One of the questions that is repeatedly asked in the various Facebook groups relating to DIY APS is “Will I gain weight?”, to which there are generally a load of responses that say… “It depends”.
As I started the year at getting on for my heaviest weight, with 106.7kg (235lbs/16st 11lb) showing on the scales after two years of soo much food, drink and being sedentary, it was time for me to get rid of some of it. Time to go from fat to fit.
In summary, throughout January I’ve gone from 106.7kg to 97.6kg ( 214lbs/15st 5lb), a loss of 9.1kg, whilst looping, so it can be done, and it turned out to be easier than I expected. And if that’s all you came to find out, then there’s the answer.
But if you are further interested in how, use of an APS during this and the effect it has on other markers, read on…
I decided on principles similar to across between the Newcastle Diet and Low Carb, in other words, a low carb, calorie restricted diet, with lots of exercise. My exercise choices were based on previous training I’ve done with various trainers and discussions I’ve had with them.
As a 6’5″ bloke, most of the calculators you can find reckon that maintenance calories for me would be around 2500 a day, so the first thing required in that was to decide on a calorie deficit, that would kick things off. I decided that 1,600-1,800 a day (roughly a third reduction) should be okay, and from Jan 1st, that’s been the daily intake. In addition, I’ve taken advantage of “Dry January”, giving my body a chance to recover from December and stop my liver from being distracted.
I’ve lost weight effectively in the past when eating Low Carb, high fat, reasonably high protein, so it made sense to continue with this approach in this case. This generally means less than 30g carbs per day, around 100g-150g protein and the rest of the calories in various fats. I’ve also tried to maintain a reasonable level of veg in the diet, supplementing with salad leaves and spinach, amongst other things. We’ve then spent Sunday afternoons cooking food that can be packaged up and taken to work so that I know exactly what I’m eating and how the macronutrients fit together. We’ve used the Weight Loss Resources website to help with this as it allows you to create recipes and understand the macronutrients. An example is shown below.
With my diet and intake decisions made, it was on to exercise, and the decision here was a combination of resistance training (weights usually two or three times a week) to maintain or improve muscle mass, “26 + 2” hot yoga (probably best known as Bikram Yoga – again two or three times per week) for both cardiovascular and muscle mass benefits, and additionally, spinning, which was again once or twice a week, purely from a cardiovascular benefit.
The intention was to do some form of exercise six days in seven, and I’ve managed to maintain this. I’ve also tried to maintain more than 10,000 steps per day, to give a bit of extra edge to the exercise story.
All the while this was being done, I’ve been running OpenAPS as my DIYAPS to maintain everything.
As mentioned in the introduction, my weight has dropped from 106.7kg to 97.6kg. That’s a drop of 9.1kg (21lbs/1st 6lbs) in just a month, which shows that there was plenty to lose. It has also resulted in a change in waist size from Gap 38 to Gap 36, Gap being critical because I think their waist sizes are perhaps slightly on the larger side…
Resting Heart Rate
In addition, my resting heart rate has gone from around 60-62bpm to around 55-56bpm, which is a good indicator of cardio health.
According to Fitbit, I’ve averaged 11,867 steps per day, and burned an average of 3,933 calories per day (a calculation that I believe is a little out).
At my diabetes clinic, early in this process, my cholesterol showed 5.9 as the number (where above five is considered a risk), and we’ll be checking the full lipids profile in a couple of weeks to see where that’s at.
From a physical perspective, the outcomes for one month I consider to be pretty good, and I seem to be fitter.
Obviously, when losing weight and exercising hard, there are going to be implications for my diabetes.
Basal Rate changes
At the start of the programme, my basal insulin profile was 55.9 units. At the end it’s 26.4 units. That’s a reduction by more than half, which is quite incredible. The two profiles overlayed on top of each other can be seen below. What’s perhaps quite interesting is that the calculated insulin sensitivity has hardly changed, going from 3.5 mmol/l/u to 3.7 mmol/l/u.
