Endocannabinoid contributions to the perception of socially relevant, affective touch in humans
Department of Biomedical and Clinical Sciences, Center for Social and Affective Neuroscience, Linköping University, Linköping, Sweden
Department of Psychiatry, Cumming School of Medicine, University of Calgary, Calgary, AB Canada
Hotchkiss Brain Institute, University of Calgary, Calgary, AB Canada
Mathison Centre for Mental Health Research and Education, University of Calgary, Calgary, AB Canada
Department of Cell Biology and Anatomy, University of Calgary, Calgary, AB Canada
Abstract
Social relationships are central to well-being. A subgroup of afferent nerve fibers, C-tactile (CT) afferents, are primed to respond to affective, socially relevant touch and may mitigate the effects of stress. The endocannabinoid ligand anandamide (AEA) modulates both social reward and stress. We thus hypothesized that AEA levels would be associated with the perceived pleasantness of affective touch in humans. Across two studies, we explored perceptions of affective, socially relevant touch and general affective stimuli. In study 1, adult participants (N = 101) were recruited based on presence (CM+) or absence (CM−) of documented childhood maltreatment (N = 52 CM+; N = 49 CM−). In study 2, healthy individuals were randomized to receive an inhibitor of fatty acid amide hydrolase (FAAH; PF-04457845) to increase AEA levels (n = 16) or placebo (n = 29). Outcomes included self-report ratings of touch pleasantness and intensity, valence and arousal ratings of affective images, and plasma levels of endocannabinoids AEA and 2-AG, cortisol, and oxytocin. In study 1, higher AEA levels were associated with a reduced preference for affective, CT-optimal touch. In study 2, pharmacological elevation of AEA resulted in reduced preference for affective touch. These effects were specific to social processing, as AEA levels were not related to ratings of affective images. In contrast to our hypothesis, elevated AEA was associated with reduced pleasantness ratings of CT-optimal, affective touch. This provides novel, in-human data linking AEA to social processing, adding nuance to the rationale for its use as a potential novel therapeutic target in disordered in social processing.
Untitled section
Subject terms: Social neuroscience, Emotion
Article notes
Untitled section
Received 2024 Nov 2; Revised 2024 Dec 13; Accepted 2025 Jan 10; Issue date 2025 Apr.
Introduction
Touch is a vital component of social attachment in mammals [1, 2]. In humans, forms of gentle touch often occur during close affiliative interactions and carry positive affective weight [3]. This likely involves contributions from specialized afferent nerves called C-tactile (CT) fibers (Löken et al. [4], etc). CT afferents convey touch signals to key brain regions, such as posterior insula, which also respond to the mere observation of gentle, affective touch in others [5]. Both experienced and observed affective touch can similarly impact psychophysiological measures of emotion processing [6, 7]. Affective touch can also mitigate feelings of social exclusion [8], and may play a role in stress regulation. For example, heart rate and skin conductance, often used as indicators of stress arousal, are attenuated during touch from a familiar human partner [9–11]. Some studies have begun to explore potential neuromodulatory systems subserving touch processing, including opioids [12] and oxytocin [13]. However, to date, few studies have explored how endocannabinoid (eCB) function may contribute to touch processing in humans.
The eCB system is believed to play a key role in modulating both stress reactivity and social behavior [14–16]. This system is comprised of the endogenous ligands N-arachidonoylethanolamine (anadamide; AEA) and 2-arachidonoylglycerol (2-AG), which bind to the CB1 receptor, mostly expressed in brain, and the CB2 receptor, primarily expressed in the body. Of the eCB ligands, potentiation of AEA activity in particular has garnered significant interest as a potential novel therapeutic mechanism via inhibition of its main degradative enzyme fatty acid amide hydrolase (FAAH). Clinical trials exploring the impact of FAAH inhibition have been completed or are ongoing for a range of disorders characterized by dysregulation of stress and social processing, including social anxiety disorder [17], post-traumatic stress disorder (EudraCT 2020-001965-36), and autism spectrum disorder [18]. Evidence from healthy humans supports the link between AEA and stress reactivity, as elevated AEA can attenuate stress reactivity and promote regulation of negative emotion [19, 20]. However, there is a paucity of information on the role of AEA in social processing in humans, particularly in non-clinical populations.
