Thank you! Is it possible to track night HRV (from Oura) and morning HRV (by Polar H10) in one HRV4Training app account? (or it is only possible to have 2 apps and accounts)?
I believe the point you are trying to make is that effective recovery should be measured as a continuous time series, and that measuring it only overnight introduces some bias.
However, there are logical inconsistencies to your post. You claim: "When eating later or relatively close to bedtime, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just digestion."
Why couldn't I apply this "logic" to any other thing that increases my heart rate, and lowers my HRV. For example:
"When drinking alcohol relatively close to bedtime, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just alcohol metabolism."
"When becoming ill with cold or flu, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just your body's response to illness."
You go on to argue that in order to measure our resilience, we should "measure far from stressors". So.. as long as we recover from a stressor, that's evidence, in your mind, that we responded well to the stressor?
If I measure far away from the stressor of late night alcohol, for any individual, they will recover. Is this evidence that no harm was done?
hi Steve, thank you for your message. No, I do not think a continuous time series is beneficial, especially when it comes to HRV, due to the many artifacts, technological limitations and oversimplification made in terms of interpretation of the data (see here: https://marcoaltini.substack.com/p/a-quick-note-on-continuous-heart)
When it comes to measurement timing, I think the measurement should be taken far from stressors but within a meaningful time window of course, typically a few hours after the stressor (research shows that even for a hard workout, that's how long it takes to bounce back, with shorter times required for more fit individuals). High alcohol intake, sickness, etc. - all leave a mark for much longer as these are strong negative stressors (you certainly don't need a night measurement to see their impact on your physiology, if you have poisoned your body in the evening, then 10 hours later your heart rate will still be elevated outside of your normal range, for example).
My statement about digestion is there to highlight that at times people are quick to associate a lower HRV with something negative, but there is nothing negative in digesting. If you measure your heart rate and HRV while you have a good time with your friends, it's a similar story. The goal is not apathy (never stress the body so that HRV doesn't reduce). Hence my recommendation to simply measure as we've done for 50 years, with a standardized morning routine, which typically helps us avoiding to overemphasize negatively late stressors the way often people end up doing with night data.
A re-normalization within a few hours (e.g. after sleep) means indeed that we responded well, that's typically how we look at various stressors or training specifically (i.e. the lack of a chronically lower value over time, as we e.g. repeat exposure to a stressor without having fully recovered). What I talk about is a re-normalization by the time we take a daily, morning measurement (maybe I was taking this for granted as this is how normally people use HRV).
I hope this clarifies some of the points I was trying to express in the blog above.
I appreciate your thoughtful response and the insights you’ve shared. However, I respectfully disagree with your assertion that "there is nothing negative in digesting." A 2020 study (https://pmc.ncbi.nlm.nih.gov/articles/PMC7215804/) found that eating within three hours of bedtime led to more frequent awakenings and disrupted, lower-quality sleep. This aligns with my personal experience: eating late at night reliably reduces my total sleep duration and deep sleep, while increasing time awake as tracked by wearables.
You suggest that as long as HRV normalizes by morning, there’s no significant consequence. However, while quick recovery might indicate resilience, it doesn’t negate all potential harms. The body strives for homeostasis, but achieving it doesn’t necessarily mean a stressor was benign.
For example, using the same logic:
I could eat unlimited candy, and as long as my fasting glucose is normal by morning, I’ve "fully recovered."
I could drink heavily each evening, and as long as my blood alcohol content is below the legal limit by morning, there’s "no problem."
In both scenarios, the timing of measurement conveniently obscures potential harm. By focusing solely on recovery metrics, we risk ignoring the cumulative and long-term effects of behaviors that may still impact health.
I hope this offers a constructive perspective, and I appreciate the dialogue.
thanks Steve, I understand what you are saying here and agree that not only the recovery but also the reactive phase can be informative under the right circumstances, and while knowing the context.
Here I'm trying to find a balance between the obsession with continuous tracking (and the many artifacts involved, false positives, psychological implications linked to data that is often just misinterpreted, etc.) and 'looking at the big picture' when using metrics that have obvious limits (like HRV) but are marketed as some sort of 'perfect 24/7 stress index'.
