Halo Blog

Announcements, musings and more, from Halo HQ.


Altium i10 – Our results are in!

Posted June 23, 2016 by in Uncategorized

As regular readers will know, we had Mike Duff, one of our coaches, test the Altium i10 personal hypoxic device over a month, in the run up to the Girona Cycling Festival races, in the first week of June.  This was done over the course of two blogposts, which can be found here:

1 – Introduction of the planned tests

2 – Device progress report

Here is his report on the quantitative tests undertaken, where he looks at the results from the four tests he set himself, over the month of testing of the Altium i10.


The Tests 

Now that the Hill Climb TT, Nocturn Cobbled Crit Race and Gran Fondo Road Race are complete, this year’s edition of the Girona Gran Fondo is complete.  As you’ll have gathered from our social media channels and blogs, I LOVE this event, and it holds a special place in my heart.  You can go back a year and read my posts on it from last year if you wish, starting here.

I decided to use it in my testing of the Altium i10, in addition to an FTP test and a [Garmin Edge 520-estimated] VO2 max, and a segment time test.  Comparing results to last year would lend a bit of qualitative, real-life comparison to the three aforementioned numerical tests.

FTP (Functional Threshold Power)
Prior to the month run-up to the Girona Cycling Festival, in which I did a 30 day Altium i10 set of 15 sessions every second day, I had an FTP of 248 watts.  By the end of the month of training incorporating the Altium i10 sessions, I had an FTP of 282 watts.  So in that final month of training, I had an increase in FTP 13.7%. Looking back at last year’s training logs, this is a faster monthly growth rate than any I achieved last year.  I of course am no scientist, so will not venture to argue this is down solely to the Altium, but am simply presenting the numbers here, for what they are.

Technological sidenote: Interestingly this FTP is exactly where I was last year at this time, but one main different is that I was using a Powertap Heart Rate band to estimate power, which is known to give on average higher-than-reality readings.  This year, I am using Powertap P1 pedals, which are significantly more accurate, so I like to tell myself that this year’s 289, is better than last year’s 289! (This is the kind of thing one says to oneself whilst suffering on a climb…)

VO2 Max
As I outlined in my opening blog post, in the absence of a full metabolic lab setup, I used my Garmin Edge 520’s VO2 max estimation algorithm, derived from Heart Rate (bpm) and Power Meter (w) readings, to look at my VO2 max before, during, and after the month-long set of Altium i10 sessions.

Before undertaking the month of training, which included the Altium i10 set of 15 sessions, my VO2 max reading was 58 (ostensibly “Superior” according to the Cooper Institute, though nowhere near the world record of 97.5 set by a Norwegian cyclist in 2012).  By the end of the month, it had increased to 60 – an increase of 3.4%.

Testing undertaken at the University of West Scotland during the development of the device showed this level of improvement in riders (between 3-4%), and was mirrored by a [lab] test either side of the month long test period by the people at Triathlon 220 Magazine, which you can read here.

So whilst I can’t attest to having performed a rigorously scientific test here, at least my numbers, estimates as they are, are showing both what the Altium research, and independent tests have shown.  An improvement, over a month, of between 3 and 4% of my VO2 max.

Segment Test
As I wrote a few weeks ago, I elected to use the climb to the Santa Pellaia pass, from the northern approach, over the Gavarres hills behind Girona, as a segment test, adding in some “real road” testing.  It’s a 6km climb, which averages 4%, is pretty consistent in gradient, and is almost always completely traffic-free, so good for “putting the hammer down”.  Here is the leaderboard (spoiler: I did not manage to displace Robert Gesink from the KOM).

When I rode the climb just before starting the Altium i10 month, I managed a time of 15:46, at that point my fastest ever, putting out 271 watts at 22.3 km/h, and an average heart rate of 168.

Two weeks later, mid-way through the Altium i10 month, I managed a time of 15:14, again my fastest ever, putting out 294 watts, at 23 km/h, at an average heart rate of 172 bpm.

So I shaved 32 seconds off my segment time, in 2 weeks, which is certainly not a small margin, but more importantly I think, I also showed an increase in my output/input ratio, i.e. how many watts I am able to generate for the work my heart is undertaking.  Looking at the first ride my ratio was 1.613 (w/bpm), whereas on my faster ride, I was getting just over 5% more watts, for each beat of my heart (an O-I ratio of 1.709).  If you want to read a bit more about O-I ratios, have a look here on Coach Joe Friel’s blog.

