Initial observations on the Freestyle Libre

I know there are a lot of Libre threads already. I have a few bits to condense into a single post in the hope that it will help someone considering their options for the first time.

I'd looked at the FreeStyle Libre write-ups about how you can use the Libre with your phone and monitor your blood glucose practically in real-time. Just what I need, I thought. I can eat a known-weight test dose of various foods with different carbohydrate loads and see which ones stay under, say, 10 mmol/L and come down to a baseline within two hours or maybe a bit longer.

Edited to add a warning - the following paragraphs are mistaken. The Libre App will in fact start a new sensor, as subsequent posts make clear. This text is retracted:


So I uninstalled the FreeStyle LibreLink Diabetes app and installed Glimp instead, which has a free app to send the Start signal and then offers a perfectly good monitoring interface. Well done Glimp, I'd have binned the sensor otherwise.

I made a Dropbox account and linked Glimp to it, and I can now see on my PC the spreadsheet of the blood glucose readings it's registered. They're stored in raw mg/dl which convert to mmol/l when divided by 18.0182.

Each sensor consistently reads higher and lower as blood glucose goes higher and lower. Using four or so blood-stick readings provides a conversion to bring the Glimp display into line with what the blood-sticks say. That, in my opinion, just puts your blood-stick meter's bias straight into your Glimp readings, but hey-ho people want consistency and that's how to get it. Me, I want to see the line float up and down, that's all.

I'm quite content to buy just the one sensor. Within two weeks I'll have trialed 50g doses of anything which has carbohydrate in it, from cane sugar to rice (white), rice (grim), spaghetti, potatoes, carrots and bread (various), one dose at noon and another at dusk. The shape of the resulting curves will dictate my diet for the next decade, all for £50 quid.

@Binary - I've just been back to the Freestyle website and checked, and you're quite right. I didn't read enough of the manual.

Here's what I saw...

... at which point my eyes misted with fury and I looked no further. That wording is atrocious and I make no apology to Freestyle for getting it wrong. The manual goes on to directly contradict the bit I just quoted.

The net effect of the additional text I didn't find is that you CAN start the sensor with the App if you only intend to continue reading with the App.

Here's my first day's result.

My sleeping BG is pretty steady around 5.5mmol/L, I woke and registered a blood-stick reading which is a bit lower than the chart. The morning involves quite a bit of walking - 3km or so - and at noon I take my first carbohydrate hit in three months, an apple and half a wholemeal bun. The gradient up shows it has to work to get the carbs digested, it's all over in 2.5 hours, the rise is 4 mmol/L, 72mg/dl, not a bad result for 33g carbohydrate. My ketosis is definitely over for the next fortnight.

4pm I try a Costa sweetened pineapple smoothie, pure glucose onna stick. The upward gradient is practically vertical, 5 mmol/L for the same 33g carbohydrate dose but instantly assimilated, the same 2.5 hour recovery.

The third peak is from 8pm, it's 170g of banana with close to the same carbohydrate load. Far less effect as it digests more slowly.

I'll try root vegetables tomorrow.

Hi @DunePlodder - no insulin or any other form of medication. Both sides of my ancestors used it though, and getting overweight while living on sandwiches was just plain silly.

I had cheese and 40g of biscuits last night and surprisingly it acted much the way the banana did earlier. Today I'll try cookery.

This is a lesson in why you shouldn't stay up until 1am and then eat 40g (five) Jacobs crackers with some cheddar while watching James Burke.

@rom35 I'll do that but it might be early next week, I'm away for the weekend. The BBC has a recipe for "Next level spaghetti carbonara" and tips the ingredients into any UK supermarket account's basket so that was simple, I'd not seen the option before.

Once this two week experiment is over I may well go back into ketosis for a couple more months. I have a strange suspicion it sharpened my memory and thinking. I'll be traveling abroad for a month after that so the option disappears.

That slow overnight digestion of just five crackers stands in contrast to a proper time of day, when a total 450 calories (32g carbohydrate) of soup and a wholemeal bread roll is fully dealt with in two and a half hours.

If I may I'll update this thread, after a third visit to the Diabetes Nurse. The gap between January's and now is five months.

I'm advised all my blood test results are now within her acceptable range, my liver is no longer fatty, my BMI is 23 and I should stop my dieting before I appear fixated. She'll see me again at Christmas.

It's an improvement but I don't feel out of the woods. It's possibly as good as I could hope for though. Much of the improvement is a consequence of my joining this site. Thank you everyone.

I implanted a Freestyle Libre yesterday afternoon and it showed only minor variation over a 24-hour baseline on just salad and yoghurt.

Today I've tried hard to get as big a reaction as I can manage.

At noon I ate 100g rolled porridge oats microwaved in 350ml whole milk for 5 minutes: 74g carbohydrate in 592 calories.

The chart is a screen capture from my phone running Glimp. The porridge excursion lasted four hours, which may be because it's a slow release carbohydrate - I hope that's what it is.

At six thirty this evening I had battered cod and chips from the chippie: 110g carbohydrate in 1,500 calories, as best I can guess.

As a jump away from five months' keto it's dramatic. On the other hand my trousers are now UK 34 instead of UK 42 and I have no intention of buying replacements for the new ones, so I don't intend any further tests like that.

This second rise above base level lasts a huge seven hours, again with a slow release carbohydrate.

Neither of the meals exceeded a 5 mmol/L rise. I find that a bit of a relief, staying below a peak 10 mmol/L is better than I'd anticipated.

What I'll try next will be smaller quantities of swifter acting carbohydrate and see whether the rise returns to a base rate within two hours.

Is the long duration on these charts unacceptable, or is it just what happens with large slow-release carbohydrate loads like that?

@Norman Stanley - your results will be interesting to see. Doing comparative meals while avoiding carbohydrate delayed my attempts. I'm finding my results quite different after losing weight, the peak is far closer to the baseline than it was previously. I think I know how to reduce the duration of each raised glucose level too - eat smaller portions!

Has anyone ever described the horror of a Cornish Cream Tea? I was in Old Town on the Isles of Scilly yesterday and the graph showing my response to two scones with jam (and cream) is a disgrace. From memory - I can't look at it much less post it here - it peaked at 13 mmol/L. A brief invocation of Gordon Bennett wouldn't be out of place.

I'll post the screenshot tomorrow when I get it transferred off my phone.