It’s interesting to see the two, and note the variation. What I’d also note is that as my needs have dropped and I’ve moved onto a lower carb diet, the profile looks a lot more like the traditional circadian profile, which raises questions over whether all carb absorption is fully accounted for as carb absorption in using Autotune, or may be being picked up in basal adjustment.
My total daily dose for the week preceding the programme was 68.4u while in the last week of the programme, it was 27.2u, which correlates with the drop in basal insulin required.
Time in Range
What’s also interesting to look at is the Time in Range (TIR) and percentile charts. Using the clinically agreed version of TIR, ie 70-180mg/dl/3.9-10mmol/l, we can see that the combination of low carb, high exercise has resulted in a TIR of 94.5%. I put that nearly all down to the low carb diet rather than the exercise or calorie restriction.
It’s worth calling out the 3.5% of time spent above 10mmol/l. These have all occurred when I’ve taken the pump and loop off for the hot yoga, and in a significant number of cases have been the Dexcom G6 reading significantly higher than blood results, as detailed here. Suffice to say that if the Dexcom was accurate during the hot yoga sessions I’d estimate that that number would be nearer 1%, and this can be seen in the spike that occurs between 8pm and 10pm on the Percentile Report graph.
For comparison, the prior month, which included Christmas had a TIR of 84.6%, with an average carb consumption daily of 237g, which was still a pretty incredible outcome.
The other thing to look at is the percentage of lows. They’ve mostly been linked to aerobic exercise, and seem to demonstrate that even when you are, in theory, in ketosis (when the urine strips show that at least), spinning still activates any insulin on board rather effectively, and means that care is needed in the run up to that type of exercise.
Thoughts on DIYAPS use while doing this
As I’ve already mentioned, I use OpenAPS. When I’m doing something like this, it’s extremely useful, as it runs Autotune overnight and has adjusted my basal insulin levels down throughout the month. Personally, I think this takes the risk out of the reduced need for basal I’ve experienced. If you weren’t able to use a system like this, it would be up to you to monitor the changes that were going on and adjust accordingly, which would have been a painful task in this case. I know that this doesn’t run true for everyone though.
It’s also helpful with aerobic exercise, as the “Exercise” function allows you to reduce basal needs ahead of starting the aerobic action, without needing to do more than set a temp target, which is simply an IFTTT button in my case.
Overall, throughout this, the glucose levels have remained reasonably flat and steady, which also helps with the weight loss.
And in answer to the original question…
The takeaway from this month long observation is that “Yes you can lose weight on a DIYAPS system“.
The reality is that you can do it on a pump or MDI as well if you choose to.
The key to this wasn’t whether I’m attached to an APS, a pump or giving myself insulin via MDI. It was making the decision to do it and sticking to that decision. I think that a low carb approach helps as there is less risk of hypo and therefore having to treat them, which reduces the amount of “top-up” food that is needed, but overall, it’s about not cheating and sticking to the plans.
Great job well done Tim. I suspect there’s a very good correlation between long term health and insulin useage for us type ones. The best possible BG control using minimal insulin is the way to go.
Great piece of work Tim – huge amount of interesting data on show. What are your plans from here in relation to the above?
Very inspiring! I’m going to start a more modest program and see where it goes. Thanks for the great article.
Fine achievement Tim, and thank you for another great post.
Just for fun, how far from accurate do you think your Fitbit would be when it reported 3,933 calories burnt / day?
As a rule of thumb, Mayoclinic says you will need a deficit of 7,700 calories for each kg of weight you lose.
7,700 x 9.1kg = 70,070 calorie deficit during your January, which was 30 days long (2nd-31st) = 2,336 cal/day deficit.
2,336 + 1,600 cal/day ingested = 3,936 cal/day burnt, implying the Fitbit algorithm underestimated your daily burn by a huge 3 calories. Sounds like an acceptable margin of error to me.
Thanks for doing the maths. It’s certainly pretty close, however, I think the average consumption is nearer 1,700 than 1,600 per day, so I’d argue it’s still off a bit. Having said that, it’s a pretty good estimate.