Given to Tri: Training

I needed a way to track my Endless Pool workouts, so I built one

2025-05-27T19:00:00-06:00

I mentioned this briefly in my race report about this year’s Ironman 70.3 St. George, but for the past few weeks I’ve been swimming in an Endless Pool, since the Teton County Rec Center’s lap pool remains, alas, closed for the foreseeable future. It’s been an effective training tool, but one of the annoying things about it is that it’s hard to track my workouts—there are no “laps,” so all I can do is time-based sets, and even then, I have trouble keeping track in my head of how many sets I’ve done, at what pace, and for how long. So, I built myself a little tool to do this: I call it SwimSetter, and it’s a simple web app to keep track of my sets in the Endless Pool.

Using it is easy: Just bring your phone to the pool in a waterproof case, leave it on the side, open the app, set a timer or a stopwatch on your watch, and then just swim. Whenever you stop for a rest, enter how long your last set lasted and at what pace it was, and repeat until you’re done with your workout. The app will add up your total time, distance, and average pace, which you can then enter into a workout log or a manual Strava activity or wherever you want.

It’s a very simple tool that answers a very simple question for me: “Have I swam enough today?” But it might be useful to you too, so here it is.

TrainerRoad Launches Zwift Integration: Here’s How It Works

2025-02-12T08:00:00-07:00

I’ve been using TrainerRoad for all my cycling training, with great results, since even before I got into triathlons. I love the training plans and all the machine learning magic they apply to them, but doing long workouts indoors while looking at nothing more than a graph and some numbers is mentally challenging, to say the least. Or as a certain cyclist would say: “It builds character.”

There are days when completing a two-hour (or more!) cycling workout indoors requires more mental endurance than I have available, especially in the winter, when I can’t ride outdoors to break the monotony. On those days, even a little bit of distraction helps. Zwift is a good one; I enjoy the visuals and collecting route badges and achievements while exploring Watopia definitely helps pass the time. Over the years I’ve tried the two common ways of using TrainerRoad with Zwift: Either by running them side by side, or by recreating the workouts manually and doing them in Zwift. Both methods work reasonably well, but they’re a bit of a hassle to set up and each comes with their own set of downsides.

About a year ago, though, Zwift announced a new Training Connections API, which enables third-party training platforms to send cycling and running workouts over to Zwift, so that users can do them there. Since that announcement I’ve been waiting with bated breath for TrainerRoad to integrate with it, and it finally launched earlier this week. I was invited to try it out while it was in closed beta, so I’ve had the chance to do about a dozen TrainerRoad workouts in Zwift over the past few weeks; here are a few first impressions, which are overall very positive.

How it works

Just to state the obvious: You’ll need to be a subscriber in both platforms to make this work (although it might work with a lapsed Zwift account if you stay under the monthly 25 km limit). If you are, setting this up is quite simple, all you need to do is go to the Connections page in the TrainerRoad website, hit the “Connect” button under Zwift, log into Zwift, and authorize the connection.

Once this is done, TrainerRoad will push any workouts that you have scheduled for today and tomorrow into Zwift, and today’s workout will show up in the “For You” carousel in Zwift’s Home screen. If you don’t have an active training plan, TrainerRoad will push a recommended workout instead; this is based on TrainerRoad’s TrainNow feature, which uses machine learning to suggest a workout based on your recent training history. The syncing happens automatically: If you put a workout on your calendar, it’ll show up in Zwift almost instantly (or at least quickly enough that it’ll be in Zwift by the time you switch apps). Similarly, if you select an alternate workout, or accept plan adaptations suggested by TrainerRoad’s Adaptive Training system, those changes will be immediately reflected in Zwift. That does mean that if you need to manage your training plan, such as selecting alternates or moving workouts around, you’ll still need to do that on TrainerRoad before opening Zwift, but it all works seamlessly (and TrainerRoad is working to add an option to automatically accept adaptations and FTP adjustments).

With your connection set up, sending TrainerRoad workouts in Zwift is pretty easy. If you have a a workout scheduled for today in your calendar…

Today’s workout: “Bashful -1,” a sixty-minute VO₂ max workout.

…then simply open Zwift…

…select it from either the Home or the Workouts screen, pick a route to do it in, and off you go.

You’ll also get a TrainerRoad kit, which looks 🔥 with the Giro Vanquish helmet and the Fire Socks.

