Closed loops and eyes. What are the potential links?

Closed loops and eyes. What are the potential links?
Closed loops and eyes. What are the potential links?

As many will be aware, the NHS is undertaking an England based population pilot of hybrid closed loop systems, to try and determine what are the best outcomes and see if the technology is up to scratch.

It’s being undertaken at centres across England, where patients are being signed up to use one of three systems available in the UK.

  • Medtronic 780G
  • CamAPS-FX
  • Tandem T:Slim X2 with Contorl-IQ

(I know the image includes Omnipod 5, but that’s because I’m just re-using something)

There have been numerous statements on Social Media from those participating in the pilot, talking about the significant reductions in average glucose levels that they’re seeing, and discussions with various clinicians suggests that many of the participants are those who struggle the most to manage their diabetes, in the hope that this will prove out the technology at multiple levels.

There is, however, one question that use of two of these three systems poses.

Both the Medtronic and Tandem offerings have restrictive targets, with Tandem offering a target of 110mg/dl (6.1 mmol/l) and Medtronic 100mg/dl (5.5mmol/l), whereas CamAPS offers a wide range (from around 4mmol/l to 11mmol/l). Now, why might that be of concern?

According to the 2019-2020 National Diabetes Audit, around about 41% of people with type 1 in the UK have an HbA1c of >70 mmol/mol (or 8.5% in old money). That’s an estimated average glucose level of 11mmol/l (198 mg/dl). That’s quite a big difference from the offerings we’re talking about here.

And as soon as you put someone on a commercial APS system with a target like this, it’s going to do its damnedest to ensure that it maintains glucose levels at these numbers.

The question for me here (and others) is “How fast will someone’s glucose levels drop once they start using a commercial system?”.

Why is this important?

Well, if the population that is using these systems fall into those with higher HbA1c levels, we know that a rapid drop in HbA1c can cause significant issues to the eyes, as outlined here. In fact, it stipulates that a >2% reduction in HbA1c is a major risk factor.

What could this mean for those participating in the trials? I assume that this is something that teams are well aware of and have been monitoring as the new users embark on their journeys, and that the significant drops in average glucose levels we’ve seen reported in social media by some participants are accompanied by regular eye checks to ensure there are no major side effects.

But it raises a bigger question on technologies like this. Where these systems may well work for those who have previously struggled to maintain lower glucose levels, why don’t they allow for a slower reduction in average glucose levels?

We know that the lower limit was imposed to meet FDA concerns about hypos, and the costs of trials, but surely higher limits should therefore not be an issue?

Wouldn’t it make sense to set these systems up to limit the possibility of known effects of rapid glucose level drops from the word “go” rather than wait for epidemiological data and wider population trials to see if anything happens? Or is that an issue with the initial clinical trials that were undertaken where the participants all started with relatively low values for HbA1c?

Either way, it certainly raises the prospect of an additional piece of information relating to the outcomes of the NHS England closed loop pilot. One that certain manufacturers should perhaps be watching carefully…


  1. This is exactly why I wrote an article on this very topic. Too many HCPs and DSNs don’t know that a rapid drop in Hba1c causes DMO and other issues. Why do I know this because it happened to me when my Hba1c dropped from 10.8 to 7.7 in 3 months 10 years ago and when I went to see my Optom he said what happened between a perfect OCT and now 3 months on and I said I dropped my Hba1c to 7.7 as encourage to and set up by my DSN. I am actually part of a focus group on this and I get so many stories of this happening and it needs to stop. But because of the work I am doing along with many charities enough is enough and no one should be losing their sight, jobs and driving licences because of mistakes and lack of knowledge of HCPs in the diabetic sector. Don’t worry Tim I am on it. Many conversations already had.

  2. I hadn’t heard about the link in diabetic retinopathy and a1c reduction before. I got LASIK last year and had been worried that my diabetes could make it troublesome for my vision, but I had been closed looping since 2016 and my a1c was already down

  3. When I first went on a pump 4 years ago I was not warked about eye effects. With the pump, no closed loop, I dropped my A1c from around 70 to 45 in about 6 months. My retinopathy got worse and I needed a lot of laser to stabilise my eyes. Would that have happened anyway. Who knows, The main point is I was not given any advice about controlling the speed to reduction and the possible affects.

  4. Hi! Is it realistic to have drop Hb1ac levels like this? I am on a closed loop since more than four years and took a while. Because of switching first to a pump and then later to loop. One remains almost the same eat behavior (like high calories) and insulin activity curve which makes it hard even with drop below 7,0%…

    • It would depend on where you started. If you had a 10% starting point and dropped to 8% as a result, then yes.

    • When I first got my pump my
      Hba1c was 10.8 it was set up so that my levels would be between 5-8 and after 3 months my Hba1c was 7.7 then I lost some of my sight and started treatment of eye injections for DMO which I still have monthly 10 years later.

Leave a Reply to Leo Cancel reply

Your email address will not be published.