There is a signalling system in the body that most people have never heard of, even though it sits in every vertebrate and regulates most of what has to do with balance. It is called the endocannabinoid system, and over the past fifteen years one of the most consistent findings in stress research has been that prolonged stress alters it.

This is a review of what has actually been shown, and an equally careful account of where the research runs out.

A system built on demand

The endocannabinoid system consists of receptors, the body’s own signalling molecules, and the enzymes that build and break them down. The receptors were mapped during the 1990s: CB1 was described in 1990, CB2 in 1993. The first endogenous signalling molecule, anandamide, was identified in 1992, named after the Sanskrit word for bliss.

One detail makes the system unusual. Most signalling substances are manufactured in advance and stored in vesicles until needed. Anandamide is not. It is built on demand, directly from the fatty acids of the cell membrane, at the moment it is required, and broken down shortly afterwards by the enzyme FAAH.

This is therefore not a system designed to maintain a level. It is designed to respond to a situation, and that is precisely why sustained load can hit it so hard.

What stress does: the mechanism, step by step

The relationship between stress and endocannabinoids is today among the best mapped in the field. In short: stress lowers anandamide levels, and chronic stress produces a downregulation of CB1 receptors in virtually every brain region examined.

What makes it interesting is that the mechanism is known in detail, not merely as a correlation. In a study published in the Journal of Neuroscience in 2015, Gray and colleagues showed how it works:

1. Stress releases CRH, corticotropin-releasing hormone.

2. CRH activates the CRHR1 receptor in the amygdala.

3. CRHR1 activation produces a rapid increase in FAAH activity, the same enzyme that breaks down anandamide.

4. Anandamide levels in the amygdala fall.

5. The inhibitory tone anandamide normally exerts weakens, the HPA axis is activated more readily, and anxiety behaviour increases.

The researchers also showed the reverse: when FAAH was blocked, the effect did not occur. This is therefore not a consequence of stress in general, but a specific enzymatic event.

The field as a whole is summarised by Morena, Patel, Bains and Hill in Neuropsychopharmacology in 2016, a review that remains the standard reference.

And it is not only animal models

The objection to all of the above writes itself: these are rats and mice.

But there is human data, and it points the same way.

In a PET study published in Molecular Psychiatry in 2013, Neumeister and colleagues examined 25 untreated individuals with post-traumatic stress disorder, 12 trauma-exposed controls and 23 healthy controls. Participants underwent magnetic resonance imaging and a PET scan with the radiotracer [11C]OMAR, which binds to the CB1 receptor.

The result:

  • The PTSD group had more CB1 receptors, measured brain-wide, 19.5 and 14.5 per cent higher than the two control groups respectively (p = 0.001)
  • The PTSD group simultaneously had lower levels of anandamide
  • The three measures together, CB1 binding, anandamide and cortisol, classified around 85 per cent of cases correctly

This is the picture of a system that has turned up its reception because the signal is failing.

But the direction is not the same as in the animal literature. There, chronic stress produces a downregulation of CB1 in virtually every brain region examined. Here, availability is instead higher in people with PTSD. Whether the difference is one of species, of time course, or of an established diagnosis not being the same thing as ongoing stress, is unsettled.

Why it matters beyond PTSD research

PTSD is an extreme, and nothing here should be read as suggesting that ordinary working stress is the same thing. But the mechanism described, CRH, FAAH, anandamide, is not diagnosis-specific. It is the body’s general stress machinery, and that connection is the whole point: sustained vigilance is not a feeling, it is a biochemical state that can be measured.

It also gives physiological grounding to a hypothesis long discussed at the margins of the field. Ethan Russo proposed in 2004, and reconsidered in 2016, that an underperforming endocannabinoid system may lie behind conditions that lack clear findings but share features, migraine, fibromyalgia, irritable bowel. He called it clinical endocannabinoid deficiency. The hypothesis has been criticised as difficult to test. Stress research gives it at least a credible way in: here is a documented mechanism by which such a deficit could arise.

Where the research runs out

Here the text must change register, because it is at the next step that most articles on the subject go off the rails.

Nothing above shows that anything external repairs this.

What exists is two separate observations pointing towards one another:

The first: prolonged stress lowers the body’s anandamide via a known enzymatic mechanism in animal models, and people with PTSD have lower anandamide levels than controls. The latter is a comparison between groups at one point in time, not a measurement of stress having lowered the level.

The second: in a clinical study by Leweke and colleagues, published in Translational Psychiatry in 2012, in which cannabidiol was compared with the antipsychotic drug amisulpride, serum levels of anandamide rose significantly in the CBD group, and the increase correlated with clinical improvement. The researchers proposed that CBD slows the breakdown of anandamide rather than replacing it.

The conclusion that lies nearest to hand, that the one should therefore remedy the other, has not been tested clinically by anyone. Leweke’s study concerned schizophrenia, not stress. It was small. The mechanism is disputed, some researchers hold that the effect runs via binding proteins rather than FAAH inhibition.

That two findings point towards one another is not a third finding. It is a hypothesis waiting for someone to test it.

Why the gap exists

It is worth a closing reflection that the question is so obvious and yet untested.

Part of the explanation is structural. Cannabis was placed in 1961 in the strictest category of the UN convention and in 1970 in Schedule I in the United States, a category defined as having no accepted medical use.

Research required not only a licence but raw material from a single federally licensed cultivation. That monopoly was not broken until 2021.

Most of the controlled human studies on cannabidiol have been published since 2010.

The field is therefore not thoroughly investigated and awaiting summary. It is young. And this question, what happens to a stress-depleted endocannabinoid system if one supplies something that slows the breakdown of anandamide, is among the more immediate ones still without an answer.

This article describes published research on the body’s own endocannabinoid system. It does not constitute medical advice and makes no claims about health effects of cannabidiol or any other product.