Okay I heard back from the tech rep at RxFoods. He said the app is provided by Dexcom for anyone using a CGM. You can sign up via Dexcom - RxFood Patient Signup
2 Likes
I tried getting an account, they wanted too much personal information and mandated connection to my Dexcom account. Sorry, I’m tired of every corporation wanting my personal data, not going to happen in this case.
5 Likes
The result was the opposite of what I expected. I had a 4-inch raspberry frangipane (almond paste) tartlet. I made it, so I know the carb count is 46 g per tart. I cut it in half.
The half in a small cereal bowl was analyzed as 48 g carb – twice as much as the actual carbs.
The same half on a large dinner plate was analyzed as 82 g – nearly double the actual carbs!
(While I’m typing this out, a notification pops up: “Information provided by RxFood Co. is for information purposes only and should not be used for medical decision-making.” Got that right!)
Still, I could see how it might be helpful for a restaurant meal, although my guesswork has usually been close enough.
5 Likes
No apologies needed, I absolutely respect being vigilant with one’s data.
That’s some excellent testing! I’d think one could calibrate to find the best plate size.
I have no stakes here and certainly if one has accurate systems for carb counting/estimation no need to do anything else.
For me, a chronic avoider of measuring and tracking, it’s helping. I am honing better guesses. I just looked at my stats 2 years ago seeing my time in range data sent to my doc was in the low 50%s. Since switching in January to a Tandem/Dexcom system and using this app, I am in the high 60%s regularly and had a few 2 week periods above 70%. And I feel better overall.
3 Likes
Please share the recipe!
4 Likes
@Beacher I agree your results seem off. I’d of expected the larger plate would shrink the # of carbs by 1/2 or so. Is it possible the distance from the plate mattered? Alternately, that other items in the frame were taken into account? Of course, the actual carbs for a “home made” pastry are unlikely to be interpreted correctly, depends too much on how much sugar, whether its a low carb variety, how much flour of what type used, etc.; though if you used “industry” standards ingredients there’s no explanation I can think of.
1 Like
Yes, @CatLady, let’s get down to brass tacks…I want some of the pastry too! I “promise” I’ll dose for it!
3 Likes
I’ve used various recipes for the tart shell/frangipane combo, but this one was based on the King Arthur Flour recipe for Bakewell Tart, with a whole bunch of modifications. For the pastry I used all-purpose flour, increased the confectioner’s sugar to 1/2 cup, and used the whole egg. (This makes a nice semi-crumbly French sweet tart pastry dough that’s very easy to work with and doesn’t break up as much as classic high-sugar recipes. It also tends not to shrink in baking.) I used 4-inch tart pans, so had leftover pastry for sugar cookies. In the frangipane I used 2 whole eggs, only 1 cup of almond flour, vanilla paste instead of almond extract, and added about 1/3 cup ground blanched almonds (not additional almond flour) for texture. I didn’t use the jam but pressed 5 or 6 raspberries (frozen, so they don’t squish) into the tops. (If you’re using fresh, spread a little frangipane in the cooled tart shell first so the berries don’t bleed into the pastry, then top with the remaining frangipane.) Baked at 375 instead. (So why even bother with the King Arthur recipe, you ask?)
6 Likes
It was a starting point for you to build upon.
3 Likes
Tried it, deleted it; it requires a Dexcom “connection” to use.
I wish there was something like this that worked and didn’t try to rip me off. If I can prove it works I’m happy to pay, but I need 30 days free use to determine that first.
Hey I am not shilling the RxFood app, but after 55 years of T1D it’s the first time I have stayed with anything on tracking my food. Others do great with old school carb counting, I always dropped the ball.
Between this and my new Tandem T-slim started in January, I’ve gone from the mid 50s for percent of time in range to the high 60s and some weeks in the 70%.
I don’t know the ins and outs of how it works, from what I saw on the info form, DexCom is not required, that’s only for the CGM to be able to send back to the app how my numbers behaved for the 4 hours after. Here’s two recent entries
I have not been asked to pay for anything. I am reading where I signed up here https://rxfood.com/dexcom/patients
Whether or not you have a Dexcom CGM yet, the form below will walk you through the first steps of getting started on your journey to better understanding how the food you eat affects your glucose.
I can’t say what happens w/o a Dexcom cause I have one. Maybe it is limited now for Dexcom users? I dont know. If anyone is keen for answers I have an email contact.
6 Likes
That’s terrific!
4 Likes
I got a notification today that the latest Dexcom G7 update for Android includes an AI-powered meal photo analysis! Screenshots below.
Unfortunately it doesn’t sound like it estimates carbs or anything…
I try not to use AI if I can help it (the environmental impact makes me cringe). For the sake of good old diabetes experimentation, though, I might give it a shot!
2 Likes
I tested FoodRx for a few days and found that for very simple, whole foods, like an apple, it pretty much nailed the carb count, which is reasonable because the difference between a small and large grocery store apple isn’t significant, for most. But give it a mixed food, like a Reuben Sandwich or even a salad with dressing, and I wouldn’t trust the carb advice.
I have also been testing out the new G7 app and although it didn’t supply a suggested carb count it did keep track of BG at the time of the meal and 2 hours post-prandial, along with a photo of the meal and the ability to note details. Others apps have done this but it’s nice to have it incorporated into the G7 app itself. Here is a screenshot - I’ve had to split it into two photos to display the data. Oh, I could have entered my estimated carb count and even the amount of insulin I took for this meal. I could see how this might be helpful for those new to a Dexcom and learning carb counting.
5 Likes
I didn’t notice that!! Thanks for pointing it out! That alone might make it worth using for me, just to see if my bolus strategy for similar meals is truly working.
A few more tidbits. I never expect the app to deduce everything from the photo, I always list as many of the food items I can type in.
For dining out, the newer version of the app lists nearby restaurants which perhaps it indexes menus (?) Wildly guessing but it seems to respond with a menu item name. My wife and I had dinner (carb heavy) at Boston Pizza (wings, small pizza, artichoke dip), and I used a photo of all the plates we would share planning to indicate for each my portion was 50%— but was surprised when it did that automatically - though in the picture you see her plate and hands. Still interesting to me.
In my previous carb guessing mode I would have missed the mark and ended with an evening high, I stayed under 7 all night long.
2 Likes