Here, we aimed to determine whether the eCB system is involved in the perception of socially relevant, affective touch, as well as whether these effects are specific to social stimuli or also evident more broadly in affective processing. In the first study, we assessed how individual variation in the eCB system impacted the perception of CT-optimal affective touch in a heterogenous participant sample with or without documented histories of childhood maltreatment (CM) [21]. We then aimed to confirm our findings in a second study in which AEA levels were pharmacologically elevated with a FAAH inhibitor (PF-04457845; [20]) in healthy adults who completed the same affective touch and affective image tasks.
Methods
Study 1 methods
Overview
The study was approved by the Regional Ethics Review Board in Linköping, Sweden (Dnr 2015/256-31). Participants were recruited between March 2017 and July 2020 for a study assessing the relationship between documented childhood maltreatment and substance use disorder histories on affective processing, with the primary results published previously [21]. Extensive details regarding the population and other experimental procedures can be found in [21] and Supplemental Materials.
Participants
Demographic information can be found in Table 1. In brief, patients were recruited based on documented history of childhood maltreatment (CM). The N = 101 population consisted of n = 52 CM+ and n = 49 CM− matched on other demographic factors. In addition to the prospective CM assessment obtained via electronic medical records, all participants also completed the Childhood Trauma Questionnaire [22].
| Study 1 | ||||
|---|---|---|---|---|
| Childhood maltreatment (N = 52) | No childhood maltreatment (N = 49) | Total (N = 102) | p-value | |
| Sex: Female | 37 (63.8) | 25 (51.0) | 62 (60.8) | p = 0.10 |
| Age | 29.1 (3.6) | 27.9 (4.3) | 28.6 (4.0) | p = 0.06 |
| CPRS (Psychiatric symptoms) | ||||
| Depression scores (MADRS) | 5.9 (4.7) | 3.7 (3.7) | 4.9 (4.4) | p = 0.01 |
| Anxiety scores | 6.9 (4.4) | 5.1 (3.7) | 6.1 (4.2) | p = 0.03 |
| CPRS total scores | 12.3 (8.4) | 8.3 (7.1) | 10.5 (8.0) | p = 0.01 |
| DERS (Emotion regulation) | 38.5 (15.7) | 33.4 (13.9) | 36.2 (15.1) | p = 0.08 |
| AUDIT | 6.2 (4.9) | 5.1 (5.1) | 5.7 (5.0) | p = 0.27 |
| DUDIT | 2.0 (5.4) | 3.1 (6.4) | 2.5 (5.9) | p = 0.31 |
| CTQ (Childhood trauma) | ||||
| Physical abuse | 9.1 (4.5) | 5.9 (2.2) | 7.6 (4.0) | p < 0.001 |
| Sexual abuse | 9.6 (6.6) | 5.1 (0.5) | 7.5 (5.4) | p < 0.001 |
| Emotional abuse | 11.8 (5.6) | 8.2 (4.6) | 10.2 (5.5) | p < 0.001 |
| Physical neglect | 8.8 (4.0) | 6.9 (3.6) | 7.9 (3.9) | p = 0.014 |
| Emotional neglect | 12.5 (5.4) | 9.5 (5.0) | 11.1 (5.4) | p = 0.004 |
| Total score | 51.8 (20.8) | 35.6 (13.7) | 44.4 (19.6) | p < 0.001 |
Behavioral tasks
Behavioral tasks were completed in a fixed order, with the Observed Touch task completed first, and the Affective Image task completed after a 15 min break. This order was necessary due to the subsequent tasks, which included a stress task, to avoid carry-over effects of stress. The other tasks are reported elsewhere [21].
Observed touch task [6, 7]
Participants viewed 6 s videos depicting a left arm resting on a black background, being stroked by a hand at a rate of 3 cm/s (“slow” or CT-optimal) or 30 cm/s (“fast” or CT-non-optimal). Duration of videos was kept constant such that videos of 3 cm/s touch included two strokes, while videos of 30 cm/s included six strokes, producing videos of equal length (6 s). After each video, participants provided ratings of perceived pleasantness (−10 to +10) and intensity (−10 to +10) of the touch. The task took ~12 min to complete. While the study herein only employed Observed touch, neural and psychophysiological responses to Observed touch and Experienced touch are highly related [5, 6].
