Supplementary Material for Ghahremani et al.Behavioral and neural markers of cigarette craving regulation in young-adult smokers during abstinence and after smoking

Figure S1. Mean craving ratings given after each video presentation for each task condition before and after smoking a cigarette for the sub-sample that received fMRI (N=21). Main effects of smoking cue, proximal/distal instruction, and cigarette smoking were observed; no significant interactions were found.Error bars reflect one standard error of the mean.

Figure S2.Parametric modulation analysis results showing regions that modulate with self-reported craving on a trial-by-trial basis during abstinence. Regions include rostral anterior cingulate, medial prefrontal cortex, lateral orbitofrontal cortex, posterior cingulate, dorsal and ventral striatum, lateral occipital cortex. Image shows thresholded z-statistic map overlaid on a group-averaged high-resolution anatomical image. R=right (images are displayed in radiological orientation; right is left). Color bar indicates z-statistic range.

Figure S3. Overlap of cue-induced craving (smoke vs. non-smoke cues) and trial-by-trial correlation with self-reported craving (parametric modulation analysis) during abstinence. Regions include ventral and dorsal striatum, rostral anterior cingulate, medial prefrontal cortex, bilateral orbitofrontal cortex, and lateral occipital cortex. Statistical conjunction analysis was computed using methods described by Nichols et al (2005).

Figure S4. Parameter estimate plots showing the interaction of abstinence/post-smoking and cues (smoking vs. non-smoking videos). Data presented are from four independent clusters emerging from the voxel-wise analyses shown in Figure 5 and listed in Table S1. Error bars reflect one standard error of the mean.

Figure S5. Mean craving ratings for five participants from days in which they did not smoke between test sessions. To determine test-retest effects of task performance independent of smoking, five participants participated in an extra day of testing in which all protocol procedures were followed except for cigarette smoking. Results indicated a trend towards increased craving from Time 1 to Time 2. Error bars reflect one standard error of the mean.

Figure S6. Sex differences in response to smoking vs. non-smoking cues. Females (N=10) showed greater activation than males (N=11) with respect to smoking vs. non-smoking videos. The strongest activation was found in caudate, thalamus, anterior cingulate cortex, and left superior/middle frontal gyrus (cluster-corrected at Z>2.3, P<0.05)

Table S1. Brain regions of significant activation. Z-statistics and X, Y, and Z MNI coordinates (mm) are from the location of peak voxel activation within each cluster. Local minima peaks are listed under max cluster peaks (indicated by voxel size). Z-statistic maps were whole-brain cluster-corrected at Z>2.3, P<.05.

Location / Voxels / Z-stat / X / Y / Z
Smoking cues > Non-smoking Cues
Midbrain / 7214 / 4.57 / -8 / -8 / -12
Thalamus / 3.73 / 2 / -8 / 8
Medial Prefrontal Cortex / 3.14 / -4 / 48 / -8
Bilateral Caudate / 4.57 / 10 / 18 / 0
3.83 / -12 / 14 / 0
Anterior Cingulate Cortex / 4.29 / 2 / 34 / 2
Bilateral Lateral Occipital Cortex/Middle Temporal/Fusiform Gyrus / 4585 / 5.88 / -44 / -66 / 6
3660 / 6.83 / 42 / -68 / 8
Bilateral Superior Parietal Lobule / 1095 / 4.5 / -34 / -50 / 52
939 / 5.02 / 34 / -44 / 60
Left Supramarginal Gyrus / 626 / 4.82 / -64 / -24 / 22
Posterior Cingulate Cortex / 586 / 4.27 / 2 / -36 / 30
541 / 4.28 / 56 / -28 / 26
Non-smoking cues Smoking cues
Precuneus/Cuneus/Occipto- Temporal/Lingual Gyrus / 4861 / 5.02 / -4 / -80 / 26
Bilateral Precentral Gyrus / 1762 / 4.35 / -48 / -16 / 56
3420 / 4.48 / 34 / -16 / 48
Right Superior Temporal Gyrus / 3.3 / 56 / -26 / -4
Close > Far
Right Frontal Pole/Superior Frontal Gyrus / 3244 / 4.12 / 18 / 56 / 30
3.69 / 6 / 58 / 4
Left Superior Frontal Gyrus / 3.99 / -6 / 34 / 54
Left Ventral Striatum / 3.37 / -8 / 8 / -6
Pre- > Post-Smoking (Smoking cues > Non-smoking cues)
Anterior Cingulate Cortex / 2532 / 3.9 / 0 / 12 / 38
3.01 / -2 / 44 / -6
Superior Frontal Gyrus / 3.06 / -2 / 56 / 26
Left Superior/Middle Temporal Gyrus / 2366 / 3.84 / -54 / 2 / -4
Left Amgydala / 2.9 / -20 / 0 / -18
Right Superior/Middle Temporal Gyrus / 991 / 3.68 / 62 / -2 / 4
Thalamus / 679 / 3.71 / 2 / -6 / 6
Right Caudate / 3.33 / 12 / 22 / 0
Parametric modulation with craving ratings
Anterior Cingulate Cortex / 12094 / 5.02 / 2 / 36 / 4
Bilateral Lateral Occipital Cortex / 4764 / 5.62 / -42 / -68 / 12
3432 / 4.93 / 50 / -70 / 8
Precuneus/Posterior Cingulate / 2614 / 4.22 / 2 / -54 / 30
Conjunction of parametric modulation and Smoking cues > Non-smoking cues
Anterior Cingulate Cortex / 2701 / 4.29 / 2 / 34 / 2
Bilateral Caudate / 3.31 / 10 / 16 / 0
3.37 / -8 / 10 / -2
Bilateral Orbitofrontal Cortex / 3.1 / -28 / 10 / -18
3.43 / 26 / 14 / -18
Right Lateral Occipital Cortex / 2300 / 4.76 / 44 / -66 / 6
Right Inferior Temporal/Fusiform Gyrus / 3.84 / 44 / -48 / -12
Left Lateral Occipital Cortex / 2175 / 5.15 / -44 / -70 / 10

