Heart rate vs power output and distance.

[gianlino]

Hello. I would like to figure out a formula predicting the heart rate as a function of power output (or speed) and distance. The 500m split of the rower is convenient for that. However a lot of data is needed to get started and if anyone would be willing to participate and send me some data that would be great.

For instance for an 8000 m's row at a constant 2.00 mn per 500 m pace I get, (after substracting 100 from the HR)

35,40,43,45,48,49,51,51,52,55,55,56,58,60,60,60

Any formula would presumably involve constants linked to the individual, and having datas from several people would help a lot.

Thx.

Gianlino

[Ralph Earle]

FWIW, from 27DEC2004 (Age 61):
TIME PACE DISTANCE HR CUM. DIST M/HB
02:00.0 02:15.8 442 132 442 1.67
04:00.0 02:14.2 447 130 889 1.72
06:00.0 02:15.3 443 132 1332 1.68
08:00.0 02:14.7 445 135 1778 1.65
10:00.0 02:13.9 448 137 2226 1.64
12:00.0 02:17.6 436 137 2662 1.59
14:00.0 02:15.6 442 139 3104 1.59
16:00.0 02:14.6 446 142 3550 1.57
18:00.0 02:13.6 449 144 3999 1.56
20:00.0 02:13.3 450 146 4449 1.54
22:00.0 02:13.7 449 147 4898 1.53
24:00.0 02:14.8 445 148 5343 1.50
26:00.0 02:14.9 445 151 5788 1.47
28:00.0 02:11.4 457 155 6245 1.47
30:00.0 02:03.5 486 159 6731 1.53
M/HB [Meters per heartbeat] = 1.71 - 0.0000365*CUMDIST; R^2 = 0.95

From 21SEP2004:
CUM. DIST PACE HR M/HB
1000 02:22.4 138 1.53
2000 02:22.5 141 1.49
3000 02:22.3 144 1.46
4000 02:22.4 147 1.43
5000 02:30.2 139 1.44
6000 02:41.6 130 1.43
7000 02:43.7 128 1.43
8000 02:45.1 127 1.43
M/HB = 1.51 - 0.0000128*CUMDIST; R^2 = 0.87

From 23AUG2004:
CUM. DIST PACE HR M/HB
500 02:26.4 129 1.59
1000 02:25.6 132 1.56
1500 02:25.9 135 1.52
2000 02:27.3 136 1.50
2500 02:29.8 133 1.51
3000 02:33.9 129 1.51
3500 02:32.5 132 1.49
4000 02:30.3 134 1.49
4500 02:39.2 132 1.43
5000 02:42.3 131 1.41
5500 02:35.0 132 1.47
6000 02:36.3 130 1.48
6500 02:34.7 132 1.47
7000 02:40.7 131 1.43
7500 02:41.4 128 1.45
8000 02:32.5 133 1.48
M/HB = 1.55 - 0.0000151*CUMDIST; R^2 = 0.76

5Ks [note the date of the last one]:
Date Pace Avg HR M/HB
09/25/03 02:05.5 157 1.52
10/28/03 02:19.4 143 1.50
11/04/03 02:15.8 140 1.58
03/25/04 02:13.6 136 1.48
09/19/04 02:24.3 142 1.46
12/18/04 02:10.1 147 1.57
03/10/13 02:28.9 126 1.60

From 31JUL2004:
CUM DIST PACE HR
1000 02:12.4 144
2000 02:13.2 149
3000 02:13.8 149
4000 02:16.1 152
5000 02:15.2 153

[gianlino]

Thank you very much Ralph.
I'll go over this tom: our time difference is 12 hours!

Best
Gianlino

[gregsmith01748]

Hi,

I've been logging HR data along with pace for erg sessions for the past 2 years. The data summaries are blogged here:

http://indoorsportservices.co.uk/forum/ ... 58567d799d

Feel freee to look through it. There are steady state low rate sessions, time trials from 500m to Full Marathons and everything in between.

If you want raw data for any session, PM me and I will send you the excel files.

[gianlino]

Hello Greg

I went to look at your blog and I suppose raw data would be easier for me to handle.
Ideally I would need data for the same distance ran at 5 distinct steady paces with HR every 500 ms.
I am sure you have that in store but you will find it more easily than me. Thanks a lot.

Gianlino

[Ralph Earle]

Here are the 500-meter data from yesterday's 8K in 44:25.9 = @2:46.6, 18spm, 10m/stroke.

The split data are the tenths of a second (i. e., "6" = 2:46.6); the heart rate (minus 100) is at the end of the interval. (It was a little elevated from trying to hit each at exactly 2:46.6.)

DIST 2:46.X HR-100
500 6 5
1000 6 10
1500 6 9
2000 6 13
2500 6 14
3000 6 15
3500 5 14
4000 6 16
4500 5 18
5000 7 20
5500 6 21
6000 8 19
6500 5 22
7000 8 23
7500 7 24
8000 5 25

[gianlino]

Thank you Ralph. Your last data are perfect.
A first version of the formula I have in mind is of the form HR = A +BP +CPd where P is the power, d is the distance, and A,B,C are constants depending on the individual. I believe A should be the actual resting rate, measured just before rowing.
The bias you mention about the HR increasing to get exactly 2 46 6 is not that bad since it should be uniform all along.

