Figures and data in Human endogenous oxytocin and its neural correlates show adaptive responses to social touch based on recent social context

Figures
Tables
Additional files

4 figures, 1 table and 8 additional files

Figures

Figure 1

Download asset Open asset

Functional magnetic resonance imaging (fMRI) experiment setup.

(A) (1) Indwelling catheter in female participant’s left arm (arrow); (2) participant’s male partner or unfamiliar stranger caressed the participant, following audio prompts; (3) serial blood samples were collected from the catheter. (B) Structure of fMRI experimental session with a serial sampling of plasma oxytocin and cortisol. Rectangle depicts the time course of the experiment, with approximate elapsed minutes shown above (yellow clock symbol). Two functional runs with partner and stranger touch, in counterbalanced order, were separated by ~27 min. Three baseline oxytocin (OT) samples (1, 5, 9) and three serial samples were collected for each run (2-4, 6-8). Clock symbol indicates the time in minutes. Blue dots in the vial symbols indicate oxytocin samples, the green dots indicate cortisol samples. The first functional run was preceded by the acquisition of an anatomical (T1-weighted) image, while between functional runs additional anatomical and functional scans were acquired: T2-weighted anatomical image, diffusion-weighted imaged, and resting-state. See also *Materials and methods* below.

Figure 2

Download asset Open asset

Endogenous hormone (OT and cortisol) changes and covariant brain responses.

(A) Familiarity, order, and sample timepoint influenced plasma OT levels: familiarity, order, and sample timepoint interacted (F_{(3, 183.180)}=3.034, p=0.031) as did familiarity and order (F_{(1, 183.169)}=11.216, p=0.001), with increased OT when the partner was the interactant in the first encounter. The contribution of the sample timepoint lay in a dip and recovery during stranger touch only when preceded by partner touch (p=0.027). (B) OT-BOLD covariance in the hypothalamus and dorsal raphe was driven by a greater decrease for stranger touch during the initial encounter. (C) Familiarity and % OT change interacted in parietotemporal BOLD clusters along right SMG/AG, TP, and mPFC extending to ACC (3dLME model, p<0.002), reflecting more positive relationships between BOLD signal and OT change for partner (the higher the BOLD, the greater the degree of OT change). (D) Familiarity and order interacted in plasma cortisol levels (F_{(1, 130)}=54.89, p<0.001), with stranger touch eliciting a greater cortisol increase compared to partner touch, reflected in a main effect of familiarity (F_{(1, 130)}=15.67, p<0.001). There was also a main effect of sample timepoint (F_{(2, 130)}=3.16, p=0.045), with levels generally declining over the session. (E) BOLD signal change in regions including mPFC/ACC covaried as a function of cortisol levels, with partner >stranger (p<0.002), and a subset of mPFC voxels covarying with both OT (partner >stranger, second encounter) and cortisol (partner >stranger, initial encounter). (F) Mean values for OT and cortisol showed an interaction with familiarity and order factors over the session (F(_{1, 178.355)}=10.565, p=0.001), with higher OT but lower cortisol in the partner first condition as compared to stranger first. OT = oxytocin, BOLD = blood-oxygen-level-dependent, f(ΔOT)=as a function of the change in OT, 3dLME = 3-dimensional linear mixed effects, SMG/AG = supramarginal gyrus/angular gyrus, TP = temporal pole, mPFC = medial prefrontal cortex, ACC = anterior cingulate cortex. All maps thresholded at p<0.002, corrected.

Figure 3

Download asset Open asset

Scatterplots showing the relationship between blood-oxygen-level-dependent (BOLD) signal and change in oxytocin (OT) during the first encounter in the hypothalamus and Raphe nuclei.

The linear mixed-effects modeling weighted the BOLD signal with an OT covariate (not visualized in the scatterplot); These showed that all standardized residuals fell within ±3 standard deviations from the trendline.

Figure 4

Download asset Open asset

Hormone-independent BOLD responses, BOLD covariance with temporal OT pattern, and BOLD covariance with pleasantness ratings.

