How PODIUM Made These Calls: Five Design Decisions, Defended

Most of these calls were the harder option at the time. The defaults are easier to ship and easier to explain. They're also wrong about half the time, in ways the athlete never sees but pays for at mile 21.

What follows is the defense of each. The alternative we considered, the evidence behind the choice, and the trade-off that comes with it. Cross-checked against the live engine, not against the marketing copy.

The decision: the carb prescription is min(research_target, capacity, sport_ceiling), with a gate-keyed safety floor that overrides the result on long sessions where the smallest would put the athlete in bonk territory.

The alternative we considered: a baseline × duration × intensity multiplier cascade. A coefficient stack feels intuitive because each input nudges the result up or down, and it's straightforward to implement.

The evidence: a multiplier cascade can't represent the case where capacity is the binding constraint. It just multiplies anyway and overshoots. The constraint model surfaces "your gut is the limit today" as a first-class output, not a hidden ceiling buried inside an unmoving coefficient. Smith et al. (2013) tested 51 cyclists across 12 dose levels and found peak performance predicted around 78 g/h, but only when subjects could actually tolerate that rate. The intolerance was the limiter, not the prescription. Jeukendrup's 2014 personalized guidelines built the same logic into duration-banded recommendations, and the four-constraint model extends it by carrying capacity through every session.

The trade-off: the constraint model is more complex to explain on a marketing page. It's less complex to reason about once you've seen it run. Tap the audit drawer in the app and the four numbers are visible. Whichever is smallest wins, and the engine names the binding constraint so the athlete knows whether the limit is the work, their gut, or the ceiling.

Full mechanics: How Many Carbs Per Hour and How PODIUM's Algorithm Works.

The decision: a graduated 5-tier multiplier applied to the athlete's sweat-profile sodium baseline. 0.85× below 59°F, 1.00× from 59 to 72°F, 1.25× from 72 to 82°F, 1.50× from 82 to 91°F, and 1.75× above 91°F. Hard-capped at 1,500 mg/h after the multiplier.

The alternative we considered: a binary cap. Hot day gets one fixed bump; everything else stays at baseline. This is the simpler approach, and at first glance it looks adequate.

The evidence: Buono 2007 showed that sweat sodium concentration scales with sweat rate non-linearly. As flow goes up, sodium secretion increases proportionally more than reabsorption. A binary cap underdoses a 95°F day because it treats it the same as a 78°F day, and it has no answer at all for a 45°F day, which should run a lower sodium target because sweat rate drops with cold. The graduated curve handles every condition the athlete will actually encounter, in both directions.

The 1,500 mg/h hard cap is downstream of Hoffman & Stuempfle 2015, which identified over-drinking (forced fluid intake to chase a beverage that's hard to drink) as the dominant risk factor for exercise-associated hyponatremia. Beverages above ~1,150 mg/L (~50 mmol/L) of sodium drive over-drinking. Capping sodium output keeps the prescription drinkable, which keeps the athlete from drowning the math.

Full mechanics: Fueling in the Heat.

The decision: acclimated athletes get the same heat multiplier as un-acclimated athletes on the same day. The curve doesn't reduce for heat exposure history.

The alternative we considered: reduce the multiplier for athletes who've been training in heat for several weeks, on the assumption that their sweat is now less salty.

The evidence: Buono 2018 documented that heat-acclimated athletes do sweat at a lower sodium concentration (~34% drop). The catch is that they also sweat in much higher volume (often roughly doubled). Composition × volume nets to more total sodium loss, not less. Reducing the multiplier for an acclimated athlete would underdose them exactly on the day the math matters most.

This is one of the calls where the intuitive answer is the wrong one. "Acclimated runners need less salt" sounds right at the concentration level, and it's wrong at the absolute level. The engine works in absolute milligrams per hour, not in concentration, because that's what the athlete actually has to drink.

The trade-off: the reduced-for-acclimated version is easier to demo and harder to defend against the underlying physiology.

The decision: the carb prescription and the sodium prescription run as parallel calculations. There is no carb-coupled sodium bonus, no sodium-driven carb adjustment.

The alternative we considered: add bonus sodium when carb intake rises, on the basis that SGLT1 (the glucose transporter) co-transports sodium with glucose at a 1:1 ratio.

The evidence: SGLT1 does need sodium to move glucose, but the sodium isn't consumed in the process: it cycles back into circulation. The body's endogenous gut sodium supply (~18 g/day from biliary and pancreatic secretions, per gastrointestinal physiology textbooks) dwarfs the demand SGLT1 generates at any carb intake rate PODIUM would prescribe. Direct experimental work confirms it: Gisolfi 1995, Jeukendrup 2009, and Fordtran 1967 all varied beverage sodium concentrations and found no meaningful change in fluid or glucose absorption attributable to sodium content. The 2023 GSSI review concluded the same.

The sodium prescription is driven by what the athlete is losing through sweat. The carb prescription is driven by what the work demands. They get drawn from the same beverage in the field, but the math doesn't link them, because the physiology doesn't link them.

Full mechanics: Sodium Isn't Just for Cramps.

The decision: the constraint model applies identically to every athlete. No sex coefficient, no luteal-phase modifier, no cycle-phase adjustment.

The alternative considered: apply a fractional coefficient that lowers the prescription for female athletes. The intuition is appealing because average body mass differences between male and female athletes are real.

The evidence: absorption is transporter-limited, and transporter capacity is sex-independent. Jeukendrup et al. (1997) measured peak exogenous glucose oxidation across body sizes and found no difference attributable to mass. The bottleneck is intestinal transporter density, not body weight. Lukasiewicz et al. (2024) modeled female endurance athletes and concluded they may need more exogenous carbohydrate than males during racing efforts, not less, because glycogen stores deplete proportionally faster at race-pace intensity. The hormonal-cycle substrate-shift effect, when it shows up, is small and largely abolished by exogenous CHO ingestion (Bailey 2021).

The trade-off: coefficient-based fueling is easier to market because it appears tailored. The risk runs asymmetrically: a small over-fuel costs nothing and is recoverable in an afternoon. Chronic under-fueling costs periods, bone density, and years of training adaptation. RED-S is downstream of the underfueling case, not the overfueling case. The defensible call is the one that biases toward the cheaper failure mode.

Full mechanics: Fueling for Female Runners.

These five calls have one thing in common. They're not the easy choices. Most of them are harder to explain than the defaults. But each one is what the evidence supports, and each one keeps the prescription right on the day the math matters most.

The engine is the sum of these decisions. The athletes are the people who get to run on the result.

// FREE RESOURCE

The First Marathon Fueling Protocol

The exact fueling blueprint to execute your first 26.2 miles with zero guesswork.

• •Carb & sodium guidelines
• •Race week and race day fueling timeline
• •Gut training program

Get Your Free Guide