This is the thirteenth installment of our look at the increasingly high placebo response that is plaguing clinical trials in analgesia and psychiatry. Read the rest of the posts in the series here.
Until now, our Placebo Problem series has focused exclusively on the placebo response in adults. Today we turn to another population: kids. Not surprisingly, due to ethical issues, very few studies have examined placebo effects in pediatric populations. One 2013 literature review calculated that of the roughly 2000 PubMed citations that investigated the placebo effect, only about 2.5 percent discussed it with respect to children and adolescents. Nevertheless, pediatric placebo responses have been observed in a number of conditions, including asthma, atopic dermatitis, autism, depression, epilepsy, hypertension, and migraine.
On the whole, the randomized controlled trial (RCT) literature indicates that placebo response rates are higher in children and adolescents than adults, but that drug responses do not differ by age. Younger children seem to exhibit greater placebo responses than older kids. For example, in a meta-analysis of pediatric major depressive disorder, kids less than 12 years old had a placebo response rate of 54 percent, while those older than 12 responded to placebo 45 percent of the time. For comparison, similar studies of adults suggest the placebo response rate is roughly 34 percent. This negative correlation between placebo response and age is also present in other illnesses, including epilepsy, migraine, attention deficit hyperactivity disorder (ADHD), anxiety, obsessive compulsive disorder (OCD), and painful functional gastrointestinal disorders.
Taken to its logical conclusion, if placebo responses are greater in kids than adults, yet drug responses are similar, then extending information about drug efficacy gained from adult studies to pediatric populations may be problematic. Specifically, it may lead to an overestimation of drug efficacy in children.
However, the one study that has actually experimentally tested the relationship between placebo response magnitude and age, by using an experimental model of placebo analgesia in both children and adults, reported no difference in placebo response between the two groups. In the study, participants received two identical creams on their forearm. They were informed that one cream would provide pain relief and told the other was a control. During the conditioning phase, the painful stimulation at the site of the “analgesic” was lower than the stimulation at the “control” site. The next day, identical stimulus intensities were used to test for placebo analgesic effects. The researchers found no difference in the magnitude of placebo analgesia experienced by adults and children, a result at odds with the rest of the literature. Clearly, more research is needed to definitively determine the true association between placebo response and age.
As we discussed in our third post in this series, the two main psychological mechanisms that are responsible for placebo response are expectancy and conditioning. It has been hypothesized that these same mechanisms also play a role in pediatric populations. The few studies that have directly examined the role of expectancy in placebo response in kids have found that it does indeed play a role.
For example, in one study, 49 children between the ages of six and nine were randomly assigned to an “analgesia expectation” group, who was told that an applied placebo lotion could reduce pain, and a “control expectation” group, who was told the lotion is necessary to measure pain but wouldn’t affect pain levels. Both groups then received a heat pain paradigm on their forearms. The kids who were told the lotion could relieve pain experienced increased heat pain threshold and tolerance compared to those whose expectancy was not manipulated.
Case studies and observational studies of infants and children suggest, albeit indirectly, that, just as in adults, conditioning also plays a role in pediatric placebo responses. For example, infants who have received repeated heel lancing without anesthetic for blood glucose monitoring showed higher pain levels during subsequent procedures than infants receiving the procedure for the first time, indicating that prior experiences can influence pain levels.
Interestingly, the experiment described above that found no difference in the magnitude of placebo responses between kids and adults did find that experience during the conditioning procedure had a larger effect on the magnitude of the placebo response in children than adults, suggesting that conditioning may play a greater role in the pediatric placebo response.
A third possible mechanism that may be at work in pediatric placebo responses is a phenomenon known as “placebo by proxy” where parent beliefs can affect outcomes. Numerous studies have shown that parent beliefs and stress levels can exert measurable effects in their children. Principally, placebo by proxy can be explained in two ways. The first is characterized by a shift in parental perception. Parents expect children to get better and therefore perceive that they are getting better. This is especially relevant in trials that use parent-reported outcomes. In the second case, a behavioral change in the parent, such as acting less worried after treatment is administered, changes the behavior of the child, who then experiences less pain.
Next week, in our penultimate post in the Placebo Problem series, we’ll take a brief look at the history the placebo and how it’s been used through the ages.