Thoughts on placebo and nocebo effects.

Nocebo effects are hugely under-appreciated.

What’s the placebo effect and why is it important?

Everyone in the healthcare-industrial complex acknowledges and understands the placebo effect - the power of suggestion to positively impact illness and / symptoms.

It is why clinical trials of new drugs should include a placebo arm - to distinguish the “true” effect of the new drug from the benefits which accrue simply because of the expectation by the recipient that they will get better¹. Recipients allocated² to the placebo arm should receive a version of the product without the active ingredient, which is indistinguishable - both in terms of appearance and taste³ - from that received by the participants in the “active” group.

The idea is that the effect (and safety) of the product as observed in the trial is then compared to those seen in the placebo group to measure the “true” effect.

The “gold standard” design provides that nobody involved in the trial - patient, doctor, study managers, analysts etc - is meant to know what any particular patient received until the “blind is broken”, as this could bias the recording and analysis of data.

The main reason for this is that during a trial, a number of judgment calls are made by various parties involved, and knowledge of treatment allocation by individual subject could consciously or subconsciously affect this process, introducing bias.

Examples of this include (but are certainly not limited to):

• How diligently adverse events leading to withdrawal are followed-up to see what happens to the subjects experiencing them

• Whether adverse events are adjudged to be related to the study drug or not

• Whether symptoms are treated as suggesting further investigations (in the case of the covid injectables - PCR testing) are warranted

• Especially for symptom-driven conditions which lack objective measures (eg for depression, pain and so on), the recipient’s response to the treatment given could be affected quite profoundly by knowledge of which group they are in.

Notably, the Phase 3 trial of the Pfizer / BioNtech product was not actually double-blind - even though many people think it was. It was instead “observer-blinded”; people at the site drew up the allocated treatment from the vials provided (so they had to know what each patient received), the identity of which was then (supposedly) kept secret from the recipient⁴.

Josh Guetzkow (

Research Rebel

) has written about that - and the resultant pitfalls - here:

The Pfizer Vaccine Trial Was Not Double Blind

(Some trials of non-pharmaceutical interventions also include something akin to a placebo, even if it isn’t called one; a good example is arthroscopic knee surgery, when the surgeon “tidies” up some ropey-looking cartilage as a treatment for knee pain.

It turns out that in some trials of this treatment, putting the patient to sleep, making an incision into the knee, but not actually entering the joint - a so-called “sham” procedure - was just as effective as the standard procedure. This isn’t the only such example which suggests that symptomatic benefits from some operations owe more to patient expectations and the healing power of medical attention than to the biological or surgical rationale of the procedure.)

Placebo problems

Having established how important including a placebo is, and how crucial a double-blind design is, what circumstances can jeopardise these essential constituent elements of good design?

The most obvious one is that the subject could work out what treatment they are receiving. This can happen in a number of ways - even if the placebo has identical taste and / or appearance to the active product:

• They might get or not get an obvious side-effect

• They might get or not get a benefit which makes it obvious which group they are in.

Remember above I said that knowledge of treatment allocation was especially important for conditions which lack objective measures and rely on subjective assessments of well-being? This especially applies in the case of trials for anti-depressant drugs.

Some people have called this an “enhanced placebo effect”⁵. I mention this in this essay about SSRI inhibitors:

As good evidence for this effect, please see this elegant 2017 study, in which:

“patients were all given an antidepressant, but half told that it was a placebo and half told the truth, those who were been told that they had received an antidepressant showed twice the change in anxiety and depression scores compared to those who believed they had received the placebo”.

Another example of similar confounding of results by the placebo effect not working as it should was mentioned in this substack post:

Semaglutide (Wegovy) Is Not a Knee Pain Drug

The bottom line (in relation to the trial discussed) was that semaglutide reduced weight and knee pain. To be fair, the authors of the NEJM paper didn’t (in the abstract at least - I don’t have access to the full paper) claim that there is a beneficial effect on knee pain independent of the weight loss, though plenty of news sources hinted at it:

(I have little doubt that the manufacturers will try to use this to move their product into those with joint pain but who perhaps aren’t that obese.)

Regardless, John Mandrola is actually making a point here about placebo effect (independent of benefits mediated through weight loss), since pain is another one of those entirely subjective things which is extremely susceptible to placebo effect, and in this study it was patently obvious who was on the active drug, from their gastrointestinal symptoms.

