Levemir 1x daily

I have never used levemir before. I have one sample pen of it that expires this month. I have been using lantus 1x daily for several years, on and off at first and then consistently. I take lantus in evening, usually around 10… I stay pretty steady basal wise throughout the day generally, but then, oddly, often drop into the 70s or 60s in evening, which is odd because that’s when the lantus should be wearing off instead.

Anyway it seems to me that trying levemir 1x daily in evening might be a good option. Does anyone else actually do this or are you all twice a day with levemir? Should I start with the same dose as I used with lantus?

Thanks for any ideas

I take levemir. When I swapped to it from Aphidra, I used the same dose and didn't need to adjust.

I take 15 units at night, and 8 units in the morning. Ie. twice a day dosing.

But some days I just take the 15 units at night and none in the morning. My blood sugars tend to be much more stable during the day (and need less insulin - maybe something to do with being active in the daytime) than overnight.

All I can recommend is that you try it starting with just the evening dosing, and then you can add in a morning dose if you find you need it....

I actually really like levemir (better for me than Aphidra or Humalin-N, which are the other 2 I've tried).

sam, i used to use lantus but switched because the injection site would be sore for ages. i really like levemir. it is painless and keeps me quite steady throughout the day. like you, i also drift downwards in the evenings, usually around 8pm till i finish teaching at about 10. ive always thought it weird, too.
i never tried once a day with levemir, just started with two shots. i had read that levemir was a bit less potent but started with the same dosage to be safe, actually having to almost double my number of units-i was taking like 3 units of lantus then (those were the days of just being a little diabetic!), and with the levemir i had to take 5, 2 in the morning and 3 at night.
that must have been almost two and half years ago. this summer i was taking 3 in the am and 5 in the pm from the heat. now with the cooler weather im at 5 in the am and 7 at night.

Levemir is not designed to be taken once per day. It binds to albumin and as long as this binding is kept the Levemir is not active. With a specific likelyhood the binding will dissolve over time. In contrast to Lantus - that binds to fatty tissue - the likelyhood for this unbinding is not constant. In fact the likelyhood to unbind is higher in the first 12 hours. In lab tests with healthy persons this behaviour can be studied for all insulins. Healthy people get an injection of Levemir and their blood glucose is constantly monitored. In response to a drop of the blood glucose the test team will infuse glucose to work against that drop. This way the blood glucose of the healthy person is kept steady. The consumption of glucose plotted against time will show how active the insulin was and when. For Levemir you will get a curved shape - as to expect:

You can see that two shots every 12 hours have the potential to combine to a more evenly distributed activity curve. While the first shot fades out the other fades in. In my opinion this works much better than Lantus which can be prone to having gaps in the 24 hour coverage. Something that is much less likely with 2xLevemir per day.

Hi Holgar,

Thanks for this.

For me though I may not need to take it twice a day (my pancreas is still not totally kaput and seems to function better in the afternoons and evenings). I take my evening dose about 10 pm. This should theoretically tide me over until about 10 am the next day.

I typically wake up with a blood sugar in the high 70s low - mid 80s. I typically don't eat breakfast these days and even without dosing in the morning can keep my blood sugars in the range of 75 - 90 most of the time....

Once I eat by lunchtime / mid-afternoon I use Novorapid to cover carbs (or if no / very low carb) then I can continue with no action up until bedtime.

However, regardless of what I do or eat I really do need the nighttime dose to keep morning blood sugars on target.

I like the 2x/day dosing because it allows one to fine tune the dose better in terms of day/night requirements.... and I believe that many people find they do not need 50% in day and 50% in night, it is an uneven split.

What isn't clear in this curve is that many believe that all basal insulins like Lantus and Levemir act more rapidly at lower doses. The explanation I've heard is that the insulin injection goes into a shape like a sphere. The volume (dose) of the insulin is the volume of the sphere, but the absorption is only through the surface of the sphere. This as the dose increases, the proportional absorption falls. I think for most this would only apply marked to doses under 10 units.

Hey Sam:

As others have stated, Levimir is not really intended for 24hrs, but from what you describe, this might actually work to your advantage since you're going low in the evenings. You could fine tune your dosages and injection timings to account for this.

