Anneal

Blog Recovery

How often should I deload? A cadence by training age

6 min read

"How often should I deload?" is the question every intermediate lifter eventually types into a search bar, usually around week six of a block when the bar starts feeling slower than it should. The honest answer is that it depends on your training age, your goal, and how your body is responding to the current load. There are defensible starting points, and there are signals that override the starting point on any given week. This post covers both. If you want the underlying mechanics first, here is what a deload week is and why it works.

The short answer

If you want a single number to anchor on, here is the cadence band most lifters land in:

Modifiers: if you are in a cut or under heavy life stress (poor sleep, high work load, a new baby, a move), shorten the cadence by one step. A cutting intermediate often runs closer to the advanced cadence. The numbers above are starting points, not destinations.

Why training age changes the answer

The reason cadence shifts with experience is that the recovery curve shifts with experience. Beginners recover from most sessions within a day or two. Each session you put in lands on a mostly-rested system, so accumulated fatigue stays low for weeks at a time. You can push hard, miss a planned deload, and keep progressing.

Intermediates accumulate fatigue across the week rather than within a session. Your Tuesday lower body session is still partly bleeding into your Friday lower body session, even when you feel recovered. By week four or five, those overlapping recovery windows start stacking, and a deload is what gives them room to drain.

Advanced lifters accumulate fatigue across the block. Heavier weights, more total volume, and tighter proximity to failure all carry longer recovery tails. The same training volume that would have been productive at year two is now riding closer to your maximum recoverable load. Cadence tightens because the gap between adaptive stimulus and overreaching stimulus narrows. The starting-point numbers above reflect that, but they are still starting points. Your signals get the final call.

Why the calendar is not enough

Two lifters at the same training age can need deloads on different weeks. One is sleeping eight hours, eating in a small surplus, and training in a quiet home gym. The other is sleeping six hours, cutting, and squeezing sessions around a demanding job. Same program on paper, different recovery context.

The calendar gives you a default. Autoregulation gives you the override. Without the override, you either deload too often (and lose adaptive stimulus) or too late (and grind into a stall). The middle ground is to plan a base cadence and let signals move the date forward or back when they show up.

The recovery models worth knowing

A few concrete models make the override easier to apply. None of these are required reading, but they sharpen the intuition behind "your body will tell you."

Mujika fatigue decay. Mujika and Padilla showed that fitness and fatigue decay at different rates after training stress. Fitness has a half-life around 21 days; fatigue decays much faster, on the order of 5 to 7 days. A deload exploits the gap. Drop training stress for a week and fatigue empties out faster than fitness erodes, so you return to the next block with most of your strength gain intact and most of the fatigue gone.

ACWR (acute:chronic workload ratio). Hulin and colleagues formalized this in sports-injury research. Take your last 7 days of training load and divide by your trailing 28-day average. Ratios above roughly 1.5 sit in an elevated injury-risk band. Ratios below 0.8 suggest you are detraining. The healthy band is around 0.8 to 1.3. If you spike the acute window for a few weeks in a row, your ACWR climbs, and a deload week is what brings the ratio back into range.

RIR drift. If you log Reps in Reserve set by set, the drift between planned and actual RIR is one of the cleanest early signals you have. Planned RIR 2 sets that start landing at RIR 0 across a week mean your fatigue is masking your real capacity. Same lift, same weight, fewer reps in the tank. RIR explained covers the scale and how to track it set by set.

Stagnation. Two or three sessions in a row of flat or regressing numbers on the same lift, with no obvious explanation (sleep, food, or a missed session), is your block telling you it is done. This is the lowest-tech signal and usually the easiest to act on.

When to deload sooner or later than the table says

Look for stacks of signals, not isolated bad days. Deload now if two or more of the following are true for two weeks running: sets are grinding earlier than they should, the bar is slower out of the bottom, joint noise is louder, sleep or mood has dropped, your logged RIR is drifting below plan, or your numbers are flat across multiple sessions. One bad workout is just a bad workout. A stack across two weeks is the cue.

You can also delay a planned deload if the signals are not there. If your numbers are still moving, RIR is stable against your plan, your sleep is fine, and the bar feels snappy, push one more productive week and deload the following one. Let the signals drive the date.

How Anneal flags deload timing

Anneal's tagline is "knows when to deload," and the cadence question is what the deload system is built to answer.

The app evaluates eight signals on a rolling basis. Mesocycle length sets the base calendar cadence by experience level. Accumulated overreach watches the trailing volume against your ACWR-style baseline. Missed sessions, grinding sets (driven by RIR drift), scheduled cycles, stagnation across two to four sessions, and a couple of profile-tuned readiness signals each contribute their own threshold. When enough of them cross at once, Anneal surfaces a deload suggestion, explains which signals fired, and offers to drop your working weights for the week automatically.

The progression tuning layer shifts every threshold by your experience and goal, which is why a cutting intermediate gets a tighter cadence than a bulking one on the same program. A calendar-only deload misses the lifter who is grinding by week four. A signal-only deload misses the lifter who feels fine but is accumulating fatigue ahead of an injury. Combining both is what catches the timing right.

Wrap-up

Pick a base cadence from the short-answer list above. Trust it for four or five weeks. Let signals override it when they stack.