He cites this meta-analysis which suggests that the placebo effect in trials for osteoarthritis (as measured in the placebo groups of a range of trials, comparing the outcomes to baseline) is particularly powerful.

Results:

We identified 198 trials with 193 placebo groups (16 364 patients) and 14 untreated control groups (1167 patients) that met our inclusion criteria. These included a range of therapies (non-pharmacological, pharmacological and surgical treatments). Placebo was effective at relieving pain (ES 0.51, 95% CI 0.46 to 0.55 for the placebo group and 0.03, 95% CI –0.13 to 0.18 for untreated control).

Placebo was also effective at improving function and stiffness. The pain-relieving effect increased when the active treatment effect (β = 0.38, p<0.001), baseline pain (0.006, p = 0.014) and sample size (0.001, p = 0.004) increased, and when placebo was given through injections/needles (0.144, p = 0.020).

Conclusion: Placebo is effective in the treatment of OA, especially for pain, stiffness and self-reported function. The size of this effect is influenced by the strength of the active treatment, the baseline disease severity, the route of delivery and the sample size of the study.

The effect size observed was 0.51. In other words, placebo halved (on average) whatever scores of pain the subjects were reporting. As Mandrola points out, that is greater than the observed effect sizes of many expensive routine drugs and other interventions, but notably those come with significant risks and side effects.

So what’s the nocebo effect all about?

• The placebo effect is the positive effect on symptoms from something given with the expectation that it will make the recipient better.

• The nocebo effect is the deleterious effect on symptoms and well-being caused by something done which creates an expectation that the person will either (1) become ill when they were previously well, or (2) will become more ill than they were before.

Despite much research on this phenomenon, the medical establishment very rarely discusses nocebo, whereas they discuss the placebo effect extensively.

The likely reason for this is that the context of such discussion is highly likely to be around things that medicine has done which has harmed people, and that is not generally something which doctors like to talk much about.

How powerful is this effect?

In other articles I have written (see here, here and here) I have cited these fascinating examples:

1. In 2007 it was recorded that a man’s blood pressure fell to dangerously low levels following an overdose, and could not be restored to normal by healthcare workers; however, upon learning that he had taken placebo, he rapidly recovered.

2. A man developed a full-blown cold when he was (incorrectly) told he had received “active broth” during an experiment about a century ago attempting to transmit the cold; upon being told he had actually had the “sterile broth” his cold immediately disappeared.

3. A group of children in Portugal in 2006 fell ill with a mystery illness including rashes, dizziness and difficulty breathing, which was eventually attributed to a reaction to a storyline in a teen soap opera.

4. In 2006 at a large hospital in the USA, an outbreak of whooping cough was suspected and many developed the symptoms expected⁶. However it was later confirmed that nobody had whooping cough, the tests were wrong, and health care workers had ordinary respiratory diseases like the common cold.

I have now done quite a bit of reading and thinking around nocebo effect. It seems clear to me that negative thoughts and expectations can manifest in symptoms which not just appear real to the subject experiencing them, but ARE real:

• The gentleman mentioned above didn’t just say he had a cold, he looked like he had a cold to others.

• The healthcare workers didn’t just suspect they had whooping cough - many of them actually had bad raking coughs which to them were much worse than a cough associated with a cold.

• The man who’d taken the overdose of placebo didn’t just feel faint, his blood pressure was objectively lower.

• The school children weren’t turned away from doctors and dismissed as being hysterical. Many had rashes and were investigated for serious illnesses.

At this point it is worth acknowledging that there is substantial overlap between the concepts of nocebo effect and psychogenic illness. Certainly the former, when it appears in groups of people, appears to be concordant with descriptions of events called “mass psychogenic illness”, or “epidemic hysteria” - of which there are many in recorded history.

Do have a read of this essay in which I discuss the 2020 dancing videos in the context of the outbreak of hysterical dancing in the middles ages in Europe:

Nocebo effects / psychogenic illness are extremely susceptible to group network amplification.

It will be obvious to those who read the above that these effects can easily be amplified and become self-reinforcing in a positive feedback loop.

Emotions are contagious. We see that at every social event we attend. An atmosphere of jollity and conviviality becomes self-reinforcing. The same is true of negative emotions, ie fear and anxiety; once you see people around you expressing fear, the natural response is to share in that emotion, as your psyche tells you that there is danger. Others then pick up on that, and the more people who are afraid, the stronger the signal and the more new people are drawn into the circle of fear, strengthening the signal so that yet more are drawn into it.