Els Angels Hill Climb TT

Halo's male racers, Mike (first in start line) and Seb (second), at the race start.

Halo’s two male 30+ racers: Mike (first in start line) and Seb (second), at the race start.

Race day!  The infamous Els Angels (often ominously mispronounced “Hells Angels”) Hill Climb Time Trial is always held on the second day of the Girona Cycling Festival.  It is gruelling, and hence nice to get out of the way at the beginning of the week’s races.




Mike on the podium at the Girona Gran Fondo hill climb, 2015

Mike on the podium at the Girona Gran Fondo hill climb, 2015

Last year, I really impressed myself with a top 10 performance, coming 2nd in my age group (to the overall winner), and 6th overall.  I did the 10.1km climb of the ~400m pass in 24:35 seconds, eclipsing my prior fastest time in training of 28:05, and smashing my goal of 27:00.  I was fitter than I thought.  So it seemed like a good test, to try to better that time, a year on, with a similar year of training under my belt, and the same supplement and nutrition regime.

Louis Meintjes's (now mine) MTN Qhubeka '15 TDF edition Cervélo S5

Louis Meintjes’s (now mine) MTN Qhubeka ’15 TDF edition Cervélo S5 (yes, I know I keep putting pictures of the #starship in my blogs)

Disclosure! I did ride a different bike this year, choosing to ride an aero Cervelo S5, with 60mm deep section wheels.  Having modelled the route and myself using Best Bike Split (such a good modelling tool its been acquired by Training Peaks), it emerged that [allegedly] because of the two short sections of descending on the climb up to the monastery at Els Angels, the numbers stacked up in favour of my aero bike with aero wheels, despite my Rocky Mountain Solo 30CR’s being slightly lighter, with stiffer Mavic R-Sys wheels.  So off I went, on a slightly counter-intuitive bike!

Screen Shot 2016-06-23 at 13.09.36

Top 15 for 2016 GGF Hill Climb – includes all 3 of our Haloistas!

The result? This year, I shaved 28 seconds off of last year’s time, coming in at 24:07.  There were more 30-39 year olds this year (including a very recently retired pro, this year’s winner), so my time only got me 4th in my age group this year, and I once again sat 6th overall.

But the test is not about placings!  I was just under 30 seconds faster, and a year older.  That is enough for me (insert broad smiley emoticon here).  I won’t compare my output/input ratio in this case, since my power readings were only estimated in the 2015 ride.  My heart rate readings however are comparable, and I can see that on a day when it was a degree hotter (27C) my maximum heart rate was actually 8 bpm lower, and my average heart rate was the same, despite going almost 30 seconds faster.

Below is the 2015 Strava file. And the blogpost I wrote the afternoon of the event last year is here:

Here is the 2016 Strava file:

So there we are, a positive result in each of the four tests I set myself (and the device).  I can’t attest that any of these were double-blind, controlled, lab-grade tests, but they are what they are.  Caveats aside, they are real-world observations, by a keen cyclist and coach – nothing more, nothing less.  I would be very comfortable to write here, on the basis of the peer-reviewed research undertaken by the Altium i10 team at the University of Western Scotland, the tests undertaken by bigger players in the endurance sport media, and the results I observed using my everyman tests, that I believe this device played a part in my performance improvements this year.  For one of the tests, I was a whole year older, for the others, the improvement shown was very quick – two weeks in the case of the Segment test, and four weeks in the case of the FTP and VO2 max tests.

End note
I didn’t include the Girona Gran Fondo Nocturn Crit Race or Road Race in the testing protocol, since both events involve multiple riders on course at the same time (making control of the test difficult), and the fact that the Road Race route changed slightly this year, but I was also VERY happy with my results this year.  Despite a stronger, more numerous field of riders this year, I managed to come 5th (tied for 4th) this year, compared to 8th last year, and at the end of a 3.5 hour race that was undertaken at a much higher work effort than last year’s.  I averaged a heart rate of 157 bpm this year, compared to only 83 last year.  I also managed to remain in the lead group til the end, and was able to bury myself twice to chase back to them, which last year was impossible.  This is all anecdotal – but feels like improvement.  And that I can carry that feeling into my continued training this year, is very important for next year!