Once you finish the workout, you’ll be presented with a post-workout survey to record your RPE on a scale of 1–10. This gets sent back to TrainerRoad and converted into their 1–5 scale, although you can always update the rating in TrainerRoad afterwards. Then, you can end and save your activity, or keep riding if you want. Once you save the activity, it’ll get synced back to TrainerRoad and any other services you have connected, such as Strava or Garmin Connect.

This ended up mapping to “moderate” in TrainerRoad’s scale.

What happens if you can’t complete the workout for some reason? For example, if it was too hard, or you ran out of time? Then you can go into TrainerRoad and answer the post-workout survey with the reason why. This “struggle survey” may pop up automatically, or if you select “I did not pass” from the regular post-workout survey.

The reason you select may determine how TrainerRoad adapts upcoming workouts. For example, “time” may not trigger adaptations, but “fatigue” or “intensity” might.

What works well

I’ve been testing this integration quite a bit over the past few weeks and it all works pretty great, but there are a handful of things that have sold me on it, although not all of them are specific to TrainerRoad. In no particular order:

The Companion app and controllers are game changers

In TrainerRoad, you can use the on-screen controls to do things like pause the workout, adjust the intensity up and down, or toggle erg mode, but for me, that usually involves reaching over my handlebars or aerobars and tapping a small target on my iPad’s screen with sweaty fingers while I’m fighting for my life. In the past I’ve experimented with things like using a Stream Deck as a TrainerRoad controller, but using the Zwift Companion app on my phone (which I have attached to my Garmin mount using this MagSafe adapter) or the Zwift Click makes it a lot easier to control the workout.

I mostly use them for shifting during free rides, but it’s pretty great to be able to adjust workout intensity right from my aerobars.

I like the arches at the end of every interval

I swear looking at the timer tick down in TrainerRoad during a hard interval only makes it slow down, so I like the way the end of intervals are presented in Zwift as arches that you ride through—somehow I can convince my legs that I can make the interval end sooner if I just pedal harder. I can’t explain why it works, but it does.

Must go faster… must go faster…

TrainerRoad workouts work great as pre-race warmups

I don’t like spinning mindlessly in the pen before a race in Zwift. For one, it’s not enough of a warmup for me, but also it’s not recorded, so I can’t track its training load or anything. I’ve found that short TrainerRoad workouts like “Clyde” (a twenty-minute threshold workout) work great as a warmup, but in the past I’ve had to time them properly to make sure I have enough time to finish it, save it, switch over to Zwift, pair my sensors, and join the event. This integration makes warmups seamless: Now I can simply schedule my warmup in my TrainerRoad calendar, do it in Zwift, and hit the “Join Event” button to teleport to the starting pen whenever I’m ready. The workout gets saved and sent to TrainerRoad automatically, separately from the event I’m doing.

I really didn’t have to do this right after that VO₂ max workout, but I needed the screenshot. Let’s just say that I was well and thoroughly warmed up for the Tour de Zwift Stage 6.

Precise power targets

Until now, workout targets in Zwift were rounded to the nearest 5 W, so if for example an interval called for 248 W, it’d show up as 250 W on the screen. This may not sound like a big deal, but for people with low FTPs, that rounding may be impactful. In any case, if a workout calls for 248 W, dammit, I want to do 248 W. This is a change that has been requested in the Zwift Forums for a few years now, so I’m impressed TrainerRoad convinced Zwift to do it.

Pausing a workout doesn’t stop the recording

I don’t like to pause in the middle of a workout, but sometimes I can’t avoid it (for example, if I get paged at work). If the pause is particularly long, I like to spin for a few minutes before restarting the workout so I can get my heart rate back up, but on TrainerRoad, none of this is recorded while the workout is paused. In Zwift, pausing the workout simply puts you in free ride mode and continues the recording, which helps me track my training load and fatigue more accurately. Once I’m ready, I can restart the workout right from where I left off.

The workout data is out of the way

This is a matter of personal preference, but I do like that all the information about the workout, like upcoming intervals and remaining time, is at the edges of the screen, so I can keep them out of sight if I want to, simply by focusing on my avatar and the scenery. The list of intervals on the left side of the screen has been somewhat divisive in the TrainerRoad Forum because it’s so hard to parse, but I actually like that it’s hard to parse. Sometimes I don’t want to know how much of the workout I have left and it’s much harder to ignore the timers and blue bars on TrainerRoad since they’re front and center.

