Fifty grams in the skull
Cerebrospinal fluid helps support the brain mechanically. An idealised calculation can show how much buoyancy reduces the apparent weight of an immersed body. Sleep research reveals oscillations in this fluid, but does not provide a formula for converting hours asleep into millilitres of brain “cleaning”.
Where does the fifty-gram example come from?
Archimedes’ principle makes the buoyant force equal to the weight of displaced fluid. In the example below, a uniform 1,400 g body with a density of 1.040 g/cm³ is fully immersed in fluid with a density of 1.007 g/cm³. Its resulting apparent weight is equivalent to about 44 g. The densities are assumptions for this exercise; the result does not measure the load on an individual’s skull base. Source.
Displaced volume comes from the volume of the immersed body. Total cerebrospinal-fluid volume cannot be used as a percentage of brain immersion. The geometry, pressures, membranes and supporting tissues within the skull require a much more complex model.
A fully immersed body
Physics exerciseV = m / ρbody; Fg = m × g; Fb = ρfluid × V × g; apparent weight = Fg − Fb. The model uses g = 9.81 m/s². Equivalent mass is apparent weight divided by g.
What do we know about cerebrospinal fluid during sleep?
Fultz and colleagues observed coupling between electrical brain activity, the haemodynamic signal and cerebrospinal-fluid oscillations during NREM sleep in 13 participants. They described slow oscillations around 0.05 Hz, or roughly one cycle every 20 seconds. The study did not measure an individual’s waste removal in millilitres as a function of sleep duration, or validate a personal multiplier for beta-amyloid clearance. Source.
How the signals can relate
Conceptual illustration with arbitrary amplitudes. These curves are not participant recordings and are not calibrated in flow, volume or amounts of substances removed. The buttons change the illustration only.
Electrical activity · Haemodynamic signal · Cerebrospinal fluid
The mechanisms and measurement of brain clearance remain under investigation. A mouse experiment published by Miao and colleagues in 2024 reported reduced tracer clearance during sleep and anaesthesia. Plá and colleagues challenged the interpretation in 2025, and Franks and Wisden replied to those criticisms. This dispute does not establish a universal rule for humans.
The approximate 500 mL per day found in physiology literature refers to cerebrospinal-fluid production over a full day. It is not the volume of a nightly brain “wash” and cannot calculate waste clearance from age, body weight, sex or hours slept.
Why this model is not a medical test
Cerebrospinal-fluid loss can be associated with a headache that changes with posture. Such a headache requires clinical assessment. This exercise cannot calculate the effects of a lumbar puncture, fluid loss or intracranial pressure, and does not establish safety thresholds.
Sources and limits of the evidence
Sources checked: 16 October 2026. An educational explanation and idealised model; this material has not undergone clinical review. Sleep evidence will be revisited when studies change the conclusions.