FN-06Field note 6 of 6 · Tart cherry, in real life

During a heavy training block, does two weeks of daily tart cherry concentrate move my soreness or my sleep quality by at least 0.5 points?

Two co-primary outcomes, zero movement on either, and a dosing footnote that matters more than the result.

Null

Dates

Jan 05 – Jan 25, 2026

Duration

21 days

Pre-registered primary

Soreness (0–10) + sleep quality (1–5), co-primary

FN-06 · 01

The question

Tart cherry is the rare recovery supplement with actual trials behind it: Montmorency concentrate carries anthocyanins with plausible anti-inflammatory and antioxidant activity, plus a small amount of melatonin, and studies in marathoners and team-sport athletes have reported less soreness and better sleep. It is also the rare supplement my mom approved of without a hearing, because it is essentially aggressive juice.

The question I could afford to run: one 30 mL dose of concentrate daily for two weeks, spanning the heaviest club block of my winter, against a one-week baseline. Two co-primary outcomes, soreness and sleep quality, a decision I already regret and will get to. The heavy block was deliberate: soreness needed headroom to fall, and January was going to supply soreness whether I experimented or not.

FN-06 · 02

Pre-registration

Pre-registration · sealed recordLocked Jan 04, 2026 · 18:55, before the baseline week

Protocol: 30 mL Montmorency tart cherry concentrate in a glass of water, once daily with dinner. Everything else unchanged through the block.
Duration: 7-day baseline, then 14 days on, spanning the January club block.
Primary outcome: Two co-primaries: next-morning soreness (0 to 10, half steps) and sleep quality on waking (1 to 5, half steps).
Secondary outcomes: Estimated time to fall asleep, minutes
Morning resting heart rate
Counts as null: Under 0.5 points of change on a co-primary is a null for that outcome; both null makes the entry a null. Honesty clause, recorded now: the studies mostly used 30 mL twice daily, often with a pre-loading week. I am using once daily with no pre-load, because concentrate is expensive and the realistic version of this habit is the one worth testing. Noted here so I cannot pretend otherwise later.

Saved as a dated, unedited note before day one. The goalposts above are the goalposts the data was judged against; nothing in this record was touched after the first measurement.

FN-06 · 03

The protocol

Fourteen evenings, fourteen doses, zero misses; compliance is easy when the protocol tastes like dessert vinegar and takes eleven seconds. The block itself cooperated by being genuinely heavy: six practices a week plus a weekend tournament in the middle of the intervention window.

Both ratings happened before getting out of bed, soreness first, sleep second, same order every day to keep the two opinions from negotiating with each other. Resting heart rate logged from the chest strap as a free objective bystander.

Dose

30mL/day

Studied dose

30×2mL/day

Compliance

14/14doses

Pre-load

none

FN-06 · 04

The log

The actual daily data, charted with the intervention window shaded, plus the raw log as a table under each chart. Same arrays feed both, so they cannot disagree.

Primary outcomeNext-morning soreness, daily0–10 scale

Baseline mean 5.00, intervention mean 5.00. Not approximately: to two decimals, on my coarse little scale, identical.

›View the raw log · 21 rows

Next-morning soreness, daily, in 0–10 scale. Full daily log.
study day	Date	Phase	Soreness (0–10 scale)
1	Jan 05	baseline	5
2	Jan 06	baseline	4.5
3	Jan 07	baseline	5.5
4	Jan 08	baseline	5
5	Jan 09	baseline	4.5
6	Jan 10	baseline	5
7	Jan 11	baseline	5.5
8	Jan 12	intervention	5
9	Jan 13	intervention	5.5
10	Jan 14	intervention	4.5
11	Jan 15	intervention	5
12	Jan 16	intervention	5.5
13	Jan 17	intervention	5
14	Jan 18	intervention	4.5
15	Jan 19	intervention	5
16	Jan 20	intervention	5.5
17	Jan 21	intervention	4.5
18	Jan 22	intervention	5
19	Jan 23	intervention	5
20	Jan 24	intervention	5.5
21	Jan 25	intervention	4.5

Primary outcomeSleep quality on waking, daily1–5 scale

The second co-primary, equally unmoved: 3.07 before, 3.07 during. Resting heart rate spent the whole block between 57 and 59 bpm.