HRV can respond with an increase or a decrease to a certain stressor, and can respond in the same way to positive and negative stressors, hence I frame it in a way where many of these aspects are (in my view) handled better, avoiding an overly reactive (and often misguided approach). While it is true that achieving homeostasis does not signal that previously there was a negative stressor, the contrary is also true, when looking at this data. We can only see changes, not meanings associated with these changes (as I've written elsewhere, otherwise we can stop eating altogether, and our HRV will be as high as ever - but that's hardly a good thing, is it).
This does not mean that there isn't a different use when data is properly contextualized, but in my opinion less data is more helpful when it comes to this specific application (I am not sure if you have used morning measurements in the past, but many of the stressors mentioned are captured easily: https://pmc.ncbi.nlm.nih.gov/articles/PMC8659706/pdf/sensors-21-07932.pdf ). Given the terrible quality of optical measurements (10-20% of data is garbage even during the night, while 70-80% of the data is garbage during the day), it's too easy to get fooled by wearables and associated estimates (as well as by the complexity of our physiology, back to digestion or lack of food intake), hence my concerns and 'push' for intentional measurements using good protocols.
If our habits are not good, they will quickly impact our recovery metrics
Thanks again for your input, I do appreciate the dialogue as well, it gives me food for thought and will certainly be part of my thinking for future writings. All the best
Yes I have used Morpheus training, which relies on morning HRV. I do find that it is more accurate in predicting training capacity.
Eating late at night does reduce sleep quality, on average, for me. Reduced sleep quality does, on average, reduce morning HRV for me, though not as much as what wrist based optical monitors suggest. I appreciate that this is just my own reaction, and others might be less sensitive.
Hi Marco, as the data is there in Oura’s feed about the measurements of HRV over the night, and you have worked with Oura, can’t you pluck out the morning reading (or ask them to add it to their API)?
hello Shane, morning data is different from what you normally have during the last part of the night, which might also be particularly messy due to the impact of REM sleep (which includes very high variation in autonomic activity). Body position will also differ. You can use the ring for a morning measurement, with this workaround: https://www.hrv4training.com/blog/using-the-oura-ring-for-morning-hrv-measurements - even though it's probably more practical just to measure with HRV4Training directly.
Great Insights marco , well done in explaining HRV measurement in a nutshell <3
Thank you! Is it possible to track night HRV (from Oura) and morning HRV (by Polar H10) in one HRV4Training app account? (or it is only possible to have 2 apps and accounts)?
thanks Oleg! Only in two separate apps.
I believe the point you are trying to make is that effective recovery should be measured as a continuous time series, and that measuring it only overnight introduces some bias.
However, there are logical inconsistencies to your post. You claim: "When eating later or relatively close to bedtime, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just digestion."
Why couldn't I apply this "logic" to any other thing that increases my heart rate, and lowers my HRV. For example:
"When drinking alcohol relatively close to bedtime, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just alcohol metabolism."
"When becoming ill with cold or flu, your heart rate will be higher and your HRV lower, impacting the night averages. This is not good or bad, it is just your body's response to illness."
You go on to argue that in order to measure our resilience, we should "measure far from stressors". So.. as long as we recover from a stressor, that's evidence, in your mind, that we responded well to the stressor?
If I measure far away from the stressor of late night alcohol, for any individual, they will recover. Is this evidence that no harm was done?
hi Steve, thank you for your message. No, I do not think a continuous time series is beneficial, especially when it comes to HRV, due to the many artifacts, technological limitations and oversimplification made in terms of interpretation of the data (see here: https://marcoaltini.substack.com/p/a-quick-note-on-continuous-heart)
When it comes to measurement timing, I think the measurement should be taken far from stressors but within a meaningful time window of course, typically a few hours after the stressor (research shows that even for a hard workout, that's how long it takes to bounce back, with shorter times required for more fit individuals). High alcohol intake, sickness, etc. - all leave a mark for much longer as these are strong negative stressors (you certainly don't need a night measurement to see their impact on your physiology, if you have poisoned your body in the evening, then 10 hours later your heart rate will still be elevated outside of your normal range, for example).
My statement about digestion is there to highlight that at times people are quick to associate a lower HRV with something negative, but there is nothing negative in digesting. If you measure your heart rate and HRV while you have a good time with your friends, it's a similar story. The goal is not apathy (never stress the body so that HRV doesn't reduce). Hence my recommendation to simply measure as we've done for 50 years, with a standardized morning routine, which typically helps us avoiding to overemphasize negatively late stressors the way often people end up doing with night data.