Affective image task [19, 23]
Positive, neutral, and negative images were selected from the International Affective Picture System based on normative ratings of valence and arousal. The task consisted of 12 images per stimulus type (positive, neutral, negative). Participants viewed a single image for 6 s and then rated valence from −4 to +4 and arousal from 0 to 9. The task took ~15 min to complete.
Biological samples
Blood samples were collected before and after each task. Both tasks reported here (Observed Touch Task and Affective Image Task), were completed prior to administration of other tasks (e.g., stress, fear conditioning), with primary results published elsewhere [21]. We calculated an AUC for the relevant biological variables (AEA, 2AG, cortisol). The eCBs AEA and 2-AG were extracted and analyzed using liquid chromatography tandem mass spectrometry (LC-MS/MS), as previously published and described in [19, 20]. Endocannabinoid values were log-transformed due to nonnormality of the distribution; these transformed values were used in all subsequent analyses. Plasma cortisol levels were obtained using the DetectX Cortisol Enzyme Immunoassay kit (Arbor Assays, Ann Arbor, MI, USA) according to manufacturer instructions.
Statistical analysis
Analyses of touch perception were carried out using repeated measures analysis of variance (RM-ANOVA) with Touch Type (CT-optimal/slow vs CT-non-optimal/fast) as a within subject factor for ratings of Intensity or Pleasantness, individually. Responses to affective images were analyzed using RM-ANOVA with Image Type (Positive, Neutral, Negative) as a within-subjects factor. Group (CM+, CM−) was entered as a between-subjects factor and AEA and 2-AG levels were entered separately as covariates. However, Group (CM+, CM−) was not a significant factor in any of the analyses and so was dropped from the model to achieve parsimony. Significant effects were followed up with Tukey’s post hoc comparisons.
To further explore the relationships between touch preference and biological measures, we created differences scores, as previously described [6]. Specifically, CT preference was calculated as [score for CT-optimal touch] – [score for non-optimal touch]. Analyses were carried out using SPSS and graphed using GraphPad Prism.
Study 2 methods
Overview
This double-blind, placebo-controlled study was approved by the Linköping Regional Ethics Review Board and the Swedish Medical Products Agency (EudraCT 2016-005013-47). The current data is an unpublished subset of a larger experimental medicine trial with primary outcomes reported previously [20]. Briefly, qualifying participants were randomized to receive the FAAH inhibitor PF-04457845 4 mg/day or placebo for 10 days. On days 9 and 10 of dosing, participants completed the tasks described herein, in addition to those described elsewhere [20]. Critically, the tasks described below (Observed touch task, Affective image task) always followed the non-stress control task (see Supplemental Materials for more information). Behavioral tasks were carried out as described in Study 1 “Methods”.
Participants
Participants over 18 years old were recruited through flyers and online advertisements from July 2017 to May 2018. Eligibility was confirmed at a screening session, followed by informed consent, and then medication distribution. Sessions were completed on consecutive days (days 9 and 10 of dosing) at 12.00 or later. Though 30 participants per treatment group were originally included, a pharmacy error resulted in only n = 16 receiving active drug as confirmed by plasma drug levels. Critically, those individuals are indistinguishable from the placebo-allocated group [20]. Demographics are in Table 1.
Drug intervention
PF-04457845 is a highly selective, orally available irreversible inhibitor of FAAH originally developed for pain and insomnia. PF- 04457845 tablets (4 mg) and visually indistinguishable placebo were provided by Pfizer (Groton, VT). Blinding and randomization were carried out by a contractor (Oriola, Stockholm, Sweden) insulated from the investigators. Labeled medication boxes were delivered to the Linköping University Hospital pharmacy. Adherence was confirmed by analysis of PF-04457845 in plasma at study completion. Participants and study personnel were blinded to drug administration until study completion.
Procedures
Blood samples were collected at T −15, T0, T+15, T+30, T+45 relative to administration of a stress or control task, with stress and control tasks completed on separate days (see Supplemental Materials for more details). Data presented here all came from the control task day, as the relevant tasks (Observed Touch Task and Affective Image Task, described below) were always completed post-control task. Biological samples were collected and analyzed in an identical manner as Study 1, with the addition of oxytocin analysis, as described [13].
