Project Progress

Completed and upcoming experiments are listed here. Follow the main page for the latest updates.


  • Measure basic properties of Afrezza and the FreeStyle Libre glucose monitor, so that total replacement of mealtime insulin injections and finger blood glucose tests can be performed. Benchmark the Freestyle Libre against finger strip testing. (July 16-18)
  • Find the optimal dose of Lantus basal insulin to use with Afrezza. Examine the effects of increasing and decreasing the dose of Lantus. (July 19 – August 22)
  • Investigate if Levemir basal insulin can outperform the dose of Lantus identified in the previous phase. Identify the optimal dose of Levemir and examine what practical differences exist between Lantus and Levemir. (August 23 – September 10)
  • Assess an experimental strategy attempting to switch from Levemir to Tresiba without days of high glucose. Identify the optimal dose of Tresiba, dosing once-daily at a time that can change each day. Compare how the optimal dose of Tresiba performs with Afrezza, against the other two basal insulins. (September – October)
  • Attempt to exclusively use the FreeStyle Libre to confirm if finger pricking blood tests can be completely replaced. If this is not possible, identify the extent that drawing blood can it be replaced. Results

In progress:

  • Identify if it is possible to surpass the HbA1c results of early US Afrezza users – without using a basal pump, or a CGM. Results
  • Work out if an already low starting HbA1c of 6.5%, can be improved much further – or is there a practical limit? (My starting A1c is equal to the lowest starting A1C of Afrezza users publishing their results online) Results
  • Using regular peak flow measurements, identify if there is any change in lung performance over time while using Afrezza. Results 
  • Observe if HbA1c results have improved at an interim test (late November), followed by another test at least six weeks after this test. Continue with HbA1c testing at regular intervals. Results
  • Conduct the Glycomark 1,5AG historic glycemic variability test, and evaluate the results in the context of the HbA1c result.
  • Continue using the optimal basal type and dose identified in the previous phases. Establish the optimal way to use Afrezza optimising for multiple parameters – user simplicity, minimise number of daily Afrezza doses, tightest glucose control and avoiding hypoglycaemia. Perform this for all typical scenarios – small/large meals, low/high GI and high fat foods, small/large corrections with/without a meal, extended multi-course meals.


I would like to continue the project and will conduct as many of the following additional experiments as I can afford to. Please see the donate page if you’d like to help make these possible:

  • Compare the effects of a single large dose of Afrezza against multiple doses spaced at varying time intervals. Identify optimal usage scenarios for both strategies.
  • Examine the interplay of low and high intensity exercise while Afrezza is active, as well as between meals on Tresiba.
  • Extrapolate the results of previous phases to generate user-friendly principles to follow for the optimum use of Afrezza using a FreeStyle Libre, and finger pricking testing.
  • Rigidly follow the previously generate usage principles to benchmark their usability and their effect on glucose control. Identify potential practical problems and refine the usage principles if required.
  • Experiment with using Tresiba 6 times per week, introducing a totally needle-free day. Ascertain if using Tresiba every second day is possible with Afrezza, and compare the performance against daily dosing. This strategy would cut the remaining injections down to just 182 per year.
  • Enumerate all of the practical constraints when living with Type 1 diabetes using previous treatments, when compared to a non-diabetic. Evaluate the optimum treatment protocol identified in earlier phases against each of these constraints. Use this to objectively quantify the quality of life improvement and identify the constraints that remain.

Wish List:

The additional cost of obtaining and running a CGM on top of the high cost of Afrezza makes the experiments completely beyond our means. If sufficient donations are received, they can be performed:

  • Obtain and use a Dexcom G5 continuous glucose monitor to investigate if glucose control is further improved with a CGM. Identify what the optimal alert settings are for Afrezza, and compare user simplicity of this strategy against the FreeStyle Libre. Are there practical consequence of sacrificing accuracy for alerts?
  • Using the Dexcom, further refine the protocol to attempt a sustained HbA1c below 4.9%, while simultaneously attempting to minimise hypoglycaemic events. Enumerate what additional constraints on the previously identified optimal protocol are required. Evaluate these new constraints against the potential impact on quality of life.