Edited by:Zafiris J. Daskalakis, University of Toronto, Canada
Reviewed by:Paul Croarkin, UT Southwestern Medical Center, USA; Tarek Rajji, Centre for Addiction and Mental Health, Canada
*Correspondence:Ziad Nahas, Department of Psychiatry and Behavioral Sciences, Institute of Psychiatry, 67 President St., Charleston, SC 29425, USA.; e-mail:
This is an open-access article subject to an exclusive license agreement between the authors and the Frontiers Research Foundation, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are credited.
Mentalization, a developmentally derived ability to affectively and cognitively infer the mental state of others is crucial to sociality and is enhanced by propitious genetics and secure attachments (Choi-Kain and Gunderson,
Although little is known about the neurophysiology of such experiences, there is evidence that oxytocin, a neuropeptide, promotes positive social interactions and preferences (Hurlemann et al.,
Theory of mind (ToM) is a research and conceptual rubric of an “imagined involvement” that was derived from studies of autism. Mentalization is a higher order capacity, either deficient in clinical populations or progressively enhanced with better caretaking and affective attunement. The failure to evolve mentalization partially explains pervasive developmental disorders (Ozonoff et al.,
A “Reading the Mind in the Eyes test” (RMET) (Baron-Cohen et al.,
To date, no studies have investigated the functional neuroanatomy of mentalization in depression nor the effects of oxytocin in modulating brain activity in depressed individuals. Accordingly, we used the RMET task and an intra-nasal oxytocin challenge to investigate mentalization in unmedicated depressed patients and healthy controls. We hypothesized that depressed patients may show greater limbic activity during the mentalization task compared to matched controls, and that oxytocin will lead to a significant decrease in ventro-limbic activity when compared to placebo.
This is a randomized double-blind crossover design to investigate the acute effects of intra-nasal oxytocin on mentalization in unmedicated depressed patients and matched controls. Primary outcome measures consisted of functional blood oxygen level dependent MRI (BOLD fMRI) scans. Secondary outcome measures were behavioral measures (speed of reaction time). This study was approved by the Medical University of South Carolina (MUSC) Institutional Review Board funded by a grant from the Hope for Depression Research Foundation, and collaboratively conducted with the MindBrain Consortium at Summa Hospitals, Akron, Ohio.
Subjects were recruited from the MUSC outpatient psychiatric services, the Mood Disorders Program (MDP) and through local advertisement. We screened 15 Healthy and 20 Depressed subjects and enrolled 22 adult subjects. All subjects signed written informed consent. Depressed patients met criteria for major depressive episode by Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) criteria of less than 24 months in duration with an antidepressant treatment history form (ATHF) severity one or less failed trials. Healthy controls were screened for any active DSM-IV-TR Axis I diagnosis (with the exception of caffeine or nicotine abuse). Subjects were off all psychotropic medications for at least 2 weeks prior to first scanning session. Subjects were excluded if they were medically unstable or had an active neurodegenerative or epilepsy disorder. All presented data are from subjects who completed a baseline clinical assessment visit along with two scanning sessions over a maximum of 3 weeks period.
At each scanning session, subjects underwent two identical blood oxygentation dependant (BOLD) fMRI scans before and approximately 10 min after inhalation of study drug (20 IU oxytocin or placebo per nostril; Lee-Silsby Pharmacy, Cleveland, OH, USA). The oxytocin dose is comparable to that used previously in human studies of behavior (Epperson et al.,
We adapted the reading the mind in the eyes paradigm (RMET) for presentation in the scanner and an event-fMRI design. We used the Integrated Functional Imaging System (IFIS-SA) (Invivo, Orlando, FL, USA) to visually present the paradigm and record responses of the participants. The RMET task consisted of three conditions: affect attribution of faces cropped around the eyes in rectangular shape (referred in the text as RMET), gender identification of the same photographs, and a basic motor cued response. Possible attribute adjectives, gender attributes (“male,” “female”) and motor cues (“yes,” “no”) were positioned at the four corners with optimal or normed responses equally randomized for all presentations. Participants pressed a hand-pad button to indicate their attribute of choice. On the hand-pad, the left and right index fingers corresponded to the top left and top right corners of images, and the left and right thumbs corresponded to the bottom left and bottom right corners of images. This entire paradigm consisted of 32 events for each condition (96 events total presented in a fixed-randomized order) and lasted approximately 769 s. The duration of each event lasted 7 s, separated by a jittered inter-stimulus black screen [inter-stimulus interval (ISI) range 6–4753 ms, ISI average = 1008 ms].