Behavioral tasks
Behavioral tasks (Observed Touch Task and Affective Image Task) were identical to Study 1. The Observed Touch Task was always completed last.
Statistical analysis
All analyses were performed as described for Study 1, with the exception that treatment (FAAHi or PBO) was entered as a between-subjects factor. AEA was not entered as a co-variate in this analysis, as AEA levels were significantly different between groups.
Results
Study 1 results—individual variation in AEA levels
Demographics
The average age was 28.4 ± 4.0 years, and 56% (n = 57) was female. Demographic factors did not differ across groups.
Biological samples
There were no significant group differences in AEA (p = 0.10; Figs. 1C), 2-AG (p = 0.66), or cortisol (p = 0.54) between Groups (CM+, CM−). Oxytocin levels were not measured in Study 1.
Observed touch task
Affective picture task
Valence ratings
There was a main effect of Image Type (F(2,154) = 172, p < 0.001, partial η2 = 0.69; Fig. 2A), but no effect of AEA (p = 0.22) or Image Type × AEA interaction (p = 0.66). Post hoc follow-up tests revealed that Positive pictures were rated more positively than Neutral (t(1,99) = 28.6, p < 0.001) and Negative (t(1,99) = 37.8, p < 0.001) images, while Neutral images were rated more positive than Negative images (t(1,99) = 23.2, p < 0.001).
Study 2 results—pharmacological elevation of AEA
Demographics
The average age was 25 ± 7.7 years old and 45% of the sample was female. Groups (PBO, FAAHi) did not differ in age, gender, or any other demographic factor.
Biochemical measures
As expected, AEA levels were significantly higher in the FAAHi group compared to PBO (F(1,38) = 82.3, p < 0.001, partial η2 = 0.68; Fig. 3A). There were no group differences in peripheral levels of 2-AG (p = 0.14), Cortisol (p = 0.40), or Oxytocin (p = 0.76).
Observed touch task
Affective Image Task
Valence ratings
There was a main effect of stimulus type (F(2,86) = 613, p < 0.001, partial η2 = 0.93; Fig. 4A), but no effect of treatment (p = 0.26) or treatment × stimulus interaction (0.47). Post hoc tests revealed that Positive images were rated more positively than Neutral (t(44) = 22.9, p < 0.001) and Negative images (t(44) = 18.3, p < 0.001), and Negative images were rated more negatively than Neural (t(44) = 18.3, p < 0.001).
Discussion
The eCB system is a critical modulator of stress responses [15] and is also believed to play a role in balancing social approach and avoidance behaviors [16]. Accordingly, the eCB system is proposed as a novel therapeutic target in disorders characterized by impairments in social processing, including social anxiety [17] and autism spectrum disorder [18]. However, evidence linking the eCB system function to social behavior in humans is limited. Here, we explored the role of the eCB system, and AEA in particular, in the perception of socially relevant, CT-optimal touch. In a heterogenous clinical population, we find that greater AEA is related to less preference for affective, CT-optimal touch. We confirmed the direction of this relationship in a subsequent study in which healthy adults received a pharmacological intervention that elevated AEA via FAAH inhibition. Again, elevated AEA was associated with reduced preference for CT-optimal touch. Across studies, the effect of elevated AEA was specific to social processing, as there was no impact of AEA on processing of emotional images.
Our results are contrary to the idea that AEA may facilitate prosocial behavior, as we find that more AEA is related to less preference for socially relevant, CT-optimal touch. However, most preclinical literature has explored the effects of elevated AEA via FAAH inhibition in a “deficit state”—that is, in animal models of chronic stress or following induction of specific ASD-like traits [24–28]. The few studies that have administered FAAH inhibitors under naïve conditions corroborate our findings [29, 30]. Similarly, most human studies have investigated AEA function in clinical populations with dysregulated social processing [17, 18, 31], but data exploring AEA function in relation to social processing in non-clinical human populations are largely absent.