You can turn off workout instructions

I really don’t like the way Zwift displays workout instructions smack dab in the center of the screen. It’s very obnoxious and distracting, and it’s one of the reasons why I don’t like to do Zwift’s built-in workouts, since there’s no way to turn them off in the settings. However, if you turn off workout instructions in TrainerRoad’s settings, they won’t sync over, so they don’t appear in Zwift. Wonderful.

All Zwift activities get synced back to TrainerRoad

While TrainerRoad will only push scheduled cycling workouts to Zwift, once the connection is enabled every activity completed in Zwift, including runs, will be synced back to TrainerRoad. This is a good thing for a couple of reasons: First, it means that all your Zwift activities, including workouts (whether or not they come from TrainerRoad), races, group rides, and free rides, will be analyzed by TrainerRoad and taken into consideration for AI FTP detection and Red Light Green Light fatigue detection; and second, because this removes Strava as a dependency to get all these activities into TrainerRoad, so there’s one less reason to keep it around after their recent policy changes.

What could be better

All that said, there are a few things that I would love to see changed, most of them on Zwift’s side. Again, in no particular order:

Scheduled running workouts don’t sync

TrainerRoad has full triathlon plans that include runs and swims, so I’d love to see the runs sync over to Zwift so I can do them on my treadmill. I don’t expect this to happen anytime soon—those runs are RPE-based, so there’s no target pace to sync over, and I imagine it’d be a big project for TrainerRoad to start tracking or calculating running pace zones (AI FTP but for running would be incredible, though). At least completed Zwift runs do sync back to TrainerRoad, so there’s that.

For now I’ll just continue to recreate these workouts manually in Intervals.icu to sync them to Zwift. I’ve also been told that Zwift integration is coming soon to Runna, which would be even better, even though it’d mean paying for yet another service.

The Kickr Climb doesn’t work during workouts

I bought a Kickr Climb last year and have enjoyed using it during Zwift races and events; I love big climbs, so I like the immersion it adds to my rides. I also like the way it works in TrainerRoad—obviously there’s no terrain to simulate there, so it simply sets the grade based on the percentage of FTP of each interval. I’ve found that the small changes in position from the bike tilt can make a big difference in comfort and engagement during long indoor workouts, so I was disappointed to find out the Climb doesn’t work at all in Zwift during workouts, despite it having been requested for years in the Zwift Forums. I would love if the Kickr Climb continued to follow the terrain in Zwift even in erg mode. In other words, if I’m doing an interval at 200 W and hit a 15% hill, then I want to continue to do 200 W but with the bike tilted 15%.

There’s no way to rewind a workout or repeat an interval

While Zwift makes it easy to skip the current interval if you need to bail out, there’s no way to rewind the workout to repeat an interval. This is something I do in TrainerRoad sometimes—if I have to pause my workout for whatever reason, I try to do so during a recovery interval, but if I have to do it in the middle of a work interval, I like to go back and repeat it. And yep, this is another feature request that goes back years and years, with multiple threads about it on the Zwift Forums.

FTP changes don’t sync between platforms

TrainerRoad workouts done in Zwift are based on the FTP set in TrainerRoad, but TrainerRoad can’t actually update the FTP saved in my Zwift profile, or vice versa. This is mostly a cosmetic issue; having a different FTP in Zwift only affects the power zones in the various graphs, but I still wish I could keep my FTP synced in both platforms whenever one of them detects a change, if nothing else so I don’t have to remember to do it manually.

Erg mode in Zwift is slower to respond

Say what you will about TrainerRoad, but they have erg mode down to a science. Zwift responds slower to changes in power, and it’s very noticeable, particularly in VO₂ max workouts where the difference in power between the work and recovery intervals is usually large. Whereas the power ramps up very quickly in TrainerRoad, I’ve seen it take almost ten seconds in Zwift, and you’ll likely see the “More power!” prompt quite often. It hasn’t really detracted from any of my workouts, but it can be impactful in workouts with short, intense intervals. (On the other hand, a lot of people would argue that you shouldn’t do VO₂ max intervals in erg mode, and at least the Companion app makes it easy to turn it off when necessary.)

Despite these few nitpicks, I think doing TrainerRoad workouts in Zwift is a game changer. It removes most of the friction needed to set it up and I get the best of both worlds: TrainerRoad’s excellent, ML-driven adaptive training plans, with the visuals and gamification of Zwift. It could simply be the novelty of it, but it has made a noticeable difference in my ability to power through long workouts indoors. The VO₂ max workout I used for the screenshots in this post is a good example: It wasn’t particularly tough, just a 3.7 in TrainerRoad’s difficulty scale, but even then I think I would have struggled mentally to complete it in TrainerRoad. My winter blues are pretty rough this year, so I’ll take all the help I can get.

