The Roles of Placebo and Nocebo in Medication Effects and Why Subjective Reports Are Unreliable

By Aly | First published January 11, 2020 | Last modified December 11, 2020

Abstract / TL;DR

The placebo effect is a phenomenon in which beneficial effects occur due to positive expectations. The nocebo effect, which is the opposite of the placebo effect, is when adverse effects occur due to negative expectations. The placebo and nocebo effects are powerful psychological phenomena that can cause real biological effects, in turn producing true therapeutic benefits and adverse effects. They accompany the real effects of medications, and while often underestimated, can be substantial in strength and importance. As an example, about 75% of the mood benefit of antidepressants is due to the placebo effect rather than the medications themselves. The placebo and nocebo effects can result in misattribution of effects to drugs. They can also amplify and worsen the severity of the true adverse effects of drugs. Aside from placebo and nocebo, faulty perceptions and coincidence can result in erroneous perceptions and misattributions of drug effects as well. Due to the problems with subjective observations, anecdotal reports and uncontrolled clinical studies are unreliable means of quantifying subjective drug effects and should be considered with caution. Only well-designed controlled clinical trials can truly demonstrate the effects of medications. The claimed benefits and adverse effects of medications encountered in anecdotes and subjective reports are often not really due to the medications themselves but are rather due to our minds and unrelated causes.

The Placebo Effect

The placebo effect is a clinical phenomenon that can be summarized as positive changes caused by one’s expectations. It’s something that’s often underestimated and discounted by people. But it’s a powerful phenomenon in many different symptom areas, especially subjective symptoms (e.g., mood, pain, sexual desire, etc.). Moreover, many people misunderstand the placebo effect. They believe that it means a person is just imagining things. Often the mere mention of the placebo effect is taken as an insult. But the placebo effect can cause real changes and benefits. And it’s a person’s friend, not one’s enemy or something that makes a fool of them.

The placebo effect has readily measurable biological effects (Mora et al., 2011; Marchant, 2016; Pinch, 2016; Resnick, 2017). It can literally change the brain. The changes are sufficiently significant that in many instances the placebo effect has therapeutic value and is being considered for use in medicine (Mora et al., 2011; Marchant, 2016; Resnick, 2017). Placebos have been found to be effective even when people are told that they’re getting one (Marchant, 2016; Resnick, 2017). And some placebos, for example larger sugar pills, have been shown to be more effective than others (Straus & von Ammon Cavanaugh, 1996). This may sound absurd, but it’s true. Researchers have theorized that lifelong conditioning with pills may result in subconscious inclinations towards benefits independently of one’s beliefs and expectations, potentially explaining the phenomenon (Resnick, 2017; Harvard, 2017). In any case, the placebo effect is stronger when people know that they’re getting a real drug (Mora et al., 2011). The placebo effect is likewise stronger for active placebos—placebos that produce noticeable side effects (Mora et al., 2011). Hence, really believing in the drug effect does appear to help strengthen the beneficial influence of placebo.

Placebo effects don’t occur for everything (Pinch, 2016; Resnick, 2017). As one researcher noted, they will not shrink a tumor or heal a broken bone. But when it comes to areas in which placebo effects do occur, like subjective symptoms, their influence can be substantial (Pinch, 2016). In a landmark study in the 1980s, post-operative patients experiencing pain were randomized to either morphine or an “overt dose of a substance described as a powerful painkiller” that was in fact saline solution (salt water) (Pinch, 2016). The patients given the saline solution reportedly experienced the same degree of pain relief as those given morphine (Pinch, 2016). In another famous study, conducted in 1950, women suffering from pregnancy-related nausea and vomiting were given a substance that they were told would cure their symptoms (Benson, 1997). Instead, they were given ipecac, a drug that induces nausea and vomiting. Despite this, their nausea and vomiting disappeared, and one patient’s stomach contractions even returned to normal. Thus, while the placebo effect doesn’t occur for all symptom areas, it can be powerful in those in which it does occur. As such, it would be a clear mistake to ignore the possibility of the placebo effect in scenarios in which it may plausibly occur.

For pain conditions and generalized anxiety disorder, the placebo effect has been estimated to be responsible for about 50% of the effect of medication therapy (Mora et al., 2011). In another well-known study on the placebo effect, it was shown that the decrease in pain symptoms caused by placebo could be blocked by naloxone, a μ-opioid receptor antagonist (Marchant, 2016). Subsequent studies showed that the placebo effect induces the release of endorphins, endogenous μ-opioid receptor agonists, and decreases the activity of brain areas involved in pain (Marchant, 2016). Other studies have shown placebos to affect other neurochemical pathways as well, for instance the endocannabinoid system and the dopamine system (Marchant, 2016). One study even found that a placebo for morphine activated the opioid system while a placebo for ketorolac—a weaker painkiller similar to ibuprofen—activated the endocannabinoid system (Marchant, 2016). Per one researcher, “placebos can modulate the same biochemical pathways that are modulated by drugs” (Marchant, 2016). And not all placebos have the same biological effects, with their actions changing with differences in expectations.