Interestingly, we did not find an effect of FAAH inhibition, and subsequent increases in AEA, on oxytocin levels. This is in contrast to preclinical findings showing that AEA may mediate the pro-social effects of oxytocin [32]. However, we have previously shown that context can significantly impact touch-related dynamic increases in oxytocin [13]. Thus, a more temporally controlled assessment of dynamic AEA and oxytocin changes, particular in relation to observed versus experienced touch, could provide more insight into this potential association in humans.
One possible interpretation of our unexpected results is that we hit a “ceiling effect” on the stress mitigating effects of AEA and touch, as both are proposed as potential stress buffers. Humans with elevated AEA via genetics [19] or pharmacology [20] are protected against stress-induced increases in negative affect. Pharmacological elevation of AEA specifically has been shown to attenuate subjective and autonomic stress reactivity [20]. Affective touch also attenuates psychophysiological stress reactivity [11, 12, 33]. Thus, AEA and affective touch may have similar stress buffering capabilities. As a result, in non-stress conditions, there may be a ceiling effect on the positive attributes of CT-optimal touch, as we find here. What remains to be determined is how this relationship changes in the context of acute or chronic stress, where AEA and touch may each have more pronounced affective impact.
This state-dependent effect on social behavior has been described in relation to the opioid system [34]. For instance, motivation for social touch is increased following stress exposure, and for those pretreated with an opioid agonist, subjective wanting of touch is further facilitated [12]. However, neither opioid stimulation nor blocked impacts touch perception or wanting in the absence of stress [35]. Interestingly, while negative states can drive increased wanting and hedonic responses to touch, appetitive states do not necessarily impact touch processing [33]. In contrast, other pharmacological interventions, including MDMA, can increase the perceived pleasantness of affective, CT-optimal touch in the absence of stress [36]. MDMA is also proposed as an intervention for psychiatric indications characterized by dysregulation of social processing, such as ASD [37]. Thus, it remains to be seen how these different interventions— MDMA and elevated AEA—could promote therapeutic outcomes in ASD populations, particularly in relation to social processing.
Our findings are limited by the cross-sectional and between-subjects design of our studies, precluding comparison in how dynamic changes in AEA may impact touch perception, particularly in stressful contexts. Moreover, the touch task employed herein is advantageous due to the controlled nature of the stimuli but may not be representative of naturalistic touch encounters [38]. Furthermore, other social processing modalities were not addressed, limiting the generalizability of our findings. Lastly, the sample size in Study 2 was small, requiring replication of these effects in larger cohorts. However, our results provide clear next steps in exploring the role of the eCB system, and AEA in particular, in social processing and adds much needed human data in this domain.
Supplementary information
Acknowledgements
We thank Åsa Axén and Gisela Öhnström for outstanding management of research participants. We would also like to thank Yunii Cho, Daniella Miro, Roosa Haataja, Valter Hammar, Evelina Larsson, Johan Lindé, and Ana Asratian for invaluable help with data collection.
Funding
This work was funded by the Swedish Research Council 2013–2024 to MH (2013–07434) and the Brain & Behavior Research Foundation NARSAD Young Investigator Grant 27094 to LMM. Open access funding provided by Linköping University.
Competing interests
MNH, MH are Associate Editors and LMM is the Social Media Editor at Neuropsychopharmacology. The other authors have nothing to declare.
Footnotes
Footnote Group
Supplementary information
The online version contains supplementary material available at 10.1038/s41386-025-02053-y.
References
Untitled section
References
- 1.Dunbar RIM. The social role of touch in humans and primates: behavioural function and neurobiological mechanisms. Neurosci Biobehav Rev. 2010;34:260–8.
- 2.Suvilehto JT, Cekaite A, Morrison I. The why, who and how of social touch. Nat Rev Psychol. 2023;2:606–21.
- 3.Croy I, Luong A, Triscoli C, Hofmann E, Olausson H, Sailer U. Interpersonal stroking touch is targeted to C tactile afferent activation. Behav Brain Res. 2016;297:37–40.
- 4.Löken LS, Wessberg J, Morrison I, McGlone F, Olausson H. Coding of pleasant touch by unmyelinated afferents in humans. Nat Neurosci. 2009;12:547–8.