Will I switch completely away from TrainerRoad? Probably not—there’s still a time and place for it, since I’m sure there’ll be times where I’ll want to do something other than Zwift during a workout. But for days when I just need a little visual stimulus, this integration works as well as I could have hoped, and then some—TrainerRoad went above and beyond, even successfully cajoling Zwift into addressing some long-standing feature requests. Not bad at all.

Understanding My Sweat & Sodium Loss Rates

2023-08-06T08:00:00-06:00

At some point during our drive to the medical tent after I dropped out of Ironman Coeur d’Alene, Mike, the race official who took me there, brushed off some of the salty residue on my trisuit and commented, “Looks like you’ve lost some salt there.” I was in such bad shape I don’t remember anything else about our conversation, but that one comment has stuck with me as I’ve tried to understand what went wrong and how to prevent it in the future.

I’m not a heavy sweater. I’m not one to come back from a run drenched in sweat or leave a puddle under my trainer after a workout, and since I’ve never had issues with muscle cramps or any other signs of sodium depletion before, I assumed that meant my sodium loss wasn’t high. In hindsight, Mike’s comment and the salt stains on my trisuit in some of my past race photos made me realize I was wrong to assume that—it’s possible I’m actually a salty sweater, and that sodium depletion contributed to my problems in Coeur d’Alene. To address that and make sure I can come up with a good hydration strategy, I’ve spent the past few weeks exploring ways to calculate my sweat and sodium loss rate.

After some research, here’s what I’ve learned.

Precision Fuel & Hydration’s sweat test

Precision Fuel & Hydration offers a sweat test, which takes only 45 minutes and doesn’t require exertion, and works like this:

A quick, non-invasive sweat test that doesn’t require a long workout is compelling. Unfortunately, it has to be done in person, and the nearest test locations are in Boulder, Colorado, and St. George, Utah, both about an eight-hour drive away, so that’s out of the question for now. (Although I am racing Ironman 70.3 St. George next year, so I might schedule a test while I’m there.)

That said, Precision Fuel & Hydration has a huge nutrition and hydration knowledge base, where I’ve learned a few things that I wasn’t aware of, like that you can, in fact, sweat while swimming. What! I hadn’t considered the possibility that I may have started my bike leg at Coeur d’Alene already depleted after swimming for an hour and a half.

Another great resource they have is this sweat loss rate calculation, which involves weighing yourself naked before and after a workout, measuring the fluids you consume during the activity, and entering those values in an Excel spreadsheet. (I don’t have Excel, so I recreated it in Google Sheets; feel free to make a copy for your own use.)

I’ve been measuring my sweat loss a couple of times a week, in various conditions, during multiple types of workouts, and collecting the results in that spreadsheet; so far, my sweat rate is averaging about 1.17 L/h. To put that into context, a study of the sweat rate of Ironman-distance triathletes showed a mean sweat rate of 1.4 L/h amongst the participants.

However, without knowing my sodium loss, this only tells me half the story, and since I won’t be able to do this sweat test for a while, I looked at a couple of alternatives.

Gatorade Gx Sweat Patch

Gatorade sells these Gx Sweat Patches, which they claim can be used to calculate sweat and sodium loss with the help of the Gatorade Gx app. It works by sticking one of the patches on your forearm and going for a workout. As you sweat into the patch, the dye-filled channels fill up, allowing the app to scan the pattern with the phone’s camera and estimate the sweat and sodium loss.

Image: Gatorade

I tried these patches last year and didn’t get any valid results: the first patch didn’t capture enough sweat for a valid reading, and the app refused to scan the second patch. They’re relatively cheap, at $24.99 for a two-pack, so for the sake of being thorough, I decided to give them another shot, hoping that the experience had gotten better since the last time I used them.

After creating an account in the app, I created a new “sweat profile” for a 60-minute indoor ride. With that set up, I placed one of the patches on my forearm and started a TrainerRoad workout called Black, “an hour of aerobic endurance riding ranging from 50–80% FTP with no recovery,” which I did indoors with my fans off to ensure I worked up enough sweat to get a reading. Roughly 30 minutes into the workout, sweat started filling up the orange line and it kept growing steadily until I finished. The blue line that measures the sodium concentration never filled up at all, though.

The Gx Sweat Patch at the end of my workout.

