What this chapter is for

This chapter is a way to observe, carefully and safely, what a modest change in meal timing does for one person: hunger, energy, sleep and a simple measure of attention, with optional ketone and glucose readings. It is an observation log, not a treatment, a weight-loss plan or a challenge. It goes no further than the 16:8 schedule, and it is written for healthy adults only.

The four weeks

Keep the same foods and roughly the same amount you normally eat. Do not try to eat less; the point is to change timing only, so that what you observe can be attributed to timing. Keep the end of the window fixed and move its start, so that the evening, and therefore sleep, stays the same.

The daily log

Fill in one row each day. Paper works as well as a spreadsheet.

The reaction-time test is a short version of the psychomotor vigilance test, widely used in sleep research: for three minutes, a stimulus appears at random intervals and you respond as fast as you can; responses slower than 355 milliseconds count as lapses (Basner et al. 2011; Dinges and Powell 1985). Several free applications offer it. Do it at the same clock time each day, ideally in the late morning, which in week 3 falls near the end of the fast.

Stop rules

Eat, and stop that day’s fast, if you notice any of: faintness or dizziness on standing; palpitations; confusion or difficulty concentrating that worries you; shakiness, sweating or sudden weakness; vomiting; a severe headache. Record the day as a warning day.

Stop the whole programme, and talk to a doctor, if warning signs occur on two days in any one week, if you faint, or if any symptom is severe or does not settle after eating.

Stop also if the log begins to affect how you feel about food: if you find yourself thinking about eating most of the day, feeling guilt after meals, eating much more at night to make up, or wanting to push the fast longer than planned. Those are signs that the practice is not good for you, whatever the numbers say.

Reading your four weeks

import random

# Twenty-eight days of an invented example log, one row per day.
rng = random.Random(5)
WEEKS = {1: ("12:12", 12), 2: ("14:10", 14), 3: ("16:8", 16), 4: ("16:8", 16)}
log = []
for day in range(1, 29):
    week = (day - 1) // 7 + 1
    window, fast_hours = WEEKS[week]
    log.append({
        "week": week,
        "fast_h": fast_hours - (rng.random() < 0.2) * 2,                 # some days end early
        "sleep_h": round(rng.gauss(7.0, 0.5), 1),
        "hunger": min(10, max(0, round(rng.gauss(3 + (week == 3) * 1.5, 1.2)))),   # 0-10
        "energy": min(5, max(1, round(rng.gauss(3.4, 0.7)))),            # 1-5, afternoon
        "lapses": max(0, round(rng.gauss(4.0, 1.5))),                    # 3-minute test
        "warning": rng.random() < 0.03})                                 # dizziness ...

print("week  plan    fast (h)  sleep (h)  hunger  energy  lapses  warning days")
for week, (window, _) in WEEKS.items():
    rows = [r for r in log if r["week"] == week]
    mean = lambda k: sum(r[k] for r in rows) / len(rows)
    print(f"{week:4d}  {window:6s}  {mean('fast_h'):8.1f}  {mean('sleep_h'):9.1f}  "
          f"{mean('hunger'):6.1f}  {mean('energy'):6.1f}  {mean('lapses'):6.1f}  "
          f"{sum(r['warning'] for r in rows):12d}")
# week  plan    fast (h)  sleep (h)  hunger  energy  lapses  warning days
#    1  12:12       11.7        7.1     2.7     3.9     3.4             0
#    2  14:10       13.1        7.0     2.7     3.9     4.7             1
#    3  16:8        15.1        7.1     4.7     3.3     4.1             0
#    4  16:8        16.0        7.0     2.4     3.3     3.6             0

In this invented example, the person’s fasting hours followed the plan (a few days ended a little early). Hunger before the first meal rose in the first week of 16:8 and fell back the week after, a common report as people adjust. Energy in the afternoon was a little lower in the two 16:8 weeks. Lapses moved up and down from week to week by more than any trend. There was one warning day, in week 2: by the rules, that day’s fast should have ended at once; one day does not stop the programme.

The rule for week 4. At the end of week 3, compare it with week 1. Keep 16:8 for week 4 only if all of the following hold; otherwise return to 12:12:

1. there were no warning days in week 3;
2. average sleep was not more than half an hour shorter than in week 1;
3. average energy was not lower by a full point or more;
4. hunger before the first meal was usually 6 or below;
5. you would like to continue.

In the example, all five hold, so week 4 continued with 16:8.

After the four weeks

Whatever you decide, the log is worth keeping. If you stop, your notes are the record of how the change felt. If you continue with 16:8, an occasional week of the log, perhaps once every few months, shows whether anything has changed. If you have a medical check-up, take the log with you. This book does not suggest going further than 16:8; Chapter 6 explains why longer fasts belong to a different category of risk.