Images were acquired using a 3T MRI scanner (Intera, Philips Medical System, Netherlands) with a SENSE parallel imaging head-coil. A sagittal reference image was first acquired, to guide positioning of the functional scans along the anterior commissure-posterior commissure line. For scanning during the RMET paradigm, echoplanar (EPI) transverse images were acquired with the following parameters: 414 volumes, repetition time = 1867 ms, echo time = 30 ms, flip angle = 90°, field of view = 208 mm, matrix = 64 × 64, SENSE factor = 2, 32 slices, 3.25 mm with no gap, yielding a voxel size of 3.3 × 3.3 × 3.25 mm3.
Descriptive statistics were performed using SPSS 16.0 for Macintosh (SPSS Inc., Chicago, IL, USA) and included mean, standard deviation, Student's
fMRI image data was first transferred to a workstation where it was converted into Analyze format using MRICro1
Individual responses collected by IFIS during RMET were collated, checked for missing responses and collated in an Excel spreadsheet. Reaction Time (RT) and Accuracy of Response (AC) were noted. Final analysis included mean, standard deviation, Student's
For mixed models examining the effects of drug by time, pre-drug administration conditions were coded in the dataset as “0,” post-drug were coded as “1,” placebo conditions were coded as “0” and oxytocin conditions were coded as “1.” Hierarchical Linear Modeling (HLM) was implemented using PROC MIXED in SAS (SAS Institute Inc., Cary, NC) to assess the linear effect of the interaction of “time” (pre to post-drug administration) and “treatment type” (oxytocin versus placebo) on reaction time and accuracy of response between groups (thereby by testing a series of 2 × 2 × 2 mixed models). HLM has been shown to handle nested models with serially dependent data points and randomly distributed missing values appropriately (Raudenbush and Chan,
Time-series statistical analysis was carried out using FILM with local autocorrelation correction (Woolrich et al.,
Eight unmedicated depressed subjects (eight females, four African-Americans, four Caucasians) and nine matched controls (eight females, five African-Americans, four Caucasians) completed all assessments required for this study. The mean age for depressed subjects was 35.5 years (SD = 10.62), ranging from 26 to 60, and for controls, 36.4 years (SD = 11.40), ranging from 25 to 59. Mean years of education post high school were 2.56 (SD = 1.95) for depressed subjects and 4.22 (SD = 1.79) for controls. Co-morbidities of depressed subjects included a past history of alcohol dependence (
Depressed | Controls | Significance | |
---|---|---|---|
Age | 0.86 | ||
Sex | 8 Female | 8 Female, 1 male | |
Race | 4 African-American, 4 White | 5 African-American, 4 White | |
Marital status | 5 Single, 2 married, 1 separated | 3 Single, 4 married, 2 separated | |
Education (years post high school) | 0.087 | ||
Employment | 8 Full time, 1 part time | 4 Full time, 3 part time, 1 homemaker | |
HRSD | 0.000 | ||
IDS-SR 30 items | 0.000 | ||
STAI Y-2 | 0.000 | ||
ANPS-seeking | 0.161 | ||
ANPS-fear | 0.003 | ||
ANPS-care | 0.000 | ||
ANPS-anger | 0.290 | ||
ANPS-play | 0.206 | ||
ANPS-sadness | 0.000 | ||
ANPS-spirituality | 0.057 | ||
MSCEIT total | 0.596 | ||
Empathy quotient | 0.074 |
Figure
Pre-placebo | Pre-oxytocin | Post-placebo | Post-oxytocin | |
---|---|---|---|---|
RT RMET | ||||
Matched controls | 4396.3 ± 1499.8 | 4198.5 ± 1728.6 | 4092.5 ± 1560.8 | 4222.8 ± 1293 |
Depressed | 3945.1 ± 1260.7 | 4201.6 ± 1169.3 | 3878.5 ± 1328.3 | 4119.8 ± 1297.2 |
RT gender attribution | ||||
Matched controls | 2168.7 ± 958.3 | 2126.1 ± 853.3 | 2023.2 ± 950.7 | 1936.4 ± 710.3 |
Depressed | 1673.7 ± 651.6 | 1787.4 ± 717.6 | 1648.3 ± 774.5 | 1771.3 ± 880.8 |
RT motor cued response | ||||
Matched controls | 1416.8 ± 592 | 1464.4 ± 730.2 | 1363.1 ± 791.9 | 1351.7 ± 465.6 |
Depressed | 1214.7 ± 472.9 | 1257.6 ± 505 | 1134.6 ± 461.8 | 1240.5 ± 395.9 |
AC RMET | ||||
Matched controls | 60.6% | 59.4% | 64.2% | 67.9% |
Depressed | 65.8% | 67% | 67% | 70.6% |
AC gender attribution | ||||
Matched controls | 89.4% | 82.8% | 87.8% | 90% |
Depressed | 87.2% | 88.1% | 83.8% | 86.9% |
AC motor cued response | ||||
Matched controls | 98.1% | 94.1% | 93.1% | 97.8% |
Depressed | 96.9% | 97.5% | 95.6% | 97.2% |
There were no significant changes with accuracy of response (AR) across variables or conditions. Table
The corresponding brain atlas coordinates and the fMRI activations for healthy controls and depressed subjects are presented in Table 3.