Based on meta-analyses, the placebo effect is estimated to be responsible for about 75% of the mood benefit of conventional antidepressants in depression (Mora et al., 2011). The differences in improvement between antidepressants and placebo are statistically significant but small, and without the placebo effect, antidepressants might not provide clinically important benefit in depression (Mora et al., 2011; Healy, 2007). Indeed, licensed antidepressants barely outperform placebos in head-to-head comparisons and often fail to do so even in large trials (Mora et al., 2011). It seems clear that a person who responds more strongly to placebo is one who is more fortunate, because their degree of genuine improvement will be greater. Although placebo is now being considered for use as a standalone therapy in medicine, technically it’s already being employed and to great benefit.

The placebo effect is significant when it comes to sexual desire as well. In clinical trials with bremelanotide (brand name Vyleesi), a recently approved drug for female sexual dysfunction, about 25% of women given bremelanotide experienced an increase in sexual desire compared to about 17% of women who were given placebo (Jaminson, 2019). In addition, 35% of women given bremelanotide experienced a decrease of distress about their libido, relative to 31% given placebo (Jamison, 2019). Findings were similar in the clinical trials for Intrinsa, the testosterone patch for sexual dysfunction in women (Intrinsa label). In these studies, 51% of women treated with the testosterone patch experienced an increase in sexual desire, while 35% given placebo had an improvement in sexual desire. In addition, 46% of women on Intrinsa had a significant increase in the frequency of sexually satisfying events, while 35% in the placebo group did. As such, the placebo effect can cause response rates for improvement in sexual desire and activity that rival in magnitude those of drugs with actual pharmacologically activating effects on sexual desire.

For reasons that should now be obvious, anecdotes and subjective reports are not reliable evidence of effect when it comes to claimed subjective benefits of drugs. In the transfeminine community, a major example of this is reports of mood benefits and improved sexual desire with progesterone. Indeed, my recent article on progestogens and sexual desire, a review of the available literature and evidence, found no indication of benefit and instead evidence for the opposite (Aly, 2020). The claimed benefits of progesterone and other progestogens on sexual desire in anecdotal/subjective reports, as well as the general widespread use of alternative medicine and non-evidence-based therapies like homeopathy, Reiki (energy healing), crystal healing, and most herbal supplements, can probably be traced in considerable part to the real benefits of the placebo effect. Indeed, it has been said that the history of medicine up to the 1850s was the history of the placebo effect (Benson, 1997).

The Nocebo Effect

The nocebo effect is the opposite of the placebo effect—negative changes caused by expectations. It is a phenomenon as well and is similarly powerful (Mora et al., 2011).

The following case study is an example of the nocebo effect (Love, 2019):

Help me, I took all my pills.” A 26-year-old man in the emergency room fell to the floor, and an empty prescription bottle dropped from his hands.

Mr. A, as he was referred to in his 2007 case study, was still conscious, but pale-faced, drowsy, and lethargic. He told doctors he had impulsively swallowed 29 capsules after a fight with his girlfriend—the pills were an experimental drug for depression he had been given in a clinical trial.

They were clearly having an effect: Mr. A’s blood pressure dropped abnormally low, and his heart rate was a rapid 110 beats per minute. After four hours on an IV, he showed little improvement.

Then, a doctor who oversaw the clinical trial arrived. He said that Mr. A was part of the placebo group—the patients who are selected at random, and without their knowledge, to receive a pill that did nothing at all. Any symptoms he was having weren’t coming from the pills, but from his mind.

Mr. A “expressed surprise then almost tearful relief.” Within 15 minutes, his blood pressure rose and his heart rate returned to normal. He was fine. “Mr. A’s hypotension appears to have occurred as a result of the placebo overdose,” the study wrote. Another name for this phenomenon: the nocebo effect.

The nocebo effect contributes substantially to drug adverse effects, often closely resembling the pattern of side effects of the actual drug (Mora et al., 2011). In addition, the effect plays a major role in non-adherence and discontinuation (Mora et al., 2011). In clinical studies, generally only a small proportion of reported side effects are actually due to the drug, and most reported symptoms are non-specific/unrelated adverse effects (Mora et al., 2011). We know this thanks to control groups (e.g., Table—note the lack of difference in incidence between groups for many of the side effects). A study showed that most of the adverse effects associated with statins in patients experiencing them were due to the nocebo effect (Wood et al., 2020; The Guardian, 2020; BBC, 2020). Like the placebo effect, the nocebo effect is biologically potent and has been demonstrated to change the brain (Mora et al., 2011).