Once I finished my workout, I fired up the app again to scan the patch. It took me a couple tries to get it to scan, and then the app asked me to confirm a few additional details about the workout (duration, intensity, temperature, fluids consumed) before showing me the results of my new sweat profile: a sweat rate of 28–31 oz/h (0.83–0.92 L/h) and a sodium concentration of 9–23 mg/oz (304–778 mg/L). The app describes this sweat rate and sodium concentration as “moderate” and “low,” respectively.

I had no way to confirm if the sodium concentration reading was accurate, but I weighed myself and my bottles before and after the workout to compare the sweat rate. Plugging the results in my spreadsheet returned a sweat rate of 0.89 L/h, within the range the Gx app estimated, so it seemed right on the money.

The next day, I created a new sweat profile for an indoor run and used my second patch while doing a 60-minute run on my treadmill at a fairly fast pace (for me) of 5:02/km, with the fans off. Like in the previous test, it took about 30 minutes to start seeing the orange line, but once again the blue one didn’t appear at all. After scanning the patch with the app, it returned a sweat rate of 28–31 oz/h (0.83–0.92 L/h) and a sodium concentration of… 0 mg/oz. Zero. Nothing. This is the one thing I want to know, and this app couldn’t give me a number.

According to the Gx app, my sweat is pure distilled water.

Was the sweat rate accurate, at least? Well… no. My own sweat rate calculation was 1.44 L/h, well outside the range estimated by the Gx app, and closer to my sweat rate on previous runs.

These faulty results from the second patch make me question the results from the first one, even though the sweat rate from that one matched my own measurement—for all I know that was just a coincidence. Considering this was the second set of patches I tried and I still couldn’t get reliable results, my conclusion is: don’t waste your time and money on these things.

LEVELEN Athlete Sweat Test Kit

Last but not least, I tried the LEVELEN Athlete Sweat Test Kit. It’s much pricier than the Gx sweat patches, at $189 for the multisport kit (which contains two tests, one for the bike and one for the run), and requires mailing the sweat samples to a lab for testing. They also sell cheaper single-test kits and starter kits that include a personalized electrolyte mix based on your test results. I ordered the multisport kit, which arrived in the mail a few days later and contained the following:

Two sweat patches, which consist of a small piece of gauze on a transparent adhesive film.
Two plastic test vials to put the gauze in after performing the test workout.
Plastic tweezers to put the gauze in the tubes without contaminating them.
Alcohol wipes to clean your arm before applying the patches.
Some #SweatTest stickers to place on top of the patches to promote it on social media (which is an actual step in the instructions).
Two information cards to write down the test results (weight before and after, fluids consumed, etc.)
A postage-paid padded envelope to mail the tests back.

The contents of the LEVELEN Athlete Sweat Test Kit.

The test works very similar to the Gx sweat patch: you wipe your arm with the alcohol swab, wait for it to dry, then stick one of the sweat patches on your forearm and go do a workout. Once you’re done, carefully remove the patch, use the tweezers to pull the gauze off, put it in the test vial, and seal it. Once you finish both tests, place the vials in the padded envelope with the information cards and drop them in a mailbox.

Unlike the Gatorade Gx app, which tries to estimate your sweat rate for you, the LEVELEN test requires you to calculate it manually, using the same method used by Precision Fuel & Hydration: you weigh yourself naked before and after the workout, measure any fluids consumed, and write that data down on the cards provided, along with things like workout duration, temperature, and humidity. (Tracking this information was pretty straightforward since I already had my sweat loss spreadsheet, so I recommend setting that up ahead of time.)

They also have a few guidelines for the test workouts themselves:

For the most representative results, you should perform the sweat test while training in the activity, intensity, and environmental conditions representative of the competition you are training for (e.g. run on a hot day for marathon training). Triathletes should test separately on the bike and run.

Tests should be done during the first sweat/workout of the day, and multiple test should be completed on separate days. Workout duration should be at least 45 minutes.

Unlike the Gx Sweat Patch, the LEVELEN patch goes on top of the forearm.

For the run, I did a 60-minute run on my treadmill at a 5:09/km pace, roughly the same pace I did at Ironman 70.3 Arizona last year. Although they recommend trying to reproduce race conditions, I ran indoors to avoid contaminating the sample with sunscreen. It was reasonably warm inside, though, at 27°C, and I did the workout with my fans off, so I got a pretty good sweat going.

The following day, I repeated the process with Rendezvous, a fun TrainerRoad bike workout consisting of “50 minutes of continuous progression from 60% FTP all the way up to 100% FTP via 10% increases every 10 minutes.” With an intensity factor of 80%, it was more intense than the workout I did with the Gx patch and, like the run, closer to Ironman 70.3 intensity. As in the previous tests, I did it indoors with the fans off.