(A) The temporal pattern of OT-BOLD changes in bilateral precuneus preceded sampling by 2.5 min, with retrosplenial cortex also emerging at 2 min, showing greater OT-BOLD covariance for stranger touch during the initial encounter. (B) ITG/TP was sensitive to differences in touch pleasantness ratings (red) for partner and stranger (mean pleasantness partner >stranger, p<0.001), with BOLD increasing with partner vs stranger pleasantness differences (Δpleas) during stranger touch in the second encounter (scatterplot). ITG/TP clusters also showed an interaction between familiarity and ΔOT (yellow), and between partner/stranger presentation order and ΔOT (green) BOLD here was greater in individuals with smaller OT change during the stranger-second condition. (C) BOLD changes in somatosensory and insular and PO cortices, as well as bilateral amygdalae, across all touch conditions, independently of familiarity of the person delivering touch, order, and OT levels (all ps <0.002). (D) Beta values reflecting the BOLD signal change in the left amygdala sensitive to partner-stranger differences (main effect of familiarity, F_(1,16) = 5.8, P=0.02), greater for the stranger in the first encounter. BOLD = blood-oxygen-level-dependent, PO = posterior operculum, OT = oxytocin, ITG = inferior temporal gyrus, TP = temporal pole, mPFC = medial prefrontal cortex, f(Δpleas)=as a function of the change in pleasantness ratings. All maps thresholded at p<0.002, corrected.

Tables

Table 1

Linear mixed-effects modeling with factors familiarity (partner, stranger), order (first or second encounter), with peak OT changes as a covariate.

All contrasts are thresholded at p<0.002, cluster-size thresholded at α = 0.05 FWE for n=27 complete functional datasets. For each cluster under each contrast heading, size in voxels, location, maximum F score, and MNI coordinates (x, y, z) are given.

