Lumen - let the conversation begin 😋

amwassil · 2020-02-18 00:03:03 UTC

Home from work and did a couple of Lumen breath samples. First tested for BrAce with my Ketonix. Although I’m only looking for a yes/no, I’ll mention that my BrAce was 10.4 ppm immediately prior to the first Lumen sample. Result: 3. I did a second test about 6 minutes later. Result: 4. Repeated 15 minutes later. BrAce: 10.7; Lumen: 3. That’s pretty stable for BrAce, by the way.

To give the Lumen the benefit of the doubt, I have been taking Dayquil for a couple of days to suppress my cough at work. It contains HFCS and probably other carbs. So let’s say that’s showing up as the carb portion of my metabolic burn. While in ketosis? That would be interesting. I’m not going to take any more Dayquil as of 4pm PST. I will repeat the breath tests again when I get up a 5am tomorrow. If this is what’s happening, then the glucose should be burned out by then. I should reasonably expect the Lumen to show a 1 or 2.

On the other hand, maybe the Lumen is not actually measuring anything. Instead, it’s performing some esoteric calculation of what it ‘thinks’ my metabolic fuel ‘should be’ based on the incessant questions it continues to ask me. I’m going to start responding to the questionaires with contradictory responses to see if the AI will implode.

Can’t compute…

Please, don’t do that, Dave…

amwassil · 2020-02-18 00:09:19 UTC

I’m going to give it a few days, at least until my current illness is over. I’m sure that’s screwing up stuff. But yes, I am going to write to Lumen support and ask them how to turn off the AI. I just want the device to measure O₂ in and CO₂ out, calculate and display the ratio. I don’t need nor want the lifestyle, meal plans, exercise, and all the rest of the crappola. Just the numbers.

RightNOW · 2020-02-18 01:59:54 UTC

Oh heck. Here I thought this thread title meant there were other hydro garden nuts on my keto forum. LOL

That thing sounds really interesting, if they can turn off the nanny-element.

EZB · 2020-02-18 03:34:48 UTC

Let me know what you find. They haven’t responded to me yet. Tomorrow is some kind of calibration day. I won’t be following the garbage about carbs.

Meerkatsandy · 2020-02-18 03:44:00 UTC

Have you found any other users on youtube or other blogs? I youtubed lumen a few weeks ago and they had mostly “testimonials,” aka advertising it as the best thing since sliced bread.

amwassil · 2020-02-18 04:19:01 UTC

7:45pm BrAce: 10.1 ppm; Lumen 4 (mostly carbs). I ate 7 grams of carbs, 64 grams of fat and 25 grams of protein at supper finishing 2+ hours ago. Lumen says I’m burning mostly carbs. So, assuming the device is actually measuring something rather than just performing math mambo, how could this be?

An hypothesis occurred to me. I’m in ketosis and theoretically I shouldn’t be burning glucose beyond a very low baseline (if at all) producing only a small amount of CO₂ (if any). And most of that small amount of added glucose should be out of my blood by now. So where is the CO₂ coming from that is raising the RER ratio? It might not be from glucose burn at all.

The BrAce readings today suggested a source. When an acetoacetate molecule breaks down it releases a molecule of acetone and a molecule of CO₂. My hypothesis is that the CO₂ released increases the RER ratio. CO₂ is released in equal amounts as acetone, so I’ve got an additional 10.1 ppm CO₂ being measured. So what I would expect is that the higher the BrAce value the more ‘mostly carbs’ or even ‘all carbs’ is what Lumen is going to report.

Does that make sense? Is it plausible? Does 10 ppm seem large enough to skew the results? I guess that depends on the sensitivity of the device.