›View the raw log · 21 rows

Sleep quality on waking, daily, in 1–5 scale. Full daily log.
study day	Date	Phase	Sleep quality (1–5 scale)
1	Jan 05	baseline	3
2	Jan 06	baseline	3.5
3	Jan 07	baseline	3
4	Jan 08	baseline	2.5
5	Jan 09	baseline	3
6	Jan 10	baseline	3.5
7	Jan 11	baseline	3
8	Jan 12	intervention	3
9	Jan 13	intervention	3
10	Jan 14	intervention	3.5
11	Jan 15	intervention	2.5
12	Jan 16	intervention	3
13	Jan 17	intervention	3.5
14	Jan 18	intervention	3
15	Jan 19	intervention	3
16	Jan 20	intervention	2.5
17	Jan 21	intervention	3.5
18	Jan 22	intervention	3
19	Jan 23	intervention	3
20	Jan 24	intervention	3.5
21	Jan 25	intervention	3

FN-06 · 05

What happened

Soreness Δ

0.00pts

Sleep Δ

0.00pts

RHR range

57–59bpm

Pre-reg bar

0.5pts

Soreness: baseline mean 5.00, intervention mean 5.00. Sleep quality: baseline 3.07, intervention 3.07. Resting heart rate: flat, 57 to 59 bpm across the entire block. It is rare for data to be this cooperative about being boring; both co-primaries landed not just under the half-point bar but at zero, to the resolution of the instruments. The block was heavy, exactly as designed, so soreness had plenty of room to fall (I averaged a flat 5) and declined the invitation.

Before this becomes a takedown of a fruit, the dosing paragraph, which was in the sealed record from the start. I used half the studied dose: 30 mL once daily versus the 30 mL twice daily most trials used. I skipped the pre-loading window: trials commonly start dosing four to seven days before the stress event, and I began cold. And I have no idea what the anthocyanin content of my particular bottle was, because concentrates vary several-fold and my brand does not publish an assay.

So the precise claim this null supports is narrow: this brand, at this dose, on this schedule, in me, did nothing detectable. That is a null on my protocol, not a refutation of the molecule. The distance between those two sentences is roughly the width of the entire supplement industry, and learning to see that gap clearly was worth more than the answer.

Half the studied dose is not a smaller experiment. It is a different experiment.

FN-06 · 06 · Mandatory in every entry

Why this is weak evidence

This section is the module’s actual thesis. The result above is the least important thing on this page; the list below is why.

The protocol was sub-study from day one

Half dose, no pre-load, unknown anthocyanin content. A null under those conditions is fully compatible with the compound working as advertised at trial protocols. This was declared in the pre-registration, which converts it from an excuse into a finding about real-world dosing.

Two co-primaries was a design error

Registering two primary outcomes doubled my chances of a spurious 'hit' had either scale wobbled past 0.5 by luck. I got away with it because both landed at zero, but the win would have been weaker than it looked, and I knew better at lock time. Logged as a mistake, kept as a lesson.

The detection floor is high

Coarse half-point scales, 14 days, n of 1: a true average benefit of, say, 0.3 points is invisible to this design. It would also be worth almost nothing to me at that size, which is why the bar sat at 0.5, but 'below my bar' and 'absent' are different claims.

A heavy block cuts both ways

The load supplied soreness signal, but also soreness variance: my daily ratings swing a full point on training load alone, and that noise floor buries small effects. A calmer block would have had less signal and less noise; I chose signal and paid in noise.

No marker of exposure

With creatine, the scale confirmed the compound arrived. Here I have no manipulation check at all: no way to know whether meaningful anthocyanin levels ever reached my bloodstream. The null could live in the bottle, in my gut, or in the physiology, and this design cannot say which.

FN-06 · 07

What I would do differently

01The studied protocol or nothing: 30 mL twice daily, with a five-day pre-load before the tournament in the middle of the block.
02An ABA structure: baseline, two weeks on, washout, two weeks on again. Two intervention blocks make a flat line much harder to dismiss as bad luck.
03A brand with a published anthocyanin assay, so the exposure is a number instead of a hope.
04One primary outcome, chosen by what I actually care about, which on reflection is sleep. The co-primary hedge was me trying to give the juice two chances to succeed, and it deserves to be named as such.

FN-06 · 08

The verdict

The verdict, for me

Null

Null, twice. I finished the bottle, it tastes fine in the way that expensive vinegar tastes fine, and I did not buy another. If I ever re-run this it will be at the studied dose with a pre-load, and the pre-registration will say so in the first line.

The non-verdict, for you

When a supplement 'did not work for you,' check your protocol against the trial protocols before you blame the molecule, and when a supplement 'worked,' check the same thing before you credit it. Half the studied dose is not a smaller experiment; it is a different experiment. Mine was, it says so on the label now, and that sentence is the entry.

Previous · FN-05Foam rolling: the nullOpen Field Notes · FNBack to the results ledgerOpen

FN-06 · Cross-references

Where this entry leads

The group evidence, the mechanism, and the instruments behind this experiment live in the other modules.

Truth Engine · TEThe tart cherry audit: anthocyanins on the recordOpen Science Library · SLInflammation is a signaling cascade, not a vibeOpen

Education, not diagnosis. This is a student-authored science platform. Nothing here replaces a physician, a physical therapist, or an athletic trainer. Sudden severe pain, numbness, an inability to bear weight, or visible deformity means stop reading and get seen.