Main effect: Familiarity >
Cluster (size)	Peaks Locations	F (x, y, z)
#1 (362)	Left Middle Frontal Gyrus	25.45 (-44, 52, 4)
		23.25 (-44, 25, 37)
		20.78 (-32, 52, 28)
		18.87 (-38, 61, 1)
		18.07 (-50, 43, 19)
		12.07 (-44, 13, 40)
	Left Inferior Frontal Gyrus	24.79 (-56, 37, 10)
		20.27 (-59, 25, 22)
#2 (234)	Right Middle Frontal Gyrus	33.33 (43, 46, 28)
		24.87 (34, 43, 40)
		18.90 (46 25, 37)
		16.93 (31, 28, 55)
	Right Superior Frontal Gyrus	29.72 (22, 43, 37)
		16.03 (25, 58, 31)
		13.16 (19, 28, 40)
#3 (128)	Left Middle Occipital Gyrus	28.69 (−17,–107, 4)
		18.25 (−26,–98, –5)
#4 (107)	Right Middle Frontal Gyrus	21.16 (40, 46, 13)
		18.20 (40, 58, 13)
		14.56 (25, 61, 28)
	Right Superior Frontal Gyrus	17.77 (31, 58, 19)
#5 (87)	Left Cerebellum	21.42 (−41,–65, –38)
		19.14 (−29,–62, –35)
#6 (73)	Right Angular Gyrus	19.18 (43, -65, 52)
	Right Middle Occipital Gyrus	14.86 (34, -62, 37)
#7 (72)	Right Inferior Occipital Gyrus	23.59 (25, -95, -8)
		15.83 (37, -95, -2)
#8 (65)	Right Cerebellum	15.40 (34, -68, -32)
		13.24 (40, -74, -53)
		13.11 (40, -62, -50)
Main effect: Site Palm >Arm
Cluster (size)	Peaks Locations	F (x, y, z)
#	Left Postcentral Gyrus	204.14 (50, -29, 61)
		131.45 (−41,–26, 49)
	Left Precentral Gyrus	126.07 (−35,–17, 64)
	Left Supplementary Motor Area	58.04 (−8,–2, 52)
	Left Superior Parietal Lobule	30.54 (−29,–59, 70)
	Left Superior Frontal Gyrus	26.92 (−20,–2, 70)
	Right Supplementary Motor Area	24.60 (10, -2, 52)
#2 (488)	Right Postcentral Gyrus	122.40 (55, -23, 55)
	Right Superior Parietal Lobule	17.13 (28, -53, 67)
#3 (444)	Right Cerebellum	119.12 (19, -53, -23)
#4 (304)	Right Precentral Gyrus	63.12 (37, -11, 67)
	Right Supplementary Motor Area	28.37 (16, 7, 67)
	Right Superior Frontal Gyrus	12.31 (19, -2, 55)
#5 (199)	Right Cerebellum	76.11 (19, -59, -50)
#6 (89)	Left Cerebellum	32.62 (−20,–53, 29)
#7 (72)	Right Middle Frontal Gyrus	19.27 (31, 40, 22)
	Right Superior Frontal Gyrus	17.35 (22, 46, 22)
Main effect: OT
Cluster (size)	Peaks Locations	F (x, y, z)
#1 (93)	Right Middle Temporal Gyrus	25.77 (49, 1, -20)
	Right Inferior Temporal Gyrus	16.66 (61, -14, 35)
		14.87 (55, -17, -23)
Interaction: Familiarity*OT
Cluster (size)	Peaks Locations	F (x, y, z)
#1 (585)	Right Superior Orbital Gyrus	37.78 (19, 55, -5)
	Left Anterior Cingulate Cortex	28.17 (-8, 49,–2)
		21.94 (-8, 49, 10)
	Right Anterior Cingulate Cortex	24.01 (10, 49, 13)
	Left Superior Frontal Gyrus	23.55 (-17, 61, 10)
	Left Mid Orbital Gyrus	23.50 (1, 55, -2)
	Right Mid Orbital Gyrus	19.24 (10,70, -11)
	Right Superior Medial Gyrus	18.75 (13, 64, 16)
#2 (436)	Right Middle Temporal Gyrus	36.99 (64, -17, -14)
	Right Medial Temporal Pole	34.12 (40, 16, -32)
	Right Inferior Temporal Gyrus	33.89 (49, -5, 29)
		21.52 (61, -14, 35)
		19.36 (52, -17, -26)
#3 (261)	Right Angular Gyrus	26.54 (46, -59, 34)
		25.40 (58, -56, 25)
		22.59 (43, -53, 25)
		21.22 (58, -62, 37)
	Right Middle Occipital Gyrus	15.73 (43, -74, 31)
#4 (185)	Left Inferior Temporal Gyrus	32.65 (-41, 4,–38)
		23.30 (−56,–5, –38)
	Left Middle Temporal gyrus	24.21 (-47, 4,–26)
	Left Medial Temporal Pole	13.62 (-53, 16,–32)
#5 (144)	Left Angular Gyrus	35.84 (−50,–65, 49)
		21.85 (−56,–56, 34)
	Left Inferior Parietal Lobule	13.96 (−41,–59, 58)
#6 (138)	Left Cerebellum	25.48 (−53,–59, –35)
		15.48 (−35,–77, –35)
		15.48 (−35,–77, –35)
		14.08 (−50,–68, –44)
		12.64 (−47,–74, –32)
		12.27 (−47,–62, –48)
#7 (92)	Right Superior Frontal Gyrus	39.89 (22, 28, 61)
		27.84 (22, 16, 67)
#8 (79)	Left Middle Frontal Gyrus	25.33 (-41, 25, 52)
		23.11 (-35, 40, 43)
#9 (62)	Right Inferior Temporal Gyrus	24.67 (43, -11, -32)
	Right ParaHippocampal Gyrus	15.74 (28, -8, 35)
	Right Fusiform Gyrus	14.45 (31, -2, -44)
Interaction: Order*OT
Cluster (size)	Peaks Locations	F (x, y, z)
#1 (195)	Right Inferior Temporal Gyrus	26.44 (43, -11, -32)
		15.30 (61, -14, -35)
	Right Middle Temporal Gyrus	22.79 (49, 1, -20)
Interaction: FamiliarityOrderOT
Cluster (size)	Peaks Locations	F (x, y, z)
#1 (111)	Right Superior Occipital Gyrus	29.02 (25, -98, 19)
	Right Cuneus	22.76 (19, -104, 10)
#2 (81)	Left Angular Gyrus	25.20 (−47,–59, 34)

