Conceived and designed the experiments: NDV GJW. Performed the experiments: FT GJW RZG NAK. Analyzed the data: NDV DT CW. Contributed reagents/materials/analysis tools: JSF DT. Wrote the paper: NDV JSF.
Dr. Volkow reports no competing interests; Dr. Tomasi reports no competing interests; Dr. Wang reports no competing interests; Dr. Fowler reports no competing interests; Dr. Telang reports no competing interests; Dr. Goldstein received consultation fee from Medical Directions, Inc. and honoraria fee from Federal Judicial Center and the Gruter Institute for Law and Behavioral Research; Dr. Klein reports no competing interests; Mr. Wong reports no competing interests. This does not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials.
Gender differences in vulnerability for cocaine addiction have been reported. Though the mechanisms are not understood, here we hypothesize that gender differences in reactivity to conditioned-cues, which contributes to relapse, are involved.
To test this we compared brain metabolism (using PET and 18FDG) between female (n = 10) and male (n = 16) active cocaine abusers when they watched a neutral video (nature scenes) versus a cocaine-cues video.
Self-reports of craving increased with the cocaine-cue video but responses did not differ between genders. In contrast, changes in whole brain metabolism with cocaine-cues differed by gender (p<0.05); females significantly decreased metabolism (−8.6%±10) whereas males tended to increase it (+5.5%±18). SPM analysis (Cocaine-cues vs Neutral) in females revealed decreases in frontal, cingulate and parietal cortices, thalamus and midbrain (p<0.001) whereas males showed increases in right inferior frontal gyrus (BA 44/45) (only at p<0.005). The gender-cue interaction showed greater decrements with Cocaine-cues in females than males (p<0.001) in frontal (BA 8, 9, 10), anterior cingulate (BA 24, 32), posterior cingulate (BA 23, 31), inferior parietal (BA 40) and thalamus (dorsomedial nucleus).
Females showed greater brain reactivity to cocaine-cues than males but no differences in craving, suggesting that there may be gender differences in response to cues that are not linked with craving but could affect subsequent drug use. Specifically deactivation of brain regions from “control networks” (prefrontal, cingulate, inferior parietal, thalamus) in females could increase their vulnerability to relapse since it would interfere with executive function (cognitive inhibition). This highlights the importance of gender tailored interventions for cocaine addiction.
Gender differences in the vulnerability for substance use disorders including cocaine
dependence have been reported
The mechanisms underlying the reported gender differences in the vulnerability for
cocaine use disorders are likely to be multiple including differences in
pharmacological sensitivity to cocaine
To test this hypothesis we compared the regional brain metabolic changes induced by
cocaine-cues between female and male cocaine abusers. We used Positron Emission
Tomography (PET) and 2-deoxy-2[18F]fluoro-D-glucose
(18FDG) to measure brain glucose metabolism (marker of brain
function)
Self-reports of craving were increased by the cocaine-cues video; the repeated factorial ANOVA showed a significant condition effect (F = 5, df 1, 24; p<0.03) but the interaction was not significant (F = 2, p<0.17). Females tended to show larger increases (3.4±2 vs 5.8±3) than males (3.0±3 vs 3.7±3) but the gender difference was not significant. Similarly the cocaine-cues increased the scores on the CCQ both in females (29±9 vs 38±12) and males (30±12 vs 34±15) but the gender difference was not significant.
Whole brain metabolism showed a significant cue by gender interaction effect (df 1, 24 p<0.05). In females whole brain metabolism significantly decreased (−8.6%±10; p<0.03) with the cocaine-cues video exposure (35.6±4 µmol/100 g/min) as compared to neutral (32.3±3 µmol/100 g/min); whereas in males whole brain metabolism tended to increase (+5.5%±18) with cocaine-cues exposure (36.2±5 µmol/100 g/min) as compared to neutral (34.9±5 µmol/100 g/min) but in the males this effect was not significant.
The SPM analysis on the absolute metabolic images in the females showed significant differences for the uncorrected threshold pu<0.005 for decreases in right midbrain (MNI coordinates for x (left to right), y (anterior to posterior) and z (top to bottom) of 3, −9, 15), left anterior cingulate (BA 24; MNI coordinates −14, 13, 30) and left lateral orbitofrontal cortex (BA 47; MNI coordinates −51, 24, −5). In males SPM did not detect any significant difference in absolute metabolism even when the threshold was reduced to pu<0.05. The cue by gender interaction showed that the differences were significant for pu<0.005 for greater cues-induced decreases in the females in anterior cingulate gyrus (BA 24; MNI coordinates −14, 13, 30 and BA 32 −18, 16, 38) and in left inferior frontal gyrus (BA 47, MNI coordinates −45, 39, −11).
The SPM results were very similar when daily cocaine use was entered as a covariate than when it was not. Here we present the results for the analysis after adjusting for the amount of cocaine used (daily cocaine use) since this was higher for the males than for the females.
