Caffeine-Timed Workday Builder

The last-dose deadline

The hard line in the grid comes from Gardiner and colleagues, a meta-analysis of 24 crossover trials in healthy adults aged 18 to 65, in Sleep Medicine Reviews in 2023. Across the 37 effect sizes in its total-sleep-time analysis, from 340 participants, caffeine cost about 45 minutes of sleep on average, and the loss grew the closer to bed the dose landed. Three products came out of that model: a 47 mg cup of black tea, for which no cut-off could be identified; a 107 mg coffee, which needs 8.8 hours; and a 217.5 mg serving of pre-workout, which needs 13.2.

Two qualifications. The cut-offs cover total sleep time only: timing did not reliably move how long people took to fall asleep. And only the two upper figures are the paper’s, so deadlines for other doses are interpolated between them, a curve fitted to two points, best read to the nearest half hour. The dependable part is the direction: doubling the dose adds roughly four hours rather than doubling the wait. Milligrams and hours are all it responds to, so a pouch, an energy drink and a strong tea at the same dose share a deadline.

The dip, and what chronotype actually does to it

The afternoon slump is not caused by lunch. Monk, Buysse, Reynolds and Kupfer ran 12 people through a standardised routine with lunch replaced by hourly liquid supplements, and the dip still appeared in 5 of them. Reyner and colleagues, with 12 drivers on a simulator, called it largely unrelated to lunch in 2012. Monk’s account is that the dippers had a stronger, later-peaking 12-hour component in their temperature rhythm. That is the standard explanation and it rests on twelve people.

Chronotype gets in through circadian phase, not the dip directly. Lack and colleagues, running 10 morning types and 12 evening types through a modified constant routine in 2009, found temperature and melatonin markers 2 to 3 hours later in the evening types, so a later phase should carry the dip later. But nobody has measured the dip’s clock time in morning versus evening types. The evidence that morning types dip harder is thinner than it sounds, resting on a 1989 study with 5 morning and 3 evening types, and Monk’s 12 people were tested for morningness as the explanation of who dipped, which it was not. So the shift here is deliberately smaller than that 2 to 3 hour gap, because your wake time already carries most of the phase information. It is a construction, not a measurement, and if your slump lands elsewhere your observation beats it.

The second dose sits 45 minutes ahead of the band, splitting the difference between routes: a caffeine solution peaked at about 30 minutes in ten healthy men (Blanchard and Sawers, 1983), while in an 84-person comparison gum peaked at 44 to 80 minutes and capsules at 84 to 120 (Kamimori and colleagues, 2002). Move it earlier for capsules.

Two rules this builder ignores

Wait 90 to 120 minutes after waking before the first coffee. Antonio and colleagues went through this one in the Journal of the International Society of Sports Nutrition in 2024 and were direct: there is no evidence that caffeine on waking causes an afternoon crash, or that delaying it prevents one. Caffeine does lift cortisol, but it lifts it whenever you take it, and hard morning exercise produces the same rise without anyone proposing you avoid that. Their one real caveat is about effect size: after a full night there is less adenosine to displace, and delaying pushes the afternoon dose towards the deadline.

The 90-minute work block descends from Kleitman’s basic rest-activity cycle, described in sleep and then proposed to continue through waking. Even in sleep it was not a constant: the 20 adults in Aserinsky and Kleitman’s 1953 recordings had intervals from 1 hour 10 minutes to 3 hours 50. Waking studies disagree, and the largest, Neubauer and Freudenthaler with 60 people tested every ten minutes for nine hours, found no 90-minute periodicity in anything. Work in blocks of 90 if they suit you. It is a rule of thumb, not a measured unit.

What this cannot tell you

Your own half-life. The average in healthy adults is about five hours, individual values from roughly 1.5 to 9.5. Low-dose oral contraceptives stretched it from 5.37 to 7.88 hours in a 9 against 9 comparison (Abernethy and Todd, 1985); smoking shortened it from 6.0 to 3.5 in 13 against 13 (Parsons and Neims, 1978). The bedtime estimate uses 4 and 6 hours, which brackets the average, so at either end of the real spread it is wrong in the obvious direction.

Whether the dose does anything for performance. The ISSN position stand (Guest and colleagues, 2021) puts the ergogenic range at 3 to 6 mg per kilogram, roughly 210 to 420 mg for a 70 kg adult. Workday doses sit well under that, which is fine: most people dose to feel like themselves, not to perform. The US Food and Drug Administration describes up to 400 mg a day as not generally associated with negative effects in healthy adults, which is where the builder warns.

And it is arithmetic on times you typed in, not advice about your health. Caffeine interacts with a number of medications and deserves a different conversation in pregnancy, or with an anxiety or heart condition. If your sleep or your energy has been wrong for more than two weeks, raise it with a clinician rather than with a spreadsheet.

If the dose on your grid is a pouch

One Roon pouch is 80 mg of caffeine with 60 mg of L-theanine, 25 mg of Dynamine (methylliberine) and 5 mg of theacrine, from $9.16 per tin. It is worth naming only because it is an exact number where a mug of coffee is a guess: pick 80 mg and the deadline you get is the one that applies. No pharmacokinetic study of caffeine pouches has been published, so the 45-minute lead time borrows the gum figures.