z-Score | x | y | z | Left/right | Region | Brodmann area |
---|---|---|---|---|---|---|
5.7 | −46 | 14 | 26 | Left | Middle frontal gyrus | Brodmann area 9 |
5.57 | 22 | −64 | 54 | Right | Superior parietal lobule | Brodmann area 7 |
5.51 | −6 | 6 | 60 | Left | Superior frontal gyrus | Brodmann area 6 |
5.4 | 32 | −62 | 52 | Right | Superior parietal lobule | Brodmann area 7 |
5.19 | −20 | −64 | 58 | Left | Superior parietal lobule | Brodmann area 7 |
4.87 | −32 | −84 | −10 | Left | Inferior occipital gyrus | Brodmann area 18 |
4.5 | −10 | 26 | 56 | Left | Superior frontal gyrus | Brodmann area 6 |
4.39 | −10 | 12 | 6 | Left | Caudate | Caudate Body |
4.32 | −6 | 54 | 40 | Left | Medial frontal gyrus | Brodmann area 9 |
3.99 | −12 | 50 | 28 | Left | Superior frontal gyrus | Brodmann area 9 |
3.91 | −10 | 58 | 22 | Left | Superior frontal gyrus | Brodmann area 10 |
3.85 | 56 | −40 | −4 | Right | Middle temporal gyrus | Brodmann area 21 |
5.79 | −46 | 26 | −10 | Left | Inferior frontal gyrus | Brodmann area 47 |
5.21 | −54 | 22 | 8 | Left | Inferior frontal gyrus | Brodmann area 45 |
4.52 | −52 | −48 | 2 | Left | Middle temporal gyrus | Brodmann area 22 |
3.79 | −42 | 10 | −36 | Left | Middle temporal gyrus | Brodmann area 21 |
5.29 | 10 | 18 | 40 | Right | Cingulate gyrus | Brodmann area 32 |
4.83 | 36 | 22 | 0 | Right | Insula | Brodmann area 13 |
4.66 | −22 | −58 | 44 | Left | Superior parietal lobule | Brodmann area 7 |
4.55 | 30 | −4 | 50 | Right | Middle frontal gyrus | Brodmann area 6 |
4.35 | −10 | 8 | 2 | Left | Caudate | Caudate Head |
4.3 | −58 | 22 | 4 | Left | Inferior frontal gyrus | Brodmann area 45 |
4.14 | −48 | 14 | −161 | Left | Superior temporal gyrus | Brodmann area 38 |
4.11 | −40 | 28 | −10 | Left | Inferior frontal gyrus | Brodmann area 47 |
4.04 | −10 | 14 | −6 | Left | Caudate | Caudate head |
2.3 | −14 | 10 | −22 | Left | Anterior cingulate gyrus | Brodmann area 25 |
3.93 | −6 | 6 | −8 | Left | Caudate | Caudate head |
3.82 | 12 | −4 | −20 | Right | Parahippocampal gyrus | Brodmann area 34 |
2.66 | 14 | −2 | −22 | Right | Amygdala | |
3.73 | 28 | 32 | −18 | Right | Inferior frontal gyrus | Brodmann area 47 |
3.72 | −8 | 0 | −6 | Left | Lentiform nucleus | * |
3.67 | −10 | 32 | −6 | Left | Anterior cingulate | Brodmann area 24 |
3.67 | 30 | −24 | −24 | Right | Parahippocampal gyrus | Brodmann area 35 |
4.24 | −40 | −30 | −10 | Left | Caudate | Caudate tail |
3.94 | −50 | 8 | −18 | Left | Superior temporal gyrus | Brodmann area 38 |
3.7 | −56 | −8 | −6 | Left | Superior temporal gyrus | Brodmann area 22 |
2.7 | −4 | 18 | −12 | Left | Anterior cingulate gyrus | Brodmann area 25 |
2.69 | −4 | 48 | −14 | Left | Anterior cingulate gyrus | Brodmann area 32 |
No significant differences | ||||||
3.99 | 50 | −27 | 10 | Right | Superior temporal gyrus | Brodmann area 41 |
3.76 | 16 | −30 | 11 | Right cerebellum | ||
3.72 | 60 | −21 | 9 | Right | Superior temporal gyrus | Brodmann area 42 |
3.65 | 47 | −31 | 15 | Right | Superior temporal gyrus | Brodmann area 41 |
3.62 | 51 | −23 | 10 | Right | Transverse temporal gyrus | Brodmann area 41 |
3.62 | 51 | −30 | 15 | Right | Superior temporal gyrus | Brodmann area 41 |
3.99 | 58 | 10 | −20 | Right | Superior temporal gyrus | Brodmann area 38 |
3.79 | 54 | 8 | 2 | Right | Insula | Brodmann area 13 |
3.77 | 56 | 12 | 6 | Right | Precentral gyrus | Brodmann area 44 |
3.73 | 50 | 18 | −14 | Right | Inferior frontal gyrus | Brodmann area 47 |
3.71 | 58 | 8 | −10 | Right | Superior temporal gyrus | Brodmann area 22 |
