves me with a ridiculous I:C ratio.

EDIT: Using the formula my IC ratio should be 1:11, which is pretty typical. I suspect the average I:C ratio is around 1:10 for most people.

2 Likes

I seriously do appreciate everyone’s input. I hope that is clear even in the absence of emoticons.

I am glad to know from FUD that even if Humalog and Novolog are interchangeable for some or even lots of people, that it isn’t necessarily interchangeable for others.

On MDI, pre-motherhood when my A1C’s were lower, I was SO aggressive with Humalog. I’d carb count, round up, add a few units for a snack at the tail of Humalog, and pre-bolus 25-60 minutes before eating. Then I’d space my eating out over the course of three hours to match its most effective time. Then the tail of the Humalog (for lunch) would coincide with Lantus’ over-effectiveness for the day (bedtime Lantus shots dosed to match and head off dawn phenomenon) so even though I’d be feeding the Lantus, I probably assumed my afternoon snack was mostly feeding the Humalog tail. Well, now that I’m on a well-tested basal regimen, there’s no afternoon low to feed and my concept of Humalog (after five years of struggling) has probably become more accurate. On MDI, my meals had an Up and a Down (bg), a nice bell curve that I tried to flatten with aggressive dosing, pre-bolusing and monitoring (soooo many fingersticks). Now, being quantity limited to preserve site integrity, I was trying to accomplish the same feat with a third of the insulin. It was just go Up and Stuck, not Up and Down. If I am moving non-stop, the Humalog works closer to my recollection of it…but sometimes I have to actually sit at my desk and get some work done. And then the Humalog just does…seemingly nothing.

So…today taking 4.5 units of Novolog for lunch (that would have been 8-10 on Humalog MDI on a work day), only pre-bolusing by 10 minutes, and seeing a defined Up and Down, albeit a higher spike than optimal, while sitting at my desk and not walking at all through the four hours of observation, was CRAZY AMAZING. Humalog never did that. Not on its best day. I’ll be curious to see if I need to cut down my meal doses a similar percentage to account for activity as I did on Humalog…or if Novolog is different in that regard in my body.

Maybe something in my chemistry has changed. Or maybe I’m someone who just reacts differently to Novolog than Humalog. Don’t know. I’ll keep trying, observing, posting.

I am curious to see how hormones impact my basal effectiveness since I’m going to see The Drop soon. I’m already seeing weird things overnight that might pose problems in the coming weeks. IF basal works similarly between Novolog and Humalog (I have no complaints about Humalog for basal that I know of) through hormone swings, that will be quite interesting given the difference in their effectiveness for corrections and carbs so far.

EDIT: I want to add that I’m sure this thread looks incredibly OCD. And it is. But after years of feeling like I’m having to inject more and more and more potential energy into my body that may or may not decide to kick in…and when it does it is hard to stop it…and I can’t feel the lows…I’m careful. Very careful. I just want to find a treatment regimen that gives me some semblance of cause and effect. A T1D level of cause and effect. It doesn’t have to be perfect, but it HAS to get better than it’s been.

3 Likes

Not for me. The insulin I need to correct X mg/dL downwards doesn’t equate to the insulin I would need for carbs that otherwise will raise me X mg/dL. I was surprised when I first observed this, but that’s just the way it is.

3 Likes

But you’re eating carbs on both sides of the equation, so that carb loss to liver is happening whether you inject before or after you eat. And it is therefore the same and doesn’t affect the validity of the equation. That is just math. Which is the point I was trying to make.

If you’re willing to do an experiment, this is one that could be done pretty quickly. Eat the same small meal, and measure delta BG when you inject before eating, and delta BG when you inject after eating. I suspect the same amount of insulin will have pretty much the same effect in either case. And that is all that equation is saying.

I’ve been changing insulin’s recently so I’ve been measuring these things again and using a CGM recently, and the numbers I’ve come up with work pretty well in that equation. So all I know is it works for me.

1 Like