- 5.Morrison I, Björnsdotter M, Olausson H. Vicarious responses to social touch in posterior insular cortex are tuned to pleasant caressing speeds. J Neurosci Off J Soc Neurosci. 2011;31:9554–62.
- 6.Mayo LM, Lindé J, Olausson H, Heilig M, Morrison I. Putting a good face on touch: facial expression reflects the affective valence of caress-like touch across modalities. Biol Psychol. 2018;137:83–90.
- 7.Ree A, Morrison I, Olausson H, Sailer U, Heilig M, Mayo LM. Using facial electromyography to assess facial muscle reactions to experienced and observed affective touch in humans. J Vis Exp. 2019;145. 10.3791/59228.
- 8.von Mohr M, Kirsch LP, Fotopoulou A. The soothing function of touch: affective touch reduces feelings of social exclusion. Sci Rep. 2017;7:13516.
- 9.Spapé MM, Harjunen V, Ravaja N. Effects of touch on emotional face processing: a study of event-related potentials, facial EMG and cardiac activity. Biol Psychol. 2017;124:1–10.
- 10.Triscoli C, Croy I, Olausson H, Sailer U. Touch between romantic partners: being stroked is more pleasant than stroking and decelerates heart rate. Physiol Behav. 2017;177:169–75.
- 11.Triscoli C, Croy I, Steudte-Schmiedgen S, Olausson H, Sailer U. Heart rate variability is enhanced by long-lasting pleasant touch at CT-optimized velocity. Biol Psychol. 2017;128:71–81.
- 12.Massaccesi C, Willeit M, Quednow BB, Nater UM, Lamm C, Müller D, et al. Opioid-blunted cortisol response to stress is associated with increased negative mood and wanting of social reward. Neuropsychopharmacol Off Publ Am Coll Neuropsychopharmacol. 2022;47:1798–807.
- 13.Handlin L, Novembre G, Lindholm H, Kämpe R, Paul E, Morrison I. Human endogenous oxytocin and its neural correlates show adaptive responses to social touch based on recent social context. eLife. 2023;12:e81197.
- 14.McLaughlin RJ, Hill MN, Gorzalka BB. A critical role for prefrontocortical endocannabinoid signaling in the regulation of stress and emotional behavior. Neurosci Biobehav Rev. 2014;42:116–31.
- 15.Morena M, Patel S, Bains JS, Hill MN. Neurobiological interactions between stress and the endocannabinoid system. Neuropsychopharmacology. 2016;41:80–102.
- 16.Wei D, Allsop S, Tye K, Piomelli D. Endocannabinoid signaling in the control of social behavior. Trends Neurosci. 2017;40:385–96.
- 17.Schmidt ME, Liebowitz MR, Stein MB, Grunfeld J, Van Hove I, Simmons WK, et al. The effects of inhibition of fatty acid amide hydrolase (FAAH) by JNJ-42165279 in social anxiety disorder: a double-blind, randomized, placebo-controlled proof-of-concept study. Neuropsychopharmacology. 2021;46:1004–10.
- 18.Klein ME, Bangerter A, Halter RJ, Cooper K, Aguilar Z, Canuso CM, et al. Efficacy and safety of JNJ-42165279, a fatty acid amide hydrolase inhibitor, in adolescents and adults with autism spectrum disorder: a randomized, phase 2, placebo-controlled study. Neuropsychopharmacology. 2024;50:480–7.
- 19.Mayo LM, Asratian A, Lindé J, Holm L, Nätt D, Augier G, et al. Protective effects of elevated anandamide on stress and fear-related behaviors: translational evidence from humans and mice. Mol Psychiatry. 2020;25:993–1005.
- 20.Mayo LM, Asratian A, Lindé J, Morena M, Haataja R, Hammar V, et al. Elevated anandamide, enhanced recall of fear extinction, and attenuated stress responses following inhibition of fatty acid amide hydrolase: a randomized, controlled experimental medicine trial. Biol Psychiatry. 2020;87:538–47.
- 21.Perini I, Mayo LM, Capusan AJ, Paul ER, Yngve A, Kampe R, et al. Resilience to substance use disorder following childhood maltreatment: association with peripheral biomarkers of endocannabinoid function and neural indices of emotion regulation. Mol Psychiatry. 2023;28:2563–71.