Holger does a great job of describing the efficacy curve. I'm one of those who take two Levimir shots over 24hrs, and when I have the dose dialed in right I get amazingly flat basal rates. In my case, I learned that it works best for me if my nigh-time dose is 25% greater than daytime. I inject at 8PM and 7AM.

As with everything about this annoying disease, you'll only know through trial and error. Good luck and let us know if it works for you.


I've used Levemir in the evenings now for several years. It takes a bit of experimentation to come up wiht a dosage that works for you. I use Novo-Rapid before meals,a nd the combination seems to be very effective. As long as I don't forget to take my insulin, I stay nicely in the 5 - 10 mmol/l range.

I think the effect the people are feeling is from the peak of the Levemir activity and its position. Actually the activity curve is like a vector shape. The higher the dosage the longer the duration - and the higher and the later the peak. The smaler the dosage the shorter the duration - and the lower and the nearer the peak. You can see that in the graph for the 0.2 plot in contrast to the 0.6 plot. This experience that the peak is coming earlier with small dosages might cause the believe that Levemir acts more rapidly at lower doses. Well, peak wise it does.

The sphere argument sounds plausible but I have a hard time to believe that insulin is just sitting in a pocket somewhere. Perhaps for one hour but for longer durations I have my doubts. Actually the Lantus does not dissolve quickly because it directly binds to the fatty tissue of the injection site. Levemir in contrast should dissolve with tissue fluid to the blood stream and on its way it will bind to albumin of the red blood cells.

I think the most important information is that 2 x Levemir can give you good coverage at the price of 2 moderate peaks. So it might be necessary to eat something before bed to prevent going low from the peak at night. I have used Lantus too and found that its more even activity curve can not compensate for its deficits in coverage. I even tried 2 x Lantus per day and it was not pleasant. There is no absolute truth here. Ask your endo for sample vials and stick with the insulin that works best for you.

Insulin in an early stage is often helpful to relief the Beta cells from the stress of working against spikes in blood glucose. It is often argued that this can help to preserve beta cells. For example they can focus on cell repairs and reproduction if their production capabilities are needed less frequently.

Newly diagnosed diabetics and T2 often report they are doing fine with one shot of Levemir. But we do not know if their Beta cells have to compensate for the missing basal coverage in the second half of the day. This might put more stress on their Beta cells than necessary in my opinion.

This is fascinating data, Holger, and explains why under my current Levemir dosing regimen I can stay flat through the night simply through a 25% increase in dose at 8PM, but when I use my Omnipod I have to increase my basal rates several times through the night. The Levemir peak takes care of that.

I echo your conclusion: there is no absolute truth: what works best is what works best for you.


Thanks holgar. I had only thought of it from the perspective of normal blood sugars (faint less than 95, posteating less than 120) regardless of endogenous or exogenous insulin. Not from perspective of resting pancreas add much as possible. Raises question does it matter if internal or external insulin source if have excellent control. Internal source could be safer in terms of preventing lows, assuming not on sulfonylurea drugs.

I take 7 units at 7 pm now, and 1-2 units at around 2-3 am depending on how I feel and my bg levels. I had switched from Lantus to Levemir due to weight gain and I haven't noticed any difference between the two, the dose didn't change but I have lowered my basal dose overall. Neither of them has ever kept me stable. I was on one dose of Lantus at night at first and then split that due to dp.

Bernstein argues that Levemir (and Lantus) overnight have an accelerated action due to being cleared faster. This is why he splitting Levemir with a dose at night and a second dose in the morning no more than nine hours later. You won't get the same accelerated effect from a pump.

Thanks Brian!

I had done MDI using 4 times a day NPH for the basal for a couple decades, and was pretty good at that.

I found Lantus 1 time a day, was markedly inferior and less flat than NPH 4x a day.

I find Lantus twice a day, to be similar or maybe a little better than NPH 4x a day.

The TDD of basal in all the above schemes was the same, it was just the absorption timing I was tweaking.

I haven't used Levemir so can't comment there.

Holger showed a graph of how un-flat Levemir activity is, but Lantus activity curve is seriously unflat too. Levemir curve actually looks similar to Lente from the days of old, and Lantus curve looks kinda like the UltraLente curve from days of old.