Brain Fog or Just Tired? A Five-Question Check

What this tool actually does

It runs five answers through a branch, in a fixed priority order, and lands on one of four ordinary explanations. There is no score and no scale, because a score would imply a measurement and nothing here is measured. The first rule outranks everything below it: if the fog arrived alongside a new medication, an illness in the last month or a large change in diet, or if it switched on over days rather than drifting in over months, the tool stops and points at a clinician. It does not then offer an alternative explanation, because at that point any explanation it offered would be a guess dressed up as a result.

The second rule: early afternoon, plus rest that helps at least somewhat, lands on the ordinary daily dip. The third: rest that works, on something that reads as sleepiness rather than fog, lands on accumulated sleep debt. Everything else lands on “not sortable”, which is a real answer and the one this tool reaches most readily when the complaint is fog that a full night of sleep does not touch.

Where the numbers come from

Monk 2005, a review in Clinics in Sports Medicine, is the source for the claim that the early-afternoon dip is not caused by lunch. It collects the experiments in which the dip appears in people who ate nothing and could not tell the time. Being a review, it has no sample size of its own, and it is cited here as a summary of that literature rather than as a trial.

Reyner and Horne 1997, in Psychophysiology, ran 12 sleepy drivers through a two hour monotonous afternoon drive in a simulator. On placebo, driving incidents and EEG measures of sleepiness both peaked in mid afternoon. 200 mg of caffeine cut incidents to 34 percent of placebo, and 200 mg followed by a fifteen minute nap cut them to 9 percent. Twelve people is a small study and a simulator is not an office, so treat the direction as solid and the size as loose.

Van Dongen and colleagues, in Sleep in 2003, held 48 healthy adults aged 21 to 38 at 4, 6 or 8 hours in bed for 14 consecutive nights. Attention and reaction speed degraded steadily across all 14 days in the restricted groups, while self-rated sleepiness levelled off after the first few. Belenky and colleagues, in the Journal of Sleep Research the same year, restricted 66 volunteers for 7 nights and then gave 3 recovery nights at 8 hours in bed: lapses on the psychomotor vigilance task did not return to baseline. Johns 1991 built the Epworth Sleepiness Scale on 180 adults, 30 controls and 150 sleep clinic patients, by asking how likely a person is to doze in eight ordinary situations rather than how tired they feel. That design decision is the evidence behind question three.

What it cannot tell you

Anything about cause. It cannot rule a single thing in or out. Sleep debt, a daily rhythm and a long list of medical explanations all produce the same complaint, and the only reason these five questions are worth asking is that they cheaply separate two of those from the rest. The rest are left exactly where they were.

It has no idea how much you actually sleep, because it did not ask, and Van Dongen’s result is precisely that you would not report it accurately if it had. It does not know your age, your shift pattern, your alcohol, or anything that would show up in routine bloodwork. The “not sortable” outcome is not a failure mode. It is the honest answer to a large share of real inputs, and this tool reaches it more often than a quiz designed always to deliver a satisfying verdict would.

One note on dose

The dip is the only branch here where caffeine has trial support, and the trials used 200 mg, which is more than most people picture when they say a small dose. A smaller dose is further from what was tested, not closer to it. Nothing in the performance literature applies at these amounts either: the International Society of Sports Nutrition position stand (Guest and colleagues, 2021) puts the ergogenic range at 3 to 6 mg/kg, roughly 210 to 420 mg for a 70 kg adult. And caffeine does not repay sleep debt, it covers it, which is what makes Belenky’s recovery data worth remembering. If this has lasted more than two weeks, it is worth raising with a clinician.