This might relate to the initial test for calibration being done fasted. Although unless the programmers knew that someone in long term ketosis would have a higher quantity of acetone/CO₂ than someone who’s eating SAD and simply fasted overnight, they might not have allowed for that difference. That might be significant.

amwassil · 2020-02-18 13:51:49 UTC

I took two Lumen samples this morning an hour apart (4am and 5am). In ketosis. First: 4, second: 3. This after 10 hours of fasting since supper (6pm) yesterday. 12 hours after last taking Dayquil. I am starting to suspect that Lumen is not useful when in ketosis. Or maybe RER itself is not particularly useful when in ketosis due to additional CO₂ generated by ketone breakdown skewing the ratio. This is not encouraging.

EZB · 2020-02-18 14:21:04 UTC

So they got back to me when I told them I am a diabetic and eating a zero-carb diet, which makes it impossible to follow their recommendations, since even 20g of carbs would send me to a blood-glucose of 200. Here is their response:

We do not recommend the use of Lumen for people with diabetes for several reasons:

Firstly, the insulin resistance or lack of insulin accompanied by diabetes affects the metabolism and therefore the Lumen measurements and will cause unclear results.
As glucose can not enter the cells and be utilized as a source of energy, stored body fat is mobilized to produce energy.
However, the use of insulin increases fat storage and decreases fat oxidation. If a person with type 2 diabetes injects insulin it will most likely reflect in higher Lumen levels as the cells will metabolize predominantly glucose.
Those are some reasons why the results for diabetic people can be skewed.

Secondly, the nutrition recommendations provided to you might not be suitable for people with diabetes for instance the higher carb recommendations on calibration day. We don’t recommend to use the Lumen nutrition recommendations as a medical treatment for people with diabetes. A medical diet needs to be recommended and supervised by a medical professional such as registered nutritionist or your doctor. Lumen can not replace medical treatment.

If you still wish to use Lumen to track your current metabolic state, I strongly recommend following your regular carb intake on calibration day. You can take the measurements as advised on calibration day.
However, it is important to keep in mind that due to the reasons mentioned above the measurements will not necessarily correlate to your food intake.

When I backed this product, there was NO mention that it would be unusable for a diabetic, who is likely the MAIN customer on a ketogenic diet to track fat burning…

ctviggen · 2020-02-18 16:30:45 UTC

If this actually works, it could be useful. For instance, I’m trying to eat a few carbs after my workouts. How many can I eat? (Ideally, the exact amount I need to, in order to replace glycogen; how much that is, is unknown.)

I have tried eating high stearic acid infused mashed potatoes, and was able to keep in ketosis. But that’s only one measure.

@EZB, that sucks about the Lumen not being useful for you.

amwassil · 2020-02-18 17:17:07 UTC

@ctviggen By my experience so far and the response to the email from @EZB it’s beginning to look like Lumen is not useful when in ketosis.

EZB · 2020-02-18 17:26:51 UTC

It would be interesting to see if somebody can do the carb up for the calibration time and then see how Lumen behaves in ketosis from then on.

amwassil · 2020-02-19 00:44:34 UTC

4pm today BrAce: 3.4 ppm; Lumen 2 (burning mostly fat). This about an hour after finishing my shift at Walmart. I have an idea of what might be happening here and a way to test it.

I’m guessing that at rest or fairly low level of physical activity, there is an excess of acetoacetate breakdown and release of sufficient CO₂ to skew RER upwards. At increasing levels of physical activity, there should be less excess of both acetoacetate and CO₂, hence less skewing of RER. My n=1 so far seems to suggest this. When my BrAce has been 10+ ppm my Lumen readings have been skewed upwards. My last BrAce was much lower and my Lumen reading less skewed.

Although it was not busy during the first 3 hours of my shift, the final 3-4 hours were very busy. So physically, it was somewhat demanding. Since I’m fairly well fat-adapted, I suspect that my energy requirements were being met efficiently. That is, not much excess acetoacetate left over to break down into acetone and CO₂. Hence, the lower BrAce reading. And RER less skewed upwards.