For both workouts, I weighed myself and my water bottles before and after and entered the numbers in my spreadsheet before writing them down on the information cards.

After dropping the envelope in my nearest mailbox, I emailed LEVELEN to let them know my tests were on the way, and they immediately replied that I would get the results by email within one or two weeks of receipt.

I received a detailed six-page PDF from LEVELEN in my email exactly two weeks later, as promised. The document outlines my sweat rate as follows:

	Bike	Run
% dehydrated	1.4%	1.3%
Gross sweat loss	1.5 L	1.3 L
Fluids replaced	430 mL 29% replaced	330 mL 26% replaced
Sweat rate	1.48 L/h 62nd percentile (moderate)	1.28 L/h 49th percentile (moderate)

Unsurprisingly, these numbers match the ones on my spreadsheet, but what I really wanted to know was my sweat composition. LEVELEN provided the following breakdown:

	Bike	Run
Sodium concentration	33 mmol/L 759 mg/L	30 mmol/L 690 mg/L
Sodium loss	1,129 mg/h	882 mg/h
Chloride concentration	40 mmol/L 1,420 mg/L	38 mmol/L 1,360 mg/L
Chloride loss	2,096 mg/h	1,747 mg/h
Potassium concentration	8.7 mmol/L	8.4 mmol/L
Salt (NaCl) loss	¾ teaspoon per hour 46th percentile	½ teaspoon per hour 34th percentile

That sodium concentration (759 mg/L for the bike and 690 mg/L for the run) is within the range given by that first Gx sweat patch, and the percentiles are interesting—perhaps I’m not that salty of a sweater after all. However, looking at these numbers, it’s clear I wasn’t adequately replenishing fluids and electrolytes at Coeur d’Alene. LEVELEN recommends replenishing up to 75% of the sweat loss and 100% of electrolyte losses; I only consumed about 500 mL of fluids and 200 mg of sodium per hour, so I was nowhere near close to hitting those numbers. They also recommend preloading fluids and salt in the 36 hours leading up to the race, which I didn’t do.

The rest of the report goes on to make recommendations for fluid and electrolyte replenishment based on my results, using their level 3 electrolyte drink mix (the levels indicate the amount of sodium in the mix, with level 3 having a moderate amount). LEVELEN recommends I drink 35 oz (1.035 L) of their drink mix per hour, containing 500 Calories and 925 mg of sodium. That would certainly accomplish the goal of replenishing 75% of sweat loss and 100% of sodium loss, but 500 Cal/h seems bonkers to me—that’s 125 grams of carbs per hour, far more than the 80 g/h I’ve been training and racing with, and more than I think I can stomach!

I might be able to train my gut over time to handle that many carbs, but for now I’d rather stick with the Maurten 320 drink mix I’ve been using. Thankfully, LEVELEN’s recommendations clearly explain how to preload and replenish fluids and sodium, and in which quantities, so I’m confident I can supplement Maurten with additional sodium and water as needed.

Although this sweat test kit is pricier than the Gatorade Gx sweat patches, and it takes a few weeks to do the tests and get the results, it was absolutely worth the investment. I got the exact information I needed to start making much-needed changes to my fueling and hydration strategy.

What’s next?

The workouts I did for my LEVELEN tests didn’t quite match race conditions, so my next step is to wait for an appropriately warm day, do some bike rides and runs at race pace, and get a more accurate sweat rate for those conditions. With that, and the sodium concentration range I got back from LEVELEN, I can figure out how much water and sodium I need to supplement my race hydration and nutrition.

I haven’t decided yet how I’ll do that, so I’ve ordered a few things to try out: Precision Fuel & Hydration’s PH 1500 tablets, which I could use for sodium preloading before the race, and I could put them in my Maurten bottles on the bike; their electrolyte capsules, which would be easy to carry and take during the run; and BOA Endure Blast sodium sprays, which I could easily carry in a top tube bag to use on the bike and have the advantage of being sold at the Ironman race expo, so I could stock up there if I have to.

I’m not sure these sweat test results fully answer the question of what went wrong at Coeur d’Alene, so I’ll keep researching that, but at least now I’m at a good starting point to experiment with sodium supplementation and see how my body reacts to it. If nothing else, I feel optimistic I can dial in my nutrition and hydration plan to prevent any sodium-related issues in future races.