Knowledge about possible adverse effects of a drug, for instance via information from the drug label, media, internet, and talking with other people, considerably increases the likelihood that a person will experience the relevant adverse effects (Mora et al., 2011). This has been suggested as a possible contributing factor to the adverse effects of 5α-reductase inhibitors like finasteride and dutasteride (Kuhl & Wiegratz, 2017; Maksym et al., 2019). 5α-Reductase inhibitors do block synthesis of certain neurosteroids, and this could indeed result in clinically significant adverse effects. But knowledge of such actions, and by extension the nocebo effect, likely serves to amplify such effects. Accordingly, one study showed that when men were told beforehand about possible sexual dysfunction with finasteride, the rate of reported sexual side effects tripled (44% vs. 14%). As another example, one not related to hormone therapy, it has been hypothesized that a nocebo effect with gluten intake may be fueling the popularity of the gluten-free diet (Resnick, 2017).

The nocebo effect is why we need to be cautious when interpreting reported adverse effects of drugs. Anecdotal reports are not reliable evidence of effect either with beneficial subjective effects or with adverse psychological effects. Nor are small uncontrolled clinical studies reliable evidence for such effects. Well-designed controlled trials, namely randomized controlled trials (RCTs), are needed to properly demonstrate cause and effect. As an example, we’ve often seen alarmingly high rates of depression with cyproterone acetate in small uncontrolled studies. But when we look at one of the few high-quality controlled studies of cyproterone acetate and mood—which in addition to having a control group used a reliable and validated measure of mood—such effects virtually disappear (Green et al., 2002; Green et al., 2004). The nocebo effect, besides in the case of cyproterone acetate and 5α-reductase inhibitors, is also of relevance to other drugs used in transfeminine hormone therapy, one of the most notable being spironolactone. A major nocebo phenomenon has likely developed against spironolactone in the transfeminine community in recent years due to the efforts of a single prominent individual. Quality clinical evidence for the claimed adverse effects of spironolactone is largely absent despite widespread use and research into the medication. In actuality, spironolactone has been well-tolerated even at high doses in clinical studies (e.g., Table).

Other Considerations

The placebo effect by definition does not involve faulty perceptions or misattribution. But these are important additional contributors to perceived effects of drugs as well (Mora et al., 2011). And they must be taken into account when interpreting claimed drug effects similarly. Oftentimes the change a person experiences and attributes to a drug isn’t due to the drug at all but is due to coincidental circumstances and hence is caused by something else entirely. Sometimes a person’s subjective observations or memory may be erroneous, and they may believe that something has changed in a given way when that may not actually be the case. Feminization is a good example of this, which occurs slowly and can be hard to accurately measure on a subjective level over the long period of time in which it occurs. Despite the limitations of anecdotal/subjective observations, people have a tendency to put much more confidence in their observations and attributions than is warranted. This is so even when they may know on some level that their claims are weak, perhaps a way of defending their ego or credibility. Unfortunately, the mind is imperfect and fallible, and the preceding problems are additional reasons why anecdotal/subjective reports are unreliable and should be interpreted cautiously.

Further Reading

  • Healy, D. (2007). The New Anecdotes. Ethical Human Psychology and Psychiatry, 9(3), 131–137. [DOI:10.1891/152315007782792687]
  • Mora, M. S., Nestoriuc, Y., & Rief, W. (2011). Lessons Learned from Placebo Groups in Antidepressant Trials. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1572), 1879–1888. [DOI:10.1098/rstb.2010.0394]
  • Benedetti, F. (2014). Placebo Effects: From the Neurobiological Paradigm to Translational Implications. Neuron, 84(3), 623–637. [DOI:10.1016/j.neuron.2014.10.023]
  • Marchant, J. (2016). Placebos: Honest Fakery. Nature, 535(7611), S14–S15. [DOI:10.1038/535S14a]
  • Pinch, B. (2016). More Than Just a Sugar Pill: Why the Placebo Effect is Real. Science in the News. Harvard Graduate School of the Arts and Sciences. [URL]
  • Resnick, B. (2017). The Weird Power of the Placebo Effect, Explained. Vox. [URL]
  • Kirsch, I. (2019). The Placebo Effect in the Treatment of Depression and Anxiety. Frontiers in Psychiatry, 10, 407. [DOI:10.3389/fpsyt.2019.00407]