3.76 | −4 | 36 | 16 | Left | Anterior cingulate | Brodmann area 24 |
3.75 | 6 | 34 | 26 | Right | Cingulate gyrus | Brodmann area 32 |
3.72 | −4 | 36 | 20 | Left | Anterior cingulate | Brodmann area 32 |
3.7 | 6 | 38 | 30 | Right | Medial frontal gyrus | Brodmann area 6 |
3.73 | −56 | −30 | 10 | Left | Superior temporal gyrus | Brodmann area 41 |
3.73 | −64 | −52 | 24 | Left | Superior temporal gyrus | Brodmann area 22 |
3.69 | −58 | −48 | 32 | Left | Supramarginal gyrus | Brodmann area 40 |
3.65 | −52 | −32 | 8 | Left | Superior temporal gyrus | Brodmann area 22 |
3.54 | 30 | 58 | −6 | Right | Superior frontal gyrus | Brodmann area 10 |
3.52 | 44 | 36 | 28 | Right | Middle frontal gyrus | Brodmann area 9 |
3.41 | 28 | 44 | 34 | Right | Middle frontal gyrus | Brodmann area 8 |
3.35 | 34 | 52 | 12 | Right | Middle frontal gyrus | Brodmann area 10 |
3.33 | 24 | 48 | 22 | Right | Superior frontal gyrus | Brodmann area 9 |
No significant differences | ||||||
No significant differences | ||||||
3.51 | 0 | 58 | 28 | Left | Medial frontal gyrus | Brodmann area 9 |
3.41 | −18 | 56 | 30 | Left | Superior frontal gyrus | Brodmann area 9 |
3.39 | 4 | 60 | 34 | Right | Medial frontal gyrus | Brodmann area 8 |
3.44 | −2 | 58 | 38 | Left | Medial frontal gyrus | Brodmann area 8 |
3.47 | 2 | 56 | 32 | Left | Medial frontal gyrus | Brodmann area 6 |
3.59 | −2 | 34 | 54 | Left | Superior frontal gyrus | Brodmann area 6 |
3.6 | 8 | 6 | 56 | Right | Medial frontal gyrus | Brodmann area 6 |
During the RMET task, at baseline and prior to inhaling study drug, depressed and normal control subjects showed statistically significant activations of the visual cortex, associated visual areas, the fusiform gyrus, and various limbic areas, along with the anterior cingulate. The depressed group showed significant activation in the right anterior cingulate gyrus (BA32), insula and middle frontal gyrus (BA6), left superior parietal lobe (BA7) and head of the caudate. The two groups were not significantly different in intensities or areas of activation on the mentalization task. They did however differ when contrasting RMET minus gender identification (see Figure
During RMET, there was significant activation with oxytocin compared to placebo in bilateral superior temporal gyrus (BA38, 41, and 22), right insula, right precentral gyrus (BA44), bilateral cingulate gyri (BA32), left anterior cingulum (BA24), left supramarginal gyrus (BA40), right inferior frontal gyrus (BA47), right superior and middle frontal gyri (BA8, 9, and 10) (see Figure
During RMET, there was significant activation with oxytocin compared to placebo in right parahippocampal gyrus extending to the amygdala, inferior frontal gyrus (BA47), left head and tail of the caudate and superior temporal gyrus (BA22 and 38) (see Figure
To our knowledge, this is the first study investigating intra-nasal oxytocin in unmedicated depressed and matched controls using fMRI. We employed RMET, a visual perceptual task that involves more affectively valenced decision making and mental attributions of other people. The task emphasizes affective “involvement” rather than cognitive “appraisal.” At baseline, the two groups were significantly different in their distress symptoms and capacity to empathize. Along with activating attentional and visual networks, RMET also involved anterior ventromedial regions. Oxytocin differentially affected normal and depressed cohorts. In controls, oxytocin enhanced activation of ventromedial, amygdala, parahippocampal and semantic associative areas. Depressed subjects showed increases in higher order cognitive areas and insula. Oxytocin also showed opposite effects in reaction times with increased speed of response in healthy controls and a slowing among the depressed.