Additional files

Supplementary file 1 Paired T-tests for partner vs stranger during each of 2 functional runs (first, second), modeled with linear mixed effects and weighted by the individual change in OT levels as covariates. All contrasts are thresholded at P<0.002, cluster-size thresholded at alpha = 0.05 FWE for n=27 complete functional datasets. For each cluster under each contrast heading, size in voxels, location, maximum T score, and MNI coordinates (x, y, z) are given. *=region of interest analysis.: https://cdn.elifesciences.org/articles/81197/elife-81197-supp1-v1.docx
Download elife-81197-supp1-v1.docx
Supplementary file 2 Paired T-test for partner vs stranger in partner first group, modeled with linear mixed effects and weighted by the individual mean cortisol levels as a covariate. All contrasts are thresholded at P<0.002, cluster-size thresholded at alpha = 0.05 FWE for n=18 complete functional datasets. For each cluster under each contrast heading, size, location, maximum T score, and MNI coordinates (x, y, z) are given.: https://cdn.elifesciences.org/articles/81197/elife-81197-supp2-v1.docx
Download elife-81197-supp2-v1.docx
Supplementary file 3 Regressor created by linear interpolation of serial OT samples, convolved with canonical HRF and modeled with factors toucher (partner, stranger) and order (first or second encounter), at time points 2 and 2.5 min preceding plasma sample collection. All contrasts are thresholded at P<0.002, cluster-size thresholded at alpha = 0.05 FWE for N=23 complete functional datasets. For each cluster under each contrast heading, size, location, maximum F score, and MNI coordinates (x, y, z) are given.: https://cdn.elifesciences.org/articles/81197/elife-81197-supp3-v1.docx
Download elife-81197-supp3-v1.docx
Supplementary file 4 Paired T-test for partner vs stranger in partner first group, modeled with linear mixed effects and weighted by individual difference in pleasantness ratings as a covariate. All contrasts are thresholded at P<0.005, cluster-size thresholded at alpha = 0.05 FWE for n=18 complete functional datasets. For each cluster under each contrast heading, size, location, maximum T score, and MNI coordinates (x, y, z) are given.: https://cdn.elifesciences.org/articles/81197/elife-81197-supp4-v1.docx
Download elife-81197-supp4-v1.docx
Supplementary file 5 Conjunction analyses showing common activations for partner and order factors (partner, stranger, first, second). All contrasts are thresholded at P<0.002, cluster-size thresholded at alpha = 0.05 FWE for n=37 functional datasets. For each cluster under each contrast heading, size in voxels, location, maximum T score, and MNI coordinates (x, y, z) are given. Negative BOLD in boldface.: https://cdn.elifesciences.org/articles/81197/elife-81197-supp5-v1.docx
Download elife-81197-supp5-v1.docx
Supplementary file 6 Participants included in each analysis based on analyzable data. OT = Oxytocin, CORT = Cortisol, LME = Linear Mixed Effects Model (neuroimaging), Regr.=Regressor (neuroimaging).: https://cdn.elifesciences.org/articles/81197/elife-81197-supp6-v1.docx
Download elife-81197-supp6-v1.docx
MDAR checklist: https://cdn.elifesciences.org/articles/81197/elife-81197-mdarchecklist1-v1.docx
Download elife-81197-mdarchecklist1-v1.docx
Source code 1 Afni_peak_sourcecode.: https://cdn.elifesciences.org/articles/81197/elife-81197-code1-v1.zip
Download elife-81197-code1-v1.zip