MATERIALS AND METHODS

Imaging data analysis

The image time-course for each participant was first realigned to compensate for small head movements (Jenkinson et al, 2002), and all non-brain matter was removed using FSL’s brain extraction tool. Data were spatially smoothed using a 6-mm full-width-half maximum Gaussian kernel. Registration was conducted through a three-step procedure, whereby EPI images were first registered to the matched-bandwidth high-resolution structural image, then to the MPRAGE structural image, and finally into standard space (Montreal Neurological Institute (MNI) avg152 template), using 12-parameter affine transformations(Jenkinson and Smith, 2001). Registration from MPRAGE structural images to standard space was further refined using FNIRT nonlinear registration (Andersson et al, 2007a, b). Statistical analyses at the single-subject level were performed in native space, with the statistical maps normalized to standard space prior to higher-level analysis.

Whole-brain statistical analysis was performed using a multi-stage approach to implement a mixed-effects model treating participants as a random effects variable. Statistical modeling was first performed separately for each imaging run. Regressors of interest (“close smoke”, “far smoke”, “close nonsmoke”, “far nonsmoke”) were created by convolving a delta function representing video onset times with a canonical (double-gamma) hemodynamic response function with width equivalent to the duration of the video presentation (15 s). The task instruction period for each trial was also modeled. Motion parameters were included as covariates of no interest to account for variance associated with residual motion. These included three translation and three rotation parameters as well as framewise displacement and DVARS exceeding a run-specific threshold (75th percentile plus 1.5*interquartile-range).

To examine trial-by-trial relationships between self-reported craving and fMRI activation, we performed a separate analysis in which a parametric modulation covariate (Buchel et al, 1998) was added, indicating each participant’s response for each trial during abstinence. The covariate was demeaned for the regression model. To compute the overlap of cue-induced craving (smoke vs. non-smoke cues) and these parametric modulation results, we conducted astatistical conjunction analysis using methods described in Nichols et al (2005)with a height threshold of Z=2.3.

For all analyses, time-series statistical analysis was carried out using FILM (FMRIB's Improved Linear Model) with local autocorrelation correction (Woolrich et al, 2001)after highpass temporal filtering (Gaussian-weighted LSF straight line fitting, with sigma = 33 s).

Contrast images for runs within each scanning session were combined using a fixed effects analyses. To determine effects of smoking (pre- to post), pairwise, fixed effects analyses comparing contrast images from the two sessions were first conducted for each subject. These images were then submitted for group analyses using random effects analyses.

For group analyses, the FMRIB Local Analysis of Mixed Effects (FLAME1) module in FSL was used (Beckmann et al, 2003; Woolrich et al, 2004). Z (Gaussianised T) statistic images were thresholded using cluster-corrected statistics with a height threshold of Z > 2.3 and a cluster probability threshold of p < 0.05, whole-brain corrected using the theory of Gaussian Random Fields (Worsley et al, 1992). All group analyses were subjected to robust outlier deweighting (Woolrich, 2008). Anatomical locations of activations were identified using the Harvard-Oxford Probabilistic Atlas, which is included in the FSL software package, and the sectional brain atlas of Duvernoy and Bourgouin (1999).

REFERENCES

Andersson J, Jenkinson M, Smith S (2007a). Non-linear optimisation. FMRIB technical report. Vol TR07JA1.

Andersson J, Jenkinson M, Smith S (2007b). Non-linear registration, aka Spatial normalisation FMRIB technical report. Vol TR07JA2.

Beckmann CF, Jenkinson M, Smith SM (2003). General multilevel linear modeling for group analysis in FMRI. Neuroimage20(2): 1052-1063.

Buchel C, Holmes AP, Rees G, Friston KJ (1998). Characterizing stimulus-response functions using nonlinear regressors in parametric fMRI experiments. NeuroImage8(2): 140-148.

Duvernoy HM, Bourgouin P (1999). The Human brain: surface, three-dimensional sectional anatomy with MRI, and blood supply, 2nd completely rev. and enl. edn. Springer: Wien ; New York, 491pp.

Jenkinson M, Bannister P, Brady M, Smith S (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage17(2): 825-841.

Jenkinson M, Smith S (2001). A global optimisation method for robust affine registration of brain images. Med Image Anal5(2): 143-156.

Nichols T, Brett M, Andersson J, Wager T, Poline J-B (2005). Valid conjunction inference with the minimum statistic. NeuroImage25(3): 653-660.

Woolrich M (2008). Robust group analysis using outlier inference. Neuroimage41(2): 286-301.

Woolrich MW, Behrens TE, Beckmann CF, Jenkinson M, Smith SM (2004). Multilevel linear modelling for FMRI group analysis using Bayesian inference. Neuroimage21(4): 1732-1747.

Woolrich MW, Ripley BD, Brady M, Smith SM (2001). Temporal autocorrelation in univariate linear modeling of FMRI data. Neuroimage14(6): 1370-1386.

Worsley KJ, Evans AC, Marrett S, Neelin P (1992). A three-dimensional statistical analysis for CBF activation studies in human brain. J Cereb Blood Flow Metab12(6): 900-918.