Dopamine Fasting

What it is

The fast sits inside an eight-step framework Lembke spells out as an acronym. Data: what is used, how much, how often. Objectives: why the person uses it, stated in their own terms. Problems: the costs. Abstinence: the four weeks. Mindfulness: staying with the discomfort rather than escaping it. Insight: what the abstinence reveals. Next steps: what the person wants afterwards. Experiment: returning, or not, and watching what happens.

Two clinical claims attach to the four weeks. A month is said to be usually the minimum needed to reset the reward pathway, and things typically get worse for the first fortnight before they improve. To her credit she also reports that about one patient in five does not respond to the fast at all, and treats that as data rather than omitting it.

The safety carve-out is hers and it is exact. She writes that she never suggests a dopamine fast to people who might suffer life-threatening withdrawal from stopping suddenly, naming severe alcohol, benzodiazepine and opioid dependence, and says that for those patients medically monitored tapering is necessary. That sentence appears once, on a single page, and is repeated neither in the chapter summary nor in the figure that sets out the acronym, which are the two places a skimming reader takes the protocol from. This publication treats the carve-out as part of the concept and never as an appendix to it.

In effect

On the duration, the two halves of the claim have to be separated, because the book does not separate them.

The lower bound is partly supported. The claim that a fortnight is not enough rests on imaging work by Nora Volkow showing dopamine transmission still below normal at two weeks of abstinence. That is a real finding, and it does the work of ruling out two weeks.

The upper claim is not supported. That four weeks is sufficient rests on a single study of heavy-drinking men with major depression, in which most no longer met criteria for clinical depression after a month. The citation itself is unstable in the book, with the author who gets named in the text absent from the bibliography and appearing as second author in the endnote. Lembke does list the confounds herself: the therapeutic milieu, spontaneous remission and the episodic nature of depression. The generalisation is the real difficulty. One sample of heavy-drinking men with major depression is carried across to entirely different reward pathways, withdrawal courses and populations, and the same number arrives at the other end. So usually the minimum and often sufficient carry nothing. They are clinical impression in the grammar of a finding, and the four weeks should be cited as her protocol rather than as an established duration.

The nearest thing to a test of the consumer version is a null. Adam Gazzaley and Larry Rosen went looking for evidence that digital detoxes work and reported that an extensive examination of the literature turned up only two peer-reviewed articles examining their effectiveness. Their verdict is that there is no evidence extended technology detoxes work, and they reject the practice over several pages in the chapter where they are supposed to be recommending things. Their mechanism for the null is the useful part: you feel better for an evening or a weekend, the detox ends, and you return to exactly the same foraging behaviour with a backlog, so returning detoxers find themselves more buried than before. Abstinence without a change to the environment is, on their account, structurally a crash diet, and their alternative is changed defaults rather than a period of heroic withdrawal. That is not a refutation of a supervised clinical protocol aimed at a dependence pattern, and it is a real constraint on abstinence offered as a general remedy for attention.

What it does not say

It does not say four weeks resets the reward pathway. It says a clinician uses four weeks, and the evidence behind that figure rules out two weeks without establishing four.

It does not transfer across behaviours. Nothing in the record bridges from heavy drinking to gaming, pornography, food or phones, and the identical number appearing for all of them is an artefact of the telling rather than a finding.

It does not apply to everyone who wants to try it, and this is the part that can cause harm. People with severe alcohol, benzodiazepine or opioid dependence should not stop suddenly. Withdrawal in those cases can be life-threatening and requires medical supervision. Any account of the dopamine fast that drops this both endangers the reader and misrepresents its author.

It does not promise a response. Around a fifth of Lembke's own patients do not respond at all.

And it is not the same thing as the popular dopamine fasting that circulated from technology circles, which is a different proposition with a different rationale and should not be conflated with a clinical abstinence trial.

---