Detecting sickness

While looking into all that, i noticed, that my watch doesn't sync stress correctly. It shows some blue bars (rest) at times where the app makes them (orange) stress. If you are experiencing same issues, i started a new thread about this. https://forums.garmin.com/forum/into-sports/health-fitness/vivoactive-3-aa/1320677-stress-value-not-correctly-synced-calculated

Well, it is difficult to find something on the stress day chart... but when looking into the average chart for 7 days, is when I noticed the trend. By the third day that I found out the trend, I looked at the stress day chart and no matter I was resting on bed all day trying to recover, the stress chart was all orange... in fact my RHR was also elevated those days. Other days that I saw a higher stress average, I can relate it to a very active day or a day with a 2 hours ride at level 4-5 HR area. It can also depend if you have the watch tight to your wrist, or loose, so I tried to have it tight since the plastic band is kind of elastic.

Fascinating conversation, folks.

Witnessing the impact of illness on your body via your stress data is pretty exciting, and a good sign for the technology. Spotting colds and flu in the early stages - often even before someone realizes they're getting sick - is a pretty common occurrence during our Firstbeat Lifestyle Assessments, which utilizes a professional-grade ecg HRV-recorder. Seeing similar feedback from a consumer product is fantastic, although, I wonder how many people think to put these puzzle pieces (illness and physiological stress).

The thing that really killed my interest in the Garmin "stress" measure was that it is completely, totally different between my VS3 and my VA3, which suggests that the HRV data is not being collected correctly, or interpreted correctly, by at least one of these devices. If not both.

Regarding the interpretation, I'm fairly confident that you won't be able to slap on two different devices and get the same stress feedback - and that's a good thing. As you wear your watch, the analytics engine is learning about you, your patterns, ranges, baselines, etc... and while every device can say 'something' out of the box, as you wear the device the quality of feedback should improve over time. In other words, the interpretation becomes more personalized.

In other words, identical sensor data could be interpreted and displayed differently.

Tricky stuff, but sounds like things are heading in a very good direction, and worth keeping in mind that this type of feedback - which has only been available for a short while - is continually being improved upon.

Regarding the interpretation, I'm fairly confident that you won't be able to slap on two different devices and get the same stress feedback - and that's a good thing. As you wear your watch, the analytics engine is learning about you, your patterns, ranges, baselines, etc... and while every device can say 'something' out of the box, as you wear the device the quality of feedback should improve over time. In other words, the interpretation becomes more personalized.

I can sort-of get behind the idea that the watch needs time to adjust its baseline values, etc. But here's a typical "stress" recording from my VA3, which I've worn 24/7 since January.



I defy anybody to draw any physiological conclusions from this kind of trace. Sometimes there is at least some connection to my heart rate, which suggests that something is actually being measured. But a recording like the one above -- what can anybody conclude from that?

It's practically impossible to make sense of a graphic like that without knowing more about you and your day, but without knowing anything, it looks reasonably sensical to me. That is to say that there is a sensibility to the patterns. It would be a lot simpler if your later sleeping stess/recovery levels were a hair lower - below that 25 mark where which turns them in to blue "rest" lines.

Your levels are lower during the night, although, perhaps not as low as they could be (possibly a low baseline is shifting the whole chart bit higher than normal, idk). A few pick ups and calm downs in the morning. An activity session which resulted in a heightened activated state immediately afterwards... which eventually subsided... some ups and downs in the evening, which seem to drop off as you slow down and get ready for sleeping... possibly some particularly good evening routines that get you ready for rest, because you very quickly get into recovery after your sleep marker.

Of course, lots of speculation there on my part, and your are a 1000000 times better positioned to make sense of it.

You also might just be a particularly "sympathetic dominant" guy, some people are.

I can sort-of get behind the idea that the watch needs time to adjust its baseline values, etc. But here's a typical "stress" recording from my VA3, which I've worn 24/7 since January.



I defy anybody to draw any physiological conclusions from this kind of trace. Sometimes there is at least some connection to my heart rate, which suggests that something is actually being measured. But a recording like the one above -- what can anybody conclude from that?

Too high values compared to my typical charts. But well, I spend most of my work days seated in front of a computer. Like HermanB mentioned, the data itself needs to be correlated to your lifestile to have a meaning. I think is like the voltage value, you really need the "potential differential" from the second cable (ground or neutral), in order to know the value of it. When I first looked at the stress value with my VA3 it did not tell me anything at all, until I started looking for patterns on my feelings.

I've never found the slightest correlation between events in my life, and the stress trace, with one exception: it is higher after really intense exercise. Even in the trace I posted earlier, which is essentially random spikes for 24 hours, the spikes are a little bit higher after exercise. I've heard the same thing about alcohol, although I rarely indulge, so I can't really comment.

My worry is that it's all too easy to find patterns in data retrospectively, if you go looking for it.

I've never found the slightest correlation between events in my life, and the stress trace, with one exception: it is higher after really intense exercise. Even in the trace I posted earlier, which is essentially random spikes for 24 hours, the spikes are a little bit higher after exercise. I've heard the same thing about alcohol, although I rarely indulge, so I can't really comment.

My worry is that it's all too easy to find patterns in data retrospectively, if you go looking for it.

From what I have read, this can also help on understand over-training and finding when is not enough recovery time in order to be more efficient on the exercise.

From what I have read, this can also help on understand over-training and finding when is not enough recovery time in order to be more efficient on the exercise.

Maybe so; but I think I would want to see published results from reputable sports physiologists before I started planning my training on this basis. I'm not saying it isn't right -- perhaps it is. But I'd like to see scientific evidence.