In your case one has HR = 108 + 1.1*d. Since the Power at 2.46.6 is P = 75.6, one gets C = 1.1/75.6 = 0.015. The value I got for myself is 0.016...

By the way, do you have any idea what your actual resting HR is? If you do then we can get A and B and then you can check if your formula works at various speeds...

All the best

Gianlino

[Ralph Earle]

My resting HR is 48. (Max is 154.)

Before the 8K I had warmed up 500m @3:07 and had two false starts of 226m and 77m when the HR monitor stopped registering. (Pin wasn't in all the way.)

When I started the 8K my HR was 100.

Do you want more 8Ks? I could do them @2:42.2 (18.5 spm; 10m/s), @2:37.9 (19 spm), @2:33.8 (19.5) and [maybe!] @2:30.0 (20).

[gianlino]

I believe the resting heart rate when you actually begin is not the proper HR to use. What I do is take it over 3 mns before starting the warm up.
It i usually between 55 and 65 depending on what I did before. Using the values you gave for your min and max, it should be between 56 and 64.

Taking 60 as an intermediate value, a tentative formula would be

HR = 60 + 0.63*P + 0.015*P*d

If you get close to higher HR's like 140+, I guess this formula would collapse.

[Ralph Earle]

I've been retired for two years, but when I was working I used to check my HR around 3 PM most days. It was consistently 51bpm. So perhaps use that? (Just checked -- it was 52bpm this past last minute.)

And yes, the formula will have to break down when one goes anaerobic. But below that pace, it would be most interesting if there were a "universal constant" of approximately 0.015 for cardiac drift among experienced rowers, regardless of age. (I'm 71 -- how old are you?)

[Ralph Earle]

Here are the 500-meter data from today's 8K in 43:15.4 = @2:42.2, 18.5spm, 10m/stroke.

Before a 500m warm-up @ 3:07.7, HR = 62; at the finish, HR = 83. At the start of the 8K, HR = 65.

The split data are the tenths of a second (i. e., "2" = 2:42.2); the heart rate (minus 100) is at the end of the interval.

DIST 2:42.X HR-100
500 2 8
1000 1 9
1500 2 13
2000 4 14
2500 2 16
3000 2 18
3500 2 18
4000 2 19
4500 2 22
5000 2 23
5500 2 25
6000 2 26
6500 2 28
7000 3 28
7500 2 29
8000 2 29

HR = A +BP +CPd

Is the last term correct -- (Constant*)Watts*Distance? If so, interesting notion!

PS: I just noticed you have PM enabled, so better to take this off-line and not bore the spectators?

[gianlino]

I don't know what "PM enabled" means nor how we could go offline...
I omitted a factor 2 yesterday and your drift constant was 0.031.
For today's work it is 0.035. So either it oscillates or the dependence on P is not linear.
Anyway, we can know that with a few more similar results, hopefully with a wider range of speeds.
Best. G.

[Ralph Earle]

By "enabled" I meant that your profile includes the option to send you private messages. (I'm assuming no one else is interested in my 8K split data.) But I just tried to send you one, and was told you don't exist! My e-mail address is [firstlast] @ hawaii.rr.com -- substitute my name and remove the spaces.

[Carl Watts]

I don't want to put you off but the more you think about it the more variables there are to try and factor in. I would have an educated guess and say its not possible to come up with any formula that has a degree of accuracy to be useful.

The more I have tracked my results over the years the more convinced of this I become. Even using the "Performance Analysis" in RowPro which is immediatley flawed as it uses pace vs HR as a linear function when it should be looking at Watts vs HR, how do you factor in the improvement in fitness ? it becomes just a relative number that said my performance had improved 30% when I compared results over 2 years appart.

Distance and HR is interesting, or should I say time spent in the Aerobic threshold region. Even though your HR may hit the same maximum for a row the longer you keep it at the maximum the slower the recovery time.

Certainly you can arrive at calculations for performance improvement figures using the right variables and the average HR but even then temperature and SPM also has a big effect even for a distance or time rowed at an identical pace. SPM figures from the current monitor are no good to use, they are way out percentage wise and the error becomes worse and worse the lower the rating, you would have to count the total number of strokes for the row to be able to even begin to start the math with any accuracy.

Just my thoughts.

[gianlino]

Hello Carl

Thank you for sharing your thoughts.

Given the current temperature in Paris where I am, I was indeed reminded of the role of temperature.
I also noticed the role of SPM and obviously fitness counts, not to mention the shape of the day. That's obviously too much for an ideal formula anyone could use at any time. I hope this is what you meant. Now these are not lethal obstacles for my purpose. Indeed I am looking for a formula which predicts your heart rate evolution if you row at constant speed in optimal conditions for you. Namely reasonnable temperature, optimal SPM and drag factor, great shape that day etc. Of course the constants of such a formula would depend on the individual, but if the form is simple, then there would be very few constants, 3 or 4 max, which might evolve over time but not that fast. So given my limited ambition, I am not pessimistic.

As for the SPM counter, I don't agree with you. I have been practising at 14 SPM following a metronome set at 56 spm, and noticed the counter settled at 14 after 3 or 4 strokes and I think at 20 SPM it is reliable too. Maybe older models (C) are better!

Anyway, thanks for reminding me to be reasonnable. I hope your HR's will follow soon, even though you seem to think it would be a waste.

All the best

Gianlino