Short answer: the behaviour sometimes helps, the explanation attached to it is false, and confusing the two is why so many attempts fail.
Worth separating carefully, because "it's a myth" and "it's useless" are not the same claim, and most coverage collapses them.
What the term was supposed to mean
"Dopamine fasting" was coined in 2019 by Cameron Sepah, a psychologist, as a label for a cognitive-behavioural technique: identify compulsive behaviours, deliberately restrict them for defined windows, notice what drives them, and rebuild deliberate choice.
That is a real, boring, evidence-adjacent intervention. It has nothing to do with dopamine levels. Sepah has said so publicly and repeatedly — the name was memorable, not literal.
The name escaped. The method stayed behind.
What it turned into
The version that spread treats dopamine as a substance you accumulate and then purge. Under that reading, more deprivation equals more purging, so the logical endpoint is a grey room: no screens, no music, no conversation, no interesting food, sometimes no food.
There's no mechanism there. Dopamine isn't consumed and replenished like a fuel. It's a signalling molecule involved continuously in motivation, movement, and learning, and abstaining from enjoyable things does not drain or reset it.
Reviewers have been consistent. A 2024 literature review concluded the practice remains scientifically unproven and warned that extreme versions push people toward isolation, anxiety, and undereating. Harvard Health and the Cleveland Clinic have published their own corrections. The general position among researchers who study dopamine is somewhere between "that's not what this molecule does" and mild exasperation.
So why does anyone report it working?
Because the restriction protocol does several real things that have nothing to do with the stated mechanism.
It interrupts automaticity. Most compulsive use isn't chosen; it's triggered. Removing the trigger for a while makes the behaviour visible, and visible behaviour is available for decision.
It changes the reinforcement schedule. Unpredictable, frequent rewards are the ones that train hardest. Restriction breaks the schedule even when it doesn't reduce the total.
It usually smuggles in sleep. People who do a detox weekend tend to sleep more. Sleep is one of the few inputs with a defensible link to reward-system function, and it's doing quiet work that the detox takes credit for.
None of that requires a dopamine reservoir to exist. All of it survives dropping the false explanation.
Why the false explanation actively hurts
If it were merely wrong-but-harmless, this would be pedantry. It isn't.
It predicts the wrong dose. If severity is the treatment, more severity is better, and the endpoint is deprivation of things that were protective — exercise, food, people.
It predicts the wrong timescale. A tank refills over days. A reward setpoint shifts over weeks to months. People do seven grim days, feel roughly the same, and conclude they're broken. They aren't; they used a model with the wrong clock.
It hides the actual variables. The things that matter — how predictable your rewards are, how much effort gates them, how you're sleeping — are invisible under a framework where the only variable is abstinence. What's actually moving is a setpoint, not a level.
It creates a rebound story. When the fast is over, people return to everything at once and interpret the surge as proof they were "detoxed." It's the same misreading, one loop later.
A more honest version of the same instinct
The instinct behind dopamine fasting is sound: something in the pattern of my rewards is off, and I want it back under my control.
Keep that. Drop the chemistry.
Which turns one vague question — how do I reset? — into three answerable ones. Which of my rewards arrive unpredictably? Which arrive with no effort in front of them? Am I sleeping enough for anything else to matter?
Those have answers. They can be measured, which is what Myo does, and they don't require a grey room.
The reward system doesn't drain and refill. It moves a setpoint, under load, over months — and it moves back the same way.