Dissociation between individual differences in self-reported pain intensity and underlying fMRI brain activation

Abstract

Pain is an individual experience. Previous studies have highlighted changes in brain activation and morphology associated with within- and interindividual pain perception. In this study we sought to characterize brain mechanisms associated with between-individual differences in pain in a sample of healthy adolescent and adult participants (N = 101). Here we show that pain ratings varied widely across individuals and that individuals reported changes in pain evoked by small differences in stimulus intensity in a manner congruent with their pain sensitivity, further supporting the utility of subjective reporting as a measure of the true individual experience. Furthermore, brain activation related to interindividual differences in pain was not detected, despite clear sensitivity of the Blood Oxygenation Level-Dependent (BOLD) signal to small differences in noxious stimulus intensities within individuals. These findings suggest fMRI may not be a useful objective measure to infer reported pain intensity.

Fig. 1. Individual differences in the subjective experience of pain are further supported by differences between pain sensitivity classes in their ability to discriminate small differences in stimulus temperature via subjective reports.

Compared to participants from the High (red bars) and Moderate Pain Sensitivity classes (yellow bars), participants from the Low Pain Sensitivity class (green bars) needed a significantly greater increase in temperature from 43 °C to report changes in their perceived pain intensity and unpleasantness (p values: low – moderate: p_intensity = 0.017, p_{unpleasantness} = 0.003; low – high: p_intensity = 0.009, p_{unpleasantness} = 0.0005), as shown by post-hoc two-sided Dunn’s tests (A). Conversely, the highly sensitive class required a significantly larger decrease in temperature from 49 °C to report a change in perceived unpleasantness in relation to low or moderate sensitivity classes (p values: low – high: p_{unpleasantness} = 0.001; moderate – high: p_{unpleasantness} = 0.003), as shown by post-hoc two-sided Dunn’s tests (B). The smallest temperature changes to achieve discrimination in sensation from the sensation of a reference temperature, i.e., 43-ascending discrimination thresholds and 49-descending discrimination thresholds, are represented on the y axis. Error bars represent standard error of the mean. * represents p values < 0.05; ** represents p values < 0.01. Source data are provided as an xlsx Source Data file. n = 101 participants.

Fig. 2. Univariate fMRI analysis reveals widespread brain activation associated with high intensity heat stimulation, but no relationship with perceived pain intensity.

A Individual ratings of pain intensity in response to high intensity heat stimulation ranged from 0.07 to 10, with most people providing averaged ratings of pain intensity below 5. Mean and standard error represented by the bar plot. n = 101 participants. B Effect of high intensity heat stimulation on brain activation. Areas of increased activation included cerebellum (Cereb), thalamus (Thal), putamen (Put), primary somatosensory cortex (SI), secondary somatosensory cortex (SII), insula (Ins), anterior cingulate cortex (ACC), and dorsolateral prefrontal cortex (DLPFC). Areas of decreased activation included the amygdala (Amy) and hippocampus (Hippo), as well as posterior cingulate cortex (PCC) and precuneus (Prec). C There was no relationship between perceived pain intensity and brain activation associated with high intensity heat stimulus. Source data are provided as an xlsx Source Data file and activation maps are available on Github.

Fig. 3. Univariate analysis revealed widespread brain activation associated with high intensity cold stimulation (0.5 °C), but no relationship with perceived pain intensity.

A Ratings of pain intensity in response to high intensity cold stimuli ranged from 0 to 6.82. Mean and standard error represented by the bar plot. n = 73 participants. B Effect of high intensity cold stimulation on brain activation. Areas of increased activation included the putamen (Put), caudate nucleus (Cau), secondary somatosensory cortex (SII), insula (Ins), anterior cingulate cortex (ACC), and dorsolateral prefrontal cortex (DLPFC). Areas of decreased activation included bilateral amygdala and hippocampus (Amy/Hippo), primary somatosensory cortex (SI), posterior cingulate cortex (PCC) and precuneus (Prec). C There was no relationship between perceived pain intensity and brain activation associated with high intensity cold stimulation. Source data are provided as an xlsx Source Data file and activation maps are available on Github.

Fig. 4. Univariate analysis revealed widespread brain activation associated with high intensity auditory stimulation (90 dB) and related to perceived intensity.

A Ratings of perceived auditory intensity in response to high intensity auditory stimuli ranged from 0 to 7.94. Mean and standard error represented by the bar plot. n = 97 participants. B Effect of high intensity auditory stimulation (90 dB) on brain activation. Areas of increased activation included putamen (Put), caudate nucleus (Cau), secondary somatosensory cortex (SII), primary auditory cortex (AI), insula (Ins), anterior cingulate cortex (ACC), and dorsolateral prefrontal cortex (DLPFC). Areas of decreased activation included bilateral amygdala and hippocampus (Amy/Hippo), primary somatosensory cortex (SI), posterior cingulate cortex (PCC) and precuneus (Prec). C High perceived intensity was associated with a greater increase in brain activation associated with high intensity stimulation in areas such as AI, SII, and insula and a greater decrease in brain activation in PCC and precuneus. Source data are provided as an xlsx Source Data file and activation maps are available on Github.

Fig. 5. Effect of graded increases in intensity of heat stimulation on brain activation.

A Average ratings of pain intensity associated with heat stimulation significantly differed between high and low intensity stimulation, as shown by a two-sided paired sample t-test (p = 2.2 e⁻¹⁶). n = 101 participants. Increased brain activation in response to high (B) and low (C) intensity heat stimulation and differences between the two intensities of stimuli (D) are observed in areas such as the putamen (Put), the caudate nucleus (Cau), the thalamus (Thal), the primary somatosensory cortex (SI), the secondary somatosensory cortex (SII), the insula (Ins), dorsolateral prefrontal cortex (DLPFC), and the anterior cingulate cortex (ACC). Decreased activation in response to the same stimuli is especially present in the precuneous (Prec) and the posterior cingulate cortex (PCC). Error bars represent standard error of the mean. *** represents p values < 0.001. Source data are provided as an xlsx Source Data file and activation maps are available on Github.

Fig. 6. NPS expression significantly differs between low (47 °C) and high (48 °C) intensity heat stimuli, as shown by a two-sided paired sample t-test (p = 1.021 e⁻⁰⁸).

The average NPS expression, represented by the bars, is higher in high intensity heat stimulation compared to low intensity heat stimulation. In addition, the violin plots, which represent the distribution of individual NPS expression in each intensity, show a greater density of individuals with higher NPS expression in the high intensity heat stimulation than in the low intensity heat stimulation. Finally, the error bars represent one standard error of the mean. *** represents p values < 0.001. n = 101 participants. Source data are provided as an xlsx Source Data file.