Virtual Everesting (Cycling)

I’ve been doing a lot of work to educate myself on exercise physiology. I not only feel that it’s my responsibility as a coach to learn these things but also I just am so curious about the capacity of human performance. Studying my own training and performance is one way to explore questions I have that aren’t directly addressed in research. Also, research is a population-level approach that doesn’t take into account my individual differences! 

Everesting is a challenge to climb and descend the same mountain as many times as required to hit 29,029 feet of elevation gain, which is the height of Mount Everest. I first heard of this in 2020 during the pandemic, and I completed my first cycling Everest on Hawk Hill in San Francisco in 2021 (report here). 

The idea for this Everest was planted in my head by seeing a vEverest (virtual Everest) on Strava (I think I follow Hells 500). Then I wrote a piece a couple weeks ago about off-season ideas, which included Everesting. But the true catalyst for doing this was I just really wanted to know if my heart rate and/or power would drift downward after 10 hours of work! This wasn’t the perfect experiment: I bonked halfway through and that disrupted my power output; I only have wrist-based heart rate; every ~90 minutes I got a 10-minute break as my avatar freely descended the mountain and I stood up and walked around for injury prevention. But I’ll work with what I have, and overall the conditions were pretty controlled so the data is pretty clean. It also just so happened that I had a free day with no meetings or commitments or coaching scheduled, and that is extremely rare for me, so I made a pretty last-minute decision to go for it. It also helped that I’ve been recovering from a groin strain so my running has been limited and the running training doesn’t feel compromised by this effort.

As mentioned, I wasn’t really sure I’d ever do this or if so, when. It was swirling in my head that this day would be my opportunity, but I didn’t really prepare well. In retrospect, the worst part of the preparation was that I did not carbo load.  

The night before I set aside six shirts and four pairs of pants in anticipation of sweating a lot and switching clothes. I poured a pre-determined amount of Naak and Tailwind into portions for my bottles. I gathered 90-gram Precision gels, SiS gels, Gus, and bananas. I planned to have toast for “lunch” at the halfway point. The overall fueling plan came out to ~75 grams of carbs per hour. I figured this would be plenty as I planned to stay mostly in Zone 2 and I’ve gone for plenty long rides and runs with only this amount, or sometimes less. I charged my watch and checked the U-turn functionality in Zwift (turns out you have to hold the key down not just click it once). I double checked the vEveresting rules on the website. No “smart” trainer required though reddit said otherwise — good because I don’t own a fancy trainer, just a single-resistance wheel-on one. I set my alarm for 3am.

The Alpe du Zwift is a simulation of the famous Alpe d’hHuez in France. There are 21 switchbacks. Zwift (the virtual cycling platform I did this on — by the way, I did it on the free trial, which lets you bike “25km” once a month for free but actually doesn’t stop you after 25km as long as you continue the activity) gives you average watts and split time for each of these 21 switchbacks alongside your PR. To Everest the Alpe, one must climb and descend it 8.5 times.

The setup from the back!
Prepared.
So lucky for cat support always.

I love riding in the dark. The first 6 hours passed quite quickly. I started out very conservatively. It was not difficult to make each climb faster than the last, and I decided this would be my secondary goal (primary was to finish). The traffic lights three stories below me switched red to yellow to green to red to yellow to green in the corner of my eye. I thought about how many times that day they would switch, how many cars and people, all doing different things, on different missions, would pass through that intersection. Around halfway up each climb the virtual Alpe would get snowy and I’d think about the snowy, very exciting Bengals-Steelers game that I’d watched until way too late the night before. No music, just the whirr of the trainer, the power numbers, and these thoughts. 

Things got interesting with about 10 switchbacks to go on the fourth climb (about 6.5 hours in). I started feeling a bit nauseous and shaky, so I decided to stop the fueling for a bit thinking maybe I’d overdone the carbs or salt and put in headphones to distract myself. I did pretty well and made it to the top alright, but then on my 10-minute descent break I started feeling really bad — nauseous, a little lightheaded, shaky, cold, heart rate not recovering when I stopped. Honestly, not unlike how I’d felt in the Grand Canyon. I tried to eat the toast like I’d planned but it just was not happening. I turned down the fan to its lowest setting since my wet shirt was cooling me maybe too much. (It was a very small fan and I’d expected to sweat way  more, so this really surprised me.) I got back on the bike and pedaled the best I could while also being conservative because I was alone at home and didn’t want to pass out without anyone knowing. I got maybe halfway up the climb and started feeling so awful I decided it was in my best interest to stop and take another break. I forced some gel and Kevin came home. He told me I was okay. We decided I’d bonked but was mentally intact and fine just needed more carbs. It’s so interesting that I get nausea as a symptom, but now I know that’s a symptom for me personally to look out for as hypoglycemia. Anyway, he encouraged me back on the bike, and I upped my carb intake to about 100 grams per hour. I told myself this was a low point and I’ll get a second wind eventually. About an hour later I started feeling much better. It was almost like a switch turned back on in my legs and suddenly I could output the power again without my heart rate even changing. Though my fifth summit was the slowest, I negative split the rest of them and my final summit was the fastest one of the day. 