In the females SPM analysis on the normalized metabolic images
(pu<0.001) showed decreases in anterior cingulate (BA 24, BA 32),
frontal (BA 4, BA 8), posterior cingulate (BA 23,), inferior parietal (BA 40),
thalamus (ventro lateral) and midbrain. There were no regions that showed
increases in metabolism (
Comparison correspond to Neutral > Cocaine-cues for uncorrected
threshold pu<0.001 cluster >200 voxels. There were no
regions where metabolism was higher during the Cocaine-cues than the
Neutral conditions.
Brain Region | BA | x | y | z | T-score | Z-score | Cluster size |
Cingulate Gyrus | 31 | 6 | −42 | 30 | 5.38 | 4.32 | 1817 |
Posterior Cingulate | 23 | −4 | −38 | 28 | 5.13 | 4.17 | |
Superior Frontal | 8 | −24 | 40 | 48 | 5.33 | 4.20 | 3716 |
Anterior Cingulate | 32 | 10 | 50 | −4 | 4.97 | 4.08 | |
Anterior Cingulate | 24 | −10 | 36 | 8 | 4.94 | 4.06 | |
Midbrain | 2 | −8 | −14 | 5.06 | 4.13 | 1758 | |
Thalamus | Ventral Post Lateral | −22 | −20 | 4 | 4.52 | 3.81 | |
Inferior Parietal | 40 | 54 | −34 | 56 | 4,64 | 3.88 | 557 |
Postcentral Gyrus | 2 | 56 | −16 | 52 | 4.53 | 3.82 | |
Precetnral Gyrus | 4 | 64 | −12 | 40 | 4.36 | 3.70 |
The clusters show the location of anatomical region including Brodmann Area (BA) and the coordinates in the Montreal Neurological Institute coordinates x (left to right), y (anterior to posterior) and z (top to bottom), and the corresponding statistical measures at the voxel level (T and Z scores) along with the cluster size for number of voxels. There where no regions where (Neutral < Cocaine-cues). Note that in the males none of the clusters was significant at pc<0.05.
In the males SPM analysis on the normalized metabolic images
(pu<0.001) showed no significant differences. Reducing the
threshold of pu<0.005 identified an area that showed increases in
the right inferior frontal gyrus (BA 44/45) but this region did not survive the
cluster correction for multiple comparison (
SPM analysis to assess the gender by condition interaction showed that the
genders differed significantly (pu<0.001). Females had
significantly greater decreases than males during the cocaine-cues when compared
with neutral in frontal regions (BA 4, BA 6, BA 8, BA 9), in anterior cingulate
(BA 24), posterior cingulate (BA 23, 31) and inferior parietal (BA 40) (
Comparison correspond to Females > Males for uncorrected threshold pu<0.001 cluster >200 voxels. There were no regions where males had larger changes than females.
Brain Region | BA | X | y | z | T-score | Z-score | Cluster size |
Middle Frontal Gyrus | 6 | −28 | 28 | 54 | 5.67 | 4.47 | 1185 |
8 | −22 | 24 | 38 | 5.15 | 4.18 | ||
6 | −32 | 14 | 62 | 4.97 | 4.08 | ||
Posterior Cingulate | 23 | −10 | −28 | 32 | 5.23 | 4.23 | 760 |
Cingulate Gyrus | 31 | 8 | −44 | 30 | 4.89 | 4.04 | |
Inferior Parietal Lobule | 40 | 58 | −30 | 54 | 5.19 | 4.21 | 625 |
Middle Frontal Gyrus | 6 | 38 | 8 | 62 | 4.85 | 4.01 | |
Precentral Gyrus | 4 | 66 | −8 | 30 | 4.52 | 3.81 | |
Superior Frontal Gyrus | 9 | −12 | 62 | 28 | 4.92 | 4.92 | 823 |
Anterior Cingulate | 24 | −10 | 36 | 12 | 4.87 | 4.02 | |
Anterior Cingulate | 24 | 10 | 38 | 12 | 4.83 | 4.00 | |
Thalamus | −7 | −13 | 5 | 4.57 | 3.84 | 48 |
The clusters show the location of anatomical region including Brodmann Area (BA) and the coordinates in the Montreal Neurological Institute coordinates x (left to right), y (anterior to posterior) and z (top to bottom), and the corresponding statistical measures at the voxel level (T and Z scores) along with the cluster size for number of voxels. Females showed significantly larger changes in regional metabolism (decreases) than males and there were no brain regions where males had larger changes than females. Note that we report on clusters >45 voxels in order to identify the regions within the large cluster that emerged when we used the preset >200 voxels level.
SPM voxel wise correlation between changes in craving and changes in normalized metabolism were not significant in females or males for pu<0.001. Reducing the threshold to pu<0.005 showed in males a significant negative correlation in a cluster region (5403 voxels) located in the posterior cerebellum (uvula and culmen); centered at MNI x,y,z coordinates of −36, −65, 24. There were no significant correlations in the females with craving.