My previous breath samples with BrAce at 10+ ppm and Lumen 3-4 (burning carbs and mostly carbs) were taken at rest. The samples this morning before I went to work were less than 10 ppm, but still above 6 ppm.

How to test? Log BrAce and Lumen breath tests over a period of time to determine if there is a direct correlation between them. That is: BrAce down / Lumen down; BrAce up / Lumen up. Do the sampling at different times and intervals of physical exertion to produce different BrAce ppm results.

I have not done an extensive search of the literature, but what I’ve seen so far has been with athletes and exercised individuals, not at rest. I will continue to search and if anyone would care to do so as well, please post links to any/all studies of ketosis/RER at rest. Thanks.

amwassil · 2020-02-19 00:51:04 UTC

I agree it would be interesting. If someone wants to go for it, great! But if Lumen is actually measuring the ratio of O₂ in and CO₂ out, I don’t think it would make much if any difference. This device was apparently not designed with ketosis in mind, but rather for folks eating SAD or lower/low carb versions of SAD. In other words, there is always ‘supposed’ to be a mix of fat/carb burn. In ketosis we’re removing the carb portion. Maybe that’s causing the AI engine to get confused and inflate the value of whatever CO₂ it’s detecting. Just a guess.

EZB · 2020-02-19 00:52:38 UTC

Even in ketosis, there is some glucose metabolism from gluconeogenesis

Don_Q · 2020-02-19 03:37:01 UTC

No it does not make sense, nor is it plausible. $50,000 metabolic carts don’t seem to have that problem measuring RER. See the study on RER in ketogenic diets linked about 20 posts up.

The only issue I can recall with RER measurements in a metabolic ward was in a fasting study. After like 8 days of water fasting, the subjects started exhibiting RER’s of 69 or something like that, which is not theoretically possible. It was something to do with some strange feature of ketone metabolism in super deep ketosis. If I recall they did the chem math on paper, and made the adjustment in a second column as “adjusted RAR”. But I’m reasonably sure that deep ketosis caused underestimation, which was how they spotted it. I’ll try to dig it up.

Assuming each of the six “zones” of the Lumen, are increments of 5 between 70 and 100, you shouldn’t be above “1” on that meter, but there is no possible way you should be above “2”, unless you’re eating your bodyweight in protein daily.

This does assumes that the 1-6 scale corresponds to an RER value in some way. Does it?

amwassil · 2020-02-19 06:15:35 UTC

@Don_Q I don’t think Lumen is comparable to your $50K metabolic cart. It may or may not actually be measuring and calculating RER directly and/or accurately. I’m trying to establish, using BrAce as a marker whether or not the Lumen is internally consistent with the values of BrAce. This to test my hypothesis and to see if we can actually get any useful results from this device. My initial assumption is that BrAce is somewhat consistent with level of activity. That may or may not be true.

I guess Lumen’s 1-5 numbers corresponds to RER thus: 1=< .76 … 5=>.94 (the range .7-1.0/5).

As I mentioned above all the studies I’ve seen so far were done on exercized subjects, not at rest. That could make a difference, or not. From the Abstract of the study you link above:

The main objective of this research was to determine the effects of a long-term ketogenic diet, rich in polyunsaturated fatty acids, on aerobic performance and exercise metabolism in off-road cyclists.

Also, I want to mention some possible confounders. I have been sick pretty much continuously for about 2 months with cold/flu combo. I had been (but not as since 4pm yesterday) medicating myself heavily with both Dayquil and Nyquil to control coughing. I suspect the illness has been and is affecting my ketone levels, including BrAce, which is currently quite a bit lower than when I was testing several times daily last spring/summer. 20-40 ppm then, 5-10 ppm now. Also, it may still take more time for whatever is in that cough syrup to get fully metabolized out of me. For example, I just took a 2 hour nap, tested immediately afterwards: BrAce 9.4 and Lumen 2. That doesn’t support my hypothesis! So maybe I’m just getting random numbers right now.