A re-normalization within a few hours (e.g. after sleep) means indeed that we responded well, that's typically how we look at various stressors or training specifically (i.e. the lack of a chronically lower value over time, as we e.g. repeat exposure to a stressor without having fully recovered). What I talk about is a re-normalization by the time we take a daily, morning measurement (maybe I was taking this for granted as this is how normally people use HRV).
I hope this clarifies some of the points I was trying to express in the blog above.
take care
I appreciate your thoughtful response and the insights you’ve shared. However, I respectfully disagree with your assertion that "there is nothing negative in digesting." A 2020 study (https://pmc.ncbi.nlm.nih.gov/articles/PMC7215804/) found that eating within three hours of bedtime led to more frequent awakenings and disrupted, lower-quality sleep. This aligns with my personal experience: eating late at night reliably reduces my total sleep duration and deep sleep, while increasing time awake as tracked by wearables.
You suggest that as long as HRV normalizes by morning, there’s no significant consequence. However, while quick recovery might indicate resilience, it doesn’t negate all potential harms. The body strives for homeostasis, but achieving it doesn’t necessarily mean a stressor was benign.
For example, using the same logic:
I could eat unlimited candy, and as long as my fasting glucose is normal by morning, I’ve "fully recovered."
I could drink heavily each evening, and as long as my blood alcohol content is below the legal limit by morning, there’s "no problem."
In both scenarios, the timing of measurement conveniently obscures potential harm. By focusing solely on recovery metrics, we risk ignoring the cumulative and long-term effects of behaviors that may still impact health.
I hope this offers a constructive perspective, and I appreciate the dialogue.
thanks Steve, I understand what you are saying here and agree that not only the recovery but also the reactive phase can be informative under the right circumstances, and while knowing the context.
Here I'm trying to find a balance between the obsession with continuous tracking (and the many artifacts involved, false positives, psychological implications linked to data that is often just misinterpreted, etc.) and 'looking at the big picture' when using metrics that have obvious limits (like HRV) but are marketed as some sort of 'perfect 24/7 stress index'.
HRV can respond with an increase or a decrease to a certain stressor, and can respond in the same way to positive and negative stressors, hence I frame it in a way where many of these aspects are (in my view) handled better, avoiding an overly reactive (and often misguided approach). While it is true that achieving homeostasis does not signal that previously there was a negative stressor, the contrary is also true, when looking at this data. We can only see changes, not meanings associated with these changes (as I've written elsewhere, otherwise we can stop eating altogether, and our HRV will be as high as ever - but that's hardly a good thing, is it).
This does not mean that there isn't a different use when data is properly contextualized, but in my opinion less data is more helpful when it comes to this specific application (I am not sure if you have used morning measurements in the past, but many of the stressors mentioned are captured easily: https://pmc.ncbi.nlm.nih.gov/articles/PMC8659706/pdf/sensors-21-07932.pdf ). Given the terrible quality of optical measurements (10-20% of data is garbage even during the night, while 70-80% of the data is garbage during the day), it's too easy to get fooled by wearables and associated estimates (as well as by the complexity of our physiology, back to digestion or lack of food intake), hence my concerns and 'push' for intentional measurements using good protocols.
If our habits are not good, they will quickly impact our recovery metrics
Thanks again for your input, I do appreciate the dialogue as well, it gives me food for thought and will certainly be part of my thinking for future writings. All the best
Yes I have used Morpheus training, which relies on morning HRV. I do find that it is more accurate in predicting training capacity.
Eating late at night does reduce sleep quality, on average, for me. Reduced sleep quality does, on average, reduce morning HRV for me, though not as much as what wrist based optical monitors suggest. I appreciate that this is just my own reaction, and others might be less sensitive.
Thanks for your reply.
Hi Marco, as the data is there in Oura’s feed about the measurements of HRV over the night, and you have worked with Oura, can’t you pluck out the morning reading (or ask them to add it to their API)?
hello Shane, morning data is different from what you normally have during the last part of the night, which might also be particularly messy due to the impact of REM sleep (which includes very high variation in autonomic activity). Body position will also differ. You can use the ring for a morning measurement, with this workaround: https://www.hrv4training.com/blog/using-the-oura-ring-for-morning-hrv-measurements - even though it's probably more practical just to measure with HRV4Training directly.