The mentalization task is a novel paradigm in this population. It is relatively complex and recruits several complementary limbic, attention and high-level executive networks. Pathological states such as depression degrade mentalization (Uekermann et al.,
Although we did not see any significant difference between groups with RMET contrast, we did find higher anterior subgenual cingulate activation and anterior temporal pole during RMET minus gender attribution in depressed, suggesting that mentalization preferentially linked processing of perceptual inputs to internal visceral emotional experiences (Figure
Our fMRI data reveals that intra-nasal oxytocin has a differential effect upon the brain of depressed and normal people, as if in the state of depression, oxytocin “recruits” a more emotionally reflexive appraisal state that is less affectively reflective. Oxytocin increased primarily emotional circuits as in the cingulate (BA24) and insula in depressed patients during RMET. These increases appear to be in the same general directions as baseline activation patterns (Figure
Number of subjects is relatively small although the total scans analyzed is 80. Individual variability and total sample size cannot identify main responders. The exclusion of one male subject in healthy controls did not significantly change the results. The effect size on behavioral response (RT) is small and perhaps due to relatively small single dose of oxytocin and the intra-nasal route. Added to this are the small size of the olfactory nasal mucosa, the small instillation volume and the relatively fast turnover of cerebrospinal fluid which will render the total amount of active oxytocin that reached the brain small (Pontiroli,
In summary, we have shown that depressed subjects perform a mental attribution task differently than healthy matched controls. In addition, a single dose of intra-nasal oxytocin has distinct effects on functional activity of networks involved in a mental attribution task and enhances cognitive and emotional appraisal. These effects appear to be dependent on the distress levels of individuals, their likely capacity to empathize and the underlying homeostasis of the affective primary brain processes. It is not clear from our study what role, if any, oxytocin may have in modulating and perhaps treating depressive symptoms. Other studies of this important component of depressive pathology are needed with likely longer exposure and repeated oxytocin administration.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Study funded by the Hope for Depression Research Foundation (ZN). Study was also made possible with general funds from NIMH 1K08MH70615-01A1 (ZN), the Mood Disorders Program, the Brain Stimulation Laboratory, the Center for Advanced Imaging Research at the Medical University of South Carolina (MUSC). The authors would like to thank Lee-Silsby Pharmacy in Cleveland, Ohio for providing the intra-nasal oxytocin and matching placebo, Kimberly Porter, Rph at MUSC Investigational Drug Services for randomization and dispensing drug, and the South Carolina Research Authority in Charleston, South Carolina for imaging storage and support.
1http://www.sph.sc.edu/comd/rorden/mricro.html
2www.fmrib.ox.ac.uk/fsl