Got the Alpe jersey a few times!
Tired but good post-Everest!

In the end I did 137 miles and 29,292 feet of elevation. It took me 14.5 hours (~13 of which was actual riding time — the rest was descending time plus my 12-minute additional break on climb 5). I took in 70 grams of carbs per hour, excluding the first hour (in retrospect, I did poor calculations at 3am and was likely only taking in about 50 grams per hour in the first 6 hours). I did best with only about 500 mL hydration an hour despite being indoors on a low fan setting. 

I imported the Garmin data into R and ran some statistics and made some graphs. I need to play around with it more, but this is some preliminary info, thanks to a new power meter! 

My average cadence while pedaling was 86. My average power while pedaling was 131 watts (126 including the coasting). My average heart rate was 144 (141 including coasting). I wanted to look closer at my heart rate recovery, but I was moving around too much and variably during the times off the bike that I think the data isn’t reliable. I do know it was higher while I was bonked. Fun fact: my accumulated work throughout the ride was about 6000 kJ, which is enough to power an electric car for about 5 miles or a 10-watt lightbulb for about 170 hours straight! 

I had a fun time checking normality with histograms and Q-Q plots before doing significance tests. I don’t think the data is normal–more like bi-modal–, but that’s ok. Just comparing averages and medians is good enough for me. I (aka ChatGPT and R) split the ride up into 4 equal time sections: early (first 3.5 hours), middle 1 (3.6-7.2 hours), middle 2 (7.3-10.8 hours), and end (10.9 hours to end). Below is a table and the corresponding graph comparing heart rate, power, cadence, pedal torque efficiency, and pedal smoothness over the time sections. 

Table 1: Metric averages over time.
Graph 1: Metric averages with SD over time.

I also made a box plot of heart rate, power, and cadence during each time section. This data excludes any times cadence or power was 0 for sure, but I’m not sure what else it excluded. (When I imported the .fit file into R it imported as 8 tibbles and this is only analysis of the longest tibble that did have time stamps from 0 to 14 hours 30 minutes, so I’m optimistic it includes a lot, but either way it’s the best I got right now.)   

Graph 2a: Box plot of heart rate over time intervals.
Graph 2b: Box plot of cadence over time intervals.
Graph 2c: Box plot of power over time intervals.
Graph 3: Power on cadence over time (red is earlier time points, purple is later).

I think another way of thinking about pedal efficiency is to look at whether the power-cadence graph shifted over time. I’ve put that visualization here — the red dots are earlier in the session and purple are later. What I perceived was that earlier on I could use cadence to achieve a higher power (maybe you can imagine the bundle of red toward the right of the graph). As it got later (and I assume I got more neuromuscularly tired), it was harder to spin for a higher cadence so I would use the next gear (leading to a lower cadence) to get to the higher power. It seems like maaaaybe at the later times the lower cadence was giving higher power and the higher cadence was giving lower power compared to at earlier times, but I think overall it doesn’t seem like much drifted or changed over time. No stats anyway.

You can look up the completed, verified Everestings in the Hall of Fame. This one isn’t up yet but the one I did in 2021 is. In 2021 I remember noticing the huge discrepancy between the number of male and female finishers, and this time around I again realized how large it is. There have only been 600 female virtual Everesting finishers (single Everesting or more) and way over 2000 male ones (I couldn’t figure out the exact number because the page only went to 1000 entries). I couldn’t filter by age either, which would have been interesting to look at.

I really enjoyed this experiment and may do it again at some point. My key takeaways were:

  • Carbs are a must if you are pushing your limits. For me personally this meant not just some carbs but a lot more carbs than I thought at first (90+ grams per hour). Also, I need to carbo load for events over 5 hours. Otherwise, metabolism is a limiter on my performance if I’m pushing upper Zone 2.
  • I can sustain upper Zone 2 heart rate and power for a long time without much (if any) drift.
  • There was some neuromuscular fatigue towards the end as it became harder to maintain a higher cadence, but it was not as much deterioration as I thought it would be.
  • All miserable times come to an end. The second wind can be glorious.

2 thoughts on “Virtual Everesting (Cycling)”

Leave a Comment

Your email address will not be published. Required fields are marked *