|
nif:isString
|
-
The initial study sample consisted of 60 subjects. They were first tested for their ability to smell androstadienone in order to be able to participate in the fMRI study. We excluded 21 subjects because of anosmia to androstadienone. From the 39 remaining subjects, one did not show up for his fMRI scan and one subject was excluded from further analysis after she reported having a bisexual orientation. One of our inclusion criteria was having a heterosexual orientation which was assessed by means of self-report. Thus, finally, a total of 21 women (mean age = 24.7, SEM = 1.4 years) and 16 men (mean age = 24.7, SEM = 1.1 years) participated in the fMRI study. They were all right-handed, heterosexual volunteers and were not using any medication. Females used no oral contraceptives and all had regular menstrual cycles. Normal olfactory function of the participants was ascertained by means of the “Sniffin’ Sticks” test battery [22]. An estimation of the detection threshold for androstadienone (Steraloids Inc., Newport, RI, USA) was obtained for each participant with a three-alternative forced-choice detection test consisting of six concentrations of androstadienone dissolved in propylene glycol (10 mM, 1 mM, 0.1 mM, 0.001 mM, 0.0001 mM, 0.00001 mM), with one of the three glass jars containing androstadienone and the other two the solvent, propylene glycol (Fluka, Sigma-Aldrich Chemie Gmbh, Munich, Germany). The three glass jars, containing 10 ml of liquid each, were presented in ascending order until subjects had correctly identified androstadienone in two consecutive trials, which was then used as an estimation of the detection threshold for this participant. During the screening session, subjects were asked to report the intensity (on a scale from 0 to 10) and quality of the 10 mM, “high” concentration (on a scale from −5 to +5). MRI sessions were planned a few days after the day of screening. Participants were asked not to wear any perfume on the day of scanning and not to eat or drink anything other than water one hour before scanning. Subjects were instructed to breathe through their mouths during the odor stimulations and females were scanned during the second or third week of their menstrual cycle, which was determined by means of self-report. Subjects gave their informed consent, according to the Declaration of Helsinki, and the study was approved by the Ethics Committee of the Medical Faculty, University of Dresden (application number EK373122009).
Subjects were exposed to two different olfactory stimuli during the scanning sessions: 1) androstadienone and 2) 1-butanol (Merck Chemicals, Darmstadt, Germany). Androstadienone, which were diluted in propylene glycol to three different concentrations: 10 mM “high,” 0.1 mM “medium” and 0.001 mM “low,” was used to determine whether there would be gender-specific effects in hypothalamic activation as well as whether there would be any dose-dependent effects of the steroid on hypothalamic activation. The medium concentration was based on the mean detection threshold for androstadienone. Butanol was diluted in propylene glycol to a concentration of 0.01 mM and was used as a control odor. The volume of each solution used during the fMRI experiments was 20 ml. Olfactory stimuli were delivered through a tubing system to the subjects’ nostrils by means of an air-dilution olfactometer [23]. Briefly, in the olfactometer setup used in our study, air is obtained from a regular clean air outlet in the wall and directed via computer-controlled electro-pneumatic valves to 5 gas-washing bottles, which contain the liquid odorants (in our case: androstadienone in three different concentrations) and the control fluids (butanol and water). When the air enters the olfactometer, a constant airflow is achieved by using a ball-flow-meter, The air is then directed into a solenoid operated three-way pneumatic switching valve. In its turned-off-state, the air leaves the valve through the port which functions as an exhaust and maintains a steady airflow. When the airflow-valve is turned on, the air leaves this valve passing the normally open port, which is split into 5 connections. This separates the air into 5 streams and distributes it over five solenoid operated three-way pneumatic switching valves. The air is saturated with odorant as it travels through the porous frit at the bottom of the gas-washing bottles and forms small air-bubbles. Upon exiting the 5 gas washing chambers, the odorized or control air passes through 5 individual Teflon tubes to the delivery section. Just before entry into the subject’s nose, the gas-flow is reunited using t-fittings into a single flow line inside a 5 cm Teflon hose. The olfactometer uses a continuous airflow design and as a result, regardless of how many valves are simultaneously opened, the resulting airflow at the delivery unit always matches the one entering the device avoiding air-puffs at the outlet. With a total air flow of 2 L per minute, during “ON” periods every 2 seconds odors where delivered during 1 second, while during “OFF” periods subjects received odorless air. The presentation order of odors (3 × androstadienone, 1 × butanol) was randomized across subjects. After each odor session, subjects were asked to report intensity and valence of the three different androstadienone concentrations and the control odor butanol on a 10-point scale.
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0040993.t001 Androstadienone threshold estimations during the screening session. Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0040993.t002 Intensity ratings during screening and scan sessions on a scale from 0 to 10. AND = androstadienone; St.dev. = standard deviation.
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0040993.t003 Valence ratings during screening and scan sessions on a scale from −5 to +5. AND = androstadienone; St.dev. = standard deviation.
Scans were performed on a 1.5 Tesla scanner (Siemens Magnetom SONATA; Siemens Healthcare, Erlangen, Germany). A gradient echo (GE) echoplanar imaging (EPI) sequence was used for functional imaging. The parameters included a 10×15 cm2 field of view, TR of 2240 msec, TE of 40 msec, a 90° flip angle, 27 slices, voxel resolution of 2.3×2.3×3 mm. The scanning volume was centered on the hypothalamus. Before each imaging session an automated local shimming technique was used to reduce susceptibility artifacts. A scanning session consisted of four subsessions, during each of which 108 images were acquired, lasting 4.06 min. Each odor session consisted of 6 alternating “ON-OFF” cycles over 108 scans in a classical block design. For co-registration with the functional images a T1-weighted scan was obtained (3D IR/GR sequence, TR = 2180 ms, TE = 3.93 ms).