This network effect is an important feature of the outbreaks of mass psychogenic illness which were in evidence during the Portuguese soap-opera incident, as well as in the whooping cough pseudo-epidemic; the latter example illustrating that the effect is not limited by age, gender, education status or occupation.

So, what elements during the covid event might have been relevant in this regard?

1. The most obvious one is the rapid spread of emotive messaging via the internet, especially via social media applications.

It must be remembered that not only did governments spend $billions on promoting fear messaging, they suppressed any dissenting voices.

The internet allows the spread of such messaging by near-instantaneous amplification; this speed removes the possibility of any intervening critical commentary.

Before the digital age, it would have taken some time to create articles for newspapers, posters, pamphlets and so on, but today nothing needs to be printed. Moreover, the messaging during the covid event was crafted with the input of professionals adept at instilling the emotions intended by the instigators.

2. Testing.

Propaganda alone was probably insufficient to spread fear. In order to propagate a pandemic, people with ordinary respiratory illnesses or no symptoms at all had to be made to fear the “novel virus”. The tool by which this was achieved was the mass administration of over-sensitive and under-specific tests.

As “positive cases” were found, a number of perverse incentives created a positive feedback loop involving more testing (especially of “contacts”), more “cases” demanding more testing, more “cases” being found and so on.

3. The widespread use of the numbers from this machine:

The widespread and frequent use of portable pulse oximetry devices was widely promoted, and the numbers obtained from them treated (together with testing data) with a huge amount of reverence.

This was contrary to all previous medical teaching which has emphasised the treatment of the patient’s symptoms and overall wellbeing, with test results being an adjunct to treatment decisions.

This was a misguided and dangerous change in practice, for the following reasons (aside from the fact there was no novel virus causing a novel illness):

• The machines used in the community were - in many cases - cheap imports with questionable reliability and accuracy.

• Even the high-end machines give variable readings according to precise placement, skin tone, ambient light conditions and so on.

• A range of “normal” readings - especially in the elderly with run-of-the-mill respiratory illnesses has never been elucidated - so we have no idea what the significance of lower readings really is.

• During fear and anxiety peripheral blood supply will vasoconstrict, probably leading to lower readings.

• Peripheral oxygen saturation readings are in any case known to underread central oxygenation amd this must be even more so with accompanying fear and anxiety.

• A high temperature (a normal feature of regular seasonal respiratory illnesses) shifts the oxygen dissociation curve such that peripheral oxygen readings as measured by pulse oximetry may appear lower but oxygen delivery to tissues be perfectly satisfactory⁷.

Needless to say, all the ingredients necessary for another feedback loop are present in the above.

Imagine the poor individuals, terrified by the government propaganda, who then “test positive” perhaps from a regular seasonal illness, and start measuring their “sats” using one of these devices. They find it varies quite a lot depending on exact placement, and the excursions they observe are quite wide, the number frequently dipping into the low 90s which, they are told, is the danger area.

This then causes more anxiety - which lowers the readings - and causes them to perform more frequent testing, resulting in observing yet more low readings. Anxiety, and associated breathlessness result. The patient’s GP is called, an assessment is made over the phone, as GPs aren’t actually seeing patients face-to face.

Because of the positive test, breathlessness and “low sats” the patient ends up in hospital “just in case” and because the GP doesn’t want to be blamed if things deteriorate.

In hospital, because of the “positive test”, a number of ultimately harmful things could have ensued, which I am not going to expand upon here (this piece is long enough as it is and many others have written extensively on that topic), other than to state that whatever harms flowed from this event, they would have been chalked up to “covid”, yet were in fact iatrogenic in nature.

1. It is for reasons of safety assessment too. The rate of events occurring in the treated group can be compared against those in the placebo group; some adverse events are simply sporadic things that happen by chance.

2. Treatment allocations should be done randomly to create groups which are as similar as possible.

3. For oral medication.

4. Practicalities may have dictated the choice of design to some extent, but it was awfully convenient, wasn’t it?

5. To some this could be a slightly misleading term, as it is something which affects the active group.

6. Nearly 1,000 health care workers at the hospital in Lebanon, N.H., were given a preliminary test and furloughed from work until their results were in; 142 people, including Dr. Herndon, were told they appeared to have the disease; and thousands were given antibiotics and a vaccine for protection. Hospital beds were taken out of commission, including some in intensive care.

7. https://pmc.ncbi.nlm.nih.gov/articles/PMC7848791/