The results from this study corroborate our hypothesis of greater brain reactivity to conditioned cocaine-cues in female than in male cocaine abusers even though the self-reported craving responses did not differ between the genders. Females when compared with males showed enhanced brain reactivity (as assessed by changes in brain glucose metabolism) to the cocaine-cues when compared with the neutral condition. The responses were also qualitatively different between genders; whereas in females the cocaine-cues significantly decreased whole brain metabolism in males it was associated with non-significant increases. In addition, the analysis of the normalized metabolic images, which increases the sensitivity to detect regional effects, showed that in the females the cocaine-cues elicited relative decreases in prefrontal cortex, anterior and posterior cingulate gyrus, inferior parietal lobe, thalamus and midbrain whereas in males the only significant difference was an increase in the right inferior frontal cortex (BA 44, 45) that did not survive cluster correction for multiple comparison. The fact that the gender differences in brain reactivity were significant after covarying for the doses of cocaine used indicates that the gender differences were not driven by differences in severity of drug use between the genders.
Traditionally executive control has been ascribed to prefrontal regions, mainly the
dorsolateral prefrontal cortex, dorsal anterior cingulate cortex/medial superior
frontal cortex and inferior frontal cortex
In a prior brain imaging study done in healthy controls in whom we exposed
participants to food cues and asked them to cognitively inhibit craving we showed
that whereas males were able to inhibit limbic brain activation by food-cues,
females were unable to do so
In the current study females tended to show greater craving than males when exposed
to cocaine-cues but this difference was not significant. Using a similar
cocaine-cues video a prior study reported higher levels of craving in females than
in male cocaine abusers
Our findings differ from prior imaging studies (using fMRI and PET CBF measures) that
showed activation of limbic brain regions with exposure to cocaine-cues (scripts
constructed to evoke craving or cocaine-cues videos)
Most clinical studies have focused on the effects of therapeutic interventions to
reduce craving
In this respect, therapeutic interventions to increase executive function
including impulse control may help patients develop coping skills to abstain
from using drugs when exposed to cocaine-cues. The findings from this study
suggest that these strategies may also benefit by considering gender differences
since the mechanism leading to relapse may differ for men and women
We recently showed that in cocaine abusers the stimulant medication
methylphenidate (MP) interfered with the reduction in metabolism triggered by
cocaine cues in cocaine abusers
A limitation for this study is the use of conscious awareness of craving as the
dependent variable. However, our study does not enable us to assess if genders
differ in unconscious responses to conditioned-cues. The studies were done in
the midfollicular phase, which is a time at which there may be a greater
reactivity to reward and prediction of reward and this reactivity in turn
appears to be modulated by estradiol
This study provides evidence of greater brain reactivity to cocaine cues in female than in male cocaine abusers but no differences in craving responses. Females, but not males showed decreased metabolic activity in brain regions implicated in top-down control network when exposed to cocaine-cues. Further studies to evaluate the cognitive consequences of these responses to cues are necessary to determine if they interfere with inhibitory control and to help guide gender tailored treatment interventions in cocaine use disorders.
This study was carried out at Brookhaven National Laboratory and approved by the local Institutional Review Board (IRB of record: Committee on Research Involving Human Subjects (CORIHS); Study #: IRBnet #91581; CORIHS ID #2007-4835; BNL IRB #404) and written informed consent was obtained from all participants.
Twenty six active cocaine abusers (16 M and 10 F) who responded to an
advertisement were studied. Subjects fulfilled DSM-IV criteria for cocaine
dependence and were active users for at least the prior 6 months (free-base or
crack). Exclusion criteria included current or past psychiatric disease other
than cocaine or nicotine dependence; past or present history of neurological,
cardiovascular or endocrinological disease; history of head trauma with loss of
consciousness greater than 30 minutes; and current medical illness. Written
informed consent was obtained from all subjects.
MalesN = 16 | FemalesN = 10 |
|
|
Age | 43±5 | 42±8 |
|
Education | 12.7±2 | 13.0±3 |
|
Current Smokers | 14 of 16 | 8 of 10 |
|
Cigarettes per day | 12±9 | 11±6 |
|
Cocaine Initiation | 25±6 | 22±8 |
|
Cocaine grams/d | 4.0±2 | 2.4±1 |
|
Years abuse | 18±7 | 20±6 |
|
Age first Rehabilitation | 32±5 | 24±11 |
|
CCS | 32±13 | 35±12 |
|
Body Mass Index (BMI) | 26±4 | 24±4 |
|
To assess the subjective experience of craving we used an analog scale
(1–10) for self-reports of “cocaine craving” and the brief
version of the CCQ
PET scans were conducted with a whole-body, high-resolution positron emission
tomograph (Siemens/CTI ECAT HR+, with 4.6×4.6×4.2 mm NEMA
(National Electrical Manufacturers Association) using 18FDG. Details
about the methods for scanning have been published
The data were analyzed using Statistical Parametric Mapping (SPM2)
We thank David Schlyer, David Alexoff, Don Warner, Paul Vaska, Colleen Shea, Youwen Xu, Lisa Muench, Barbara Hubbard, Pauline Carter, Karen Apelskog and Linda Thomas for their contributions.