@EZB Yes, always some glucose burn. But Lumen value 1 covers all the territory from .7-.76, so as Don says, it should be showing ‘1’ all the time unless we purposely carb up.

Don_Q · 2020-02-19 04:05:24 UTC

I wish! I’ve got the budget model. I’ve been underwhelmed by it’s RER measurements, which is why I am skeptical of the lumen.

The study I linked earlier took a baseline for all athletes at rest. The “ketogenic” subjects blew a .76. Bear in mind though, they had to eat a 15% carb diet in this study (don’t ask me why they call that ketogenic), and given that these were 180lb athletes at 11% bodyfat, I’d guess that TDEE was easily 3000kcals. This would give us 450kcals, which is 112g carbs. I don’t think anybody here is eating 112g carbs!

Here is the table I was talking about where they had to adjust the RQ (same thing as RER but under strict measurement and resting protocols). The chart increases in size for clarity if you click on it. The RQ adjustment was way bigger than I remembered! By day 18 of water fasting, these guys were blowing a .628!!! That’s how they knew the readings from the $50,000 metabolic cart were bullshit. They made the adjustments on paper, and after day 1, these fellas all stayed at ~0.70 for the duration of the 18 days. In short though, being in deep ketosis caused underestimation rather than over-estimation.

Don_Q · 2020-02-19 04:27:04 UTC

I think that makes sense. I’m not trying to bust your chops or belittle the Lumen. My rig sucks for RER too. It had me at .75 while 3 days into a water fast. That’s not even humanly possible. Even the one’s in metabolic wards struggle with RER once deep ketosis comes into play!

I’m stuck with my rig, but you’re not stuck with the Lumen. Either it can measure the RER of people in ketosis or it can’t. It seems like it’s giving weird readings that don’t make sense to either of us. I think the Lumen folks have an obligation to answer questions about whether it’s capable of measuring RER of keto folks, and if not I think they have an obligation to refund your money. Or maybe a software patch can adjust the algorithm for ketosis, I don’t know. I think it’s worth asking them this stuff sooner rather than later if return is an option though. I’m just trying to throw in what little I know about RER, but lord knows I’m not an expert on any of this stuff.

I’m guessing there are plenty of carb burners on the waiting list for whom it might be a better fit, and who’d be thrilled to get a refurb model ahead of schedule. It’s possible the algorithm is optimized for them.

amwassil · 2020-02-19 04:51:47 UTC

@Don_Q Thanks for pointing out those are rest values. Assuming Lumen numbers represent this:

1: =<.76 2: >.76 <.82 3: >.82 < .88 4: >.88 < .94 5: >.94 =<1.0

I want to admit that my first sampling was a total shock! How could I possibly have an RER >.82 and/or >.88 when I have consumed sub-15 grams of carbs nearly every day for the past 3+ years and many days only 10 grams or less! The Ketonix measurements confirmed ketosis. So either the device is junk or there’s CO₂ coming from somewhere other than fat burn. Before concluding the Lumen is junk, I want first to eliminate the alternative possibilities. So, where could other non-fat burn CO₂ come from? Acetoacetate breakdown. Dayquil/Nyquil and/or Fisherman’s Friend.

I’m going after acetoacetate breakdown first because I think that’s the least likely source. 10 ppm is a very small amount. I’m plotting sample results on a spreadsheet and already, it’s looking somewhat random. Maybe it will look less random with more samples. We’ll see.

I’m hoping that the combination of being ill and consuming overdoses of OTC meds to keep myself able to go to work will ultimately prove to be the source of the skewing. My last previous sample after my nap earlier supports this possibility. I had a powerful ‘break’ in the current illness yesterday and overnight. I feel a lot better today than in more than a week. Maybe when I get over this, the Lumen samples will go to 1 and stay there. I certainly agree with you that’s where it should.

amwassil · 2020-02-19 05:15:23 UTC

The question of the moment!