- 22.Gerdner A, Allgulander C. Psychometric properties of the Swedish version of the childhood trauma questionnaire-short form (CTQ-SF). Nord J Psychiatry. 2009;63:160–70.
- 23.Lang PJ, Greenwald MK, Bradley MM, Hamm AO. Looking at pictures: affective, facial, visceral, and behavioral reactions. Psychophysiology. 1993;30:261–73.
- 24.Alteba S, Mizrachi Zer-Aviv T, Tenenhaus A, Ben David G, Adelman J, Hillard CJ, et al. Antidepressant-like effects of URB597 and JZL184 in male and female rats exposed to early life stress. Eur Neuropsychopharmacol. 2020;39:70–86.
- 25.Wei D, Dinh D, Lee D, Li D, Anguren A, Moreno-Sanz G, et al. Enhancement of anandamide-mediated endocannabinoid signaling corrects autism-related social impairment. Cannabis Cannabinoid Res. 2016;1:81–9.
- 26.Qin M, Zeidler Z, Moulton K, Krych L, Xia Z, Smith CB. Endocannabinoid-mediated improvement on a test of aversive memory in a mouse model of fragile X syndrome. Behav Brain Res. 2015;291:164–71. Available from: https://pubmed.ncbi.nlm.nih.gov/25979787/
- 27.Kerr DM, Gilmartin A, Roche M. Pharmacological inhibition of fatty acid amide hydrolase attenuates social behavioural deficits in male rats prenatally exposed to valproic acid. Pharmacol Res. 2016;113:228–35.
- 28.Doenni VM, Gray JM, Song CM, Patel S, Hill MN, Pittman QJ. Deficient adolescent social behavior following early-life inflammation is ameliorated by augmentation of anandamide signaling. Brain Behav Immun. 2016;58:237–47.
- 29.Ribeiro FT, de Serro-Azul MIS, Lorena FB, do Nascimento BPP, Arnold AJT, Barbosa GHL, et al. Increased endocannabinoid signaling reduces social motivation in intact rats and does not affect animals submitted to early-life seizures. Front Behav Neurosci. 2020;14:560423.
- 30.Seillier A, Advani T, Cassano T, Hensler JG, Giuffrida A. Inhibition of fatty-acid amide hydrolase and CB1 receptor antagonism differentially affect behavioural responses in normal and PCP-treated rats. Int J Neuropsychopharmacol. 2010;13:373–86.
- 31.Karhson DS, Krasinska KM, Dallaire JA, Libove RA, Phillips JM, Chien AS, et al. Plasma anandamide concentrations are lower in children with autism spectrum disorder. Mol Autism. 2018;9:18.
- 32.Wei D, Lee D, Cox CD, Karsten CA, Peñagarikano O, Geschwind DH, et al. Endocannabinoid signaling mediates oxytocin-driven social reward. Proc Natl Acad Sci USA. 2015;112:14084–9.
- 33.Massaccesi C, Korb S, Skoluda N, Nater UM, Silani G. Effects of appetitive and aversive motivational states on wanting and liking of interpersonal touch. Neuroscience. 2021;464:12–25.
- 34.Loseth GE, Ellingsen DM, Leknes S. State-dependent μ-opioid modulation of social motivation. Front Behav Neurosci. 2014;8:430.
- 35.Løseth GE, Eikemo M, Leknes S. Effects of opioid receptor stimulation and blockade on touch pleasantness: a double-blind randomised trial. Soc Cogn Affect Neurosci. 2019;14:411–22.
- 36.Bershad AK, Mayo LM, Van Hedger K, McGlone F, Walker SC, de Wit H. Effects of MDMA on attention to positive social cues and pleasantness of affective touch. Neuropsychopharmacology. 2019;44:1698–705.
- 37.Danforth AL, Grob CS, Struble C, Feduccia AA, Walker N, Jerome L, et al. Reduction in social anxiety after MDMA-assisted psychotherapy with autistic adults: a randomized, double-blind, placebo-controlled pilot study. Psychopharmacology. 2018;235:3137–48.
- 38.Maallo AMS, Duvernoy B, Olausson H, McIntyre S. Naturalistic stimuli in touch research. Curr Opin Neurobiol. 2022;75:102570.