Managing Type 1 Diabetes
metabolism | sports & fitness
Last year Alex Collins was diagnosed with Type 1 diabetes. Prior to his diagnosis Alex was frequently engaged in different types of exercise and physical activity. After his diagnosis his doctor mentioned that he might have a hard time exercising and controlling his blood sugar to prevent hypoglycemia. In this talk, presented at the London QS meetup group, Alex described his process for tracking and understanding the data that affects his day-to-day life so that he could live his life normally without a high risk of complications.
In August last year I was diagnosed with Type 1 diabetes which is a life changing event. So just a quick summery for you who aren’t familiar with the condition basically what Type 1 diabetes does is it kills off the insulin production in your body. So in order to absorb sugar into your bloodstream which you get from any carbohydrates you need to inject insulin into yourself. So basically I have to do the job of my pancreas and if I don’t get it right and my blood sugar’s too high in the long run I have a high level of suffering from horrible complications of blindness and heart disease and strokes and stuff like that. On a daily basis if my blood sugar goes to low that could be dangerous and collapse and lose consciousness and some people die from those so it’s a serious thing.
So exercise is a challenge for people with Type 1 because of this hypo thing, so when the human body exercises all the cells in the body become much more sensitive to insulin that’s in the body. So for a normal healthy person, if you go for a run your body will stop producing insulin and your blood sugar will stay stable. Whereas for me because I inject insulin once a day which is only in my system for the whole day, when I start exercising I become more sensitive to the insulin and my blood sugar drops. So that’s a risk.
And on the first day that I was diagnosed I said can I do exercise and he said it’s going to be a lot more difficult. And actually a lot of doctors will say to people just don’t do any exercise for a while.
So these are three photos of stuff that I like doing. So one of them is kind of ski mountaineering where you’re in remote places far away from any medical help and it’s dangerous. Another one is climbing; same kind of thing. And another one is running, and I like running away from civilization.
So you can see I’ve got a pretty big motivation to learn to control my condition. I actually read an article about this event in the Times or the Telegraph or something a week after I was diagnosed, and I immediately started coming to this because I thought it would be useful to learn how to control my condition.
So my aims basically are to be able to do the things that I love; all those activities and also to control my blood sugar so that I can live normally without a high risk of complications in the years down the line.
So QS, very helpful. I collect lots of data all the time so I do blood sugar tests several times a day. I have to do them before I eat. I often do them a couple of hours after I eat and I do them around exercise, and that’s to check that I’m managing my blood sugar correctly.
So I generate all that data and that’s the data that I’m trying to kind of get right. so I’m trying to get those numbers to be between four and eight and that’s just a proportion of sugar in my blood.
So the main variables for me that effect blood sugar are how much sugar I eat, so that’s grams of carbohydrate in my food. So everything I eat I know or I guess how many grams of carbohydrate are in it, and then I inject insulin before every meal to allow me to absorb that. And then exercise as I explained makes you more sensitive to insulin so that’s another very important variable that affects my blood sugar.
So I record all of this on a mobile app, and I’m very thankful I got diagnosed now and not 20 years ago before mobile apps, because doing it on a bit of paper would be a real pain in the arse. So these are just a screen print of my app. You know home page, history and I generate all my data points there. I generate 20 or 30 data points a day.
The commercially available apps for things like Type 1 diabetes are very useful. You can customize to an extent what data you put into them. they give you some simple analytics like graphs but they don’t do everything.
So what I do use it for is I can see stats about what my blood sugars been doing in recent history. So this is my average blood sugar per day going back one reading a day for a month and I’m trying to get it between five and six, so that’s the kind of ideal range.
I also use it to kind of give me high frequency information about how my medication and food are linked. So every night I look at two bits of data; one how many grams of carbs I’ve eaten that day, and then two, what medication I’ve taken.
So this is my thing that I take once a day, so that’s always the same, and this is my insulin thing that I take before every meal. So all I do is every day add up how much insulin I’ve taken, how many carbs I’ve eaten, divide the carbs by the insulin and then that’s my sensitivity to carbohydrate. So at the moment I’m taking 1 unit of insulin per g5 grams of carbs I eat. Two weeks ago I was taking 1 unit of insulin per 60 grams of carbs I eat. So it varies quite a lot and that’s why I get these high frequency readings from the app.
Before I talk about this, so QS is you know you generate this data and you can be very scientific about it but there are limitations to how powerful you know data analysis is. So one really useful thing about the app and because I’m looking at it and getting daily feedback about how my body reacts to food or exercise or insulin I can develop a pretty good intuition about how to manage my diabetes. So the app is good for that because it gives me the high frequency feedback and it reminds me what I’ve done that day and I can reflect on that in the evening and get better at managing my condition.
None of the less I knew as soon as I was diagnosed I became very curios as I imagine most people here would be about how the condition works because I wanted to solve the problem.
So I used Excel to get a much more customize look at the data. So you can use it for analytics and this is kind of a fan-chart of my blood sugar. So 50% of my readings are between these red bits, 90% are between these and that’s my average, so that kind of tells me how I’m doing.
When I first started do it I thought I could do a multi-analysis and kind of come up with this incredibly complex and clever algorithm that tells me exactly how much insulin to take based on what exercise I’ve been doing in the past, how hot it was, how stressed, all this kind of stuff.
It can’t do that but you can kind of get these different graphics that help you figure out what’s been going on. So this one is actually a really powerful one. This green line is my sensitivity to insulin. So here I was on a two week ski holiday and I became much more sensitive to insulin and if anyone has Type 2 diabetes, do lots of exercise and your blood sugar will go down. It really comes through clearly in all my data.
So again, these are just different graphs. This is average glucose when I’m doing lots of exercise. I was curious about what effected my glucose levels when I woke up, so I plotted different things on a chart, have a look at it and draw some very simple observations from looking at the stuff.
One of the things about self-tracking for me is that I’m very very motivated to do it because I’ve got diabetes, but none the less there is a limited time in the day so it has to be as simple as possible.
And I started off by doing a very complicated VBA program but it was too costly for me to run, so now I just use a spreadsheet and that gives me stuff.
I don’t do anything scientifically. I don’t do proper hypothesis testing or anything like that because the data is not high quality enough and I don’t have enough of it.
I won’t talk about exercise. I just learnt through trial and error of how many grams of carbohydrates I had to eat depending on what exercise I’m doing. So I’ve got different rules of thumb and I count up to 40 grams if I’m running or 60 gram if that’s how much I need as I’m running.
And actually talking about being motivated to do stuff to collect the data, since I was diagnosed I started running to see how my body reacted to exercise and because I wanted to get more and more data I was running for longer and longer, and ended up running an ultra-marathon in December which was 50 miles.
The disadvantage of using a phone is I get distracted, you know, you check your phone at work and you put the data on the phone and then you’re reading a personal email and that’s annoying.
It’s been really successful and I’ve learned a huge amount about my condition through both digging into the data with Excel and kind of trying to spot trends in what’s happening and so the daily feedback that I get from all the records that I do.
So QS has had a positive impact. I do everything that I used to do, so all that stuff I showed you on the pictures I still do and I still love doing it, and I got a Guinness World record last week end. I was the fastest person to ever run the marathon in an animal costume.
I’ve done loads of different analysis and I don’t have time to show much here. But at the moment I’m in what’s called a honeymoon phase where I’ve still got some insulin production in my pancreas and my immune system is killing that off, but for the time being I do still produce some of my own insulin. So it means that my blood sugar has a certain amount of or my body has a certain ability to control my blood sugar but not enough. 13:13