Friends in pain: pain tolerance in a social network

In this issue of the Scandinavian Journal of Pain, Engebretsen and colleagues [1] report on the association between friendship and pain tolerance. This work is notable in a number of respects. First, it adds to a small but rapidly growing literature on social factors and pain [2], [3], [4], [5]. As one of the three pillars of the biopsychosocial model of pain [6], social modulation of pain has received far less empirical attention than biology and psychology, despite the known importance of social factors as risk factors for chronic pain development and in predicting outcomes. Second, the current effort represents the first in-depth social network analysis of pain, an intriguing methodology that has been successfully applied to traits including loneliness, obesity, smoking, and alcohol consumption [7]. Third, it features an extremely large sample size of phenotyped individuals – 997 adolescents living in Tromsø, Norway – with a participation rate of 93%. And finally, its methods section features some of the most complex algebraic equations ever to appear in a journal with the word “pain” in its title.

The authors invited all first-year students (ages 15–17) at all eight upper secondary schools in Tromsø to be tested for pain thresholds and tolerance as part of the Tromsø Study: Fit Futures I project. The adolescents were tested for heat pain threshold and tolerance, pressure pain threshold and tolerance (on the fingernail and trapezius), and pain tolerance on the 3°C cold-pressor test, in that order. In addition, all participants were asked to identify up to five of their friends, defined as the five individuals they had spent the most time with in the preceding week. This pain and friendship information was combined with demographic variables and the lifestyle factors smoking and physical activity.

Although results varied somewhat according to the pain modality and measure, they can be summed up as follows. The pain tolerance of ones friends was positively correlated with ones own pain tolerance, or, in other words, there was a tendency for friendships among adolescents with similar pain tolerance. Strikingly, when the data were analyzed by sex this was found to be true in boys but not girls; in fact, only same-sex male friendships were associated with pain. Smoking and physical activity affected pain in the predicted directions, but controlling for these factors did not abolish the friendship effect. Popularity (defined as the centrality of an individual within the network of friends) had complex effects on pain, and thus the current study generally supported but did not confirm previous findings that pain tolerance was higher in individuals with a larger social network [8].

An obvious potential explanation of the effect being specific to same-sex male friends was that since the boys were tested in sequence, they may have been competing and highly motivated to be able to brag to their peers (friends or not) about being able to endure the test to its (105 s) cut off. As a male, even many years past high school, I can confirm this “peer pressure”, having once been tested for pain in Tromsø myself in front of people I know and would see again. More relevant than my personal anecdote might be a recent report that males observing their male friends in pain increased the friend’s pain threshold and tolerance [9]. The current authors confirmed that there was indeed a significant effect of testing sequence; controlling for this reduced but did not abolish the effect.

As in other studies of social network effects, the finding that pain tolerance is “assortative” [10] can be attributed to one or both of two explanations: homophily or social transmission. An explanation based on homophily – birds of a feather flocking together – would suggest that individuals select friends based on similarity in pain tolerance. An explanation based on social transmission would suggest that friends could influence each other’s pain tolerance. In this particular case, neither explanation is obvious. Unlike obesity or smoking, the pain tolerance of a friend or potential friend is not directly observable, nor is it clear what exactly would be transmitted from one person to another.

The authors speculate that individuals might select friends with similar lifestyles, and one or more lifestyle characteristic might in turn be associated with pain tolerance. As the most obvious candidates here – smoking, physical activity and parents’ socioeconomic status – have been ruled out, it is not clear what that characteristic might be, although the authors speculate that it likely relates to whatever makes boys be perceived as “tough” or “macho”. If, on the other hand, future research determines that the effect is somehow due to social transmission, this could explain a large fraction of the long-appreciated and robust variability in pain sensitivity [11], [12]. Pain can be socially transmitted in rodents [13], [14] and humans [15], although these phenomena likely only occur in real time. It is well known that the lion’s share of the variability in biological traits in general [16] and pain tolerance in particular [17] is explained by genetic and unshared environment; shared environment plays a negligible role. In other words, variation is due to friends, not parents.

Distinguishing between homophily and social transmission as explanations of the observed results will likely require longitudinal studies. In a fascinating recent study performed on adolescents on a 3-week hiking expedition in the Arctic, no evidence of either pain homophily or contagion was found, but males (but not females) in pain (largely due to injuries obtained during the expedition) endured reduced popularity [18]. Studies like this and Engebretsen et al. [1] suggest we have just scratched the surface of an extremely interesting set of questions, the answers to which will contribute importantly to our emerging appreciation of social influences on pain.