Nutrition Timing and the 10-Day Adjustment Rule

Nutrition Timing and the 10-Day Adjustment Rule

Why the best AI coaching systems change calories slowly, on purpose, instead of reacting to every noisy weigh-in, travel day, and gut fluctuation.

The clearest practical lesson in the available coaching record is blunt: after a few days of slightly lower water intake during travel, the client was “backed up for a couple days,” and Justin’s response was to anticipate the pattern rather than chase it with constant new tweaks. That is the mechanism in one phrase: bowel-content lag. In prep, the sharp thesis is this: if you adjust nutrition before the signal is real, you end up treating noise as physiology, and the athlete pays for it with overcorrection.

The most useful coaching move is not to do more. It is to wait long enough for the body to show you what actually changed.

The problem isn’t lack of data. It’s reaction speed.

AI coaching makes it easy to see everything: daily weigh-ins, photos, steps, water, food logs, digestion notes, training output. That visibility creates a new failure mode. Coaches can now overfit to short-term variation.

A scale jump after a salty meal can look like fat gain. A flat look after travel can look like a failed carb setup. A constipated midsection can look like the need for more drastic food cuts.

But the coaching note in the source points to a simpler chain:

1. Water intake dipped a little during travel.
2. Back-up followed for a couple of days.
3. Distention showed up.
4. The coach interpreted it as a predictable prep issue, not an emergency.

That is the kind of causal sequence you want an AI system to recognize. Not “numbers changed, therefore change the plan,” but “inputs shifted, so wait for the lagging outcome.”

Why timing matters more than the tweak itself

In physique prep, nutrition changes often have delayed visibility. The immediate response to a change in water, fiber, food volume, or meal timing is not always body composition. It is often gut function, fullness, and scale noise.

The practical implication is straightforward: if you make three changes in three days, you won’t know which one mattered.

That’s where over-adjustment does damage:

• Too many calorie cuts can flatten training before the real issue is clear.
• Too many refeeds or “fixes” can hide the signal you were trying to observe.
• Too many fiber or fluid changes can worsen the very distention you were trying to solve.

The record from Justin’s coaching style supports the opposite approach: identify the likely mechanism, set a timeline, and avoid panic changes. In the quoted note, the idea was not to reinvent the plan every morning. It was to expect that the gut “dries out faster in prep,” that food backs up, and that distention can build as a result. That is a timing problem first, not necessarily a calorie problem first.

A better rule: match adjustment speed to the delay in the signal

A good AI coach should not be calibrated to the fastest available metric. It should be calibrated to the slowest meaningful signal.

If the issue is bodyweight trend, wait for trend. If the issue is digestion, wait for digestion pattern. If the issue is training performance, wait for repeated performance decline, not one rough session.

This matters because different signals move on different clocks. Daily scale weight can swing immediately. Digestive outcomes may lag by a day or two. Visual changes may lag even longer. If the coach responds to each one as if it were equally trustworthy, the athlete gets pulled in contradictory directions.

In practice, that means:

• Don’t change calories because one weigh-in was off.
• Don’t change meal timing because one meal sat heavy.
• Don’t change the whole plan because travel changed water intake for 24 hours.

The goal is not stubbornness. The goal is respecting latency.

Nutrition changes should be staged, not stacked

The biggest mistake I see in AI-assisted coaching is stacked intervention.

A coach sees a problem and changes:

• calories,
• cardio,
• meal timing,
• water,
• and fiber, all at once.

Then the athlete improves or worsens, and nobody knows why.

Justin’s note points toward a different logic. If prep distention is partly a gut-filling and constipation issue, the response should be conservative and specific. Adjust one lever, then allow enough time to observe whether the lever actually affected the outcome.

That is especially important when the athlete is already close to the edge of adaptation. In harder prep phases, the margin for error is smaller. A small, unnecessary change can make an already sensitive athlete feel worse fast.

So the right coaching question is not “What can we change today?” It is “What single change has the highest chance of solving the actual problem without obscuring the next read?”

AI can help, but only if it resists the urge to be helpful

This is where AI coaching is most useful and most dangerous.

Useful, because it can surface trends across weeks that a human coach may miss in a busy week. Dangerous, because it can generate confident action from weak evidence.

A competent system should be able to say:

• “The trend is not established yet.”
• “That is likely travel noise.”
• “Hold the current setup until the lagging indicators settle.”
• “Do not stack another change on top of a recent change.”

That sounds unexciting. It is also what keeps a coaching system honest.

The temptation is to make every check-in actionable. But not every check-in deserves an intervention. Sometimes the correct move is to preserve the current setup long enough for the body to answer.

The coaching takeaway for today’s daily angle

For nutrition adjustment timing, the winning habit is patience with structure.

Make changes when you need to, but make them in a way that can be evaluated.

That means:

• choose one variable,
• give it enough time to show an effect,
• read the delayed signal,
• and avoid the reflex to “fix” what is still unfolding.

In physique coaching, over-adjustment is rarely a sign of intensity. It is usually a sign that the coach is reading the wrong time horizon.

If AI is going to improve coaching, it will not be by making more changes faster. It will be by making fewer changes, later, and for better reasons.

Sources Used

• raw/_consumed/2026-05-31/kahunas-export/2026-05-31-w19-24m/clients/david_lamartina___members-tlssnsjthkmnhfqcscszce25acz_vhdm_x2_xdlpx_i.json
• raw/_consumed/2026-05-26/troponin_supplements_kb.md
• raw/_consumed/2026-05-26/troponiniq_kb.md
• modules/08-voice/kahunas-coaching-deep-voice.md