Figure data removed from full text. Figure identifier and caption: 10.1371/journal.pone.0040993.g001 Hypothalamus activation for the interaction effect of the factors “group” and “androstadienone concentration”.F-contrast, testing for group (men and women) differences in hypothalamus activation, dependent on androstadienone concentration condition. The color bar depicts the F statistic and image coordinates (x = 2; y = −6; z = 0) are in Montreal Neurological Institute brain atlas space.
fMRI data were analyzed using SPM8 (Wellcome Trust Centre for Neuroimaging, http://www.fil.ion.ucl.ac.uk/spm). Functional images were realigned to the mean image, co-registered with the individual T1 scan and normalized to MNI (Montreal Neurological Institute) space using segmentation. Finally, images were spatially smoothed with a 4 mm full width half maximum (FWHM) isotropic Gaussian kernel. This setting was chosen not to exceed the size of the anatomical structure of interest, the hypothalamus. In order to identify the effect of odor stimulation, first level linear contrast images were entered into a general linear model (GLM), generating statistical images. These images displayed the effect of odor stimulation (“On” blocks versus “Off” blocks) in each participant. Next, these contrast images were entered into second-level group analyses to investigate the effects of group (men versus women) and odor concentration. In order to address the main questions of the study, i.e. whether there are any sex differences in hypothalamic activation upon smelling androstadienone, and whether the hypothalamus would respond differently in the two sexes to the three androstadienone concentrations, a flexible factorial model was applied. The factors “gender,” having two levels (females and males) and “androstadienone,” having three levels (three different concentrations) were defined, as well as the factor “subject” (accounting for multiple within-subject observations), in order to investigate the interaction effects of gender and concentration. Furthermore, to correct for subjective differences in perceived odor concentrations, the behavioral variable of subjective ratings of odor intensity, obtained after each odor session, was included as a covariate in the flexible factorial model. Sex differences for each of the three androstadienone concentrations were analyzed separately using two-sample t-tests. We defined a hypothalamic region of interest (with Marsbar, [24]) by modeling brain activation across all subjects and androstadienone concentrations, and built a region of interest for the cluster of hypothalamic activation (peak voxel MNI-coordinates: x = 2, y = −6, z = 0; volume size = 250 mm3). This region of interest was applied at the second level to investigate differences between groups and concentrations. For illustrative purposes we extracted contrast values for each concentration condition of androstadienone, using this hypothalamic region of interest. It was ensured that the reported cluster fell within coordinates defined as pertaining to the hypothalamus in earlier studies [19], [20], [25]-[29]. The statistical threshold was set at p<0.05, using family-wise error (FWE) correction for multiple comparisons in small volumes.
Figure data removed from full text. Figure identifier and caption: 10.1371/journal.pone.0040993.g002 Sex differences in hypothalamus activation. (A) Hypothalamus activation in women, compared to men (directional t-contrast), who were exposed to the high concentration of androstadienone. (B) Hypothalamus activation in men, compared to women (directional t-contrast), who were exposed to the medium concentration of androstadienone. The color bar depicts the t statistic and image coordinates (x = 2; y = −6; z = 0) are in Montreal Neurological Institute brain atlas space.
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0040993.t004 One-sample and two-sample t-tests, with directional t-contrasts testing for sex differences in the hypothalamus, separately for the three different androstadienone concentration conditions. AND = androstadienone; FWE = family-wise error; * = statistically significant at p<0.05; coordinates are in Montreal Neurological Institute (MNI) brain atlas space.
Figure data removed from full text. Figure identifier and caption: 10.1371/journal.pone.0040993.g003 Contrast values for the three different androstadienone concentrations.Contrast values were extracted from the hypothalamus (region of interest at MNI coordinates x = 2; y = −6; z = 0, sphere diameter of 8 mm). Contrast values were determined for each of the three androstadienone concentration conditions and depicted separately for men and women. Error bars reflect the 95% confidence interval.
Additionally, we investigated a set of predefined olfactory regions of interest for possible sex differences in activation, i.e. amygdala, piriform cortex, insula, the orbito-frontal cortex (OFC) and the anterior cingulate cortex (ACC) since it has been suggested that these brain areas are implicated in the processing of androstadienone [20], [30]. For these analyses, the dataset was smoothed differently, applying a kernel of 6 mm. This setting was chosen in order to increase the signal-to-noise ratio and to account for the larger anatomical variability in cortical brain regions, compared to the (much) smaller hypothalamus. The anatomical borders for each area of interest were defined by using overlays generated by the Wake Forest University (WFU) Pickatlas [31] toolbox for SPM. Applying these regions of interest as mask images, the data were analyzed with respect to the effects of each odor (androstadienone and butanol) and for sex differences in response to smelling androstadienone (across concentrations) or butanol. The statistical threshold was set at p<0.05 (FWE corrected).
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0040993.t005 Main effect analyses for the two odors androstadienone and butanol with one-sample t-tests investigating whole group brain activations in predefined olfactory regions of interest. * = statistically significant at p<0.05; FWE = family-wise error; L = left; R = right; ACC = anterior cingulate cortex; OFC = orbito-frontal cortex; coordinates are in Montreal Neurological Institute (MNI) brain atlas space.
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}